Publications on Membrane Receptors and Their Ligands by ...

Publications on

Membrane Receptors and Their Ligands

by

Peter Robert Schofield

Garvan Institute of Medical Research

Submitted to University of New South Wales

for consideration for the award of the Degree of

Doctor of Science

February 1996

Table of Contents

1. Contents of Submission 3

2 . Applicant's Role in the Submitted Work 4

3 . Research Areas Covered in the Submission 8

I. Ligand-gated Ion Channel Receptors 9

II. Other Studies of Receptors and Ligands 21

A. G Protein-coupled Receptors 21

B . Enzymes Involved in Neurotransmission 24

C. Recombinant Follicle Stimulating Hormone 25

D. Genes Involved in Brain Dysfunction 27

4 . Applicant's Declaration 29

5 . List of Publications 3 0

6 . List of Awards and Invited Lectures at National and

International Conferences 4 2

7 . Enclosed Publications 4 6

U N S W

3 0 O C T 2001 LIBRARY

1. Contents of Submission

The published work contained in this submission represents the results of a series of

studies in the field of membrane receptors and their ligands. These studies were

initiated in late 1985, as part of my postdoctoral studies, and reflect two significant

career decisions, namely to work in mammalian molecular biology and study

biological specificity by undertaking molecular studies of receptor structure and

function. The submission contains 54 items consisting of 42 scientific papers, 6

reviews and 6 book chapters which are detailed as publications 17 to 70 in the List of

Publications. All of this work has been published and available for public criticism

with the exception of paper 70 which is 'in press'.

The List of Publications also includes a Citation Analysis that comprises the current

(1993) Science Citation Index (SCI) Impact Factors for the relevant journals and

indicates the total number of citations for each specific paper. The average SCI Impact

Factor of the joumals in which refereed papers were published is 10.7 (42 papers and

5 review joumals). Up until the end of 1995, the complete citation count of all works

was 4,645 with 3 works (publications 17, 23 and 32) being citation classics, having

1,075, 488 and 610 citations, respectively. In addition, 2 papers have over 300

citations, 6 have over 100 citations and 8 have over 50 citations. A list of awards

based on the work contained in this submission and invited lectures at National and

International Conferences is also included.

I have chosen not to include the 16 scientific papers and book chapters that comprised

my PhD thesis studies and a brief postdoctoral period as this work deals with studies

of symbiotic nitrogen fixation and the nodulation and nitrogen fixation genes encoded

by the symbiont Rhizobium trifolii. The details of this work and citation analysis are,

however, contained as publications 1 to 16 in the List of Publications.

2 . Applicant's Role in the Submitted Work

The research covered in this submission commenced in late 1985 when I undertook postdoctoral studies with Professor Peter Seeburg, initially at Genentech Inc., the leading US biotechnology company, and subsequently at the Centre for Molecular Biology at the University of Heidelberg. This work (publications 17 to 44) represents studies into the ligand-gated ion channel receptors, G protein-coupled receptors and enzymes involved in neurotransmission. For those papers in which I was the first author (publications 17, 18, 28 and 31) or equal first author (23 and 34), I designed and executed the experiments, and played a seminal role in the analysis and interpretation of the data and was responsible for the preparation of the resultant publications. In the case of the two equal first author publications (publications 23 and 34) the work involved in these studies involved contributions by both authors within their relative fields of expertise. For example in paper 23,1 cloned, analysed and prepared expression constructs of the a2 and a3 G A B A A receptor subunits which were expressed and analysed electrophysiologically analysed by Dr Ed Levitan at the MRC Molecular Neurobiology Unit in Cambridge.

The sequence of research involving the cloning and analysis of the various G A B A A receptor subunits was such that following the original isolation of subunits at Genentech as part of a collaboration with Professor Eric Barnard's group in Cambridge, Peter Seeburg and I moved to the Zentrum fur Molekulare Biologie (ZMBH) in Heidelberg. This allowed many new areas of investigation to be initiated by the new students and postdoctoral researchers recruited to Seeburg's laboratory. Indeed, I brought many newly identified cDNA clones from the States that subsequently formed the basis of a number of projects. In particular, one rack of newly isolated cDNAs became the basis of several important papers (34, 40 and 44) concerning receptor heterogeneity. During this time my role in the laboratory changed

and I provided far more practical and intellectual leadership of the projects. This is

reflected in teaching situations in which I would typically be the second (papers 22, 34

and 39) or a lessor author (papers 35, 43 and 44) or overall leadership situations in

which I would typically be the second last author (papers 21, 30, 32, 37 and 40) or in

one instance (paper 26) where my role was such that I was senior (last) author.

Several studies were undertaken as part of collaborative efforts, such as those on the

enzyme enkephalinase which were conducted with Dr Bernard Malfroy, a protein

chemist who had purified the enzyme (papers 19, 24 and 29). In these molecular

studies I provided all of the molecular expertise both in terms of teaching and assisting

Dr Malfroy as well as conducting many of the experiments. Analysis and publication

were joint tasks. Other collaborative studies in which I played lessor roles through the

provision of specific molecular constructs involved the cloning of the gene encoding

the amidating enzyme (paper 20) and chromosomal localisation studies (paper 41).

Several invited reviews were published during this time (papers 25, 36, 38 and 42)

and these were as either single author or first author publications reflecting my

principal role in developing the concepts and writing these papers. In addition, a

number of invited conference proceedings and book chapters were published (papers

27, 33 and 39) which were based entirely on my either first or second last authored

contributions.

My return to Australia saw me undertake the leadership of a project to develop and

characterise recombinant human follicle stimulating hormone, initially at Pacific

Biotechnology Ltd. and, subsequently at the Garvan Institute of Medical Research. In

all of these papers and invited book chapters (46, 47, 50, 59 and 62), my role is

reflected in holding the senior authorship and being responsible for the overall design,

execution and interpretation of the research. Two collaborative studies in which our

research materials were critical are represented by lesser authorships (papers 45 and

68).

At the Garvan Institute, I established a research group in 1991 which continued the

study of ligand-gated ion channel receptors with the emphasis being on the Glycine

receptor. In all of the work published from this period my role in overall design,

execution and analysis is reflected in by my holding of the senior (last) authorship

position (papers 48, 49, 52, 55, 58, 61, 64, 67 and 70). This also applies to the

various invited reviews and book chapters that were published (papers 51, 53, 56 and

65). My group's role in a significant international collaboration was reflected in my

second last authorship position (paper 60). A small project on G protein-coupled

receptors was also under my direction and this is represented by either my joint

authorship (papers 54 and 66) or my senior authorship (paper 63). Finally,

collaborative studies aimed at establishing research projects in the areas of psychiatric

genetics and dementia resulted in the first two publications, in one (paper 56), my role

as senior author reflected the role of my group in conducting and analysing the

research while the other (paper 69) represented a contribution to a previously

established project.

For each of the papers published as senior author, it is important to note that these

studies represented the training of postgraduate students and postdoctoral researchers

and that while the work was under my overall direction, these individuals made major

contributions in both the practical and intellectual aspects of these papers. In

particular, it is appropriate to note the role of PhD students, Leonora Bishop (papers

47, 50, 59 and 62) and Sundran Rajendra (papers 52, 56, 58, 61, 64, 65 and 67), and

postdoctoral research fellows, Dr Robert Vandenberg (papers 48, 49, 51, 52, 53 and

67), Dr Andrea Townsend-Nicholson (papers 54, 63 and 66) and Dr Joseph Lynch

(56, 58, 60, 61, 64 and 70). Finally, I would like to acknowledge the ongoing

collaboration with Professor Peter Barry and his laboratory at the University of New

South Wales (papers 48, 52, 56, 58, 61, 64 and 67).

3 . Research Areas Covered in the Submission

The research that I have conducted in the field of membrane receptors and their ligands

falls into one of two related areas as outlined below.

I . Ligand-gated Ion Channel Receptors

II . Other Studies of Receptors and Ligands

A . G Protein-coupled Receptors

B . Enzymes Involved in Neurotransmission

C. Recombinant Follicle Stimulating Hormone

D . Genes Involved in Brain Dysfunction

The majority of my work from 1985 to the present has concentrated on the inhibitory

ligand-gated ion channel receptors (Section I), initially focusing on the G A B A A

receptor and more recently on the Glycine receptor. Research into the G protein-

coupled receptors (Section ILA) has been a lessor interest but has spanned the same

period and included studies on adrenergic, muscarinic, opioid and adenosine

receptors. My other research activities have been largely based on developing

opportunities and applying my expertise to specific projects that arose whilst at

Genentech (Section TLB), Pacific Biotechnology (Section II.C) and the Garvan

Institute (Section II.D). I will now consider each of these Sections separately.

I . Ligand-gated Ion Channel Receptors Publications 17, 21, 23, 26, 27, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 48, 49, 51, 52, 53, 55, 56, 58, 60, 61, 64, 65, 67 and 70

Summary Through the cloning and expression of the G A B A A receptor subunits, these studies have revealed the existence of the ligand-gated ion channel receptor superfamily. The inhibitory G A B A A and glycine receptors are multi-subunit proteins and receptor subunit and subunit-subtype diversity has been shown to be an important characteristic by which structural and, more importantly, functional diversity is encoded. Using the glycine receptor as a model, the ligand-binding domains of the receptor have been mapped by site-directed mutagenesis. Heritable mutations in the glycine receptor that cause startle syndromes have led to the identification of the residues that mediate signal transduction in those sequences that flank the ion channel lumen. Together, these studies have provided a detailed molecular description of both the structural and functional architecture of the inhibitory ligand-gated ion channel receptors.

A Ligand-gated Ion Channel Receptor Superfamily The research interests I developed during my thesis studies concerned the understanding of the molecular basis of specific biological recognition and signalling events. I therefore made two significant career decisions before undertaking a postdoctoral research position with Peter Seeburg at Genentech Inc., the leading US biotechnology company. Firstly, I decided to work in eukaryotic, and specifically mammalian, molecular biology in areas of importance to biomedical research as I felt that this area tended to lead much of modem biology. Secondly, I decided to focus

10 my research interests on the understanding of specificity by undertaking molecular studies of receptor structure and function.

At Genentech, my primary research project concerned the cloning of the a l and p subunits of the major inhibitory neurotransmitter receptor, the gamma-aminobutyric acid type A receptor (GAB A A R). This project was undertaken as a collaborative endeavour with the group of Professor Eric Barnard at the MRC Molecular Neurobiology Unit in Cambridge. Degenerate oligonucleotide probes, based on partial amino acid sequences obtained from purified receptor protein were used to clone the cDNAs encoding the receptor subunits. I then generated full length cDNAs which were used for expression studies in Xenopus oocytes. In addition to defining the primary structure of this important neuroreceptor, this work also revealed the existence of the ligand-gated ion channel receptor superfamily (17, 27 and 33). The work was published in Nature with our model of the G A B A A R forming the cover illustration of the issue of the journal. The paper was the 26th most cited life-sciences paper to be pubUshed in 1987 (Current Contents 33: 3-18, 1989) and is now a citation classic having received almost 1,100 citations. While some similarities were thought to exist between the various receptors that encoded the receptors that operated as ligand-gated ion channels, it was completely unexpected that such a clear protein superfamily would emerge. The back-to-back publication of the cloning and expression of the G A B A A R with the sequence of the strychnine binding subunit of the glycine receptor, isolated by Professor Heinrich Betz's laboratory in Heidelberg, further reinforced this view of receptors occurring in protein superfamilies. These studies were also occurring contemporaneously with those being conducted by other groups, both at Genentech and elsewhere in which it was being demonstrated that apparently unrelated receptors that signalled via G protein coupling also comprised a receptor superfamily of G protein-coupled receptors (see Section I I .A) .

11

Receptor Subunit Diversity Experiments initiated at Genentech immediately revealed the high degree of molecular diversity of the GAB A A R. A novel cDNA, which was shown to encode the a 2 subunit was isolated during the screening to obtain full length a l cDNAs (23). Further screening of cDNA libraries allowed the isolation of an a 3 cDNA as well. The presence of additional receptor subunit cDNAs was consistent with pharmacological reports of receptor heterogeneity (23). This observation encouraged the further low stringency homology screening of brain cDNA libraries and resulted in the isolation of a number of additional subunit cDNAs. These included cDNAs for the a4, P2, p3 and yl subunits (34, 40, 44). In addition, the screening of human brain cDNA libraries allowed the isolation of cDNAs encoding the various human cDNAs (31). Together, these studies demonstrated that receptor subtype heterogeneity was a feature of the GAB AA R and more generally of the ligand-gated ion channel receptor superfamily. This was a profound and unexpected conclusion.

At this point in my postdoctoral studies, my supervisor. Professor Peter Seeburg moved to the Centre for Molecular Biology (ZMBH) at the University of Heidelberg. I was the only person to move with Peter Seeburg and the newly established laboratory devoted the majority of its research effort to the pursuit of the analysis of the structural and functional basis of molecular diversity at the G A B A A R. This opportunity allowed me to both continue my own experimental program but also to assume an increasing significant role in the direction of the entire laboratory and mentoring of new students and postdoctoral research fellows. This is reflected in a large number of publications in which I was either second author or more significantly second last author.

There were three notable achievements of my work in Heidelberg. Firstly, the definition and characterisation of multiple G A B A A R subunit subtypes. The initial

12

isolation of the a 2 and a 3 subunits allowed the co-expression of the three alpha

subunits with the existing p subunit and electrophysiological examination of these

recombinant receptors in Xenopus oocytes. This work, which was undertaken as a

collaboration with Dr Ed Levitan in Eric Barnard's group allowed us to both

demonstrate that receptor heterogeneity occurs at the molecular level and with resultant

different biological properties, in this instance up to thirty-fold differences in agonist

sensitivity, depending upon which particular alpha subunit was present in the

recombinant receptor (23). The four GAB A a R sequences, plus that of the strychnine

binding subunit of the glycine receptor revealed the presence of a highly conserved

octapeptide sequence located in the second transmembrane domain of the receptor

(21). Thus, many library screening experiments were conducted to complement the

existing collection of novel subunits which were brought from the States. This

subsequently allowed the isolation of full length cDNA clones encoding the bovine, rat

and human a4 (40), a5, a6, p2 and p3 (34) subunits. Again, the expression of

various cDNA clones was important as this allowed the demonstration of the

functionally distinct properties conferred by specific receptor subunit subtypes (34,

42). Much of the characterisation of these cDNA clones such as a5, a6 and others

continued in the Seeburg laboratory after my retum to AustraUa.

The second area of significance was the identification and analysis of novel receptor

subunits, especially those encoding essential functions at the G A B A A receptor. As

mentioned above, the yl subunit (44) had already been isolated by low stringency

homology screening and the use of degenerate oligonucleotides probes, designed

against the invariant octapeptide sequence of transmembrane domain two, allowed the

isolation of additional cDNA clones. As before, each of these cDNA clones was

functionally characterised so that the unique properties that were conferred upon a

recombinant G A B A A R could be determined. The isolation of the yl and 72 subunits

(32) was unexpected as the dogma of the time described G A B A A RS as containing

13

only a and (3 subunits. However, despite earlier suggestions (17), recombinant GAB A A Rs were not potentiated by the addition of benzodiazepines (23) one of the most important classes of drugs that act at these receptors. Therefore, using the mammalian cell expression systems, the development of which is outhned below (26), we sought to define which subunits could mediate benzodiazepine responsiveness. As part of a group effort we showed that the y subunits, and the 72 subunit in particular, mediated the benzodiazepine response (32). In my view this paper, which was published as a letter to Nature represents the conciseness and brevity that can only be achieved when a series of experiments answers all of the important questions of an investigation. In five paragraphs we described the cloning of the 72 subunit, its role in forming benzodiazepine binding sites, its role in forming benzodiazepine potentiated channels, its anatomical distribution and the implications of this novel subunit on the then current state of knowledge of ligand-gated and more specifically GAB AA RS. It is therefore not suprising that this paper has received over 600 citations.

Similar studies were undertaken to characterise the yl subunit's role in mediating the benzodiazepine response (44). In this instance, a less classical pharmacology was observed, and these results again demonstrated that the functional heterogeneity of GAB A A Rs would far exceed that which had been previously reported, or even conceived. The cloning of the y class of subunit allowed my colleagues in Heidelberg to continue to explore the role of receptor subtype heterogeneity on the functional properties of G A B A A RS. In particular, more detailed analysis of subunit contributions to benzodiazepine selectivity was pursued by (the late) Dr Dolan Pritchett through several high profile publications.

The use of low stringency homology screening also allowed the isolation of additional receptor subunit classes which had not previously been predicted. Iris Killisch, a PhD student working under my supervision, isolated a novel subunit termed 5 (37). In

14

characterising the anatomical distribution of the 5 and 72 subunits, Dr Brenda Shivers was able to conclude that these two receptor subunits are expressed in distinct and non-overlapping neuronal subpopulations (37). Thus, having established some of the functional bases of receptor subtype diversity, we now were also able to provide an anatomical basis as well. These results were especially significant as many in the field had speculated that the number of possible functional receptors based on various combinations and permutations of the known receptor subunits would be extremely large, possibly in the thousands. Studies such as ours (37) demonstrated that such concepts were extremely simplistic. For example, if 5 and 72 subunits were not co-localised, it would not be possible to have functional G A B A A RS that contained both receptor subunits. Thus, although receptor subtype heterogeneity was a new phenomenon of great theoretical and practical significance, it would turn out to be a tractable, not intractable, problem. Such discussions formed much of the office and coffee-room dialogue and resulted in more detailed considerations of these issues being published (38, 42).

The chromosomal localisation of the first of these receptor subunit genes (41) demonstrated that the genes are dispersed throughout the genome. More recent studies have in fact revealed that there are at least three functional receptor loci located on chromosomes 4, 5 and 15. At each loci, the co-ordinated expression of a specific a , P and Y subunit subtype gene results in the functional expression of a specific G A B A A

R. Thus, chromosomal loci also reduce the heterogeneity of receptor subtype heterogeneity. This work also kindled an ongoing interest in the genetics of receptors.

Functional Analysis of Receptor Subtypes The third area in which I played a major role was in the development and practical use of the human embryonic kidney '293' cell line being used as the system of choice for the combined pharmacological and electrophysiological analysis of cloned ion channel

15

genes. This work was initiated at Genentech where Dr Cori Gorman had developed a high efficiency mammalian cell expression system that was particularly suited to the analysis of transient transfections. This system relied on the trans-activation of the human cytomegalovirus (CMV) promoter/enhancer by adenovirus gene products that were responsible for transforming the human embryonic kidney 293 cell line. In Heidelberg, this work was one of the project areas that were transferred to new post doctoral fellows in the laboratory. In this instance, a fellow postdoc Dr Dolan Pritchett undertook the expression studies and then in collaboration with a PhD student, Harald Sontheimer, in Dr Helmut Kettenmann's laboratory, these cells were electrophysiologically analysed by whole cell patch-clamp recording. This was the first published use of the CMV/293 cell system for electrophysiological analysis and this system has now become the method of choice for mammalian cell expression and analysis of ion channels (26). This work also led to the identification and use of homomeric receptors (pentameric receptor complexes comprising a single subunit subtype only) as a powerful way of analysing the structure and function of the ligand-gated ion channel receptors (26). The use of the homomeric assay was of critical importance in actually establishing the ligand selectivity and therefore the receptor class of novel receptor subunits. Thus, analysis of the 6 and y l subunits as homomeric channels was essential to prove that the 5 subunit was a member of the GABAA R (37).

Glycine Receptor Subunits and Expression Studies Located on the same floor of the Centre for Molecular Biology was Heinrich Betz's group who had independently cloned the other major inhibitory neurotransmitter receptor, the glycine receptor (Gly R). This physical proximity of the two laboratories enabled a large number of collaborative research projects to be undertaken. These projects included the identification of receptor subunit diversity in the Gly R. I identified the human Gly R a l subunit via low stringency homology screening for

16

related G A B A A R subunits and a PhD student, Gabi Grenningloh, cloned the a2 subunit by screening spinal cord libraries with the Gly R a l cDNA (43). Again, the use of chromosomal localisation studies was included as part of the receptor subunit characterisation (43). The other significant project conducted in collaboration with the Betz laboratory was the demonstration that homomeric a l Gly Rs have properties almost indistinguishable from spinal cord Gly Rs when examined in either the classical Xenopus oocyte system (30) or our recently developed CMV/293 system (35). These studies established the a l Gly R as a model system for characterising the mechanism of action of all ligand-gated ion channel receptors. In this way, the Gly R a l subunit has come to occupy a niche analogous to the homomeric receptors formed by the nicotinic acetylcholine al subunit.

My work on the molecular biology of the G A B A A R and Gly R was recognised in 1990 by the presentation of the A. W. Campbell Award by the Australian Neuroscience Society for the best contributions by a member of the Society over the first five postdoctoral years.

Glycine Receptor Agonist and Antagonist Binding Sites After my return to Australia in late 1988, and whilst employed at Pacific Biotechnology, I held an honorary appointment at the Garvan Institute of Medical Research. Combined with external grant support, T commenced an independent research program to describe the detailed basis of the mode of action of the ligand-gated ion channel receptors. This work was undertaken by a postdoctoral research fellow Dr Robert Vandenberg. In order to establish an internationally competitive research program, we focused on the use of the homomeric Gly R and studies aimed at identifying ligand binding sites at the receptor (48). The need for ongoing electrophysiological analysis led to the initiation of a longstanding collaboration with Professor Peter Barry and members of his laboratory including Dr Chris French. Our

17

results demonstrated that the three loop model of ligand binding, which had been

proposed by Changeux's group for the nicotinic acetylcholine receptor, was applicable

to the other members of the ligand-gated ion channel receptors (48).

I commenced full time employment with the Garvan in late 1991 following the demise

of Pacific Biotechnology and played an active role in the Neurobiology Division's

submission for the Institute's National Health and Medical Research Council

(NHMRC) block grant renewal. At this time I developed and expanded the research

program on the Gly R. We further explored the role of the three loop ligand binding

model as outlined above and were successful in identifying an agonist specific binding

site in loop three (49). A third paper involved the formal refutation of a theoretical

model of ligand binding and receptor specificity in which the invariant disulfide

bonded loop was proposed to form the agonist binding site of the receptor. By site-

directed mutagenesis and functional expression we were able to disprove this

hypothesis (52).

My research group published several key papers in high profile journals on the

molecular basis of ligand (agonist and antagonist) binding at the Gly R (48, 49, 52).

This work thus established me as an independent player in the field of molecular

neurobiology of receptors. Invited reviews (51) and book chapters (53) helped in

consolidating this position. This output aided in achieving a highly successful

Institute block grant review, and my appointment to the NHMRC Fellowship Scheme

as a Senior Research Fellow.

Our research continued to explore the residues involved in mediating agonist binding.

These studies relied on input from a jointly (with Professor Peter Barry) supervised

PhD student Sundran Rajendra and from a postdoctoral research fellow Dr Joe Lynch

who had joined the Garvan following the NHMRC block grant review. Most

18

significantly this work demonstrated that loop three of the Gly R a l subunit is, in fact, the principal ligand binding site at the Gly R (67). This study complemented the earlier characterisation that we had undertaken of this domain (48,49, 56) and added the use of immunohistochemical localisation studies to examine the ability of various mutant Gly Rs to be expressed on the cell surface. The paper was complemented by our data forming the cover illustration of the issue of The EMBO Journal (67).

Genetics of Startle Syndromes The expansion of the Gly R project included an increasing emphasis on the use of genetics. The localisation of the Gly R a l subunit gene to chromosome 5q32 (55) suggested that this gene may be the locus of the human neurological disorder hereditary hyperekplexia or familial startle disease. Similarly, the localisation of the Gly R p subunit gene to murine chromosome 3 (70) suggested that this locus may be responsible for the defective glycinergic transmission in the mouse mutant spastic. 1994 saw the success of several groups in the characterisation of the molecular mechanisms by which mutations at the Gly R cause inherited startle syndromes in man (hyperekplexia) and in mice {spasmodic and spastic). We initiated a collaboration with Dr Stephen Ryan of the University of Texas at San Antonio and Dr Rita Shiang and (the late) Professor John Wasmuth of the University of California at Irvine. Together this group demonstrated that hyperekplexia was caused by mutations of the a l subunit of the Gly R with two different mutations, to Gin and Leu of Arg 271 being identified in different pedigrees. My group undertook to conduct the functional characterisation of these mutations (58, 61) which were originally predicted to disrupt chloride ion flow through the receptor channel. The characterisation of these mutations initially revealed that mutations of Arg 271 did not cause a major disruption to the chloride flux or agonist binding but rather disrupted the agonist activation of the channel. In other words, the hyperekplexia mutations revealed a residue involved in signal transduction at the Gly R.

19

Studies of the hyperekplexia mutations provided fundamental insights into the mechanisms of signal transduction in the ligand-gated ion channel receptors. For example, the Arg 271 mutations that underlie hyperekplexia result in reduced glycinergic transmission via altering single channel conductances (61). Even more suprisingly, these mutations convert agonists such as taurine and p-alanine into competitive antagonists (61) and the competitive antagonist, picrotoxin, into an allosteric potentiator (64). Such a combination of effects seems possible from a single amino acid residue only if it occupies the role of a key integration site within the GlyR. Exploration of the molecular basis of these properties is now defining the domains and specific amino acids of the receptor that mediate the conformational change necessary for signal transduction. Indeed, it is suprising that all of the mutations that have been identified as causing hyperekplexia have been localised between transmembrane domains 1 and 3 of the receptor. Moreover, each of the known mutations Ile244Asn, Arg271Leu, Arg271Gln, Lys276Glu and Tyr279Cys appear to disrupt signal transduction, leading to the conclusion that both the intracellular and extracellular sequences that flank the ion permeation pore (transmembrane domain 2) are involved in mediating the allosteric transition of the receptor protein that results in signal transduction.

In collaboration with Stephen Ryan my group provided the functional expression data that allowed the demonstration that the mutation in the spasmodic mouse was due to a single point mutation in the a l subunit of the Gly R (60). In this case, the resulting amino acid change, from an Ala to Ser at position 52 also resulted in the disruption of the signal transduction capacities of the receptor. The work with Stephen Ryan has led to us developing stronger interests in the use of animal models, and current research projects are using these mice, and the allelic variant oscillator, to more fully characterise the functional roles of Gly Rs in vivo. Our preliminary data indicate that

20

the oscillator mouse represents a null allele of the Gly R gene. My interests in, and

contributions to, the genetics and functional characterisation of the effects of Gly R

deficits resulted in being invited to write a review for the Perspectives on Disease

section of Trends in Neurosciences (65).

Over the past decade, I consider that my work in the area of ligand-gated ion channel

receptors has fundamentally changed the field of receptor research, and cite as

evidence for this the following. My model of the G A B A A receptor (figure 5 from

paper 17) has been reproduced directly in standard textbooks such as Review of

Medical Physiology, 14th edition, 1989, W.F. Ganong, Prentice-Hall International

Figure 2 from paper 17 has been used in standard textbooks such as Ionic Channels of

Excitable Membranes, 2nd edition, 1992, B. Hille, Sinauer. Likewise, key

publications (papers 17, 32 and 43) have been cited in the Selected Reading or

References sections of standard textbooks such as Principles of Neural Science, 3rd

edition, 1991, E.R. Kandel, J.H. Schwartz & T.M. Jessell, Elsevier, (paper 32) and

Ionic Channels of Excitable Membranes, 2nd edition, 1992, B. Hille, Sinauer (papers

17, 32 and 43).

21

II. Other Studies of Receptor and Ligands

Publications 18, 19, 20, 22, 24, 25, 28, 29, 36, 45, 46, 47, 50, 54,

57, 59, 62, 63, 66, 68 and 69

Summary

The work described in this section outlines the contributions that I have made to

understanding receptor diversity through the cloning and functional expression of

receptor subtypes of the G protein-coupled receptor superfamily (Section II. A). These

studies have been extended to define the ligand binding domains within the adenosine

receptor class of this superfamily. My other research activities have been largely

based on developing specific projects that arose in specific laboratories. This included

cloning and expression studies on enzymes involved in neurotransmission (Section

II.B), expression of recombinant human follicle stimulating hormone (FSH) and

investigating the role of the carbohydrate residues attached to FSH in signal

transduction (Section II.C) and genetic linkage studies of genes involved in bipolar

affective disorder and Alzheimer's Disease (Section n.D). I will now consider each of

these Sections separately.

A. G Protein-coupled Receptors

Publications 18, 22, 28, 36, 45, 54, 63 and 66

G Protein-coupled Receptors form a Superfamily

At Genentech, several collaborative projects were initiated to clone the \i opioid

receptor subtype using oligonucleotide probes based on peptide sequences. Only one

such collaboration was fruitful and allowed the cloning of a putative opioid receptor

(28). By this time it had been realised that all members of the G protein-coupled

receptor superfamily were seven transmembrane proteins (36). Unfortunately,

22 sequence analysis of the putative opioid receptor revealed that this candidate did not fulfil the criteria for a seven transmembrane domain protein and this protein, a member of the immunoglobulin superfamily, was probably involved in cell contact and adhesion. In parallel with degenerate oligonucleotide screening for novel G A B A A receptor subunits, I commenced studies, initially with Dr Ernie Peralta at Genentech and subsequently with a PhD student, Thomas Braun at the ZMBH to isolate novel G protein-coupled receptors. This work resulted in the cloning of the human [32 adrenergic receptor gene (18) and the rat muscarinic M4 (m3) receptor (22). Our work provided further direct support for the growing realisation that each class of G protein coupled receptor was, and would be, characterised by the presence of multiple related receptor subtypes.

As part of the project on recombinant human follicle stimulating hormone undertaken at Pacific Biotechnology, I again collaborated with Thomas Braun and Rolf Sprengel from the ZMBH group on the characterisation of the role of the amino-terminal leucine-rich repeats of the FSH receptor in determining hormone specificity (45). In this instance, the ability to provide recombinant FSH, free of contaminating LH (luteinising hormone) was essential in understanding where the ligand binding site was located in the FSH receptor.

Adenosine Receptor Subtypes At the Garvan Institute, a project was initiated by Professor John Shine, in which I was a co-investigator, to use our respective skills in the isolation of receptor subtypes to clone novel members of the G protein-coupled receptor superfamily. The primary emphasis of this project was to identify novel receptors for known neuropeptides, but one of the early successes was the isolation of the family of adenosine receptors. In particular, the human Al, A2a and A2b receptor subtypes were cloned by the Garvan group. After the successful block grant funding renewal, John transferred to

23

adenosine receptor project to my sole supervision in 1993. Working with Dr Andrea

Townsend-Nicholson, a postdoctoral fellow, I extended my interests in molecular

pharmacology and we commenced a series of structure/function studies using the

human A1 adenosine receptor subtype. Our initial and important conclusion was to

identify a single residue, Thr 277, in the seventh transmembrane domain that mediated

specific agonist binding (54). These studies have been extended to include a

structure/function mutagenesis scan by alanine replacement of the entire seventh

transmembrane domain and of other residues of potential ligand specificity throughout

the entire A1 receptor. The first of these studies has been submitted for publication

having identified Thr 270 and Pro 285 as key ligand binding residues, while the

second study is nearing completion, having identified Pro 86 as a key residue

involved in signal transduction.

In collaboration with Elizabeth Baker in Professor Grant Sutherland's group in

Adelaide, we have localised the chromosomal map positions of all four adenosine

receptor subtype genes, with the A2b and A1 receptor genes having been published

(63, 66) and the map positions and genomic structure of the A2a gene having been

submitted and the A3 gene in preparation. This work was significant in that earlier

reports in the literature (subsequently retracted) were shown to be wrong by our

studies.

The collective efforts of the Garvan group have led to us being the first group to have

cloned and expressed all four human adenosine receptor subtypes. The current

research, which I have directed, has now complemented this information with a

complete picture of the chromosomal map positions and information on genomic

structure of the various genes. However, I believe that our contributions to the

understanding of ligand binding and receptor signalling will be of greatest significance

to the field of G protein-coupled receptor research.

24

B . Enzymes Involved in Neurotransmission

Publications 19, 20, 24, 25 and 29

At Genentech, I was also involved in two research projects which involved the cDNA

cloning, expression and analysis of genes involved in mediating the biosynthesis of

neuropeptide transmitters (the enzyme that causes carboxyl-terminal amidation of

peptides) and the inactivation of the enkephalins and other peptides (neutral

endopeptidase or enkephalinase). Two approaches were taken to the cloning of the

enzyme encoding peptidyl alpha mono-oxygenase (PAM). One was screening of

expression libraries using polyclonal antisera, which I undertook, and the other was

by oligonucleotide screening based on limited peptide sequences. The latter approach

was successful and I further contributed by the cloning of full length cDNAs that were

subsequently used in the expression studies of this enzyme (20).

The studies on the enzyme enkephalinase were conducted with Dr Bernard Malfroy,

who had characterised and purified the enzyme prior to joining the staff of Genentech.

Bernard had sought my assistance to teach and guide him in the use of molecular

techniques so as to be able to clone and express this important enzyme. Competing

against another group, we successfully cloned the rat cDNA (19). This facilitated the

cloning of the human cDNA (24) and the expression of enzymatically active protein in

mammalian cells (29). While these studies were based on the opportunities for

collaborative research and have not been further pursued, they were important to my

overall scientific development as a molecular neurobiologist, since they allowed me to

explore and develop my knowledge and skill base, in collaboration (19, 20, 24, 29),

or alone (25), in this new field of endeavour.

25

C. Recombinant Follicle Stimulating Hormone

Publications 45, 46, 47, 50, 59, 62 and 68

I returned to Australia to join Pacific Biotechnology Ltd in late 1988. Whilst at the

company, I led a team on the development of recombinant human follicle stimulating

hormone (FSH) as a drug for the treatment of infertility. Urinary FSH is already

widely used in treating infertility, in particular as part of in vitro fertilisation (IVF)

programs, but at that time there was no source of recombinant hormone. This project

resulted in the development of recombinant human FSH proceeding to the preclinical

(animal testing) stage (46,47) but the under-capitalisation of the Pacific Biotechnology

resulted in its closing in 1991 and the project being scaled down and transferred back

to the Garvan Institute. Currently, the partners of the Co-operative Research Centre

for Biopharmaceutical Research are continuing the development of recombinant human

follicle stimulating hormone with the aim of undertaking clinical trials.

The research component of my work at Pac Bio involved the analysis of the role that

the carbohydrate residues play in determining the biological properties of glycoprotein

hormone. The ability to produce large amounts of recombinant human FSH free of

other glycoprotein hormones allowed the demonstration that the charge carried by the

pi isoforms (determined by carbohydrate structure and content) direcdy correlates with

the in vitro biological properties of the hormone (46, 50). The purified hormone was

also of use in determining the sites of the FSH receptor that were involved in ligand

binding (45). These studies led to the development of Leonora Bishop's PhD research

program in which she investigated the role of specific asparagine-linked carbohydrate

residues in determining the biological properties of the hormone. The combined

analysis of genetically modified forms of recombinant human FSH using both in vitro

(59) and in vivo (62) assays, which had not previously been attempted for any of the

glycoprotein hormones, provided important new insights into the role of the (3 subunit

26

carbohydrate residues in determining hormone potency. The conclusions that were

made in these published studies and in, as yet, unpublished studies led to the

collaborative extension of these investigations in the related glycoprotein hormone

thyroid stimulating hormone (TSH) (68).

27 D. Genes Involved in Brain Dysfunction Publications 57 and 69

In 1993 I assumed the Garvan's responsibility for a small collaborative project aimed at identifying the genes that cause or predispose individuals to the neuropsychiatric disorder, bipolar affective disorder (manic depressive illness). This work was being conducted with Associate Professor Philip Mitchell, a psychiatrist who had recruited approximately 500 individuals from 24 large multigenerational pedigrees with a high density of bipolar disorder. The laboratory work was all conducted at the Garvan Institute and had previously focussed on the testing of specific candidate genes, notably the dopamine receptor subtypes, as being causative of bipolar disorder. In the first instance I extended the scope of these studies by the use of highly polymorphic microsatellite markers for the candidate gene for the alpha subunit of the stimulatory G protein (57).

In response to the call by the NHMRC for the establishment of a 'Network for brain research into mental disorder', I played a key role in the preparation of the successful application. My role in this Network is as coordinator of the Genetic Linkage Consortium and this has seen an increased effort in programs directed to identifying the genes that cause bipolar disorder as well as familial Alzheimer's Disease. I have now focussed on the use of a whole genome scan approach for the bipolar disorder project with specific focus on areas of potential linkage that have been identified by other groups. To this end, we have completed an exclusion map of chromosome 16 and are preparing this work for publication. We are currently extending these studies and have submitted a manuscript in which a potential schizophrenia locus is shown not to be causative of bipolar disorder. We have also completed studies examining the role of chromosome 21 markers at the PFKL locus in which, assuming heterogeneity, we find support for the existence of a gene involved in causing bipolar disorder. In

28

the area of familial Alzheimer's Disease, we have been involved in the screening of

numerous small Australian pedigrees for the presence of mutations. One such early

onset Alzheimer's Disease pedigree possesses the common Val 717 mutation in the

amyloid precursor protein (APP) (69). We have also identified another pedigree

which has a novel mutation in the APP gene and are currently examining the effect of

this mutation on amyloid production prior to publishing the results of this study.

Other studies are aimed at investigating the role of the recently identified presenilin

genes in the amyloid cascade hypothesis of Alzheimer's Disease.

Although I have only published two relatively minor papers in the field of brain

dysfunction, it is my intention to develop this area as an important area of

investigation. The skills that I have developed and applied, particularly in the analysis

of the ligand-gated ion channel receptors, in combination with the human genetic

(pedigree) resources that have been acquired over the last two years, will allow me to

make significant contributions to the further understanding of brain function and

dysfunction. In particular, as we now know much more about the precise workings

of the brain's key signalling molecules and how disruption can lead to neurological

disease, it seems timely to extend my studies on the brain to include the analysis of

higher cognitive functions, such as mood and cognition, through the investigation of

the genes that are causative and contributory to the development of bipolar disorder

and Alzheimer's Disease. It is entirely possible that this work will lead me away from

my chosen field of membrane receptors and their ligands. However, this may not

necessarily be the case as for example, the presenilin genes that play a role in the

etiology of Alzheimer's Disease appear to most closely resemble an ion channel or

receptor molecule. Whatever the outcome, understanding the molecular functions of

genes that cause brain dysfunction will also provide significant insights into brain

function.

30

5 . List of Publications

Use of the Science Citation Index

The current (1993) Science Citation hidex Impact Factors for papers pubhshed in refereed journals are indicated. These values represent the average number of times a paper, published in that particular joumal will be cited. The Number of Citations has been determined by searching the particular paper in the Science Citation Index (Dialog Scisearch, January 1996).

1. Shine, J., Scott, K.F., Fellows, F.F., Djordjevic, M.A., Schofield. P.R.. Watson, J.M. and Rolfe, B.G. (1983) Molecular Anatomy of the symbiotic region in R. trifolii and R. parasponia. In: Molecular Genetics of the Bacteria-Plant Interaction. A. Puhler (Ed.), Springer Verlag, Berlin, Heidelberg pp 204-209.

SCI Impact Factor N/A Number of Citations 3

2. Watson. J.M. Schofield. P.R.. Ridge. R.W.. Djordjevic, M.A., Rolfe, B.G. and Shine, J. (1983) Molecular cloning and analysis of a region of the Sym plasmids of Rhizobium trifolii encoding clover nodulation functions. In: Plant Molecular and Cellular Biology, New Series Vol XII, R.B. Goldberg (Ed.), Alan R. Liss, New York, pp. 303-318.


3. Schofield. P.R.. Djordjevic, M.A., Rolfe, B.G., Shine, J. and Watson, J.M. (1983) A molecular liricage map of nitrogenase and nodulation genes in Rhizobium trifolii. Molecular and General Genetics 192: 459-465.

SCI Impact Factor 3.2 Number of Citations 64

4. Schofield. P.R.. Ridge. R.W., Rolfe, B.G., Shine, J. and Watson, J.M. (1984) Host-specific nodulation is encoded on a 14 kb DNA fragment in Rhizobium trifolii. Plant Molecular Biology 3: 3-11.


31

5. Scott, D.B., Court, C.B., Ronson, C.W., Scott, K.R, Watson, J.M. Schofield. P.R.. and Shine, J. (1984) Organisation of nitrogen fixation and nodulation genes on a Rhizobium trifolii symbiotic plasmid. Archives of Microbiology 139: 151-157


6. Rolfe, B.C., Scott, K.F., Schofield. P.R.. Watson, J. M., Plazinski, J and Djordjevic, M.A. (1985) Genetic and molecular analysis of host range nodulation genes in Rhizobium trifolii and the Parasponia Rhizobium strain ANU 289. In: Advances in Molecular Genetics of the Bacteria-Plant Interaction. A. A. Szalay and R.P. Legocki (Eds.) Media Services, Cornell University, Ithaca, N.Y. pp. 4-48.


7. Djordjevic, M.A., Schofield. P.R.. Ridge, R.W., Bassam, B.J., Plazinski, J., Watson, J.M. and Rolfe, B.G. (1985) Rhizobium nodulation genes involved in root hair curling (Hac) are functionally conserved. Plant Molecular Biology 4: 147-160.


Watson, J.M. and Schofield. P.R. (1985) Species-specific, symbiotic plasmid-located repeated DNA sequences in Rhizobium trifolii. Molecular and General Genetics 199: 279-289.


9. Schofield. P.R.. and Watson, J.M. (1985) Conservation of nif- and species-specific domains with repeated promoter sequences from fast-growing/ /z/zoM'Mm species. Nucleic Acids Research 13: 3407-3418.


10. Djordjevic, M.A., Schofield. P.R.. and Rolfe, B.G. (1985) Tn5 mutagenesis Rhizobium trifolii host-specific nodulation genes result in mutants with altered host range ability. Molecular and General Genetics 200: 463-471.


11. Watson, J.M. lismaa, S.E. and Schofield. P.R. (1985) Repeated DNA sequences in fast-growing Rhizobium species. In: Advances in Molecular Genetics of the Bacterial-Plant Interaction. A. A. Szalay and R.P. Legocki (Eds.) Media Services, Cornell University, Ithaca, N.Y. pp. 22-26.


32

12. Schofield. P.R. and Watson, J.M. (1986) DNA sequence of Rhizobium trifolii nodulation genes reveals a reiterated and potentially regulatory sequence proceeding nodABC and nodFE. Nucleic Acids Research 14: 2891-2903.


13. Djordjevic, M.A., Innes, R.W., Wijffelman, C.A., Schofield. P.R.. and Rolfe, B.C. (1986) Nodulation of specific legumes is controlled by several distinct loci in Rhizobium trifolii. Plant Molecular Biology 6: 389-401.


14. Rolfe, B.C., Innes, R.W., Schofield. P.R.. Watson J.M. Sargent, C.L., Kuempel, P.L., Plazinski, J., Canter-Cremers, H. and Djordjevic, M.A. (1986) Plant-secreted factors induce the expression of R, trifolii nodulation and host-range genes. In: Nitrogen fixation research progress. H. J. Evans, P.J. Bottomley and W.E. Newton (Eds.) Martinus Nijhoff, Dordrecht, pp. 79-85.


15 Schofield. P.R.. lismaa, S.E., Watson, J.M., Dudman, W.F. and Gibson, A.H. (1987) Symbiotic gene exchange between Rhizobium trifolii strains in soil. In: Temperate Pastures; their production, use and management. J.L. Wheeler, C.J. Pearson and G.E. Robards (Eds.) AWC Technical Publication/CSIRO Australia, pp. 208-210.


16. Schofield. P.R.. Gibson, A.H., Dudman, W.F. and Watson, J.M. (1987) Evidence for genetic exchange and recombination of Rhizobium symbiotic plasmids in a soil population. Applied and Environmental Microbiology 53: 2942-2947.


33 Publications for consideration for the degree of Doctor of Science

17. Schofield. P.R.. Darlison, M. G., Fujita, N. Burt, D. R., Stephenson, F. A. Rhee, L. M., Rodriguez, H., Ramachandran, J., Glencorse, T.A., Reale, V. Seeburg, P.H. and Barnard, E.A. (1987) Sequence and functional expression of the G A B A A receptor shows a ligand-gated receptor superfamily. Nature 328: 221-227. Article included cover illustration. A citation classic and the 26th most cited life-sciences paper of 1987. (Current Contents 33: 3-18. 1989)


18. Schofield. P.R.. Rhee, L.M. and Peralta, E.G. (1987) Primary structure of the human beta-adrenergic receptor gene. Nucleic Acids Research 15: 3636. SCI Impact Factor 3.8 Number of Citations 15

19. Malfroy, B. Schofield. P.R.. Kuang, W. - J., Seeburg, P.H., Mason, A. J. and Henzel, W.J. (1987) Molecular cloning and amino acid sequence of rat enkephalinase. Biochemical and Biophysical Research Communications 144: 59-66. SCI Impact Factor 3.3 Number of Citations 154

20. Hipper, B.A., Park, L. P., Dickerson, I.M., Keutmann, H.T., Thiele, E.A., Rodriguez, H., Schofield. P.R. and Mains, R.E. (1987) Stmcture of the precursor to an enzyme mediating carboxyl-terminal amidation in peptide biosynthesis. Molecular Endocrinology 1:777-790. SCI Impact Factor 6.4 Number of Citations 137

21. Grenningloh, G., Gundelfinger, E., Schmitt, B., Betz, H., Darlison, M.G., Barnard, E.A., Schofield P.R. and Seeburg, P.H. (1987) Glycine vs. GABA receptors (Scientific Correspondence) Nature 330: 25-26. SCI Impact Factor 22.3 Number of Citations 39

22. Braun, T., Schofield P.R.. Pritchett, D. B., Shivers B.D. and Seeburg, P.H. (1987) A novel subtype of muscarinic receptor identified by homology screening. Biochemical and Biophysical Research Communications 149: 125-132. SCI Impact Factor 3.3 Number of Citations 44

34 23. Levitan, E.S.,* Schofield. P.R..* Burt, D.R., Rhee, L.M., Wisden, W.,

Koehler, M., Fujita, N., Rodriguez, H., Stephenson, F. A., Darlison, M. G., Barnard, E. A., and Seeburg, P. H. (1988) Structural and functional basis for G A B A A receptor heterogeneity. Nature 335: 76-79. * ESL & PRS are equal first authors. SCI Impact Factor 22.3 Number of Citations 488

24. Malfroy, B., Kuang, W.-J., Seeburg, P.H., Mason, A.J., and Schofield. P.R. (1988) Molecular cloning and amino acid sequence of human enkephalinase (neutral endopeptidase). Letters 229'. 206-210. SCI Impact Factor 3.3 Number of Citation 111

25. Schofield. P.R. (1988) Carrier-bound odorant delivery to olfactory receptors. Trends in Neuroscience 11: 471-472. SCI Impact Factor 17.3 Number of Citations 9

26. Pritchett, D.B., Sontheimer, H., Gorman, C.M., Kettenmann, H., Seeburg, P.H. and Schofield. P.R. (1988) Transient expression shows ligand gating and allosteric potentiation of G A B A A receptor subunits. Science 242: 1306-1308. SCI Impact Factor 21.1 Number of Citations 170

27. Barnard, E.A., Darlison, M.G., Fujita, N., Glencorse, T. A., Levitan, E.S., Reale, V., Schofield. P.R.. Seeburg, P.H., Squire, M. and Stephenson, F.A. (1988) Molecular biology of the G A B A A receptor. In: Advances in Experimental Medicine & Biology Vol 236, 'Neuroreceptors and Signal Transduction'. S. Kito, T. Segawa, K. Kuriyama, M. Tohyama and R. W. Olsen (Eds.) Raven Press, New York, USA. pp. 31-45. * Authors listed in alphabetical order SCI Impact Factor N/A Number of Citations 6

28. Schofield. P.R.. McFarland, K. C., Hayflick, J. S., Wilcox, J. N., Cho, T. M., Roy, S., Lee, N. M., Loh, H.H. and Seeburg, P.H. (1989) Molecular characterization of a new immunoglobulin-superfamily protein with potential roles in opioid binding and cell contact. The EMBO Journal 8: 489-495. SCI Impact Factor 13.2 Number of Citations 44

35 29. Gorman C.M., Gies, D., Schofield. P.R.. Kado-Fong, H. and Malfroy, B. (1989) Expression of enzymatically active enkephalinase (neutral endopeptidase) in mammalian cells. Journal of Cellular Biochemistry 39: 277-292.


30. Schmieden, V., Grenningloh, G., Schofield. P.R. and Betz, H. (1989) Functional expression in Xenopus oocytes of the strychnine binding 48-kD subunit of the glycine receptor. The EMBO Journal 8: 695-700. SCI Impact Factor 13.2 Number of Citations 92

31. Schofield. P.R.. Pritchett, D.B., Sontheimer, H., Kettenmann, H. and Seeburg, P.H. (1989) Sequence and expression of human GAB A a receptor a l and 61 subunits. FEBS Letters 244: 361-364. SCI Impact Factor 3.3 Number of Citations 59

32. Pritchett, D.B., Sontheimer, H., Shivers, B.D., Ymer, S., Kettenmann, H., Schofield. P.R. and Seeburg. P.H. (1989) A novel GAB A A receptor subunit important for benzodiazepine pharmacology. Nature 338: 582-585. SCI Impact Factor 22.3 Number of Citations 610

33. Barnard, E.A., Burt, D.R., Darlison, M.G., Fujita, N., Levitan, E.S., Schofield. P.R.. Seeburg, P.H., Squire, M. and Stephenson, F.A. (1989) Molecular biology of the GAB AA receptor. In: Allosteric modulation of amino acid receptors: Therapeutic implications. Fidia Research Foundation Symposium Series Vol. 1. E.A. Barnard and E. Costa (Eds.) Raven Press, New York, USA. pp. 19-30. * Authors listed in alphabetical order. SCI Impact Factor N/A Number of Citations 19

34. Ymer, S.,* Schofield. P.R..* Draguhn, A., Werner, P., Koehler, M. and Seeburg, P.H. (1989) G A B A A receptor B subunit heterogeneity: Functional expression of cloned cDNAs. The EMBO Journal 1665-1670. * S Y & PRS are equal first authors. SCI Impact Factor 13.2 Number of Citations 300

36 35. Sontheimer, H., Becker, C.-M., Pritchett, D.B., Schofield. P.R..

Grenningloh, G., Kettenmann, H., Betz, H. and Seeburg, P.H. (1989) Functional chloride channels by mammalian cell expression of rat glycine receptor s u b u n i t . 2 : 1491-1497. SCI Impact Factor 17.3 Number of Citations 91

36. Schofield. P.R. and Abbott. A. (1989) Molecular pharmacology and drug action: structural information casts light on ligand binding. Trends in Pharmacological Sciences 10: 207-212. SCI Impact Factor 16.2 Number of Citations 9

37. Shivers, B.D., Killisch, I., Sprengel, R., Sontheimer, H., Koehler, M., Schofield. P.R. and Seeburg. P.H. (1989) Two novel GAB A A receptor subunits are expressed in distinct neuronal subpopulations. Neuron 3: 327-337. SCI Impact Factor 17.3 Number of Citations 356

3 8 . Schofield. P.R. ( 1 9 8 9 ) The G A B A A receptor: molecular biology reveals a compex picture. Trends in Pharmacological Sciences 10: 4 7 6 - 4 7 8 .


39. Ymer, S., Schofield. P. R.. Shivers, B.D., Pritchett, D.B.,Lueddens, H., Koehler, M., Werner, P., Sontheimer, H., Kettenmann, H. and Seeburg, P.H. (1989) Molecular studies of the GABKp^i&c&i^iov. Journal of Protein Chemistry 8: 352-355. SCI Impact Factor 1.6 Number of Citations 0

40. Ymer, S., Draguhn, A., Koehler, M., Schofield. P.R. and Seeburg, P.H. (1989) Sequence and expression of a novel G A B A A receptor a subunit. FEBS Letters 25S: 119-122. SCI Impact Factor 3.3 Number of Citations 82

41. Buckle, V.J., Fujita, N., Ryder-Cook, A.S., Derry, J.M.J., Barnard, P.J., Lebo, R.V., Schofield. P.R.. Seeburg, P.H., Bateson, A.N., Darlison, M . G . and Barnard, E . A . (1989) Chromosomal localization of G A B A A receptor subunit genes: Relationship to human genetic disease. Neuron 3: 647-654. SCI Impact Factor 17.3 Number of Citations 76

37 42. S d i o f i e l d , ^ ^ , Shivers, B.D. and Seeburg, P.H. (1990) The role of

receptor subtype diversity in the CNS. Trends in Neurosciences 13: 8-11. SCI Impact Factor 17.3 Number of Citations 73

43. Grenningloh, G., Schmieden, V., Schofield. P. R.. Seeburg, P.H., Siddique, T., Mohandas, T.K., Becker, C.-M. and Betz, H. (1990) Alpha subunit variants of the human glycine receptor: primary structures, functional expression and chromosomal localization of the corresponding genes. The EMBO Journal 9: 771-776. SCI Impact Factor 13.2 Number of Citations 103

44. Ymer, S., Draghun, A., Wisden, W., Werner, P., Keinaenen, K., Schofield. P.R.. Sprengel, R., Pritchett, D.B. and Seeburg, P.H. (1990) Structural and functional characterization of the yl subunit of GABAA/benzodiazepine receptors. The EMBO Journal 9: 3261-3267. SCI Impact Factor 13.2 Number of Citations 146

45. Braun, T., Schofield. P.R. and Sprengel. R. (1991) Amino-terminal leucine-rich repeats in gonadotropin receptors determine hormone selectivity. The EMBO Journal 10: 1885-1890. SCI Impact Factor 13.2 Number of Citations 88

46. Schofield. P.R.. Ce^a-Poljak, A., Albrecht, M.F., Stuart, M.C. and Hort, Y.J. (1992) Biological activities of recombinant human FSH. In: Follicle Stimulating Hormone; Regulation of secretion and molecular mechanisms of action. Serono Symposia USA, Proceedings. (Eds.) Hunzicker-Dunn, M. and Schwartz, N.B. Springer-Verlag, New York, USA. pp 330-334. SCI Impact Factor N/A Number of Citations 3

47. Smith, G.M., Bishop, L.A., DeKroon, R., Wright, G., Cerpa-Poljak, A. and Schofield. P.R. (1992) Purification and characterisation of recombinant human FSH. In: Follicle Stimulating Hormone; Regulation of secretion and molecular mechanisms of action. Serono Symposia USA, Proceedings. (Eds.) Hunzicker-Dunn, M. and Schwartz, N.B. Springer-Verlag, New York, USA. pp 335-338. SCI Impact Factor N/A Number of Citations 3

48. Vandenberg, R.J., French, C.R., Barry, P. H., Shine, J. and Schofield, P.R. (1992) Antagonism of ligand-gated ion channel receptors: Two domains of the glycine receptor a subunit form the strychnine binding site. Proceedings of the National Academy of Sciences U.S.A. 89: 1765-1769. SCI Impact Factor 10.3 Number of Citations 32

38

49. Vandenberg, RJ., Handford. C.A. and Schofield. P.R. (1992) Distinct agonist- and antagonist-binding sites on the glycine receptor. Neuron 9: 491-496. SCI Impact Factor 17.3 Number of Citations 31

50. Cerpa-Poljak, A., Bishop, L.A., Hort, Y.J., Chin, C.C.K., Dekroon, R., Mahler, S.M., Smith, G.M., Stuart, M.C. and Schofield P.R. (1993) Isoelectric charge of recombinant human follicle-stimulating hormone isoforms determines receptor affinity and in vitro bioactivity. Endocrinology 132: 351-356. SCI Impact Factor 4.3 Number of Citations 10

51. Vandenberg, R.J. and Schofield. P.R. (1993) The importance of being inhibited: brain GABAA and glycine receptors. Today's Life Sciences 5: 20-26.

Article included cover illustration. SCI Impact Factor N/A Number of Citations 1

52. Vandenberg, R.J., Rajendra, S., French, C.R., Barry, P.H. and Schofield. P.R. (1993) The extracellular disulfide loop motif of the inhibitory glycine receptor does not form the agonist binding site. Molecular Pharmacology 44: 198-203. SCI Impact Factor 5.4 Number of Citations 9

53. Vandenberg. R.J, and Schofield. P.R. (1994) Inhibitory ligand-gated ion channel receptors: Molecular biology and pharmacology of GABAA and glycine receptors. In: Handbook of Membrane Channels: Molecular and Cellular Physiology (Ed.) Peracchia, C. Academic Press, San Diego, USA. pp317-332. SCI Impact Factor N/A Number of Citations 0

54. Townsend-Nicholson. A. and Schofield. P.R. (1994) A threonine residue on the seventh transmembrane domain of the human A1 adenosine receptor mediates specific agonist binding. Journal of Biological Chemistry 269: 2373-2376. SCI Impact Factor 6.8 Number of Citations 14

39

55. Baker, E., Sutherland, G.R. and Schofield. P.R. (1994) Localization of the glycine receptor a l subunit gene (GLRAl) to chromosome 5q32 by FISH Genomics 22: A9l-A9?>.


56. Rajendra, S., Lynch, J.W., Vandenberg, R.J., Pierce, K.D., French, C.R., Barry, P.H. and Schofield. P.R. (1994) The functional structure of the human inhibitory glycine receptor. In: Studies in Honour of Karl Julius Ullrich; An Australian Symposium. (Eds.) Poronnik, P., Cook, D.I. and Young, J.A. Wild and Woolley Press, Sydney, Australia, pp 53-56.


57. Le, F., Mitchell, P., Vivero, C., Waters, B., Donald, J., Selbie, L., Shine, J. and Schofield. P.R. (1994) Exclusion of close linkage of bipolar disorder to the Gs-a subunit gene in nine Australian pedigrees. Journal of Affective Disorders 32\ 187-195.

SCI Impact Factor L7 Number of Citations 0

58. Rajendra, S., Lynch, J.W., Pierce, K.D., French, C.R., Barry, RH. and Schofield. P.R. (1994) Startle disease mutations reduce the agonist sensitivity of the human inhibitory glycine receptor. Journal of Biological Chemistry 269: 18739-18742.


59. Bishop, L.A., Robertson, D.M., Cahir, N and Schofield. P.R. (1994) Specific roles for the asparagine-linked carbohydrate residues of recombinant human follicle stimulating hormone in receptor binding and signal transduction. Molecular Endocrinology 722-731.


60. Ryan, S.G., Buckwalter, M.S., Lynch, J.W., Handford, C.A., Segura, L., Shiang, R., Wasmuth, J.J., Camper, S.A., Schofield. P.R. and O'Connell, P. (1994) A missense mutation in the gene encoding the a l subunit of the inhibitory glycine receptor causes the spasmodic mouse phenotype. Nature Geneticsl: 131-135.


40 61. Rajendra, S., Lynch, J.W., Pierce, K.D., French, C.R., Barry, P.H. and Schofield. P.R. (1995) Mutation of a single amino acid in the human glycine receptor transforms p-alanine and taurine from agonists into competitive antagonists.//ewrow 14: 169-175.


62. Bishop, L.A., Nguyen, T.V. and Schofield. P.R. (1995) Both of the beta subunit carbohydrate residues of follicle stimulating hormone determine the metabolic clearance rate and in vivo potency. Endocrinology 136: 2635-2640. SCI Impact Factor 4.3 Number of Citations 1

63. Townsend-Nicholson, A., Baker, E., Sutherland, G.R. and Schofield. P.R. (1995) Localization of the adenosine A2b receptor subtype gene (ADORA2B) to chromosome 17pl 1.2-pl2 by FISH and PCR screening of somatic cell hybrids. Genomics 25: 605-607. SCI Impact Factor 5.4 Number of Citations 3

64. Lynch, J.W., Rajendra, S., Barry, P.H. and Schofield. P.R. (1995) Mutations affecting the glycine receptor agonist transduction mechanism convert the competitive antagonist, picrotoxin, into an allosteric potentiator. Journal of Biological Chemistry 270: 13799-13806. SCI Impact Factor 6.8 Number of Citations 0

65. Rajendra, S. and Schofield. P.R. (1995) Molecular mechanisms of inherited startle syndromes. Trends in Neuroscience 18: 80-82. SCI Impact Factor 17.3 Number of Citations 1

66. Townsend-Nicholson, A., Baker, E., Schofield, P.R. and Sutherland, G.R. (1995) Localization of the adenosine A1 receptor subtype gene (ADORAl) to chromosome lq32.1. Genomics 26: A23-A25. SCI Impact Factor 5.4 Number of Citations 4

67. Rajendra, S., Vandenberg, R. J., Pierce, K.D., Cunningham, A.M., French, P., Barry, P.H. and Schofield. P.R. (1995) The unique extracellular disulfide loop of the glycine receptor is a principal ligand binding element. The EMBO Journal 14: 2987-2998.

Article included cover illustration. SCI Impact Factor 13.2 Number of Citations 0

41

68. Grossmann, M., Szkudlinski, M.W., Tropea, J.E., Bishop, L.A., Thotakura, N.R.. Schofield. P.R. and Weintraub. B.D. (1995) Expression of human thyrotropin in cell lines with different glycosylation patterns combined with mutagenesis of specific glycosylation sites: Characterization of a novel role for the oligosaccharides in the in vitro and in vivo bioactivity. Journal of Biological Chemistry 270: 29378-29385.


69. Brooks, W.S., Martins, R.N., De Voecht, J., Nicholson, G.A., Schofield. P.R.. Kwok, J.B.J., Fisher, C., Yeung, L.U. and Van Broeckhoven, C. (1995) A mutation in codon 717 of the amyloid precursor protein gene in an Australian family with Alzheimer's disease. Neuroscience Letters 199: 183-186.


70. Handford, C.A., Lynch, J.W., Baker, E., Webb, G.C., Ford, J.H., Sutherland. G.R. and Schofield. P.R. (1995) The human glycine receptor p subunit: Primary structure, functional characterisation and chromosomal localisation of the human and murine genes. Molecular Brain Research (in press).

SCI Impact Factor 3.8 Number of Citations N/A

46

7 . Enclosed Publications

GABA RECEPTOR TURE

Reprinted from Nature, Vol. 328, No. 6127, pp. 221-227, 16 July 1987 © Macmillan Magazines Ltd., 1987

Sequence and functional expression of the GABAA receptor shows a ligand-gated receptor super-family Peter R. Schofield*^ Mark G. Darlison% Norihisa Fujita^ David R. Burf F. Anne Stephenson^ Henry Rodriguez*, Lucy M. Rhee\ J. Ramachandran*, Vincenzina Reale% Thora A. Glencorse% Peter H. Seeburg'^ & Eric A. Barnard^* * Genentech, Inc., Department of Developmental Biology, 460 Point San Bruno Boulevard, South San Francisco, California 94080, USA t MRC Molecular Neurobiology Unit, MRC Centre, Hills Road, Cambridge CB2 2QH, UK

Amino-acid sequences derived from complementary DNAs encoding the a- and p-subunits of the GABA/benzodiazepine receptor from bovine brain show homology with other ligand-gated receptor subunits, suggesting that there is a super-family of ion-channel-containing receptors. Co-expression of the in vitro-generated a-subunit and jB-subunit RNAs in Xenopus oocytes produces a functional receptor and ion channel with the pharmacological properties characteristic of the GABA^ receptor.

7-AMINOBUTYRIC acid ( G A B A ) , the major inhibitory neurotransmitter in the vertebrate brain, mediates neuronal inhi-bition by binding to the GABA/benzodiazepine (GABAA) receptor and opening an integral chloride channel. The structure of this receptor is of particular interest because it contains a variety of binding sites for pharmaceutically significant drugs which interact allosterically with the G A B A agonist site or the receptor channel. These include anxiolytic (benzodiazepines), anti-convulsant (barbiturates), anxiogenic (/S-carbolines) and convulsant (picrotoxin) agents (reviewed in refs 1 and 2). The receptor complex has been affinity-purified from bovine cerebral cortex with all of those binding sites intact^-"^. The subunit structure of the complex has been shown^'® to be a2P2- Photo-affinity labelling experiments on the pure receptor have demon-strated that the a-subunit (apparent relative molecular mass 53,000; Mr 53K) carries the benzodiazepine binding site and the )3-subunit (apparent M^ 57K) the G A B A binding site^'^. Limited structural similarity is suggested by the recent finding® that only one of about twenty®'^ monoclonal antibodies raised so far against this receptor recognizes both the a - and the

chain. We have used this purified receptor to generate peptide

sequences, allowing the isolation of cloned cDNAs encoding both subunits. We here report the deduced amino-acid sequences of two GABA^-receptor subunits, which we find to have homologies in both primary sequence and predicted transmem-brane topology. Comparison of these sequences with those of other neurotransmitter receptors identifies structural features which appear characteristic of the chemically-gated ion chan-nels. Further, we demonstrate by functional expression in Xenopus oocytes that the cloned subunit cDNAs encode the entire GABAA receptor with its full pharmacological profile. cDNA and protein sequences The GABAA receptor was purified to homogeneity from bovine cerebral cortex by benzodiazepine-affinity chromatography, essentially as described previously'^. The purified preparation bound [^H]muscimol with high affinity and comprised only a-and )8-subunits when analysed by SDS-polyacrylamide gel elec-trophoresis under reducing conditions, as previously found^*'^. Gas-phase micro-sequencing of the entire protein failed to yield detectable signals, presumably due to blocked N-termini. But

t To whom correspondence should be addressed. § Present addresses: Laboratory of Molecular Neuroendocrinology, ZMBH, Universitat Heidel-berg, Im Neuenheimer Feld 282, D-6900 Heidelberg, FRG (P.R.S. and P.H.S.); Department of Pharmacology, University of Maryland School of Medicine, 655 West Baltimore Street, Baltimore, Maryland 21201, USA (D.R.B.)

cleavage of the whole receptor by cyanogen bromide and HPLC separation of the peptides resulted in several amino-acid sequen-ces, of which three (CN8.9, CN10.8 and CN10.9) were used for the design of synthetic oligodeoxyribonucleotide probes (see Fig. 1 legend). In addition, the a-subunit protein was isolated by electroelution from SDS-polyacrylamide gels and, after tryp-tic cleavage, a further a-specific peptide sequence (T33) was obtained.

Bovine brain and calf cerebral cortex cDNA libraries, con-structed® in phage AgtlO, were screened with the synthetic pro-bes. Hybridizing clones were analysed by DNA sequencing and found to be of two types. One specified an open reading frame which contained three of the chemically-determined peptide sequences, including the one specific for the a-subunit (CN8.9, CN10.8 and T33; see Fig. 1). The other type of clone did not contain the a-specific sequence (T33) but did contain the coding sequence for peptide CN10.9 (not found in the a-subunit cDNA) and was therefore designated as encoding the /3-subunit. The nucleotide and deduced amino-acid sequences of both subunits are shown in Fig. 1.

These cDNAs code for polypeptides with deduced protein sizes of 456 amino acids for the a-chain and 474 amino acids for the /3-chain. In both polypeptide sequences the N-terminal 25-30 residues constitute typical signal peptides^. As no N-terminal sequence was obtained from the purified GABAA recep-tor, our assignments for the actual lengths of these signal pep-tides (Fig. 1) remain tentative. The proposed mature subunit sizes (a-subunit, 429 residues, Mr48,800; ;8-subunit 449 residues, Mr 51,400) agree (within the error limits of SDS gel elec-trophoresis of hydrophobic polypeptides) with the values found^ for the deglycosylated GABAA receptor subunits (a-subunit, 44K; /3-subunit, 55K). Structural interpretations The deduced protein sequences (for alignment see Fig. 2a) and hydropathy profiles^® (Fig. 2b) reveal striking similarities in the structure and general architecture of both GABAA receptor subunits, suggesting their common evolutionary origin. The overall sequence identity of the mature subunits is 35%, while the homology based on conservative substitutions'' rises to 57%. Four hydrophobic sequences of membrane-spanning length (designated as M1-M4) are present in each subunit. A large N-terminal hydrophilic domain, likely to be extracellularly located, contains several potential sites for iV-linked glycosyla-tion (2 in the a-subunit, 3 in the )3-subunit). Carbohydrate attachment to the receptor is known to occur in vivo'^'^'^^, and would account for the differences between the observed'*'® and

1 CACGCAGAGTCCATGATGGCTCAGCCCGCTCAGACCGAGTGAGTGAGCGGCGGAGCGAGGACGCCCCTCCGCTGCCGGCGCGCCGGGACTCGCGGACTCG

101 CGGACTCGCGCCGGCTTCCAGCTCTACACGATTTTCTCTCGCAGACTTTTCCCGGGTCTGGAGCGATCCTGTGCCCAGAGGGGGCCCGAGCTGCACAAGC

Pro Asp Thr Phe Phe His Asn Gly lys Lys Ser Val Ala His Asn Met Thr Met 575 CCG GAT ACC TTT TTC CAC AAT GGA AAG AAG TCA GTG GCA CAC AAC ATG ACC ATG

Pro CCG

140 Pro CCA

Asn AAC

Lys AAG

Leu CTC

Leu CTG

Arg CGC

lie ATC

Thr Glu 650 ACA GAG

Asp GAC

Gly GGC

Thr ACT

Leu TTG

Leu Tyr CTG TAC

Thr Met Arg Leu Thr Val Arg Ala Glu Cys ACC ATG AGA CTG ACG GTG AGA GCC GAA TGT

Pro CCG

140 Pro CCA

Met ATG

His CAT

leu TTG

Glu GAG

Asp GAT

Phe TTC

Pro Met 725 CCT ATG

Asp GAT

150 Ala GCC

His CAT

Ala GCC

Cys Pro TGC CCT

Leu Lys Phe Gly Ser Tyr Ala Tyr ThrÂrg CTA AAA TTT GGA AGC TAT GCT TAT ACA AGA

Ala GCA

Glu GAA

Val GTT

Val GTT

Tyr TAT

170 Glu GAG

Trp TGG

Thr Arg 800 ACC AGA

Glu GAG

Pro CCA

Ala GCA

Arg CGC

Ser Val TCA GTG

180 Val Val Ala Glu Asp Gly Ser Arg Leu Asn GTT GTA GCC GAA GAT GGG TCA CGC CTG AAC

190 Gin CAG

Tyr TAC

Asp GAT

leu CTT

Leu CTT

Gly GGA

Gin CAA

Thr Val 875 ACA GTA

Asp GAT

200 Ser TCT

Gly GGA

H e ATT

Val Gin GTT CAG

210 Ser Ser Thr Gly Glu Tyr Val Val Met Thr TCT AGT ACA GGA GAG TAT GTT GTT ATG ACA

Thr ACT

His CAT

Phe TTC

His CAC

Leu CTG

220 Lys AAG

Arg AGA

1640 TGGTAATTCCCATCTGCTTTATTGCCTCIGTCTTAGAGAATTTGAAGGTTTCCNATTTCCATAATTCATATAAGAACAACAGACCCCTGTCTGGCCGTC

1740 CGGAGCAGATCAGAGCATACAGCTGAGAGFLCGGGATTCTGAGAGAGGGAGCCAGAGAGCAAAATCATGACAGAAGGAGACAGAAGGGAAGAGAGAGAAAA

1840 GGAGGGGGAAGTAGATCCAAAGATAGGAGGAAAAGTAGAAGAAAAATCCAACTTAACTCCAGGTCATTTGTAGATATATATTTCCAAATATTCTAGAAAA

1940 AAAAMAGATACTGTATATGTCAAAAATATTTTTATGTGAAGGTGTTTAAAAGAATATAATGTTTAATGAAGAAACATTTAAAAAAATCTATGTCTTTAT

2040 TGCACAGATGATGATGTGTTTCTCACATTTCTGTTTTGTATTTTAAACCTATGIATAGCTTTAACAGTTTGTNCCAAAGCTCAAGATCCCCATTCTTTC

2140 TCTTTTAAGAAATACCTAGGGCATTATTTTGTTGTTAAATGCTATTTTAAAATTTATGGAAATTGCATAGGCAAAGGTGCAGTGTCTCACAGTAAGAGCA

2240 CATTTAATCCAATGGAGACAAATGCTTTAAATAGGCTGCTTTACTGTCATCTGAGCTTTTACCAGTAGACTCAATGAGGAATCATTCTAACAGATATATA

2340 CACTCAATAAAACTAAAAAAAAAA

CN8.9 (M) A S K I X T P D T F F H N G X K (S) V A (I) (T) (S)

CN10.8 (M) P N K L L H I T E D G T I. L Y(T) (R) CN10.9 (M) X P H E N I L L S T L E I K(N)(E) T33 X G S Y A Y T

I K(N)(E)

F

1 TGCATTCCTTGAATCTTCGCAGAAAAGACAATTCTTTAATCAGAGTTAGTA ATG

-10 li Leu Ser Phe Pro »al Met H e Ala Met Val Cys Cys Ala Hfs

88 CTC TCT TTC CCT GTG ATG ATT GCC ATG GTC TGT TGT GCA CAC

Lys Glu Thr Va) Asp Arg Leu Leu Lys Gly Tyr Asp H e Arg 163 AAA GAG ACA GTG GAC AGA TTG CTC AAA GGA TAT GAC ATT CGC

<0 50 Val Gly Met Arg H e Asp Val Ala Ser H e Asp Met Val Ser

238 GTC GGG ATG AGG ATC GAT GTC GCC AGC ATA GAC ATG GTC TCC

Trp Thr Val Gin Asn Arg Glu Ser Leu Gly Leu TGG ACA GTA CAA AAT CGA GAG AGT TTG GGG CTT

Ser Ala Asn Glu Pro Ser Asn Met Ser Tyr Val AGC GCC AAC GAG CCC AGC AAC ATG TCA TAC GTG

30 Leu Arg Pro Asp Phe Gly Gly Pro Pro Val Asp TTG CGG CCG GAC TTT GGA GGG CCC CCC GTG GAC

Glu Val Asn Met Asp Tyr Thr Leu Thr Met Tyr GAA GTG AAC ATG GAT TAT ACT CTC ACC ATG TAT

^ e Gin Gin Ser Trp Lys Asp Lys Arg Leu Ser Tyr Ser Gly H e Pro Leu Asn Leu Thr Leu Acn J^n Arn ,313 n c CAG CAG TCT TGG AAA GAC AAA AGG CTT TCT TAT TCT GGA ATC ccl TO AIS ETC A M CU GIC W SGG G?G

90 100 110 ?AT rlS I!! ^ ^ Gly Val Thr Val Lys

388 GCT GAT CAA CTC TGG GTA CCA GAC ACC TAC TTT CTG AAT GAC AAG AAA TCC H I GTG CAT GGG GTT ACA GTG AAA

120 130 Asn Arg Met lie Arg Leu His Pro Asp Gly Thr Val Leu Tyr Glv Leu Aro H e Thr Thr Thr Ala ai» rv. >•.»

463 AAT CGC ATG ATC CGC CTG CAT CCT GAT GGI ACA GTT CTC TAC GGA CTC CGI I k ACC J M GJ? GCG W Are

Met Asp Leu Arg Arg Tyr Pro Leu Asp Glu Gin Asn Cys Thr Leu Glu lie Glu Ser Tvr Glv Tvr 538 ATG GAT CTA CGA AGA TAT CCT CTG GAT GAA CAA AAC T K ACC CTG GAG ATT S M AGT TAT GGC T «

170 180 11? rlj lif!: I;;? ^^r Gly Val Asn Lys lie Glu Leu Pro

613 GAC ATT GAG TTT TAC TGG AAT GGA GGA GAG GGC GCA GTA ACA GGA GTG AAT AAA ATT GAG CTC CCT (

190 200

'1= l-J'S Het Val Ser Lys Lys Val Glu Phe Thr Thr Gly Ala Tyr Pro Arg Leu Ser I 688 ATT GTT GAT TAC AAG ATG GTG TCC AAA AAG GTG GAA TTC ACA ACA GGG GCA TAC CCA CGA CTG TCA I

?r? lir Ijl êr Thr Leu lie Thr H e Leu Ser Trp 763 CGT CTA AAA AGA AAC ATC GGT TAC TTC ATT TTG CAA ACC TAC ATG CCT TCC ACA CTG ATT ACA A H CTG TCG TGG

I""" Tyr Asp Ala Ser Ala Ala Arg Val Ala Leu Gly H e Thr Thr Val Leu Thr Met Thr 838 GTG TCG TTT TGG ATC AAT TAT GAC GCA TCT GCA GCC AGA GTT GCA CTA GGA ATC ACC ACA GTG CTG ACA ATG ACC

ij^ It!; i!: !:?* '•>•5 '•'•o yr Val Lys Ala H e Asp H e Tyr Leu Met Gly 913 ACC ATC AGC ACT CAC CTC AGG GAG ACC CTG CCA AAG ATT CCT TAT GTC AAA GCG ATT B K A H T « CTG ATG GGT

„oo T^^ iji Sis Pê Tyr H e Phe Phe Gly Lys Gly Pro Gin 988 TGC TTC GTG TTT GTG TTC CTG GCT CTG CTG GAA TAT GCC H T GTA AAT TAC ATC TTT TTT GGJI AAG K C CCT C M

320 330 ^^^ '•>'5 *sn Lys Leu Glu Met Asn Lys Val Gin Val Asp

1063 AAG AAA GGA GCT GGC AAA CAA GAC CAG AGC GCT AAT GAG AAG AAC AAA CTG GAG ATG AAT AAA GTC CAG GTC GAC

,,,0 JJJ SI? U s t®" *''9 ' sn Glu Thr Ser Gly Ser Glu Val Leu Thr Gly Val 1138 GCC CAC GGC AAC ATT CTC CTC AGT ACC CTG GAG ATC AGG AAC GAG ACG AGC GGC TCI GAA GTG CTC ACG GGC GTG

370 380

iS;? Iv. II!'' Ser Ala Ser H e Gin Tyr Arg Lys Pro Met Ser Ser Arg Glu 1213 GGA GAC CCC AAG ACC ACC ATG TAC TCC TAC GAC AGC GCC AGC ATC CAG TAC CGC AAG CCA ATG AGC AGC CGC GAG

390 400 4J0 Gly Tyr Gly Arg Ala Leu Asp Arg His Gly Ala His Ser Lys Gly Arg H e Arg Arg Arg Ala Ser Gin Leu Lys

1288 GGG TAC GGG CGC GCG CTG GAC CGG CAC GGC GCG CAC AGC AAG GGG CGC ATC CGC AGA CGC GCC TCG CAG CTC AAA

,,,, 'I® 5er H e Asp lys Trp Ser Arg Met Phe Phe Pro H e Thr Phe Ser 1363 GTC AAA ATC CCC GAC TTG ACT GAC GTG AAT TCC ATA GAC AAG TGG TCC CGA ATG TTC TTC CCC ATC ACC H T TCT

Leu Phe Asn Val Val Tyr Trp Leu Tyr Tyr Val His OC* 1438 CTT TTT AAC GTC GTT TAT TGG CTT TAC TAT GTA CAC TAA GGTCTGTCCCATGGTTCCTAAACGACTTCmCCTCTTCTTTTGGTT

1524 TTTAACCCCACAGGTCCCTAGCAGCGATACTGCTGTGCTTTTTGAGGTAAGAGATTCAGCCATCCAATTGGTTTTCGGTCTTGCATATCAAHTTATTAC

1624 TGCACCATGTTTACTTCAAAAAGAAACGAAAACTATTTTTTTCCAGTCTACTGTGGTCCAGGTTATTGGCTCTTTCAGAGCTTTGTTAATTGCCATGTN

1724 ACAAACACACACACACAAAGAGAGAAATTAGATAGGTAGATCTTTAGCAGTTCTAGTCGCCCTGGATTTTATTATTTTTTAAATGCAAGTGAAAAGAGGG

1824 CTTAGCTCTCGGTGTGCATGGCCTCCTGGTAAACTGTAACAATCTCATGCTGCCAAAAAACAATACAATTTCCAGGATCTCAGAAGAAAAAAAACAAAGC

1924 CATTGATAAGTTACAGATTTCCAGGAAGAAGGGAAAACAAAAAGGAGCCCAGAAGAAGAAGGAAGATTCCTTTCCACTCTCCATCNCCCCAGGTCCATT

2024 GTAGGAGTCCCTGCTCCACAGAGAAAGAGGGAGAGATAGGCAAGCTGGAACCCAGGTTATACAAACTCAGCCTGCCATCTTCACCCCACCTTTAAGCCAG

2124 ATATTCTGAAAAATGCTTCTTTCTCAGACTTCCAGGGAGATTTTAAAAATTAACATGAGAATCCAACATGGTAGCCATTGGGAACATnTTCTTCTCTTG

2224 TTAAGAATGAGTTTGGGGGGCAGTGATTAATTTTGTAAACTATTTCAAATGAGCAGCCCTGAACTTATTTCCTTAGCCTGGAATCACnGACCTCATGCC

2324 ATGAAGCTATGGCGTCACACTGTCCTAGTTCCTCCTTCAGAGATTAAGAGAAAGGGGGCTGGGGCAGGTGAGCTTTGAAGAAGGGAATGGTAAGAATTTT

2424 TGCTTGAGAAAAATATTGCCCTTTCCATGAGGAAGGCAAGACTTACATGCTGTGAAAAGGGTGCCTGGGCAGnACCTGAACTGCCAAATCTTTAATCAA

2524 ATAGTATATATTTATTTCTTAGCTTTCCAGAGAGCAGTATCTAATGGTGACAAGCTAGTTGCACAGCGCTAGTTTCTCAGCTAGCACAGTATGCCTCTGC

2624 TCTCTGAGCTCTGCATTCCAACACATGAAATGAATGCGTCAGGAAATGCACATATACCaCTTTAGTATCATTTCCCTTCAAGGCCAGATTTCTGTTTTA

2724 AAGAAAACCTTCAAAACAGCAAAACCACAGATAACAACATCACTGAGCCTTCAGACATTTACAACAGCCCCCAAATGACAGCCCTTCTCTACCATCATTA

2824 CTGCTTCTGCCTTATCTCTGCTCATGGGTTACTTTGCTCATCATAGAAGGAGTATTCCCTCAAGAGCTCAACACACAATTCTAATCAAATCCAACCAACT

2924 ACAGTTGAGATTCATCTCATTTATGTCAATAAAGCCTCATTATTTTTTTGCAGTGTGTGTTCAGCCTGAGTATGTATATAAGAATTC

Fig. 1 DNA sequence and predicted amino-acid sequence of the bovine GABAa receptor subunits: a, a-subunit; b, /S-subunit. Chemically-determined peptide sequences are overlined, with residues differing from those predicted indicated by dotted lines. Potential signal sequence cleavage sites are indicated by arrows. Amino-acid sequence numbering starts at the proposed mature N-terminal residue, the presumptive signal sequences being indicated by negative numbering The proposed membrane-spanning hydrophobic sequences are indicated by solid bars and putative extracellulariy-located AT-glycosylation sites are indicated by hatched bars. An asterisk denotes the first upstream in-frame stop codon in the 5' untranslated sequence. Methods. Affinity-purified G A B A A receptor protein was subjected to cyanogen bromide cleavage and the gel-eluted a-subunit was digested with trypsin; peptides were resolved by HPLC and subjected to gas-phase micro-sequencing'^''. The sequences used for probe construction and clone identification are shown in the boxed insert. For peptides cleaved by cyanogen bromide, the initial Met is assumed. Peptide CN10.9 arose by acidic cleavage at Asp 337. Oligo-dT-primed cDNA libraries were constructed in AgtlO, using adult brain or calf cerebral cortex poly(A"')RNA isolated in guanidine isothiocyanate, using EcoRl-Xho\ adaptors or £coRI linkers respectively. Libraries were screened with [y-^^P]ATP-labelled oligonucleotides and DNA from positive phage was sequenced from both strands by the chain termination method, using complete cDNAs or £coRI restriction fragments, combined with specific oligonucleotide priming** ^ The tt-subunit sequence presented is from a single adult cDNA clone. For the ^-subunit sequence, the two largest cDNA clones obtained begin at nucleotide positions 160 and 207 in b. The shorter clone (starting at position 207) lacks a T at position 1,042, this probably being an artifact of reverse transcription of the mRNA. Both of these clones contain - 1 . 5 kilobases (kb) of additional 3'-untranslated region that was not sequenced. A specifically-primed cDNA library was constructed (primer, 5'-A G G A T T T C T T G T C A T T C A G A - 3 ' ) using adult brain poly(A"^) RNA and the complete 5' end of the yS-subunifcDNA sequence was obtained from one of these clones. No differences in the sequences (where sampled) were found in the cDNA clones from the calf or the adult libraries. The nucleotide sequences of the subunit cDNAs presented are shorter in length than their corresponding mRNAs, as determined by Northern analysis of bovine brain RNA (data not shown),

and all the cDNAs lack their cognate 3' poly-A sequences.

a QABARa GABAR tlACh R a

GABARa GABARP nAChRa

QABARa GABARP nAChRa

GABARa GABARP nAChRo

GABARa GABARP nACh R o

GASARa GABARP nAChRa

GABARa GA8ARP nACh R a

GABARo GABARp nAChRo

GABARa GABARp nACh R a

GABARa GABARP nAChRa

GABARa GABARp nACh R a

L\(Uy(N|N*OODFAl|y]KFT KvITTJotIiIq H i llJwT P P - - Al F K S v|c|e I I V T hF pI F

I S f t g S J ] ; ; .

YApfrlATlilYTPNLAROD LLslllLExflNETSGSEV

a :

rovGDPKTlrlMYSY oIsIa s I 0 y r k[p1ms s re g ekodkkQfItIedi di[s|di sgkp g[p3p mo f h

L 0 R "h G A „ s ; G RfHR "r nns o L r " T To S k S m ^ S S s l ® KHPEVKSAJ EG1]JKY||AJETMIK1SDQES N0A A E EWKYVAMV MDHII LIL A vlFlMfTv

>H@ .0@G

200 300

. ACh R a

^v/Hl. nA 100 200 300

Amino acids

400-

300-

cc £

< 2 0 0 -

100-

400-

300-1

cc

§ 200-<

100-

1

1 /

1 /

1 • / \

/

/

100 200 300 400

GABA R a

100 200 300

GABA R ^

400

Fig. 2 Homologies o f bovine G A B A ^ receptor a- and )8-subunits and the bovine nicotinic acetylcholine receptor a-subunit. a, Primary sequence homology: identical residues are boxed. The overall sequence identity between each G A B A ^ receptor subunit and the bovine nAChR a-subunit is definite but low: 18.5% ( a ) and 15% ()3). b. Hydropathy profiles computed according to Kyte and Doolittle'"; window size is 17 residues, plotted with a 1-residue interval. Similar profiles were obtained when the hydropathy index of Hopp and Woods'^^ was used, c, Diagonal matrix ( D I A G O N ) conservative homology comparison, computed"*^ using the mutation data matrix MDMyg, o f the bovine muscle nAChR a-subunit*'^ versus (left) the a-subunit or (right) the ^-subunit o f the G A B A ^ receptor. The points correspond to mid-points o f 21-residue spans giving a double-matching probability o f <0.0005. Sequence regions o f strong similarity between the pair by this criterion (including strictly conservative replacements as well as identities) are revealed by unbroken diagonal segments. Very similar plots were obtained when other subunits o f the bovine muscle nAChR were used instead of its a-subunit. Solid bars indicate proposed signal sequences (b) and

transmembrane domains (b and c) .

Fig. 3 A loop-containing structure common to the GABAA and vertebrate nicotinic ACh receptor subunits. This loop is formed by presumed disulphide bonding between Cys 139 and Cys 153 (GABAA a-subunit numbering). Positions identical (or with one highly conservative variant, as shown) in all nAChR subunits and the )S-subunit of the GABA^ receptor are marked with uppercase symbols; h denotes that a hydrophobic residue only is found at that position. A hypothetical interaction between the C-terminal constant Tyr and hydrophobic side-chains of the loop may occur. In the one invertebrate nAChR subunit sequence known^", all of the positions marked are occupied similarly except for 152 (T), 154 (I) and 162 (F) (GABA^ receptor a-subunit numbering). In the bovine GABA^ receptor a-subunit all of this also holds, except that the three residues underlined differ from those in the fS-subunit. The disulphide-bridged loop can readily form (in a model)

a /8-structure, with a hairpin turn centred at the invariant Pro.

predicted molecular weights of the natural and cDNA-deduced G A B A A receptor subunits respectively.

Interestingly, no sequence homology is apparent in the region, presumably intracellularly located, w^hich connects M3 and M4. This region is of different length in the two subunits (87 residues in the a-subunit, 124 in the /3-subunit) and contains in the j8-subunit a unique cAMP-dependent serine phosphorylation consensus'^ sequence (Arg-Arg-Arg-Ala-Ser). Receptor super-family The domain distribution in the G A B A A receptor subunits is remarkably similar (Fig. 2fe) to that seen'"*-' in the various subunits of the nicotinic acetylcholine receptor (nAChR). All the polypeptides of these two classes are similar in size and contain an identical number and distribution of predicted hydro-phobic transmembrane regions. There is some homology between the sequences of either the a- or jS-subunits of the G A B A A receptor and any one of the bovine nAChR subunits; this homology is localized to several specific regions of the sequence (Fig. 2c). For the a-subunit of the G A B A A receptor, this is most notable for the Ml region and for some segments in the N-terminal hydrophilic domain, and in these limited regions sequence identity is 34% and homology (conservative'^ replacements) is 62%. For the )8-subunit the assumed membrane regions M1-M3 show the same levels of homology (Fig. 2c).

As a consequence of these homologies, several structural features emerge as common to both receptor types and may well prove of general functional significance for all ion channel receptors. Most notably, such shared features include specific conformational motifs within the extracellular ligand-binding domain, as well as structural specifications for the gated channel. Thus, a major feature is an extracellular )8-structural loop which can be formed by the disulphide bonding of two conserved cysteine residues (positions 139 and 153, G A B A A receptor a -subunit numbering). This loop (Fig. 3), already known to be

formed in vivo in the nAChR'^*'^, plus an adjacent region is now seen to contain three other invariant amino-acids (Pro 147, Asp 149 and Tyr 162) as well as strongly conserved residues at numerous other positions; these features are preserved in all the polypeptides of the G A B A A receptor and the 20 known ' - ' ® (including neuronal) nAChR subunits. This loop also contains an iV-linked glycosylation site which is conserved between the G A B A A )3-subunit and the nAChR subunits, and has been shown to be used in vivo in the latter^®-'^.

Although containing channels of mutually exclusive ion selec-tivity, G A B A A receptor and nAChR subunits share several features of their transmembrane sequences which are presumed to be important for channel function. Without exception, the Ml region in all these subunits contains a proline residue at an identical position and embedded in a conserved protein environ-ment, the consensus sequence being Pro-Cys-X-Leu (where Pro 1 is invariant, Cys 2 is replaced by Ser in the G A B A FI-subunit or Thr in an invertebrate nAChR subunit^", Leu 4 is replaced by Met in the G A B A a-subunit and X is any amino acid). The invariant proline residue is expected to introduce a bend in the transmembrane a-helix. The ensuing protrusion into the channel lumen might keep the channel closed in the absence of neurotransmitter. A second general feature of these channels, regardless of charge preference, is the abundance of threonine and serine residues in the M2 region of all of the subunits: there are 8 of these in M2 in each G A B A A receptor subunit, or 32 per receptor (assuming the 02^2 stoichiometry'-®). These residues may be important in the formation of a hydro-philic lining of the channel, conducive to ion flow. Indeed, there is chemicaP^'^^ and mutagenesis^^ evidence on the nAChR which has implicated the M2 segment in the ion channel structure.

Notably lacking in either subunit of the G A B A A receptor are certain features which have been considered of functional importance in the nAChR, in particular a proposed amphipathic helix (MA) located in the latter between the M3 and M4 domains (which has been proposed to line the channel lumen^"*), and two adjacent cysteines in the extracellular domain at positions 192-193 in every nAChR a-subunit (proposed to be in a channel-gating switch'^). Their absence in the G A B A A receptor indicates that these proposed structures are not a general feature of chemically-gated ion channels.

The structural similarities discussed above, however, strongly indicate that a super-family of chemically-gated ion channel receptors exists. This family is expected to include other amino-acid activated channels such as those gated by glycine or by excitatory amino-acid neurotransmitters. In fact, a concurrent report^^ on the sequence of a 48 K subunit of the rat spinal cord glycine receptor shows marked similarities to the transmembrane distribution and some of the other structural features discussed above. The presence of both anionic and cationic channel types within this family (including, even, nicotinic receptor anionic channels, as seen in the invertebrate Aplysia^^) suggests that the integral ion channel and the site for its activation can be discrete and separable domains of these proteins. This super-family can be compared in its generality with that of the G-protein-coupled receptors, which includes the various j8-adrenergic, muscarinic and opsin types^^"^®. In line with their entirely different transduc-tion mechanisms, these two protein super-families are not related at all in either sequence or domain pattern. Expression in the oocyte system mRNAs encoding membrane receptors can be translated con-veniently and efficiently in the Xenopus oocyte, which can fully assemble them in its membrane^"'^^ This system can produce functional G A B A A receptors when total mRNA from chick or rat brain is micro-injected^^'^^. To confirm the authenticity of the cloned a- and /8-5ubunit cDNAs as encoding the complete G A B A A receptor, we prepared a-subunit and )8-subunit RNAs by inserting the cloned cDNAs into plasmids containing the bacteriophage SP6 promoter and performing in vitro transcrip-

®

- 1 | 4 M - 0 - 3 M M — I 3 0 s

-auM

D

- 3 | J M + S U M P T X

-3MM 8 0 m i n w a s h

- 3 p M - 3 1 J M + 2 0 p M C L Z

- I p M -1pM + 2 5 | j M P B

V ( m V )

- i a Fig. 4 Expression of the GABA^ receptor in the Xenopus oocyte, a, GABA responses recorded from individual oocytes co-injected with the a- and y3-subunit-specific RNAs derived from the cloned cDNAs. Downward deflections indicate inward currents (see scale); the duration of the GABA application is indicated by the horizontal bar. A, Membrane conductance change evoked by 1 |jlM (i) or 0.3 FJIM (ii) GABA. B, Membrane conductance change evoked by: (i) 1 jxM GABA; (ii) 1 |xM GABA plus 5 )xM bicuculline methobromide (Bic); (iii) 1 p-M GABA plus 10 (xM bicuculline methobromide. C, Membrane conductance change evoked, in another RNA-injected oocyte, by: (i) 3 ^tm GABA; (ii) 3 (iM GABA plus 5 |xM picrotoxin (PTX); (iii) 3 [aM GABA, after washing out the picrotoxin for 80 min. D, Membrane conductance changes in response to: (i) 3 (xM GABA; (ii) 3 |xM GABA plus 20 jtM chlorazepate (CLZ). The latter was present also for 10 min before GABA application. E, Membrane conductance changes evoked by: (i) 1 |xM GABA; (ii) 1 j j l M GABA plus 25 (jiM ( - ) pentobarbital (PB). Note that no response to GABA was ever observed in numerous experiments (not illustrated) in which either the a- or )3-subunit-specific RNA was injected alone, b. Quantitative analyses of oocyte expression. A, GABA log-dose/conductance relationship. Ordinate, change in peak membrane conductance (g, in jiS); abscissa, GABA concentration (|xM, log scale). Typical data obtained from a single oocyte co-injected 48 h previously with a- and )8-subunit-specific RNAs. B, Measurement of the reversal potential for GABA (used at 1 |xM) in a single oocyte, co-injected with a- and ^-subunit-specific RNAs. The ordinate is g, expressed as a fraction of g measured at - 6 0 mV. The reversal pgtential is -27.5 mV. Methods. The entire a-subunit cDNA was subcloned by its £coRI adaptor sequence into pSP65 (ref. 34) (pbGRasense). The j8-subunit cDNA was constructed by using a Hiwdlll (restriction site in M13 polylinker)-£agI fragment from the specifically-primed cDNA clone and an Eagl-Bglll fragment from the longer of the two oligo(dT)-primed cDNA clones. Both the HindlU and Bglll sites were filled in with T4 DNA polymerase and the fragments ligated at the EagI site. The ligation product was gel-eluted and ligated to pSP65 (pbGR)3sense). RNA transcripts were obtained after cleavage of both template DNAs with / / m d l l l and transcribing in vitro with SP6 RNA polymerase^". Transcripts were capped with m'G(5')ppp(5')G. The synthetic RNA was injected into Xenopus oocytes (each subunit-specific RNA concentration was 0.25 ng n r ' ; total volume injected per oocyte, —30 nl); the oocytes were then incubated at 19 °C for 2 to 4 days in modified Earth's medium''". All recordings were made at 20-23 °C in amphibian Ringer solution"^. GABA-evoked currents were recorded by a two micro-electrode voltage-clamp technique at - 6 0 mV holding potential, with superfusion where indicated of GABA and other drugs"^. Chord conductances

were calculated, where g is shown.

tion^". The two pure R N A s produced were micro-injected, singly or together, into Xenopus oocytes and pulses of G A B A were subsequently superfused over the oocyte while recording mem-brane currents under voltage-clamp conditions. Only when both RNAs were injected together was any current response to G A B A application observed (Fig. 4a ) , indicating correct functional assembly. This response was large and immediate at, for example, 0.3 or 1 |jlM G A B A concentration, and (at a holding

potential of - 6 0 mV), the threshold for an observable current change was in the range of 10"^ M. The peak inward current was reproducibly dose-dependent (Fig. 4fe); the half-maximal G A B A concentration was 1.8 j j l M , noticeably lower than that (30-60 ixM) observed with rat" or chick^^ brain m R N A . Ele-vated expression from the pure m R N A s may account for the reduced desensitization, relative to that seen after rat brain m R N A injection^^, observed with a 1-min G A B A application.

H,N H,N

mm mm O • • • • • 4

a Fig. 5 A schematic model for the topology of the GAB A^ receptor in the cell membrane. Four membrane-spanning helices in each subunit are shown as cylinders. The structure in the extracellular" domain is drawn in an arbitrary manner, but the presumed )3-loop (Fig. 3) formed by the disulphide bond predicted at cysteines 139 and 153 is shown. Potential extracellular sites for iV-glycosylation are indicated by triangles. Those charged residues which are located within or close to the ends of the membrane domains are indicated as small circles with charge marked. The site for cAMP-dependent serine phosphorylation, present only in the )3-subunit, is marked by an encircled P. It is proposed that two copies of each of these subunit structures are complexed in the receptor molecule so as to align the membrane-spanning domains, only some of which will

form the inner wall of a central ion channel.

This desensitization may plausibly be modulated by serine phos-phorylation at the cAMP-dependent site found in the )8-subunit; such a system could become overloaded at an exceptional recep-tor density. GABA responsiveness was totally absent in appropriate controls or in oocytes injected with the a - or the /3-subunit RNA alone.

The current-voltage curve for the RNA-induced GABA response was regular in form and showed a reversal potential of -27.5 mV (Fig. 4b), which corresponds to the equilibrium potential of chloride in the Xenopus oocyte^^. This evidence, plus recent further work by E. Levitan and E. A. Barnard (to be published) showing that the extracellular chloride concentra-tion dependence and other properties are as predicted, demon-strates that an anion channel, formed by the injected RNAs, is opened by GABA in the oocyte membrane, as at brain synapses. Although the non-injected oocyte can also produce, upon activa-tion of its intrinsic muscarinic receptors, a fast inward chloride current^^ the channel induced here is a new and different one, as the two currents, quite apart from their entirely different controlling pharmacologies, are unlike in their voltage depen-dencies and other channel characteristics (to be described else-where).

The pharmacology of the oocyte-expressed receptor was com-pared with that of the native G A B A A r e c e p t o r ' C u r r e n t flow

in the oocyte-expressed receptor was strongly blocked (Fig. 4a) by either bicuculline, a highly-selective competitive antagonist^ or by picrotoxin, a convulsant blocking agent of the GABA-activated channel'-^® which shows slow binding reversibility. As expected, the blockade by bicuculline was readily reversible upon wash-out (not shown), whereas that of picrotoxin was only slowly reversible (Fig. 4a). A benzodiazepine (chloraze-pate, at 20 |xM) and barbiturates (for example, (-)pentobarbital at 25 |xM) allosterically potentiated the GABA response, the latter more strongly. Benzodiazepines are known to achieve their potentiation by increasing the frequency of channel opening, whereas barbiturates do so by prolonging the channel lifetime'-^^

Quantitative comparison of the responses observed with G A B A and its modulators cannot be made in detail to those on native (neuronal) receptors, as the electrophysiology of bovine neuronal G A B A receptors has not been reported. However, the evidence presented here establishes that the a - and /3-subunits are necessary and sufficient to form the functional receptor for G A B A with its integral chloride channel. The characteristic binding sites of the G A B A A receptor complex, as known in brain membranes or in the isolated protein, are all formed by the transcription of the two cloned cDNAs and the assembly of the subunits. Moreover, allosteric potentiation of the channel opening by barbiturates and by benzodiazepines is present, as is true for the native receptor. Structural model In the light of the similarities of the GABAA receptor and the nAChR, we propose the following model for the structure and function of the GABAA receptor complex (Fig. 5). In both subunits the four hydrophobic domains transverse the mem-brane. These 16 helices must contribute to or stabilize the walls of the channel. Some of these must be placed so as to form the 5.6 A-bore lumen of this channel"-^®. The N-terjnini are assigned to the extracellular side (as is generally found with receptors which contain signal peptides). This topology would place the C-termini just outside the extracellular surface and locates the cAMP-dependent phosphorylation site in the interior, where it must be if functional. We cannot exclude at this stage, however, a variant of this model, in which the M4 helix lies in the membrane surface (as has been established crystallographically for the reaction centre protein in the chloroplast membrane^®), allowing the C-terminus to be intracellular.

The lining of the channel is postulated to be rendered con-ducive to ion flow by the presence of threonine and serine side-chains, largely from the M2 helix (as discussed above), as well as by the two basic and two acidic residues (probably paired) within the assigned transmembrane domains.

A marked clustering of positively-charged side-chains occurs at both ends of the postulated membrane-spanning domains (Fig. 5). These charged residues are presumed to form the chan-nel mouth. In an a2iS2 receptor complex, there would be 26 positive charges and only 4 negative charges on the extracellular side. Most of these clustered basic residues are absent or become of opposite charge in the corresponding nAChR locations. Such clusters of positive charge in the mouth of the channel would act as an anion-selective filter and increase the driving force for chloride flow upon opening of the channel gate. This model is in agreement with the channel reversibility behaviour", channel conductance studies which suggest that at least two anion bind-ing sites are associated with the channeP', and the strong positive influence of halide ion concentration on the binding of some of the specific ligands for this receptor^

On a more speculative note, the gating mechanism may be operated through a dual configuration of the Ml sequence within both subunits. A flexure in Ml is introduced by the unique proline residue noted to be in an invariant motif; Ml could, therefore, be repositioned upon neurotransmitter binding to open the full channel lumen. The binding sites for the various

G A B A and benzodiazepine agonists and antagonists are prob-ably located in the large N-terminal extracellular domains, which contain a total of 10 potential iV-glycosylation sites per receptor complex. Such binding site locations are supported by the finding^^ that deglycosylation of the rat G A B A a receptor modifies the binding of benzodiazepine agonists and antagon-ists. Thus, G A B A binding to the )3-subunits^ of the receptor complex would induce a conformational change, exposing some of the positively-charged residues at the channel mouth and possibly shifting the configuration of M l , resulting in chloride ion flux. Allosteric modulat ion of these conformational changes must be produced by the binding of benzodiazepines or barbitur-ates. The region l inUng M3 and M4, which shows n o homology between the a - and /3-subunits and contains the phosphorylat ion

Received 26 May; accepted 16 June 1987.

1. Olsen, R. W. & Venter, J. C. (eds) Benzodiazepine/GABA Receptors and Chloride Channels: Structural and Functional Properties (Liss, New York, 1986).

2. Turner , A. J. & Whittle, S. R. Biochem. J. 209, 29-41 (1983). 3. Sigel, E., S tephenson, F. A., Mamalaki , C. & Barnard, E. A. J. bioL Chem. 258, 6965-6971

(1983). 4. Sigel, E. & Barnard , E. A. J. bioL Chem. 259, 7219-7223 (1984). 5. Casalohi , S. O., Stephenson, F. A. & Barnard, E. A. J. bid. Chem. 261, 15013-15016 (1986). 6. Mamalaki , C., Stephenson, F. A. & Barnard, E. A. EMBO J. 6, 561-565 (1987). 7. Schoch, P. et al. Nature 314, 168-171 (1985). 8. Huynh, T. V., Young , R. A. & Davis, R. W. in: DNA Cloning: A Practical Approach, Vol. 1

(ed. Glover, D. M.) 49-78 ( IRL, Oxford , 1985). 9. Von Heijne, G., Nucleic Acids Res. 14, 4683-4690 (1986).

10. Kyte, J. & Doolit t le, R. F. J. molec. Biol. 157, 105-132 (1982). 11. Schwartz, R. M. & Dayhoff , M. O. in Atlas of Protein Sequence and Structure, Vol. 5, Suppl .

3 (ed. Dayhoff , M. O.) 353-358 (Nat . Biomed. Res. Fdn., Washington D C , 1978). 12. Sweetnam, P. M. & Tal lman, J. F. Molec. Pharmacol. 29, 299-306 (1986). 13. Feramisco, J. R., Glass, D. B. & Krebs, E. G. J. biol Chem. 255, 4240-4245 (1980). 14. Kubo , T. et al. Eur. J. Biochem. 149, 5 -13 (1985). 15. Popot, J-L. & Changeux, J-P. Physiol. Rev. 64, 1162-1239 (1984). 16. Criado, M., Sarin, V., Fox, J. L. & Lindstrom, J. Biochemistry 25, 2839-2846 (1986). 17. Kao, P. N. & Karl in, A. / biol. Chem. 261, 8085-8088 (1986). 18. Goldman , D. et al. Cell 48, 965-973 (1987). 19. Conti-Tronconi , B. M., Hunkapi l ler , M. W. & Raftery, M. A. Proc. nam. Acad. Sci. U.S.A.

81, 2631-2634 (1984). 20. Hermans-Borgmeyer, I. et al EMBO J. 5, 1503-1508 (1986). 21. Hucho, P., Oberthiir , W. & Lottspeich, F. FEBS Lett. 205, 137-142 (1986).

site, is a potential location for other sites, as yet unknown, mediating intracellular control of channel activity.

The testing of such a model , and the establishment of the mechanisms of channel function and receptor modulation, should be feasible n o w by means of mutational analysis of the cloned a- and )8-subunit c D N A s , coupled with single channel recording in the oocyte expression system.

N.F. holds an E M B O Long-term Fellowship, D.R.B. held a Fogarty Senior International Fellowship and F.A.S. holds a Royal Society University Research Fellowship. We thank Michael Squire (Cambridge) for expert technical help with some of the D N A sequencing, Carole Morita (Genentech) for help with computer graphics and Mary Wynn (Cambridge) for word processing.

22. Giraudat , J., Dennis , M., He idmann , T., Chang , J.-Y. & Chargeux, J.-P. Proc. natn. Acad. Sci U.S.A. 83, 2719-2723 (1986).

23. Imoto, K.. el aL Nature 324, 670-674 (1986). 24. McCarthy, M. P. el al. A Rev. Neurosci. 9, 383-413 (1986). 25. Grenningloh , G. et al. Nature 328, 215-220 (1987). 26. One , J. K. & Salvaterra, P. M. / . Neurosci. 1, 259-270 (1981). 27. Dixon, R. A. F. et al. Nature 321, 75-79 (1986). 28. Kubo , T. et al. Nature 323, 411-416 (1986). 29. Nathans , J . & Hogness, D. S. Cell 34, 807-814 (1983). 30. Sumikawa, K., Houghton , M., Emtage, J. S., Richards, B. M. & Barnard, E. A. Nature 292,

862-864 (1981). 31. Barnard, E. A., Miledi, R. & Sumikawa, K. Proc. R. Soc. Lond. B 215, 241-246 (1982). 32. Smart, T. G., Constant! , A., Bilbe, G., Brown, D. A. & Barnard, E. A. Neurosci. Lett. 40,

55-59 (1983). 33. Houamed , K. M. et al. Nature 310, 318-321 (1984). 34. Melton, D. A. et al. Nucleic Acids Res. 12, 7035-7056 (1984). 35. Dascal , N., Landau , E. M. & Lass, Y. J. Physiol. 352, 551-574 (1984). 36. Study, R. E. & Barker, J. L. Proc. natn. Acad. Sci. U.S.A. 78, 7180-7184 (1981). 37. Bormann, J., Hammil l , O. P. & Sakmann , B. J. Physiol. 385, 243-286 (1987). 38. Unwin, P. N. T. Nature 323, 12-13 (1986). 39. Deisenhofer , J., Epp , O., Miki, P., Huber , R. & Michel, H. Nature 318, 618-624 (1985). 40. Rodriguez, H. J. Chromatog. 350, 217-225 (1985). 41. Messing, J., Crea , R. & Seeburg, P. H. Nucleic Acids Res. 9, 309-321 (1981). 42. Hopp , T. P. & Woods , K. R. Proc. natn. Acad. Sci. U.S.A. 78, 3824-3828 (1981). 43. Staden, R. Nucleic Acidi Res. 10, 2951-2961 (1982). 44. Barnard, E. A. & Bilbe, G. in Neurochemistry: A Practical Approach (eds Turner , A. J. &

Bachelard, H.) 243-270 (IRL, Oxford , 1987). 45. Van Renterghem, C. et al. Molec. Brain Res. 2, 21-31 (1987).

Printed in Great Britain by Turnergraphic Limited. Basingstoke. Hampshire

Volume 15 Number 8 1987 Nucleic Acids Research

Primary structure of the human beta-adrenergic receptor gene

Peter R.Schofleld^ Lucy M.Rhee and Ernest G.Peralta^

Departments of ^Developmental and ^Molecular Biology, Genentech, Inc., 460 Point San Bruno Boulevard, South San Francisco, CA 94080, USA Submitted January 14, 1987 Accession no. Y00106

A human genomic DNA library (1) was screened with two overlapping 60-mer oligonucleotide probes based on the hamster beta-adrenergic receptor (BAR) sequence (nucleotides 547-606 and 583-642, respectively) (2). One clone, >.hBAR17 was identified and a 12 kb EcoRI restriction fragment subcloned (ph6AR3). 2305 bp of the hybridizing region was sequenced (Fig. 1). Analysis of the predicted 413 amino acid sequence revealed seven hydrophobic transmembrane domains, characteristic of guanine nucleotide-protein coupled receptors. The entire coding sequence is encoded on a single exon as is also seen in the hamster gene. While the promoter region of this gene is not obvious, the 200 bp 5' of the initiation codon are 77% conserved with the hamster BAR cDNA sequence indicating that these sequences are most likely transcribed (by comparison, coding sequences are 87.5% conserved). The predicted human and hamster BAR proteins are 88% homologous, while the turkey BAR product (3) is 50% conserved with the human sequence.

1 GAATTCATGCAK:GTTTCTGTGTTGGACAGGGGTGACTTTGTGCCGGATGGCTTCTGTGTGAGAGCGCGCGCGAGTGTGCATGTCGGTC 1 2 1 GCTGTGGTTCGGTATAAGTCTAAGCATGTCTGCCAGGGTGTATTTGTGCCTGTATGTGCGTGCCTCGGTGGGCACTCTC^ 2 4 1 TGCCTTGAGACCTCAAGCCGCGCAGGCGCCCAGGGCAGGCAGGTAGCGGCCACAGAAGAGCCAAAAGCTCCCGGGTTGGCTGGTAAGCACACCACCTCCAGCTTTAG^^ 3 6 1 CCCAGGGTAGCCGGGAAGCAGIWTGGCCCGCCCTCCAGGGAGCAGTTGGGCCCCGCCCGGGCCAGCCTCAGGAGAAGGAGGGCGAGGGGAGGGGAGGGAAAGGGGAGGAGTGCCTCGCC 4 8 1 CCTTCGCGGCTCCCGGCGTGCCJ^TTEGCCGAAAGTTCCCGTACGTCACGGCGAGGGCAGTTCCCCTAAAGTCCTGTGCACATAACGGGCAGAAC^ 6 0 1 ACGGGCTGGAACTGGCAGGCACCGCGAGCCCCTAGCACCCGACAAGCTGAGTGTGCAGGACGAGTCCCCACCACACCCACACCACAGCCGCTGAATGAGGCTTCCAGGCGTCCGCTCGCG

1 M G Q P G N G S A F L L A P N 7 2 1 GCCCGCAGAGCCCCGCCGTGGGTCCGCCTGCTGAGGCGCCCCCAGCCAGTGCGCTTACCTGCCAGACTGCGCGCCATGGGGCAACCCGGGAACGGCAGCGCCTTCTTGCTGGCACCC^

1 6 R S H A P D H D V T Q Q R D E V W V V G M G I V M S L 1 V L A I V F G N V L V I 8 4 1 AGAAGCCATGCGCCGGACCACGACGTCACGCAGCAAAGGGACGAGGTGTGGGTGGTGGGCATGGGCATCGTCATGTCTCTCATCGTCCTGGCCATC^

S E T A I A K F E R L Q T V T N Y F I T S L A C A D L V M G L A V V P F G A A H I L 9 6 1 ACAGCCATTGCCAAGTTCGAGCGTCTGCAGACGGTCACCAACTACTTCATCACTTCACTGGCCTGTGCTGATCTGGTCATGGGCCTC

9 6 M K M W T P e N F W C E F H T S I D V L C V T A S I E T L C V I A V D R Y F A I

1 0 8 1 ATGAAAATGTGGACTTTTGGCAACTTCTGGTGCGAGTTTTGGACTTCCATTGATGTGCTGTGCGTCACGGCCAGCATTCAGACCCTGTGCGTC 1 3 6 T S P F K Y Q S L I , T K N K A R V I I L M V H I V S G L T S F L P I Q M H W Y R

1 2 0 1 ACTTCACCTTTCAAGTACCAGAGCCTGCTGACCAAGAATAAGGCCCGGGTGATCATTCTGATGGTGTGGATTGTGTCAGGCCTTACCTCCTTC 1 7 6 A T H Q E A I N C Y A N E T C C D F F T N Q A Y A I A S S I V S F Y V P L V I K

1 3 2 1 GCCACCCACCAGGAAGCCATCAACTGCTATGCCAATGAGACCTGCTGTGACTTCTTCACGAACCAAGCCTATGCCATTGCCTCTTCCATCGTGTCCTTC 2 1 6 V F V Y S R V F Q E A K R Q L Q K I D K S E G R F H V Q N L S Q V E Q D G R T G

1 4 4 1 GTCTTCGTCTACTCCAGGGTCTTTCAGGAGGCCAAAAGGCAGCTCCAGAAGATTGACAAATCTGAGGGCCGCTTCCATGTCCAGAACCTTAGCCAGGTGGAGCAGGATGGGCGGACGGG 256 H G L R R S S K F C L K E H K A L K T L G I I M G T F T L C U L P F F I V N I V

1561 CATGGACTCCGCAGATCrrCCAAGlTt;TGCTTGAAGGAGCACAAAGCCCTCAAGACGTTAGGCATCATCATGGGCACTTTCACCCRC 2 9 6 H V I Q D N L I R K E V Y I L L N W I G Y V N S G F N P L I Y C R S P D F R I A

1 6 8 1 C A T G T G A T C C A G G A T A A C C T C A T C C G T A A G G A A G T T T A C A T C C T C C T A A A T T G G A T A G G C T A T G T C A A T T C T G G T T T C A A T C C C C T T A T C T A C T ^ 3 3 6 F Q E L L C L R R S S L K A Y G N G Y S S N G N T G E Q S G Y H V E Q E K E N K

1 8 0 1 TTCCAGGAGCTTCTGTGCCTGCGCAGGTCTTCTTTGAAGGCCTATCGGAATGGCTACTCCAGCAACGGCAACACAGGGGAGCAGAG^ 3 7 6 L L C E D L P G T E D F V G H Q G T V P S D N I D S Q G R N C S T N D S L L

1 9 2 1 CTGCTGTGTGAAGACCTCCCAGGCACGGAAGACTTTCTGGGCCATCAAGGTACTGTGCCTAGCGATAACATTGATTCACAAGGGAGGAATTC 2 0 4 1 AGTTTTTCTACTTTTAAAGACCCCCCCCCCCAACAGAACACTAAACAGACTATTTAACTTGAGGGTAATAAACTTAGAATAAAATTCTAAAATTCTA 2 1 6 1 A T C C T T C T G C C T T T T T T A T T I T T T T A A G C T G T A A A A A G A G A G A A A A C T T A T T T G A G T G A T T A T T T G T T A T T T G T A C A G 2 2 8 1 GAGCTTTAGTCCTAGAGGACCTGAGTC I

ACKNOWLEDGEMENT We thank Drs. D.J. Capon and P.H. Seeburg for support and encouragement and the

Genentech DNA synthesis group for oligonucleotides.

REFERENCES 1. Lawn, R.M., era/. (1978) Ce//15, 1157-1174. 2. Dixon, R.A.D., er fl/. (1986) Nature 221,15-19. 3. Yarden.Y., era/. (1986) Proc. Narl. Acad. Sci. USA 83, 6195-6199.

3636 © IRL Press Limited, Oxford, England.

Vol. 144, No. 1, 1987 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

April 14, 1987 Pages 59-66

MOLECULAR CLONING AND AMINO ACID SEQUENCE OF RAT ENKEPHALINASE

Bernard MalfroyPeter R. Schofield^, Wun-Jing Kuang^, Peter H. Seeburg^ Anthony J. Mason^, and William J. Henzel^

1 1 3 Departments of Pharmacological Sciences , Developmental Biology , and Molecular Biology Genentech, Inc.,

460 Point San Bruno Boulevard, South San Francisco, CA 94080

R e c e i v e d F e b r u a r y 2 3 , 1987

cDNA clones enccxiing rat enkephalinase (neutral endopeptidase, EC 3.4.24.11) have been isolated in XgtlO libraries from both brain and kidney mRNAs and the complete 742 amino acid sequence of rat enkephalinase is presented. The enzyme possesses a single transmembrane spanning domain near the N-terminal of the molecule but lacks a signal sequence. Because enkephalinase has its active site located extracellularly and is thus an ectopeptidase, we suggest that the N-terminal transmembrane region of the enzyme anchors the protein in membranes and that the majority of the protein, including the carboxy terminus, is extracellular. Enkephalinase, a zinc-containing metallo enzyme, displays homology with other zinc metallo enzymes such as carboxypeptidase A, B and E, suggesting enzymatic similarities in these enzymes. ® 1987 Academic Press, inc.

Enkephalinase (neutral endopeptidase; EC 3.4.24.11) is a cell membrane-associated enzyme which hydrolyses the Gly^-Phe^ amide bond of enkephalins in vitro and in vivo (1-4). Thus, pharmacologically useful inhibitors of enkephalinase such as thiorphan (5) and phosphoryl-Leu-Phe (6) have been shown to produce naloxone-reversible analgesia in mice. In addition, acetorphan, a parenterally active enkephalinase inhibitor (7), displays analgesic properties in humans (8). Enkephalinase preferentially hydrolyses peptide bonds on the amino side of hydrophobic residues, and its substrate specificity is sufficiently broad that it is able to hydrolyse, in vitro, many short neuropeptides and peptide hormones (9-11). Although initially characterized in the brain (1), enkephalinase was later found to be present in many tissues, in particular the kidney (12), where it was shown to be identical to an enzyme identified several years before using the B chain of insulin as subsu-ate (13), the so called "neutral endopeptidase" (14-17). We have isolated cDNAs encoding enkephalinase from both rat brain and kidney mRNAs and present the complete 742 amino acid sequence of rat enkephalinase.

MATF.RIAl.?> AND METHQPS

Purification and sequence analysis of rat kidney enkephalinase: Rat enkephalinase was purified as described (17) by differential solubilization of kidney membranes with Triton X-lOO, followed by DEAE Sephadex chromatography, Concanavalin A-Sepharose chromatography, and hydroxylapatite chromatography. Fractions eluted from the hydroxylapatite column and containing enkephalinase activity were loaded onto a 10x10 mm Concanavalin A-Sepharose column, and the enzyme was eluted in several 1ml fractions with 5 mM, pH 7.4 HEPES buffer containing 0.1% Triton X-100 and 500 mM methyl-a-D-Glucopyrannoside. Eight 1ml fractions, containing a total of 450 ^ig of protein as determined by the Coomassie blue method using bovine serum albumin as standard, were obtained- One of these fractions containing an estimated 100 jig of protein was concentrated to 200 |il using a Centricon 30 device, and loaded onto a Superose 6 (Pharmacia) column equilibrated in 5 mM, pH 7.4 phosphate/Na

0006-291X/87 $1.50 Copyright © 1987 by Academic Press, Inc.

All rights of reproduction in any form reserved.

Vol. 144, No. 1, 1 987 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICAT IONS

buffer containing 150 mM NaCl and 0.1% Triton X-100. The column was eluted in the same buffer at a flow rate of 200 |il/min and the fractions obtained (300 pJ) were assayed for enkephalinase activity (18). Aliquots of fractions displaying enkephalinase activity were subjected to SDS-polyacrylamide (7.5%) electrophoresis and the gels, stained with coomassie blue, revealed a unique band of molecular weight ca 90 kD. These fractions were used for N-terminal amino acid analysis and for Lysine-C proteinase digest, of enkephalinase. N-terminal sequence was obtained by applying the purified enzyme to a vapour phase protein sequencer (Model 470A, Applied Biosystem) (19), equipped with an on-line amino acid PTH analyzer (Model 120A, Applied Biosystem). It was never possible to obtain reliable protein sequence after 15 cycles. For Lysine-C proteinase digest, the purified enkephalinase, as eluted from the Superose 6 column was digested overnight with the proteinase (10 ng per |ig of enkephalinase) and the peptides released were separated by HPLC on a Synchrom 2x100 mm, 300 A C4 column, eluted with a linear gradient of 0 to 70% 1-propanol (1% per min) in 0.1% trifluoroacetic acid at a flow rate of 400 }il per minute. The peptides displaying a large 280 nm to 214 nm absorbance ratio were rechromatographed on the same column, using acetonitrile instead of 1-propanoI, and their amino acid sequence was determined using the vapour phase protein sequencer. Molecular cloning: RNA was prepared from rat kidney or brain (20) and polyadenylated mRNA was obtained by oligo(dT) cellulose chromatography. High complexity cDNA libraries were constructed in XgtlO (21)as described elsewhere (22). Three libraries were generated: random 8-mers were used to prime rat kidney poly(A)'^ mRNA, generating a library of « 6X106 clones of size > 500 bp; oligo(dT) primed libraries of both rat kidney and brain poly(A)+ mRNA, « 5x10^ and 6x10^ clones of size > 1500 bp, respectively, were also generated. Oligonucleotide probes were designed based on the amino acid sequence of peptide KC2-18-4. A pool of 32x18-mers, 5'-YTGYTGNGTCCACCARTC-3' (Y = T, C; R = G, A; N = G, A, T, C) (short probe), covering all non-coding strand possibilities of the protein sequence DWWTQQ was used with the base composition independant method (23) and a single long probe (45-mer) 5'-GAAGTTGTTGGCGGACTGCTGGGTCCACCAGTCGACCAGGTCGCC-3', complementary to the entire sequence of peptide KC2-18-4 except for its C-terminal lysine residue was also used. Both the short and long probes were used to screen 5x10^ clones of the randomly-primed rat kidney cDNA library. Hybridization was carried out with ^^p.phosphorylated probes in 5xSSC, 20% formamide at room temperature. Filters were washed at room temperature in 0.5xSSC, 0.1% SDS for the long probe and in 3M tetramethylammonium chloride at 51 ° C for the short probe. Inserts of X. phage were subcloned into an Ml3 derivative and sequenced by the chain-termination method (24,25). DNA- and RNA-blot analysis: Sprague-Dawley rat DNA, isolated by standard procedures, was digested with either PstI, EcoRI or Bglll and electrophoresed on a 1% agarose gel. After Southern u-ansfer (26) onto nitrocellulose the blot was hybridized with a ^^p.iabelled (27) 466 bp EcoRI-Bglll cDNA fragment of clone XK3, in 5xSSC, 50% formamide at 42 °C. The blot was washed at 65 °C in O.lxSSC, 0.1% SDS, and exposed for 60 h. For Northern blots, polyadenylated mRNA (5 |ig) was electrophoresed in a formaldehyde-1.2% agarose gel (28) and blotted onto nitrocellulose paper (29). The blot was hybridized with the 1459 bp EcoRI insert of :\.K3 in 5xSSC, 50% formamide at 42 and washed in O.lxSSC, 0.1% SDS at 65 ®C.

ANP PISCUS?>ION Rat kidney enkephalinase was purified to homogeneity and subjected to amino acid sequence analysis either directly or after Lysine-C-proteinase digestion, followed by purification of peptide fragments generated by HPLC (Fig. 1). One such peptide, KC2-18-4, containing two tryptophan residues, was used to design oligonucleotide probes for screening cDNA l ibr^es . Both a complementary short pool (32x18 mers) containing all possible nucleotide assignments, for screening by the base composition independant method (23), and a 45-mer long probe (30) were synthetized and used to screen, in duplicate, a randomly primed rat kidney cDNA library constructed in ?Lgil0. Two positive clones were identified and the largest, XK3, was sequenced. This clone contains an open reading frame that encodes all peptide fragments sequenced, except the N-terminus. This cDNA done was in turn used as a probe to screen oligo(dT) primed XgtlO cDNA libraries constructed from either rat brain or kidney poly (A)+ mRNA. Two clones, XK2 and >.K5, were identified from the kidney library, and two clones, XBIO and XB16, were identified from the brain library. Clones XK2, XK5 and ABIO contain a stop codon for the reading frame identified in clone X.K3, while clone XB16 encodes the N-terminal protein sequence. Thus a complete reading frame coding for the enkephalinase molecule was obtained from overlapping clones, from brain and kidney (Fig. 2 a, b). Brain and kidney cDNA clones overlap by a total of 2008 base pairs (bp) of the 3243 bp sequence. Two nucleotide sequence differences were detected. At nucleotide 1917, a T occurs in kidney clone XK2 and in brain clone XBIO while

60


9 2 . 5 -6 6 . 2 -

4 5 -

T 20 Minutes

P e p t i d e s N- t erra ina l KC8 KC2-12-6 K C 2 - 1 8 - 4 K C 2 - 1 9 K C 2 - 3 1 K C 2 - 3 2

Amino Acid S e q u e n c e A s p - I l e - T h r - A s p - I l e - A s n - A l a - P r o - L y s - P r o - L y s - L y s - L y s - G l n - A r g L e u - L e u - P r o - G l y - L e u - A s p - L e u - A s n - H i s - L y s G l u - A r g - I l e - G l y - T y r - P r o - A s p - A s p - I l e - I l e - S e r - A s n - G l u - A s n - L y s X X X - G l y - A s p - L e u - V a L - A s p - T r p - T r p - T h r - G l n - G l n - S e r - A l a - A s n - A s n - P h e - L y s G l u - G l u - G l u - T y r - P h e - G l u - A s n - I I e - I l e - G l n - A s n - L e u - L y s H i s - G l n - A s n - X X X - P h e - S e r - X X X - G l u - I l e - A s n - G l y - L y s A l a - V a l - V a l - G l u - A s p - L e u - I l e - A L a - G l n - I l e - A r g - G l u - V a l - P h e - I I e - G l n - T h r - L e u

XXX: no r e s i d u e i d e n t i f i e d . Figure 1. Rat kidnev enkenhalinase purification and protein sequence analysis. (A) Coomassie blue-stained SDS-poIyacrylamide gel of purified enkephalinase. The positions of molecular weight standards are indicated. (B) Lysine-C proteinase map of purified enkephalinase. Upper trace is absorbance at 280 nm and lower trace absorbance at 214 nm. Arrows indicate peaks that were rechromatographed and sequenced. (C) Amino acid sequence of native enkephalinase (N-tenninal) and of Lysine-C proteinase peptides.

a C is found in a second kidney clone, X.K3; at nucleotide 2930, a G or an A is found in either the kidney (XK2) or brain (XBIO) clones, respectively. Southern blot analysis (26) of rat genomic DNA using a small coding region cDNA probe was undertaken. The multiplicity of hybridyzing bands in only some restriction enzyme digests (Fig. 3a) suggests the presence of several intron sequences but not multiple genes. In the 3'-untranslated region of the cDNA clones there is over 650 bp of overlapping kidney and brain sequence with total homology, except for the single nucleotide change at position 2930. Since the 3'-untranslated region is typically least strongly conserved between related genes, this, and the Southern blot data, suggest that a single gene encodes both brain and kidney enkephalinase. The nucleotide differences are thus probably errors of reverse transcriptase. This identity of kidney and brain enkephalinase cDNA's fully supports the earlier contentions that the kidney and brain enzymes are identical (15,16,31).

61


A

0 500 1000 1500 2000 2500 3000 3500 1 [ 1 1 1 1 1 1 S ' l - H 1 1 3 '

1 2167 XB^e

395 1269 XK3 I 1

1090 2457 X K 5 t 1

1791 2984 XK2 I 1

1850 3243 > . 8 1 0 I i

^ 1 0 ' , ' ^ e t G l j / A r - a S e r G l u S e r G l n M e t A s p ! l e T h r A s p I l e A s n A l j P r o L y s P r o L v s L v s L y s G l n A r o ' T r D

1 GCGGAGATmGCAAGTGGCGAAGCTGGACCGMGTGCAGGCGCAAGCTGCTGAGCGGCTGAGGCGAGGGATTTTAGGrGATGGGAAGATCAGAAAGTCAGATPGATATTACTGATATCAATGCTCCAAAGCCGAAGAAGAAACAGCGATGG

I 3 0 I 5 0 6 0

T h r P r o L e u G l u n e S e r L e u S e r V a l L e u V a l L e u L e u L e u T h r n e l l e A l a V a l T h r H e t i l e A l d L e u T y r A l a T h r T / r A s p A s p G l y l l e C y s L y s S e r S e r A s p C y s I l e L y s S e r A l a A l a A r g L e u I l e G l n A s n M ^ 15 t A L T C C A C T G G A G A T C A G C C T T T C T G T G C T C G T C T T G t T C C I G A C T A T C A T A G C T G T G A C A A T G A T T G C T C T C T A T G C A A C r T A T G A T G A T G G T A T T T G C A f i A T C A T C A G A C T G C A T A A A A T C A G C T G C T C G A C T G A T C C A G A A C A T G G A T

8 0 9 0 1 0 0 1 1 0

A l a S e r A l a G l u P r o C y s T h r A s p P h e P h e L y s T y r A l a C y s G l y G l y T r p L e i i L y s A r q A s n V a l l l p P r o G h i T h r S e r S e r A r a T y r S e r A s n P h e A s p I l e L e u A r q A s p G l u L e u G l u V a l I l e L e u L y s A s p V a l L e u G l n G l u 3 U 1 G C C T C T G C T G A G C C A T G T A C G G A C T T C T T C A A A T A T G C T T G T G G A G G C T G G T T G A A A r n r A A T G T C A T r C C T G A G A C C A G T T f X C G A T A C A G T A A T T T T G A C A T T C T A A G A G A T G A A C T A G A A G T C A T T

1 3 0 K / X / / 1 1 1 0 1 5 0 1 6 0

P r o L y s T h r G l u A s p l l e V a l A l a V a l G l n L y s A l a L y s T h r L e u T y r A r g S e r C y s n e A s n G l t i S e r A l a l l e A s p S e r A r q G l v G l y G l n P r o L e u L e u T h r L e u L e u P r o A s p n e T v r G l y T r p P r o V a l A l a S e r G l n A s n T r p 4 5 1 C C U A A A C T G A G G A C A T A G T A G C A G T G C A G A A A G C A A A A A C T T T G I A C A G A T C A T G T A T A A A T G A A T C T G C T A T T G A T A G C A G A G G T G G G C A A C C T C T G C T C A C A C T G T T A C C A G A T A T A T A ^

lau 190 ?00 \ / / / / A 210 O l u G l n T h r T y r G I y T h r S e r T r p T h r A l a G l u L y s S e r l l e A l a G l n L e u A s n S e r L y s T y r G l y L y s L y s V a l L e i i l l e A s n P h e P h e V a l G l y T h r A s p A s p L y s A s n S e r T h r G l n H i s I l e n e H i s P h e A s p G l n P r o A r q L e u

6 0 1 G A A C A A A C A T A T G G T A C T T C T T G G A C A G C T G A G A A A T C T A T T G C A C A A C T G A A T T C T A A A T A T G G G A A A A A G G T C C T C A T T A A T T T T T T T G T T G G C A C T G A T G A T A A G A A T T C T A C C C A G C A T A T A A T T C A T T T T G A C C A G C C T a ^ ^

2 3 0 2 4 0 2 5 0 2 6 0

G t y L e u P r o S e r A r g A s p T y r T y r G l u C y s T h r G l y l l e T y r L y s G l u A l a C y s T h r A l a T y r V a l A s p P h e M e t n e S e r V a l A l a A r Q L e u I l e A r q G l n G l u G l n A r a L e u P r o I l e A s p G l u A s n G l n L e u S e r L e u G l i i M e t ^ ^ 7 5 1 G O L C I C L C T T C C A G A & A C T A C T A T G A G T G T A f A G G A A T A T A T A A A G A G G C T T G C A C A G C A T A T G T G G A T T T r A T G A T T T r T G T G G C C A G A C T G A T T C G f C A G G A A C A A A G A T T G C r T A T T G A T G A A A A C C A G C T C T C T T T G G A A A T G A A T

X / y / ' / X i m ? 9 0 ,300 \ / / / / A 3 1 0

L y s V a l M e t G l u L e u G l u L y s G l u I l c A l a A s n A l a T h r T t i r L y s P r o G l u A s p A r q A s n A s p P r o H e t L c i i L e u T y r A s n L y s M e t T h r L e u A l a L y s L p i i G l n A s n A ^ v < P r o P h e S e r T r p S e r 9 0 1 AAAG ITATGGAATTGGAAAAAGAAATTGLXAATGCCACAACTAAALCAGAAGACCGAAATGACCCAATGCTGCTTTATAACAAAATGACATTGGCCAAGCTCCAAAATAACTTCTCTCTGGAGATCAATGGGAA^^^

EZZZZ23 ?0 330 , 340 , 3S0 360 A s n P h e T h r A s n G l u I l e M e l S e r T h r V a l r t s n n e A s t i l l e G l n A s n ' t n j t r i u G l ^ ^

l O b l A A T T r t A C A A A T G A A A T C A T G T C A A C T G T G A A T A T T A A T A T r C A A A A T G A C G A A G A A G T G G T T G T T T A T G C T L X A G A A T A T T T A A r X A A A C T T A A O f C T A T T C T T A r C A A A T A T T C T C f r A G A C ^ ^ ^

3 / 0 38U 3 9 0 aOO 4 1 0

' ' ' ' • ' ^ ' • " S p L c i j V a l b e r S e r L e u S e r A r q A s r J y r L y s G h i S e r A r q A s n A l a P h e A r q L y s A l a L e u T y r G l y T h r T M r S e r G l u T h r A l a T h r T r p A r q A r q 1^:01 T r C A T A A T G G A r C T T G T A A G C A G C C T C A G C C G A A A C T A C A A G G A G I C C A G A A A T G C T T T C C G C A A G G C C C T T T A C G G G A C T A C A T C C G A A A C T G f A A C C T G G A G A C G G T G T G C C A A C T A C G T C A A r G G G A A C A T G G A G A A T ^ ^

j l 3U 4 4 0 « 5 0 4 6 0

A r g L e L T y r V a l G l u A l d A l a P h e A l a G l y G l u S e r L y s ' H i r v n V ^ l G T ^ ^

1 3 5 1 A G G t T T T A T G T G G A A G C A G L T T T T G t T G G A G A G A G C A A G r A C G T G G T T G A A G A r T T G A T T G G A C A A A T C C G t G A A G T T T T T A T T C A G A C T T T A G A T G A C C T C A C T T G G A T G G A T G r T G A G A C A A A A A A G A A A G C T G A A

|iZk_ 4UU ^ /)9U 500 510 A l a l l e L y b G h i A r y l l c G l y T y r P r o A s p A s p l l e i l e S c r A s n G l u A s n L y s ' L e u A ^ i n A s n G l u T y r L e u G U i L e u A s n T y r L y s G l u G l u G l u T y r P h e G l u A s n I l e l l e G l n A s n L e i i L v s P h e S e r G l n S e r L v s G l n L e u L y S

1 5 0 1 GCAATTAAAGAAAGGArTGGCTATCXTGATGACATCATCTCCAATGAGAATAAACTGAATAATGAGTATCTTGAGTTGAACTACAAGGAAGAGGAATACTTTGAGAACATAATTCAAAATTTGAAATTCAGCCAAAGCAAGCAGCTAAAG

5311 54U ^>50 5 6 C

' • ^ ' ' : ? " ? " J ^ ' " ' - J ^ ^ ^ ' ' * ^ P ' - y 5 ' ' s p G l u T r p n e S e r G l y A 1 a A l a V a l V a l A 5 n A l a P h e T y r S e r S e r G l y A r q A s n G l n n e V a l P h e P r o A l a G t y n e L e u 1 6 5 1 ' ^ ^ ( ^ C l L L G A O A A A A G G T G G A C A A A G A l G A G T G G A T A A G r U G C G t G G G G G T A G r C A A T G C A T T T T A T T f X r W G G C A G A A / J C A G A T C G T t - T T C C C r G r C G G C A T T T T G C A G C r r C C A

5 ^ 0 bJ!0 5 9 0 6 0 0 6 1 0

T y r G l y G l y l l e G l y N t I V a U l e G l y H i s G l u 11 e T h r l l i s G l y P h e A - i p A s p A s n G l y A r q A s n P h e A s n L y s W c W A s p L ^ ^

W O l T A T G G G G G C A T C G G C A T G G T C A r C O G A C A I G A A A l C A t A C A T G G C T T T G A T G A C A A T G G C A G A A A T T T T A A C A A A G A T G G A G A C C T C G T T G A C T G G T G G A C T C A G C A G T C T G C A A A T A A T T T C A A A G A C C A A T C C C A G T G T A T G G T ^

C

n ^ N W N 6 3 0 6 4 U e-^O 6 6 0

1 9 5 1 L A G T M T G G A M L T T T A L A T G G G A L C T A G C A G G T G G A C A G C A T C T C A A T G G A A T T A A C A C A C T A G G A G A A A A T A T T G C T G A T A A T G G A G G G A T T G G C C A A G C A T A C A G A G C C T A T C A G A A T T A J G T T A A A A A G A A T G

„ ^ 1 WU I 6 9 0 7 0 0 7 1 0

^ l U l ^ I L L C l G G A L T T G A C L T L A / a w C A A A C A A C T A T T C I T C T T G A A C T T T G C C C A G G T G T G G T G T G G A A C C T A C C G G C C A G A G T A T G C A G T C A A T T C C A T T A A A A C A G A T G T A C A C A G T C C T G G W A T T T ^

' ^ U 7 3 0 7 4 0

A s i i S e r A l d G l u P h e A l a A s p A l a P l i e H i s C y s A r g L y s A s n S e r T y r M e t A s n P r o G l u A r q L y s C y s A r g V a l T r p O P *

2 2 5 1 ' w t r c I G L T G A G T T T G C G G A T G C C T T T C A T T G C C G C A A G A A C T C A T A C A T G A A T C C A G A A A G G A A A T G T C G G G T T T G G T G A T C T T C A C A G G A A G T G G A G C A T C C A T G G C A G G A C T C G C C A A A G C C A C A G A A A C A G G A A G T C T T C C C T C A G

2 4 0 1 AGAA IGTGGGCCC:GGGAAGTTTCTTCAGCTTLTTGGGGGAAATTCAC»GA( ;ATGAGCACGA( ; I :TAACAAAAATGAAATTAGATTATTAAAACCGCTGTGAATGAAAGGGGAGAAAACCTACGATCTAGCAAATCAATCACTTCACTGTGT

2 5 5 1 ' W A T A A T T A C C T T C C A A t G G T A A T A T T A C t G T T a i L T T t l G G T T C T C A C A G A G A C r G C A G C T T T C A T G C T G T C T G T A G A G A A C A G T G T T A A C A C T T A A A G C A G G T T A T G A C T T C T G A T C A A G A G G A G G A A G A C G C T

2 7 U 1 ' ^ C M A A G T A C A G A T T T G t C T C T C M G t A C T C A t T T T T G T T T G C A A C A T T C A G C T C C T T C A A A A T T C T C C C A A A G A A C r C C C A T G C A T A C T G T G G C C T T r A G G C T C r T G r A G T G T G G A A

25151 C L A C A T C A T T T T A G T T T G A G C A C T C T T A G A G C T T A A A C T A G A G A G T C T G A A A T G G T I C C G C C A T T T A C C C A C T T G A G T G G T G T T G A G A C T C T T C A G C C C C C T A C A G A T T T T T G A G C A A T T T C T T G C T C T C G C T G C C C C T C A G A C T T ^ ^ ^

A

3001 T T T T A A A G G A T T T G T A G T A A T G T A T A A A A A A L A T T C T A T A T T T A A T R A T T A A C T A C A C A T G A C C A A A T A A A L : : A T T G C T A T A G G T A A T C A T T G A A T A T T G A C A T T A T A T G G C C A A G A T A G A T A G T T A A G A A G A T C T G T A A C A T G A T G T G C

3 1 5 1 A G A I b A A A A n r G A A A C T T T T T A A G C t T G T A A A T G A T A T T G L l G A A A A T C T T G A A A C A C A A A C T C T G G G G T G A G C A T T A C C A T T G A A C A G T T G

62

Structure of the Precursor to an


Northern blot analysis of poly(A)+ mRNA from rat kidney and brain reveals a complex pattern of hybridization (Fig.

3b). In kidney, the major message is 3.4 kilobases (kb) in size, with other messages of 6.5, 6.0, 3.2 and 2.6 kb in

size. The brain messages are at least 10-foId less abundant and only two mRNA's, 6.5 and 3.4 kb in size, are detected.

This difference in intensity is in agreement with the much higher specific activity of enkephalinase in kidney as

compared to brain (12,16). Potendal polyadenylation signals, ATTAAA and AATAAA at nucleotides 2485 and

3063, respectively (Fig. 2b), may correspond to the 2.6 and 3.2 kb kidney mRNA's.

The enkephalinase cDNA potentially encodes a 750 amino acid polypeptide, starting from Mer^ as numbered in Fig.

2. This reading frame has 78 bp of 5'-untranslated sequence and 912 bp of 3'-untranslated sequence. An in-frame stop

codon is located 6 bp upstream of the AUG codon for Mer^ . A second, alternate initiation codon ( M e f l ) is found

downstream of the first ATG (Met"^). Neither of these two potential initiation codons, at Met"^ or Mer^ , conform

closely to the Kozak rule for predicting the site of protein initiation (32). Whereas most eukaryotic genes are

translated using the first AUG encountered in the mRNA, a number of examples where a second AUG is used have

been noted (32). N-terminal protein sequence data obtained using purified enkephalinase suggests that the second

initiation codon (Mef^) is used for the in vivo synthesis of mature enkephalinase. The polypeptide starting with

Aspl at its amino terminus would be 742 amino acids in length, yielding a molecular weight of 85 kD. Molecular

weight estimates of enkephalinase, obtained through SDS polyacrylamide gel electrophoresis, have yielded values in

the 90-94 kD range (11). Since enkephalinase is known to be a glycoprotein, the difference between these estimates

and the value deduced from the protein sequence presumably reflects N-Iinked glycosylation. There are six potential

N-linked glycosylation sites (Asn-X-Thr/Ser) in the rat enkephalinase sequence (Fig. 2b). The failure to detect an

asparagine residue in peptide KC2-31 (Fig. Ic) suggests that this residue, which forms a potential glycosylation site,

is indeed glycosylated.

Enkephalinase has been shown to be an ectopeptidase, with its active site in the extracellular space (11,33). The

hydropathic profile of the enzyme (Fig. 2c) shows the presence of a 23 amino acid long hydrophobic domain (residues

21-43) that is likely to span the lipid bilayer. Additionally, a very strong stop transfer sequence (34), PKPKKKQR,

- . 1 1 1 •• • I 1 " 1 1

A A /\A n A '

1 1 1 1 1 1 1 ~

3.0 JZ 2.0 ra 1.0 Q. O 0.0 "O -1.0 X -2.0

-3.0 100 200 300 400 500 600 700

Figure 2. Nucleotide and deduced protein sequence for rat enkephalinase cDNA. (A) Schematic representation of enkephalinase mRNA. Untranslated sequences are represented by a line; coding sequences are boxed. The scale is in nucleotides from the 5' end of the longest cDNA clone. Overlapping cDNA clones used in sequence determination are shown below the diagram of the mRNA structure. (B) Nucleotide and predicted amino acid sequence of rat enkephalinase. Nucleotides are numbered at the left, and amino acids are numbered throughout The amino acid sequences determined by protein sequencing (see Fig. 1) are overscored. The amino acid sequence at position 592-608 was used for designing oligonucleotide probes. The protein sequence is numbered from Asp^ as this is the N-terminal residue of the mature protein. Two potential initiation codons are Mef l and Met"^ and are underscored, and an inframe 5' stop codon (TAG) is indicated by asterisks. The 8 amino acid stop-transfer sequence (PKPKKKQR) is indicated by a black bar and the putative 23 amino acid signal sequence/transmembrane spanning domain is indicated by an open bar. Six potential N-Iinked glycosylation sites are shown by cross-hatched bars. Potential poly(A) addition signals ATTAAA and AATAAA are underlined. (C) Hydropathy analysis of enkephalinase protein sequence. The method of Kyle and Doolittle (47) was used with a window length of 10 residues and a jump of 2 residues. Hydrophobic regions (positive values) demonsu-ate the presence of a single transmembrane spanning domain between residues 21 to 43.

63


23.5 —

9.5 — 6.7 —

4.4 —

2.3 — 2 . 0 —

1.4 —

Figure 3. Southern and Northern blot analysis of rat DNA and RNA. (A) Southern blot hybridization analysis of rat DNA, digested with PsU (lane a), EcoRI (lane b) or Bglll (lane c) and hybridized with a 466 bp cDNA probe. Size standards (kb) are Hindlll-cleaved XcI857 DNA. (B) Northern blot analysis of mRNA encoding rat enkephalinase. Rat kidney poly(A)+ mRNA (lane a) and rat brain (minus cerebellum) poly(A)+ mRNA (lane b) were hybridized with a 32p-iabelled 1459 bp cDNA probe (see Fig. 2) and exposed for either 4 or 80 h at -70 ® C. Size standards (kb) are from a RNA ladder (BRL).

is located at residues 8-15 on the N-terminal side of the transmembrane spanning region. This stop transfer sequence probably accounts for the fact that enkephalinase is very tighdy bound to membranes (13,17). All these features suggest that the enkephalinase N-terminus is located in the cytoplasm whereas the majority of the protein (704 amino acids, which would contain the active site) including the C-terminus, extends into the extracellular space. The single transmembrane domain would therefore be required to act as a signal sequence as well as a transmembrane spanning sequence. Several membrane-bound proteins possess those features proposed for enkephalinase, namely an intracellular N-terminal sequence and an extracellular C-terminal sequence, a single hydrophobic domain acting as both a signal sequence and transmembrane-spanning sequence, and a stop-transfer sequence located N-terminally of the transmembrane-spanning sequence. These include transferrin receptor (35), asialoglycoprotein receptor (36), HLA-DR-associated invariant chain (37), and influenza virus neuraminidase (38), although the latter possesses an uncleaved signal sequence. Enkephalinase is a zinc-containing metalloenzyme (13,39,40). Because it hydrolyzes peptide bonds at the amino side of hydrophobic residues in its substrates, it has been suggested that enkephalinase resembles two other zinc-containing peptidases, carboxypeptidase A and thermolysin (5,9-11,13,14,40-43). We compared the sequence of enkephalinase with that of bovine carboxypeptidases A (44), B (45) and E (46). The carboxy terminal region of enkephalinase (from residues 615 to 738) can be aligned with the N-terminal domain of all three carboxypeptidases (Fig. 4). The best alignment is with carboxypeptidase A (22% homology, plus 14% conservative changes). A weaker homology was also detected with the N-terminal region of thermolysin (12% homology, from residues 634 to 742 of enkephalinase; not shown). These homologies suggest that the C-terminal domain of the enkephalinase molecule may be functionally related to the N-terminal domains of the carboxypeptidases. The availability of cloncd enkephalinase will facilitate an understanding of the substrate specificity and function of this enzyme.

64


Enkephalinase 610 KcySQCMVyg Carboxypeptidase A 1 ARSTNTFN Carboxypeptidase B 1 TTGHS Carboxypeptidase E 1 RPgEDGISF

Y Y YE

EYH

GNFTWDLAGGQHGNG^NTLGEN IADNGGIIG ATYHTLDEIYDFMDLLVAEHPOLVSKLQIG %YNNWETIEAWTEQVASENPDLISRSMG IRYPELRE AGVSVWLQCAAVSRIY^G

YVK

LL KN

YVLKFSTG-G VGKP-G

R S F EG R E L L0L E LS DN P|G

EE|§LL-SNR SN® VHEPGE

'^FM^G 'EFKYIGNMLH

Enkephalinase 67 5 K£LFFLNFA(;:f7w)::GT®RPL Carboxypeptidase A 70 SREWlTUATqvWFAKKFl^ Carboxypeptidase B 67 AREWISPAFCq^VREAVR Carboxypeptidase E 73 GNEAVGRELLIFLAQ0:.CNEj^KGNE-

TTO-VHSPGNLFB IFLB^ K AI LDjSf D ll^^lJvgN PI^ T^FLDKLDFYVLPWNIDGY IYTWTTNRMWRK

NSYMNPERK TRSVTSSSL TRSTRAGSS

CRW? CVGVDAN CTGTDLN

IVOLIHNTRIH{I)^PSI^PDG(^.K|^SQLGELKDWFVGRSNAOGIDLN

Figure 4. Sequence homologies of enkephalinase and bovine carhoxynentidases A. B and E. Amino acids are shown in the one-letter code and numbered at the left. Identical residues are boxed whilst conservative changes, based on the following groupings: D-VM, DE, RKH, FWY, TS. GA, QN, P, C, are dotted.

ACKNOWLEDGMENTS ;We thank the Genentech DNA Synthesis group for oligonucleotides. Dr. Ellson Chen for his help and advice during DNA sequencing, and Drs. Bryan Finkle, Ben Borson, Claude Gros and Jean-Charles Schwartz for helpful discussions and encouragemenL

RRFRRRNCF.S 1. Malfroy B., Swerts JP., Guyon A., Roques B.P. and Schwartz J.-C. (1978) Nature 276,523-526. 2. Schwartz J.-C, Malfroy B. and De La Baume S. (1981) Life Sci. 29,1715-1740. 3. De La Baume S., Yi C.C., Schwartz J.-C., ChailletP., Marcais-CcUado H. and Costentin J. (1983) Neuroscience 8.143-151. 4. Schwartz J.-C. (1983) Trends Neurosci. 6,15-18. 5. Roques B.P., Foumie-Zaluski M.C., Soroca E., Lecomte J.M., Malfroy B., Llorens C. and Schwartz J -C

(1980) Nature 288,286-288. 6. Altstein M., Bachar E., Vogel Z. and Blumberg S. (1983) Eur. J. Pharmacol. 91,353-361. 7. Lecomte J.M., Costentin J., Vlaiculescu A., Chaillet P., Marcais-Collado H., Llorens-Cortes C., Leboyer M.

and Schwartz J.-C. (1986) J. Pharmacol. Exp. Therap. 237,937-944. 8. Horas P., Bidabe A.M., Caille J.M., Simonnet G., Lecomte J.M. and Sabathie M. (1983) Amer. J.

Neuroradiology 4, 653-655. 9. Matsas R., Fulcher I.S., Kenny A.J. and Turner A.J. (1983) Proc. Nad. Acad. Sci. U.S.A. 80,3111-3115.

10. Gafford J.T., Skidgel R.A., Erdos E.G. and Hersh L.B. (1983) Biochemistry 22,3265-3271. 11. Kenny A J. (1986) TIBS 11,40^2. 12. Llorens C. and Schwartz J.-C. (1981) Eur. J. Pharmacol 69,113-116. 13. Kerr M.A. and Kenny A.J. (1974) Biochem. J. 137,477-488. 14. Almenoff J., Wilk S. and Orlowski M. (1981) Biochem. Biophys. Res. Commun. 102,206-214. 15. Fulcher I.S., Matsas R., Turner A J. and Kenny A J. (1982) Biochem. J. 203, 519-522. 16. Malfroy B. and Schwartz J.-C. (1982) Biochem. Biophys. Res. Commun. 106,276-285. 17. Malfroy B. and Schwartz J.-C. (1984) J. Biol. Chem. 259,14365-14370. 18. Llorens C., Malfroy B., Schwartz J.-C., Gacel G., Roques B.P., Roy J., Morgat J.L., Javoy-Agid F. and Agid

Y. (1982) J. Neurochem. 39,1081-1089. 19. Hewick R.M., HunkapUIer M.W., Hood LB. and Dreyer W.J. (1981) J. Biol. Chem. 256,7990-7997. 20. Kaplan B.B., Bernstein S.L. and Gioio A.E. (1979) J. Biochem. 183,181-184. 21. Huynh T.V., Young R.A. and Davis R.W. (1985) in DNA Cloning Vol. 1, A Practical Approach (ed. Glover

D.) IRL, Oxford, 49-78. 22. Wood W.Letal. (1984) Nature 312,330-337. 23. Wood W.I., Gitschier J., Lasky L.A. and Lawn R.M. (1985) Proc. NaU. Acad. Sci. U.S.A. 82,1585-1588. 24. Messing J., Crea R. and Seeburg P.H. (1981) Nucleic Acids Res. 9,309-321. 25. Sanger F., Nicklen S. and Coulson A.R. (1977) Proc. NaU. Acad. ScL U.S.A. 74, 5463-5467. 26. Southern E.M. (1975) J. Molec. Biol. 98,503-517. 27. Taylor J.M., Illmensee R. and Summers S. (1976) Biochim. Biophys. Acta 442,324-330. 28. Dobner P.R., Kawasaki E.S., Yu L.Y. and Bancroft F.C. (1981) Proc. Nad. Acad. Sci. U.S.A. 78,2230-2234. 29. Thomas P.S. (1980) Proc. NaU. Acad. Sci. U.S.A. 77,5201-5205. 30. Ullrich A., Berman C.H., DuU TJ., Gray A. and Lee J.M. (1984) EMBO J. 3,361-364. 31. Relton J.M., Gee N.S., Matsas R., Turner A.J. and Kenny AJ. (1983) Biochem. J. 215,519-523. 32. Kozak M. (1984) Nucleic Acids Res. 12,857-872, 33. Schwartz J.-C., Giros B., Gros C., Llorens C. and Malfroy B. (1984) in Proceedings International Union of

Pharmacology 9th Congress of Pharmacology (eds. Mitchell J.F., Paton W. and Turner P.) Mc Millan Press Ltd., London, vol. 3,277-283.

65


34. Blobel G. (1980) Proc. Natl. Acad. Sci. U.S.A. 77,1496-1500. 35. Schneider C., Owen M.J., BanviUe D. and WilUams J.G. (1984) Nature 311,675-678. 36. Holland E.G., Leung J.O. and Drickamer K. (1984) Proc. NaU. Acad. Sci. U.S.A. 81,7338-7342. 37. Strubin M., Mach B. and Long E.O. (1984) EMBO J. 3.869-872. 38. Fields S., Winter G. and Brownlee G.G. (1981) Nature 290,213-217. 39. Malfroy B., Llorens C., Schwartz J.-C., Soroca E., Roques B.P., Roy J., Morgat J.L., Javoy-Agid F. and Agid

Y, (1981) in Advances in Endogenous and Exogenous Opioids (eds. Takagi H. and Simon EJ.) Elsevier, Amsterdam, 191-194.

40. Orlowski M. and Wilk S. (1981) Biochemistry 20,4942-4950. 41. Malfroy B. and Schwartz J.-C. (1985) Biochem. Biophys. Res. Commun. 130,372-378. 42. Hersh L.B. and Morihara K. J. (1986) Biol. Chem. 261,6433-6437. 43. Pozsgay M., Michaud C., Liebman M. and Orlowski M. (1986) Biochemistry 25,1292-1299. 44. Bradshaw R.A., Walsh K.A. and Neurath H. (1971) Biochemistry 10,938-972. 45. Titani K., Ericsson L.H., Walsh K.A. and Neurath H. (1975) Proc. Nad. Acad. Sci. U.S.A. 72,1666-1670. 46. Fricker L.D., Evans C.J., Esch F.S. and Herbert E. (1986) Nature 323,461-464. 47. Kyte L and Doolittle RP. (1982) J. Moiec. BioL 157,105-132.

66

structure of the Precursor to an Enzyme Mediating COOH-Terminal Amidation in Peptide Biosynthesis

Betty A. Eipper, Larry P. Park, Ian M. Dickerson, Henry T. Keutmann, Elizabeth A. Thiele, Henry Rodriguez, Peter R. Schofield, and Richard E. Mains

Department of Neuroscience (B.A.E., L.P.P., I.M.D., E.A.T., R.E.M.) The Johns Hopkins University School of Medicine Baltimore, Maryland 21205 Endocrine Unit (H.T.K.) Massachusetts General Hospital Boston, Massachusetts 02114 Genentech Inc. (H.R., P.R.S.) South San Francisco, California 94080

Many bioactive peptides terminate with an amino acid a-amide at their COON terminus. The enzyme responsible for this essential posttranslational mod-ification is known as peptidyl-glycine a-amidating monooxygenase or PAM. We identified cDNAs en-coding the enzyme by using antibodies to screen a bovine intermediate pituitary Xgtll expression li-brary. Antibodies to a /?-galactosidase/PAM fusion protein removed RAM activity from bovine pituitary homogenates. The 108,207 dalton protein predicted by the complete cDNA is approximately twice the size of purified RAM. An NHz-terminal signal se-quence and short propeptide precede the NHa ter-minus of purified RAM. The sequences of several RAM cyanogen bromide peptides were localized in the NHz-terminal half of the predicted protein. The cDNA encodes an additional 430 amino acid intra-granular domain followed by a putative membrane spanning domain and a hydrophilic cytoplasmic do-main. The forms of RAM purified from bovine neu-rointermediate pituitary may be generated by endo-proteolytic cleavage at a subset of the 10 pairs of basic amino acids in the precursor. High levels of RAM mRNA were found in bovine pituitary and cer-ebral cortex. In corticotropic tumor cells, levels of RAM mRNA and pro-ACTH/endorphin mRNA were regulated in parallel by glucocorticoids and CRF. (Molecular Endocrinology 1: 777-790, 1987)

INTRODUCTION

Although the importance of posttranslational process-ing of precursors to bioactive peptides is well known,

0888-8809/87/0777-0790$02.00/0 Molecular Endocrinology Copyright © 1987 by The Endocrine Society

very little is known about the many enzymes involved in converting inactive precursors into bioactive prod-ucts. We have focused our studies on the a-amidation of bioactive peptides (1-3). Approximately half of the known neuronal and endocrine peptides (e.g. gastrin, neuropeptide Y, vasopressin, CRF) have an a-amidated amino acid at their COOH terminus and the presence of this a-amide moiety is generally essential for biolog-ical activity. Alpha-amidation appears to be uniquely associated with bioactive peptides; a number of novel bioactive peptides have been identified based solely on the fact that they terminate with an amino acid a-amide (4, 5).

Prohormone sequences consistently contain a -Gly residue to the COOH-terminal side of the residue a-amidated in the product peptide. Using a synthetic peptide, o-Try-Val-Gly, Bradbury et al. (6-8) identified «-amidation activity in porcine pituitary extracts; reac-tion products included D-Tyr-Val-NHa, with the amide nitrogen derived from the amino group of Gly and glyoxylate. Although the amidation reaction requires a peptide substrate with a COOH-terminal glycine, it is capable of producing peptides terminating with a wide variety of amino acid a-amides.

Requirements for amidation activity in a number of different tissues are very similar (1-3, 6-15). The en-zyme purified from bovine neurointermediate pituitary requires the presence of molecular oxygen, catalyzes the consumption of ascorbic acid, and is inhibited by the addition of divalent metal ion chelators; activity is restored by the addition of CUSO4 (1-3). The enzyme has therefore been referred to as peptidyl-glycine a-amidating monooxygenase or PAM. Dopamine jS-hy-droxylase, a key enzyme in the synthesis of catechol-amines, exhibits similar requirements. Two forms of PAM were purified from bovine neurointermediate pi-tuitary: PAM-A, mol wt 54,000 and PAM-B, mol wt 39,000. Sequence analysis and peptide mapping indi-

777

MOL ENDO-1987 778

Voll No. 11

cated that the two were closely related. PAM Is local-ized to secretory granules and is cosecreted with pro-ACTH/endorphin-related peptides from corticotropic tu-mor cells in response to CRF or cAMP (13,14). Levels of PAM activity are regulated in parallel with levels of prohormone in corticotropes and melanotropes (15).

RESULTS AND DISCUSSION

Identification of cDNAs Encoding PAM

PAM purified from bovine neurointermediate pituitary was used to generate rabbit polyclonal antisera. When cross-linked to Protein A-Sepharose resin, PAM anti-serum 36 removed equivalent amounts of PAM activity and ^^ l-labeled PAM protein from solution (16), con-sistent with the identity of the enzyme and the major protein. Purified PAM was linked to activated CH-Seph-arose and used to prepare affinity purified PAM anti-serum for screening of a bovine intermediate pituitary Xgtl 1 expression library.

Purified PAM-B was used for NH2-terminal amino acid sequence determination (Table 1). Since there were ambiguities at a number of positions in the NH2-terminal sequence, data were obtained for a series of cyanogen bromide fragments of PAM-A and PAM-B (Table 1). The reduced and alkylated cyanogen bromide fragments were fractionated by gel filtration; fractions containing cyanogen bromide peptides ranging in size from 12,000 to 2000 daltons were further purified by reverse phase-HPLC. The amino acid sequences obtained for five cyanogen bromide peptides showed no significant ho-mology to amino acid sequences in the National Biomedical Research Foundation Protein Sequence Da-tabase. Oligonucleotides corresponding to three sepa-

rate regions of sequence were synthesized (Table 1). A bovine intermediate pituitary Xgt11 cDNA library

was prepared from 5 g poIy(A)" RNA. Complementary DNAs of greater than 500 base pairs (bp) were selected and a total of 2.3 x 10® recombinant phage were obtained. When screened with the nick translated cDNA for pro-ACTH/endorphin, approximately 14% of the recombinant phage were positive, as expected for this tissue. Upon screening 360,000 recombinant phage with affinity purified PAM antibody 36, four positive phage were identified and plaque purified. Phage XPAM-1 contained the largest insert and the EcoRI fragments generated from XPAM-1 [0.4-kilobase (kb), 0.7-kb, and 2.2-kb fragments] were subcloned, nick translated, and used to probe Southern blots of EcoRI digests of the other XPAM phage (16); the phage were found to represent overlapping clones (Fig. 1A).

The fusion proteins produced by XPAM-1, -3, and -5 were quite similar in size (Fig. 2), despite the fact that the 2.2-kb fragment of XPAM-1 cross-hybridized with much smaller (1.4 kb) EcoRI fragments from XPAM-3 and -5. The fusion protein produced by XPAM-2 was slightly smaller than the fusion protein produced by XPAM-1, -3, or -5, suggesting that the regions missing in XPAM-2 were protein coding and that the 5'- to 3'-orientation of the EcoRI fragments was as shown in Fig. 1. The fact that oligonucleotides 6 and 8 cross-hydridized with the 0.7 kb EcoRI fragments of XPAM-1, -3, and -5, but not with XPAM-2, -3A, or -5A also indicated that the portion of the 0.7-kb fragment adja-cent to the 0.4-kb fragment was a protein-coding re-gion.

The EcoRI fragments of XPAM-1 and XPAM-5 were subcloned and sequenced by the dideoxy chain termi-nation method (18-20) using the strategies outlined in Fig. IB. Although XPAM-1 contained the poly(A) tail, partial sequence analysis demonstrated that the com-

Table 1. Amino Acid Sequence Data for PAM and its Cyanogen Bromide Fragments Location in

Peptide Sequence Predicted Protein

NH2-Terminal and B.E36B PheLysGluThrThrArgSerPheSerAsnGluCysLeuGlyThr 31-45 NH2-terminal and B.E38B SerPheSerAsnGluCysLeuGlyThrThrArgProVallleProlleAsp 37-53

HR7 ProGlyValThrProLysGlnSerAspThrTyrPheCys 64-76 B.F28 AspGluGluAlaPheVallieAspPheLysProArgAlaSerThr 84-97 HR6 MetMetSerValAspThrVallleProProGlyGlyLvsValValAsnSerAsplleSerCysHisTyrLysLvsTyrPro 202-228

(0LIG06) CN2 HisValPheAlaTyrArqValHisThrHlsHisLeuGlyLysValValSerGlyTyrArqValArqAsnGlyGlnTrpThrLeulle 230-258

(0LIG08) HR3 GluAlaLysHlsAlaValSerPheMetThrCysThrGlnAsnValAlaProAspllePheArq 319-339

(0LIG07) For each analysis, PAM was purified from 400 frozen bovine neurointermediate pituitaries (2). The NHa-terminal sequence of PAM-B was determined for four different preparations; for the NHa-terminal sequence, multiple residues were consistently found at every cycle. Data are consistent with the presence of the two major NH2-terminal sequences indicated; approximately equivalent amounts of Glu and Tyr were consistently observed at position 41. The CNBr peptides sequenced came from the pools indicated: CN2 (8500-14,500 daltons); HR3 (6400-8000 daltons); HR6 (3500-4800 daltons); B.E36B and B.E38B (3100-4800 daltons); HR7 (1700-2900 daltons); B.F28 (1500-2200 daltons). Oligonucleotide probes (OLIGO 6, 7, and 8) corresponding to the preliminary versions of the underlined amino acid sequences were synthesized.

Cloning of PAM cDNA 779

PAM c D N A c lones i so la ted f rom B I L - 4

A. PAM ant ibody sc reen ing

> P A M - I > P A M - 2 > P A M - 3

X P A M - 5

0 . 4 k b , 0 . 7 k b ,

(3A)

(5 A)

2 . 2 k b

Bi Sequencing s t r a t e g y ; r e l a t i o n s h i p of cDNA to pep t ides and o l i g o n u c l e o t i d e s

(J 5 0 0 1000

X P A M - I

1500 — I —

2 0 0 0 2 5 0 0 3 0 0 0 3 5 0 0 — I 1 1 1

E C ^ R I E C ^ R I

H h

X P A M - 5 ECOR I ECOR I

\ A .

X P A M - 6

P e p t i d e s

ECOR I ^ ECOR I

0 N - T E R

B HR 7

Q B.F28

H R 6 HR3 m

CN2

Ol igonuc leot ides 0 L I G 0 6 0 L I G 0 7

OLIGO 8 Fig. 1. XPAM Clones Obtained by Screening cDNA Library with PAM Antibodies

A, Four phage identified with affinity purified PAM antiserum 36 were plaque purified from the Xgt11 bovine intermediate pituitary cDNA library. Phage DNA was digested with EcoRI and the sizes of the fragments released were determined by agarose gel electrophoresis. The EcoRI fragments of XPAM-1 were subcloned into Bluescript (Stratagene), nick translated, and used to probe Southern blots of EcoRI digests of the other phage. The fragments were initially placed in 5'- to 3'-order based on the molecular weight of the fusion proteins produced in E. coli strain CAG456 cells and on the ability of synthetic oligonucleotides (Table 1) to hybridize with the restriction fragments. Phage XPAM-3A and -5A arose during the plaque purification of XPAM-3 and -5. B, The sequencing strategies used for XPAM-1, -5, and -6 are indicated. The relationship of the cyanogen bromide peptides sequenced and the synthetic oligonucleotides to the cDNA are indicated.

MOL ENDO-1987 780

VoM No. 11

CON WT P-1 P-2 P-3 B Ab G P G R G P G R G R G P

kD

2 0 0 -

1 1 6 - m t j k

9 2 -

m

6 6 -

4 5 -

3 1 -

Fig. 2. Production of Fusion Protein by XPAM Clones Ten-milliliter cultures of E. coli strain CAG456 cells were infected with wild type Xgt l l (WT), XPAM-1 (P-1), XPAM-2 (P-2), or

XPAM-3 (P-3), and induced with 10 mw isopropyl-,8-D-thiogalactopyranoside; after 2 h at 37 C cells were pelleted and dissolved in boiling SDS-gel sample buffer (17) containing 0.3 mg/ml phenylmethylsulfonylfluoride. Uninfected CAG456 cells served as the control (CON). Aliquots equivalent to 5% of the sample were fractionated by gel electrophoresis on 8% polyacrylamide, 0.2% N, A/'-methylenebisacrylamide slab gels (17), transferred to nitrocellulose, and visualized with PAM antibody 36 (P) or a mouse monoclonal antibody to /3-galactosidase (G) (Promega, Madison, Wl) and ^^^l-labeled Protein A. The visualization procedure was identical to that described for screening the Xgtl 1 library. Location of molecular weight markers is indicated.

plete 5'-region was not contained within XPAM-1. In order to obtain a cDNA clone encoding the NHa-terml-nus of the protein, the original library was rescreened with 0.4- and 0.7-kb EcoRI fragments of XPAM-1. Screening of 360,000 recombinant phage revealed 29

phage positive for both cDNA fragments and four pos-itive for the 0.4-kb fragment of XPAM-1 but not for the 0.7-kb fragment of XPAM-1. A subset of the positive phage were plaque purified and characterized by South-ern blot analysis. An 0.8-kb EcoRI fragment cross-


hybridizing with the 0.4-l<b fragment of XPAM-1 (from XPAI\/I-6, Fig. 1B) was subcloned and sequenced and found to encode the NH2 terminus of the protein. The bovine intermediate pituitary library contained approxi-mately 2500 times as many copies of pro-ACTH/en-dorphin cDNA as PAM cDNA, in good agreement with previous estimates based on protein and peptide puri-fication (1,2).

Characterization of Fusion Protein

The molecular weights of the fusion proteins (21, 22) produced by bacteria infected with XPAM-1, 2 and 3 were determined. Extracts of infected bacteria were fractionated by sodium dodecyl sulfate (SDS)-polyacryl-amide gel electrophoresis, transferred to nitrocellulose, and visualized with antibody to /S-galactosidase or PAM (Fig. 2). Although the fusion proteins produced were clearly heterogeneous, the largest protein recognized by the PAM antibody had a molecular weight slightly greater than that of myosin, indicating that a fusion protein with a PAM-related region of approximately 100,000 daltons was being synthesized. PAM purified from bovine neurointermediate pituitary is only half this size (1, 2). Upon lysis with a buffer containing Na TES and NaCI, the fusion protein remained insoluble. At-tempts to solubilize the fusion protein by extraction of the pellet with buffer containing 1% NP-40, Triton X-100, 3-[(3-cholamidopropyl)dimethylammonio]-1-pro-panesulfonate, or Tween-20 were unsuccessful. Time course studies demonstrated the same heterogeneous collection of fusion proteins when incubation times were as short as 1 h.

In order to verify that the cloned cDNA did indeed encode the enzyme PAM, XPAM-1 encoded fusion proteins were partially purified by gel filtration in the presence of SDS and used to generate antibodies. The same major band was seen when Western blots of bovine neurointermediate pituitary were visualized with the fusion protein antibody or with PAM antibody 36. Fusion protein antibodies were linked to Protein A Sepharose resin and assayed for their ability to bind PAM enzyme activity from preparations of purified PAM and from bovine pituitary homogenates (Table 2). One of the fusion protein antibodies obtained bound enzyme activity from preparations of PAM-A, PAM-B, and crude bovine neurointermediate pituitary homogenate. The

fact that antibodies to fusion protein recognize bovine pituitary PAM activity demonstrates that the cDNAs obtained do encode an a-amidation enzyme.

Sequence of PAM cDNA

The nucleotide sequences of XPAM-1, -5, and -6 were assembled to form a 3724 bp cDNA (Fig. 3). A single potential initiator (23) Met occurs at position 134 and is preceded by a GC rich (84% G+C) 5'-nontranslated region and followed by a putative signal sequence; the only other Met codon in this region (position 67) is followed immediately by a stop codon. The nucleotide sequences of the two 0.7-kb fragments examined were identical except for a single base difference at position 954, resulting in substitution of an Arg (XPAM-5) for a His as shown for XPAM-1. The sequences of all five cyanogen bromide fragments of PAM were contained within the 0.8- and 0.7-kb fragments. The sequence of the 1.4-kb fragment of XPAM-5 was contained within that of the 2.2-kb fragment of XPAM-1 except for the presence of a 54 bp insertion in the 2.2-kb fragment of XPAM-1 (positions 2823-2876). The 54 bp insertion in XPAM-1 encodes an 18 amino acid peptide and may represent use of alternate acceptor sites in the splicing of an intron; the nucleotide sequence in this region is compatible with the occurrence of alternate splicing (24). The cDNAs sequenced contain open reading frames encoding 972 or 954 amino acids (proteins of 108,207 daltons or 106,195 daltons, respectively). The 675 bp 3'-nontranslated region contains multiple stop codons in all three reading frames and is relatively AT rich (62% A+T). The poly(A) tail in XPAM-1 is preceded by a consensus poly(A) addition site (AATAAA) 16 bp before the poly(A) tail (25,26). No homologous repeated sequences were found in XPAM-1 using either the nucleotide or amino acid sequence. No significant ho-mologies were found to sequences in the GenBank (Update 48), European Molecular Biology Laboratories, or National Biomedical Research Foundation data-bases.

Structure of PAM Precursor

The key features of the proteins encoded by these cDNAs are summarized in Fig. 4. The single potential initiator Met is followed immediately by a hydrophobic

Table 2. Binding of PAM Activity to Antibody Resins

Resin PAM Activity (pmol/h) % Input Activity Bound to Resin

Resin Buffer PAM-A PAM-B Pituitary

Supt PAM-A PAM-B Pituitary Supt

Preimmune 0.001 0.03 0.002 0.06 3 0.3 2 Ab 36 (bovine PAM) 0.78 1.07 1.00 1.58 31 34 29 Ab 46 ((8-gal-PAM fusion) 0.001 0.55 0.16 0.52 58 25 19

Aliquots of PAM-A (0.95 pmol/h), PAM-B (0.65 pmol/h), or 100,000 x g bovine neurointermediate pituitary supernatant (2.78 pmol/ h) were incubated with Protein A-Sepharose resin to wfiicli immunoglobulins from the antisera indicated had been covalently cross-linked (16). Aliquots (100 /il 1:10 slurry) of the antibody resins were tested. PAM activity bound to the resin was assayed in duplicate; duplicates varied less than ±15%. This experiment was repeated nine times in various forms with similar results, looking at depletion of PAM activity from the supernatant or appearance of PAM activity bound to the resin.

MOL ENDO • 1987 Vol 1 No. 11 782

30 . . 60 . . 90 , GCGGGCCK:TGAGCGCGCGCGGCGCTGCTGTGTGGCCGCCGCGGGGACGCCCGCCGCGCCCCGGGCCATGMGTAGCGGCTGCCGGCGGCGCCGCGCCTCTCGGTCC

1 120 . ^ . 150 . . 180 . . 210 . CGGAGCGCCGAGGACATGGCTGGCITTCGGAGCCTGCTAGTTCTTCTCCTCGTGTTTCCCAGTGGCTGTGTGGGCTTCCGAAGCCCACXTTCTGTCTTTAAGAGGTTTAAAGAAACTACC

MetAlaGlyPheArgSerLeuLeuValLeuLeuLeuValPheProSerGlyCysValGIyPheArgSerProLeuSerValPheLvsArgPheLvsGluTh^ 10 20

240 . , 270 . . 300 . . 330 AGATCATTTTCCAATGAATGTCTTGGTACCACCAGACCAGTAATTCCTATCGATTCATCAGATTTTGCGTTGGATATTCGCATGCCTGGGGTCACACCCAAACAGTCTGATACGTACITC ArKSerPheSerAsnGluCysLeuGlyThrThrArgProVallleProIleAspSerSerAspFheAlaLeuAspIleArgMetProGlyValThrProLysGlnSerAspThrTyrPhe

AO 50 60 70

360 . . 390 . . 420 . . A50 TGCATGTCAGTGCGTTTGCCAATGGATGAGGAAGCCTTCGTGAITGACTTCAAACCCCGTGCCAGCATGGATACTGTCCATCATATGTTACTGTTTGGATGCAATATGC CYSMetSerVaLArgLeuProMetAspGluGltiAlaPheVallleAspPheLYsProArgAlaSerMetAspThrValHisHlsMetLeuLeuPheGlvC^

80 90 100 110

'•SO • • 510 . . 540 . . 570 . GGAAATTACTGGTTTTGTGATGAAGGCACCTGTACAGATAAAGCCAATATTCTCTATGCCTGGGCAAGAAATGCTCCCCCAACGAGACTCCCCAAAGGTGTTGGATICAG^^ GlyAsnTyrTrpPheCyBAspGluGlyThrCysThtAspLysAlaAsnlleLouTyrAlaTrpAiaArgAsnAlaProProThrArgLeuProLysGLyValGlyPheArgV

120 130 140 150

600 . , 630 . . 660 , . 690 GAGACTGGAAGCAAATACTTTGTACTCCAAGTACACTATGGGGATATTAGTGCTTTTAGAGATAATCACAAGGACTGCTCTGGTGTTTCCTTACACCTCACACGCCTGCC^^ GluThrGlySerLysTyrPheValLeuGlnValHisTyrGlyAspIleSerAlaPheArgAspAsnHisLysAspCysSerGlyValSerLeuHisLeuThrArgLeuProGlnProLeu

160 170 180 190

720 . . 750 . . 780 . . 810 ATTGCTGGCATGTACCTTATGATGTCTGTTGACACTGTTATACCACCAGGAGGGAAAGTGGTGAATTCTGATATTTCATGCCATTATAAAAAGTATCCAATGCATGTCTTTGCCT^^ IleAlaGlyMetTyrLeuMetMetSerVaUspThrVallleProProGlyGlyLysValValAsnSerAspIIeSerCYsHlsTyrLysLyaTyrProMetHisValFheAlaTYrAri^

200 210 220 • • 230

S'lO . . 870 . . 900 . . 930 . . G GTICACACTCACCATTTAGGCAAGGTAGTCAGTGGATACAGAGTAAGAAATGGACAGTGGACCCTGATTGGACGCCAAAGCCCCCAACTGCCACAGGCCTTCTACrc^ ValHlsThrHisHlsLeuGlYLysVaLValSerGlvTvrArKVaLArKAsnGlyGlnTrpThrLeuIleGlYArgGlnSerProGlnLauProGlnAlaPheTvrProVa

240 250 260 270 Arg

960 . . 990 . . 1020 . . 1050 GTGGATGTGAGTTTTGGTGACATCCTGGCAGCAAGATGTGTGTTCACTGGGGAAGGGAGGACAGAAGTCACACACATTGGTGGCACGTCTAGTGATGAAATGTGCAACTTATACATTATG ValAspValSerPheGlyAspIleLeuAlaAlaArgCysValPheThrGlyGluGlyArgThrGluValThrHisIleGlyGlyThrSerSerAspGluMetCysAsnLeu^

280 290 300 310

1080 . . 1110 . . 1140 . . 1170 TATTACATGGAAGCCAAGCATGCAGTTTCTTTCATGACCTGTACACAGAATGTAGCTCCAGATATATTCAGAACCATACCACCAGAGGCCAATATTCCAATTCCTGTGAAATCCGA^ TyrTYrMetGluAlaLY3HisAlaValSerFheMetThrCv5ThrGlnAsnV«1A1«PrnAsr.Il^PhBAr»T>,rT1«PrnPrnr,1nA1«^

320 330 340 350

1200 . . 1230 . . 1260 . . 1290 GTTATGATGCATGGACATCACAAAGAAACAGAGAACAAAGACAAGACTTCTTTACTACAGCAGCCCAAACGAGAAGAAGAAGGAGTGTTAGAACAGGGTGATTTCTATTCACTACTTTCC ValMetMetHisGlyHisHisLysGluThrGluAsnLysAspLysThrSerLeuLeuGlnGlnProLysArgGluGIuGluGlyValLeuGluGlnGlyAspPhaXyrSerLeuLeuSer

360 370 ^ ^ 3 8 0 390

1320 . . 1350 . . 1380 . . 1410 AAGCTGCTAGGAGAAAGGGAAGATGTTGTTCATGTGCATAAATATAATCCTACAGAAAAGGCAGAATCAGAGTCTGACCTGGTAGCTGAGATTGCAAACGTAGTCCAAAAGAAGGATCTT LysLeuLeuGlyGluArgGluAspValValHisValHisLysTyrAsnProThrGluLysAlaGluSerGluSerAspLeuValAlaGluIleAIaAsnValValGlnLysLysAspLeu

400 410********* 420 430 ^ ^ Fig. 3. Complete Nucleotide Sequence of PAM cDNA

The nucleotide sequences obtained for XPAM-1, -5, and -6 have been used to construct the cDNA sequence and to predict the amino acid sequence. The entire molecule was sequenced on both strands of the DNA, as diagramed In Fig. 1B. The amino acid sequence is numbered below the line from the initiator Met (marked by a large downward arrow). Amino acid residues underlined coincide with peptide sequence information obtained from cyanogen bromide fragments (Table 1). Pairs of basic amino acids are marked by a pair of dark triangles. Potential sites for N-linked glycosylation are indicated with dots. The poly(A) addition signal is boxed. In \PAM-5, position 954 is G, resulting in substitution of Arg for His; in addition, in XPAM-5, the 54-bp segment

Cloning of PAM cDNA i 783

1<|40 . . 1470 . . 1500 . . 1530

GGTCGGTCCGATACCAGAGAGAGTGCAGAACAGGAGAGGGGCAATGCTATTCTTGTCAGAGACAGAATTCACAAATTCCACAGACTAGTGTCTACCTTGAGGCCTGCAGAGAGCAGAGTT

GlyArsSerAspThrArgGluSerAlaGluGlnGluArgGlyAsnAlalleLeuValArsAspArglleHisLysPheHisArgLeuValSerThrLeuArgProAlaGluSerArgVal

440 450 460 470

1560 . . 1590 . . 1620 . . 1650

CTGTCGTTACAGCAGCCCCTACCTGGTGAAGGCACCTGGGAACCAGAACACACAGGAGATTTCCATGTAGAAGAGGCACTGGATTGGCCTGGAGTATACTTGTTACCAGGCCAGGTTTCT

LeuSerLeuGlnGlnProLeuProGlyGluGlyThrTrpGluProGIuHisThrGlyAspPheHisValGluGluAlaLeuAspTrpProGlyValTyrLeuLeuProGlyGlnValSer

480 490 500 510

1680 . . 1710 , . 1740 . . 1770

GGGGTGGCTCTGGACCCTCAGAATAATCTCGTGATTTTCCACCGAGGTGACCACGTCTGGGATGGAAACTCTTTTGATAGCAAGTTTGTTTACCAGCAAAGAGGTCTTGGACCAATTGAA

GlyVaLAlaLeuAspProGlnAsnAsnLeuValllePheHisArgGlyAspHisValTrpAspGlyAsnSerPheAspSerLysPheValTyrGlnGlnArgGlyLeuGlyProIleGlu

520 530 540 550

1800 . . 1830 . . 1860 . . 1890

GAAGACACTATTCTTGTCATAGATCCAAATAATGCTGCAGTACTCCAGTCCAGTGGAAAAAATCTGTTTTACTTGCCACATGGCTTGAGTATAGATAAAGATGGAAATTATTGGGTCACA

GluAspThrlleLeuVallleAspProAsnAsnAlaAlaValLeuGlnSerSerGlyLysAsnLeuPheTyrLeuProHisGlyLeuSerlleAspLysAspGlyAsnTyrTrpValThr

560 570 580 590

1920 . . 1950 . . 1980 . . 2110

GACGTGGCGCTTCATCAGGTGTTCAAACTAGATCCAAAGAGTAAAGAAGGCCCTCTGCTAACCCTGGGAAGGAGCATGCAACCAGGCAGTGACCAGAATCACTTCTGTCAGCCCACCGAT

AspValAlaLeuHisGlnValPheLysLeuAspProLysSerLysGluGlyProLeuLeulhrLeuGlyArgSerMetGlnProGlySerAspGlnAsnHisPheCysGlnProThrAsp

600 610 620 630

2040 . . 2070 . . 2100 . . 2130

GTGGCTGTGGATCCAGACACCGGAACCATCTATGTGTCAGATGGCTACTGCAACAGTCGCCTXGTGCAGTTTTCACCAAGTGGAAAATTCATCACACAGTGGGGAGAAGCGTCTCTAGAG

ValAlaVaLAspProAspThrGlyXhrlleTyrValSerAspGlyXyrCysAsnSerArgLeuValGlnPheSerProSerGlyLysPhelleThrGlnTrpGlyGluAlaSerLeuGlu

640 650 660 670

2160 . . 2190 . . 2220 . . 2250

AGCAGTCCTAAACCAGGCCAGTTCAGAGTTCCTCACAGCTTGGCCCTGGTGCCTCCCCTGGGCCAGCTGTGTGTGGCAGACCGGGAAAACGGTCGGATCCAGTGTTTCAAAACCGACACC

SerSerProLysProGlyGlnPheArgValProHisSerLeuAlaLeuValProProLeuGlyGlnLeuCysValAlaAspArgGluAsnGlyArglleGlnCysPheLysThrAspThr

680 690 700 710

2280 . . 2310 . . 2340 . . 2370

AAAGAATTTGTGCGCGAGATTAAGCACCCATCGTTTGGAAGAAATGTGTTTGCAATTTCATACATACCAGGTTTGCTCTTTGCGGTGAATGGAAAGCCTTACTTTGAGGACCAAGAACCC

LysGluPheValArgGluIleLysHisProSerPheGlyArgAsnValPheAlalleSerTyrlleProGlyLeuLeuPheAlaValAsnGlyLysProTyrPheGluAspGlnGIuPro

720 730 740 750

2400 . . 2430 . . 2460 . . 2490

GTGCAAGGATTTGTGATGAACTTTTCCAGCGGGGAAATTATCGATGTCTTCAAGCCAGTGCGCAAGCACTTCGACATGCCCCACGACATTGCTGCGTCCGAGGACGGGACCGTGTATGTC

ValGlnGlyPheValMetAsnPheSerSerGlyGluIlelleAspValPheLysProValArgLysHisPheAspMetProHisAspIleAlaAlaSerGluAspGlyThrValTyrVal

760 ••••••••• 770 ^ ^ 780 790

2520 . . 2550 . . 2580 . . 2610

GGAGACGCTCACACCAACACCGTGTGGAAGTTCACCTCGACCGAAAAAATGGAACATCGATCAGTTAAGAAGGCTGGCATTGAGGTTCAGGAAATCAAAGAATCCGAGGCAGTTGTTGAA

GlyAspAlaHisThrAsnThrValTrpLysPheThrSerThrGIuLysMetGIuHisArgSerVaLLysLysAlaGlylleGluValGlnGlulleLysGluSerGluAlaValValGlu LysLysfl

A Ao 800 810 ^ - ^ 8 2 0 830

2640 . . 2670 . . 2700 . . 2730

ACCAAAATGGAGAACAAGCCCGCCTCCTCAGAATTGCAGAAGATACAAGAGAAACAGAAGCTGGTCAAAGAGCCGGGCTCCGGAGTGCCGGCTGTTCTCATTACAACCCTTCTGGTTATT

ThrLysMetGluAsnLysProAlaSerSerGluLeuGlnLysIleGlnGIuLysGlnLysLeuValLysGluProGlySerGlyValProAlaValLeuIleThrThrLeuLeuVallle

840 850 860 870

2760 . . 2790 . 2820 . . 2850

CCTGTGGTTGTCCTGCTGGCCATTGCCTTATTTATTCGGTGGAAAAAATCAAGGGCCTTTGGAG S>TTCTGAACGTAAACTTGAGGCCAGTTCGGGAAGAGTTCTGGGAAGACTTAGAG3A

ProValValValLeuLeuAlalleAlaLeuPhelleArgTrpLysLysSerArgAlaPheGlyAspSerGluArgLysLeuGluAlaSerSerGlyArgValLeuGlyArgLeuArgGLy

880 J k t ^ ^ 910

Fig. 3, part 2 (continued)

MOL E N D 0 . 1 9 8 7 784

V o l ! No. 11

2880 . . 2910 . . 2940 . . 2970

AAGGGAGGCGGAGGCCTAAACCTGGGCMCTTCTTCGCCAGCCGGMG(KK:XACAGCCGGMGGGGTTCGACCGCCTCAGCACGGMGGMGCGACCAGGAGAMGATGATO^ LysGlyGlyGlyGlyLeuAsnLeuGlyAsnPhePheAlaSerArgLysGiyTyrSerArgLysGlyPheAspArgLeuSerThrGluGlySGrAspGLnGluLysAspGluAspAlaSer

920 ^ ^ ^ ^ 940 950

3000 . . 3030 . . 3060 . . 3090

GAGTCCGAAGAAGAGTACTCGGCCCCGCCGCCCGCCCCTGCCCCTTCCTCCTGAGAAACTGGGTTTTCTTTTAGGCTGACGAGACTTACCAAGGATGCCAGCTTCCTTTCCCCTTGAGCA GluSerGluGluGluTyrSerAlaProProProAlaProAlaProSerSer

960 970

3120 . . 3150 . . 3180 . . 3210

CGTTGAGCGTTTTGCGTATTTAACTGTAAACTGTACCCATCTGTGTGGGACCGTACACCTTTTATTTACTTCCXTTGGGATTAGTTGGCTTCTGTTCCTAGTTGAGGAGTTTCCTGAAAG

3240 . . 3270 . . 3300 . . 3330

TTCATTCATCGTGCCATTGTCTTTATATGAACATAGGCTAGAGAAGTGATCCTCTTCTTCCGTCACACTAGTCACTTAGGGATGGAAGGTXTGCTCATCTGCATTTCTGAGACTTTTCTG

3360 . . 3390 . . 3420 . . 3450

TAGTTTGTAAATAACTCCATTCTCTGCTTGAACACAGTATTCTCCCAGTAGCACTTCCATTGCCAGTGTCTTTCITTGGTGCCTTTCCTGTTCAGCATTCTCAGCCTGTGGCAGTGAAGA

3480 . . 3510 . . 3540 . . 3570

GAAACTTTGTGCTACATGACAACAAAGCTGCTAAATCTCCTATTTTTTTAAAATCACTAACATTATATTGCAACAAGGGAAAGAAAAAAGTCTCTATTTAAATTGTTTTTTTTTTAATXT

3600 3630 3660 3690

CCTTCCTCAGTTGGTGTGTTTTGGGGATGTCTTATTTTTAGATGGTTACACTGTTAGATCACTATTTTCAGAATCTGAATGTAATTTGTGlf^TAAifeTGTTTTCAGAGCATTAAAAAAA lAAijsi

3720

AAAAAA

Fig. 3, part 2 (continued)

amino acids

200

I^^NHg-termini KR of PAM KK

400

KR KK

600 1 1 r

800 RK 18 a m i n o a c i d "1 i - .

insert ion in X P A M - I j L

signal s e q u e n c e

T HTHH

RK KK KK RK

RKi

HGHH p o t e n t i a l m e m b r a n e

s p a n n i n g d o m a i n

l ikely C - t e r m i n u s l ike ly C - t e r m i n u s

o f P A M - B of P A M - A

hydroph i l i c c y t o p l a s m i c d o m a i n w i t h p o t e n t i a l p h o s p h o r y l a t i o n s i t e s

Fig. 4. Summary Diagram of Key Features of PAM The locations of the putative signal sequence and membrane spanning domain are indicated. All potential paired basic cleavage

sites are indicated, as are the NH2 termini and likely COOH termini of PAM-A and PAM-B. The sequences of two His-rich regions potentially involved in copper binding are indicated.

segment characteristic of a signal sequence (27-29). Based on all of the accepted criteria (27-29), the only likely site for signal peptide cleavage is between Gly ° and Phe^^ until the exact site has been experimentally determined, the amino acid sequence has been num-bered with the initiator Met as position 1.

Protein sequence information from intact PAM-B and

from two cyanogen bromide peptides (Table 1; dem-onstrated two populations of molecules differing in length at the NHa-terminus. One could be attributed to endoproteolytic cleavage after Lys ®Arg®°, and another after Arg ®; the cleavage after Arg®® appeared slightly predominant, constituting approximately 60% of the total. This heterogeneity could have arisen, at least in


part, from the use of frozen bovine pituitaries as the source of the protein purified for sequence analysis. The short segment between Phe^ and the NHg-terminal residues appears to represent a "pro-peptide" removed after cleavage of the signal sequence. It is not yet clear whether the presence of this pro-peptide affects enzy-matic activity.

Ten pairs of basic amino acids occur In the protein. Generation of the COOH termini of PAM-A and PAM-B could result from endoproteolytic cleavage at pairs of basic amino acids (Fig. 4). PAM-B could be produced by cleavage at (producing a 347 amino acid protein with a mol wt of 38,934) and PAM-A by cleavage at Lys' ^ -Lys'*^^ (producing a 401 amino acid protein with mol wt of 45,046). In this scheme the peptide segment present in PAM-A and absent from PAM-B is acidic (calculated pi = 4.6) and would account for the fact that PAM-A binds to diethylaminoethyl-cellulose at neutral pH while PAM-B does not (2). Two clusters of His residues occur in the part of the protein encoding the active enzyme (Fig. 4) and may be in-volved in the interaction of the enzyme with copper (30). This region contains one of the two potential sites for N-linked glycosylation (Asn'*^^-Pro-Thr^^^); however, the presence of a Pro between the Asn and Thr makes glycosylation at this site unlikely (31). Consistent with this structure bovine PAM-B does not bind to lectin affinity resins (2).

The open reading frame continues for another 430 residues after the putative COOH terminus of PAM-A before reaching an extremely hydrophobic 24 amino acid segment that very likely constitutes a transmem-brane domain (amino acid residues 864-887). Two pairs of basic amino acids occur in the protein segment between the COOH terminus of PAM-A and the putative transmembrane domain; this segment of the protein would be exposed to the intragranular milieu and these sites could constitute endoproteolytic processing sites. Since this region of the protein has never been exam-ined, it is not known whether the remaining potential site for N-linked glycosylation (Asn^^^-Phe-Ser^®^) is used.

The calculated hydrophobicity index for the putative transmembrane segment is 2.9; any peptide segment of this length with a hydrophobicity index of greater than 1.6 is highly likely to be membrane-spanning (32, 33). Immediately after the putative transmembrane do-main is a cluster of basic amino acids (-Arg®®®-Trp-Lys-Lys-Ser-Arg®^^). Such clusters of basic amino acids are characteristic of the cytoplasmic side of membrane spanning domains and are believed to serve as a stop transfer signal (32, 33). If the hydrophobic segment does serve as a transmembrane domain, the remaining COOH-terminal region of the protein (residues 888 to 972) would be positioned in the cytoplasm. Preliminary studies have demonstrated the presence of substantial amounts of membrane-associated PAM activity in mouse corticotrope tumor cells and rat anterior pitui-tary, with lesser amounts of membrane associated PAM activity in rat neurointermediate pituitary (May, Victor, personal communication).

The COOH-terminal domain contains three or four pairs of basic amino acids and has a very unusual amino acid composition. In XPAM-1 the open reading frame after the putative transmembrane domain en-codes a total of 85 amino acids before reaching an in frame stop codon (3050); in XPAM-5 this region is 67 amino acids long. The 54 bp insertion that distinguishes XPAM-1 and XPAM-5 occurs immediately after the pu-tative transmembrane domain and encodes an 18 amino acid segment containing one of the pairs of basic amino acids (Fig. 4). The COOH-terminal domain of both proteins is very rich in charged amino acids (19% Lys or Arg, 17% Asp or Glu for XPAM-1) and in Ser or Thr (16%), and Gly or Pro (20%). Based on the sub-strate specificity of cyclic nucleotide-dependent protein kinases, Ca'^'^'-calmodulin-dependent protein kinase, and protein kinase C, the COOH-terminal domain contains several potential phosphorylation sites (Ser892.934.942j (34_38).

Northern Blot Analysis of PAM mRNA

In order to determine the approximate size of the mRNA encoding PAM and its tissue localization, total RNA from a variety of bovine tissues was fractionated on denaturing agarose gels, transferred to Nytran, and hybridized with probes prepared from the 0.7-kb and 2.2-kb fragments of XPAM-1 (Fig. 5). The highest levels of PAM mRNA were found in the neurointermediate lobe of the pituitary and in the cerebral cortex; lower levels were found in the anterior pituitary and none was detectable in the liver. Levels of PAM mRNA in bovine hypothalamus were similar to those in cerebral cortex; levels in cerebellum were below the limit of detection. Some evidence of size heterogeneity was observed, with the major mRNA species being approximately 3.7 kb. Both cDNA probes hybridized with mRNA species of similar size in rat pituitary and brain and in mouse AtT-20 cells.

Previous studies had demonstrated that levels of PAM enzyme activity in AtT-20 corticotropic tumor cells were down-regulated by glucocorticoids in parallel with levels of immunoactive hormone (14). In order to deter-mine whether this parallel regulation of prohormone and processing enzyme reflected parallel regulation at the level of mRNA, AtT-20 cells were treated with the synthetic glucocorticoid dexamethasone or the secre-tagogue CRF; total RNA was fractionated on denatur-ing agarose gels and analyzed with PAM cDNA and pro-ACTH/endorphin cDNA (Fig. 6). Levels of PAM mRNA and pro-ACTH/endorphin mRNA were both di-minished approximately 2-fold by pretreatment with dexamethasone and stimulated approximately 2-foId by application of CRF. Thus the parallel regulation of hor-mone and PAM enzyme activity observed previously can be accounted for by parallel alterations in the levels of the two mRNAs.

CONCLUSIONS

PAM represents the first posttranslational processing enzyme whose structure has been fully elucidated.

MOL ENDO-1987 786

Vol 1 No. 11

A B

k b

7 . 4 -

4 . 4 -

2 . 3 7 -

1 . 3 5

C L N A C L N A C L N A Fig. 5. Northern Blot of Various Bovine Tissues

Duplicate aliquots of 20 fig total RNA from bovine cerebral cortex (C), liver (L), neurointermediate pituitary (N), and anterior pituitary (A) were fractionated on an agarose-formaldehyde gel and hybridized w/ith the nick-translated 2.2-kb fragment (A) or 0.7-kb fragment (B) of XPAM-1. Similar results were obtained with the 0.4-kb fragment from XPAM-1. In C the blot shown in Panel A was stripped and hybridized with the pro-ACTH/endorphin cDNA probe. The positions of the molecular weight markers are indicated at the left.

Studies of regulatory processes affecting neuronal and endocrine cells producing a-amldated peptides can now be expanded to include effects on a processing enzyme as well as effects on synthesis and secretion of the

peptides themselves. At least in corticotropes, levels of PAM mRNA are subject to regulation by secreta-gogues. Since it can be estimated that a granule con-tains only a few molecules of PAM, regulation of PAM


c

o

o

u . X

c r a >

o o

PAM

ase E (39), the precursor to PAM contains many pairs of basic amino acids; endoproteolytic processing may play a role in the tissue-specific maturation of posttrans-iational processing enzymes as well as their peptide substrates. The presence of a membrane-spanning do-main in the COOH-terminal region of the precursor to PAM could allow expression of functional enzyme on the cell surface. The hydroxyl-rich hydrophilic cyto-plasmic domain with several potential phosphorylation sites is similar to the COOH-terminal cytoplasmic do-mains of rhodopsin and the (8-adrenergic receptor (38). The fact that precursors to a series of growth factors related to EGF have COOH terminally located mem-brane spanning domains and can be expressed on the cell surface (40, 41) raises the intriguing possibility that enzymes traditionally regarded as functioning within secretory granules may also play a role in cell surface-mediated interactions.

M A T E R I A L S AND M E T H O D S

Sequence Analysis of PAM Peptides

PAE

Fig. 6. Parallel Regulation of PAM and pro-ACTH/Endorphin mRNAs in Corticotropic Tumor Cells

AtT-20 cells were grown in regular growth medium or pre-treated in medium containing 1 ^m dexamethasone for 72 h. Cells were then exposed for 5 h to cgntrol medium or medium with 1 nM dexamethasone or 100 nM CRF. Secretory rates in the various treatments (percent cell hormone content secreted per hour) were as follows: control, 3.5; CRF, 9.9; dexameth-asone, 2.0. Total RNA (10 ^g/lane) was analyzed as in Fig. 5 and probed with the 0.7-kb fragment of XPAM-1 (1.0 x 10® cpm//ug) and with the pro-ACTH/endorphin cDNA probe (0.2 X 10® cpm/^g) with the amounts of the two probes adjusted to give comparable intensities for the pro-ACTH/endorphin and PAM mRNAs. Similar results were obtained in two other experiments.

expression may regulate production of bioactive a-amidated peptides. The structure of the predicted pre-cursor to PAM has important implications for the role of this enzyme. Like the precursor to another granule-associated peptide-processing enzyme, carboxypeptid-

PAM was purified from 400 bovine neurointermediate pituitar-ies as described (2). Fractions containing purified PAM-B (typ-ically 20 to 70 Mg) were dried under vacuum and dissolved in 0.2 ml 70% formic acid for cleavage with cyanogen bromide. Cyanogen bromide (10 mg) was added and the samples were incubated at room temperature for 12 h in the dark; a second addition of cyanogen bromide (4 mg) was made and the peptides were dried under vacuum after 12 h. Peptides were dissolved in 100 n\ 0.5 M Tris HCI, pH 8.5, 6 M guanidine HCI, 10 mM dithiothreitol, and incubated at 37 C for 2 h; samples were chilled, made 20 mw in iodoacetamide, and incubated 2 h on ice in the dark. In some analyses, the cyanogen bromide fragments were alkylated first with 10 mM [2-^H]iodoacetic acid (Amersham, Arlington Heights, IL; 144 mCi/mmol) for 2 h followed by reaction with 10 mM unlabeled iodoacetic acid for 1 h. After reduction and alkylation, the peptides were fraction-ated on a column of Sephadex G-50 Superfine (38 x 1 cm), equilibrated and eluted with 10% formic acid. The size frac-tionated cyanogen bromide peptides were pooled, dried under vacuum, and purified on a Bondapak CI 8 column (7.5 x 300 mm; Waters Associates, Bedford, MA) equilibrated with 0.1% trifluoroacetic acid and eluted with a series of linear gradients to 80% acetonitrile in 0.1% trifluoroacetic acid. Peptides were detected by absorbance at 220 nm. Edman degradations were carried out using the Applied Biosystems (Foster City, CA) model 470A/120A gas phase sequencer (42, 43). To calibrate recoveries, an aliquot was hydroiyzed in 6 N HCI (24 h, 110 C) and subjected to amino acid analysis on the Beckman (Fuller-ton, CA) model 6300 analyzer. Oligonucleotide probes corre-sponding to preliminary sequences of three cyanogen bromide peptides were synthesized on a BioSearch (San Rafael, CA) model 8600 DNA synthesizer using the criteria of Lathe (44) to select codons.

Construction and Screening of cDNA Libraries

PAM antiserum 36 was generated by subcutaneous injection of a rabbit with 9 ^g PAM-B (2) dissolved in 0.9% NaCI, 0.75% SDS by boiling, and emulsified with complete Freund's adju-vant; the rabbit was boosted once with a similar sample of PAM-B and twice with a mixture of PAM-A plus PAM-B. For affinity purification, the protein in a pool of purified PAM-A plus PAM-B from 400 bovine neurointermediate pituitaries was linked to 350 mg activated CH-Sepharose 4B (Pharmacia, Piscataway, NJ). Antiserum was affinity purified as described

MOL END0.1987 788

Vol1 No. 11

(45). The basic procedures of Snyder et al. (21) were followed In preparing and screening the cDNA library. Total RNA was prepared from bovine intermediate pituitaries (46); 5 ng poly(A)" RNA were used for synthesis of cDNA using the Amersham cDNA synthesis system (47). Complementary DNAs larger than 500 bp were selected by gel filtration, ligated to 1 lg Xgtl l arms (Promega Biotec, Madison, Wl), and packaged using the Gigapack Gold in vitro X DNA packaging kit (Stratagene, San Diego, CA). A total of 2.3 x 10® recom-binant phage were obtained; of the total phage, 11% were nonrecombinant. For the initial screening with affinity purified PAM antiserum, six 150-mm dishes were plated at a density of 63,000 recombinant phage per dish. To localize positive phage, isopropyl-|8-D-thiogalactopyranoside-coated nitrocellu-lose filters (Schleicher & Schuell, Keene, NH) were applied for 8 h, washed for 3 x 10 min at room temperature in TBS (50 mM Tris HCK 150 mw NaCI, pH 7.5) and blocked for 1 h at room temperature in TBS containing 0.1 mg/ml BSA (TBS/ BSA) and 0.1% Tween-20. Primary antibody was diluted in TBS/BSA plus Tween-20 and allowed to incubate overnight at 4 C or for 4 h at room temperature. Filters were then washed 3 x 10 min at room temperature in TBS/BSA plus Tween-20 and [ ^®l]labeled Protein A (10® cpm/ml) was added in TBS/BSA for 2-4 h at room temperature. Filters were washed for 10 min at room temperature in TBS/BSA, for 2 x 10 min in TBS/BSA containing 0.1 % NP-40, for 10 min in TBS/ BSA before visualization by autoradiography at - 70 C with an intensifying screen. In the initial screen with affinity purified PAM antiserum 36,18 positive plaques were identified. Upon rescreening, only six remained positive; the four phage giving the strongest signals (XPAM-1, 2, 3, and 5) were plaque purified and further characterized. In the second round of screening the library, nick translated 0.4 kb and 0.7 kb XPAM-1 cDNA were used to screen replica nitrocellulose lifts pre-pared from four 150-mm plates (90,000 recombinant phage/ plate); plaques hybridizing with either or both cDNA probe were identified and a subset were plaque purified, character-ized by Southern blot analysis, and a likely full length phage (XPAM-6) was further characterized.

Cloning and cDNA Sequence Analysis

The EcoRI fragments of XPAM-1, -5, and -6 were subcloned into Bluescript (Stratagene) and sequenced by the dideoxy-chain termination method (18-20). Single stranded sequencing was performed as described by Stratagene. Double stranded sequencing was performed as described by New England Biolabs (Beverly, MA). To sequence the 2.2-kb EcoRI fragment of XPAM-1, nested sets of deletions extending inwards from the 5'- and 3'-termini of the cDNA were generated with an £xolll/Mung Bean Nuclease kit from Stratagene. To sequence the 0.8-kb EcoRI fragment of XPAM-6, deletions with Nde\ and Nar\ were prepared. The EcoRI site separating the 0.8- and 0.7-kb fragments is in peptide HR6, and both EcoRI fragments hybridized with 0LIG06. The sequence across the EcoRI site separating the 0.7- and 2.2-kb fragments was obtained by subcloning the appropriate Ssp\-Xbal fragment and using a synthetic oligonucleotide primer. Synthetic oligonucleotide primers or deletions using unique restriction sites were used to direct sequencing to a particular region of the molecule.

Fusion Protein Antibodies

from the wild type and XPAM-1 samples were analyzed on slab gels and visualized with Coomassie brilliant blue. The fractions from the XPAM-1 sample containing proteins ranging in size from myosin to ;8-galactosidase were pooled and used for immunization of rabbits. Three female New Zealand white rabbits (2 kg) were immunized with one-sixth of the fusion protein (0.5 A280 U each) from a 50-ml culture emulsified in Freund's complete adjuvant. Rabbits were given five booster injections of the same amount of fusion protein in incomplete Freund's adjuvant at intervals of 2-4 weeks.

The ability of various antisera to recognize PAM was as-sayed by cross-linking the immunoglobulin fraction to Protein A-Sepharose (Pharmacia, Piscataway, N J ) as described (16). In addition to assaying the ability of the antibody resin to remove PAM activity from solution (16), the appearance of resin-bound PAM activity was determined. Resin prepared from Ab 46 had a significant amount of PAM activity (0.15 pmol/h for 10 ix\ packed resin); this PAM activity is thought to be due to circulating (15) rabbit PAM bound to the immuno-globulin. In order to reduce this background PAM activity, immunoglobulins from rabbit Ab 46 were precipitated with ammonium sulfate, dialyzed into 3 M K S C N , 0.1 M Tris HCI, pH 7.2, and fractionated on a Sephacryl S-300 column (1 x 50 cm) in the same buffer. The fractions containing proteins larger than 80,000 daltons were dialyzed into 50 mw NaTES, pH 7.4, and linked to protein A-Sepharose (16). After this treatment, resin prepared from Ab 46 contained less than 1% as much PAM activity as before the treatment. The preimmune serum and Ab 36 were not KSCN-treated before linking to protein A-Sepharose; the Ab 36 resin exhibited PAM activity.

Northern Blot Analyses of Various Bovine Tissues

Total RNA was prepared by the method of Chirgwin et ai. (46). Duplicate aliquots of total RNA were fractionated on a 1% agarose gel containing 2.2 M formaldehyde, transferred to Nytran (Schleicher & Schuell), and hybridized with the nick-translated probes. Blots were prehybridized for 2-4 h at 42 C in 5x SSC (20x SSC is 3.0 M NaCI, 0.3 M Na citrate, pH 7.0), 4x Denhardt's (50x Denhardt's is 1% Ficoll, 1% polyvinylpyr-rolidone, 1 % BSA in water), 50% deionized formamide, 20 mM Na phosphate, pH 6.8, 0.1% SDS, 0.1 mg/ml sonicated carrier herring sperm DNA. Hybridization was carried out by adding boiled nick-translated probe (-10® cpm/ml; SA, 0.5 to 2 x 10® cpm//xg DNA) and continuing the incubation at 42 C for 16-24 h. Filters were washed 2 x 30 min in 2x SSC, 0.1% SDS at room temperature, and 2 x 30 min in 0.1 x SSC, 0,1% SDS at 50 C before autoradiography at - 70 C with an intensifying screen. Probes were stripped off by incubation in 50% form-amide, 1X SSC at 70 C for 60 min, and the blot was reprobed with a second probe. Sample application position is marked by a bar and the standards are shown at the left. For molecular weight determination, the 18 S and 28 S ribosomal RNAs were visualized with acridine orange and an RNA ladder (Be-thesda Research Laboratories, Gaithersburg, MD) fractionated in an adjacent lane was visualized by hybridization to nick-translated wild type XDNA.

Determination of pro-ACTH/Endorphin-Related mRNA and Peptides

For production of fusion protein antibody, cell pellets from 50 ml cultures of Escherichia coli strain CAG456 cells infected with wild type Xgtl 1 or XPAM-1 were lysed by freeze-thawing and sonication in 1.5 ml 100 mM Na TES, pH 7.4, 10 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, and centrifuged at 10,000 X g for 10 min; very little fusion protein appeared in the soluble fraction. The cell pellets were dissolved by boiling in 2.0 ml SDS sample buffer (17) and were individually applied to a column of Sephacryl S-300 SF (1 x 53 cm) equilibrated and eluted with 2x running gel buffer. Aliquots of fractions

AtT-20/D-16v mouse corticotropic tumor cells were grown as described (14). Parallel wells were harvested in guanidine thiocyanate for extraction of RNA (46) or in 50% acetic acid for RIA of pro-ACTH/endorphin-derived peptides (48). Secre-tory rates (% cell hormone content secreted/hour) were deter-mined by RIA of 16 K fragment in spent medium and cell extracts (48). Equal aliquots of total RNA were fractionated on formaldehyde agarose gels as described above. Levels of pro-ACTH/endorphin mRNA were determined using nick-trans-lated pMKSU probe (49).


Acknowledgments

We thank Drs. Randy Reed, Phil Heiter, Keith Peden, and Hugh Niall for their help and encouragement, and Dr. Wun-Jing Kuang for DNA sequencing.

Received August 3,1987. Accepted September 8,1987. Address requests for reprints to: Dr. Betty A. Eipper, De-

partment of Neuroscience, The Johns Hopl<ins University School of Medicine, 725 North Wolfe Street, Baltimore, IVIary-land 21205.

Supported by NIH Grants DK-32949 and DK-32948 and National Institute on Drug Abuse Grants DA-00097 and DA-00098.

REFERENCES

1. Murthy ASN, Keutmann HT, Eipper BA 1987 Further characterization of peptidylglycine a-amidating monoox-ygenase from bovine neurointermediate pituitary. Mol En-docrinol 1:290-299

2. Murthy ASN, Mains RE, Eipper BA 1986 Purification and characterization of peptidylglycine a-amidating monoox-ygenase from bovine neurointermediate pituitary. J. Biol Chem 261:1815-1822

3. Eipper BA, Mains RE, Glembotski CC 1983 Identification in pituitary tissue of a peptide a-amidation activity that acts on giycine-extended peptides and requires molecular oxygen, copper, and ascorbic acid. Proc Nat! Acad Sci USA 80:5144-5148

4. Tatemoto K 1983 Isolation and characterization of novel brain-gut peptides. Biomedical Res 4:[Suppl1]1-6

5. Tatemoto K, Efendic S, Mutt V, Makk G, Feistner GJ, Barchas JD 1986 Pancreastatin, a novel pancreatic pep-tide that inhibits insulin secretion. Nature 324:476-478

6. Bradbury AF, Finnie MDA, Smyth DG 1982 Mechanism of C-terminal amide formation by pituitary enzymes. Na-ture 298:686-688

7. Bradbury AF, Smyth DG 1983 Substrate specificity of an amidating enzyme in porcine pituitary. Biochem Biophys Res Commun 112:372-377

8. Bradbury AF, Smyth DG 1985 C-Terminal amide formation in peptide hormones. In: Hakanson R, Thorell J (eds) Biogenetics of Neurohormonal Peptides. Academic Press, New York, pp 171-186

9. Mizuno K, Sakata J, Kojima M, Kangawa K, Matsuo H 1986 Peptide C-terminal a-amidating enzyme purified to homogeneity from Xenopus laevis skin. Biochem Biophys Res Commun 137:984-991

10. Sakata J, Mizuno K, Matsuo H 1986 Tissue distribution and characterization of peptide C-terminal a-amidating activity in rat. Biochem Biophys Res Commun 140:230-236

11. Mollay C, Wichta J, Kreil C 1986 Detection and partial characterization of an amidating enzyme in skin secretion of Xenopus laevis. FEBS Lett 202:251-254

12. Kizer JS, Bateman RC Jr, Miller CR, Humm J, Busby WH Jr, Youngblood WW 1986 Purification and characteriza-tion of a peptidyl glycine monooxygenase from porcine pituitary. Endocrinology 118:2262-2267

13. Mains RE, Glembotski CC, Eipper BA 1984 Peptide a-amidation activity in mouse anterior pituitary AtT-20 cell granules; properties and secretion. Endocrinology 114:1522-1530

14. Mains RE, Eipper BA 1984 Secretion and regulation of two biosynthetic enzyme activities, peptidyl-glycine a-amidating monooxygenase and a carboxypeptidase, by mouse pituitary corticotropic tumor cells. Endocrinology 115:1683-1690

15. Mains RE, Myers AC, Eipper BA 1985 Hormonal, drug, and dietary factors affecting peptidyl glycine a-amidating monooxygenase activity in various tissues of the adult

male rat. Endocrinology 116:2505-2515 16. Park LP, Thiele EA, Dickerson IM, Mains RE, Eipper BA

Cloning of cDNA encoding bovine peptidyl-glycine a-ami-dating monooxygenase (PAM). In: Christiansen C, Riis BJ (eds) Highlights on Endocrinology. Norhaven Bogtrykkeri, Copenhagen, pp 130-140

17. Laemmli UK 1970 Cleavage of structural proteins during assembly of the head of bacteriophage T4. Nature 227:680-685

18. Sanger F, Nicklen S, Coulson AR 1977 DNA sequencing with chain-terminating inhibitors. Proc Natl Acad Sci USA 74:5463-5467

19. Biggin MD, Gibson TJ, Hong GF 1983 Buffer gradient gels and label as an aid to rapid DNA sequence determination. Proc Natl Acad Sci USA 80:3963-3965

20. Chen EY, Seeburg PH 1985 Supercoil sequencing: a fast and simple method for sequencing plasmid DNA. DNA 4:165-170

21. Snyder M, Sweetser D, Young RA, Davis RW, Xgt l l : gene isolation with antibody probes and other applica-tions. Methods Enzymol, in press

22. Davis LG, Dibner MD, Battey JF 1986 Basic Methods in Molecular Biology. Elsevier Science Publishing Co, New York

23. Kozak M1986 Point mutations define a sequence flanking the AUG initiator codon that modulates translation by eukaryotic ribosomes. Cell 44:283-292

24. Padgett RA, Grabowski PJ, Konarska MM, Seiler S, Sharp PA 1986 Splicing of messenger RNA precursors. Annu Rev Biochem 55:1119-1150

25. Proudfoot NJ, Brownlee GG 1976 3' Non-coding region sequences in eukaryotic messenger RNA. Nature 263:211-214

26. Fitzgerald MD, Shenk T1981 The sequence 5'-AAUAAA-3' forms part of the recognition site for polyadenylation of late SV40 mRNAs. Cell 24:251-260

27. Hafen E, Easier K, Edstroem JE, Rubin GM 1987 Seven-less, a cell-specific homeotic gene of Drosophila, encodes a putative transmembrane receptor with a tyrosine kinase domain. Science 236:55-63

28. von Heijne G 1985 Signal sequences: the limits of varia-tion. J Mol Biol 184:99-105

29. Walter P, Lingappa VR 1986 Mechanism of protein trans-location across the endoplasmic reticulum membrane. Annu Rev Cell Biol 2:499-516

30. Sigel H (ed) 1981 Metal Ions in Biological Systems. Marcel Dekker, Inc, New York, vol 13

31. Kornfeld R, Kornfeld S 1985 Assembly of asparagine-linked oligosaccharides. Annu Rev Biochem 54:631-664

32. Kyte J, Doolittle RF 1982 A simple method for displaying the hydropathic character of a protein. J Mol Biol 157:105-132

33. von Heijne G 1981 Membrane proteins: the amino acid composition of membrane-penetrating segments. Eur J Biochem 120:275-278

34. Feramisco JR, Glass DB, Krebs EG 1980 Optimal spatial requirements for the location of basic residues in peptide substrates for the cyclic AMP-dependent protein kinase. J Biol Chem 255:4240-4245

35. Wang Y, Bell AW, Hermodson MA, Roach PJ 1986 Liver isozyme of rabbit glycogen synthase. J Biol Chem 261:16909-16915

36. House C, Wettenhall RE, Kemp BE 1987 The influence of basic residues on the substrate specificity of protein ki-nase C. J Biol Chem 262:772-777

37. Bengur AR, Robinson EA, Appella E, Sellers JR 1987 Sequence of the sites phosphorylated by protein kinase C in the smooth muscle myosin light chain. J Biol Chem 262:7613-7617

38. Hall ZW 1987 Three of a kind: the /3-adrenergic receptor, the muscarinic acetylcholine receptor, and rhodopsin. Trends Neurosci 10:99-101

39. Fricker LD, Evans CJ, Esch FS, Herbert E 1986 Cloning and sequence analysis of cDNA for bovine carboxypeptid-

MOL ENDO-1987 790

Vol 1 No. 11

ase E. Nature 323:461-464 40. Teixido J, Gilmore R, Lee DC, Massague J 1987 Integral

membrane glycoprotein properties of the prohormone pro-transforming growth factor-a. Nature 326:883-885

41. Bell Gl, Fong NM, Stempien MM, Wormsted MA, Caput D, Ku L, Urdea MS, Rail LB, Sanchez-Pescador R 1986 Human epidermal growth factor precursor: cDNA se-quence, expression in vitro and gene organization. Nucleic Acids Res 14:8427-8446

42. Hewick RM, Hunkapiiler MW, Hood LE, Dryer WJ 1981 A gas-liquid solid phase peptide and protein sequenator. J Biol Chem 256:7990-7997

43 Rodriquez H, Kohr WJ, Harkins RM 1984 Design and operation of a completely automated Beckman microse-quencer. Anal Biochem 140:538-547

44. Lathe R 1985 Synthetic oligonucleotide probes deduced from amino acids sequence data: theoretical and practical considerations. J Mol Biol 183:1-12

45. Wand GS, Eipper BA1987 Effect of chronic secretagogue exposure on pro-ACTH/endorphin production and secre-tion in primary cultures of rat anterior pituitary. Endocri-nology 120:953-961

46. Chirgwin JM, Przybia AE, MacDonald RJ, Rutter WJ 1979 isolation of biologically active ribonucleic acid from sources enriched in ribonuclease. Biochemistry 18:5294-5299

47. Gubler V, Hoffman BJ 1983 A simple and very efficient method for generating cDNA libraries. Gene 25:263-269

48. Eipper BA, Mains RE 1978 Existence of a common pre-cursor to ACTH and endorphin in the anterior and inter-mediate lobes of the rat pituitary. J Supramo! Struct 8:247-262

49. Uhler M, Herbert E 1983 Complete amino acid sequence of mouse proopiomelanocortin derived from the nucleo-tide sequence of proopiomelanocortin cDNA. J Biol Chem 258:257-261

NATURE v o l . 5 NOVEMBER 1987 SCIENTIRC CORRESPONDENCE

Glycine vs GABA receptors SIR—Compar ison of the primary struc-

tures of the strychnine-binding subunit of

the glycine receptor' and both subunits of

the G A B A ^ receptor" has revealed the

existence of a gene family for neurotrans-

mitter-gated ion channel receptors which

comprises both anionic and cationic chan-

rec-

' - ' " we suggest that this conserved

will be characteristic of all

221 216 218

271 ISrf 266 THL^ 268 S6J

RCIMTY

FSTLIT $LL1V

Ml—

ILSOV ILEIVV i i i ^ I

SFWjf SFV flf SFWIf I^ESVP

YDASfl iDAAP

RNSLPKV/!VA7/ ETLPKIFY^/ ASLPK \/SYVK/

flwFIAV lYLMG IWMAV <

Y/*=VF FV-VF LL-VF

lET^T

Al ignment of transmembrane domains M l to M3 of ihe G A B A and

glycine receptor polypcpiidcs. Identical amino-acid residues are boxed,

nels including the nicotinic acetylcholine information

receptor ( n A C h R ) and presumably those

gated by amino acids involved in excita-

tory neurotransmission, such as gluta-

mate. Sequence comparison shows that

the subunits of all known members of this

gene family have the same number and

distribution of membrane-spanning hel-

ices, resulting in an identical transmem-

brane topology' The first transmem-

brane region predicted. M l , contains a

proline residue at an identical position

point ing to a common mechanism of

channel gating'

In addi t ion, the subunits of the glycine

and G A B A . , reccptor display some striking

structural similarities. First, iheir sequen-

ces are considerably more similar to each

other (347o-387o) than thev are to the

subunits of n A C h R s {i57o-2()7o).

Second, the G A B A , and glycine rec-

eptors arc most similar within the first

three membrane-spanning regions {see

figure): eight coniiguinis aniino-acit!

residues ( T T V L T M I T ) are invariant

within M2. As segment M2 is thought to

line the channel lumen of the different

ion-conducting neurotransmitter

eptors

sequence

ligand-gated chloride channels.

Third , a large number of positively-

charged amino-acid residues occurs in the

putative chloride channel-mouth domains

of both glycine and G A B A ^ receptor

polypeptides. These clustered charges

have been implicated in anion binding and

channel selectivity' ^ and may constitute a

general feature of ligand-gated anion

channels.

Finally, chloride-conducting neuro-

transmitter receptor proteins have no

amphipathic a-helix preceding the fourth

hydrophobic region, M4. Such an

amphipathic helix is found in all n A C h R

subunits, and was originally proposed to

form the inner wall of the cation channel'.

The above and further comparisons within

this receptor family have been made in

more detail in an article appearing else-

where".

These conserved features indicate a late

separation of G A B A ^ and glycine rec-

eptors during ligand-gated ion-channel

evolution. In the case of these inhibitory

channels, the acquisition of ion selectivity

probably preceded that of agonist

specificity. However, domain exchanges

by e.xon shuffling may also have con-

tributed to the genera-

tion of the present ion

channel receptor

family, since cation-

conducting G A B A

receptors and anion-

selective n A C h R s have

been described in dif-

ferent systems". Seq-

uence and structural

these latter receptors

should contribute to our understanding of

how diversity was generated within this

important super-family.

G A B R I E L E G R E N N I N G L O H

E C K A R T G U N D E L F I N G E R

B E R T R A M SCHMITT

HELNRICH B E T Z

Molecular Neuropharmacology Lab,

ZMBH, University of Heidelberg,

6900 Heidelberg, FRG

M A R K G . D A R L I S O N

E R I C A . B A R N A R D

MRC Molecular Neurobiology Unit, MRC

Centre,

Cambridge CB2 2QH, UK

PETER R . SCHOF IE I .D

PETER H . SEEBURC.

Laboratory of Molecular

Nei I roe 11 docrinology

ZMBH, University of Heidelberg,

ARTVF3 ARVAL3 ^ G L g <

/TTVLTMTT ITTVLTMTT TTVLTMTT

M 2 —

GABAR a GABAR 6 6LYR ijSK

6900 Heidelberg. FRG

C.tinnini;l»li (i n u! ,V«mri-:i i ^riMI'W) Siiu-iici.i !•./,</ .i2«.::i ::(.(i')s7i

SCIENTinC CORRESPONDENCE N A I U R f - : V O L .Ul l > 1 M l i l l<

a ( .ir.iud.ii I n III I'm, niifii .•\( ii</ Vi i < \

5 O K r i h u r W . ( I . M H O J 5. I S l 5 IS ( | W > ) (. Inu'I i . K I I III V,/(iiri-.U4. (.711-674 (m.SM ^ PiiK-rMiHirc. J S ir i 'u i l . R M I'nn niiiii Aiiul Sci

I S A 81. l .-S I ^ V d W J l > U.irii;iri.l t A . lJ.irliM>n. M Ci & Sccburg. P . I I Trends

\,i,r,,u, 10. • i i i : -5 (N ( lw7 ) Hillc H Ikhii (. hunnch of liuiiuhle Mi-nihranrs (S inuucr Sundcrl jud MassiithusclN. 0 X 4 )

Vol. 149 , No. 1, 1 9 8 7 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

November 30 , 1 9 8 7 Pages 1 2 5 - 1 3 2

A NOVEL SUBTYPE OF MUSCARINIC RECEPTOR IDENTIFIED BY HOMOLOGY SCREENING

Thomas Braun, Peter R. Schofield, Brenda D. Shivers, Dolan B. Pritchett and Peter H. Seeburg*

Laboratory of Molecular Neurobiology, ZMBH, University of Heidelberg,

Im Neuenhelmer Feld 282, 6900 Heidelberg, F.R.G.

R e c e i v e d O c t o b e r 1 3 , 1987

A new member of the protein superfamily of G-protein coupled receptors has been isolated by homology screening. By virtue of its homology with other muscarinic acetylcholine receptors and its ability to bind muscarinic specific antagonists, this muscarinic receptor subtype is designated M4. The M4 mRNA is preferentially expressed in certain brain regions. The existence of multiple receptor subtypes encoded by distinct genes in the brain has functional implications for the molecular mechanisms underlying information transmission in neuronal networks. © 1987 Academic Press, Inc.

Acetylcholine plays a key role in the neuronal mechanisms underlying

memory, learning, arousal and control of movement. The majority of

cholinergic synapses in the vertebrate central nervous system are

muscarinic. Muscarinic acetylcholine receptors (mAChR) are widely

distributed, being present in neurons of the central and peripheral nervous

systems, in cardiac and smooth muscle tissue and in exocrine glands. Their

activation results in a variety of G-protein mediated events including the

inhibition of adenylate cyclase, activation of phosphoinositide turnover,

stimulation of cGMP synthesis and regulation of potassium channels (for a

review, see ref. 1).

Existence of at least two different subtypes of the mAChR was predicted

on the basis of differential binding affinities for selective muscarinic

receptor antagonists (2-4). Recently, cDNAs for mAChRs have been isolated

from either porcine cerebral cortex or heart and were shown by functional

expression to encode two different subtypes Ml and M2, respectively (5-7).

Pharmacological evidence supports the possibility that additional mAChR

*To whom correspondence should be addressed.

0006-291X/87 $1.50 Copyright © 1987 by Academic Press, Inc.

^^^ All rights of reproduction in any form reserved.


subtypes contribute to the observed diversity in mAChR responses (8). Such

other mAChR subtypes may share structural homologies with the Ml and M2

subtypes, thus providing a basis for the isolation of cDNAs encoding

receptor subtypes by screening for homologous clones.

Sequence comparisons of the two mAChR subtypes Ml and M2 and of other G-

protein coupled receptors (e.g. the opsins and p-adrenergic receptor) show

that all G-protein coupled receptors possess a similar tertiary structure

(5, 9). Most characteristically, they all possess seven hydrophobic

membrane spanning regions. For each of these receptors, a particularly well

conserved sequence motif occurs between the transmembrane regions (TM) 1

and 2. We have utilized this intrinsic homology for screening a cDNA

library derived from rat forebrain mRNA and have isolated a novel subtype

of the mAChR.

MATERIALS AND METHODS

A AgtlO rat forebrain cDNA library (10) was screened with two 50-mer oligonucleotide pools RSpi (5' CTGCAGACRGTCACCAAMTACTTYATCACC TCCYTGGCCTGTGCTGATCT 3') (lUPAC-IUB ambiguous base code) and RSM2 (5' CTCCAGACRGTCAACAAYTACTTYYTGTTCAGCYTGGCCTGTGCTGACCT 3') and rescreened with shorter more degenerate pools RS3a-f (5' WSCYTVGCYT6YGCYGAYYT 3') and RS4a-f (5' CARACWGTBAMYAAYTAYTTYMT 3'). Filters were washed in IxSSC or 5xSSC at 42®C for both 50-mers and for the RS4a-f pool, respectively and in 3M TMACl (11) at 50°C for the RS3a-f pool. DNA sequences were obtained by the chain-termination method using M13 vectors (12).

Cellular expression data were obtained using the host vector system and transformation procedures described by Eaton et al. (13). Membrane fractions of transfected cells were assayed for ^H-QNB (quinuclidinylbenzilate) binding as described (7). In situ hybridization with ^^p.iabeiied cRNA prepared using T7 RNA polymerase utilized a subtype-specific (TM5-6) 358bp StuI-PstI DNA fragment (nt 936 - 1294) subcloned into vector pGEMII (Promega). Hybridization was performed as described (14). Following RNase digestion the sections were washed 2 times in 2xSSC, 0.05% inorganic sodium pyrophosphate for 10 minutes each at room temperature, followed by 2 washes at 50°C in O.lxSSC, 0.05% inorganic sodi um pyrophosphate for 15 minutes each prior to film autoradiography.

RESULTS AND DISCUSSION

The superfamily of G-protein coupled receptors (opsins, p-adrenergic and

muscarinic receptors) contain a highly conserved sequence between TM 1 and

2. The consensus amino acid sequence for this region is L-Q-T-V-N/T-N-Y-F-

L/I-L/T-S-L-A-C-A-D-L. We used two degenerate 50-mer oligonucleotides (RSpl

and RSM2) spanning this entire region to screen 2x10® clones of a rat

forebrain cDNA library. Two cDNA clones with a 3.8kb insert were obtained.

These clones also hybridized to the two other highly degenerate

oligonucleotide pools (RS3a-f and RS4a-f) which are based on either the 5'

or 3' sequences of the longer probes. DNA sequence analysis (Fig. 1)

126

Vol. 1 4 9 , No. 1, 1 9 8 7 BIOCHEMICAL A N D BIOPHYSICAL RESEARCH C O M M U N I C A T I O N S

1 CTTTGCTGCATTCTGACTAGTGGCCAAGCACGTGACATCCCGAACTTCTGCAAAGGTACAATAAGGGCCGACCATTTAATTTTGGATAACCAGTTGGTGTGTTCTTCCTTGGACTATGT

120 CAGAGAGTCACAATGACCTTGCACAGTAACAGTACAACCTCGCCTTTGTTTCCCAACATCAGCTCTTCCTGGGTGCACAGTCCCTCGGAGGCAGGGCTGCCCTTGGGGACAGTCACTCAG 1 ML-tThrLeuHisSerAsnSerThrThrSerProLeuPheProAsnIleSerSerSerTrpValHisSerProSerGluAlaGlvLeuProLeuGlvThrVa1ThrRln

240 TTGGGCAGCTACAACATriCACAAGAAACTGGGAATTTCTCCTCAAACGACACCTCCAGCGACCCTCTCGGGGGrCACACCATCTGGCAAGTGGTCTTCATTGCCTTCTTAACCGGCTTC 37 LeuGIySerTyrAsnlleSerGlnGIuThrGlyAsnPheSerSerAsnAsDThrSerSerAspPrQLeuGIyGlvHisThrlleTrpGlnValValPhelleAlaPheLeuThrGJvPhe

' ' ' ' ' ' 1 360 CTGbCAITGGTGACCATCATTGGCAACATCCTTGTCATTGTGGCCTTCAAGGTCAACAAACAGCTGAAGACAGTCAACAACTACTTCCTCTTAAGCCTGGCCTGTGCAGACCTGATCATC n LeuAlaLeuValThrllelJeGlyAsnlleLRuVallleValAlaPheLysValAsnLysGlnLeuLysThrValAsnAsnTyrPheLeuLeuSerLeuAJaCysAlaAspLeuIlelle

— 2 - — 480 GGGGTCATTTCCATGAACCTGTTCACTACCTACATCATTATGAACCGTTGGGCACTGGGGAACTTAGCCTGCGACCTCTGGCTCTCCATTGACTATGTGGCCAGCAATGCCTCTGTCATG

GlyVallleSerMetAsnLeuPheThrThrTyrllelleNetAsnArgTrpAIaLeuGlyAsnLeuAlaCysAspLeuTrpLeuSerlleAspTyrValAlaSerAsnAlaSerValMet 3

600 AATCTGCTGGTCATCAGCTITGACAGGTACTTTTCCATCACTAGGCCACTCACCTACCGAGCCAAAAGAACAACAAAACGAGCTGGTGTGATGATTGGTCTGGCTTGGGTCATCTCCTTT 157 AsnLeuLeuVallleSerPheAspArgTyrPheSerlleThrArgProLeuThrTyrArgAlaLysArgThrThrLysArgAIaGIyValHetlleGlyLeuAlaTrpVallleSerPhe

— 4 720 GTCCTATGGGCTCCTGCCATCTTGTTCTGGCAATACTTTGTAGGGAAGAGAACTGTGCCCCCAGGAGAATGTTTCATTCAGTTTCTGAGTGAGCCCACCATCACCTTCGGCACGGCGATC 197 ValLeuTrpAlaProAlalleLeuPheTrpGInTyrPheValGlyLysArgThrValProProGlyGluCysPhelleGlnPheLeuSerGIuProThrlleThrPheGlyThrAlalie

5 — 840 GCTGCCTTTTACATGCCTGTCACCATCATGACTATTTTATACTGGAGGATCTATAAGGAAACTGAGAAGCGTACCAAAGAGCTGGCTGGCCTACAGGCCTCTGGGACAGAAGCGGAGGCA 237 AlaAlaPheTyrMetProValThrlleMetThrneLeuTyrTrpArglleTyrLysGluThrGluLysArgThrLysGluLeuAlaGlyLeuGlnAlaSerGlyThrGIuAIaGIuAIa

960 GAAAACTTTGTCCACCCCACAGGCAGTTCTCGAAGCTGTAGCAGCTATGAACTGCAACAGCAAGGTGTGAAACGATCATCCAGGAGGAAGTACGGTCGCTGTCACTTCTGGTTCACCACC 277 GluAsnPheValHisProThrGlySerSerArgSerCysSerSerTyrGluLeuGlnG]nGlnGlyVaUysArgSerSerArgArgLysTyrGlyArgCysHisPheTrpPheThrThr

1080 AAGAGCTGGAAGCCCAGTGCCGAGCAGATGGACCAAGACCACAGCAGCAGCGACAGTTGGAACAACAACGATGCTGCTGCCTCCCTGGAAAACTCTGCTTCCTCCGATGAAGAGGACATT 317 LysSerTrpLysProSerAlaGluGlnMetAspGlnAspHisSerSerSerAspSerTrpAsnAsnAsnAspAlaAlaAlaSerLeuGluAsnSerAlaSerSerAspGluGluAspIle

1200 GGCTCAGAGACCAGGGCCATCTATTCCATTGTCCTCAAGCTTCCAGGCCATAGCTCCATCCTCAACTCTACCAAGCTACCATCCTCAGATAACCTGCAGGTGTCCAACGAGGACCTGGGG 357 GlySerGluThrArgAlalleTyrSerlleValLeuLysLeuProGlyHisSerSerlleLeuAsnSerThrLysLeuProSerSerAspAsnLeuGlnValSerAsnGluAspLeuGly

1320 ACTGTGGATGTGGAGAGAAATGCTCACAAGCTTCAGGCCCAGAAGAGCATGGGTGATGGTGACAACTGTCAGAAGGATTTCACCAAGCTTCCCATCCAGTTAGAGTCTGCTGTGGACACA 397 ThrValAspValGluArqAsnAlaHisLysLeuGlnAlaGlnLysSerMetGlyAspGlyAspAsnCysGlnLysAspPheThrLysLeuPrpIleGlnLeuGluSerAlaValAspThr

1440 GGCAAGACCTCTGACACCAACTCCTCGGCAGACAAGACCACGGCTACTCTACCTCTGTCCTTCAAGGAGGCCACGCTGGCTAAGAGGTTTGCTCTCAAGACCAGAAGTCAGATCACCAAG

437 GlyLysThrSerAspThrAsnSerSerAlaAspLysThrThrAlaThr'LeuProLeuSerPheLysGluAlaThrLeuAlaLysArgPheAlaLeuLysThrArgSerGlnlleThrLys

1560 CGGAAGAGGATGTCGCTCATCAAGGAGAAGAAGGCCGCCCAGACGCTCAGTGCCATCTTGCTAGCCTTCATCATCACGTGGACCCCCTACAACATCATGGTCCTGGTGAACACCTTCCGT 477 ArgLysArgMetSerLsuIleLysGluLysLysAlaAlaGlnThrLeuSerAlalleLeuLeuAlaPhellelleThrTrpThrProTyrAsnlleMetValLeuValAsnThrPheArg 6

1680 GACAGCTGCATACCCAAAACCTATTGGAATCTGGGCTACTGGCTGTGCTATATCAACAGCACCGTGAACCCTGTGTGCTATGCCCTGTGCAACAAAACATTCAGAACCACCTTCAAGATG 517 AspSerCysIleProLysThrTyrTrpAsnLeuGlyTyrTrpLeuCysTyrlleAsnSerThrValAsnProValCysTyrAlaLeuCysAsnLysThrPheArgThrThrPheLysMet

1800 CTCCTCTTGTGCCAGTGTGACAAAAGGAAGAGGCGCAAACAGCAGTACCAGCAGAGACAGTCGGTCATTTTTCACAAGCGAGTGCCGGAGCAGGCCTTGTAGAAAAGGGGTATGTTGAGA 557 LeuLeuLeuCysGlnCysAspLysArgLysArgArgLysGlnGlnTyrGlnGlnArgGlnSerValllePheHisLysArgValProGluGlnAlaLeu***

1920 GCAGTGACCACGCAAGCGCGTCAGCCCACACAGCCTAGCAGGAGTCTGGCGCAGGTGGGGGGAGCCATCCCTGATGGTGATAAAATGGGTTTTTATCGACCTCACGGGAGAGAAGCTACC

2040 TGTTTACTGATCCATTGAATAACTGATTTTGGTCCAATGCCAATTCAGCAGGAAAGAAGGAGAGGCATACCGCTAAACATGAAGAGATGTGTTCTGAAACAGACTTTTAAGTGGATTTTG

2160 TTTCCTCTAAGAGAAAAAGAAATTATTGTCTCAGAGCAAGTATCCTCAGAAATTGGTCTGCCTGGGTCTCTTAATTCCTATCAGCTCTGGAATCACTGGTGAGCCTCAAGGCACTAGATG

2280 CCATGTGCTCTCCCTAAGGGTCCCAAAGTGTCCATCCAGATCCCATGTGAAGCACGGCTAGCTTGAAAA

Figure 1. Nucleotide sequence of the rat mAChR M4 cDNA. The 589 amino acid M4 polypeptide within the sequenced 2341bp region is indicated. Potential N-1inked glycosylation sites are boxed and hydrophobic transmembrane domains are underlined. An inframe 5' stop codon (TAA) is indicated by asterisks.

indicated a single long open reading frame encoding a polypeptide of 589

amino acids with an Mr of 66,151 daltons. The initiation codon is assigned

to the first AT6 triplet encountered (nt 132) which also appears downstream

of an in-frame nonsense codon, TAA (nt 87-89).

The hydropathy profile of this protein predicts seven membrane spanning

regions within which the homologies to the other G-protein coupled

receptors is greatest. The predicted amino acid sequence is most homologous

to the muscarinic receptor class possessing 45.5% and 39.3% overall amino

acid identity with the porcine Ml (5) and M2 (5, 7) subtypes, respectively

(Fig. 2). Following the nomenclature of Peralta and Capon who have

characterized by functional expression the human genomic mAChR subtypes

(15), the mAChR in this report was designated M4.

127


r a t M 4

p o r c i n a M l

p o r c i n e M 3

M T L H S N 8 T T S P L F P M I S S 8 W V H S P S E A Q L P L Q T V T Q L 0 S Y N I S ^ E T G 0 F [ i ]

A P P A M N T S

M N N S

r a t M 4

p o r c i n a M l

p o r c i n e M2

5 1

9

8

S N D T S S D P L

V S P N I T V L A

S N S G L A L T S

I j ^ L V T I I G N I I T v

L s l [ ^ t | v t | g N 0 L V

L S L V T I I G N I L V

1 |v|A|f k V n K 0 L K T

L H s F K V N T E L K T

Mfv S I T I K V N R H J J ^

r a t M 4

p o r c i n a M 1

p o r c i n a M 2

101

59

5 7

V N N Y F L L S L A C A D L I

V N N Y F L L S L A C A D L I ^

V N N Y F L I T I S L A C A D L I I G V F S M N L Y T i l l y

g ^ H s m n l [ f J t t y

I G f r f F S M N L Y T T Y

I I ^ N R

L L [ M G H

T V T Y Y

W A L G

W A L G

V / P ] L G

L A C D L W L | S I )D Y V A

L A C D L W L A L D Y V A

P V V | C D L W L A L D Y V [ V

-rat M 4

p o r c i n e M 1

p o r c i n e M 3

151

1 0 9

1 0 7

S N A S V M N L L

S N A S V M N L L

S N A S V M N L L

I S F D R Y F S I J J T R P L T Y R A K R T T K R A G | V

S F D R Y F S V T R P T F I L Y R A K R T ! P R I R A I A L

S F D R Y F [ ^ V I F T ^ P L T Y I P V I K R T T K I M A

M I G L A W ^ i Ml G L A ^ T V Ml ' '

S F V L W A S F V L W A S FfTlL W A

: r a t M 4

p o r c i n e M 1

p o r c i n e M 2

201

159

1 5 7

P A I L F W Q Y

P A I L F W Q Y

P A I L F W O F T

V G K R T V P P

V G E R T V L A

V G V R T V E D

G E C [ F J | Q F L S

G 0 C Y I 0 F L S

G E C Y I Q F F F I S

P T

P 1

I T F G T A I A A F P V T I M T T L Y

I T F G T A 0 A A F Y L P V L Y

^ T F G T A I A A F Y L P v f l f l M ^ v | l Y

r a t M 4

p o r c i n e M l

p o r c i n e M 2

2 5 1

2 0 9

2 0 7

W R I Y | J ^ E T E

W R I Y R E T E

W F H L L J S L R J A S K S

t [ k E L

A r I e L

T ^ A E A E N F 0 H P T G S S R

J J P G K G G G

I 0 K D K K E P V A N Q [ E ] P V S P S L @ Q G R I V K P

G r Q A S G A Q G S E

S c s s Y E L Q RAO

S s s s S E R S dp

r a t M 4

p o r c i n e M l

p o r c i n e M 2

3 0 1

2 5 0

2 5 6

S R R K Y

P E T P P

E H N K 1 Q N

G R C I H F W F T 0 K S W K P S A F H O M D Q D H I U S

G R C | C [ ^ C C R A P R L L [ 3 A Y

^ K A P Y D A V 0 E N C V J ^ G E I E J K E S S N D I ^ T

d [ ^ w n n n d a a a

W K E E E E E D E G _ _ _ v [ i ] A V A S N M R D D E I T Q D E

L E N S A S

S M E S L T s

r a t M 4 3 5 1 s D E E D 1 G S E p o r c i n e M l 2 9 1 s E G E . E P G S E

T R A I y [ ^ i v l k l p g h s s i l n [ ^ t k l p s s d n l q v s n e d l g t

V V I K M P M

p o r c i n e M 2 3 0 6 N T V S T S L G H S K D E n[s] K Q T C I K I V T K T Q D S C T P A N T T V E L V G S

V D

V D

r a t M 4

p o r c i n e Ml

p o r c i n e M 2

4 0 1

3 1 0

3 4 3

R N A H K L

A A K j O

A Q K S M G D G D N C Q K D F T K L P I Q L E S A V D T G K T S D T N S S A D K T P P R [ ^ S P N

r a t M 4

p o r c i n e M l

p o r c i n e M 2

4 5 1

3 2 6

3 5 1

A T L P

V K R [ P

L S F 0 E A T R F A L K [ T

T R K G R E R A ^ K G Q K P R G

S G Q N G D E 0 Q N I V 0 R K I V K M I T K

R S

E

Q 1 ^ K R K

Q L A K R K

Q P A K F I ^

S L 1

S L V

K E K K A A [ ^ T L S A I L L A F I

T F L S _ L J V [ K E K K A A R T L S A I L L A F I

P P P S - ^ E K K I V T I R T I I L I A I L L A F I

r a t M 4

p o r c i n e M l

p o r c i n e M 2

T W T P Y N I M V L V N T F [ r D S

y j T W T P Y N I M V

I T W[A]P Y NFVIM V

L V ^ T F

L[| [n T F

K D

A P

C 1 P K T Y V,

C 0 P E T L V,

C 1 P N T V V,

N I L G Y W L C Y I N S T | V J N ~ P

E[L G Y W L C Y 0 N S T I N P

T I I G Y W L C Y I N S T I N P

C Y A L C N K T F

C Y A L C N I ^ F

C Y A L C N F A F R F

r a t M 4

p o r c i n e M 1

p o r c i n e M 2

5 5 1

4 2 6

4 4 9 K K

T F ^ M L L L C

T F [ r L L L L C

T F ^ H L

Q C D K R

R W L D K R

H Y K N I G A T R

K R

R W

R K

R K

Q Q Y Q Q R Q S V I F H K

I P K R P G S V H R T P S

V P E Q A L

Q C

Figure L Comparison of muscarinic receptor subtype protein sequences. Identical amino acids in the sequences are boxed. The M4 sequence has 45.5% identity with the porcine Ml sequence (5) and 39.3% identity with the porcine M2 sequence (5, 7). Gaps have been introduced to maximize the homology.

The muscarinic receptor M4 protein sequence is longer than the other

muscarinic receptor proteins at both the N-terminal region, where five

128


2: Q.

CD

TJ C D O rj

o Q. I/)

Free QNB (pM)

Figure 3. ^H-QNB binding to the M4 mAChR subtype. A) Specific ^H-QNB binding was determined from the difference between total binding of the labelled antagonist in the presence or absence of lOjxM atropine. B) Scatchard analysis (using least-squares linear regression) of the binding data gives a KD of 68 pM (n=4). Untransfected cells did not bind ^H-QNB. Bmax values varied depending upon the transfection efficiency with transfected cells expressing between 4,000-20,000 receptors. The data shown are from a representative experiment.

potential N-1inked glycosylation sites are present, and also in the

intracellular loop region connecting transmembrane domains 5 and 6. This

loop region has very low homology among all G-protein coupled receptors as

well as between different subtypes of a receptor family. The functional

significance of these highly variable and therefore unique sequences in

each receptor subtype, is unknown. However, their intracellular location

suggests that they may represent regulatory sites for signal transduction.

To demonstrate that the cDNA we isolated is a muscarinic receptor

subtype we transiently expressed the M4 cDNA in mammalian cells (13).

Membrane fractions of these cells were analyzed for ^ h - Q m b (a specific

muscarinic receptor antagonist) (1-3) binding. The presence of a high

affinity binding site for QNB on these membranes demonstrated that the M4

receptor subtype is indeed a muscarinic receptor (Fig. 3). The affinity

constant (KD= 68pM; n=4) is of the same range as that determined for either

the Ml or M2 subtypes (1, 5, 7, 15).

The expression pattern of the M4 subtype was determined in rat brain by

in situ hybridization using ^zp^-j^belled antisense cRNA derived from the

TM5-6 subtype specific sequences of the cDNA. M4 transcripts were highly

abundant in cells in all subfields of the hippocampus, the dentate gyrus,

129

Vol. 149, No. 1, 1 987 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

Figure 4. M4 mAChR mRNA expression in rat brain. Rat brain saggital sections were hybridized with a 32p_]a5elled subtype-specific cRNA probe. Brain regions showing the highest levels of expression included cerebral cortex (Cx), hippocampus (Hi), dentate gyrus (DG), thalamus (Th) and the granule cell layer of the cerebellar cortex (Cb). M4 mRNA was not highly expressed in the caudate-putamen (CPu).

the cerebral cortex and the granule cell layer of the cerebellar cortex

(Fig. 4). A strong signal was also obtained in the grey matter of the

spinal cord. We were not able to detect any specific signal in either heart

or kidney (not shown). No signals were obtained with sense strand cRNA

probes or when sections were treated with RNase before hybridization with

cRNA.

The demonstration of a family of muscarinic receptor genes suggests

that preferential coupling of a given type of G-protein (e.g. 6i, Go, Gs

etc.) (16) to specific receptor subtypes may occur. Thus, activation of

endogenous mAChRs in the cell lines NG108-15 and 1321-Nl inhibits adenylate

cyclase activity and stimulates phosphoinosityl turnover, respectively

(17), suggesting that these cell lines express a distinct receptor subtype

that couples to a specific effector G-protein. Preferential coupling of a

G-protein with a receptor subtype may occur in part by interaction with the

subtype specific sequences located intracellularly between TM5 and 5. In

support of this notion, deletion of part of the TM5-6 sequence in the (3-

adrenergic receptor abolishes its ability to regulate adenylate cyclase

activity while not affecting ligand binding (18).

Our localization studies demonstrate that the M4 receptor subtype is

expressed in specific regions of the central nervous system. The Ml subtype

130

Vol. 1 4 9 , No. 1, 1 9 8 7 BIOCHEMICAL A N D BIOPHYSICAL RESEARCH C O M M U N I C A T I O N S

has been localized to the cerebral cortex and corpus striatum (5) while the

M2 subtype is expressed in the medulla-pons and heart atrium (5). Thus, not

only does the neuronally expressed M4 subtype display a higher degree of

protein identity to the cerebral Ml subtype but it also has a similar

regional distribution in brain (e.g. cerebral cortex). The existence of

distinct genes encoding mAChR subtypes, in addition to permitting tissue-

specific gene expression, could also allow for differential gene regulation

within the same tissue or cell.

Hippocampal and cerebral cortex cholinergic systems (where the Ml and

M4 receptor subtypes are expressed) are particularly involved in the

processes of learning and memory (1). These cognitive processes are

believed to rely upon the phenomenon of synaptic plasticity. We propose

that receptor subtype heterogeneity is one mechanism by which such

plasticity is achieved. More explicitly, by controlling the relative levels

of expression of two (or more) receptor subtypes, a single cell could

potentially alter quantitatively as well as qualitatively its response to

the synaptic release of a given neurotransmitter. The availability of

molecular probes for these receptor subtypes will allow the examination of

this hypothesis.

The full extent of mAChR subtype diversity is unknown but, based on the

methods outlined here, can be addressed. Whether protein sequence

similarities between mAChR and other classes of G-protein coupled receptors

are sufficient to allow their isolation remains to be seen.

Subsequent to the completion of this work, Bonner et al. (19) reported a

similar rat muscarinic receptor subtype designated m3. Our results are

consistent with theirs except that we note three amino acid differences

between our predicted protein sequence and theirs, repectively: Ais4 > R,

R516 > C and M556 > T. These workers (19) also demonstrated higher levels

of expression of this subtype in the dentate gyrus relative to the

hippocampus as compared to our finding of similar levels of expression in

both the hippocampus and the dentate gyrus.

ACKNOWLEDGEMENTS

We thank Drs. Ernie Peralta and Dan Capon for sharing their human genomic receptor sequences prior to publication and for their encouragement, Bettina Crusius for excellent technical assistance, Jutta Rami for preparation of the manuscript and Dr. Alfred Bach for his support of this work. Predoctoral support for T.B. was funded by Boehringer Ingelheim. This work was supported by BMFT grant No. BCT0381 5 C7 to P.H.S.

REFERENCES

1. Nathanson, N.M. (1987) Ann. Rev. Neurosci. 10, 195-236.

131


2. Hammer, R., Berrie, C.P., Birdsall, N.J.M., Burgen, A.S.V. and Hulme, E.G. (1980) Nature 283. 90-92.

3. Birdsall, N.J.M. and Hulme, E.G. (1983) Trends Pharmacol. Sci. 4, 459-463.

4. Watson, M., Roeske, W.R., Vickroy, T.W., Smith, T.L., Akiyama, K., Gulya, K., Duckies, S.P., Serra, M,, Adem, A., Nordberg, A., Gehlert, D.R., Wamsley, J.K. and Yamamura, H.I. (1986) Trends Pharmacol. Sci. Suppl. 7, 46-55.

5. Kubo, T., Fukuda, K., Mikami, A., Maeda, A., Takahashi, H., Mishina, M., Haga, T., Haga, K., Ichiyama, A., Kangawa, K., Kojima, M., Matsuo, H. Hirose, T. and Numa, S. (1986) Nature 323, 411-416.

6. Kubo, T., Maeda, A., Sugimoto, K., Akiba, I., Mikami, A., Takahashi, H., Haga, T., Haga, K., Ichiyama, A., Kangawa, K., Matsuo, H., Hirose, T. and Numa, S. (1986) FEBS Lett. 209, 367-372.

7. Peralta. E.G., Winslow, J.W., Peterson, G.L., Smith, D.H., Ashkenazi, A., Ramachandran, J., Schimerlik, M.I. and Gapon, D.J. (1987) Science

236, 600-605. 8 . Giraldo, E., Hammer, R. and Ladinsky, H. (1987) Life Sci. 40, 833-840. 9. Dixon, R.A.F., Kobilka, B.K., Strader, D.J., Benovic, J.L., Dohlman,

H.G., Frielle, T., Bolanowksi, M.A., Bennett,, G.D., Rands, E., Diehl, R.E., Mumford, R.A., Slater, E.E., Sigal, I.S., Garon, M.G., Lefkowitz, R.J. and Strader, G.D. (1986) Nature 321, 75-79.

10. Malfroy, B., Schofield, P.R., Kuang, W.-J., Seeburg , P.H., Mason, A.J. and Henzel, W.J. (1987) Biochem. Biophys. Res. Gommun. 144, 59-66.

11. Wood, W.I., Gitschier, J., Lasky, L.A. and Lawn R.M. (1985) Proc. Natl. Acad. Sci. U.S.A. 82, 1585-1588.

12. Sanger, F. Nicklen, S. and Goulson A.R. (1977) Proc. Natl. Acad. Sci. U.S.A. 74, 5463-5467.

13. Eaton, D.L., Wood, W.I., Eaton, D., Hass, P.E., Hollingshead, P., Wion, K., Mather, J., Lawn, R.M., Vehar, G.A. and Gorman, G. (1986) Biochemistry 25, 8343-8347.

14. Shivers, B.D., Schachter, B.S. and Pfaff, D.W. (1986) Methods Enzymol. 124, 497-510.

15. Peralta, E.G., Ashkenazi, A., Winslow, J.W., Smith, D.H., Ramachandran, J. and Gapon, D.J. (submitted for publication).

16. Stryer, L. and Bourne, H.R. (1986) Ann. Rev. Gell Biol. 2, 391-419. 17. Harden, T.K., Tanner, L.I., Martin, M.W., Nakahata, N., Hughes, A.R.,

Helper, J.R., Evans, T., Masters, S.B. and Brown, J.H. (1986) Trends Pharmac. Sci. Suppl. 7, 14-18.

18. Dixon, .R.A.F., Sigal, I.S., Rands, E., Register, R.B., Gandelore, M.R., Blake, A.D. and Strader, G.D. (1987) Nature 326, 73-77.

19. Bonner, T.I., Buckley, N.J., Young, A.G. and Brann, M.R. (1987) Science 237, 527-532.

132

Reprinted from Nature, Vol. 335, No. 6185, pp. 76-79, 1 September 1988 © Macmillan Magazines Ltd., 1988

Structural and functional basis for GABAA receptor heterogeneity Edwin S. Levitan*t, Peter R. Schofieldtl§, David R. Burt*t, Lucy M. Rhee§, William Wisden*, Martin K6hler$, Norihisa Fujita*t, Henry F. Rodriguez§, Anne Stephenson*, Mark G. Darlison*, Eric A. Barnard* & Peter H. Seeburgt§|| * MRC Molecular Neurobiology Unit, MRC Centre, Hills Road, Cambridge CB2 2QH, U K t Laboratory of Molecular Neuroendocrinology, ZMBH, University of Heidelberg, Im Neuenheimer Feld 282, 6900 Heidelberg, FRG § Genentech, Inc., Department of Developmental Biology, 460 Point San Bruno Boulevard, South San Francisco, California 94080, USA

When y-aminobutyric acid (GABA), tlie major inhibitory neurotransmitter in vertebrate brain, binds to its receptor it acti-vates a chloride channel. Neurotransmitter action at the G A B A A receptor is potentiated by both benzodiazepines and barbiturates which are therapeutically useful drugs (reviewed in ref. 1). There is strong evidence that this receptor is heterogeneous*"'. We have previously isolated complementary DNAs encoding an a - and a /3-subunit and shown that both are needed for expression of a functional GABA^ receptor®. We have now isolated cDNAs encod-ing two additional GABAA receptor a-subunits, confirming the heterogeneous nature of the receptor/chloride channel complex and demonstrating a molecular basis for it. These a-subunits are differentially expressed within the CNS and produce, when expressed with the j3-subunit in Xenopus oocytes, receptor subtypes which can be distinguished by their apparent sensitivity to GABA. Highly homologous receptor subtypes which differ functionally seem to be a common feature of brain receptors.

Of 12 bovine brain a-subunit cDNA clones isolated®, one contained an open reading frame encoding a different 451 amino acid polypeptide (Fig. 1). The first 28 amino acids of this polypeptide A have typical features of a signal peptide^ and the predicted relative molecular mass (MJ of the mature polypep-tide is 48,000 (48 K). The corresponding amino acid sequence shows 79% identity with the previously characterized G A B A A receptor a-subunit®, but only 34% amino acid identity with the mature G A B A A receptor jS-subunit®. We therefore called the subunit that this cDNA encodes an a l subunit, with the pre-viously characterized clone encoding an a 1 subunit. The predic-ted al subunit polypeptide sequence also contained a peptide sequence that was chemically determined from the affinity-purified adult G A B A A receptor®, indicating that the new subunit is present in this receptor (Fig. 1). Re-screening of the bovine brain cDNA library revealed the existence of additional cDNAs encoding a-subunits. Ten cDNAs were analysed and the longest three sequenced. These all contained putative unspliced introns at two different locations but had the potential (after intron removal) to encode a polypeptide of 492 amino acids (Fig. 1). A 28-residue signal peptide^ and a mature polypeptide of M^ 52 K is predicted. This subunit, designated a3, displays 72% sequence identity to both the a l and al subunits. The receptor heterogeneity observed arises from different genes encoding the respective subunits and not by alternative splicing.

The presence of four hydrophobic transmembrane domains, M1-M4, deduced from both the a l and fi chains are present

t Present addresses: Department of Pharmacology, Yale University School of Medicine, New Haven, Connecticut 06510, USA (E.S.L.); Pacific Biotechnology Ltd, 72 McLachlan Avenue, Rushcutters Bay 2011, Australia (P.R.S.); Department of Pharmacology, University of Maryland School of Medicine, Baltimore, Maryland 21201, USA (D.R.B.); Biorganic Research Depart-ment, CIBA-GEIGY (Japan) Ltd, Takarazuka 665, Japan (N.F.). II To whom correspondence should be addressed at Heidelberg.

at equivalent positions in the a2 and a3 polypeptides®. The sequences of M1-M3 are extremely highly conserved in all three a-subunits with only a single conservative amino acid change ( I> M314 in the a3 sequence) over a stretch of 98 amino acids (Fig. 1). The M4 region is also highly conserved, as is the majority of the presumed extracellular large N-terminal domain, which shows 75% sequence identity in all three chains (Fig. 1). The region between M3 and M4, deduced to be intracellular, and the extreme N-terminal sequences (including the signal peptides) are very poorly conserved. Each a-subunit contains potential N-linked glycosylation sites: two in a l , three in a2 and four in a3. The hypothesis that members of the ligand-gated ion channel superfamily have a common evolutionary origin®-'" is strengthened by the homologies seen here and by the homologous positions of some of the putative unspliced intron sequences in the a3 subunit and the a, y and 8 subunits of the nicotinic acetylcholine receptor genes"*'^.

We investigated whether the three a-subunits confer different properties on the GABAA receptor by expressing each a RNA in combination with the /3-subunit RNA in Xenopus oocytes. Electrophysiological recording showed that either the a2 or the a3 polypeptide could be substituted for the a l polypeptide to yield GABA-evoked currents. These were always inhibited similarly by bicuculline and picrotoxin (Fig. la). The barbiturate pentobarbital directly activated each receptor and markedly enhanced the responses to GABA (Fig. lb). Plots of current against voltage demonstrated that receptors containing any one of the three a-subunit polypeptides had reversal potentials (E^) close to the chloride equilibrium potential'"* {E^~-10 mV, Fig. 2c), indicating that each receptor subtype forms GABA-gated chloride channels. Finally, the currents due to the (a2 + ^ ) and the (a3 + /3) receptors do not increase linearly with voltage, but show marked outward rectification (Fig. 2c). Identical voltage-dependence is seen with the (a 1 + /3) receptor (data not shown). Thus, substitution of either the al or the a3 subunit for the a l subunit does not appear to change the qualitative properties of the expressed G A B A A receptor.

Exchanging the a-subunits, however, alters the apparent sensitivity of the receptor to neurotransmitter (Fig. 3). The half-maximal dose for G A B A varies significantly between recep-tor subtypes: I.3±0.2|JLM (N = 6 oocytes) for (a2 + /3), 12± 1 |JLM (n = 6) for (al + p) and 42±8 ixM (n = 5) for (a3 + yS). The differences in G A B A sensitivity could be due to altered receptor affinities, changes in coupling of the receptor to channel gating, or desensitization kinetics. The latter seems less likely since the differences in G A B A sensitivity were consistent, despite large variations in desensitization. Moreover, for ( a l + /S) and (a3 + ^) . Hill plots were linear (data not shown). Our results demonstrate that exchanging the a-subunits can produce an apparent 30-fold change in sensitivity of G A B A A receptors expressed in oocytes.

The expressed receptors display differences in their apparent sensitivity to GABA which are indicative of receptor subtypes, and share many properties with native G A B A A receptors. These include competitive inhibition by bicuculline, voltage-depen-dent enhancement and direct activation by pentobarbital, anion selectivity, voltage-dependent gating (data not shown) and characteristic multiple single-channel conductance states (L. Blair and V. Dionne, unpublished). The expressed receptors are, however, not identical with native G A B A A receptors. Hill coefficients derived from the data in Fig. 3 are less than or equal to 1, rather than 1.4 to 2 as expected'^''®. Benzodiazepine potenti-ation was not observed, whereas a modest, response in follicle-enclosed oocytes to a single benzodiazepine had been seen previously®. Neither co-expression of all four known subunits nor the use of a range of benzodiazepines gave the receptor potentiation observed when brain poly(A)"^ RNA is expressed in Xenopus oocytes'®"'®. Additionally, transiently transfected mammalian cells expressing ( a l + ;8) subunits do not bind ben-zodiazepines (Pritchett et al, submitted). Specific post-transla-

Fig. 1 Comparison of the predicted amino acid sequences of the bqyine GABA^ receptor a2 and a 3 subunit cDNAs with the previously character-ized a l and /3-subunits®. Primary sequence iden-tities are boxed. Potential signal sequence cleavage sites' are indicated by the large arrow. Amino acid sequence numbering starts at the proposed mature N-terminal residue, the presumptive signal sequen-ces being indicated by negative numbering. The proposed membrane-spanning hydrophobic sequences and disulphide-bonded loop region® are indicated by solid bars and a dotted line respec-tively, and the putative N-linked glycosylation sites by dashed boxes. The inverted arrows in the a 3 sequence show the approximate locations of the two putative unspliced introns; the first being located after the second nucleotide of the codon encoding Rij^, the second after the third nucleotide of the codon encoding K354. The DNA sequences of the al and a 3 subunit cDNAs are available on request and have been deposited in the EMBL DNA data base under accession numbers 12361 and 12362, respectively. Methods. A bovine brain cDNA library® was screened with the [y^^P]ATP-labelled 60-mer oligonucleotide GR-5 (5'-TGCCACTGACT-TCTTTCCATTGTGGAAAAAGGTATCCGG-AGTCCAGATTTTACTTGCCAT-3') which cor-responds to a portion of the extracellular sequence of the a l subunit®. Positively hybridizing clones were sequenced by the chain termination method combined with specific oligonucleotide priming. The a 2 subunit sequence is from a single clone AbGR16 and the a 3 subunit sequence is from the clone AbGR9.3. Except for the putative unspliced intron sequences, only two nucleotide differences were detected between the three a3-subunit cDNA clones sequenced. Nucleotides 1,037 and 1,038 are GA in AbGR9.3 and AG in AbGRl. This would result in an Asn to Asp change. The unspliced intron located after nucleotide 610 was found in A bGRl, whereas that located after nucleotide 1,205 was seen in both AbGR7 and AbGR9.3. Other cDNA clones isolated from this library have also been shown to contain unspliced introns. Micro-sequencing of affinity purified GABA^ receptor peptide fragments derived by cyanogen bromide cleavage (see ref. 8) included the following mixture sequence: major sequence : M X Q K I ( C ) Y A V A V A N Y A P N L minor sequence : M X Y T L A I P T E F ( N ) L ( A )

( F ) D Q

The underlined residues are specific for the a 2 subunit (residues 333-350). Residues enclosed in parentheses have uncertain assignments; the

methionine residue is assumed.

bGABAA R al bGABAA R a2 bGABAA R a3 bGABAA R 6

-27 - 2 8 - 2 8 -25

M K K S I ' G L S D Y I . W A W T L [ F ] L S T L T G R S Y G M K T K L N S S N M Q L L L F V F L A W D P A R L V L A M I 1 T Q M S Q F Y M A G L G L L I F I L I N I L P G T T G

L G L L s | l J p V M I A M V C C A M W T V Q N R E S

176 201

239 210

255 254 279 250

295 294 319 290

330 328 356 330

Q P S L Q

Q V E S R R Q E P G D E V K Q D I G G L S P K H A P D l ^ P

D E L K D E A K N D S T D

A N E P S

F r

F T F T

M S! Y V K

R l L D R L L b G Y D N R L R P G L G R 1 L D R L L D G Y D N R L R P G L G

L D R L L D G Y D N R L R P G L G E T V | D R L L p r i G Y D F I R L R P f p T I G

E R V D S I

E ^ T E V

D A V p- E V V

D M E Y T I D V F F R Q D M E Y T I D V F F R Q D M E Y T I D V F F R Q N I M I ^ Y T | L T M Y I F M O

W K D E R L K F W K D E R L K F W | H ] D E R L K F , , W K p f i r i R L I S Y S

G P M G P M G P M GJL P

K T D r F V T S F G P V S D H F T N I Y V T S F G P V S D T K T D 1 Y V T S F G P V S D T G M R i_ D [VJAI^I D M|V S f E V

T V L R L N N L M A S K I W T N I L R L N N L M A S K I W T K I L P L , N N L L A S K I W L [N L j] t j [D|N[R V ^ D Q L j w l v

P D T F F H N G K K S V A HFIVLM'L^M P N K L L R P D T F F H N G K K S V A H I N M TIM P N K L L R P D T F F H N G K K S V A H .NJA^XJFFLP N K L L R P D TRYIFLLLNLDLK K S R M H L G ^ T LVIR|N|R M I [R

I T E D G T L L Y T M R L T Q D D G T L L Y T M R L T

I. V D M G I T J L L Y T M R4L T I L H P [DG TTVIL YIG L IRFTIT IT

V R V Q

E C P M H L E D F P M D E C P M H L E D F P M D

. E C P M H L E O F P M D AJAICJMIMJDILJR R Y|pjriD_

C P L K F G S Y A Y T [ r A C P L K F G S Y A Y T T l s

„ C P L K F G S Y A Y T T E Q & i I C l f ] L | E I Els y I ^ Y T T D D I E F Y

R E P A Y i N . A . S j L G K IN

W N G G - -

P. S V A V A

A krr D G S R L N Q Y D L D G S R L N Q Y D L D G S R L N O Y D L

L G Q T - V D S G I V Q S S T G E Y V V M T P G Q S - I G K E T I K S S T G E Y T V M T L G H V - V G T E I I R s S T G E Y V V M T V D Y - K M V S K K V E F T | T G | a 1 Y P R L S

H F H L K R K I G Y F V I Q T Y L P C I M T V I L S Q V S H F H L K R K I G Y F V I Q T Y L P C I M T V I L S Q V S

W L N R E S V P W L N R E S V P

T|H F H L K R K 1 G Y F V I Q T Y L P C I M T V I L S L SIFIRIL K RINII G Y FLL LLO T YHJILPLS T L I TIL L S

Q V S F W L N R E S V P A FWLV 5 F WIDNIY D A S AJA

S A R N S L P K V A Y A T A M D W F S A R N S L P K V A Y A T A M D W F

A V C Y A V C Y

R T V F G V T T V L T M T T L S I R T V F G V T T V L T M T T L S I R T V F G V T T V L T M T T L S I S A R N S L P K V A Y A T A M D W F| MIA V C Y R| V A L IG II IT T V L T M T TrrislT H L I R I T T I L P KRMYRV"ICLAFRID[TY L MGICJF

A F V F S A L I A F V F S A L I A F V F S A L I E F A T VIF V FFN A LIT] ELYLARFL

A T V N Y A T V N Y

V N Y V N Y

- F T K - F T K

F FJGIK

G Y G V? S

A W D A W D AWE

G P Q K K G A

V P - - E K P V N D K P E A L E M K

Q D Q S A N E K N

K K K K

IkJL V K D P L I K K N N T Y A P . T A T S E K A S V M I Q[RJ N A Y A V A V A N

- - T P A V P T K K T S T T F N I V E M N K V Q V D A H G 0 I L L S T L

Y T P N L A R G D P G L A T I A K S A Y A P N L S K - D P V L S T I S K S A Y P I N L A K - D T E F S A I S K G A V G T T

E I R N E T S G S E V L T G V G D P K T T M

367 T T I 364 T T T 393 K G A 370 Y S Y

E P K E V K E P - N K K P S T S S T S A S I Q Y

E T K P E N K P • T I I A S P K T T C V Q K P M S S R E G Y G R A L D R H G A H S K G R I R R R A S

381 P E P K K T F N S V S K I D R L S R I A F P L L F G I F N L V Y W A T Y L N 378 - - A E A K K T F N S V S K I D R M S R I V F P V L F G T F N L V Y A T Y L N 416 D I P T E T K T Y N S V S K fv\D K I S R I I F P V L F A I F N L V Y W A T Y V N 410 Q L K V K I P D L T DLV| N S I D K W S .RJM F F P I S L F NPVLV Y W IL Y|_Y V H

419 R E 416 R E 456 R E

P Q L K A P T P H Q P V L G V S P S A I K G M I R K Q

tional modification may be needed for such potentiation Alternatively, additional receptor components may be lacking. However, not all native G A B A A receptors have been found to be benzodiazepine sensitive^'"^^. Hence the expressed receptors may represent benzodiazepine-insensitive G A B A A receptors.

The expression of the G A B A A receptor a-subunits was examined by Northern blot analysis of m R N A from bovine cortex, hippocampus and cerebellum (Fig. 4). Subunit-specific probes recognize single transcripts ( a l , 4.5 kb; a2, 8.5 kb; a3 , 4.0 kb). Although the /3-subunit transcript is found in all these regions (data not shown), the a-subunit m R N A s display distinct differences both in their steady-state amounts and in their rela-

tive abundance in the brain regions examined. The a 1 subunit m R N A is clearly the most abundant one in all three regions, whereas a 2 m R N A is the least abundant. The highest levels of a l , a 2 a n d a 3 m R N A s occur in the adult cerebellum, hippocam-pus and cortex, respectively. The amount of a 3 m R N A seems to change during development, being prominent in the cerebel-lum of the 12-day-old calf and then declining. The a l and a 2 m R N A s show no obvious developmental change. The genes encoding G A B A A receptor a-subunits are therefore under specific control, allowing for regional variation in the levels of the respective subunits.

The three a-subunits of the G A B A A receptor differ in both

a 2 .(i

CONTROL BIG CONTROL BIC

r ^

1 / _ j 9 0 n A 20s 1 / _ | 3 0 n A

40s

Fig. 2 Expression of cloned G A B A a receptors in oocytes injected with (a2+)3) (left) or ( a 3 + ^ ) (right) RNAs. All data are from defolliculated oocytes, except the ( a3 + j8) traces in a and b. a, Efiects of bicuculline methobromide and picrotoxin on GABA-evoked currents. GABA applications (0.2 jaM for a 2 + )3; 2^,M. for a 3 + /3) are indicated by bars. The different concentrations were used to compensate for the different sensitivities (Fig. 3). Bicuculline (BIC), 4 |xM, or picrotoxin (PTX), 2.5 |xM, was applied for 1.5 minutes; GABA was then applied in the continued presence of each drug. The traces are from the same oocyte with the bicucul-line washout not shown, b. Effects of the barbiturate pentobarbital. 20 I.M pentobarbital (PB) directly activates the expressed receptors (no effect is seen in uninjected oocytes) and enhances the responses to GABA (0.1 jjiM for a2 + p, 1 j j lM for a3 + )3). c. Current/voltage relationships for the GABA-evoked current. Measurements were made during a voltage ramp (300 mV s~'). Currents before applica-tion of GABA (l(i.M for (a2 +P)- 30 |j.M for (q;3 + ^ ) ) were subtracted. The reversal potential is the voltage where zero current flows through the membrane. Reversal potentials (E^) were: ( a 2 + /3) £'r = - 2 4 ± 2 m V (±standard error of the mean, n = 4 oocytes); ia3 + p) £r = - 1 9 ± 2 m V ( n = 6). Similar results as a, fc and cwere also obtained with oocytes injected with ( a l + /8), E^ = -20±l mV (n = 6) (data not shown, see also ref. 8). Methods. The a \ and /3-subunit RNAs were prepared using SP6 RNA polymerase as previously described®. The a2-subunit cDNA was subcloned as a 1.8 kb Pst\-Sac\ fragment (restriction sites from the polylinker of the sequencing vector M13tgl30) into Ps/I-SacI-cleaved pSP64 (PbGRa;2sense). The fragment contains the first 1,803 base pairs (bp) of the cDNA sequence. These capped RNA transcripts co-injected with )3-subunit RNA produced poor GABA responses, presumably due to the very short 5' untranslated sequence. Accordingly, synthetic sequences were inserted providing 5' untranslated sequences based on the a l subunit sequences and a Kozak consensus sequence^®. The Pstl (polylinker)-fi<3mHI (nucleotide 68) fragment of pbGra2sense was replaced by the synthetic DNA sequence:

5 ' G M G C C C A C C A T G M G A C G A M C T O M C A G C T C C M T A T G C A G T O C T r c c n T I T O 3 ' 3 ' ACXJTCTTCGGGTGGTACTTCTGCTTTGACTTGTCGAGGmTACXSTCAACGACGAAAM 5 '

(The initiation codon is overlined). This plasmid (pbGRa2sense + K) was linearized with Pstl and the synthetic DNA sequence: 5' GGATmcrCTCXSCAGACnTITCCCmJTCTGGAGCGATCCTGTGCCCAGAGGGffi 3 '

3 ' ACGTCCTAAAAGAGAGCGTCrrGAAAAGGGCCCAGACCTCGCTAGGACAaSGGTCTCCCCCmSCTCG 5 '

inserted in the orientation shown. After linearization of the resultant plasmid (pbGRa2sense + K-l-5'ut) with 5acl, capped SP6 polymerase transcripts were obtained. These gave the GABA responses shown. For the a 3 subunit, the cDNA AbGR9.3 was subcloned as a 3.6 kb Xhol (restriction sites in cDNA adaptors) fragment into Sail cleaved pBS+ (Stratagene). A~2.0-kb Ncol fragment beginning at nucleotide 1,069 was replaced with the equivalent fragment from AbGRl. This allowed the removal of the unspliced intron of the former clone. After linearization of pbGRa3 with HmdIII, capped RNA transcripts were obtained using T7 RNA polymerase. Other details were as in ref. 8. Each RNA (10 ng) was injected into Xenopus oocytes (stages 5 and 6). Oocytes were defolliculated manually after 15-90 min in 2 m g m r ' collagenase in Ca^"^-free modified Earth's medium^", either before or 1 day after the RNA injection. After 2-10 days of incubation in modified Barth's medium supplemented with 2.5 mM pyruvate^", ooctyes were placed in a 120 |xl oblong well, impaled with 2 electrodes filled with 3M KCl and voltage clamped at - 6 0 mV. The preparation was superfused with saline at 6 ml min" ' and GABA was bath-applied with a dead-time of 4 s. No qualitative differences in the GABA responses shown were observed with and without defolliculation. All experiments were performed

on 3-6 oocytes, with consistent results. 1 0 0 - 1

Fig. 3 Exchanging a-subunits alters the sensitivity of the receptor to GABA. Dose-response curves from ooctyes injected with ( a H -/3),(a2-l-)3)or(a3 + )3) were constructed using the peak responses, to minimize the effects of desensitization. The data for (al-l-;8) and ( a3 + /3) were normalized to the response to 1 mM GABA. ( a l + p ) data were normalized to the 300 |jlM response. Each point is the mean from five or six defolliculated oocytes from at least two donors, voltage-clamped at - 6 0 mV. The half-maximal doses were obtained from Hill plots of these data. The half-maximal dose for ( a 1 -I- )8) was significantly different from those for ( a 2 + ) (P<0.001) and (a3 + /3) (P<0 .01) . Maximal responses in C P conductance were: 7.45 ± 2.40 |xS (n = 6) for (a 1 -f yS); 13.8 ±4.7 jaS

( n = 6) for {a2 + py, 67.2±27.9 jjlS (« = 5) for {a3 + p).

size and sequence. Significantly, they confer different apparent sensitivities to GABA in the expressed receptors. Pharmacologi-cal alteration of G A B A a receptor sensitivity has been shown to change both the size and duration of GABAergic synaptic poten-tials in vitro^^. It is therefore likely that neuronal responsiveness to GABAergic input is influenced by differential expression of

-

c O 80 -a M » V.

E 60 -

1 X m E 40 -

^ % • u

20 -w a» a.

0.01 II iiii|

0.1 TTTiTi?—n nrp-

10 100 mn] 1000

[GABA] (^M) a-subunit genes. Northern blot analysis reveals that this expression varies between different brain regions and possibly during development. Thus, differential expression of multiple receptor subtype genes may be used to generate a greater diver-sity of GABA responses than would be achieved with a single subtype. Such differential expression of homologous but func-

o E ox o o E o E

O X O I O o E o E

O I O O o E o E O X O I o o

S % * u

Fig. 4 Regional brain distribution of a-subunit mRNAs. Northern blots of poly(A)"*" RNA from adult, 3-month-old (3m) and 12-day-old (12d) bovine cortex (Co), hippocampus (H) and cerebellum (Cm) were probed with a l - , a2- and a3-specific oligonucleotide probes. Methods. Total RNA was isolated by a modification of the guanidinium/hot phenol method^' and poly(A)'^ RNA was extracted by oligo(dT)-cellulose chromatography. Poly(A)'^ RNA (25 (xg of each preparation) was electro-phoresed in a 1% (w/v) formaldehyde agarose gel, blotted onto a nylon membrane (Hybond-N, Amersham) and UV cross-linked. Probes used were subunit-specific oligonucleotides: a l -specific 45-mer (5'-GGGTCACCCCTGGCCA-GATTAGGTGTGTAGCTGGTTGCTGTTG-GA3'), a2-specific 45-mer (5'-ACTGGGTCTT-TTGAAAGATTCGGGGCATAATTGGCAAC-AGCCACT-3') and a3-specific 45-mer (5'-GGT-GGTTCCCACGATGTTGAAAGTGGTGCTG-GTTTTCTTGGTCGG-3'). Probes were 3' end-

labelled with [^^PJdATP using terminal deoxynucleotidyl transferase to a specific activity of 5 x 10® dpm per |xg. Blots were hybridized in 20% (v/v) formamide at 42 °C and washed in 0.2xSSC, 0.1% (w/v) SDS at 60 °C, before exposure to X-ray film with an intensifying screen. An RNA ladder (BRL) was used for size markers. Autoradiographs were exposed for either 4 h ( a l and a3) or for 24h (al).

- 9 - 5 — 7-5

— 4-5

— 2 - 2

—1-4

— 0 - 2

adult 3m 12d adult 3m 12d adult 3m 12d

t ional ly d is t inct r ecep to r sub types , k n o w n a l ready fo r the G -pro te in c o u p l e d receptors^^"^^ m a y be an i m p o r t a n t m e c h a n i s m tha t con t r ibu tes to synap t i c plast ici ty.

We t h a n k Michae l Squi re fo r he lp with D N A sequenc ing , Jeane t t e O ' B r i e n a n d Br ian Morr i s fo r some SP6 a n d m R N A samples , B r e n d a Shivers f o r t issue samples , Dr s M. G o e d e r t , I. Mar t i n , L. Blair , V. D i o n n e a n d B. Shivers for advice a n d

va luab le d i scuss ions a n d Ju t t a R a m i fo r p r e p a r a t i o n of the manusc r ip t . E.S.L. is s u p p o r t e d by the H u n t i n g t o n ' s Disease Society of A m e r i c a a n d N S F grant INT-8602995, D.R.B. he ld a Fogar ty Senior In t e rna t i ona l Fe l lowship , a n d F.A.S. ho lds a Royal Society Univers i ty Resea rch Fe l lowship . In pa r t , this work was s u p p o r t e d by a g ran t f r o m the Deu t sche Fo r schungsgeme in -schaf t to P .H.S. A u t h o r s E.S.L. a n d P.R.S. are equa l first au thors .

Received 21 March; accepted 19 July

1. Olsen, R. W. & Venter, J. C. (eds) Benzodiazepine/ GABA Receptors and Chloride Channels: Structural and Functional Properties. (Liss, New York, 1985).

2. Braestrup, C. & Nielsen, M. J. J. Neurochem. 37, 333-341 (1981). 3. Squires, R. F. el al. Pharmac. hiochem. Behav. 10, 825-830 (1979). 4. Lippa, A. s . et al. Brain Res. Bull. 14, 189-195 (1985). 5. Cooper, S. J., Karkham, T. C. & Estall, L. B. Trends pharmac. Sci. 8, 180-184 (1987). 6. Akaike, N., Inoue, M. & Krishtal, O. A. J. Physiol. Lond 379, 171-185 (1986). 7. Cash, D. J. & Subbarao, K. Biochemistry 26, 7562-7570 (1987). 8. Schofield, P. R. et al. Nature 328, 221-227 (1987). 9. von Heijne, G. Nucleic Acids Res. 14, 4683-4690 (1986).

10. Grenningloh, G. et al Nature 328, 215-220 (1987). 11. Noda, M. et al. Nature 305, 818-823 (1983). 12. Nef, P., Mauron , A., Stalder, R., Alliod, C. & Ballivet, M. Proc natn. Acad. Sci. U.S.A. 81,

7975-7979 (1984). 13. Shibahara, S. et al. Eur. J. Biochem. 146, 15-22 (1985). 14. Kusano, K., Miledi, R. & Stinnakre, J. J. Physiol. 328, 143-170 (1982). 15. Smart, T. G., Houamed, K. M., Van Renterghem, C. & Constanti , A. Biochem. Soc. Trans.

15, 117-122 (1987). 16. Sigel, E. & Bauer, R. J. Neurosci. 8, 289-298 (1988).

17. Smart, T. G., Constanti , A., Bilbe, G., Brown, D. A. & Barnard, E. A. Neurosci. Lett. 40, 55-59 (1983).

18. Houamed, K. M. et al. Nature 310, 318-321 (1984). 19. Kuriyama, K.. & Taguchi, J. J. Neurochem. 48, 1897-1903 (1987). 20. Sweetnam, P. & Tallman, J. F. Molec. Pharmac. 29, 299-306 (1986). 21. Study, R. E. & Barker, J. L. Proc. natn. Acad Sci. U.SA. 78, 7180-7184 (1981). 22. Unnerstall, J. R., Kuhar, M. J., Niehoff, D. L. & Palacios, J. M. J. Pharm. exp. Ther. 218,

797-804 (1981). 23. de Bias, A. L., Vitorica, J. & Friedrich, P. J. Neurosci. 8, 602-614 (1988). 24. Segal, M. & Barker, J. L. J. Neurophysiol. 52, 469-487 (1984). 25. Nathans, J., Thomas, D. & Hogness, D. S. Science 232, 193-202 (1986). 26. Bonner, T. 1., Buckley, N. J., Young, A. C. & Brann, M. R. Science 237, 527-532 (1987). 27. Peralta, E. G. et al. EMBO J. 6, 3923-3929 (1987). 28. Frielle, T. et al. Proc natn. Acad Sci. U.SA. 84, 7920-7924 (1987). 29. Kozak, M. Cell 44, 283-292 (1986). 30. Barnard, E. A. & Bilbe G. in Netirochemistry: A Practical Approach (eds Turner, A. J. &

Bachelard, H.) 243-270 (IRL, Oxford, 1987). 31. Maniatis, T., Fritsch, E. F. & Sambrook, J. Molecular Cloning—A Laboratory Manual (Cold

Spring Harbor Laboratory, New York, 1982).

Printed in Great Britain by Turnergraphic Limited. Basingstoke. Hampshire

Volume 229, number 1, 206-210 FEB 05639 February 1988

Molecular cloning and amino acid sequence of human enkephalinase (neutral endopeptidase)

Bernard Malfroy, Wun-Jing Kuang°, Peter H. Seeburg*"^, Anthony J. Mason* and Peter R. Schofield*""

Departments of Pharmacological Sciences, ° Molecular Biology and * Developmental Biology, Genentech, Inc.. 460 Point San Bruno Boulevard, South San Francisco, CA 94080. USA

Received 30 November 1987; revised version received 15 January 1988

We have isolated a cDNA clone encoding human enkephalinase (neutral endopeptidase, EC 3.4.24.11) in a /.gtlO library from human placenta, and present the complete 742 amino acid sequence of human enkephalinase. The human enzyme displays a high homology with rat and rabbit enkephalinase. Like the rat and rabbit enzyme, human enkephalinase con-tains a single N-terminal transmembrane region and is likely to be inserted through cell membranes with the majority

of protein, including its carboxy-terminus, located extracellularly.

Enkephalinase; Neutral endopeptidase; Metallo peptidase; cDNA cloning; (Human)

1. INTRODUCTION

Enkephalinase (neutral endopeptidase, EC 3.4.24.11) is a membrane-bound zinc-metallo pep-tidase which cleaves the Gly^-Phe'* amide bond of the opioid peptides, enkephalin's, both in vitro and in vivo [1-3]. Thus, acetorphan, a parenterally ac-tive enkephahnase inhibitor, was shown to display analgesic properties in humans [4,5]. Enke-phalinase was initially identified in rodent brain, but it was soon realized that it is also present in many peripheral organs. In particular, its activity is highest in the kidney, where it was shown to be identical to an insulin B-chain degrading enzyme [6-8] identified several years before [9], the so-called neutral endopeptidase. We have recently

Correspondence address: B. Malfroy, Dept of Pharmacological Sciences, Genentech, Inc., 460 Point San Bruno Boulevard, South San Francisco, CA 94080, USA

* Present address: Laboratory of Molecular Neurobiology, ZMBH, Universitat Heidelberg, Im Neuenheimer Feld 282, D-6900 Heidelberg, FRG

cloned enkephalinase from both rat brain and kidney [10], further demonstrating the co-identity of the brain and kidney enzymes. In addition, enkephalinase has also been recently cloned from rabbit kidney [11].

Enkephalinase activity has been detected in several human tissues including brain [12], neutrophils [13,14], kidney [15] and placenta [16]. We now report the isolation and characterization of a full-length cDNA clone encoding human enkephalinase, constructed from placenta mRNA, and present a comparison of the human, rat and rabbit enzymes.

2. MATERIALS A N D METHODS

A partial rat enkephalinase cDNA clone was used to screen a human placenta cDNA library constructed from polyadenylated mRNA, in AgtlO. This library had been previously used to clone the insulin receptor [17], The 466 base pair fcoRI-fig/II cDNA fragment of the rat clone AK3 [10] was labelled by random priming, and used to screen in duplicate 1.6 X 10^ human placental cDNA clones. Hybridization was carried out at 7>1°C in 2 x SSC, 2007o formamide, and filters were washed at 37°C in 0.5 x SSC, 0.1% SDS. Inserts of A

206 Published by Elsevier Science Publishers B. V. (Biomedical Division)

00145793/88/53.50 © 1988 Federation of European Biochemical Societies

Volume 229, number 1 FEES LETTERS February 1988 phage were subcloned in an M13 derivative and sequenced by largest clone, A H 7 , was Sequenced. Its insert, 3 1 8 1 the chain-termination method [18]. base pairs in length, contains an open reading

frame starting at an ATG initiation codon at base 21, with a TGA stop codon at base 2247 (fig.l).

3. RESULTS AND DISCUSSION This reading frame encodes a 742 amino acid long polypeptide. One of the three cDNA clones we

Using a fragment of rat enkephalinase cDNA identified differs from clone AH7, with an A in clone as a probe, we identified three positive clones place of a G at base 1413 (fig. 1). An A base would in the human placenta library we screened. The change the corresponding codon from an Ala to a

L » S 5 T R G L U S E R G L M M E T A S P ! L £ T H R A S P L L E A S N T H R P R O L Y S P R O L Y S L V S L Y S 6 L . M A R G T R P T H H P R O L E U 6 L U L L E S E R L E U S E R V * I . L T U V A L L E U L E U L E U T H R L T . E L L £ A L ( L V A L T H B H E L L L E A L A L E U T T R A T A T H R

1 G C A A G I C A G A A A G T C A G A I G G A I A T A A C T G A T A I C A A C A C I C C A A A G C C A A A G A A G A A A C A G C G A T a G A C T C C A C T G G A C A T C A G C C T C T C G G T C C T I G T C C T C C T C C T C A C C A I C A I A G C T G I G A C A A I G A I C G C A C I C T A I G C A A C C

5 0 6 0 7 0 8 0 9 0

L Y B A S P A S P G L Y L L E C V S L V S S E R S E R A S P C V S U E L Y S S E R A L A A L A A R G L E U L L E G L H A S N H E T A S P A L A T H R T H R G L U P R O C V S T H R A S P P H E P H E L Y S L V R A L A C Y S G L V G L - Y L R P L E U L V S A R B A S N V A L U E P R O G L U T H B S E R

150 T A C G A T G A I G G I A n i G C A A G T C A T C A G A C T G C A I A A A A T C A G C I G C T C G A C T G A T C C A A A A C A T G G A T G C C A C C A C T G A G C C T r G T A C A G A C I T T T T C A A A T A T G C T I G C G G A G G C r G G T F G A A A C G I A A I G T C A t T C C C G A G A C C A G C

S E R A R G T Y B G L Y A S N P M E A S P L L E L E U A H G A S P G L U L E U G L U V A L V A L L E U L Y S A S P V A L L E U G U N G L U P ( 1 0 L Y S T H R G L . L I A S P ! L E V A L A L A V A L G T N L Y S A L A L Y S A L A L E U T Y R A R G S E R C Y S L L £ A S N G L U S E R A L A L L E A S P S E R

500 T C C C G I t A C G G C A A C T T I G A C A T n i A A G A G A T G A A C I A G A A G T C G T T T r G A A A G A I G T C C T I C A A G A A C C C A A A A C I G A A G A T A T A G T A G C A G T G C A G A A A G C A A A A G C A T I G T A C A G G T C I I G T A I A A A T G A A T C I G C T A T I G A I A G C

1 5 0 1 6 0 1 7 0 1 8 0 1 9 0

A R G G L » G L . V G L U P R O L E U L E U L Y S L E U L E U P R O A S P L L E L Y R G L Y T R P P R O V A L A L A T H R G L U A S N T R P G L U G L N L Y S T Y R 6 L Y A ( . A S E R T R P L H R A L A G T U L Y S A L A L L E A L A G L N L E U A S N S E R L Y 5 T Y R G L Y L Y S L Y S V A L L E U U E

150 A G A G G T G G A G A A C C I C T A C T C A A A C T G T I A C C A G A C A t A T A I G G G T G G C C A G I A G C A A C A G A A A A C T G G G A G C A A A A A T A I G G T G C n C I T G G A C A G C T G A A A A A G C I A T I G C A C A A C T G A A I T C T A A A T A r G G G A A A A A A G I C C I T A T T

200 210 2 2 0 2i0 210

A S N L E u P H t V A L G l . Y T H R A S P A S P L Y S A S K S E R V A L A S K H l S V A l . U E H l S l L E A S P G L N P R 0 A R C L E u G L y L E U P R 0 S E R A R G A S P T Y R l Y R G L u C Y S T M R 6 L Y l L E T Y R L Y S G L U A l . A C Y S T H R A L A l Y R V A L A S P P H E H E T l L E S E R

600 A A T n G T I T G t l G G C A C T G A T G A T A A G A A T T C I G T G A A T C A T G T A A I I C A T A T T G A C C A A C C T C G A C T T G G C C T C C C T T C T A G A G A T r A C T A I G A A T G C A C T G G A A T C T A T A A A G A G G C I T G I A C A G C A I A I G T G G A T T I I A l G A T I T C T

2 5 0 2 6 0 2 7 0 2 9 0

V « L T L « A R G L E U L L E A R G G L F L G L U G L G A R G L E U P R O L L E A S P G L U A S N G L N L E U A L . A L E U G L U H E T A S N L Y S V A L H E T G L U L E U G L . U L Y S G L U L L E A L A A S N A L A T M R A L A L Y S P R O G L U A S P A R G A S N A S P P B O H E R L E U L £ U T Y ( ! A S M

7 5 0 G I G G C C A G A T T G A T T C G T C A G G A A G A A A G A T T G C C C A T C G A T G A A A A C C A G C I T G C T T T G G A A A I G A A T A A A G T I A T G G A A I T G G A A A A A G A A A T I G C C A A T G C T A C G G C T A A A C C T G A A G A T C G A A A T G A I C C A A I G C I R C T G T A T A A C

300 V / / / / / / / A 310 V / / / / / / / / M L ^ 3«0

L Y S M E T T M R L E U A L A G L N U E G L M A S N A S N P H E S E R L E U G L U I L E A S M G L Y L Y S P R O P H E S E R T R P L E U A S N P H E I H R A S N G L U I L E H E T S E R T H R V A L A S N I L E S E R I L E I H R A S N G L U G L U A S P V A L V A L V A L T V R A L A P R O G L U I Y R

9 0 0 A A G A T G A C A I I G G C C C A G A T C C A A A A T A A C T T T T C A C T A G A G A T C A A I G G G A A G C C A I T C A G C T G G I T G A A T T T C A C A A A T G A A A I C A T G I C A A C T G T G A A T A I I A G T A T T A C A A A T G A G G A A G A T G T G G N G T T T A I G C T C C A G A A T A T

3 5 0 5 6 0 3 7 0 3 8 0 ? 9 0

L E U L H R L Y S L E U L Y S P R O L L E L E U L H R L Y S T Y R S E R A L A A R G A S P L E U G L N A S N L E U M E T S E R T R P A R G P H E L L E M E T A S P L E U V A L S E R S E R L E U S E R A R G T H R T Y R L Y S G L U S E R A R G A S K A L A P H T A R G L Y S A L A L E U T Y R G L Y L H R

1 0 5 0 T I A A C C A A A C I T A A G C C C A I T C I I A C C A A A T A T T C I G C C A G A G A I C T T C A A A A T I T A A I G I C C T G G A G A I T C A T A A T G G A T C T T G T A A G C A G C C T C A G C C G A A C C T A C A A G G A G I C C A G A A A I G C T T T C C G C A A G G C C C T T I A I G G T A C A

100 MO H20 AJO NUO T H B S E R G L U I H R A L A I H R T R P A R G A R G C Y S A L A A S N T Y R V A L A S H G U Y A S N M E I G L U A S N A L A V A L G L Y A R O L E U T Y R V A L G L U A L A A L A P H E A L A G L Y G L U S E R L Y S H I S V A L V A L G L U A S P L E U I L E A L A G L N U E A R G G L U V A U P H E

1 2 0 0 A C C I C A G A A A C A G C A A C I T G G A G A C G I I G I G C A A A C T A T G I C A A T G G G A A T A T G G A A A A T G C T G T G G G G A G G C T N A T G I G G A A G C A G C A T R T G C T G G A G A G A G T A A A C A T G T G G I C G A G G A I I T G A I I G C A C A G A T C C G A G A A G T T T I I

'<50 1(60 1170 1 8 0 1 9 0

I L E G L N I H R L E U A S P A S P I E U T H R I R P N E T A S P A L A G L U I H R I Y S L Y S A R G A L A G L U G L U L Y S A L A L E U A L A I L E L Y S G L U A R G I L E G L Y T Y R P R O A S P A S P I L E V A L S E R A S N A S P A S N L Y S L E U A S N A S N G L U I Y R L E U G L U L E U A S N

1 3 5 0 A T T C A G A C N I A G A I G A C C T C A C T I G G A I G G A I G C C G A G A C A A A A A A G A G A G C T G A A G A A A A G G C C T T A G C A A T T A A A G A A A G G A T C G G C T A I C C T G A T G A C A I T G I T I C A A A T G A T A A C A A A C T G A A I A A I G A G T A C C I C G A G I T G A A C

5 0 0 5 1 0 5 2 0 5 3 0 5 1 0

R Y F I L . » S G L U A S P G L U T Y R P M E G L U A S N L L E L L E G L N A S T ( L E U L Y S P H E S E R G L N S E R L Y S G L T L L E U L Y S L » S L E U A R G G L U L Y S V A L A S P L Y S A S P G L U T R P U E S E R G L Y A L A A L A V A U V A L . A S N A L A P H E T Y B S E R S E R G L Y A R G A S N

1 5 0 0 T A C A A A G A A G A I G A A I A C I T C G A G A A C A T A A T T C A A A A T T I G A A A T T C A G C C A A A G I A A A C A A C I G A A G A A G C I C C G A G A A A A G G L G G A C A A A G A I G A G T G G A T A A G T G G A G C A G C T G I A G U A A T G C A T I I T A C I C I T C A G G A A G A A A T

5 5 0 5 6 0 5 7 0 5 8 0 5 9 0

G L N U £ V A L P M E P H O A L A G L Y I L E L E U G L N P R O P R O P H E P H E S E R A I . A G L N G I N S E R A S N S E R L E U A S N I Y H G L Y G L Y U E G L Y M E I V A L 1 L E G L Y H I S G L U I L E T M R H I S G L Y P H E A S P A S P A S N G L Y A R G A S W P H E A S N L Y S A S P G L Y

1 6 5 0 C A G A I A G T C I T C C C A G C C G G C A T R C T G C A G C C C C C C T T C I T T A G T G C C C A G C A G I C C A A C I C A T T G A A C T A T G G G G G C A T C G G C A R G G T C A I A G G A C A C G A A A I C A C C C A I G G C T T C G A T G A C A A T G G C A G A A A C N T A A C A A A G A T G G A

6 0 0 6 1 0 V / / / / / / / A 6 3 0 6II0

A S P L E U V A L A S P L R P T R P L K R G L N G L N S E R A L A S E R A S N P H E L Y S G L U G U K S E R G L N C Y S « E T V A L . T Y R G L N T Y R G T Y A S N P H E S E R T R P A S P L E U A L A G L Y G L Y G L N H L S L E U A S N G L Y L L E A S N L M R L E U G L V G N J A S N L L E A L A A S P

1 8 0 0 G A C C T C G ) T G A C T G G I G G A C I C A A C A G T C T G C A A G T A A C T T T A A G G A G C A A T C C C A G T G C A I G G I G T A T C A G I A I G G A A A C T T T T C C I G G G A C C I G G C A G G T G G A C A G C A C C R T A A I G G A A I I A A I A C A C I G G G A G A A A A C A I I G C T G A I

6 5 0 6 6 0 6 7 0 6 8 0 6 9 0

A S N G L Y G L Y L E U G L Y G L N A L A I Y R A R G A L A T Y R G L H A S N I Y R I L E L Y S L Y S A S N G L Y G L U G L U L V S L E U I E U P R O G L Y L E U A S P L E U A S N H I S L Y S G I H L E O P H E P H E L E U A S N P H E A L A G L N V A L I R P C Y S G L Y I H R T Y R A R G P R O G L U

1 9 5 0 A A I G G A G G I C I I G G I C A A G C A I A C A G A G C C T A T C A G A A T T A T A I T A A A A A G A A T G G C G A A G A A A A A T T A C T T C C T G G A C I T G A C C T A A A I C A C A A A C A A C I A T T I I I C I T G A A C T L T G C A C A G G T G I G G I G R G G A A C C I A L A G G C C A G A G

? 0 0 7 1 0 7 2 0 7 3 0 7 1 0

L Y B A U A V A L A S N 5 E H L L E L Y S L H R A S P V A L H L S S E R P R 0 G L Y A S M P H £ A R G L L E L L E G L Y T M R L E U G L N A S N S E R A ! . A G L U P K E S E R G L U A L A P H E M S C Y S A F L G L Y S A S N S E R L Y R M E L A S N P R 0 G L U L Y S L Y S C Y S A R G V A L . L R P 0 P '

2 1 0 0 L A I G C G G I I A A C T C C A L T A A A A C A G A T G T G C A C A G T C C A G G C A A I T T C A G G A T T A N G G G A C T N G C A G A A C T C I G C A G A G I T I T C A G A A G C C I I I C A C T G C C G C A A G A A T K A T A C A I G A A T C C A G A A A A G A A G T G C C G G G N I G G I G A

2 2 5 0 U T T C A A A A G A A G C A I T G C A G C C C I T G G C T A G A C T T G C C A A C A C C A C A G A A A T G G G G A A T T C T C I A A I C G A A A G A A A A T G G G C C C I A G G G G T C A C T G T A C T G A C I T G A G G G T G A I t A A C A G A G A G G G C A C C A I C A C A A I A C A G A T A A C A T

2 U O O T A C G I I G I C C I A G A A A G G G r G t G G A G G G A G G A A G G G G G T C T A A G G I C T A I C A A G T C A A T C A I T T C T C A C T G T G T A C A t A A I G C T T A A T T I C I A A A G A T A A T A I T A C I G I T T A T I T C T G T T T C I C A T A T G G I C T A C C A G T I I G C I G A T G T C

2 5 5 0 C C T A G A A A A C A A T G C A A A A C C T n G A G G T A G A C C A G G A T T T C T A A I C A A A A G G G A A A A G A A G A I G T T G A A G A A T A C A G T T A G G C A C C A G A A G A A C A G t A G G T G A C A C I A l A G I T I A A A A C A C A T I G C C f A A C I A C T A G I I T I I A C T T T T A

2 7 0 0 I T T G C A A C A T I I A C A G T C C I I C A A A A I C C T T C C A A A G A A T T C T T A T A C A C A T T G G G G C C I I G G A G C T T A C A I A G T I T I A A A C I C A T I T T T G C C A r A C A I C A G T I A T T C A T T C T G I G A T C A T n A T T T I A A G C A C T C I T A A A G C A A A A A A I

2 8 5 0 G A A I G T C I A A A A T I G t I T I t I G T T G T A C C T G C I I I G A C I G A t G C I G A G A T T C I T C A 6 G C T I C C T G C A A I I T T C T A A G C A A T T T C T I G C I C I A T C I C I C A A A A C T I G G T A n T l I C A G A G A ! I T A I A T A A A I G I A A A A A T A A T A A I I I I I A

3 0 0 0 t A r i l A A T l A I T A A C I A C A I I T A T G A G I A A C r A T I A T T A I A G G T A A T C A A I G A A T A T I G A A G I T T C A G C T I A A A A T A A A C A G T I G T G A A C C A A G A T C r A T A A A G C G A T A T A C A G A T G A A A A I T T G A G A C T A I I T A A A C T l A T A A A I C A I A

5150 I I G A t G A A A A G A T T I A A G C A C A A A C T T I A G G G

Fig.l . Nucleotide and deduced amino acid sequence of human enkephalinase. The 8 amino acid stop-transfer sequence PKPKKKQR is indicated by a black bar, and the putative 23 amino acid transmembrane spanning domain is indicated by an open bar. Six potential

jV-linked glycosylation sites are shown by hatched bars. A potential poly(A) addition signal AATAAA is underlined.

207

Volume 229, number 1 FEES LETTERS February 1988

Human Rat Rabbit

Human Rat Rabbit

Human Rat Rabbit

Human Rat Rabbit

Human Rat Rabbit

Human Rat Rabbit

Human Rat Rabbit

Human Rat Rabbit

Human Rat Rabbit

Human Rat Rabbit

Human Rat Rabbit

Human Rat Rabbit

Human Rat Rabbit

Human Rat Rabbit

Human Rat Rabbit

1 10 20 30 40 - K | S E S Q M D 1 T D I N T P K P K K K Q R W T P L E I S L S V L V L L L T ( I A V T M I A L Y A T G R S E S Q M D I T D I N 0 P K P K K K Q R W T P L E L S L S V L V L L L T I I A V T M I A L Y A T G R S E S Q M D I T D I N T P K P K K K Q R W T P L E I S L S V L V L L L T [ V | I A V T M I A L Y A T

50 60 70 80 90 Y O D G I C K S S D C I K S A A R L I Q N M D A ^ ^ E P C T O F F K Y A C G G W L K R N V I P E T S Y D D G I C K S S D C I K S A A R L I Q N M D A Q A E P C T O F F K Y A C G G W L K R N V I P E T S Y D O G I C K S S D C I K S A A R L I Q N M D A T A E P C T D F F K Y A C G G W L K R N V I P E T S

1 0 0 110 120 130 140 S R Y [ G J N F D I L R D E L E V [ ^ L K D V L Q E P K T E O I V A V Q K A K^F^L Y R S C I N E S A I D S S R Y S N F D I L R D E L E V L L K O V L Q E P K T E D L V A V Q K A K T L Y R S C I N E S A I D S S R Y S N F D L L R D E L E V I L K D V L Q E P K T E D I V A V Q K A K T L Y R S C F ^ N E I T I A I D S

150 1 6 0 170 180 190 R G G [ ^ P L L K L L P D I Y G W P V A T E J N W E Y G I J ^ S W T A E K [ ^ I A Q L N S K Y G K K V L I R G G Q P L L 0 L L P D I Y G W P V A [ S J Q N W E Q T Y G T S W T A E K S I A Q L N S K Y G K K V L I R G G Q P L L K L L P D M Y G W P V A T Q N W E Q T Y G T S W T S I A E K S I A Q L N S F I ^ Y G K K V L I

200 V G T D O K N S V G T D D K N S V G T D D K N S

210 220 230 240 V | N H[X | I H I D Q P R L G L P S R D Y Y E C T G I Y K E A C T A Y V D F M I T ^ H I I H @ D Q P R L G L P S R D Y Y E C T G I Y K E A C T A Y V D F M I M I N RI I I H I D Q P R L G L P S R D Y Y E C T G I Y K E A C T A Y V D F M I

250 260 270 280 290 S V A R L I R Q E E R L P I D E N O L | A J L E M N K V M E L E K E I A N A T J ^ K P E D R N D P M L L Y S V A R L I R Q E [ Q 1 R L P I D E N Q L S L E M N K V M E L E K E I A N A T T K P E O R N D P M L L Y A I V A H ^ L I R Q E ^ ^ L P I D E N O F T L S F V L E M N K V M E L E K E I A N A T T K [ S 1 E D R N O P M L L V

300 310 320 330 340 N K M T L A N N F S L E I N G K P F S v | j j N F T N E l M S T V N I [sjT N K M T L A jK L | Q N N F S L E I N G K P F S W S N F T N E I M S T V N I N l N K M T L A Q I O N N F S L E l N G K P F S W S N F T N E I M S T V N I N l

N E E D V V V Y A P E N E E @ V V V Y A P E N E E D V V V Y A P E

350 360 370 380 390 Y L T K L K P I L T K Y S L ^ R D L Q N L M S W R F I M D L V S S L S R T Y K E S R N A F R K A L Y G Y L T K L K P I L T K Y S P R D L Q N L M S W R F I M D L V S S L S R 0 Y K E S R N A F R K A L Y G Y LFILK L K P I L T K Y [ F ] P R D[F1O N LFFIS W R F I M D L V S S L S R T Y K I ^ S R N A F R K A L Y G

400 410 420 430 440 T T S E T A T W R R C A N Y V N G N M E N A V G R L Y V E A A F A G E S K H V V E D L I A O I R E V T T S E T A T W R R C A N Y V N G N M E N A V G R L Y V E A A F A G E S K H V V E D L I A Q I R E V T T S EFS]A T W R R C A N Y V N G N M E N A V G R L Y V E A A F A G E S K H V V E D L I A Q I R E V

450 ^ 0 470 480 490 F l Q T L D D L T W M D A E T K K [ R J A E E K A L A I K E R I G Y P D D I V S N O N K L N N E Y L E L F l Q T L O D L T W M D A E T K K K A E E K A L A I K E R I G Y P D D I |T]s N [ ^ N K L N N E Y L E L F l Q T L D D L T W M D A E T K K K A E E K A L A I K E R I G Y P D D I V S N D N K L N N E Y L E L

500 510 520 530 540 N Y K E D E Y F E N I I Q N L K F S Q S K Q L K K L R E K V O K D E W I S G A A V V N A F Y S S G R N Y K E [ ^ E Y F E N I I Q N L K F S Q S K Q L K K L R E K V O K D E W I S G A A V V N A F Y S S G R N Y K E D E Y F E N I I O N L K F S Q S K Q L K K L R E K V D K D E W I [ T I G A A I T I V N A F Y S S G R

550 560 570 580 590 N Q I V F P A G I L Q P P F F S A Q Q S N S L N Y G G I G M V I G H E I T H G F D D N G R N F N K O N Q I V F P A G I L Q P P F F S A @ Q S N S L N Y G G I G M V I G H E I T H G F D D N G R N F N K O N Q I V F P A G L L Q P P F F S A Q Q S N S L N Y G G I G M V I G H E I T H G F D D N G R N F N K O

600 610 620 630 640 G D L V D W W T Q Q S A[SJN F K E Q S Q C M V Y Q Y G N F S W D L A G G Q H L N G I N T L G E N I A G O L V D W W T Q Q S A N N F K 0 O S Q C M V Y Q Y G N F Q W O L A G G Q H L N G I N T L G E N I A G D L V D W W T Q Q S A N N F K E Q S Q C M V Y Q Y G N F S W D L A G G Q H L N G I N T L G E N I

650 660 670 680 690 D N G G [ I J G Q A Y R A Y Q N Y | J J K K N G E E K L L P G L D L N H K Q L F F L N F A Q V W C G T Y R P D N G G I G Q A Y R A Y O N Y V K K N G E E K L L P G L D L N H K Q L F F L N F A Q V W C G T Y R P D N G G I G O A Y R A Y Q N Y V K K N G E E K L L P G F T I D L N H K Q L F F L N F A Q V W C G T Y R P

700 710 720 730 740 E Y A V N S I K T D V H S P G N F R I I G T L Q N S A E F S E A F H C R K N S Y M N P E K K C R V W E Y A V N S T K T D V H S P G N F R I I G T L Q N S A E F | A D | A F H C R K N S Y M N P E @ K C R V W E Y A V N S I K T D V H S P G N F R I I G I S I L Q N S F V I E F S E A P F ^ C F P L K N S Y M N P E K K C R V V Y

Fig.2. Comparison of the amino acid sequences of human, rat and rabbit enkephalinase. The amino acid sequences of human, rat and rabbit enkephalinase are shown in this order. Numbering of amino acids refers to the human and rat sequences.

208

Volume 229, number 1 FEBS LETTERS February 1988 Thr. Since both the rat and rabbit enkephalinases contain an Ala in this position (Ala'^^^, fig-2), the presence in this clone of an A is probably due to an error of reverse transcriptase of the mRNA,

The rat enkephalinase gene contains two poten-tial A T G initiation codons, and amino-terminal protein sequence analysis has shown that the second is used [10]. Clone A H7 does not extend far enough to the 5 ' -end of the message to include the first ATG. Since neither of the other two clones we identified contained any additional 5'-sequence, we do not know if, like the rat, the human enkephalinase gene contains a first initiation codon that would be 4 bases upstream from the 5 ' -end of clone AH7. The amino-terminal sequence of human enkephaUnase has not been established, but, by analogy with the rat protein, we propose that the ATG at base 21 initiates the translation of human enkephalinase, with an aspartyl residue at its amino-terminus. We have therefore numbered the amino acid sequence of human enkephalinase starting from this aspartyl residue. However, the first ATG of the rabbit enkephalinase gene is used for initiation [11], therefore this assignment must remain tentative.

Human enkephalinase displays a high degree of homology with the rat enzyme. Of the total of 742 amino acids, 700 (94%) are conserved and, of the 42 differences, only 6 are non-conservative changes (fig.2). The rabbit enzyme also shows high homology to rat and human enkephalinase (93%) but, while no insertions a r t needed to optimally ahgn the rat and human sequences, the rabbit en-zyme contains an additional amino acid, a valine residue after Phe^^^ (fig.2).

One other important difference between human, rat and rabbit enkephalinase concerns potential N-linked glycosylation sites. Both rat and human enkephalinase contain six potential glycosylation sites (Asn-X-Ser/Thr), while the rabbit enzyme has five. The site missing in the rabbit, involving Asn^®^, is also missing in the human enzyme, but in the latter there is a sixth site, at Asn^^"' (f ig. l) . The significance of these differences is not clear. However, the fact that rat and human enkephalinase have been shown to be identical with respect to a number of properties, including optimal pH, sensitivity to inhibitors and kinetic parameters for the hydrolysis of a number of substrates [12,13,15,16], suggests that these dif-

ferences do not affect the enzymatic activity of the enzyme.

Several important domains of enkephalinase are fully conserved between rat, rabbit and human proteins. The highly charged, conformationally restrained fragment PKPKKKQR (residues 8-15), which we proposed to serve as a stop transfer se-quence [10], the hydrophobic region which follows this fragment (residues 21-43), which is likely to be a transmembrane spanning domain, as well as two fragments (residues 568-580 and 627-635), which have been proposed to serve as binding sites for the zinc atom of this metallo peptidase [11] are con-served in the three species. In addition, all cysteine residues present in the rat protein, which may be involved in disulfide bridges, are also found in the human and rabbit enzyme. These results suggest that enkephalinase is highly conserved among mammalian species.

Acknowledgement: We thank Dr Ellson Chen for his help and advice during DNA sequencing.

REFERENCES

[1] Malfroy, B., Swerts, J.P., Guyon, A.. Roques, B.P. and Schwartz, J.-C. (1978) Nature 276, 523-526.

[2] Schwartz, J.-C., Malfroy, B. and De La Baume, S. (1981) Life Sci. 29, 1715-1740.

[3] Hersh, L.B. (1982) Mol. Cell. Biochem. 47, 35-43. [4] Floras, P., Bidabe, A.M., Caille, J.M., Simonnet, G.,

Lecomte, J.M. and Sabathie, M. (1983) Am. J. Neuroradiol. 4, 653-655.

[5] Lecomte, J.M., Costentin, J., Vlaiculescu, A., Chaillet, P., Marcais-Collado, H., Llorens-Cortes, C., Leboyer, M. and Schwartz, J.-C. (1986) J. Pharmacol. Exp. Therap. 237, 937-944.

[6] Almenoff, J., Wilk, S. and Orlowski, M. (1981) Biochem. Biophys. Res. Commun. 102, 206-214.

[7] Fulcher, I.S., Matsas, R., Turner, A.J. and Kenny, A.J. (1982) Biochem. J. 203, 519-522.

[8] Malfroy, B. and Schwartz, J.-C. (1982) Biochem. Biophys. Res. Commun. 106, 276-285.

[9] Kerr, M.A. and Kenny, A.J. (1974) Biochem. J. 137, 477-488.

[10] Malfroy, B., Schofield, P., Kuang, W.J., Seeburg, P., Mason, A.J. and Henzel, W.J. (1987) Biochem. Biophys. Res. Commun. 144, 59-66.

[11] Devault, A., Lazure, C., Nault, C., Le Moual, H., Seidah, N.G., Chretien, M., Kahn, P., Powell, J., Mallet, J., Beaumont, A., Roques, B.P., Crine, P. and Boileau, G. (1987) EMBO J. 6, 1317-1322.

209

Volume 229, number 1 FEES LETTERS February 1988

[12] Llorens, C., Malfroy, B., Schwartz, J.-C., Gacel, G., Roques, B.P., Roy, J., Morgat, J.L., Javoy-Agid, F. and Agid, Y. (1982) J. Neurochem. 39. 1081-1089.

[13] Connelly, J.C., Skidgel, R.A., Schulz, W.W., Johnson, A.R., Erdos, E.G. (1985) Proc, Natl. Acad. Sci. USA 82, 8737-8741.

[14] Yuli, I. and Snyderman, R, (1986) J. Biol. Chem. 261, 4902-4908.

[15] Gafford, J.T., Skidgel. R.A.. Erdos, E.G. and Hersh, L.B. (1983) Biochemistry 22. 3265-3271.

[16] Johnson. A.R., Skidgel, R.A.. Gafford. J.T. and Erdos. E.G. (1984) Peptides 5. 789-796.

[17] Ullrich, A.. Bell. J.R., Chen. E.Y., Herrera, R., Petruzzelli, L.M., Dull, T.J., Gray, A., Coussens, L., Liao, Y.-C., Tsubokawa, M., Mason, A., Seeburg, P.H., Grunfeld, C., Rosen, O.M. and Ramachandran, J. (1985) Nature 313, 756-761.

[18] Sanger, F.. Nicklen, S. and Coulson, A.R. (1977) Proc. Natl. Acad. Sci. USA 74, 5463-5467.

210

Soluble proteins that bind hydro-phobic odorants belong to a

larger family of lipophilic-molecule carrier proteins. The odorant-binding proteins are uniquely ex-pressed in nasal tissues and are thought to play a key intermediary role in the concentration and deliv-ery of airborne odorants.

Olfactory reception in verte-brates occurs via the interactions of odorants with olfactory receptor cells located in the distal nasal 'epithelium. These neuronal cells are bipolar, possessing an axonal projection to the olfactory bulb glomeruli and a dendritic projection bearing the chemosensory cilia. The cilia are embedded within the olfactory mucosa and are believed to be the sites of expression of olfactory receptors^'^. The chemo-sensory cilia contain an odorant-sensitive adenylate cyclase^ as well as cyclic nucleotide-gated ion channels'^. Direct ion channel acti-vation by odorants has also been observed. Odorant activation of these signal transduction systems, suggests a mechanism in which olfactory receptors, coupled to G proteins (guanine nucleotide regu-latory proteins), transduce the pri-mary event of odorant binding into cellular responses^-^. Indeed, human olfactory dysfunction (anos-mia) is correlated with Gs protein deficiency® (G^ is the stimulatory G protein that activates adenylate cyclase). These data suggested a strong similarity between the mechanisms of olfactory and visual transduction and led to the widely held view that olfactory receptors are members of the G protein-coupled receptor superfamily^-^.

Using more functionally based experiments, Pelosi and co-workers described a protein that bound a labelled pyrazine deriva-tive (the potent green pepper odorant). This protein, termed odorant-binding protein (OBP), has been purified to homogeneity and shown to be a dimer of two identical ~19kDa subunits®-' . The correlation between odorant bind-ing affinities of the protein and human odour detection thresholds is consistent with a physiological role for this protein in olfaction. Primary sequence homology was originally seen between the puri-fied bovine OBP and a family of

Carrier-bound odorant delivery to olfactory receptors urinary proteins of unknown func-tion®. Moreover, this homology includes a much larger family of proteins (see below).

The recent cDNA cloning of the rat OBP^ has demonstrated that this protein is a member of the lipophilic-molecule carrier protein superfamily. This family^® includes soluble proteins that serve as car-riers for small lipophilic molecules including cholesterol, steroids and retinol. For example, retinol is transported to the retina of the eye by means of a series of carrier (retinol-binding) proteins. Like the highly lipophilic molecule retinol, many odorants are also very lipo-philic. Thus, in humans, odorant detection ability increases with the increase in chain length of related alcohols. The involvement of lipo-philic carrier proteins is especially interesting, since both the ol-factory and visual transduction systems appear to have many features in common, including now, the delivery of retinol or odorants to the receptor by a lipophilic-molecule carrier protein. Indeed, all of the mammalian lipophilic-

molecule carrier proteins identified to date appear to have an associ-ated receptor.

A cDNA encoding a protein with homology to retinol-binding protein has also been isolated by selection for frog olfactory epithelium-specific mRNAs^^. This protein is expressed specifically in Bowman's glands (BG), the source of olfac-tory mucosa. It was suggested that the BG protein could have a non-specific role in binding hydrophobic odorant molecules. Interestingly, the urinary proteins, to which OBP was first seen to share homology, have been suggested as carriers of mammalian pheromones®. A solu-ble pheromone-binding protein has been identified in insects^^ and it will be of interest to determine whether olfaction in vertebrates and invertebrates occurs in a similar manner.

Figure 1 shows the three-dimensional structure of retinol-binding protein ^ which, based on other structural determinations^''' ^ protein sequence alignments^" '"^® and secondary structure calcu-lations, would predict the three-

Peter R. Schofield Centre for Molecular Biology, Im Neuenheimer Feld 282, University of Heidelberg, 6900 Heidelberg FRG.

Fig. 1. The three-dimensional structure proposed for the lipophilic carrier protein family. The figure shows the structure determined for the human retinol-binding protein (after Newcomer et a l . ) " showing the eight ft-sheets (ribbons) and the a-helix (coiled). A hydrophobic molecule is shown inside the pocket of the protein. The hydrophobic or uncharged residues that line the hydrophobic core and lie closest to retinol in the retinol-binding protein are indicated by filled circles.

TINS, Vol. 11, No. 11, 1988 11988, Elsevier Publications. Cambridge 0378 - 5912/88/S02.00 471

Acknowledgements dimensional structure of the other The author thanks members of this protein family. Brenda Shivers for Lipophil ic-molecule ca r r i e r p ro -

critically reading the teins are characterized by the manuscript, presence of an eight-stranded up-

and-down (3 barrel core that encap-sulates the lipophilic molecule. This is further illustrated in Fig. 1 by the filled circles, which show the locations of the amino acids that line the hydrophobic core of the carrier protein.

It appears that there is only a single class of odorant-binding pro-t e i n ® ' T h i s suggests that these proteins play a general role in odorant delivery to the olfac-tory receptors. In this regard, it would be particularly interesting to examine completely anosmic patients to determine if such a phenomenon is due to the absence of OBP.

If lipophilic carrier proteins bind odorants, how are the odorants delivered to the olfactory recep-tors? Pevsner et al. determined the sites of expression of the rat OBP and showed selective local-ization to the lateral nasal gland (LNG). The main secretory duct of the LNG extends to the ostium, the narrowest place through which

inspired air passes as it enters the nose. LNG secretions appear to be atomized at the tip of the nose, allowing airborne odorants to be bound within droplets contain-ing the OBP. Bound odorants are then carried to the olfactory epi-thelium where they traverse the mucosa, delivering the odorant molecules to the ciliary receptors. The role of the OBP carrier pro-tein thus appears to be in delivery and concentration of volatile air-borne odorants. Whether the pro-tein plays an active role in odorant-specific receptor binding remains to be determined.

The molecular structure of olfac-tory receptors and the mechan-isms by which odorants are perceived remain unknown. How-ever, the carrier protein model provides a mechanism for concen-trating and delivering airborne odorants to the olfactory recep-tors. The availability of OBP anti-bodies and the cDNA clone en-coding OBP should facilitate a more detailed analysis of this model.

Selected references 1 Anholt, R. R. H. (1987) Trends

Biochem. Sci. 12, 58-62

2 Lancet, D. and Pace, U. (1987) Trends Biochem. Sci. 12, 63-66

3 Pace, U., Hanski, E., Salomon, Y. and Lancet, D. (1985) Nature 316, 255-258

4 Nakamura, T. and Gold, G. H. (1987) Nature 325, 442-444

5 Weinstock, R. S., Wright, H. N., Spiegel, A.M., Levine, M. A. and Moses, A . M . (1986) Nature 322 635-636

6 Bignetti, E. et al. (1985) Eur. J. Biochem. 149, 227-231

7 Pevsner, J., Trifiletti, R, R., Strittmat-ter, S. M. "and Snyder, S. H. (1985) Proc. Natl Acad. Sci. USA 82. 3050-3054

8 Cavaggioni, A., Sorbi, R. T.. Keen, J. N., Pappin, D. J. C. and Findiay, J. B.C. (1987) FEBS Lett. 212, 225-228

9 Pevsner, J., Reed, R. R., Feinstein, P. G. and Snyder, S. H. (1988) Science 241, 336-339

10 Pervaiz, S. and Brew, K. (1985) Science 228, 335-337

11 Lee, K-H., Wells, R. G. and Reed, R. R. (1987) Science 235, 1053-1056

12 Vogt, R. G. (1987) in Pheromone Biochemistry (Prestw/ich, G. D. and Blomquist, G. J., eds), pp. 473-527, Academic Press

13 Newcomer, M. E. etal. (1984) EAABOJ. 3,1451-1454

14 Sawyer, L. (1987) Nature 327, 659 15 Godovac-Zimmermann, J. (1988)

Trends Biochem. Sci. 13, 64-66 16 Pevsner, J., Hwang, P. M., Sklar, P. 8.,

Venable, J. C. and Snyder, S. H. (1988) Proc. Natl Acad. Sci. USA 85, 2383-2387

Hilary Anderson Department of

Zoology, University of California, Davis, CA 95616, USA.

Drosophila adhesion molecules and neural development

How do neurons navigate cor-I -rectly through the complex

tissues of the embryo to reach their appropriate targets? Studies on vertebrate neurons in vitro have shown that differential ad-hesive interactions between a growth cone and the surfaces it contacts can dramatically influence the rate and direction of neuron outgrowth. Subsequently, many molecules to which neurons can adhere and over which they can extend processes in vitro have been isolated from vertebrate em-bryos. These include molecules from the extracellular matrix and from the surfaces of neuronal and non-neuronal cells. The relative importance of these different ad-hesion molecules in guiding neuron outgrowth in vivo is still poorly understood but is being intensively investigated in many laboratories by the application to the embryo of antibodies and synthetic peptides that might block interaction of

growth cone receptors to specific adhesion molecules.

In recent months, a flurry of papers has shown that all major classes of adhesion molecules so far identified in vertebrates have their equivalents in the fruit fly Drosophila. This pro\ades concrete evidence to justify the belief that the basic mechanisms underlying neuronal pathfinding, and indeed of development in general, are the same in insects and vertebrates. It also ushers in a new era of research on adhesion molecules in which the particular genetic advan-tages of the fruit fly can be ex-ploited to identify rapidly the genes that code for these adhesion mol-ecules. The genes may then be selectively removed or altered, an approach that has so far not been possible in vertebrates.

The research on Drosophila ad-hesion molecules is only just begin-ning. Currently our knowledge is restricted largely to descriptions of

the adhesion molecules them-selves and of their expression and distribution in the embryo. How-ever, in the firm belief that experi-mental results of interest to all those working on nervous system development will soon follow, this account attempts to bring together the scattered references on D?vs-ophila adhesion molecules and to indicate some of the directions in which this research is now moving.

Subs t r a tum adhes ion molecules

Like that of vertebrates, the Drosophila extracellular matrix con-tains collagens, glycoproteins, and proteoglycans. Their possible role in promoting neuron outgrowth has been reviewed in an earlier TINS article'. One Drosophila collagen has been biochemically isolated from medium conditioned by the Drosophila Kc cell line". At leasi seven collagen genes have been identified by cross-hybridization with chicken collagen cDNA' "". Partial sequences for one of these, DCgl, indicate considerable hom-

472 (c) 1988, Elseviei Publications. Cambridge 0378 - 5912/88/S02.00 TINS, Vol. 11, No. 11, 1988

Reprint Series Q r > T T 7 1 V T 0 1 ? 2 December 1988, Volume 242, pp. 1306-1308 O L ^ l I L i l V y t l j

Transient Expression Shows Ligand Gating and AUosteric Potentiation of GABAA Receptor Subunits

DOLAN B . PRITCHETT, HARALD SONTHEIMER, CORNELIA M . GORMAN, HELMUT KETTENMANN, PETER H . SEEBURG,* AND PETER R . SCHOFIELD

Copyright © 1988 by the American Association for the Advancement of Science

Transient Expression Shows Ligand Gating and AUosteric Potentiation of G A B A A Receptor Subunits

D O L A N B . PRITCHETT, HARALD S O N T H E I M E R , C O R N E L I A M . GORMAN, H E L M U T KETTENM[ANN, PETER H . S E E B U R G , * PETER R . S C H O F I E L D

Human 7-aminobutyric acid A (GABAA) receptor subunits were expressed transiently in cultured mammalian cells. This expression system allows the simultaneous charac-terization o f ligand-gated ion channels by electrophysiology and by pharmacology. Thus, coexpression of the a and (5 subunits of the GABAA receptor generated GABA-gated chloride channels and binding sites for GABAA receptor ligands. Channels consisting of only a or p subunits could also be detected. These homomeric channels formed with reduced efficiencies compared to the heteromeric receptors. Both of these homomeric GABA-responsive channels were potentiated by barbiturate, indicating that sites for both ligand-gating and allosteric potentiation are present on receptors assembled f rom either subunit.

T " I H E G A B A A RECEPTOR CONTAINS

an intrinsic chloride ion channel that is opened (gated) by GABA, the

major inhibitory neurotransmitter in mam-malian brain. GABAA receptor activation stabilizes the neuron's resting potential ( i , 2). Channel activity can be allosterically modulated by therapeutically useful drugs, for example, barbiturates and benzodiaze-pines {1-3). The afiinity-purified receptor contains an a subunit of 53 kD, on which the benzodiazepine binding site is thought to be located, and a 3 subunit of 57 kD, which can be photoaffinity labeled by GABA agonists {"1-6).

Human ai and pi subunit-encoding cDNAs were isolated from a fetal brain library (7). The D N A and predicted poly-peptide sequences of these cDNAs reveal a high degree of conservation with the corre-sponding bovine sequences (1). The cloned

and Pi subunit cDNAs were inserted, singly or together (in tandem array), into a eukaryotic expression vector, pCIS2, that contained the human cytomegalovirus pro-

motor-enhancer (8). Human embrj^onic kid-ney cells (8) were transfected with these constructs at high efficiencies, by using a modified CaP04 precipitation technique (9, 10). Cells were harvested 48 hours after transfection for R N A isolation, for pharma-cological studies, or were used for electro-physiology. Analysis of total R N A from transfected cells showed high levels of GABAA receptor subunit m R N A but no such mRNA was seen in untransfected or mock-transfected cells (Fig. 1).

To characterize the transiently expressed heteromeric a i -I- Pi receptors, membranes

D. B. Pritchctt, P. H. Secburg, P. R. Schofield, Labora-tory of Molecular Neurocndocrinolog)', ZMBH, Univer-sity of Heidelberg, Im Neucnheimer Feld 282, 6900 Heidelberg, Federal Republic of Germany. H. Sontheimer and H. Kettenmann, Department of Neurobiology, University of Heidelberg, Im Neuen-heimer Feld 364, 6900 Heidelberg, Federal Republic of Germany. C. M. Gorman, Department of Cell Genetics, Genen-tech, 460 Point San Bruno Boulevard, South San Fran-cisco, CA 94080.

*To whom correspondence should be addressed.-

1306 SCIENCE, VOL. 242

prepared from transfected cells {11) were analyzed by binding of [^H] muscimol, a GABA analog, and [^^S]f-butyl-bicyclo-phosphorodiionate (TBPS), a GABAA re-ceptor channel-specific cage convulsant {12). Muscimol bound with a dissociation constant (Ka) of 13 ± 3 nM (n = 3) (Fig. 2), demonstrating the presence of a high-affinity binding site for Ae neurotransmitter on the recombinant a j + pi receptors. This value is close to that determined for purified bovine GABAA receptor (Ka = 12 ± 3 nM) (4). Muscimol had an average maximum binding (B^ax) of 23 ± 3 fmol per 10^ cells {n = 3), corresponding to approximately 6000 receptors per expressing cell. Similarly, TBPS displayed specific binding with B^ax values of 18 ± 3 fmol per 10^ cells, based on a Kd of 25 nM {12). Less than 2 fmol per 10^ cells of ligand ([^H] muscimol or [^^S]TBPS) bound to cells transfected with single subunit cDNAs. These values are not significantly diffisrent from those obtained

- 2 8 S

-18S

Fig. 1. Blot analysis of RNA from transfected and untransfected cells. Total RNA and polyadenylat-ed [poly(A)'^] RNA were prepared from 293 ceUs transfected with pCIS2 plasmid DNA containing a , + Pi, ai , or Pi cDNAs, respectively. No con-taminating Pi or ai CDNA sequences ( < 0 . 0 1 %) were found in the respective ai or Pi vector preparations. RNA was analyzed by electrophore-sis in a 1% agarose/7% formaldehyde denaturing gel and transferred to nitrocellulose. The blot was probed (A) with both ai and Pi, (B) with Pi, and (C) with tti subimit-specific oligonucleotides (7). Total RNA (5 |xg) was loaded in lanes 1 to 4 and 2 Jig of poly(A)'^ RNA was loaded in lane 5 to further demonstrate the absence of endogenous GABAA receptor subunits in the recipient cell line. Lanes 1 and 5, untransfected cells; lane 2, a , -transfected; lane 3, ai -f p,-transfected; and lane 4, Pi-transfected.

for nonspecific binding (Fig. 2). This may result from reduced numbers of assembled receptors or from the presence of a site of lower affinity than in ai + Pi-transfected cells. No benzodiazepine binding ([^H]flu-nitrazepam or ^H-labeled Rol5-1788) was observed. This correlates with a lack of potentiation of GABA-induced a] -h (3] channel activity by benzodiazepines {13).

Cells transiendy expressing a j + (3] re-ceptors were analyzed further with the whole-cell patch clamp technique {14). In 65% {n = 127) of die tested cells {10, 14), GABA (10"^M) induced an inward current at a resting potential of - 6 0 mV (Fig. 3). The average current response was 400 pA and the largest response was 1 nA {15). No response was seen in over 40 untransfected or vector-transfected cells. The reversal po-tential of the GABA response, determined by plotting GABA-induced currents against membrane potential under various ionic conditions, indicated that the current was conducted by C P ions.

The GABA agonist muscimol (lO^^^M) (n = 7) mimicked the effect of GABA by mducing an inward current at the resting potential of - 6 0 mV. Picrotoxin (10~''M) (« = 7), a specific channel blocker, and bi-cuculline, a competitive GABA antagonist (lO-^M) (n = 12), blocked the GABA-in-duced inward current (Fig. 3A). The barbi-turate pentobarbital (50 |JLM) increased the GABA-induced currents (Fig. 3A). The ef-fect of the barbiturate was most prominent at low (<10"^M) GABA concentrations. Pentobarbital alone did not trigger channel activity, as described in cultured spinal neu-rons {16). GABA-induced inward currents were observed at concentrations as low as 10~" M GABA in the presence of pentobar-bital, whereas GABA alone was in most cases not effective at such low concentra-tions. These results, in conjunction with those obtained from ligand-binding studies, demonstrate that transient mammalian cel-lular expression allows for the pharmacolog-ical and electrophysiological analvsis of li-gand-gated ion channel receptors.

GABA-induced current flow was also ob-served in cells expressing ai or Pi subunits alone (Fig. 3, B and C). S.uch GABA-responsive channels were not seen on trans-fection with , vector alone, but were specifi-cally generated by the expression of either of the single subunit cDNAs {17). The whole-cell currents of the homomeric channels were approximately 10% of those seen with heteromeric receptors; the average satura-tion response was 40 pA and the maximal response was 100 pA at lO' ^M GABA (Fig. 3, B and C). Currents were detected with GABA doses as low as 10"^M, and the reversal potentials for the homomeric chan-

10 15 20 25 30 35 40 45 50

I^H] Muscimol (nM)

Fig. 2. Binding isotherm and Scatchard plot of [^Hjmuscimol binding to ai + Pi-transfected cell membranes. Specific binding ( • ) (11) was determined by subtracting the binding in the presence of 1 |JLM GABA. Untransfected cells showed no specific binding ( • ) . The results shown {Kd = 13 nM, Bmax = 25 nM) were from a single transfection experiment and were deter-mined by least-squares analysis of the data. As obser\'ed from three different transfections, K^ values were 13 ± 3 nM and Bmax values were 23 ± 3 fmol.

nels indicate CI" ion flow. The low whole-cell current obtained by expression of homo-meric channels may reflect reduced efficien-cies of receptor assembly, because cells trans-fected with cDNAs encoding a single subunit synthesize the respective mRNA in amounts as high as do doubly transfected cells (Fig. 1, B and C). A change in single-channel characteristics, shorter lifetimes for example, may also contribute to the ob-served reduction in whole-cell current. Ho-momeric channel formation has not been previously detected in voltage-clamped oo-cs'tes {1, \3).

Our finding that GABAA receptors, as-sembled from either of the known subunits, can form ligand-gated CI" channels suggests that homomeric assembly into fiinctional channels may be a general propert}' of manv of the subunits of ligand-gated ion channels {17) since it is also true for one subunit of the neuronal nicotinic acet}'lcholine receptor {18). Our results indicate that a site con-served in both the a. and p subunits, possibly of lower apparent affinit}' than the unique photoaffinit}'-labeled agonist binding site on the p subunit (6), may be responsible for agonist gating of the channel. We propose that this site is located extracellular)' in a region encompassing the 85-amino acid residues from the invariant disulfide-bonded P-structural loop to the first transmembrane segment {1). In addition, the parallel phar-macolog}' of the ai and Pi subunit homo-meric channels and the heteromeric GABAA

2 DECEMBER 1988 R E P O R T S 1 3 0 7

al +pi

B al

c pi

100pA[

— — „ — — „

r

GABA {\iM) 0.5

100

100

GABA GABA+ GABA+ 30 s PB PicroTx

Fig. 3. Pharmacology of the GABA-induced cur-rent in cells transfected with either a j + Pi sub-units or the tti subunit or the subunit alone. (A) AppUcation of GABA (5 x lO^^M) to a cell transfected with the a , H- subunits and volt-age-clamped at - 6 0 mV gave rise to an inward current (left). This current was amplified and increased in duration in the presence of pentobar-bital (PB) (50 |xM) (middle). After a 5-min washout the current returned to the control level. P i c r o t o ^ (PicroTx) at a concentration of 5 X 10 " M (right) almost completely blocked the GABA-induced current. The same experimental protocol was applied to a cell transfected with the ai subunit alone (B) or with the (3, subunit alone (C). Note that the concentration used to elicit a detectable inward current in the cells expressing only a , or p, subunits was IQ'^^M GABA (left traces). Pentobarbital (50 |xiW) amplified (center traces) and picrotoxin (5 x lO^'^A^ blocked the inward current (right traces) in cells expressing either the a i or the subunit.

receptor indicates that allosteric sites for barbiturate potentiation are also present on receptors formed from either subunit.

Our results demonstrate the usefulness o f expressing neurotransmitter-gated ion chan-nels transiendy in mammalian cells. Tran-sient expression, rather than stably express-ing cell lines {19), should provide for the rapid and simultaneous electrophysiological and pharmacological characterization o f many ligand- and voltage-gated ion chan-nels.

G A G G C C G C T C G T C T C G T T C C T G A T C T C -C G G G T A C T G A G G A G A A T G T T G C C G T G -3'. Full-length cDNA clones were identified by DNA sequence analysis and subcloned into the expression vector pCIS2 {8).

8. D. L. Eaton et al., Biochemistry 25, 8343 (1986). For a detailed description of the host-vector sys-tem see C. M. Gorman, D. Gies, G. McGray, and M. Huang {J. Virol., in press).

9. C. Chen and H. Okayama, Mol. Cell Biol. 7 2745 (1987).

10. Control transfections with the Escherichia coli lacZ gene under simian virus 40 (SV40) early promoter control were assayed with the chromogenic substrate 5-bromo-4-chloro-3-indolyl-p-D-galactopyranoside (X-Gal), and color formation was determined on parafomialdehyde-fixed cells [J. R. Sanes, J. L. R. Rubenstein, J.-F. Nicolas, EMBO J. 5, 3133 (1987)]. About 40% of die cells were transfected (9). In die 293 cell line, positively staining cells were more prevalent in small cell groups, consistent with the notion that only actively dividing cells (diat is, groups of cells) are capable of transient DNA uptake and subsequent expression of tiiis DNA.

11. Forty-eight hours after transfection, cells from ten plates (10 cm) were washed twice with phosphate-buffered saline (PBS) and scraped into PBS (10 ml). After low-speed centrifugation (500^) the cell pellet was homogenized in a Polytron tissue homogenizer (Brinkmann) in 10 ml of 50 mM potassium phos-phate, pH 7.4, and centrifiiged at 50,000^ for 20

• nun. Membrane pellets were frozen at - 2 0 ° C over-night and the wash procedure was repeated three times. The final pellet was resuspended in potassium phosphate buffer, pH 7.4, containing 100 mM KCl. Homogenate equivalent to 10® cells (100 fjig of protein) was incubated in a 1-ml reaction volume widi [^HJmuscimol (Du Pont, 23 Ci/mmol) for 60 min at 4°C or ['^S]TBPS (Du Pont, 70 Ci/mmol) for 90 min at 27°C. Samples were filtered on GF/B filters with a Bio-Rad vacuum filter apparatus and washed twice with 5 ml of potassium phosphate-KCl buffer. After drying, filter-retained radioactivity was determined by liquid scintillation.

12. R. F. Squires, J. E. Casida, M. Richardson, E. Saederup, Mol. Phartnacol. 23, 326 (1983).

13. E. S. Lesntan et ai. Nature 335, 76 (1988). 14. Membrane currents were recorded with the patch-

clamp technique in the whole-cell configuration [O. P. Hamill, A. Marty, E. Neher, B. Sakmann, F. I.

Sigwortii, Pfluegers Arch. 391, 85 (1981)]. Pipettes contained 130 mM CsCl, 1 mM MgClz, 0.5 mM CaClz, 5 mM EGTA, and 10 mM Hepes. The pK was adjusted to 7.2. Cultures were continuously perfiised with a bathing (control) solution contain-ing 5.4 mM KCl, 116 mM NaCl, 0.8 mM MgClz, 1.8 mM CaClz, 11.1 mM o-glucose, 26 mM NaHCOj, and 10 mM Hepes. The pU was adjusted to 7.2. GABA was added at the concentrations indicated. Cells were only tested if they formed a gigohm seal and had a resting potential more nega-tive than - 4 0 mV.

15. Electrophysiological responses showed a wide varia-tion around the mean. The 293 cells grow in groups and, by Lucifer yellow injection, are electrically coupled. Thus, it is likely that GABA-induced mem-brane currents generated in transfected cells can also be recorded from an electrically coupled untransfect-ed cell. This would increase the number of apparent-ly transfected cells and may, in part, explain the wide variation of response amplitudes.

16. J. L. Barker and B. R. Ransom, J. Physiol. (London) 280, 331 (1978); B. Sakmann, O. P. Hamill, J. Bormann, J . Neural Transm. Suppl. 18, 83 (1983); J. Bormann and D. E. Clapham, Proc. Natl. Acad. Sci U.S.A. 82, 2168 (1985).

17. These channels were also seen in HeLa cells express-ing single GABAA receptor subunits, suggesting that no cellular component generates these channels or leads to the formation-ef-heteromeric GABAA receptors. Similar homomeric channels have recentiy been observed in oocytes with the use of high-sensitivity electrophysiological techniques [L. A. C. Blair, E. S. Levitan, J. Marshall, V. E. Dionne, E. A. Barnard, Science 242, 577 (1988)]. In addition, glycine-gated, GABA-insensitive homomeric recep-tor channels were formed when either cells (D. B. Pritchett et al., in preparation) or oocytes (H. Betz, personal communication) expressing the 48-kD rat glycine-receptor subunit were analyzed.

18. J. Boulter et ai, Proc. Natl. Acad.'Sci. U.S.A. 84 7763 (1987).

19. T. Claudio et al.. Science 238, 1688 (1987), 20. We thank B. Wingbermiihle for preparing plasmid

DNA, H. Steininger for assistance with cell culture, and J. Rami for typing the manuscript. Supported in part by the Deutsche Forschungsgemeinschaft, SFB 317 grants B/8 to H.K. and B/9 to P.H.S.

3 June 1988; accepted 9 September 1988

REFERENCES AND NOTES

1. P. R. Schofield ei ai. Nature 328, 221 (1987). 2. R. W. Olsen and J. C. Venter, Bemodiazepine-

GABA Receptors and Chloride Channels: Structural and Functional Properties (Liss, New York, 1986).

3. A. J. Turner and S. R. Whittle, Biochem. J. 209, 29 (1983).

4. E. Sigel, F. A. Stephenson, C. Mamalaki, E. A. Barnard, J. Biol. Chem. 258, 6965 (1983).

5. H. Mohlcr, M. K. Battersbv, J. G. Richards, Proc. Natl. Acad. Sci. U.S.A. 77, 1666 (1980).

6. D. Cavalla and N. H. Neff, J. Neurochem. 44, 916 (1985); L. Deng, R. W. Ransom, R. W. Olsen, Biochem. Biophys. Res. Commun. 138, 1308 (1986); S. O. Casalotti, F. A. Stephenson, E. A. Barnard, J. Biol. Chem. 261, 15013 (1986).

7. Human oi and pi subunit cDNA clones were isolated from a human fetal brain cDNA library with subunit-specific radiolabeled oligonucleotides: ai subunit, 5 ' - A C C C T G G C C A G A T T A G G T G -T G T A G C T G G T T G C T G T T G G A - 3 ' ; p, sub-unit, 5 ' -TCCCACGCCCGTGAGCACTTCA-

1308 S C I E N C E , V O L . 2 4 2

NEURORECEPTORS AND SIGNAL TRANSDUCTION

Edited by

Shozo Kito and Tomio Segawa Hiroshima University Hiroshima, Japan

Kinya Kuriyama Kyoto Prefectural University of Medicine Kyoto, Japan

Masaya Tohyama Osaka University Osaka, Japan

and

Richard W. Olsen University of Cal i fornia , Los Angeles Los Angeles, Cal ifornia

PLENUM PRESS • NEW YORK AND LONDON

Library of Congress Cataloging in Publication Data

International Symposium on Neurotransmitter Receptors (1987: Hiroshima-shi, Japan)

Neuroreceptors and signal transduction / edited by Shozo Kito . . . (etal.J.

p. cm. —(Advances in experimental medicine and biology; v. 236)

"Proceedings of the International Symposium on Neurotransmitter Receptors, held

October 6-9, 1987, in Hiroshima, Japan" —T.p. verso.

Includes bibliographies and index.

ISBN 0-306-42985-3

1. Neurotransmitter receptors —Congresses. I. Kito. Shozo, date. II. Title. III . Series.

[DNLM: 1. Neural Transmission —congresses. 2. Neuroregulators —congresses. 3.

Receptors, Sensory—congresses. 4.Synaptic Receptors —congresses.W1 AD559 v. 236

/ WL 102.8 1615 n 1987]

QP364.7.I59 1987

599'.0I88-dcI9

D N L M / D L C 88-22525

for Library of Congress CIP

Proceedings of the International Symposium on Neurotransmitter Receptors,

held October 6-9, 1987, in Hiroshima, Japan

© 1988 Plenum Press, New York

A Division of Plenum Publishing Corporation

233 Spring Street, New York, N.Y. 10013

All rights reserved

No part of this book may be reproduced, stored in a retrieval system, or transmitted

in any fprm or by any means, electronic, mechanical, photocopying, microfilming,

recording, or otherwise, without written permission from the Publisher

Printed in the United States of America

MOLECULAR BIOLOGY OF THE GABA^ RECEPTOR

E.A. Barnard^, M.G. Darlison^, N. Fujita\ T.A. Glencorse^,

E.S. Levitan^, V. Reale^, P.R. Schofield^, P.H. Seeburg^,

M.D. Squire^, and F.A. Stephenson^

MRC Molecular Neurobiology Unit, MRC Centre Hills Road, Cambridge, CB2 2QH, England

z Laboratory of Molecular Neurendocrinology, ZMBH

D-6900 Heidelberg, Federal Republic of Germany

The GABA receptor is the major molecular site of the ubiquitous inhibitory activities of the brain, being present on the great majority of mammalian brain neurones (1). Electrophysiological studies and, especially, recent patch-clamp studies on cultured neurones (2,3) have established that at these sites GABA opens a chloride channel which is integrally associated with its receptor; Further, the GABA receptor at brain synapses is known to be a site of action of several pharmacologic-ally important classes of drugs. Pharmacological and ligand-binding studies (reviewed in Ref. A) have identified at least 5 types of binding site on this receptor: (i) the GABA agonist/antagonist site; (ii) the benzodiazepine site, which itself is complex, having interactions with both anxiolytic agonists and anxiogenic "inverse agonists" (5); (iii) the picrotoxin site, where agents such as picrotoxin (5) or £-butyl-bicyclophosphorothionate (6) block the GABA-activated channel; (iv) the depressant site, recognising the CNS-depressant barbiturates and certain other depressant drugs which prolong the lifetime of the GABA-activated channel (3); this site, also, appears to be multiple, since certain steroids (7), the anaesthetic propanadid (8) and avermectin B^^ (9,10) act similarly to depressants in some but not all ways; (v) sites^binding the channel-permeating anions (but not other ions) (2,4). Each of these types of ligand site can interact allosterically with one or more of the other types (4). From this network of interactions, it can be deduced that several .of these sites can be occupied by their respective ligands

31

simultaneously and that each of the 5 or more types of site must be physically separate on the receptor structure.

THE PURIFIED GABA, RECEPTOR PROTEIN A The unitary protein structure postulated has, indeed, been

identified. Purification of the receptor in an appropriate detergent could be accomplished on a benzodiazepine affinity column (11) and the single protein finally obtained binds ligands for all of the' sites listed above (12,13). The rank order of drug potencies in these classes is preserved from the ^ vitro state to the membrane, to the purified protein state (13,14). The purified protein also shows the charac-teristic allosteric interactions between sites, although these are quantitatively lowered compared to the membrane state (12).

In the purified receptor from bovine (12,14) or chick (15) brain, two subunit classes are detectable, a(M 53,000) and 6(M 57,000). Their stoichiometry appears (15) to be this is not yet certain. The a band contains at least two a-type polypeptides, al and a2, and the 6 band overlaps another of the a class, a3. In the pure receptor the a subunits can be photoaffinity-labelled by flunitrazepam, while the B subunit is the site of photoaffinity-labelling by muscimol (16,17). The two binding sites for GABA, shown electrophysiologically as required to be occupied before channel opening occurs (18,19), are, therefore, on the 6 subunits and the benzodiazepine binding sites are on the a subunits.

MOLECULAR CLONING AND EXPRESSION

Cloning of the cDNAs encoding the receptor subunits

The GABA receptor protein from bovine cerebral cortex, purified in CHAPS/phospholipid medium (12), was used either in its entirety or after separating by gel electro-elution the a subunit. Selective cleavage by cyanogen bromide or by trypsin, and HPLC separation of the peptides produced, were followed by their gas-phase micro-sequencing (20). A set of oligonucleotide probes corresponding to various segments of the two subunits was thus designed and used for the screening of calf cortex or adult bovine brain cDNA libraries, constructed in phage AgtlO. It was necessary to employ specific oligonucleotide primer extension in the final stage to obtain the complete 5' end of the B subunit cDNA sequence. Full-length cDNAs encoding the a subunit and the B subunit were ultimately obtained (20), these being definitively identified by the chemically-determined peptide sequences encoded within them.

The deduced polypeptides of the GABA receptor each have, before the mature N-terminus, a typical short "signal sequence" which can be assumed to be removed by cleavage after membrane insertion. The calculated molecular weights of the mature a and 3 subunits are 48,800 and 51,400 daltons respectively. Values of 53,000 and 57,000 daltons were deduced from protein gels, as noted above; the difference is ascribed partly to several thousand daltons of carbohydrate subsequently attached vivo to each subunit (15) and partly to the known anomalous migration in denaturing gels of such membrane-bound polypeptides.

32

Expression in the Xenopus oocyte system

Proof that all of the cDNAs encoding the GABA^ receptor had been cloned has been obtained by translation of the corresponding RNAs in the Xenopus oocyte system. This system, as applied to the special case of receptor mRNA identification, was developed previously (21-24). Earlier, the correct assembly of the GABA^ receptor in this system was shown: when- mRNA extracted from the rat or chick brain was micro-injected into Xenopus oocytes, the receptor and its ion channel appeared in the cell membrane (4,18,25). Therefore, the cloned a and 3 cDNAs were used as templates to synthesise the corresponding RNAs in yttro, using the bacteriophage SP6 RNA polymerase/ transcription system. The two pure RNAs were micro-injected, together, into the Xenopus oocyte. The application of a low concentration of GABA produced a large and immediate conductance change in the cell membrane (Fig. 1). Control (non-injected or medium-only injection) oocytes never gave any response to GABA. After 2 days incubation high receptor/ ion channel densities were obtained.

The expressed receptor showed the characteristic sensitivities to bicuculline and picrotoxin, and the potentiation (very marked) by pentobarbital. Some of these responses are illustrated in Fig.l. A much larger quantity of either the a- or the e-subunit RNA injected alone produces similar receptor channels in the oocyte membrane (46). The results show that the cloned a and B cDNAs together encode a structure which has many of the known properties of the GABA receptor and its anion channel, while either subunit alone can produce a quite similar, but presumably artefactual receptor channel.

The channel opened has been identified as a chloride channel of the same type as that (2,3) opened in the native GABA receptor on neurones. The evidence for this is:

(1) The reversal potential in normal amphibian Ringer solution was found (20,47) to be abaut -25 mV (at 20°C), which agrees with the equilibrium potential of chloride in the Xenopus oocyte in these conditions.

Further screening produced cDNAs encoding, al, a2 and a3 subunits (47), from different genes, that exhibit approximately 80% amino—acid identity (the original a subunits). When co-expressed in the oocyte with B, their RNAs produce similar receptors, which differ functionally. Thus, the half-maximal dose for GABA is 1.2MM for a2+B, 12pM for al+6 and 42JJM for a3+B. These isoforms also differ in their regional distribution in the brain (47). Thus, subtypes of the GABA receptor can arise in the brain as different, homologous gene products.

(2) When the external chloride concentration was varied (in further studies made by E. Levitan) the reversal potential varied linearly, with a slope of -58 mV for a 10-fold increase, as predicted for chloride permeation.

(3) The series of relative anion permeabilities which has been determined here for the expressed receptor channel compares closely with that reported (2,3) for the GABA receptor channel on cultured neurones.

33

A :

\J r -1mM

lOOnA

B : -0-3MM -J 308

% ^

-ImM -1mM + SpM Bic ImM • lOgM Bic

-3MM •3MM + 5MM PTX -3(iM

80 min wash

D :

25MM PB Figure 1. Expression of the GABA^ receptor in the Xenopus oocyte. Each series of GABA responses was recorded from an individual oocyte co-injected with the al- and B-subunit-specific RNAs derived from the cloned cDNAs. The duration of the GABA application is indicated by the horizontal bar. A, Membrane conductance change evoked by 1 pM (i) or 0.3 jjM (ii) GABA. B, Membrane conductance change evoked by: (i) 1 pM GABA; (ii) 1 yM GABA plus 5 pM bicuculline methobromide (Bic); (iii) 1 ^M GABA plus 10 mM bicuculline methobromide. C, Membrane conductance change evoked, in another RNA-injected oocyte, by: (i) 3 jjM GABA; (ii) 3 jjM GABA plus 5 ^M picrotoxin (PTX); (iii) 3 jjM GABA, after washing out the picrotoxin for 80 min. D, Membrane conductance changes evoked by: (i) 1 ;jM GABA; (ii) 1 pM GABA plus 25 pM (-) pentobarbital (PB). From ref. 20, which see for further details.

34

TABLE 1. Percentage amino-acld sequence homology of subunlts in the receptor/ion channel super-family.

(i) Identical residues;

GABA a GABA & GLY

GABA a - 35% 3A%

GABA 6 - - 39%

nAChR a (muscle) 19% 15% 15%

nAChR a (brain) 15% 17%

(ii) Identical plus conservatively-substituted^

residues:

GABA a GABA 6 GLY A8K

nAChR a ' 38% 32% 37%

GABA a - 57% 56%

GABA B _ _ 59% a

The sequences of the mature polypeptides of the bovine GABA receptor(20), the rat glycine receptor 48,000-M subunit (GLY), the' bovine muscle nAChR a subunit (30) and the a-A n^ChR subunit from rat brain (28) are compared. Optimized alignments of all the pairs were made using the computer program DIAGON (see ref. 20). b

Conservatively-substituted residues are defined as those in which the mutation data matrix (MDM^g) score (31) is greater than or equal to 0.1.

35

PREDICTED STRUCTURE OF THE GABA, AND RELATED RECEPTORS A

To interpret the amino-acid sequences of the GABA receptor sub-units, hydropathy plots were generated from them by a suitable computer program (Fig. 2, top and middle panels). These provide a quantitative estimate of hydrophobic (above the line) or hydrophilic (around or below the line) character of the side-chains along the sequence.

Candidates for membrane-spanning segments are those where a clear peak of hydrophobicity occurs over a region of the order of 20 amino acids, i.e. sufficient to form a helix which can span the bilayer. Four of these putative transmembrane regions are seen, at the same relative positions in each of the subunits, being designated Ml, M2, M3 and M4. Since a signal peptide sequence is present in the cDNAs, the N-termini are predicted to be extracellular, and the region from the mature N-terminus to the start of Ml is presumed to lie outside the membrane.

It is instructive at this point to compare the sequences and hydropathy profiles with those now known for subunits of other receptors of the same class, i.e. the transmitter-gated ion channels. These are available for the nicotinic acetylcholine receptor (nAChR) and the glycine receptor. For the latter from rat spinal cord, very recently a cDNA for one of its subunits (48,000 daltons, strychnine-binding) has been cloned and sequenced by H.Betz and co-workers (26). The hydropathy plot which we can construct from their deduced amino acid sequence is strikingly similar to that of each GABA^ receptor subunit (Fig.2). There is also a strong similarity to that of the bovine (or other muscle) nAChR a subunit (20). Indeed, in all of the nAChR subunits an equivalent four hydrophobic segments, Ml-MA, at the same relative positions in the chain as in Fig. 1, have generally been recognized and regarded as membrane-spanning a-helices (27).

Alignment for maximum homology of the sequences of the subunits of these 3 receptor types reveals that the homology between the GABA and glycine receptor subunits is quite high, considering that they are from different receptors (Table 1). The overall homology between the sequences of GABA or glycine receptor subunits on the one hand and muscle or electric organ nAChR subunits on the other hand is low, of, the order of 15% for identities (Table 1). This rises to 30-40% for conservative substitutions plus identities, but this homology is not spread throughout the sequence but is localized in the Ml, M2 and M3 regions and some structures in the assigned N-terminal domain (1). One can ask whether this low (although significant) degree of homology between the GABA receptor and nAChR subunits is due to the difference that the latter are not from the central nervous system. However, a few nAChR sequences are now available from rat or chicken neuronal sources, and these do not show more homology to the brain GABA or glycine receptors. An illustration is given in Table 1 in the case of a subunit analyzed at the DNA level (28) from rat brain.

The homology in amino-acid sequence and in domain distribution leads us to conclude that the transmitter-activated ion channels together form a protein super-family. One can speculate that the excitatory amino-acid receptors, when similarly subjected to DNA. cloning, will turn out to be members of this same super-family. We should note that the GABA^ receptor sequences show no detectable homology with any known protexn sequence stored in the data-bases, other than, in this super-family. This receptor super-family can be compared in its generality to the recently discovered super-family (reviewed in ref. 29) of G-protein-linked receptors (6-adren^rgic, muscarinic and opsin tjrpes) .

36

100 200 300 400

100 200 300 400

- 2 0

-401-

100 200 300

Amino acids

400

Figure 1, Hydropathy profiles of the GABA^ and glycine receptor subunit sequences. Hydropathy profiles of the mature polypeptides (i.e. excluding proposed signal sequences) were computed according to Kyte and Doolittle (45) using a window size of 17 residues and plotted with a 1-residue interval. Solid bars indicate the positions of the proposed hydrophobic transmembrane domains.

37

A region in the N-terminal domain which shows stronger homology between the GABA^ and the other receptor types in this superfamily is found around a pair of cysteine residues, at positions 139 and 153 in the GABA^ receptor a-subunit sequence. These are maintained in all of the known subunits of the three receptor types. These residues are known to form a disulfide bond vivo in the nAChR (32,33). It was concluded (20) that such a loop structure also exists in the other ligand-gated receptor subunits. Identical residues are found in this region, at five positions in all the known subunits of these receptors-, while another seven residues there are always strongly hydrophobic (Fig. 3) . An invariant proline halfway along the loop is predicted to be in a hairpin 6-turn, and the whole loop forms a 6-structure (in a scale model) , with one face completely hydrophobic and. perhaps interacting with the invariant tyrosine at position 162.

This B-loop contains an N-linked glycosylation site in all of the sequenced vertebrate peripheral nAChR subunits (which is known to be used ^ vivo (32,33)) and also in the bovine GABA^ receptor B subunit, but not in its a subunit nor in the glycine receptor subunit. The B-loop constitutes a conserved structural motif for this group of receptors which is probably of functional significance. It is interesting that this region in the peripheral nAChRs has been implicated, from antibody-mapping studies (34), in its ACh binding. While other evidence had indicted the region near Cys 192-Cys 193 in ACh binding (35,36) these two regions may be close and both may contact the bound transmitter, but it should be noted that the disulfide postulated to be important there in the nAChR, and likewise the Cys 198-Cys 209 disulfide {postulated for the glycine receptor (28), do not denote a general feature of the trans-mitter binding in this family because cysteines are absent from this entire region in the GABA^ receptor subunits. We can speculate, rather, that the aspartate residue at position 146 (in the B-subunit) , which has been found to be invariant in all of this super-family to date, may provide the charge which binds the positive pole of ACh, GABA or glycine to their receptors. The smaller side-chain of aspartic acid and its situation on the polar face of a loop stabilised on a hydrophobic core would give a suitable fixed location for this site. This speculation can now be tested by point mutation there and oocyte expression.

A model for the structure in the membrane

Based upon the cDNA-derived sequences, a model for the GABA^ receptor in its neuronal membrane has been proposed(20). In Fig. 4, a view cutting through the molecule is shown, with one of the a and one of the B subunits illustrated. The 4 hydrophobic segments in each are assumed to be transmembrane helices. The structure of the large N-terminal extracellular domain is drawn in an arbitrary manner. Two cytoplasmic loops are shown on each subunit, one long and one very short. The location and relative sizes of the two such loops in models (reviewed in Ref. 37) of the nAChR turn out to be equivalent to these. On the GABA^ B subunit the major loop is longer and contains a concensus sequence for a site for cAMP-dependent phosphorylation (marked P) . Such a reaction is likely to mediate some form of modulation of the receptor, and should occur on the intracellular side.

Since the M4 segment of the B subunit as assigned ends only one residue from the C-terminus, a variant on the structure of Fig- 3 should be considered as a possible alternative, in which M4 lies as an a-helix in and parallel to the cytoplasmic surface of the membrane (not illustrated) . That type of helical arrangement occurs at the C-terminus of the one integral membrane protein so far subjected to high-resolution crystallographic analysis, namely the reaction centre,, protein of the

38

chloroplast membrane (39). It cannot be excluded at present that any one of the Ml-MA regions has this disposition. Indeed, Guy and Hucho (37) and Lindstrom and co-workers (35) have suggested, based upon the mapping (35,39) of various antigenic sites in the nAChR relative to the membrane, models with the M4 helix lying in the cytoplasmic plane of the membrane.

In published discussions of the nAChR structure, there has been much support (reviewed in refs. 27,37) for the existence of a fifth membrane-spanning region (MA), a postulated amphipathic a-helix (with one face polar and the other hydrophobic) to line the channel. Such a segment with an approximate 3.6-residue period of polar (mostly charged) side-chains can be identified in most of the known subunit sequences of the nAChR. We must note that, neither by inspection nor by using the hydrophobic moment analysis of Eisenberg (40), can we identify a putat-ive amphipathic segment anywhere in the GABA receptor a or 3 subunit structures. A sequence of the MA type xs likewise absent in the glycine receptor 48,000-M subunit (26). An amphipathic transmembrane helix cannot be a general feature of chemically-gated ion channels.

Conserved structural motifs

The alignment of the 3 sequences reveals several striking features in relation to this model. One of these, the disulfide-bridged loop (Fig. 4) has been noted above. Secondly, a remarkable similarity of hydroxy-rich amino-acid sequence is observed within the M2 domain in both the GABA and the glycine receptor namely the sequence Thr-Thr-Val-Leu-Thr-Met-Thr-Thr-X-Ser (where X is any residue). It is noteworthy that this same domain, M2, has been implicated by affinity-labelling (41, 42) or mutagenesis (43) results as forming part of the lining of the ion channel in the nAChR. While a similar sequence does not occur in any of the 20 nAChR subunits (peripheral and neuronal) known, several hydroxy-amino acids do occur in M2 there, also. We tentatively suggest, there-fore, that this stretch contributes to the ion selectivity of the channel.

Thirdly, in the Ml segment a proline occurs at its eighth position from the predicted extracellular end. This is invariant in all of the GABA , glycine and nAChR subunits known, including an invertebrate (44) nAChR. A proline well within an a-helix is rare and gives a bend in it of 20 - 25°. It has been suggested (20) that the backbone bond to this prolyl group gives a flexibility employed in the conformational change which opens the channel in the GABA receptor.

A

Fourthly, the model (Fig. 4) shows a high concentration and excess of positively-charged amino-acid side-chains in the immediate vicinity of the ends of the proposed transmembrane domains of both GABA^ receptor subunits. The same is true for the subunit of the glycine receptor. In contrast, most of these positions are occupied by neutral or negatively-charged residues in the nAChR. Clusters of arginines and lysines at the channel mouth, therefore, are proposed to act, in both the GABA and the glycine receptors, as an anion-concentrating device and to increase the driving force for anion flow upon opening of the channel.

We can speculate that the hydroxy—groups on the M2 and MI helices maintain the water content of the channel, and that the angles between these lumenal helices (or segments thereof) are altered in the agonist-activated conformational change, to open the limiting aperture. Point mutation and expression in the oocyte test system, as well as anti-peptide antibody site mapping (now in progress) should allow these possibilities for transmembrane topology and the channel architecture to be tested experimentally.

39

Figure 3. A loop-containing structure common to the GABA^ and vertebrate nicotinic ACh receptor subunits. This loop is formed by presumed disulphide bonding between Cys 139 and Cys 153 (GABA^ a-subunit numbering). Positions identical (or with one highly conservative variant, as shown) in all nAChR subunits and the B-subunit of the GABA^ receptor are marked with uppercase symbols; h denotes that a hydrophobic residue only is found at that position. A hypothetical interaction between the C-terminal constant Tyr and hydrophobic side-chains of the loop may occur. In the one invertebrate nAChR subunit sequence known, all of the positions marked are occupied similarly except for 152(T), 154(1) and 162(F) (GABA receptor a-subunit numbering). In the bovine GABA receptor a-subunit all of this also holds, except that the three residues underlined differ from those in the 6-subunit. The disulphide-bridged loop can readily form (in a model) a 6-structure, with a hairpin turn centred at the invariant Pro. From ref, 20.

40

HoN H.N

EXTHRACELLJULAR

^ i'k i

.. J "fef -

r TRACELUULAR

a p Figure A. A schematic model for the topology of the GABA. receptor in the membrane. Four membrane-spanning helices in each subunit are shown as cylinders. The structure in the extracellular domain is drawn in an arbitrary manner, but the presumed 3—loop formed by the disulphide bond predicted at cysteines 139 and 153 (a-subunit numbering) is shown. Potential extracellular sites for N-glycosylation are indicated by triangles and a possible site for cAMP-dependent serine phosphorylation, present only in the 6 subunit, is denoted by an encircled P. Those charged residues which are located within or close to the ends of the membrane-spanning domains are shown as small circles with charges marked. It is proposed that two such structures are complexed in the receptor molecule so as to align the membrane-spanning domains, only some of which will form the inner wall of a central ion channel. From ref. 20.

41

References

1. S. J. Enna, (1983): In: The GABA Receptors, edited by S. J. Enna, pp. 1-23. Humana Press, Clifton, New Jersey.

2. J. Bormann, 0. P. Hammill, and B. Sakmann, (1987): Mechanism of anion permeation through .channels gated by glycine and Y-aminobutyric acid in mouse cultured spinal neurones. J. Physiol., 385:243-286.

3. J. L, Barker, and D. G, Owen, (1986): Electrophysiological pharmacology of GABA and diazepam in cultured CNS neurons. In Reference 4, pp. 135-166.

4. R. W. Olsen, and J. C. Venter, editors (1986): Benzodiazepine/ GABA Receptors and Chloride Channels: Structural and Functional Properties. Alan R. Liss, New York.

5. C, L. Brown, and I. L. Martin, (1985): Modification of pyrazoloquinolinone affinity by GABA predicts efficacy at the benzodiazepine receptor. Eur. J. Pharmacol., 106:167-173.

6. C. Van Renterghem, G. Bilbe, S. Moss, T. G. Smart, A. Constanti, D. A, Brown, and E. A. Barnard, (1987): GABA receptors induced in Xenopus oocytes by chick brain mRNA: evaluation of TBPS as a use-dependent channel-blocker. Mol. Brain Res., 2:21-31.

7. J. J. Lambert, J. A. Peters, and G. A. Cottrell, (1987): Actions of synthetic and endogenous steroids on the GABA receptor. Trends Pharmacol. Sci., 8:224-227.

8. E. F. Kirkness, and A. J. Turner, (1986): The y-aminobutyrate/benzodiazepine receptor from pig brain. Biochem. J., 233:259-:264.

9. G. Drexler, and W. Sieghart, (1984): Evidence for association of a high affinity avermectin binding site with the benzodiazepine receptor Eur. J. Pharmacol., 99:269-277.

10. E. Sigel, and R. Baur, (1987): Effect of avermectin on chick neuronal GABA receptor channels expressed in Xenopus oocytes. Mol. Pharmacol, (in press).

11. E. Sigel, C. Mamalaki, and E. A. Barnard, (1982): Isolation of a GABA receptor from bovine brain using a benzodiazepine affinity column. FEBS Letts, 147:45-48.

12. E. Sigel, and E. A. Barnard, (1984): A y-aminobutyric acid/ benzodiazepine receptor complex from bovine cerebral cortex. Improved purification with preservation of regulatory sites and their regulations. J. Biol. Chem., 259:7219-7223.

13. E. Sigel, F. A. Stephenson, C. Mamalaki, and E. A. Barnard, (1984): The purified GABA/benzodiazepine/barbiturate receptor complex: Four types of ligand-binding sites, and the interactions between them are preserved in a single isolated protein complex, J, Recept. Res., 4:175-188.

14

42

F. A. Stephenson, C. Mamalaki, S. 0. Casalotti, and E. A- Barnard, (1986): The GABA receptor and its antibodies. Biochem. Soc. Symp,, 52:33-40.

15. C. Mamalaki, F. A. Stephenson, and E. A. Barnard, (1987): The GABA /benzodiazepine receptor is a heterotetramer pf homologous a and 6 subunits. EMBO J., 6:561-565.

16. S. 0. Casalotti, F. A. Stephenson, and E- A. Barnard, (1986): Separate subunits for agonist and benzodiazepine binding in the Y-aminobutyric acid receptor oligomer. J. Biol. Chem., 261:15013-15016.

17. L. Deng, R. W. Ransom, and R. W. Olsen, (1986): [^HjMuscimol photolabels the y-aminobutyric acid receptor binding site on a peptide subunit distinct from that labelled with benzodiaze-pines. Biochem. Biophys. Res. Commun., 138:1308-1314.

18. T. G. Smart, K. M. Houamed, C. Van Renterghem, and A. Constanti, (1986): mRNA-directed synthesis and insertion of functional amino acid receptors in Xenopus laevis oocytes. Biochem. Soc. Trans. 15:117-122.

19. J. Bormann, and D. E. Clapham, (1985): y-Aminobutyric acid receptor channels in adrenal chromaffin cells: A patch-clamp study. Proc. Natl. Acad. Sci. USA., 82: 2168-2172.

20. P. R. Schofield, M. G, Darlison, N. Fujita, D. R. Burt, F. A. Stephenson, H. Rodriguez, L. M. Rhee, J. Ramachandran, V. Reale, T. A. Glencorse, P. H. Seeburg, and E. A. Barnard, (1987): The brain GABA receptor: cloning and functional expression of the cDNAs encoding its subunits. Nature, 328:221-227.

21. K. Sumikawa, M. Houghton, J. S. Emtage, B. M. Richards, and E. A. Barnard, (1981): Active multi-subunit ACh receptor assembled by translation of heterologous mP-NA in Xenopus oocytes. Nature, 292:862-864.

22. E. A. Barnard, R. Miledi, and K. Sumikawa, (1982): Translation of exogenous messenger RNA coding for nicotinic acetycholine receptors produces functional receptors in Xenopus oocytes. Proc. Roy. Soc. Lond. B, 215:241-246.

23. T. G. Smart, A. Constanti, G. Bilbe, D. A. Brown, and E. A. Barnard, (1983): Synthesis of functional chick brain GABA-benzodiazepine barbiturate/receptor complexes in mRNA-injected Xenopus oocytes. Neurosci. Lett., 40:55-59.

24. E. A. Barnard, D. Beeson, G. Bilbe, D. A. Brown, A. Constanti, K. Houamed, and T. G. Smart, (1984): A system for the translation of receptor messenger-RNA and the study of the assembly of functional receptors. J. Recept. Res., 4:681-704.

25. K. Houamed, K. Constanti, T. G. Smart, G. Bilbe, D. A. Brown, E. A, Barnard, and B. M. Richards, (1984): Expression of functional GABA, glycine and glutamate receptors in Xenopus oocytes injected with rat brain mRNA. Nature, 310:318-321.

26 G. Grenningloh, A. Rienitz, B. Schmitt, C. Methfessel, M. Zensen, K. Beyreuther, E. D. Gundelfinger, and H. Betz. (1987): Complementary DNA sequence of the strychnine-binding subunit of the glycine receptor and homology with nicotinic acetylcholine receptor proteins. Nature, 328:215-220.

43

27. J-L. Popot, and J-P. Changeux, (1984): Nicotinic receptor of acetylcholine: structure of an oligotneric integral membrane protein Physiol. Rev., 64: 1162-1239.

28. D. Goldman, E. Deneris, W. Luyten, A. Kochhar, J. Patrick, and S. Heinemann, (1987): Members of a nicotinic acetylcholine receptor gene family are expressed in different regions of the mammalian central nervous system. Cell, 48:965-973.

29. . T. I. Bonner, N. J. Buckley, A. C. Young, and K. R. Brann, (1987): Identification of a family of muscarinic acetylcholine receptor genes. Science, 237:527-532.

30. M. Noda, Y. Furutani, H. Takahashi, M. Toyosato, T. Tanabe, S. Shimizu, S. Kikyotani, T. Kayano, T. Hirose, S. Inayama, and S. Numa, (1983): Cloning and sequence analysis of calf cDNA and human genomic DNA encloding a-subunit precursor of muscle acetylcholine receptor. Nature, 305:818-823.

31. R. M. Schwartz, and M. 0. Dayhoff, (1978): In: Atlas of Protein Sequence and Structure, edited by M. 0. Dayhoff, Vol. 5, Suppl, 3, pp. 353-358. National Biomedical Research Foundation, Washington, D.C.

32. B. M. Conti-Tronconi, M. W. Hunkapiller, and M. A. Raftery, (1984): Molecular weight and structural nonequivalence of the mature a subunit of Torpedo Californica acetylcholine receptor. Proc. Natl. Acad. Sci. USA., 81:2631-2634.

33. M. Criado, V. Sarin, J. L. Fox, and J. Lindstrom, (1986): Evidence that the acetylcholine binding site is not formed by the sequence a 127-143 of the acetylcholine receptor. Biochemistry, 25:2839-2846.

34. D. J. McCormick, and M. Z. Atassi, (1984): Localization and synthesis of the acetylcholine-binding site in the a chain of the Torpedo Californica acetylcholine receptor. Biochem. J., 224:995-1000.

35. S. Ralston, V. Sarin, H. L. Thanh, J. Rivier, J. L. Fox, and J. Lindstrom, (1987): Synthetic peptides used to locate the a-bungarotoxin binding site and immunogenic regions in a subunits of the nicotinic acetylcholine receptor. Biochemistry, 26:3261-3266.

36. A, Karlin, P. N. Kao, and M. DiPaola, (1986): Molecular pharmacology of the nicotinic acetylcholine receptor. Trends Pharmacol. Sci., 7:304-308.

37. R. H. Guy, and F. Hucho, (1987): The ion channel of the nicotinic acetylcholine receptor. Trends Neurosci., 10:318-321.

38. J. Deisenhofer, 0. Epp, K. Miki, R. Huber, and H. Michel, (1985): Structure of the protein subunits in the photosynthetic reaction centre of Rhodopseudomonas viridis at 3 A resolution. Nature, 318:618-624.

39. E. F. Young, E. Ralston, J. Blake, J. Ramachandran, Z. W. Hall, and R. M. Stroud, (1985): Proc. Natl. Acad. Sci. USA., 82:626-630.

44

AO. D. Eisenberg, E. Schwarz, M. Komaromy, and R. Wall, (1984): Analysis of membrane and surface protein sequences with the hydrophobic moment plot. J. Mol. Biol., 179:125-142.

41. J. Giraudat, M. Dennis, T. Heidmann, J-Y. Chang, and J-Y. Changeux, (1986): Structure of the high-affinity binding site for non-competitive blockers of the acetylcholine receptor: Serine-262 of the 6 subunit is labelled by [ H] chlorpromazine, Proc. Natl. Acad. Sci. USA., 83:2719-2723.

42. F. Hucha, W. Oberthllr, and F. Lottspeich, (1986): The ion channel of the nicotinic acetylcholine receptor is formed by the homologous helices Mil of the"receptor subunits. FEES Letts, 205:137-142.

43. K. Imoto, C. Methfessel, B. Sakmann, M. Mishina, Y. Mori, T. Konno, K. Fukuda, M. Kurasaki, H. Bujo, Y. Fujita, and S. Numa, (1986): Location of a 6-subunit region determining ion transport through the acetylcholine receptor channel. Nature, 324:670-674.

44. I. Hermans-Borgmeyer, D. Zopf, R-P. Ryseck, B. Hovemann, H. Betz, and E. D. Gundelfinger, (1986): Primary structure of a developmentally regulated nicotinic acetylcholine receptor from Drosophila. EMBO J., 5:1503-1508.

45. J. Kyte, and R. F. Doolittle, (1982): A simple method for displaying the hydrophathic character of a protein. J. Mol. Biol., 157:105-132.

46. L. A. C. Blair, E. S. Levitan, J. Marshall, V.E. Dionne and E.A. Barnard, (1988): Single subunits of the GABA receptor form ion channels with properties of the native receptor. Science (in press) .

47. E. S. Levitan, P. R. Schofield, D. R. Burt, L. M. Rhee, W. Wisden, M.- Kohler, N. Fujita, H. Rodriguez, F. A. Stephenson, M. G. Darlison, E. A. Barnard, P. H. Seeburg, (1988): Structural and functional basis for GABA receptor heterogeneity. Nature, (in press).

45

The EMBO Journal vol.8 no.2 p p . 4 8 9 - 4 9 5 , 1989

Molecular characterization of a new immunoglobulin superfamily protein with potential roles in opioid binding and cell contact

Peter R.Schofield^ ^ K.C.McFarland\ Joel S.Hayflick^ Josiah N.Wilcox^ Tae Mook Cho^, Sabita Roy^, Nancy M.Lee^, Horace H.Loh^ and Peter H-Seeburg^ "*

Genentech, Inc., Departments of Developmental Biology' and Molecular Biology^, 460 Point San Bruno Blvd., South San Francisco, CA 94080, USA, ^Departments of Pharmacology and Psychiatry, School of Medicine, University of California, San Francisco, CA 94143-0450, USA and ''Laboratory of Molecular Neuroendocrinology, ZMBH, University of Heidelberg, Im Neuenheimer Feld 282, 6900 Heidelberg, FRG

^Present address: Pacific Biotechnology Limited, 74 McLachlan Avenue, Rushcutters Bay NSW 2011, Australia

^Present address: Institute of Advanced Biomedical Research, Oregon Health Sciences University, 3181 S.W. Sam Jackson Park Rd., Portland, OR 97201, USA

Communicated by P.H.Seeburg

A purified opioid-binding protein has been characterized by cDNA cloning. The cDNA sequence predicts an extra-cellularly located glycoprotein of 345 amino acids. This protein does not possess a membrane-spanning domain but contains a C-terminal hydrophobic sequence characteristic of membrane attachment by a phosphatidyl-inositol linkage. It displays homology to the immuno-globulin protein superfamily, featuring three domains that resemble disulfide-bonded constant regions. More specifically, the protein is most homologous to a subfamily of proteins which includes the neural cell adhesion molecule (NCAM) and myelin-associated glycoprotein (MAG) and one subgroup of the tyrosine kinase growth factor receptors comprising the platelet-derived growth factor receptor (PDGF R), the colony-stimulating factor 1 receptor (CSF-1 R) and the c-kit protooncogene. These sequence homologies suggest that the protein could be involved in either cell recognition and adhesion, peptidergic ligand binding or both. Key words: cDNA cloning/cell adhesion molecule/immuno-globulin superfamily/opioid receptor/phosphatidylinositol membrane linkage

Introduction

The opioids are a highly diverse group of drugs, which include both a large series of plant-derived alkaloids and many peptides in the mammalian brain. Opioid receptors are likewise heterogeneous, with at least three different classes, differing in their selectivities for alkaloids, or the opioid peptides (Hollt, 1986). Numerous reports of the solubilization and purification by affinity chromatography of opioid binding sites from mammalian brain have been made (Bidlack etal, 1981; Cho etal, 1983; Gioannini et al, 1985; Maneckjee et al, 1985). However, few have reported purification of opioid-binding proteins to apparent

©IRL Press

homogeneity (Gioannini et al., 1985; Simonds et al., 1985; Cho et al., 1986) and, to date, no molecular characterization of these purified proteins has been achieved.

Cho et al (1986) purified a 58-kd opioid-binding protein to apparent homogeneity. Unlike the opioid-binding preparations described by others, the purified protein (Cho et al., 1986) bound ligands only when reconstituted with acidic lipids (Hasegawa et al., 1987). Neither the protein alone, nor the lipids bound opioid ligands to a significant degree, but in combination high affinity binding and selectivity for alkaloid opioids was observed. The binding affinities of ligands to this reconstituted material are lower than values seen for brain membranes, but the rank order of ligand affinity to both preparations is highly correlated (Cho etal, 1983; Hasegawa etal, 1987).

We have undertaken a molecular characterization of this protein and report its primary sequence as deduced from cDNA clones derived from bovine brain. The sequence is homologous to various members of the immunoglobulin (Ig) protein superfamily, especially to those molecules involved in cell adhesion.

Results

Peptide sequences and identification of cDNA clones The affinity-purified protein failed to yield any N-terminal protein sequence, presumably due to a blocked N terminus. Therefore, cyanogen bromide digestion was used to obtain peptide fragments. Four HPLC-purified peptides were subjected to gas-phase microsequencing and the sequences obtained were:

1 . M I Q N V D V Y D E G P Y T X S V Q T 2. M A I E N K G H I S T L T F F ( X V S E K D Y G ) 3. M ^ E F Q W F K E D T R L A T 4. M X X V T V X Q G E S A T

Sequences 1, 2 and 3 were used for oligonucleotide probe design. Duplicate filters of 1.5 X 10® clones of a bovine brain cDNA library were screened at low stringency with ^^P-labelled oligonucleotides designed against peptides 1 and 3, respectively. Clones hybridizing to both sequences were re-screened with an oligonucleotide probe designed against peptide 2. A single clone XBOM106 was positive to all three probes and its 1.8 kb insert was sequenced. A large open reading frame encoding a 318 amino acid polypeptide contained the three peptide sequences used for probe construction as well as the fourth peptide sequence that was obtained from the purified protein (Figure 1).

The open reading frame of XBOM106 lacked an initiation codon. Therefore, the 1749 bp insert of this cDNA clone was ^^P-labelled by random priming and used to re-screen the entire cDNA library at high stringency. No new hybridizing phage were obtained. A specifically primed cDNA library was therefore constructed from tiie same bovine brain mRNA using two different anti-sense

4 8 9

P.R.Schofield et at.

1 CGCCCCCGGCCTCCGCGCGTCCCCGCGCCCCGCGCCGCGTCCCCGGAGCAGGCGGGAGCGCCCGTCTGCCCGCTCCCCGTGCGCCCCG

89 AGCCTGCGCCGCGGGCGCTCCGCGGGCCGGGCCGGCCGCCAGTCGCCGTCCGGGCTGCGGCGAGGACGCCAGGCACCCACCGCTCCGGAG

179 GATCGTCGGACTTGGCCGCGGCTGCGGGTCCCCGCGGTTGGAACTTTTTACCGCCTTGGCGTTCCAGATCAGAGACTCGGGCTGGTTTCT

269 AAATACATCTGTACAT^TGTGATCTAACTAAAACATTCTCTCCTCTCCTTGAAGCGGCAGCTTTTT7rftkAT^

359 GACACTTGCCTGGCTTCTTTGGGTTATAAACTTTTGArGCCAGACCTCGGCAGGACGTGCCCAACTGATrGTTAAAGTAkAT

449 AGAGGCAGAGAAAGAACAACCGTGACCTTCATCCCCGCACCTTCCTCCTTCCTCTCCCTTCCCGGCCGCCTCCCTCTTCCCCAGGGGAGC

539 ATCGAGAAAGCGCTTTTTTGGrreCAGGAGGGGGCATCTGGTTGTGCCAGGCTGGAAAGCTGAGGCAGGATCTGAGGAAGAACAGTAGAC

629 TCCGGAGAGCCTGGACTCCGCTTCTCCTTACCCCCCTCTCCGCTTCTGGCTGGTTCCAGCCCCTGCCGCCCTCCCCGCGCGGCTGCCCGA

1 MetGlyValCysGlySerLeuPhe 719 GACCCGCTCGCGTGCGTGCGAGGACCGAGCCGCCGCCGGAGTTCTGGGAAGTTGTGGCTGTCGAGGATGGGGGTCTGTGGGTCCCTGTTC

9 GlnProTrpLysCysLeuValValValSerLeuArgLeuLeuPheLeuValProTh^lyValProValArgSerGlyAspAlaThrPhe 809 CAGCCCTGGAAGTGCCTCGTGGTCGTGTCTCTCAGGCTGCTGTTCCTTGTACCCACAGGAGTGCCCGTGCGCAGCGGAGATGCCACCTTT

39 ProLysAlaMetAspAsnValThrValArgGlnGlyGluSerAlaThrLeuArgCysThrlleAspAspArgValThrArgValAlaTrp 899 CCCAAGGCGATGGACAACGTGACGGTCCGGCAGGGGGAGAGCGCCACCCTCAGATGTACCATAGATGATCGGGTCACCCGGGTGGCCTGG

••••••••• 69 LeuAsnArgSerThrlleLeuTyrAlaGlyAsnAspLysTrpSerlleAspProArgValllelleLeuValAsnThrProThrGlnTyr

989 CTGAACCGCAGCACCATCCTCTACGCCGGGAATGACAAGTGGTCCATAGACCCTCGAGTGATCATCCTGGTGAACACGCCAACCCAGTAC •

99 SerlleMetlleGlnAsnValAspValTyrAspGluGlyProTyrThrCysSerValGlnThrAspAsnHisProLysThrSerArgVal 1079 AGCATCATGATCCAGAACGTGGACGTGTACGACGAGGGCCCCTATACCTGCTCTGTGCAGACGGACAACCACCCCAAGACCTCCCGTGTC

129 HisLeuIleValGlnValProProGlnlleMetAsnlleSerSerAspValThrValAsnGluGlySerSerValThrLeuLeuCysLeu 1169 CACCTCATCGTGCAGGTCCCTCCCCAGATCATGAACATCTCCTCAGATGTCACCGTGAATGAGGGGAGCAGCGTGACCCTGCTGTGTCTT

159 AlalleGlyArgProGluProThrValThrTrpArgHisLeuSerValLysGluGlyGlnGlyPheValSerGluAspGluTyrLeuGlu 1259 GCTATCGGCAGACCAGAGCCAACGGTGACGTGGAGACACCTGTCAGTCAAGGAAGGCCAGGGCTTTGTGAGTGAGGATGAATACCTGGAA

• 189 IleSerAspIleLysArgAspGlnSerGlyGluTyrGluCysSerAlaLeuAsnAspValAlaAlaProAspValArgLysValLysIle

1349 ATCTCTGACATCAAACGTGACCAGTCCGGGGAGTATGAGTGCAGCGCCTTGAATGATGTTGCTGCCCCTGACGTGCGGAAAGTAAAGATC

219 ThrValAsnTyrProProTyrlleSerLysAlaLysAsnThrGlyValSerValGlyGlnLysGlylleLeuSerCysGluAlaSerAla 1439 ACTGTCAACTACCCCCCCTATATCTCCAAAGCCAAGAACACAGGGGTCTCCGTTGGCCAGAAGGGCATCTTGAGCTGTGAAGCCTCGGCA

249 ValProMetAlaGluPheGlnTrpPheLysGluAspThrArgLeuAlaThrGlyLeuAspGlyMetArglleGluAsnLysGlyHisIle 1529 GTGCCCATGGCTGAGTTCCAGTGGTTCAAGGAAGACACCAGGCTGGCCACCGGCCTGGACGGCATGAGGATCGAGAACAAAGGCCACATA

••••••••• ••••••••• ® ••••••••• 279 SerThrLeuThrPhePheAsnValSerGluLysAspTyrGlyAsnTyrThrCysValAlaThrAsnLysLeuGlylleThrAsnAlaSer 1619 TCCACGCTGACCTTCTTCAACGTCTCAGAGAAGGATTATGGGAACTATACTTGTGTGGCCACAAACAAGCTTGGGATTACCAATGCCAGC

309 IleThrLeuTyrGlyProGlyAlaVallleAspGlyValAsnSerAlaSerArgAlaLeuAlaCysLeuTrpLeuSerGlyThrLeuPhe 1709 ATCACACTGTATGGGCCTGGAGCCGTCATTGATGGTGTAAACTCGGCCTCTAGAGCGCTCGCTTGCCTCTGGCTATCAGGGACCCTCTTT

339 AlaHisPhePhelleLysPhe***

1799 GCCCACTTCTTCATCAAGTTTTGATAAGAAACCATAGGTCTTCTGAGCAACGCCTGCTTCTCCATATCACAGACTTTACCTGCACTGCGG

1889 AGGGCCAGTTTGGGCTTCTTTCTGTTTCTGTTCTTCTTCTCAGTATTTTTTTTTTTTTTTTGGACTCTTTTCTTTGTTGATTTGATTTTT

1979 CTTTCTAGTTTGAACGAGTGGGTTTGGGGAGGGTGGGCAGGTCTACACATGTAGGATATCACTCATATGTGATGTGTCCAAACTGGAACC

2069 CATTCTCCACCCTTCCCTTCCCTTCCCGCTGCTCCCCTGGGCATTGTCACCACCAACCAACCTCCCTCCCACACCAAACCATAGTTCCAT

2159 TTGGGCAAAAACGTGCCTCGTGATAAACACCCTGAAGACACAACTTGACTTATAATGTAGTGCACAGCAAGAATTATATCCAAGTGTCCT

2249 ATTACTGTGTCTTTTAAGCTGTGGACCATTTTCCTGACTATAATGTACAGATCCCTCTCTCCATGTTTATTATGATCTAATTACATTGAG

2339 GGAACACATCCTTTCTTTCGGGAAGTTCTCTCTCTCTTCCTATGTCTCTCTCTGACTCTCTGTCTCTGTCTCTCTCTCTCTCTCTACAAA

2429 TATATGAAACATTGCCATCCTTCCTAGTCATTCTGCCCTATTTGCTTTCTCTACCAGCTATAAGGATTCGAAGTTTGGGGGTAGGCATAC

2519 AACCCAAACCTGAACATTGAGGAGTCATGTAACCGAAAATGGGGGGAAGAATCTTGAGGAGAGTATTTCTCTCGGAAAAAAAAAAAAAAA

2609 AAAAAA

Fig. 1. Nucleotide and deduced amino acid sequence of the cDNA clones. The sequence is derived from the two cDNA clones, XBOM159 (nt 1 - 1 1 6 5 ) and XBOM106 (nt 865-2614) . Potential initiation codons (ATG) in the 5'-untranslated sequence are overscored and stop codons marked with asterisks. Within the amino acid sequence the putative position of signal sequence cleavage is indicated by an arrow, potential N-linked glycosylation sites by filled dots and the peptide sequences used for probe construction are underlined. The positions of the three repeating Ig-like domains are indicated by arrowheads and the conserved cysteine residues are indicated by filled circles. The C-terminal hydrophobic sequence, characteristic of Pl-linked proteins is indicated by a dashed line.

oligonucleotide primers. This library was screened unamplified with the ^^P-labelled 1749 bp insert and a single positively hybridizing clone, XBOM159, obtained. The sequence of this 1165 bp cDNA extended the open reading frame of XBOM106 a further 27 amino acids including a putative initiation codon (Figure 1).

The bovine cDNA contains a very long 5'-untranslated region of 784 bp beginning with an extremely GC-rich sequence (90% of the first 128 nucleotides). This untrans-lated region contains six potential initiation codons (ATG), each preceding a small open reading frame of < 35 amino acids. The seventh ATG triplet (nt 785-787) , which con-

forms most closely to the consensus initiation sequence (Kozak, 1987), is preceded by an in-frame stop codon (TGA at nt 611 -613 ) and starts the long reading frame encoding the four chemically determined peptide sequences. Thus, the protein is apparenfly initiated at the seventh ATG codon, whereas most eukaryotic proteins are initiated at the first such codon in the mRNA sequence. Kozak (1987) has suggested that such RNAs, with multiple potential initiation codons, may contain an unspliced intron. However, the rat cDNA homologue (P.R.Schofield and P.H.Seeburg, unpublished) contains an initiation codon in an analogous position, indicating a correct initiation codon assignment and

490

New Ig superfamily protein

the absence of unspliced introns. The DNA sequence of clone XBOM106 terminates with a short A-rich sequence. However, the absence of a polyadenylation addition signal (A AT A A A) suggests that there are further 3'-untranslated sequences that are not encoded by this cDNA clone. Northern blot analysis The cDNA encoding the bovine protein is rare, since only one clone was identified in a high complexity bovine brain cDNA library. This result is confirmed by Northern blot analysis (Figure 2). The mRNA is expressed within whole brain and hypothalamus but not in the liver. Two distinct mRNA species of 4500 and 7200 nt were detected, as were some minor species. Their size indicates that the cloned bovine cDNA does not contain the full mRNA sequence but was primed at an internal A-rich sequence (Figure 1). We do not know if the two mRNA species seen by Northern analysis encode proteins characterized by alternate exon usage or differ in the length of their 5'- or 3'-untranslated sequences. Structural interpretations The 345 amino acid long polypeptide encoded by this cDNA has a predicted mol. v^ of only 38 kd, considerably less than the 58 kd of the purified protein. The presence of six potential N-linked glycosylation sites (Asn-X-Ser/Thr) (Figure 1) suggests that carbohydrate attachment may account for the size discrepancy. Thus, Asnjgs is a particularly strong candidate site for glycosylation, as this residue alone was not identified in the chemically determined sequence of peptide 2. By similar criteria, Asn44 may also be glycosylated.

The protein is largely hydrophilic but has hydrophobic sequences at both ends. The N-terminal sequence has the characteristics of a signal peptide, with a putative cleavage site occurring after Thrjy (von Heijne, 1986). The presence of a signal sequence and of N-linked glycosylation sites suggest that this protein is extracellularly located.

The hydrophobic C-terminal sequence constitutes the last 19 amino acids, but its proximity to the end of the protein and the absence of a stop transfer or anchoring sequence (Blobel, 1980) suggests it is not membrane spanning. However, one class of membrane-anchored proteins is characterized by the presence of C-terminal hydrophobic sequences which lack charged anchoring residues. These are the phosphatidylinositol (Pl)-linked proteins (Cross, 1987; Low and Kincade, 1985) which include the rodent Thy-1 antigen (Seki etai, 1985), trypanosome variant surface glycoprotein (VSG) (Boothroyd etai, 1980), chicken (Hemperly et al., 1986) and mouse (Barthels et al, 1987) NCAM120, human decay-accelerating factor of complement (DAF) (Caras et al., 1987), Qa-2 antigen of the major histocompatability complex (Stroynowski et al., 1987), and T cell-activating protein (TAP) (Reiser etai, 1986), reviewed by Low and Saltiel (1988). For each of these proteins, the C-terminal hydrophobic sequence is processed in the coupling of the protein to the phosphatidylinositol moiety. Lipid attachment occurs at a residue located within or adjacent to the hydrophobic sequence (Low and Kincade, 1985; Low and Saltiel, 1988; Tse etai, 1985; Ferguson et al, 1985). Apart from this hydrophobicity no strong PI-linkage consensus sequence is apparent. However, in the cases of NCAM, Qa-2 and DAF, the Pl-linkage domain is

L H B

Fig. 2. Northern blot of bovine RNA. Poly(A)+ mRNA (5 fig) from liver (L) hypothalamus (H) and whole brain (B) was subjected to Northern blot analysis using ^^P-labelled XBOM106 cDNA. Size markers (BRL) are shown in kb. Exposure was for 2 weeks. encoded on a differentially spliced exon. Moreover, any eukaryotic protein that terminates in a short hydrophobic segment is a candidate for Pl-linkage or glypiation (Cross, 1987). Such a proposal has been made for the carcino-embryonic antigen (CEA) (Oikawa etai, 1987) which terminates with a characteristic C-terminal hydrophobic sequence (Williams, 1987). We suggest that this protein is similarly processed and attached to the cell membrane by glypiation. This conclusion is further supported by homologies to other Pl-linked proteins (see below).

The hydrophilic portion of the amino acid sequence contains three internal repeating structures of —100 amino acids (Figure 3). These three sequences share - 2 5 % identity around two separate cysteine residues present in each of the repeatings. Homologies with the immunoglobulin superfamilies Comparison of the protein sequence to the Dayhoff protein database using the 'fastp' algorithm (Lipman and Pearson, 1985) revealed the highest and most significant homologies to be with two cell adhesion molecules, neural cell adhesion molecule (NCAM) (Cunningham et al, 1987) (22% identity) and myelin-associated glycoprotein (MAG) (Arquint et al, 1987; Salzer et al, 1987; Lai et al, 1987) (21 % identity). Additionally, varying degrees of sequence identity were seen with members of the extended immunoglobulin protein superfamily (Williams, 1987). These homologies were centered around the conserved cysteine residues of the repeating structure identified above. The conserved motifs form the disulfide-bonded folded regions of the immuno-globulin superfamily of proteins. Other groups of proteins that belong to this superfamily include: (i), the cell adhesion molecules NCAM (Cunningham etai, 1987), MAG (Arquint et al, 1987; Salzer et al, 1987; Lai et al, 1987), the LI glycoprotein, involved in fasciculation (Moos et al, 1988), and the intercellular adhesion molecule, ICAM-1 (Simmons et al, 1988); (ii) the carcinoembryonic antigen (CEA) (Oikawa et al, 1987), which may be involved in cell adhesion, (iii) the type in tyrosine kinase growth factor receptors represented by the platelet-derived growth factor

491

P.R.Schofield ef al.

A) COMPARISON OF OBCAM IG-DOMAINS WITH THE CONSENSUS SEQUENCES OF THE IG VARIABLE AND CONSTANT DOMAINS

OBCAM OBCAM OBCAM cons

A T L R C T I D D V T L L C L A I G G I L S C E A S A

t L C a

- V T PEP P M A P

- R V A W L N R S - T V T W R H L S - E F Q W F K E D

V W s

Q Y S I M I Q N V D V Y D E G P D E Y L E I S D I K R D Q S G E I S T L T F F N V S E K D Y G N

1 i n V d G

V cons CI cons C2 cons

V t L t C s A t L v C L V s V T L T C e a

V s G F f W V W 1 W

R Q n G

f s 8 S

L t I L t V L L V T

D s G

0 s G

Y T C S V Q T Y E C S A L N Y T C V A T N Y e C 8 a n

Y C A Y s C V H Y C A N

13-strand B

B) i CELL CONTACT MOLECULES-NEURONAL

(5-strand C

NCAM cons MAG cons LI cons cons

C V s 1 C

1 C C

P n P

P P

B-strand E (3-strand F

i i CELL CONTACT MOLECULES-LYMPHOIDAL

ICAM-1 cons t V

i i i GROWTH FACTOR TYROSINE KINASE RECEPTORS (TYPE I I I )

TKGF R 1 TKGF R 2 TKGF R 3 TKGF R 4 cons

L i e P c

t i C L V 1 c

g • 9 1 T D P

V d

V E A y P

P

V e

V n f

W s 1 k

w y

1 f t L L 1

n e t k

a R 1 K

a k

n T G

s G a G

G

Y

y y Y t

y

C t C s c f 1 c

i v CARCINOEMBRYONIC ANTIGEN

CEA 2,4,6 CEA 3,5,7 cons

V A f T C E P E Q N T T L n L S C H a A S N P p a

C N

V MACROPHAGE-LYMPHOCYTE Fc RECEPTOR

FcR cons t L C

Y L W W v N n Q 0 Y S W i n G

W n

F h N g

N R T L T L f n V T R N D T Q e L F I n I T n N s G

L n T n

Y C g i Q N Y t C q a N Y C N

k A S G

Fig. 3. Alignment of the Ig domains of the protein with members of the immunoglobulin protein superfamily. (A) The three Ig-like repeating domains are aligned and the derived consensus sequence compared with the three Ig domain consensus sequences. The positions of the relevant /3-strands are indicated. Only portions of the respective protein sequences are shown. (B) Consensus sequences derived from functionally related groups of the C2 class of Ig-related molecules. Consensus sequences used are derived from the complete amino acid sequences. Except for the first three lines which contain the OBCAM protein sequence, captial letters denote an invariant residue while lower case letters denote a residue conserved in >50% of sequences. Numbers indicate the Ig domain number of the molecule. The consensus (cons) sequences "J«-esented in i, iii, and iv are derived from the date presented only.

receptor (PDGF R), the colony-stimulating factor-1 receptor (CSF-1 R) and the putative receptor and protooncogene c-kit (Yarden et al., 1987); (iv) the leukocyte (macrophage) Fc receptors (Lewis et al., 1986; Stengelin et al, 1988). All contain immunoglobulin-like domains that display either seven anti-parallel j8-strands, characteristic of constant domains, or nine anti-parallel /S-strands, characteristic of variable domains. The seven jS-strand constant domain-containing molecules have been divided into tv^o classes; CI, which includes the immunoglobulin and T cell receptor constant domains, and C2 which includes the four other groups of Ig-related proteins, typified respectively by NCAM, CEA, PDGF R and the Fc R.

The three Ig domains of this protein were aligned and a consensus sequence derived. The consensus sequence was then compared (Figure 3A) to V, CI and C2 consensus sequences (Williams, 1987). The sequence is assigned to the C2 set of proteins based on the number of conserved residues present, and the fact that the sequences around the second cysteine residue are more V- than C-like. Additionally, the spacings between the cysteine residues are consistent with the presence of C-like domains. Within this set, consensus sequences were derived for the various C2 protein subgroups (Figure 3B). Greatest conservation of the consensus sequence was detected with the molecules involved in cell adhesion, and to a lesser degree with the type IE tyrosine kinase growth

factor receptors. The similarities between the OBCAM protein and members of the NCAM family of proteins suggest that we have isolated a new member of the cell adhesion class of molecules. The putative Pl-linkage of the protein to the cell membrane is another characteristic shared with the various Ig-related recognition and adhesion molecules, such as NCAM, Thy-l,Qa-2, TAP and perhaps CEA. Sequence similarities with the type EI tyrosine growth factor receptors suggest a role of the protein in receptor-like peptidergic ligand binding, a function consistent with the purification and biochemical characterization of the opioid-binding protein.

Identity of cloned cDNA To show that the cloned cDNA encodes the purified protein, we made use of an anti-peptide antibody raised against a portion of the cDNA-encoded amino acid sequence. An immunoaffinity antibody column was constructed using the IgG fraction of the anti-peptide antibody. Protein solubilized from rat brain membranes, that bound and was specifically eluted from this column, bound opioids when examined in the assays used for protein purification (Figure 4). Opioid binding was inhibited by the presence of a monoclonal antibody (3B4R11) (Roy etal, 1988) which was raised against the purified protein (Figure 4). These results indicate that the purified protein is encoded by the cloned cDNA,

492

N e w Ig superfamily protein

1 10 100 Antibody Concentration (nM)

0.02

Protein Concentration (yg)

Fig. 4. Co-identity of purified protein and the cloned cDNA. (A) Solubilized rat brain membrane protein, purified on an anti-peptide antibody column, binds [^HJdiprenorphine. (B) Binding was inhibited by the presence of monoclonal antibody 3B4F11.

supporting the notion that the protein has a role in opioid binding.

Discussion

The primary sequence of the purified opioid-binding protein is characterized by the presence of a signal sequence, the lack of internal hydrophobic membrane-spanning sequences and by potential N-linked glycosylation sites, suggesting that this protein is extracellularly located. Of the six potential glycosylation sites, two, Asn285 and Asn44 appear to be glycosylated, as determined by chemical sequencing. This is in agreement with binding of the solubilized protein to lectin columns and a calculated mol. wt of the protein of ~20 kd lower than that observed for the purified protein.

The purified opioid-binding protein had been thought to be the n opioid receptor (Cho et al, 1986). However, the predicted primary sequence of the cDNA shows a lack of membrane-spanning domains. This is somewhat suprising, since receptors that interact with GTP-binding proteins, including ^ e /3-adrenergic receptors (Dixon et al, 1986), the muscarinic cholinergic receptors (Kubo et al, 1986), rhodopsins (Hargrave, 1982) and the substance K receptor (Masu et al., 1987) share significant sequence and strucmral similarities including the presence of seven membrane-spanning domains. At least some types of opioid receptors are thought to associate with G-proteins, most notably the 6 receptor of NG108-15 neuroblastoma-glioma cells (Chang and Cuatrecasas, 1979) and of mammalian striatum (Law et al., 1981), but possibly also the fi receptor (Milligan et al., 1987). G-protein interactions suggest that opioid receptor proteins would belong to the same seven membrane-spanning protein superfamily. The protein sequence predicted by the cloned cDNA thus suggests that the OBCAM molecule may not be a G-protein-coupled opioid receptor. The protein may, however, have a functional role in opioid binding, as suggested by the properties of the purified protein (Cho etal., 1983, 1986; Hasegawa etal., 1987).

The primary sequence of the protein is very highly conserved between bovine and rat (data not shown). The mRNA is expressed in neurons in the brain and not detected in other tissues (manuscript in preparation). However, the distribution of the mRNA does not co-localize with the known distributions of the fi or 8 opioid receptors, as determined by in vitro autoradiography using radiolabelled opioid receptor ligands. The brain-specific expression of this highly conserved protein implies functional significance although its physiological role remains to be determined.

Pl-linked membrane-anchoring mechanisms have been variously suggested as a primitive form of membrane attachment or as a mechanism that facilitates release of the protein component upon phospholipase cleavage (Low and Saltiel, 1988). For the trypanosome VSG, phospholipase cleavage may permit specific protein release from the cell surface during the parasitic life cycle (Ferguson et al., 1985; Low and Kincade, 1985). The release of the protein compo-nent of Pl-linked proteins may simultaneously release phos-phatidylinositol, diacyglycerol or phosphatidic acid (Low and Kincade, 1985) which concomitantly effect cellular responses, in particular the mobilization of calcium (Berridge and Irvine, 1984) and, hence, the activation of protein kinase C (Nashi-zuka, 1984). The demonstration that cross-linking Thy-1 with antibodies increases the cytoplasmic free Ca "*" concentration (Kroszek et al., 1986) suggests a role for Pl-linked proteins in signal transduction. Recently, the major Fc receptor in blood has been shown to be Pl-linked (Selvaraj et al., 1988; Simmons and Seed, 1988). This receptor is capable of signal transduction, resulting in the triggering of cell-mediated killing, thus demonstrating that Pl-linked proteins can act as true signal transducing receptors and not just binding proteins. By analogy, we suggest that cell contact or ligand binding of some Pl-linked molecules, such as the protein characterized in this study, may activate membrane-bound phospholipases which could cleave the Pl-linkage molecules and initiate the Pl-mediated release of free cellular calcium and activation of protein kinase C.

The homology of the protein with the C2 class of the Ig superfamily suggests two possible functions. One is an involvement in cell recognition, contact or adhesion, demonstrated by the homology to NCAM, MAG and related molecules. The presence of a putative Pl-linkage site for membrane attachment is consistent with this notion. The other function is as a peptidergic binding protein or receptor, suggested by the homology to the type III tyrosine kinase growth factor receptors. Since the primary sequence of the protein shows homologies with two functionally different classes of immunoglobulin-related proteins, we have termed the protein OBCAM (opioid-binding protein-cell adhesion molecule). Expression of the cDNA will help in the elucidation of its role in binding opioids and in cellular recognition and adhesion.

Materials and methods

Purification and peptide sequencing Bovine brain membranes were sonicated and solubilized in Triton X-100. Opioid-binding protein was purified by 6-succinyl morphine affinity chromatography, ultrogel filtration, wheat germ affinity chromatography and preparative isoelectric focussing as previously described (Cho et al., 1986). A single 58-kd band was observed which was electroeluted and digested with cyanogen bromide overnight in 300 of 60 mg/ml CNBr in 70% formic acid. Individual peptides were resolved by separation on a Synchron CA reverse phase HPLC column, eluted with a n-propanol/

493

P.R.Schofield et al.

trifluoroacetic acid gradient and subjected to gas-phase microsequence analysis (Rodriguez, 1985). Peptides 2 and 3 were resolved by additional chromatography using a propanol/heptafluorobutyric acid gradient. A portion of sequence 2 was found in the mixture, indicated by parentheses, but was not identified as the extension of peptide 2 until confirmed by the DNA sequence. Unidentified residues are indicated by X and residues >30% below the expected yield are italicized. The proximal methionine residue is assumed.

cDNA cloning and DNA sequencing Synthetic oligonucleotide probes based on codon usage (Lathe, 1985) were constructed. Three oligonucleotides, 1 (a 42-mer against peptide 1), 2 (two 45-mers against peptide 2) and 3 (two 45-mers against peptide 3) were used, their sequences being:

1:

2:

5' ATGATCCAGAACGTGGACGTGTACGACGAGGGCCCTTACC 3'

AGP 5' ATGGCCATCGAGAACAAGGGCCACATCXCCAACCTGACCTTCTTC 3'

3: AGA, 5' ATGGCCGAGTTCCAGTGGTTCAAGGAGACACCCGCCTGGCCACC 3'

Total bovine brain RNA was prepared using guanidinium thiocyanate (Chirgwin et al., 1979) followed by ultracentrifugation over a 5.7 M CsCl cushion. Poly (A)"*" mRNA was isolated by oligo d(T)-cellulose chroma-tography (Aviv and Leder, 1972) and 5 /ig was used for library constmction. Oligo d(T7-primed cDNA was prepared as previously described (Wood et al., 1984). After EcoRl-Xhol adapters were ligated, cDNA > 1500 bp in size was selected by elution from a 6% polyacrylamide gel. The cDNA (20 ng) was ligated to XgtlO (1 ng) and packaged in Gigapack (Stratagene) packaging extracts giving a library of 1.4 X 10^ independent clones. To obtain full length CDNAS, a specifically-primed library was constructed using the same bovine brain poly(A)"'" mRNA by priming with two synthetic anti-sense oligonucleotides:

XI and X2

5' ACGGGAGGTCTTGGGGTGGTTGTCCGTC 3'

5' GGATGGTGCTGCGGTTCAGC 3'

cDNA > 200 bp was size-selected by elution from a polyacrylamide gel and 20 ng was ligated to XgtlO (1 ng) and packaged, giving a libraiy of 0.2 X 10^ independent clones. cDNA inserts were subcloned into appropriate M13 derivatives and sequenced by the chain termination method (Sanger et al., 1977; Messing et al., 1981) using specific oligonucleotide sequencing primers.

Northern blot analysis Bovine poly(A)"^ mRNA was isolated from whole brain, hypothalamus and liver as described above. RNA (5 ^g) was electrophoresed in a formaldehyde—1.2% agarose gel and transferred to nitrocellulose prior to hybridization in 50% formamide at 42°C with the -^^P-labelled insert of XBOM106. The blot was washed at 65 °C in 0.1 X SSC prior to exposure to X-ray film for 2 weeks.

Immunological analysis A synthetic peptide corresponding to amino acids 188—202 (EISDIKRD-QSGEYEC) was synthesized and used to raise rabbit polyclonal antibodies. An immunoaffinity antibody column was constructed using the IgG fraction of this antisemm (obtained by Protein-A binding). Rat P2 membranes were solubilized with 0.5% A''-octyl ;8-D-glucopyranoside, 1 mM dithiothreitol, 0.05M Tris, pH 7.4 for 1 h at 4°C and the supernatant passed through the immunoaffinity column and bound proteins eluted with a gradient of 0.1 M glycine HCl, pH 2.0. The eluate was immediately neutralized with Tris base (0.1 M) and dialysed against 0.05 M Tris, pH 7.4. Determination of the [-^HJdiprenorphine binding was carried out as described (Cho et al., 1986). The monoclonal antibody 3B4F11 (Roy et al., 1988) raised against purified opioid-binding protein was used to study the inhibition of ligand binding to the immunoaffinity purified protein.

Acknowledgements

We tiiank Heniy Rodriguez for the protein sequence analysis, Dr John Bumier for synthesis of synthetic peptides, Dr Marion Moos and Professor Melitta Schachner for sharing the LI sequence data prior to publication and for helpful discussions and Dr Andrew Smith, J.Rami and I.Baro for preparation of the manuscript. This work was supported in part by NIDA grant DA-01583, Research Scientist Award 5-K05-DA-70554 (HHL) and

grant DA-01583, Research Scientist Award 5-K05-DA-70554 (HHL) and will appear in the EMBL/Gen Bank/DDBJ/Nucleotide Sequence Database under the accession number X12672.

References

Arquint,M., Roder,J., Chia,L.-S., Down,J. Wilkinson,D., Bayley,H., Braun,P. and Dunn,R. (1987) Proc. Natl. Acad. Sci. USA, 84, 600 -604 .

Aviv,H. andLeder,P. (1972) Proc. Natl. Acad. Sci. USA, 69, 1408-1412. Bartiiels,D., Santoni,M.-J., Wille,W., Ruppert,C., Chaix,J.C., Hirsh,

M.-R., Fontecille-Camps,J.C. and Goridis,C. (1987) EMBO J., 6, 907 - 914.

Berridge,M.J. and Irvine,R.F. (1984) Nature, 321, 315-321. Bidlack,J.M., Abood,L.G., Osei-Gyimah,P. and Archer,S. (1981) Proc.

Natl. Acad. Sci. USA, 78, 636-639. Blobel,G. (1980) Proc. Natl. Acad. Sci. USA, 77, 1496-1500. BoothroydJ.C., Cross,G.A.M., Hoeijmakers,J.H.J. and Borst.P. (1980)

Nature, 288, 624-626. Caras,I.W., Davitz,M.A., Rhee.L., WeddelI,G., Martin.D.W. Jr. and

Nussenzweig,V. (1987) Nature, 325, 545 - 5 4 9 . Chang,K.-J. and Cuah-ecasas,P. (1979)7. Biol. Chem., 254, 2610-2618. Chirgwin,J.M., Przybyla,A.E., MacDonald,R.J. and Rutter,W.J. (1979)

Biochemistry, 18, 5294-5298. Cho,T.M., Ge,B.L., Yamato,C., Smith,A.P. and Loh.H.H. (1983) Proc.

Natl. Acad. Sci. USA, 80, 5176-5180. Cho,T.M., Hasegawa,j.-I., Ge,B.-L. and Loh,H.H. (1986) Proc. Natl.

Acad. Sci. USA, 83, 4138 -4142 . Cross,G.A.M. (1987) Cell, 48, 179-181. Cunningham,B.A., HemperlyJ .J . , Murray,B.A., Prediger.E.A.,

Brackenbury,R. and Edelman,G.M. (1987) Science, 236, 799-806. Dixon,R.A.F., Kobilka,B.K., Strader,D.J., Benovic,J.L., Dohlman.H.G.,

Frielle,T., Bolanowski,M.A., Bennett,C.D., Rands,E., Diehl,R.E., Mumford,R.A., Slater,E.E., Sigal,I.S., Caron,M.G., Lefkowitz,R.J. and Strader,C.D. (1986) Nature, 321, 7 5 - 7 9 .

Ferguson,M.A.J., Low,M.G. and Cross,G.A.M. (1985)7. Biol. Chem., 260, 14547-14555.

Gioannini,T.L., Howard,A.D., Hiller.J.M. and Simon.E.J. (1985)7. Biol. Chem., 260, 15117-15121.

Hargrave,P.A. (1982) Prog. Retinal Res., 1, 1 - 5 1 . Hasegawa,J.-L, Loh,H.H. and Lee.N.M. (1987) 7. Neurochem., 49,

1007-1012. Hemperly,J.J., Edelman.G.M. and Cunningham,B.A. (1986) Proc. Natl.

Acad. Sci. USA, 83. 9822-9826. Hollt,V. (1986) Annu. Rev. Phannacol Toxicol., 26. 5 9 - 7 7 . Kozak,M. (1987) Nucleic Adds Res., 15, 8125-8148. Kroczek,R.A., Gunter,K.C., Germain,R.N. and Shevach.E.H. (1986)

Nature, 322, 181-184. Kubo,T., Fukuda,K., Mikami.A., Maeda.A., Takahashi.H., Mishina.M.,

Haga,T., Haga.K., Ichiyama,A., Kangawa,K., Kojima,M., Matuso.H.. Hirose,T. and Numa,S. (1986) Nanire, 323, 411-416 .

Lai,C., Brown.M.A., Nave,K.-A., Noronha,A.B., Quarles.R.H., Bloom,F.E., Milner,R.J. and SutcliffeJ.G. (1987) Proc. Natl. Acad. Sci. USA, 84, 4337 -4341 .

Lathe.R. (1985)7. MoL Biol., 183, 1 - 1 2 . Law,P.Y., Wu,J., Koehler,J.E. and Loh,H.H. (1981)7. Neurochem., i6.

1834-1846. Lewis,V.A., Koch,T., Plutner.H. and Mellman.I. (1986) Nanire. 324.

372-375. Lipman.D.J. and Pearson,W.R. (1985) Science. 227, 1435-1441. Low,M.G. and Kincade,P.W. (1985) Nature, 318, 62 - 64. Low.M.G. and Saltiel.A.R. (1988) Science, 239, 268-275 . Maneckjee.R., Zukin.R.S., Archer,S,. MichaeKJ. and Osei-Gyimah.P.

(1985) Proc. Natl. Acad Sci. USA. 82, 594-598. Masu,Y., Nakayama.K., Tamaki.H., Harada,Y., Kuno.M. and Nakanishi.S.

(1987) Nana-e, 329, 836-883. Messing,J., Crea.R. and Seeburg,P.H. (1981) Nucleic Acids Res.. 9.

309 - 321. Milligan.G.. Streaty.R.A., Gierschik.P., Siesel.A.H. andKlee.W.A. (1987)

J.^Biol. Chem., 262. 8626-8630. Moos.M., Tacke,R., Scherer.H.. Teplow.D.. Fmeh.K. and Schachner.M.

(1988) Nature. 334. 701-703 . Nashizuka,Y. (1984) Nanire, 308. 693-698. Oikawa.S.. Nakazato.H. and Kosaki.G. (1987) Biochem. Biophys. Res.

COWHJH/J., 142, 511-518 . Reiser.H., Oettgen.H., Yeh.E.T.H., Terhorst.C.. Low.M.G.. Benaceraf.B.

and Rock-KX. (1986) Cell, 47, 365-370. Rodriguez.H. (1985) 7. Chromatogr., 350. 217-225 .

494

New Ig superfamily protein

Roy,S., Zhu,Y.X., Loh,H.H. and Lee,N.M. (1988) Biochem. Biophys. Res. Commun., 154, 688-693.

SalzerJ.L., Holmes,W.P. and Colman,D.R. (1987)7. Cell Biol., 104, 957-965.

Sanger,?., Nicklen,S. and Coulson,A.R. (1977) Proc. Natl. Acad. Sd. USA, 74, 5463-5467.

Seki,T., Moriuchi,T., Chang,H.C., Denome,R. and Silver,!. (1985)Mjtore, 313, 485 -487.

Selvaraj,P., Rosse,W.F., Snber,R. and Springer,!.A. (1988) AfeMre, 333, 565-567.

Simmons,D. and Seed,B. (1988) Nature, 333, 568-570. Simmons,D., Makgoba,M.W. and Seed,B. (mS) Nature, 331, 624- 627. Simonds,W.F., Burke,T.R.Jr., Rice,K.C., Jacobson,A.E. and KIee,W.A.

(1985) Proc. Natl. Acad. Sci. USA, 82, 4974 -4978. Stengelin,S., Stamenkovic,!. and Seed,B. (1988) EMBOJ., 7, 1053-1059. Stroynowski,!., Soloski,M., Low,M.G. and Hood,L. (1987) Cell, 50,

759-768. Tse,A.G.D., Barclay,A.N., Watts,A. and Williams,A. (1985) Science, 230,

1003-1008. von Heijne,G. (1986) Nucleic Acids Res., 14, 4683-4690. Williams,A.F. (1987) Immunol. Today, 8, 298-303. Wood,W.I., Capon,D.J., Simonsen,C.C., Eaton,D.L., Gitschier,J.,

Keyt,B., Seeburg,P.H., Smith,D.H., Hollingshead,?., Wion,K.L., Delwart,E., Tuddenham,G.D., Veher,G.A. and Lawn,R.M. (1984) Nature, 312, 330-337.

Yarden,Y., Kuang,W.-J., Yang-Feng,T., Coussens,L., Muneniitsu,S., Dull,T.J., Chen,E., Schlessinger,!., Francke,U. and Ullrich,A. (1987) EMBO J., 6, 3341-3351.

Received on August 25, 1988; revised on November 21, 1988

495

Journal of Cellular Biochemistry 39:277-284 (1989) Cellular Proteases and Control Mechanisms 79-86

Expression of Enzymatically Active Enkephalinase (Neutral Endopeptidase) in Mammalian Cells Cornelia M. Gorman, David Gies, Peter R. Schofield, Helen Kado-Fong, and Bernard Malfroy

Departments of Cell Genetics (C.M.G.. D.G.), Developmental Biology (P.R.S.), and Pharmacological Sciences (H.K.-F., B.M.), Genentech, Inc., South San-Francisco, California 94080

A cDNA encoding the rat enkephalinase protein (neutral endopeptidase; EC 3.4.24.11) has been constructed from overlapping XgtlO cDNA clones. This cDNA was inserted into an expression plasmid containing the cytomegalovirus enhancer and promoter. When transfected with this plasmid, Cos 7 cells tran-siently expressed the enkephalinase protein in a membrane-bound state. Recom-binant enkephalinase recovered in solubilized extracts from transfected Cos 7 cells was enzymatically active and displayed properties similar to those of the native enzyme with respect to sensitivity to classical enkephalinase inhibitors.

Key words: enkephalinase, neutral endopeptidase, metallo peptidase The signals conveyed by the opioid pentapeptides enkephalins when released in

their synapses are turned off by degradation of the peptides by two membrane-bound metallopeptidases, enkephalinase (neutral endopeptidase; EC 3.4.24.11) and amino-peptidase M (EC 3.4.11.2) [1,2]. Aminopeptidase M hydrolyzes the Tyr^ -Gly^ amide bond of the pentapeptides, while enkephalinase hydrolyzes the Gly^-Phe"^ amide bond, thus releasing the carboxy terminal dipeptide [3,4]. The study of the inhibitory potency of a number of peptides towards enkephalinase in its membrane-bound state helped to design potent inhibitors such as thiorphan [5] and phosphoryl-Leu-Phe [6], which proved to be invaluable tools to elucidate the physiological role of this enzyme. Thus, when these inhibitors are administered in the cerebral ventricles of rats or mice, both display naloxone-reversible antinociceptive activity, thereby demonstrating the involvement of enkephalinase in the in vivo degradation of opioid peptides, most likely the enkephalins [7].

Peter R. Schofield's present address is Laboratory of Molecular Neurobiology, ZMBH, Universitat Heidelberg, Im Neuenheimer Feld 282, D-69(X) Heidelberg, FRG.

Received April 27, 1988; accepted September 30, 1988.

© 1989 Alan R. Liss, Inc.

278:JCB Gorman et al. Although initially characterized in brain [3], enkephalinase activity was later

found to be present in many peripheral organs [8], including the kidney, where activity is highest. It was soon realized that, in this organ, enkephalinase is identical with an enzyme identified several years before using the B chain of insulin as substrate, the so-called neutral endopeptidase [9-12]. The detailed characteristics and function of this enzyme had however remained obscure. Enkephalinase has now been purified from many organs. Estimates of the molecular weight of the enzyme in these studies have ranged from 90 to 94 Kd. This variation is thought to be due to the fact that enkephalinase is a glycoprotein, and that its degree of glycosylation may vary [13].

Recently, we have isolated cDNA clones encoding rat enkephalinase in XgtlO libraries constructed from both brain and kidney mRNA [14], demonstrating the co-identity of the enzyme from both sources. In addition. Southern blot analysis of rat DNA indicated that a single gene encodes enkephalinase in the rat genome [14]. Rat enkephalinase is a 742 amino acid protein, which contains six potential sites for N-linked glycosylation. Its single putative membrane spanning region is located close to the amino terminus of the protein, suggesting that the bulk of this ectopeptidase, including its carboxy terminus, is located extracellularly [14,15].

We now report the production of enzymatically active recombinant rat enke-phalinase from mammalian cells. We show that, based on the inhibitory potency of various compounds, the recombinant enkephalinase is similar to the native enzyme.

MATERIALS AND METHODS

A full-length cDNA for the rat enkephalinase gene was constructed from two overlapping partial clones, XB16 and XK5, described in Malfroy et al. [14]. An 1,190-base pair (bp) Hindill (site in M13 poly linker)-B^/II fragment of XB16 was ligated to a 1,290-bp BgHl-EcoKL (site in cDNA adapter) fragment of XK5 at the Bglll site after blunt ending the Hindill and Ecc?!^.sites. The resulting 2,480-bp fragment was ligated into SmaI-clQ2LY&d pSP64 (prENKanti). The cDNA was further subcloned, from this intermediate, as a blunt ended 2,520-bp Hindlll-Sacl fragment into a mammalian expression vector based on a human cytomegalovirus (CMV) immediate early gene promoter, yielding the recombinant pCISrENK. The vector, described by Eaton et al. [16], had the factor Vin cDNA removed by cleavage with Clal and Hpal and was blunt ended with T4 DNA polymerase. All subclonings were performed by standard techniques [17].

Transient expression in Cos 7 cells was carried out using the calcium phosphate transfection technique of Graham and van der Eb [18] as modified by Gorman [19]. Ten ug of pCISrENK DNA was used to make 1 ml of precipitate. One microgram of plasmid DNA containing the adenovirus VA RNA genes [20] was included in the precipitate. Culture dishes (60 mm, 80% confluent) were transfected with 0.5 ml of precipitate and the cells assayed for expression 36 h later. For detection by immuno-logical methods cells were fixed with methanol:acetone (50:50) and stained with a rabbit polyclonal antibody to the rat enkephalinase protein followed by a peroxidase-conjugated second antibody (goat antirabbit from Dako Corp., Santa Barbara, CA). Detailed methods are described in Gorman et al. [20].

To assay for expression of enkephalinase, culture medium was removed, and 2 ml of 5 mM Hepes buffer (pH 7.4) containing 0.1% Triton X-100 were added to the 80:CPCM

Expression of Recombinant Enkephalinase JCB:279 cells. The dishes were gently shaken until all cells had detached. The cells were then collected in plastic tubes and kept at 4°C for at least 16 h to maximize protein solubilization before "Western" blot analysis or measurement of enzymatic activity was undertaken. For Western blot analysis, the solubilized extracts were separated by polyacrylamide gel electrophoresis under denaturing conditions and electroblotted onto nitrocellulose as described by Towbin et al. [22]. The blot was incubated with a rabbit polyclonal antibody raised against a synthetic peptide fragment (residues 212 to 227) of the rat enkephalinase polypeptide chain conjugated to soybean trypsin inhibitor, and developed using a goat antirabbit second antibody conjugated with horseradish peroxidase (Bio-Rad). Enkephalinase activity was measured by the method of Llorens et al. [23] using ^H-(DAla^, Leu^) enkephalin as substrate.

RESULTS The expression vector for the rat enkephalinase cDNA, pCISrENK, uses the

human cytomegalovirus (CMV) enhancer and promoter [24] to initiate transcription. This vector also contains a chimeric splice-intron region composed of sequences from the CMV immediate early gene and a synthetic splice acceptor [16]. Importantly, for high levels of transient expression in Cos 7 cells [25], this vector includes the simian virus 40 (sV40) origin of replication contained in the 8V40 early promoter region which, in this vector, directs expression of the mouse dhfr cDNA [26].

The full-length rat enkephalinase cDNA was constructed by appropriately fusing two partial cDNA clones, XB16 and XK5. Although the former clone was derived from brain mRNA and the latter from kidney mRNA, DNA sequence identity and Southern blot analysis had demonstrated that a single gene encodes enkephalinase in the rat genome [14]. The resulting cDNA thus contains 99 bp of 5' untranslated sequence, the entire 742 amino acid coding region, and 128 bp of 3' untranslated sequence. The structure of the enkephalinase expression construct, pCISrENK, is shown in Figure 1.

pCISrENK DNA was assayed for transient expression in COS 7 cells with the addition of the adenovirus virus-associated (VA) RNA genes [21], The addition of the VA RNA genes has been shown to increase the level of protein expression during transient expression [27]. Expression of the rat enkephalinase protein in transfected mammalian cells was assayed by immunoperoxidase staining. Approximately 20% of the cells (Fig. 2) were found to express enkephalinase transiently.

Cells were harvested and solubilized with Triton X-100. After polyacrylamide gel electrophoresis, the solubilized extracts were subjected to Western blot analysis,

cmv enhancer and promoter rat enkephalinase c'DNA

V

SV40 early poly A promoter

dhfr HBsAg c'DNA poly A

• • • • • • • • • • • • t t t t t»tt t t t t

Fig. 1. pCISrENK expression plasmid. The vector diagram includes from left to right the cytomegalo-virus enhancer and promoter [24], a splice donor-acceptor [16], the rat enkephalinase cDNA flanked by the SV40 poly(A) addition site, the SV40-dihydrofolate reductase (dhfr) transcription unit [26], and hepatitis B surface antigen poly(A).

CPCM:81

280:JCB Gorman et al.

^ # J .

Fig. 2. Immunological detection of recombinant eakephalinase on Cos 7 cells. Cells were fixed with methanol:acetone (50:50) 36 h after transfection with the pCISrENK expression plasmid and stained with a rabbit polyclonal antibody to the rat kidney enkephalinase protein. ( x 100).

116 92.5 66.2

Fig. 3. Western blot analysis of solubilized extracts f rom pCISrENK-transfected Cos 7 cells. Triton X-lOO-solubilized Cos 7 proteins were separated by polyacrylamide gel electrophoresis under denaturing conditions and electroblotted onto nitrocellulose paper. The blot was incubated with a rabbit polyclonal antibody raised against a synthetic peptide fragment of the enkephalinase protein (fragment 212-227) .

using the rabbit polyclonal antibody described above. These antibodies detected a 94-Kd protein (Fig. 3), further suggesting that the enkephalinase gene was being ex-pressed by the transfected cells. This 94-Kd protein was not detected when pre-immune serum was used.

To determine if the transfected cells were expressing enzymatically active enkephalinase, we assayed enkephalinase activity in..solubilized Cos 7 cells transfected with pCISrENK, using a specific enkephalinase substrate, ^H-(DAla^,Leu^) enkeph-alin [23], Enkephalinase activity could be detected as early as 1 day after transfection (Fig. 4). Activity was maximal (equivalent to the activity of about 2 [xg native

82:CPCM

Expression of Recombinant Enkephalinase JCB:281

2 3 Days in Culture

Fig. 4. Time course of expression of recombinant enkephalinase in pCISrENK-transfected Cos 7 cells. Enkephalinase activity was measured in solubilized Cos 7 cells or culture medium, following transfection with the pCISrENK expression plasmid. Activity is expressed as micrograms per 10" cells, assuming a specificity constant for the hydrolysis of ^H-(DAla^, Leu^) enkephalin by recombinant enkephalinase similar to its value by the native enzyme (35 /xmol"^ -min" ' ) [29]).

1 0 0 *

80 • >

t) < 0) 60

J? c 40 S Q> y

C Q. LU — 20

Phe-Gly EDTA

Thiorphan Phosphoramidon

[Inhibitor] (-log M)

Fig. 5. Effect of various inhibitors on recombinant enkephalinase activity from solubilized Cos 7 cells. Enkephalinase activit>' (duplicate determinations varying by 5 % or less) was measured using ^H-(DAla^, Leu^) enkephalin as substrate, in the presence of increasing concentrations of inhibitors.

enkephalinase, in a 10 cells dish) after 3 days and slightly decreased afterwards. We also assayed enkephalinase activity in the culture media of the transfected cells and found a low activity that became detectable from the second day after transfection (Fig. 4). In some experiments, the transfected cells were harvested without added Triton X-100, lysed using a Polytron homogenizer, and a particulate fraction was prepared by centrifugation. Enkephalinase activity was only detectable in the partic-ulate fraction (not shown) indicating that the enzyme is expressed in a membrane-associated state. Control, untransfected cells were totally devoid of enkephalinase activity.

To determine if recombinant enkephalinase activity has the same properties as the natural enzyme, we studied the effects of various known inhibitors of enkephali-nase. As shown in Fig. 5, recombinant enkephalinase activity solubilized from Cos 7 cells 5 days after transfection could be inhibited by thiorphan (ICso^ 1 nM), phosphoramidon (IC50 = 2 nM), the dipeptide Phe-Gly (ICso^ 2 /xM) and EDTA

C P C M : 8 3


(IC5o= 0.5 mM). These IC50 values agree with those obtained on the natural enzyme [5,11,12,27,28]. We also studied the effects of amidation in position P2' on the sensitivity of peptide substrates to hydrolysis. When both used at a concentration of 5 fiM, the fluorescent substrate dansyl-Gly-Tyr-Gly-NH2 was hydrolyzed at a velocity 11% ± 1% of that of the carboxylated substrate dansyl-Gly-Tyr-Gly (mean and standard deviation from 3 determinations).

DISCUSSION

We have shown that the metallo peptidase enkephalinase (neutral endopeptidase; EC 3,4.24.11) can be expressed in its membrane-bound state in mammalian cells, and that the recombinant enzyme is enzymatically active and displays properties similar to those of the native enzyme with respect to sensitivity to inhibitors.

Three major post-translational modifications are known to occur in the biosyn-thesis of native enkephalinase: formation of disulfide bridges, glycosylation, and insertion of one atom of zinc. Twelve cysteine residues are found in the amino acid sequence of rat enkephalinase, and it has been suggested that there are at least four disulfide bridges in the native enzyme [30]. The location of these disulfide bridges has not yet been determined, but it is likely that their correct formation is necessary to maintain the tertiary structure of the enzymatically active enzyme.

Native enkephalinase is a glycoprotein. The extent of glycosylation has been reported to vary depending on the source of enzyme [13]. It is not yet clear if glycosylation plays any role in the enzymatic activity of the enzyme. However, the properties of enkephalinase from kidney and intestine, two organs where the enzyme is differently glycosylated, have been reported to be identical [31], suggesting that the degree of glycosylation of the enzyme may not have a critical importance for its enzymatic activity.

The final, major post-translational modification is the incorporation of one atom of zinc per enkephalinase molecule [32], and a potential binding site for the metal has been recendy proposed [15] based on similarities with the binding site of zinc in the bacterial metallopeptidase, thermolysin. EDTA-treated enkephalinase has been shown to be easily reactivated upon addition of micromolar concentrations of zinc salts [32,33], suggesting that this step in the biosynthesis of the enzyme can occur sponta-neously, provided a high enough concentration of the metal is present.

Active recombinant enkephalinase with properties similar to those of the native enzyme with respect to its sensitivity to various inhibitors is produced by the trans-fected Cos 7 cells (Fig. 5), showing that the folding of the polypeptide chain and the formation of the disulfide chains probably occurs correctly. Furthermore, the inhibi-tion by EDTA of the enzymatic activity displayed by the recombinant enzyme implies the presence of a metal in its active site. Finally, we have found that the Triton X-100-solubilized enzyme binds to concanavalin-A (not shown), suggesting that the recombinant enzyme is glycosylated. This is also suggested by the molecular weight of the enzyme, 94 Kd (Fig. 3), which is higher than the predicted 85-Kd molecular weight of the unglycosylated enkephalinase polypeptide chain.

Native enkephalinase is a membrane-bound enzyme. The determination of the amino acid sequence of the enzyme, as deduced from cDNA clones [14,15] showed the presence of a single putative membrane spanning region. This transmembrane region, located close to the aminoterminus of the enkephalinase polypeptide chain

84:CPCM

Expression of Recombinant Enkephalinase JCB:283

(residues 21-43) is preceded by a highly charged and conformationnally restrained region, Pro-Lys-Pro-Lys-Lys-Lys-Gln-Arg (residues 8 to 15), which probably serves as a stop transfer sequence [34] to terminate the transfer of the polypeptide chain through the cellular lipid membrane. In agreement with these hypotheses, we have found that, like the native enzyme, recombinant enkephalinase as produced in the Cos 7 cells is membrane-bound. The enkephalinase gene, as deduced from molecular cloning, does not encode a recognizable signal peptide [14]. It is likely that the single transmembrane region of the enkephalinase polypeptide plays the role of both signal peptide and membrane spanning/anchoring domain. The appearance of a small but significant amount of soluble enkephalinase activity in the culture medium, 2 days after transfection of the Cos 7 cells, suggests that enkephalinase may be released from the cell membranes by proteolytic activity. That such a process could yield enzymat-ically active enkephalinase is not unexpected since solubilization of native enkephali-nase in an active form has been succesfully achieved by treating membranes with trypsin [32] or papain [30]. Additionally, it is possible to detect soluble enkephalinase activity in human plasma, cerebrospinal fluid [35] and urine (B. Malfroy, unpub-lished), suggesting that such a proteolytic solubilization of enkephalinase may also occur in vivo.

Recombinant enkephalinase as produced from Cos 7 cells displayed properties similar to those of the native enzyme as estimated by the potency of various inhibitors, with affinities ranging from nanomolar to millimolar concentrations (Fig. 5). We also found, using the fluorescent substrates dansyl-Gly-Tyr-Gly and dansyl-Gly-Tyr-Gly-NH2 [36], that recombinant enkephalinase shows a clear preference for substrates displaying a free carboxyl end in position P2', as has been shown to be the case for the native enzyme [12,29,36-38]. It has been suggested that this preference is due to the interaction of the free carboxyl group of substrates with an arginyl residue in the active site of enkephalinase [12,37-39], which would play a role similar to that of Arg 145 in bovine carboxypeptidase A [40]. The availability of a system to produce active recombinant enkephalinase will allow the examination of such questions by structure/function studies using site-specific mutagenesis.

REFERENCES

1. Schwartz JC, Malfroy B, De La Baume S: Life Sci 29:1715-1740, 1981. 2. Hersh LB: Mol Cell Biochem 47:35-43, 1982. 3. Malfroy B, Swerts JP, Guyon A, Roques BP, Schwartz JC: Nature 276:523-526, 1978. 4. Giros B, Gros C, Solhonne B, Schwartz JC: Mol Pharmacol 29:281-287, 1986. 5. Roques BP, Fournie-Zaluski MC, Soroca E, Lecomte JM. Malfroy B, Llorens C, Schwartz, JC:

Nature 288:286-288, 1980. 6. Altstein M, Bachar E, Vogel Z, Blumberg S: Eur J Pharmacol 91:353-61, 1983. 7. Lecomte JM, Costentin J, Vlaiculescu A, Chaillet P, .Marcais-Collado H, Llorens-Cortes C,

Leboyer M, Schwartz JC: J Pharmacol Exp Ther 237:937-944, 1986. 8. Llorens C, Schwartz JC: Eur J Phamnacol 69:113-116, 1981. 9. Kerr MA, Kenny AJ: Biochem J 137:477-488, 1974.

10. Almenoff J, Wilk S, Orlowski M: Biochem Biophys Res Commun 102:206-214, 1981. 11. Fulcher IS, Matsas R, Turner AJ, Kenny AJ: Biochem J 203:519-522, 1982. 12. Malfroy B, Schwartz JC: Biochem Biophys Res Commun 106:276-285, 1982. 13. Fulcher IS, Chaplin Mf, Kenny AJ: Biochem J 215:317-323, 1983. 14. Malfroy B, Schofield PR, Kuang WJ, Seeburg PH, Mason AJ, Henzel WJ: Biochcm Biophys Res

Commun 144:59-66, 1987.

CPCM:85


15. Devault A, Lazure C, Nault C, Le Moual H, Seidah NG, Chretien M, Kahn P, Powell J, Mallet J, Beaumont A, Roques BP, Crine P, Boileau G: EMBO J 6:1317-1322, 1987.

16. Eaton DL, Wood WI, Eaton D, Hass PE, Hollingshead P, Wion K, Mather J, Lawn RM, Vehar GA, Gorman C: Biochemistry 25:8343-8347, 1986.

17. Maniatis T, Fritsch EF, Sambrook J: "Molecular Cloning: A Laboratory Manual." Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press, 1982.

18. Graham FL, van der Eb A: Virology 52:456-467, 1973. 19. Gorman C: In Glover DM (ed): "DNA Cloning IL"Washington DC: IRL Press, 1985, pp 143-190 20. Gorman CM, Rigby PWJ, Lane D: Cell 42:519-526, 1985. 21. Thimmappaya B, Weinberger C, Schneider RJ, Shenk T: Cell 31:543-551, 1982. 22. Towbin H, Staehelin T, Gordon J: Proc Natl Acad Sci USA 76:4350-4354, 1979. 23. Llorens C, Malfroy B, Schwartz JC, Gacel G, Roques BP, Roy J, Morgat JL, Javoy-Agid F, Agid

Y: J Neurochem 39:1081-1089, 1982. 24. Boshart M, Weber F, John G, Dorsch-Hasler K, Fleckinstein B, Schaffner W: Cell 41:521-530,

1985. 25. Gluzman Y: Cell 23:175-182, 1981. 26. Simonsen CC, Levinson AD: Proc Nad Acad Sci USA 80:2495-2499, 1983. 27. Svensson C, Akusjarvi G: EMBO J 4:957-964, 1985. 28. Llorens C, Gacel G, Swerts JP, Perdrisot R, Fournie-Zaluski MC, Schwartz JC, Roques BP:

Biochem Biophys Res Commun 96:1710-1716, 1980. 29. Malfroy B, Schwartz JC: J Biol Chem 259:14365-14370, 1984. 30. Tam LT, Engelbrecht S, Talent JM, Gracy RW, Erdos EG: Biochem Biophys Res Commun

133:1187-1192, 1985, 31. Relton JM, Gee NS, Matsas R, Turner AJ, Kenny AJ: Biochem J 215:519-523, 1983. 32. Kenny AJ: In Barrett AJ (ed): "Proteinases in Mammalian Cells and Tissues." Amsterdam:

Elsevier/North-Holland Biomedical Press, 1977, pp 393-444. 33. Malfroy B, Llorens C, Schwartz JC, Soroca E, Roques BP, Roy J, Morgat JL, Javoy-Agid F, Agid

Y: In Takagi H, Simon EJ (eds): "Advances in Endogenous and Exogenous Opioids." Amsterdam: Elsevier/North-Holland Biomedical Press, 1981, pp 191-194.

34. Blobel G: Proc Natl Acad Sci USA 77: 1496-1500, 1980. 35. Spillantini MG, Farciallaci M, Michelacci S, Sicuteri F: IRCS Med Sci 12:102-103, 1984. 36. Malfroy B, Burnier J: Biochem Biophys Res Commun 143:58-66, 1987. 37. Malfroy B, Schwartz JC: Biochem Biophys Res Commun 130:372-378, 1985. 38. Jackson D, Hersh LB: J Biol Chem 261:8649-8654, 1986. 39. Beaumont A, Roques BP: Biochem Biophys Res Commun 139:733-739, 1986. 40. Quiocho FA, Lipscomb WN: Adv Protein Chem 25:1-48, 1971.

86:CPCM

The EMBO Journal vol .8 no.3 p p . 6 9 5 - 7 0 0 , 1989

Functional expression in Xenopus oocytes of the strychnine binding 48 kd subunit of the glycine receptor

V.Schmieden, G.Grenningloh, P.R.Schofield^ and H.Betz Zentrum fiiir Molekulare Biologie, Universitat Heidelberg, Im Neuenheimer Feld 282, 6900 Heidelberg, FRG 'Present address: Pacific Biotechnology Ltd, 74 McLachlan Ave. , Rushcutters Bay, NSW 2011, Australia Communicated by P.Seeburg

The inhibitory postsynaptic glycine receptor (GlyR) of rat spinal cord is an oligomeric transmembrane protein which forms an agonist-gated anion channel. Expression in Xenopus oocytes of its mol. wt 48 000 subunit generated glycine-gated chloride channels which were analysed by voltage clamp. The agonist and antagonist response properties as well as the desensitization characteristics of these 48 kd subunit receptors resembled GlyRs expressed from spinal cord polyCA)"*" RNA. These data indicate that the 48 kd subunit is capable of assembling into a functional receptor homo-oligomer which displays the pharmacology characteristic of the spinal cord GlyR. Key words: glycine receptor/chloride cYmrntMXenopus oocyte/strychnine/voltage clamp

Introduction In vertebrates, the amino acid glycine mediates postsynaptic inhibition of most sensory and motor pathways by increasing the chloride conductance of the neuronal plasma membrane (Curtis et al, 1968; Aprison and Daly, 1978). The plant alkaloid strychnine blocks postsynaptic chloride currents elicited by glycine and is a highly selective antagonist of the inhibitory glycine receptor (GlyR) (Young and Snyder, 1973; Betz and Becker, 1988). The mammalian GlyR has been purified by affinity chromatography and shown to contain two types of subunits of mol. wts 48 000 and 58 000, in addition to a mol. wt 93 000 receptor-associated peripheral membrane protein (Pfeiffer etal, 1982; Graham etal, 1985; Becker et al, 1986; Schmitt etal, 1987). Photo-affinity labelling experiments have localized the strychnine binding site of the GlyR on the 48 kd polypeptide (Graham etal, 1981, 1983; Pfeiffer etal, 1982). The 48 kd and 58 kd subunits are homologous transmembrane proteins (Pfeiffer et al, 1984; G.Grenningloh and P.Prior, unpub-lished) assembled in a pentameric core structure which forms the glycine-gated chloride channel (Betz, 1987; Langosch etal, 1988).

Injection of poly(A)"^ RNA, isolated from mammalian brain, into Xenopus oocytes results in incorporation of functional GlyR into the oocyte membrane (Gundersen et al, 1984; Houamed etal, 1984; Sumikawa etal, 1984). Recently, the primary structure of the 48 kd GlyR subunit has been deduced from cloned cDNA and shown to exhibit ©IRL Press

sequence and structural homology to nicotinic acetylcholine and, in particular, to G A B A A receptor polypeptides (Grenningloh et al, 1987a,b; Schofield et al, 1987). From protease digestion experiments (Graham et al, 1983) and sequence analysis (Grenningloh et al, 1987a) the glycine-and strychnine-binding sites of the GlyR have been assigned to the extracellular N-terminal domain of this protein. To show that the 48 kd polypeptide indeed carries receptor sites for agonists and antagonists, we have investigated its expression in heterologous cell systems. Here we show that injection into Xenopus oocytes of a single RNA species encoding the 48 kd subunit creates glycine-gated chloride channels. These channels exhibit pharmacological and functional characteristics typical of the GlyR produced by expression of rat spinal cord poly(A)"^ RNA.

Results Expression of the 48 kd subunit in Xenopus oocytes generates glycine-gated chloride channels Synthetic mRNA encoding the 48 kd subunit of the GlyR was generated by in vitro transcription of a recombinant pSP64 vector containing a full-length 48 kd subunit cDNA. This cDNA was assembled using oligonucleotides which encoded translational initiation signals and a portion of the signal peptide not present in the cloned cDNA sequence (see Materials and methods). Xenopus oocytes were injected with this RNA and, after 3 - 5 days, analysed for glycine responses by voltage clamp.

Of 15 oocytes tested initially, 10 exhibited large (100 nA up to several iiPC) inward currents at negative holding potentials upon bath application of 250 (or 500) fiM glycine (Figure 1). Non-injected oocytes (n = 10) were glycine-insensitive. No response was seen after exposure of injected oocytes to millimolar concentrations of the neurotransmitter amino acids glutamate and GAB A (not shown). Glycine-elicited currents produced at different holding potentials had a reversal potential of - 3 0 to - 2 0 mV, which corresponds to the equilibrium potential of chloride in oocytes. At hyper-polarizing potentials beyond - 5 0 mV the currents showed marked rectification, consistent with the voltage dependence of GlyR channel gating (Bormann et al, 1987; Faber and Korn, 1987). These data indicate that the 48 kd subunit is capable of forming functional GlyR, presumably via assembly of a homo-oligomeric receptor protein. Pharmacology of the 48 kd subunit GlyR The channels formed upon expression of the 48 kd subunit were activated not only by glycine, but also by the GlyR agonists /3-alanine and taurine (Barker and Ransom, 1978a; Betz and Becker, 1988). As shown in Figure 2, the response to glycine and jS-alanine of the 48 kd subunit receptor was indistinguishable from that observed in oocytes injected witii poly(A)"^ RNA isolated from rat spinal cord at postnatal day 20 (P20). In both cases, half-maximal responses were

695

V.Schmieden et al.

r

Glydne Glydne

-40

-50

u -70

IjjA

B

400 -

2 0 0 -

- 2 0 0 -

-400 -120 -100 - 8 0 -60 - 4 0 -20 0 20

P o t e n t i a l (mV)

Fig. 1. Current-voltage relation for glycine-induced currents in Xenopus oocytes injected with 48 kd subunit RNA. (A) Membrane currents elicited by 250 IJM glycine, recorded from a well-responding oocyte. Time-course and amplitude are indicated by bars. Downward deflections denote inward current that reverses to outward current near —30 mV. The holding potentials used are indicated in mV. (B) Current-voltage relation for another oocyte exhibiting a smaller response to 500 |tM glycine. Measurements were obtained by stepping the holding potential from - 7 0 mV to indicated potentials 10 s prior to superflision with agonist. Glycine-activated membrane currents at each potential were obtained after subtraction of leakage current.

obtained with 300-400 /xM glycine or 2 mM |Q-alanine. In all experiments, corresponding Hill coefficients were close to /7 = 3 for glycine, and n = 1 for /S-alanine (Figure 2).

A significant difference in agonist sensitivity of the two types of receptors was found in case of taurine. In accord with previous electrophysiological and ligand-binding data (see Betz and Becker, 1988), taurine was the least efficient agonist tested (Figure 3). Concentrations of > 5 mM were required for half-maximal responses (Hill coefficient n = 1.0 for 48 kd subunit channels; data not shown). A consistently stronger relative response to taurine was observed in oocytes injected with 48 kd subunit RNA than in those injected with spinal cord poly(A)^ RNA prepar-ations (Figure 3a,b). In addition, analysis of oocytes injected with poly(A)'^ RNA isolated from different developmental stages revealed particularly low taurine responses with RNA from newborn animals (Figure 3c). No significant dif-ferences in the relative potency of /3-alanine were observed

-log M Glycine -log M .B-Alanine

Fig. 2. Membrane currents elicited by glycine and /3-alanine in 48 kd subunit RNA and spinal cord poly(A)^ RNA injected oocytes. (A,B) Current responses to 500 /xM glycine and 10 mM /S-alanine of oocytes injected with 48 kd subunit RNA (A) and spinal cord (P20) polyCA)"*" RNA (B) respectively. All measurements were made at a holding potential of —70 mV. Perfusion times are indicated by bars. (C,D) Dose—response curves for glycine (C) and /3-alanine (D) of oocytes injected with P20 poly (A) RNA (X) and 48 kd subunit RNA (O). The data are plotted in semilogarithmic coordinates. Agonist concentrations eliciting half-maximal responses (400 fiM for glycine and 2 mM for /3-alanine) are indicated by arrows and are indistinguishable for both types of RNA. Currents are normalized to maximal responses observed at 10 mM glycine and 100 mM ;8-alanine respectively. The insets show Hill plots of the same data indicating Hill coefficients of about n = 3 for glycine and n = 1 for /3-alanine, respectively. This experiment was repeated six times on both oocyte populations with similar results.

between the different RNA samples in the same oocytes. These data indicate partial agonist discrimination by the two channel types investigated here and may reflect GlyR mRNA heterogeneity in the spinal cord poly (A) RNA samples.

Superfusion of injected oocytes with both glycine and taurine, or glycine and /3-alanine, revealed unexpected inter-actions between the different glycinergic agonists. In the presence of taurine or /3-alanine, glycine responses were markedly reduced compared to those elicited by glycine alone (Figure 4). Similarly, /3-alanine-induced currents were inhibited by simultaneous taurine application (not shown). These inhibitory effects could not be attributed to enhanced desensitization by the other agonist, as prior perfusion with taurine (Figure 4a) or /S-alanine (not shown) did not alter subsequent glycine responses. Analysis of /3-alanine-(Figure 4b) and taurine- (not shown) mediated inhibition over a range of glycine concentrations revealed that the second agonist reduced apparent glycine affinities. These agonist interactions were observed in oocytes injected with both spinal cord poly(A)+ RNA (PO, P20, P40) and 48 kd subunit RNA (Figure 4, and data not shown).

Strychnine is the classical antagonist of the mammalian GlyR (Curtis et al, 1968; Young and Snyder, 1973). The currents induced by glycine in injected oocytes were efficiently blocked by 20-100 nM concentrations of the alkaloid (K^ values 9 - 1 5 nM in different experiments). Comparable results were seen with 48 kd subunit RNA and spinal cord poly(A)"^ RNA (P20, P40) injections (Figure 5). Responses to |S-alanine and taurine were similarly strychnine-sensitive (not shown). The response of 48 kd

696

Glycine receptor expression

Gtycine 0.5 mM

^-Alanine 10 mM

Taurine 10 mM

50 nA

100 nA

o o ClO £ 5 3 «> " I

1 -

0.8-

0.6-

0.4 •

0.2

0

1 1 1

20 sec

• PO n=5 0 P20 n=4 m P40 n=3 fZi 48K n=7

mm ^-Alanine Taurine Fig. 3. Comparison of current responses to glycine, /3-alanine and taurine of oocytes injected witii spinal cord (P20) poly(A)"^ RNA (A) and 48 kd subunit RNA (B) respectively. Responses were determined at the indicated agonist concentrations using a holding potential of - 7 0 mV. Between agonist perfusions, oocytes were washed for 5 min with Ringer solution. Control experiments confirmed that after this washout period full agonist responses were recovered. The time-scale shown is valid for (A) and (B). (C) Mean relative responses to alanine and taurine of oocytes injected with different RNA preparations. Responses to 500 /iM glycine, 10 m M /3-alanine and 10 mM taurine were determined from records as shown in (A) and (B) using oocytes injected with PO, P20 and P40 poly (A) RNA preparations as well as with 48 kd subunit RNA. Current responses to /3-alanine and taurine are shown as a fraction of the response obtained with 500 /tM glycine and represent average values from 3—7 oocytes; bars indicate minimal and maximal deviations.

subunit RNA injected oocytes was also antagonized by picro-toxin, a drug known to block inhibitory GABA^ receptor channels (Olsen, 1982), but not [^H]strychnine binding to spinal cord membranes (Graham et al., 1982). The inhibition by picrotoxin occurred over a broader concentration range than that mediated by strychnine (Figure 5), indicating that these two antagonistic compounds act by different binding mechanisms (Hill coefficients n = 1.4 for strychnine, and n = 0.6 for picrotoxin). Addition of pentobarbital (0.24 or 2.4 mM), a drug which enhances inhibitory GABA^ responses (Barker and Ransom, 1978b), did not affect responses of 48 kd subunit expressing oocytes to either glycine (500 fiM), /S-alanine (10 mM) or taurine (10 mM). Desensitization of glycine responses In all injected oocytes, glycine currents exhibited a noticeable decline during maintained agonist application at concen-trations >300 /iM (Figure 6). In several experiments, no significant differences in this desensitization behaviour of the channels expressed from 48 kd subunit (n ^ 9) and PO, P20 or P40 spinal cord (n = 5) RNAs were found. The

kinetics of desensitization could be fitted by two exponential ftinctions characterized by time constants of 20—40 s and 250-350 s respectively. These time constants displayed no measurable voltage sensitivity over the range of membrane potentials tested ( — 120 to —50 mV). Furthermore, a similar desensitization was seen upon superfusion of high concen-trations of jS-alanine (not shown). Apparently, desensitization previously described for the GlyR of spinal cord neurons (Barker and McBurney, 1979; Hamill etal, 1983) is maintained upon expression of the ligand-binding 48 kd GlyR subunit.

Discussion Expression in Xenopus oocytes of a single GlyR subunit, the 48 kd polypeptide, resulted in formation of glycine-gated strychnine-sensitive chloride channels. The 48 kd subunit thus harbours not only agonist and antagonist binding sites as well as chloride-channel-forming domains, but is also capable of assembling into a functional GlyR. Although co-assembly with oocyte-derived polypeptide(s) cannot be excluded, the resulting receptor channels probably represent homo-oligomeric structures. In accord with this interpret-ation, expression of the 48 kd subunit in mammalian cells also results in incorporation of functional GlyR into the plasma membrane (Sontheimer et al, 1989). Further-more, biochemical and immunological (Pfeiffer et al., 1984) as well as recent cDNA cloning data (G.Grenningloh and P.Prior, unpublished) indicate considerable structural homology of 48 kd and 58 kd GlyR subunits. Thus, additional copies of the 48 kd polypeptide may substitute for the 58 kd polypeptide in the pentameric channel core structure of the GlyR (Langosch et al., 1988). Interestingly, formation of functional receptor channels upon expression of a single subunit in oocytes and mammalian cell lines has also been observed for nicotinic acetylcholine (Boulter et al., 1987) and GABA a (Blair etal., 1988; Pritchett etai, 1988) receptor proteins. However, in these cases agonist responses were much lower than those found here upon expression of the 48 kd subunit. This may reflect a particularly efficient assembly of 48 kd subunit homo-oligomers and/or a higher conductance of homomeric glycine receptor channels. Single-channel recordings will be required to distinguish between these alternatives.

The efficient expression of 48 kd subunit GlyR observed here raises the question whether homomeric forms of the receptor might also exist in vivo. Indeed, biochemical data indicate that the GlyR expressed in primary cultures of embryonic mouse spinal cord may comprise only the strychnine-binding subunit (W.Hoch, H.Betz and C.-M. Becker, in preparation). Homo-oligomeric channel formation as demonstrated in this study may also account for the recent observation of different GlyR responses upon expression of size-fractionated RNA samples isolated from different regions of the rat central nervous system (Akagi and Miledi, 1988). In light of our data, oocyte expression studies using size-fractionated mRNAs which encode multisubunit channels (see Sumikawa et al., 1984) must be viewed with some caution.

The pharmacological properties of the expressed 48 kd subunit GlyR closely resemble those of the receptor encoded by spinal cord poly(A)+ RNA. In particular, both types of receptors displayed identical dose responses to the agonists

697

V.Schmieden et al.

A

Glycine

1/

Glyc ine - Taur ine . G l ^ c ^

1/ 1/

10" 10" 10-3 10-2 Glycine (M)

Fig. 4. Inhibition of glycine-induced currents by taurine and /3-alanine. Oocytes injected with spinal cord (PO) polyCA)"*" RNA (A) and 48 kd subunit RNA (B) were superfused with 500 /iM glycine, or 10 mM /3-alanine or taurine, or combinations of these amino acids. Duration of agonist application is indicated by bars. The traces shown in (A) and (B) represent subsequent recordings from the same oocytes. (B, right) Dose response relation for glycine in the absence ( • ) and presence ( • ) of 700 fxM jS-alanine. Amplitudes were normalized to maximal responses obtained at 10 mM glycine. Half-maximal responses to glycine are shifted from - 2 9 0 iM to 700 /nM by addition of the second agonist.

LI 10 9 8 7 6 5

- log M Antagonist

Fig. 5. Inhibition of glycine responses by strychnine (A,B) and picrotoxin (C). Current responses to glycine (A,B: 500 /xM; C: 200 IJ.M) of oocytes injected with spinal cord (P20) poly(A)+ RNA (B) and 48 kd subunit RNA (A,C) were determined in the presence of the indicated concentrations of antagonist as detailed under Materials and methods. Data are plotted as fraction of the currents observed in the absence of antagonist. Antagonist concentrations producing half-maximal inhibition (IC50) are indicated by arrows (strychnine 3 0 - 4 0 nM, and picrotoxin - 0 . 8 (M). Holding potential was - 7 0 mV.

glycine and jS-alanine as well as to the classical GlyR antagonist strychnine. These data argue against the recently postulated existence of a separate strychnine-sensitive receptor for /5-alanine (Parker etal., 1988), but are consistent with previous notions that responses to /3-alanine and taurine may be exclusively mediated by GlyR and G A B A a receptors (Barker etal., 1982; Pfeiffer etal, 1982; Choquet and Korn, 1988; Horikoshi et al, 1988a).

For taurine, another glycinergic agonist, higher responses were obtained for homomeric 48 kd subunit receptors than for receptors encoded by spinal cord poly(A)"^ RNA. This difference may reflect heterogeneity of 48 kd subunit mRNA in spinal cord (Becker etal., 1988; M.L.Malosio and G.Grenningloh, unpublished) or, less likely, a selective effect of the 58 kd subunit on taurine affinity. However, although ligand affinity rank order was preserved in oocytes for both

a = 0.3948

c = + 0.3948 e / SQ

80 120

Time (sec)

Fig. 6. Desensitization of glycine-activated membrane currents in Xenopus oocytes. (A) Oocyte injected with 48 kd subunit RNA. (B) Oocyte injected with spinal cord (P20) poly(A)+ RNA. Oocytes were continuously superfused with 500 ;ttM glycine for 220 s. Desensitization of current responses (b) became apparent after ~ 12 s of agonist application. (Here, upward deflection denotes inward current.) The observed decline in membrane current was fitted by two exponential functions with time constants of 240 s (a) and 30 s (not shown) respectively. Curve c simulates the desensitization current by superposition of both e-functions. Holding potential was - 7 0 mV.

receptor types, absolute agonist affinities were much lower than seen for spinal cord GlyR both in ligand-binding studies (for review see Betz and Becker, 1988) and in electro-

698

Glycine receptor expression

physiological experiments (Barker and Ransom, 1978a; Hamill et al., 1983; Bormann et al., 1987). Whereas studies on cultures and membrane preparations from spinal cord indicate K^ values of 1 0 - 1 0 0 / M for glycine, /S-alanine and taurine (Barker and Ransom, 1978; Pfeiffer et al, 1982; Betz and Becker, 1988), half-maximal oocyte responses were obtained at 300 fiM for glycine, 2 mM for /^-alanine and > 5 mM for taurine respectively. Similar observations have been made by others using expression of poly (A) RNA preparations isolated from different regions of the rodent central nervous system (Sumikawa et al., 1984; Carpenter et al., 1988; Horikoshi et al., 1988a). This difference may reflect particularities of the oocyte system and/or lack of subunits and other components required for assembly and post-translational modification of the receptor. However, no differences were seen in antagonist affinity; the K of strychnine calculated for inhibition of oocyte agonist responses ( ~ 12 nM) is almost identical to the K^ of 10 nM deduced from [^H] strychnine-binding experiments (Betz and Becker, 1988).

Simultaneous exposure to glycine and taurine, or jS-alanine, disclosed a lack of synergism of the different GlyR agonists. Rather, glycine responses were inhibited in the presence of a second agonist. This inhibition cannot be attributed to cross-desensitization as reported previously (Horikoshi et al., 1988a), but reflects a lowering of apparent glycine affinity by the other ligand. Interestingly, both taurine and /S-alanine open oocyte-expressed GlyR channels in an apparently non-co-operative fashion (Hill coefficient n = \), whereas glycine activation exhibits a steep concentration dependence (n = 3). At least three ligand-binding subunits may therefore co-operate during glycine-induced channel opening, an interpretation consistent with a subunit stoichiometry of a-^^j deduced from crosslinking of affinity-purified GlyR preparations (Langosch etal, 1988). Consequently, taurine and /S-alanine probably affect glycine responses by mutually exclusive binding resulting in abolition of glycine co-operativity, as discussed for other allosteric proteins (Rubin and Changeux, 1966). During preparation of this manuscript, an inhibition of glycine responses by taurine has also been reported for oocytes injected with brain poly(A)+ RNA (Horikoshi et al, 1988b). Because of their widespread occurrence, taurine and /3-alanine may serve as physiological modulators of glycinergic transmission in various regions of the central nervous system (see also Aprison and Daly, 1978). It should be noted, however, that presently no corresponding data are available for slice preparations or primary cultures of spinal cord. Thus the inhibitory agonist interactions discussed above could reflect a special feature of the receptor expressed in oocytes.

Differences between the GlyRs expressed in oocytes and spinal cord neurons exist for desensitization upon prolonged agonist exposure. Both 48 kd subunit RNA and spinal cord poly (A) RNA injected oocytes displayed a slow biphasic desensitization of their glycine responses of rather similar kinetics. Hence, desensitization is an intrinsic property of the ligand-binding subunit of the GlyR. However, oocyte GlyRs desensitized only poorly compared to the receptor of embryonic spinal cord cultures. Whereas spinal cord receptors exhibit efficient desensitization at glycine concen-trations > 2 0 /AM (Hamill et al., 1983), the oocyte response declined only at agonist concentrations > 300 ^M, and after several seconds of ligand occupation. This difference might

be attributed to the existence of different ligand-binding subunits in neonatal (embryonic) and adult isoforms of the GlyR (Becker et al., 1988). However, under the conditions used here oocytes injected with poly (A) + RNA preparations isolated from neonatal and adult spinal cord exhibited a similar decline of agonist responses (V.Schmieden, unpublished). Thus, the comparatively inefficient desensi-tization of the oocyte-expressed GlyR probably reflects its lower agonist affinity. Alternatively, neuron-specific modifications which enhance desensitization may not occur in oocytes. For isolated nicotinic acetylcholine receptor or Torpedo electric organ, phosphorylation has been reported to accelerate markedly the desensitization process (Huganir et al., 1986). Similar modification reactions also might affect functional properties and/or assembly of the GlyR.

Materials and methods Isolation of poly (A) + RNA Total RNA was isolated from spinal cord and brain stem of newborn (PO), young (P20) and adult (P40) rats using the method of Cathala et al (1983). Poly (A) + RNA was selected by chromatography on oligo(dT)-cellulose (Aviv and Leder, 1972).

Construction of transcription plasmid and preparation of 48 kd subunit RNA The cDNA clones GR 1-6 and GR 1-6-1 (Grenningloh et al., 1987a) encode the entire mature 48 kd subunit and six amino acids of the authentic signal peptide. A complete coding sequence was constructed in pSP64 by combining the 67 bp EcoRl-HindUl and 425 bp HindJR fragments of clone GR 1-6-1 and the 892 bp Hindlll-EcoRl fragment of clone 1-6 with synthetic signal peptide, initiation codon and 5' untranslated sequences (Sontheimer et al., 1989). In this construct, the f coRI linker at the 5' end of clone GR 1-6-1 contributed three further amino acids, and the remainder of the signal peptide was based on the corresponding sequence found in the bovine G A B A A receptor a subunit (Schofield etal., 1987). Additionally, a consensus sequence surrounding the initiation codon (Kozak, 1984) and a synthetic 5' untranslated sequence of 67 bp were inserted. The latter was identical to the synthetic Pstl fragment previously used to express GABA^ receptor a2 subunit in oocytes (Levitan et al., 1988). The resultant plasmid was used for generating synthetic capped RNA transcripts according to Melton et al. (1984) using a transcription kit (Boehringer).

Oocyte injection and electrophysiological recordings Oocytes were obtained from anaesthetized adult Xenopus laevis, carefiilly dissected and stored in sterile Barth's medium (Colman, 1984). Before injection, the follicle cell layer was removed by treatment with 1 mg/ml collagenase (type II, Sigma) for 1 h at 20°C followed by mechanical disruption. Frozen aliquots of spinal cord poly(A)"'" RNA were diluted with water to a final concentration of 1 iig/iA, and of 48 kd subunit RNA to 0.2 iig/fi]. Volumes injected per oocyte were 100 nl for poly(A)"'" RNA, and 50 nl for 48 kd subunit RNA.

After 3 - 4 days of incubation at 19°C with daily changes of the medium, current responses were measured by voltage clamp under continuous super-fusion (0.3 ml/s) with frog Ringer solution. For determining dose-response relations, superflision was switched to desired concentrations of glycine, j3-alanine or taurine in Ringer solution. Effects of antagonists were recorded after superfijsing the oocytes with strychnine or picrotoxin 1 min prior to addition of agonist. In all experiments the oocytes were washed for 3 - 5 min before agonist readdition. The two microelectrodes used for voltage clamping were filled with 3 M KCl and had a resistance of 0.5 - 1 MOhm. For data analysis, currents were recorded by a video-recorder (j3-system) and digitally stored on a personal computer.

Acknowledgements We thank C.Udri for RNA preparation, P.Seeburg, C.-M.Becker, D.Langosch, W.Hoch and B.Sakmann for helpful suggestions and critical readings of the manuscript, and I.Veit-Schirmer and I.Baro for assistance during its preparation. This work was supported by the Deutsche Forschungsgemeinschaft (SFB 317), Bundesministerium fiir Forschung und Technologic (BCT 365/1) and Fonds der Chemischen Industrie.

699

V.Schmieden et al.

References

Akagi,H. and Miledi;R. (1988) Science, 242, 270-273. Aprison,M.H. and Daly,E. (1978) Adv. Neurochem., 3, 230-294. Aviv.H. and Leder.P. (1972) Proc. Natl. Acad. Sci. USA. 69, 1403-1412. Barker,J.L. and McBumey,R.N. (1979) Nature, 277, 234-236. Barker,J.L. and Ransom,B.R. (1978a) 7. Physiol., 280, 331-354. Barker,J.L. and Ransom,B.R. (1978b) 7. Physiol, 280, 355-372. Barker,J.L., McBumey,R.N. and Macdonald,J.F. (1982)/. Physiol, 322,

365-387. Becker,C.-M., Hermans-Borgmeyer,!., Schmitt,B. and Betz,H. (1986)

J. Neurosci., 6, 1358-1364. Becker,C.-M., Hoch,W. and Betz,H. (1988) EMBO J., 7, 3717-3726. Betz,H. (1987) Trends Neurosci., 10, 113-117. Betz,H. and Becker,C.-M. (1988) Neurochem. Int., 13, 137-146. Blair,L.A.C., Levitan,E., Marshall,!., Dionne,V.E. and Barnard,E.A.

(1988) Science, 242, 577-579. Bormann,J., HamilI,O.P. and Sakmann,B. (1987) J. Physiol, 385,

234-286. Boulter,!., Connolly,!., Deneris,E., Goldman,D., Heinemann,S. and

Patrick,!. (1987) Proc. Natl Acad. ScL USA, 84, 7763-7767. Caipenter,M.K., Parker,!. andMiledi,R. (1988)Proc. K Soc. Lond., B234,

159-170. Cathala,G., Savouret,I.-F., Mendez,B., West,B.L., Karin,M., Martial,!.A.

and Baxter,!.D. (1983) DNA, 2, 329-335. Choquet,D. and Kom,H. (1988) Neurosci Lett., 84, 329-334. Colman,A. (1984) In Hames,B.D. and Higgins,!. (eds). Transcription and

Translation: A Practical Approach. IRL Press, Oxford, pp. 4 9 - 6 9 . Curtis,D.R., H6sli,L., !ohnston,G.A.R. and !ohnston,I.H. (1986) Exp.

Brain Res., 5, 235-258. Faber,D.S. and Kom,H. (1987) J. Neurosci, 7, 807-811. Graham,D., Pfeiffer,F. and Betz,H. (1981) Biochem. Biophys. Res.

Com/mm., 102, 1330-1335. Graham,D. Pfeiffer,F. andBetz,H. (mi) Neurosci Lett., 29, 173-176. Graham,D., Pfeiffer,F. and Betz,H. (1983) Eur. J. Biochem., 131,

519-525. Graham,D., Pfeiffer,F., Simler,R. and Betz,H. (1985) Biochemistiy, 24,

990-994. Grenningloh,G., Rienitz,A., Schmitt,B., Methfessel,C., Zenssen,M.,

Beyreuther,K., Gundelfinger.E.D. and Betz,H. (1987a) Nature, 328, 215-220.

Grenning]oh,G., Gundelfinger,E.D., Schmitt,B., Betz,H., Darlison,M.G., Bamard,E.A., Schofield,P. and Seeburg,P.H. (1987b) Nature, 330, 2 5 - 2 6 .

Gundersen,C.B,. Miledi,R. and Parker,!. (1984) Proc. R. Soc. Lond., B221, 235-244.

Hamill,O.P., Bormann,!. and Sakmann,B. (1983) Nature, 305, 805-808. Horikoshi,T., Asanuma,A., Yanagisawa,K., Anzai,K. and Goto,S. (1988a)

Mol Brain Res., 4, 97-105 . Horikoshi,T., Asanuma,A., Yanagisawa,K. and Goto,S. (1988b) Mol Brain

Res., 4, 243-246. Houamed,K.M., Bilbe,G., Smart,T.G, Constanti,A., Brown,D.A.,

Bamard,E.A. and Richards,B.M. (1984) Nature, 310, 318-321. Huganir,R.L., Delcour,A.H., Greengard,P. and Hess,G.P. (1986) Nature,

321, 11A-116. Kozak,M. (1984) Nucleic Acids Res., 12, 857-872. Langosch,D., Thomas,L. and Betz,H. (1988) Proc. Natl Acad. Sci USA,

85, 7394-7398. Levitan,E., Schofield,P.R., Burt,D.R., Rhee,L.M., Wisden,W., K6hler,M.,

Fujita,N., Rodriguez,H.F., Stephenson,A., Darlison,M.G., Bamard,E.A and Seeburg,P.H. (1988) Nature, 335, 76 -79 .

Meiton,D.A., Krieg,P.A., Rebagliati,M.R., Maniatis,!., Zinn,K. and Green,M.R. (1984) Nucleic Acids Res., 12, 7035-7059.

01sen,R.W. {\9%1) Anna. Rev. Pharmacol Toxicol, 22, 245-277. Parker,!., Sumikawa,K. and Miledi,R. (1988) Proc. R. Soc. B233,

201-216. Pfeiffer,F., Graham,D. and Betz,H. (1982) J. Biol Chem., 257,

9389-9393. Pfeiffer,F., Simler,R., Grenningloh,G. and Betz,H. (1984) Proc. Natl Acad.

Sci USA, 81, 7224-7227. I»ritchett,D., Sontheimer,H., Gormann,C.M., Kettenmann,H., Seeburg,P.H.

and Schofield,P.R. (1988) Science, 242, 1306-1308. Rubin,M.M. and Changeux,!.P. (1966) J. Mol Biol, 21, 265-274. Schmitt,B., Knaus,P., Becker,C.-M. and Betz,H. (1987) Biochemistry, 26,

805-811. SchofieId,P.R., Darlison,M.G., Fujita,N., Burt,D.R., Stephenson,F.A.,

Rodriguez,H., Rhee,L.M., Ramachandran,!., Reale,V., Glencorse,T.A., Seeburg,P.A. and Bamard,E.A. (1987) Nature, 328, 221-227.

700

Sontheimer,H., Becker,C.-M., Pritchett,D.D., Schofield,P.R., Grenningloh, G., Kettenmann,H., Betz,H. and Seeburg,P. (1989) Neuron, in press.

Sumikawa,K., Parker,!, and Miiedi,R. (1984) Proc. Natl Acad. Sci USA, 81, 7994-7998.

Young,A.B. and Snyder,S.H. (1973) Proc. Natl Acad. Sci USA, 70, 2832-2836.

Received on December 13, 1988; revised on January 2, 1989

Volume 244, number 2, 3 6 1 - 3 6 4 FEB 06844 February 1989

Sequence and expression of human GABA^ receptor al and subunits

Peter R. Schofield*, Dolan B. Pritchett, Harald Sontheimer^, Helmut Kettenmann^ and Peter H. Seeburg

Laboratory of Molecular Neuroendocrinology, Zentrum fur Molekulare Biologie, Im Neuenheimer Feld 282, University of Heidelberg, 6900 Heidelberg and '^Department of Neurobiology, Im Neuenheimer Feld 364, University of Heidelberg,

Heidelberg, FRG

Received 2 January 1989

The deduced amino acid sequences of cDNA clones encoding human G A B A A receptor a 1 and subunits are presented. The human subunits display very high levels of sequence identity with the corresponding bovine receptor subunits. The cloned human G A B A A receptor subunits induce the formation of GABA-gated chloride channels when expressed in

mammalian cells.

7-Aminobutyric acid A receptor; Ligand-gated ion channel; Electrophysiology; cDNA cloning; (Human brain)

1. INTRODUCTION

Receptors for the major inhibitory neurotrans-mitter GABA (y-aminobutyric acid) are present on the majority of neurons in the mammalian brain. The G A B A A receptor contains an intrinsic chloride ion channel which is opened (gated) by the neuro-transmitter, thus stabilizing the neuron's resting potential [1]. Channel activity can be allosterically modulated by therapeutically useful drugs e.g. bar-biturates and benzodiazepines [1]. The primary structure of the cloned bovine a andsubunits has shown the existence of a ligand-gated ion channel receptor superfamily [21. Coexpression of these subunits in Xenopus oocytes produces receptors with many functional characteristics of native G A B A A receptors [2]. The clinical importance of drugs that act at this ubiquitous brain receptor makes it desirable to characterize the structure of

Correspondence address: P .H. Seeburg, Laboratory of Molecular Neuroendocrinology, Zentrum fiir Molekulare Biologie, Im Neuenheimer Feld 282, University of Heidelberg, 6900 Heidelberg, FRG

* Present address: Pacific Biotechnology Ltd, 72-76 McLachlan Avenue, Rushcutters Bay 2011, Australia

the human receptor subtypes. To this end we have cloned and sequenced the human G A B A A receptor a \ and /3\ subunits and expressed them in mam-malian cells.

2. MATERIALS AND METHODS

Full-length G A B A A receptor a-l and/?l subunit cDNAs were isolated from a human fetal brain cDNA hbrary, constructed in AgtlO by standard methods. 2 x 10® pfu of this library were screened with subunit-specific radiolabelled oligonucleotides based on the bovine cDNA sequences: al subunit, 5 ' - A C C C T -G G C C A G A T T A G G T G T G T A G C T G G T T G C T G T T G G A - 3 ' ; /?1 subuni t , 5 ' - T C C C A C G C C C G T G A G C A C T T C A G A G G C -C G C T C G T C T C G T T C C T G A T C T C C G G G T A C T G A G G A G - ^ AATGTTGCCGTG-3'. Three o-l subunit cDNA clones were' analyzed and the largest (in AhGRa'28), 4.25 kb in size, was se-quenced across the coding, and into the 3'-untranslated region. Eight cloned /?1 subunit cDNAs were obtained and rescreened with specific 5 ' - and 3'-oligonucleotides based on the known bovine/5'1 cDNA sequence. One (AhGR/?19) which hybridized to both probes was sequenced. Two/^l cDNAs positive only to the 3'-probes (AhGR^lS) contained an additional T at position 1220, unlike AhGR/?19. This missing nucleotide was added to the sequence derived from AhGR/J19 by site-directed mutagenesis, using the oligonucleotide 5'-ATGTACTCCTA-T G A C G A G C G - 3 ' , thus restoring the reading frame to com-plete alignment with the bovine /?! subunit cDNA. DNA sequences were obtained using the chain-termination method [4] and M13 vectors.

Published by Elsevier Science Publishers B. V. (Biomedical Division) 00145793/89/$3.50 © 1989 Federation of European Biochemical Societies 361

Volume 244, number 1 FEES LETTERS February 1989 1 GTGAAATCTTCAGC«AAGGAGCACGCAGAGTCCATGATGGCTCAGACCAAGTGAGTGAGAGGCAGAGCGAGG»CGCCCCTCTGCTCTG

89 GCGCGCCCGGACTCGGACTCGCAGACTCGCGCTGGCTCCAGTCTCTCCACGATTCTCTCTCCCAGACTTTTCCCCGGTCTTAAGAGATCC K Y T F

- 2 7 ••* H R K S P G L S 0 C L W A W I L L L 179 TGTGTCCAGAGGGGGCCTTAGCTGCTCCAGCCCGCGATGAGGAAAAGTCCAGGTCTGTCTGACTGTCTTTGGGCCTGGAICCTCCTTCTG

362

. • c - 9 S T L T G R S Y G ' Q P S L Q D E L K D N T T V F T R I l O B 269 AGCACACTGACTGGAAGAAGCTATGGACAGCCGTCATTACAAGATGAACTTAAAGACAATACCACTGTCTTCACCAGGATTTTGGACAGA

22 L L O G Y D N R L R P G L G E R V T e V K T D I F V T S F G 359 CTCCTAGATGGCTATGACAATCGCCTGAGACCAGGATTGGGAGAGCGTGTAACCGAAGTGAAGACTGATATCTTCGTCACCAGTTTCGGA

52 P V S O H D M E Y T I D V F F H Q S W K O E R L K F K G P M 449 C C C G T T T C A G A C C A T G A T A T G G A A T A T A C A A T A G A T G T A T T T T T C C G T C A A A G C T G G A A G G A T G A A A G G T T A A A A T T T A A A G G A C C T A T G

w c =

82 T V L R L N N I H A S K I R T P O T F F H N G K K S V A H N 539 ACAGTCCTCCGGTTAAATAACCTAATGGCAAGTAAAATCCGGACTCCGGACACATTTTTCCACAATGGAAAGAAGTCAGTGGCCCACAAC

112 M T M P N K L L R I T E O G T L L Y T M R L I V R A E C P H 529 ATGACCATGCCCAACAAACTCCTGCGGATCACAGAGGATGGCACCTTGCTGTACACCATGAGGCTGACAGTGAGAGCTGAATGTCCGATG

142 H L E D F P H D A H A C P L K F G S Y A Y T R A E V V Y E W 719 C A T T T G G A G G A C T T C C C T A T G G A T G C C C A T G C T T G C C C A C T A A A A T T T G G A A G T T A T G C T T A T A C A A G A G C A G A A G T T G T T T A T G A A T G G

172 T H E P A R S V V V A E O G S R L N Q Y O L I - G O T V O S G 809 ACCAGAGAGCCAGCACGCTCAGTGGTTGTAGCAGAAGATGGATCACGTCTAAACCAGTATGACCTTCTTGGACAAACAGTAGACTCTGGA

202 I V Q S S T G E Y V V H T T H F H L K B K I G Y F V I Q T Y 899 A T T G T C C A G T C A A G T A C A G G A G A A T A T G T T G T T A T G A C C A C T C A T T T C C A C T T G A A G A G A A A G A T T G G C T A C T T T G T T A T T C A A A C A T A C

232 L P C I M T V I L S O V S F W L N R E S V P A R T V F G V T 989 C T G C C A T G C A T A A T G A C A G T G A T T C T C T C A C A A G T C T C C T T C T G G C T C A A C A G A G A G T C T G T A C C A G C A A G A A C T G I C T T T G G A G T A A C A

252 T V L T M T T L S I S A R N S L P K V A Y A T A H O W F I A 1079 A C T G T G C T C A C C A T G A C A A C A T T G A G C A T C A G T G C C A G A A A C T C C C T C C C T A A G G T G G C T T A T G C A A C A G C T A T G G A T T G G T T T A T T G C C

292 V C Y A F V F S A L I E F A T V N Y F T K R G Y A W D G K S 1169 G T G T G C T A T G C C T T T G T G T T C T C A G C T C T G A T T G A G T T T G C C A C A G T A A A C T A T T T C A C T A A G A G A G G T T A T G C A T G G G A T G G C A A A A G T

322 V V P E K P K K V K O P L I K K N N T Y A P T A T S Y T P N 1259 G T G G T T C C A G A A A A G C C A A A G A A A G T A A A G G A T C C T C T T A T T A A G A A A A A C A A C A C T T A C G C T C C A A C A G C A A C C A G C T A C A C C C C T A A T

352 L A R G O P G L A T I A K S A T I E P K E V K P E T K P P E 1349 TTGGCCAGGGGCGACCCGGGCTTAGCCACCATTGCTAAAAGTGCAACCATAGAACCTAAAGAGGTCAAGCCCGAAACAAAACCACCAGAA

382 P K K T F N S V S K I O R L S R I A F P L L F G I F N L V Y 1439 C C C A A G A A A A C C T T T A A C A G T G T C A G C A A A A T T G A C C G A C T G T C A A G A A T A G C C T T C C C G C T G C T A T T T G G A A T C T T T A A C T T A G T C T A C

412 W A T Y L N R E P Q L K A P T P H O - * * 1529 T G G G C T A C G T A T T T A A A C A G A G A G C C T C A G C T A A A A G C C C C C A C A C C A C A T C A A T A G A T C T T T T A C T C A C A T T C T G T T G T T C A G I T C C T C 1619 T G C A C T G G G A A T T T A T T T A T G T T C T C A A C G C A G T A A T T C C C A T C T G C C T T T A T T G C C T C T G T C T T A A A G A A T T T G A A A G T T T C C T T A T T T 1709 T C A T A A T T C A T T T A A G A C A A G A G A C C C C T G T C T G

- 2 5 ••• M W T V O N R E S L G L L S F P V M I 1 G A A A A G A C A A T T C T T T T A A T C A G A G T T A G T A A T G T G G A C A G T A C A A A A T C G A G A G A G T C T G G G G C T T C T C T C T T T C C C T G T G A T G A T T

1 A S - 6 T M V C C A ' H S T N E P S N M P Y V K E T V O R L L K G Y D 89 ACCATGGTCTGTTGTGCACACAGCACCAATGAACCCAGCAACATGCCATACGTGAAAGAGACAGTGGACAGATTGCTCAAAGGATATGSC

25 I R L R P D F G G P P V D V G H R I D V A S I D M V S E V N 179 ATTCGCTTGCGGCCGGACTTCGGAGGGCCCCCCGTCGACGTTGGGATGCGGATCGATGTCGCCAGCATAGACATGGTCTCCGAAGTGAAT

55 M D Y T L T H Y F Q Q S W K D K R L S Y S G I P L N L T L 0 269 A T G G A T T A T A C A C T C A C C A T G T A T T T G C A G C A G T C T T G G A A A G A C A A A A G G C T T T C T T A T I C T G G A A T C C C A C T G A A C C T C A C C C T A G A C

85 N R V A D Q L W V P D T Y F L N D K K S F V H G V T V K N B 359 A A T A G G G T A G C T G A C C A A C T C T G G G T A C C A G A C A C C T A C T T T C T G A A T G A C A A G A A A T C A T T T G T G C A T G G G G T C A C A G T G A A A A A T C G A

115 H I R L H P O G T V L Y G L R I T T T A A C N M D L R R Y P 449 A T G A T T C G A C T G C A T C C T G A T G G A A C A G T T C T C T A T G G A C T C C G A A T C A C A A C C A C A G C T G C A T G T A T G A T G G A T C T T C G A A G A T A T C C A

145 L D E Q N C T L E I E S Y G Y T T D O I E F Y W N G G E G A 539 CTGGATGAGCAGAACTGCACCCTGGAGATCGAAAGTTATGGCTATACCACTGATGACATTGAATTTTACTGGAATGGAGGAGAAGGGGCA

175 V T G V N K I E I P Q F S I V O Y K H V S K K V E F T T G A 629 G T C A C T G G T G T T A A T A A A A T C G A A C T T C C T C A A T T T T C A A T T G T T G A C T A C A A G A T G G T G T C T A A G A A G G T G G A G T T C A C A A C A G G A G C G

205 Y P R L S L S F R L K R N I G Y F I L Q T Y M P S T L I T I 719 T A T C C A C G A C T G T C A C T A A G T T T T C G T C T A A A G A G A A A C A T T G G T T A C T T C A T T T T G C A A A C C T A C A T G C C T T C T A C A C T G A T T A C A A T T

235 L S W V S F W I N Y O A S A A R V A L G I T T V L T H T T I 809 C T G T C C T G G G T G T C T T T T T G G A T C A A C T A T G A T G C A T C T G C A G C C A G A G T C G C A C T A G G A A T C A C G A C G G T G C T T A C A A T G A C A A C C A T C

265 S T H L R E T L P K I P Y V K A I O I Y L M G C F V F V F L 899 A G C A C C C A C C T C A G G G A G A C C C T G C C A A A G A T C C C T T A T G T C A A A G C G A T T G A T A T T T A T C T G A T G G G T T G C T T T G T G T T T G T G T T C C T G

G 295 A L L E Y A F V N V I F F G K G P Q K K G A S K Q O O S A N 989 G C T C T G C T G G A G T A T G C C T T T G T A A A T T A C A T C T T C T T T G G G A A A G G C C C T C A G A A A A A G G G A G C T A G C A A A C A A G A C C A G A G T G C C A A T

325 E K N K L E H N K V Q V O A H G N I L L S T L E I R N E T S 1079 GAGAAGAATAAACTGGAGATGAATAAAGTCCAGGTCGACGCCCACGGTAACATTCTCCTCAGCACCCTGGAAATCCGGAATGAGACGAGT

G G T H 355 G S E V L T S V S D P K A T H Y S Y D S A S I Q Y R K P L S

1169 GGCTCGGAAGTGCTCACGAGCGTGAGCGACCCCAAGGCCACCATGTACTCCTATGACAGCGCCAGCATCCAGTACCGCAAGCCCCTGAGC G A H

385 S R E A Y G R A L O R H G W P S K G R I R R R A S Q L K V K 1259 AGCCGCGAGGCCTACGGGCGCGCCCTGGACCGGCACGGGGTACCCAGCAAGGGGCGCATCCGCAGGCGTGCCTCCCAGCTCAAAGTCAAG

415 I P O L T O V N S I D K W S R H F F P I T F S L F N V V Y W 1349 A T C C C C G A C T T G A C T G A T G T G A A T T C C A T A G A C A A G T G G T C C C G A A T G T T T T T C C C C A T C A C C T T T T C T C T T T T T A A T G T C G T C T A I T G G

445 L Y Y V H ••• 1439 C T T T A C T A T G T A C A C T G A G G T C T G T T C T A A T G G T T C C A T T T A G A C T A C T T T C C T C T T C T A T T G T T T T T T A A C C T T A C A G G T C C C C A A C A G 1529 C G A T A C T G C T G T T T C T C G A G G T A A G A G A T T C A G C C A T C C A A T T G G T T T T A G G T C T T G C A T A T C A G T T T T A T T A C T G C A C C A T G T T T A C T T 1619 C A A A A A G A C A A A A C A A A A A A A A A A T T A T T T T T C C A G T C T A C C G T G G T C C A G G T R A T C A G C T C T T T A A G A G C T C T A T T A A T T G C C A T G T T T 1709 A C A A A C A A A C A C A A A G A G A G A A G T T A G A C A G G T A G A T C T T T A G C A G T C T T T T C T A G T T T C C C T G G A T T T C A C T G A T T T A T T T T T T A G G G A 1799 A A A T G A A A A G A G G A C C T T G C T G T C C G C C T G C A C T G C T T C C T G G T A A A C T A T A A C A A A C T T A T G C T G C C A A A A A A A A A A A

Volume 244, number 2 FEES LETTERS February 1989 Fig . l . DNA and predicted amino acid sequence of the human GABAA receptor al (a) and fil (b) subunit cDNAs. Amino acid numbering starts at the proposed mature N-terminal residues. The putative signal sequence cleavage site is indicated by an arrow and the signal sequence is shown in negative numbering. The proposed membrane-spanning hydrophobic sequences are highlighted by overscoring, the disulfide-bonded loop region is indicated by a dotted line and the putative extracellularly located N-glycosylation sites are boxed. Asterisks denote both 3 ' - terminal stop codon and the first upstream in-frame stop codon in the 5 '-untranslated sequence.

Amino acids that differ f rom the bovine sequences are indicated above the human polypeptide sequence.

200pA

A R

u r'

GABA IpMl

2.5

Vm [mVl

100 0

-100

10

2 5

I (nAl

- 1

GABA 20 s

GABA 20 s

(pAJ ICXXD

100

10

(nAl

ImVl

1 2 5 10 25 GABA (pM)

Fig.2. Dependence of the GABA-induced current on ligand concentration, (a) The membrane current at the resting potential ( - 60 mV) was recorded f rom a cell transfected with the al +/31 subunits using the patch-clamp technique in the whole cell configuration. GABA was applied as indicated by the bar at the given concentrations. A dose-dependent inward current was observed, (b) Dose-response curve of GABA-induced currents (/) for the experiment shown in a, log-log scale (Hill plot). The slope of the curve at low GABA concentrations is approx. 1, suggesting a single-binding site for the ligand. (c) The membrane current was measured while briefly clamping membrane potential (Km) from the resting value ( - 6 0 mV) to 120, - 9 0 , - 3 0 , 0, 30 and 60 mV for a period of 200 ms with 1 s intervals. This series of voltage jumps was continuously applied. GABA was added as indicated by the bar at a concentration of 10"'' M. (d) Currents (/) before application of GABA were subtracted f rom currents in the presence of GABA for the experiments

shown in c and plotted as a function of membrane potential (Km). Kn,Cr = 0 mV, K^K"" = - 8 4 mV and K^Na"" > 80 mV. 363

Volume 244, number 2 FEES LETTERS February 1989 The a l - and y6'l-subunit cDNAs were expressed and analysed

in mammalian tissue culture cells as previously described [3]. Membrane currents were recorded, using the patch-clamp technique in the whole cell configuration.

3. RESULTS AND DISCUSSION Using oligonucleotide probes, cloned .human

G A B A A receptor al and /SI subunit cDNAs were isolated from a human fetal brain cDNA library. One of three al cDNAs was sequenced and found to be full-length. Rescreening of eight /3l cDNA clones with 5 ' - and 3'-sequence-specific oligonu-cleotide probes identified a single full-length cDNA encoding this subunit. The DNA and pre-dicted polypeptide sequence of the a I and /SI cDNAs (fig.l) reveal a very high degree of conser-vation with the corresponding bovine sequences [2]. The 456 amino acid al subunit which contains a 27 residue signal peptide displays 99% sequence identity (five amino acid differences) with the bovine al subunit. However, of these, four are located within the signal sequence. The only dif-ference in this mature protein sequence occurs in the extracellular domain and is a non-conservative replacement. The human sequence contains Trp 95 while this residue is an Arg in the bovine sequence. The 474 amino acid (31 subunit displays 98% se-quence identity (11 amino acid differences). In this case, only one of the differences is located within the 25 residue signal peptide. Of the remainder, two non-conservative replacements are found at the extreme N-terminus and the other 8 are located within the intracellular loop located between transmembrane domains 3 and 4. Of these, only two are conservative in nature. The concentration of sequence differences within the intracellular loop is consistent with this region of all members of this receptor superfamily being the most divergent. The results obtained suggest that the G A B A A receptor subunits are very highly con-served among the mammalian species.

The cloned human al and j3l subunit cDNAs have recently been expressed in mammalian cells [3]. Using this technique, we transfected human cells as described and obtained electrophysio-logical recordings 48 h after transfection. G A B A A receptors expressed in mammalian cells have many of the properties of the native receptors [3]. They conduct chloride ions, are specifically inhibited by picrotoxin and the competitive G A B A A receptor

antagonist bicuculline and are potentiated by bar-biturates. They do not display benzodiazepine responsiveness [3,5]. Single cell dose-response curves of GABA-induced currents (fig.2) had an average slope of 1 (SD = 0.1, TV = 5), indicating that binding of only one G A B A molecule per receptor caused a channel opening. This is at variance with the observed cooperativity of neurotransmitter gating of G A B A A receptor chan-nels [6-8], but in agreement with results obtained by coexpression of a- and /J-subunits in Xenopus oocytes [2,5].

Our results suggest a more complex composition of neuronal G A B A A receptors. Other hitherto unknown subunits may coassemble with al and/3l subunits to generate a full display of all known receptor properties. The recent isolation in our laboratory of cDNAs which encode proteins with partial sequence identity to the a- and /?-subunits should facilitate the experimental evaluation of the design of this important neurotransmitter receptor.

Acknowledgements: The authors thank Jutta Rami for secretarial assistance. This work was supported in part by the Deutsche Forschungsgemeinschaft, SFB 317 grant B/9 and by the Bundesministerium fiir Forschung und Technologic grant BCT 0381/5 to P.H.S.

REFERENCES [1] Olsen, R.W. and Venter, J.C. (1986) Benzodiaze-

pine/GABA Receptors and Chloride Channels: Structural and Functional Properties, A.R. Liss, New York.

[2] Schofield, P.R., Darlison, M.G., Fujita, N., Burt, D.R., Stephenson, F.A., Rodriguez, H., Rhee, L.M., Ramachandran, J., Reale, V., Glencorse, T.A., Seeburg, P.H. and Barnard, E.A. (1987) Nature 328, 221-227.

[3] Pritchett, D.B., Sontheimer, H., Gorman, C.M., Kettenmann, H., Seeburg, P.H. and Schofield, P.R. (1988) Science 242, 1306-1308.

[4] Sanger, F., Nicklen, S. and Coulson, A.R. (1977) Proc. Natl. Acad. Sci. USA 74, 5463-5467.

[5] Levitan, E.S., Schofield, P.R., Burt, D.R., Rhee, L.M., Wisden, W., Koehler, M., Fujita, N., Rodriguez, H., Stephenson, F.A., Darlison, M.G., Barnard, E.A. and Seeburg, P.H. (1988) Nature 335, 76-79.

[6] Barker, J.L. and Ransom, B.R. (1978) J. Physiol. (Lond.) 280, 331-354.

[7] Sakmann, B., Hamill, O.P. and Bormann, J. (1983) J. Neural Transm. suppl.18, 83-95.

[8] Borman, J. and Clapham, D.E. (1985) Proc. Natl. Acad. Sci. USA 82, 2168-2172.

364

Reprinted from Nature, Vol. 338, No. 6216, pp. 582-585, 13 April 1989 © Macmillan Magazines Ltd., 1989

Importance of a novel GABA^ receptor subunit for benzodiazepine pharmacology Dolan B. Pritchett*, Harald Sontheimert, Brenda D. Shivers*, Sanie Ymer% Helmut Kettenmannt, Peter R. Schofield*$ & Peter H. Seeburg*§

* ZMBH, Universitat Heidelberg, Im Neuenheimer Feld 282, 6900 Heidelberg, FRG t Institut fur Neurobiologie, Universitat Heidelberg, Im Neuenheimer Feld 364, 6900 Heidelberg, FRG t Present address: Pacific Biotechnology Ltd, 72-76 McLachlan Avenue, Rushcutters Bay 2011, Australia § To whom correspondence should be addressed

NEUROTRANSMISSION eifected by GABA (y-aminobutyric acid) is predominantly mediated by a gated chloride channel intrinsic to the GABA^ receptor. This heterooligomeric receptor^ exists in most inhibitory synapses in the vertebrate central nervous system (CNS) and can be regulated by clinically important com-pounds such as benzodiazepines and barbiturates^. The primary structures of GABA^ receptor a - and /3-subunits have been deduced from cloned complementary DNAs '"*. Co-expression of these subunits in heterologous systems generates receptors which display much of the pharmacology of their neural counterparts, including potentiation by barbiturates'"®. Conspicuously, however, they lack binding sites for, and consistent electrophysiological responses to, benzodiazepines'*'®. We now report the isolation of a cloned cDNA encoding a new GABA^ receptor subunit, termed y2, which shares approximately 40% sequence identity with a -and /3-subunits and whose messenger RNA is prominently localized in neuronal subpopulations throughout the CNS. Importantly, coexpression of the y2 subunit with a \ and /31 subunits produces G A B A A receptors displaying high-aiiinity binding for central ben-zodiazepine receptor ligands.

Cloned cDNA encoding a new G A B A A receptor subunit (Fig. la ) , designated yl, was isolated from a human fetal brain cDNA library by screening with a probe based on the conserved octameric amino-acid sequence in the second transmembrane region of G A B A A and glycine receptor subunits''" '*^. The predic-ted primary structure of 468 amino-acid residues shares many features with a- and /3-subunits (Fig. lb). These include a signal peptide, a disulphide-bonded /3-structural loop and three poten-tial N-linked glycosylation sites in the putative extracellular

FIG. 1 The human GABA^ receptor y2 subunit. a. Nucleotide and predicted • amino-acid sequence of cloned y ^ cDNA. The amino-acid sequence of the. encoded polypeptide is shown in the single letter code below the nucleotide sequence. Numbers on the left denote nucleotides and amino-acid residues. Negative numbers specify residues in the proposed signal sequence^'*. Putative /V-linked glycosylation sites are circled, the two cysteines involved in the formation of the invariant loop structure are starred, the four trans-membrane segments are underlined and the putative tyrosine phosphoryl-ation site is boxed, b. Comparison of the amino-acid sequences of human G A B A A receptor a l , and y2 subunits. The amino acids of the mature polypeptides are numbered, identical residues are boxed. Transmembrane segments are overlined. Gaps are introduced for optimal sequence alignment. METHODS. A Ag t lO human fetal brain cDNA library^^ was screened for G A B A A receptor subunits using as a probe a 96-fold degenerate synthetic 23-mer oligonucleotide encoding a conserved octameric amino-acid sequence in the second transmembrane segment of G A B A A and glycine-receptor subunits^ '* ®. A duplicate of the library was screened using a mixture of oligonucleotides which contain unique sequences found in previously cloned GABAA-receptor subunit-encoding cDNAs (refs 3, 4 and unpublished observations). Cloned cDNAs hybridizing with only the 23-mer probe were sequenced^®' ®. The y2 cDNA represents one of several new sequences identified in this manner.

a 1 C C T G A C G C T T T G A T G G T A T C T G C A A G C G T T T T T G C T G A T C T T A T C T C T G C C C C C T G A A T A

61 T T A A T T C C C T A A T C T G G T A G C A A T C C A T C T C C C C A G T G A A G G A C C T A C T A G A G G C A G G T G 121 G G G G G A G C C A C C A T C A G A T C A T C A A G C A T A A G A A T A A T A C A A A G G G G A G G G A T T C T T C T G 181 C A A C C A A G A G G C A A G A G G C G A G A G A A G G A A A A A A A A A A A A A A A G C G A T G A G T T C A C C A A A - 3 9 M S S P N 241 T A T A T G G A G C A C A G G A A G C T C A G T C T A C T C G A C T C C T G T A T T T T C A C A G A A A A T G A C G G T . 3 4 I W S T G S S V Y S T P V F S Q K M T V 301 G T G G A T T C T G C T C C T G C T G T C G C T C T A C C C T G G C T T C A C T A G C C A G A A A T C T G A T G A T G A - 1 4 W I L L L L S L Y P G F T S ' Q K S D D D 361 C T A T G A A G A T T A T G C T T C T A A C A A A A C A T G G G T C T T G A C T C C A A A A G T T C C T G A G G G T G A

7 Y E D V A S ® K T W V L T P K V P E G D 421 T G T C A C T G T C A T C T T A A A C A A C C T G C T G G A A G G A T A T G A C A A T A A A C T T C G G C C T G A T A T

27 V T V I L N N L L E G Y D N K L R P D I 481 A G G A G T G A A G C C A A C G T T A A T T C A C A C A G A C A T G T A T G T G A A T A G C A T T G G T C C A B T G A A

47 G V K P T L I H T D M Y V N S I G P V N 541 C G C T A T C A A T A T G G A A T A C A C T A T T G A T A T A T T T T T T G C G C A A A T G T G G T A T G A C A G A C G

67 A I N N E Y T I D I F F A Q M W Y D R R 601 T T T G A A A T T T A A C A G C A C C A T T A A A G T C C T C C G A T T G A A C A G C A A C A T G G T G G G G A A A A T

87 L K F ® S T I K V L R L N S N M V G K I 661 C T G G A T T C C A G A C A C T T T C T T C A G A A A T T C C A A A A A A G C T G A T G C A C A C T G G A T C A C C A C 107 W I P D T F F R N S K K A D A H - W I T T 721 C C C C A A C A G G A T G C T G A G A A T T T G G A A T G A T G G T C G A G T G C T C T A C T C C C T A A G G T T G A C 127 P N R H L R I W N D G R V L Y S L R L T 781 A A T T G A T G C T G A G T G C C A A T T A C A A T T G C A C A A T T T T C C A A T G G A T G A A C A C T C C T G C C C 147 I D A E C * Q L Q L H N F P M D E H S C * P 841 C T T G G A G T T C T C C A G T T A T G G C T A T C C A C G T G A A G A A A T T G T T T A T C A A T G G A A G C G A A G

167 L E F S S Y G Y P R E E I V Y Q W K R S 901 T T C T B T T G A A G T G G G C G A C A C A A G A T C C T G G A G G C T T T A T C A A T T C T C A T T T G T T G G T C T 187 S V E V G O T R S W R L Y Q F S F V G L 961 A A G A A A T A C C A C C G A A G T A G T G A A G A C A A C T T C C G G A G A T T A T G T G G T C A T G T C T G T C T A 207 R ® T T E V V K T T S G D Y W V M S V Y

1021 C T T T G A T C T G A G C A G A A G A A T G G G A T A C T T T A C C A T C C A G A C C T A T A T C C C C T G C A C A C T 227 F D L S R R M G Y F T I Q T Y I P C T L

1081 C A T T G T C G T C C T A T C C T G G G T G T C T T T C T G G A T C A A T A A G G A T G C T G T T C C A G C C A G A A C 247 I V V L S W V S F W I N K D A V P A R T

1141 A T C T T T A G G T A T C A C C A C T G T C C T G A C A A T G A C C A C C C T C A G C A C C A T T G C C C G G A A A T C 267 S L G I T T V L T H T T L S T I A R K S

1201 G C T C C C C A A G G T C T C C T A T G T C A C A G C G A T G G A T C T C T T T G T A T C T G T T T G T T T C A T C T T 287 L P K V S Y V T A H D L F V S V C F I F

1261 T G T C T T C T C T G C T C T G G T G G A G T A T G G C A C C T T G C A T T A T T T T G T C A G C A A C C G G A A A C C 307 V F S A L V E Y G T L H Y F V S N R K P

1321 A A G C A A G G A C A A A G A T A A A A A G A A G A A A A A C C C T G C C C C T A C C A T T G A T A T C C G C C C A A G 327 S I C O K D K K K K N P A P T I O I R P R

1381 A T C A G C A A C C A T T C A A A T G A A T A A T G C T A C A C A C C T T C A A G A G A G A G A T G A A G A G T A C G G 347 S A T I O M N N A T H L Q E R D E E I U G

1441 C T A T G A G T G T C T G G A C G G C A A G G A C T G T G C C A G T T T T T T C T G C T G T T T T . G A A G A T T G T C G 367 Y E C L D G K D C A S F F C C F - E D C R

1501 A A C A G G A G C T T G G A G A C A T G G G A G G A T A C A T A T C C G C A T T G C C A A A A T G G A C T C C T A T G C 387 T G A W H H G R I H I R I A K M D S Y A

1561 T C G G A T C T T C T T C C C C A C T G C C T T C T G C C T G T T T A A T C T G G T C T A T T G G G T C T C C T A C C T 407 R I F F P T A F C L F N L V Y W V S Y L

1621 C T A C C T G T G A G G A G G T A T G G G T T T T A C T G A T A T G G T T C T T A T T C A C T G A G T C T C A T G G A G 427 Y L

1681 A G A T G T C T G T T C T A A G T C C A C T T A A A T A A T C C T C T A T G T G G T T G A T A A G T A T C T G A A T C T 1741 G T T T C

S P N l W S T G S S V Y S T P V F S Q K M T V W l L L L t K S P G L S D C L W A W l L L L S

f - T V Q N f t E S L C L L S F P V

Q K S D D D Y E D ' L Q D - - -T N E - - -

L R P D I L R P C I I R P b F

V K P T L I H T D M Y E R V T E V K T D I F C P P V D V G M R I D

T V I I N N L L E G Y D T R I L D R L L D C Y D K E T V D R L L K G Y D

P. G P G P

S I D M

I 5 T I K V ; C P M T V

G I P L N

L N S N M V G K I L N N I M A S K l R T l L D N R V A D Q L D | K K

W I T T P N R M L R l W N D G R V L Y 5 1 . R L N M T M P N K 1. I R 1 T E D G T L L Y T M R L G V [ T V K N R M 1 ^ L H P D G T V L Y G L ^ I

H N F E D F R R Y

P H

P R E E I V Y Q T R A E V V Y E T T D D I E F Y

- - V C D T R S W R L S V V V A E D G S R 1. - - W T G V N K . E [ £

F V S N » K P 5 F T K R G Y A ® I F F G R G P Q

D K D K K K I t N P I G K S V V P E K P l . K C A S K Q D Q S

T I D I R P R S V K D P L I K It

l E K N K l E M N

I H N N A 1 A P T A 1 D A H G 1

T S V S D P It : I Y S Y D S A

. - - - Q E R , - . - - T P N

I Q Y R K P L

D E E Y C Y E C I O G K D C L A R G D P G L A T I A K S S S R E A Y G R A L D R H G

I S F F C C F E D C R T C A W R H G R I H I R I A l t I r i E P K E V l t P E T l t P P E P I l K T F N S V S I t r P S K G R l R R R A S Q L R V K I P D I T D V N S

Y W V S Y W A T Y w L Y

LETTERS TO NATURE

( ^ H ) - R o 15-1788 (nM)

FIG. 2 Saturation isotherm of pH]Rol5-1788 binding. Inset, the correspond-ing Scatchard plot (/<(j=0.97 nM, 6^3^=3.2 pmol protein). In three independent transfection experiments K^ values of l ± 0 . 2 n M and values of 2.7 ±0.7 pmol mg~^ protein were obtained. METHODS. Human a l - , )81- (ref. 11) and •y2-encoding cDNAs, subcloned individually into the expression vector pCIS2 (ref. 27), were used to transform human embryonic kidney 293 cells as described previously®' ®. After trans-fection (48 h), cells from ten plates (10 cm; 4x10® cells per plate) were washed twice with phosphate-buffered saline (PBS) and scraped into PBS (10 ml). The cell pellet (500g) was homogenized in a Polytron tissue homogenizer (Brinkmann) in 10 ml of 10 mM potassium phosphate, pH 7.4, and centrifuged (SO.OOOg 20 min). This procedure was repeated three times and the final pellet resuspended in potassium phosphate buffer, pH7.4, containing 100 mM KCI. For each ligand concentration, triplicate homoge-nates, each equivalent to 10® cells (100 jxg protein), were incubated (4 °C, 60 min) in a 1 ml reaction volume with pH]Rol5-1788 (NEN, 75 Ci mmoP^). Non-specific binding was determined in the presence of 10~® M clonazepam and was <10% of total binding. K, values of clonazepam, diazepam, and DMCM (see text) were determined in similar reactions using 10~^°-10~® M unlabelled competitor compounds. Samples were vacuum-filtered on GF/C filters with a vacuum filter apparatus (Biorad) and filters were washed twice with 5 ml homogenization buffer. After drying, filter-retained radioactivity was determined by liquid scintillation counting.

domain, as well as four transmembrane segments. The greatest sequence similarity to a - and /3-subunits occurs in the region of these transmembrane segments and includes the charged amino-acid residues thought to form the channel mouth^'^ The proposed large intracellular domain displays the highest sequence divergence among the subunits. In the y2 subunit, this domain contains a consensus sequence for tyrosine-specific pro-tein phosphorylation, suggesting a target for the cellular control of channel activity by receptor-mediated tyrosine kinases^. A putative control site for serine phosphorylation by protein kinase A is found vi'ithin the same domain of the subunit^. The overall sequence similarity to a l and subunits is 42 and 35%, respectively. This level of structural similarity is seen for the different nicotinic acetylcholine receptor subunits® and is significantly below the 70-80% identity which typically charac-terizes the variants of GABA^ receptor a-subunits^'"^ or (3-subunits (manuscript in preparation). Based on these structural considerations and the functional signature detailed below, the

y2 polypeptide is a novel GABA/benzodiazepine receptor subunit.

To determine the functional properties of the y2 subunit, the cloned c D N A was transiently expressed in cultured human cells^, either singly or together with cloned c D N A encoding human a 1 and /31 subunits". The membranes of the transfected cells were analysed for the binding of radiolabelled ligands specific for central benzodiazepine receptors^'*". High-affinity binding sites for [^H]Rol5-1788 (/iCd = 1.2±0.2 nM, = 2 ,900±300 fmolmg"^ Fig. 2) and [^H]flunitrazepam ( X j -2 n M , 2,000 fmolmg"'; data not shown) were observed on cells expressing all three G A B A a receptor subunits. The binding of these labelled compounds was completely blocked in the presence of 1 |xM diazepam (K; = 1 0 ± 2 nM), clonazepam (Ki = l . l ± 0 . 7 n M ) or methyl-4-ethyl-6,7-dimethoxy-/3-carbo-Hne-3-carboxylate (DMCM; = 2.1 ± 0.8 nM; data not shown). The Ki values of these compounds are similar to their rank order and potency determined on cerebellar membranes'^. No

FIG. 3. Ionic and pharmacological characteristics of GABA-induced membrane currents in 293 cells expressing human GABA^ receptor a l , /81 and y2 subunits. a. Current-voltage relation of GABA-induced responses (GABA, lOOfJiM). A reversal potential close to 0 mV was seen in 126 mM [CI~]o. When [Cr]o was lowered to 40 mM the reversal potential shifted to 30 mV, indicating a highly chloride-selective conductance, b, Flunitrazepam (FNZM, 1 |xM) reversibly potentiated the GABA-induced current (GABA, 5 |i,M). This potentiation was completely inhibited by Rol5-1788 (1 |jlM). DMCM (1 |xM) reduced the GABA-induced current by ~50%. This reduction was lost in the presence of Rol5-1788 (1 |xM). c. Diazepam (DZP, 1 (xM), FNZM (1 (xM) and pentobarbital (PB, 50 |xM) all reversibly potentiated the current induced by 10 (jiM GABA (// indicates a 5-min interval between subsequent applications). METHODS. Membrane currents were recorded by the patch-clamp technique in the whole cell configuration^^ using an EPC-7 patch-clamp amplifier (List Electronics, Darmstadt). Cells were clamped at - 6 0 mV. Data were filtered at 3 KHz ^ ^ (8-pole Bessel) and sampled up to 10 KHz. To determine the reversal potential of the trans-mitter-induced responses, membrane currents were measured while apply-ing a voltage ramp ( -120 to 120 mV, 1.6 s duration). The current-voltage curves were calculated by subtracting currents (I) before appl ication of GABA from currents obtained in the presence of GABA, and were plotted as a function of membrane potential ( l/J. Substances were added, as indicated by bars, at the following concentrations: FNZM, DZP, Rol5-1788, and DMCM,

NATURE • VOL 338 • 13 APRIL 1989

126 c r u u

u GABA

• FNZM •RO 15-1788

HH V.

Hh r^'r

V

H H

u

u ' l y GABA

• DMCM . R015-1788

HH HH \ j

1 |xM; PB, 50 (xM. Pipettes contained in mM: CsCI, 130; MgCIa, 1, CaCIa, 0.5; EGTA, 5; HEPES, 10; pH, 7.2. Cultures were continuously perfused with a solution containing in mM: KCI, 5.4; NaCI, 116; MgClj, 0.8; CaCIa, 1.8; D-glucose, 11.1; NaHCOg, 26.2; HEPES, 5; pH, 7.2. All measurements were performed at room temperature.

583

LETTERS TO NATURE

binding was detected when the yl subunit alone or pairs of subunits were expressed. These data show that a-, fS- and y-subunits can assemble into GABA^ receptors which contain a high-affinity benzodiazepine binding site.

The receptors formed by co-expression of three different sub-units were characterized by electrophysiology, using the whole cell patch-clamp technique" (Fig. 3). The application of G A B A induced large inward currents; G A B A sensitivity showed a Hill coefficient of ~1 , as observed upon co-expression of a- and yS-subunits' . The selectivity for chloride ions is demonstrated by a shift in the current versus vo4tage curve when changing the concentration of extracellular chloride. The GABA-induced cur-rent was blocked by the G A B A A receptor antagonist bicuculline (10"^ M) and the channel blocker picrotoxin (10"'* M) (data not shown) but was potentiated by the barbiturate pentobarbital. Importantly, these receptors display the full functional proper-ties of GABA/benzodiazepine receptors. In all expressing cells tested, responses to G A B A were consistently enhanced about

twofold by the benzodiazepine receptor agonists flunitrazepam and diazepam and reduced by 50% in the presence of the inverse agonist DMCM ' "*. These effects were completely blocked by the benzodiazepine receptor antagonist Ro 15-1788. This spec-trum of sensitivity was not observed upon co-expressing a\y2 and filyl subunit pairs. Furthermore, no responsiveness was seen on expression of the y2 subunit alone which generated homomeric receptors (data not shown) comparable in their properties to the receptors formed from single a- or /3-subunits^. Our results indicate that all three subunits contribute to the formation of GABA/benzodiazepine receptors.

The yl subunit is comparable in abundance to a- and subunits, as judged from the number of cognate cDNA clones in libraries of rat, bovine and human CNS origin and as eviden-ced by northern analysis of rat and bovine brain mRNA (unpub-lished observations). To determine the spatial pattern of y2 mRNA expression in the CNS, we performed in situ hybridiz-ation on sagittal sections of rat brain, using as a probe a radioac-

FIG. 4 Location of neurons synthesizing y2 subunit mRNA in rat brain. In situ hybridization was performed using a ^®S-labelled cRNA on frozen, paraformaldehyde-fixed brain sections. Grain clusters in the emulsion mark the location of cells containing hybridizing mRNA in stained sections. Sec-tions probed with ^®S-labelled sense RNA produced no hybridization pattern (data not shown), a, Darkfield photomicrograph of cerebellar cortex (cb, left) and neocortex (cx, right). In cerebellar cortex, y2 subunit mRNA expression is highest in the Purkinje neurons in the Purkinje cell layer (p) with an additional population of hybridizing cells (arrowheads) in the molecular layer (m) representing presumptive basket/stellate neurons. In the neocortex, numerous cells throughout layers ll-Vl express the y2 subunit mRNA highly and display no obvious lamination pattern; neocortical layers l-lll are shown. b. Bright-field photomicrograph of y2 subunit mRNA expression in all Purkinje neurons (arrowheads), c, Dark-field photomicrograph of cerebellar cortex (left) and of the inferior colliculus (IC, right). Numerous cells expressing the y2-subunit mRNA are distributed throughout the inferior colliculus, unlike the superior colliculus (not shown), d. Dark-field photomicrograph of part of the hippocampal formation showing high y2 subunit mRNA expression in granule neurons of the dentate gyrus (dg) and pyramidal neurons of the hippocampal CAS and CA4 regions. C A l and CA2 pyramidal neurons (not shown) similarly express 72-subunit mRNA. Scale bar in a (for a, c and d),

584

100 pim; in b, 30 jjim. Other abbreviations: g, granule cell layer; iv, white matter. METHODS. RNA probes were transcribed from a linearized Bluescript plasmid DNA, containing a 1,500-base pair EcoRl fragment from the coding region of rat y2 cDNA. Following plasmid linearization at unique restriction sites in the polylinker region, sense or antisense RNA was synthesized using phage T3 or T7 RNA polymerase. RNA products were radiolabeled^® to a specific activity of ~3x l0®c.p .m. |jLg" using 1.0 jj-g linearized plasmid DNA, 50 |xCi aP®S]CTP (1,000 Ci mmoP^, Amersham), 2.5 mM each of ATP, OTP and UTP, and were hydrolysed (pH 10, 37 °C, 1 h) to an average length of 100 nucleotides. For each section the RNA probe (2.5xl0®c.p.m.) was dissolved in 50 |xl hybridization solution which included 0.6 M NaCI, 50% formamide, and 40 mM ;3-mercaptoethanol. In situ hybridization was perfor-med as described previously^®. Briefly, the sections were hybridized at 42 °C for 3 days, washed twice for 20 min each at 65 °C in O . l x S S C , 0.05% inorganic pyrophosphate, 14 mM ^-mercaptoethanol, and dehydrated in alcohol. Following overnight exposure to Kodak XAR 5 film, the sections were dipped in Kodak NTB2 emulsion (diluted 1:1 in water), exposed for 8 days, and stained in 1% Fast Green and 0.5% cresyl violet. Sagittal brain structures were identified from ref. 31.

NATURE • VOL 338 • 13 APRIL 1989

lively labelled cRNA derived from cloned rat y l cDNA (manu-script in preparation). Our results show that y l mRNA is promi-nently expressed in neuronal subsets throughout the CNS which include neurons in the olfactory bulb, anterior olfactory nuclei, preoptic area, neocortex, globus pallidus, hippocampus, dentate gyrus, thalamus, inferior colliculus, substantia nigra, pontine nuclei, cerebellar cortex and cerebellar nuclei. Four of these regions have been chosen to illustrate the cellular location of yl mRNA (Fig. 4). Significantly, all of these regions contain high-affinity binding sites for benzodiazepines'^-'^ and also express a- and )3-subunit mRNAs (refs 17, 18; B.D.S., unpub-lished observations), further supporting the hypothesis that the yl subunit is an integral part of GABA/benzodiazepine receptors.

The existence of the yl subunit was not anticipated from biochemical studies as subunits of aflBnity-purified GABA/ben-zodiazepine receptor are electrophoretically resolved into only two main bands corresponding to relative molecular mass (M^) 48,000-53,000 (48K-53K) (a) and M,55-57K (/3) (ref. 1). These bands, however, are heterogeneous, consisting of variants of the a-subunits^''^ and /3-subunits, and of additional GABAA recep-tor subunits, including yl and a related y\ subunit, whose cloning preceded that of yl and which shows glial localization (manuscript in preparation). The mature -yl-polypeptide (unglycosylated, Mr«=48K) may co-migrate with a-subunits (Mr, 48-53K)'''*, which have been postulated to carry the ben-zodiazepine site on the basis of photoaflfinity labelling with flunitrazepam^'"^^. Our results on benzodiazepine responsive-ness indicate that the y- subunit contributes to the formation of the benzodiazepine site and thus may also be photoaffinity-labelled. In fact, it is probable that a flunitrazepam-labelled yl subunit would be indistinguishable from certain labelled a-subunits by present methods.

We note that other subunit combinations may also create benzodiazepine responsiveness. Indeed, recent cDNA cloning

experiments in our laboratory provide evidence for the existence of additional GABAA receptor subunits (manuscript in prepar-ation). It seems likely therefore that the true diversity of GABA/benzodiazepine receptor subtypes has been only partly revealed by classical pharmacology'^-'^'^". •

Received 31 January; accepted 3 March 1989.

1. Stephenson, F. A. Biochem. J. 249, 21-32 (1988). 2. Benzodiazepine/GABA Receptors and Chloride Channels: Structural and Functional Properties

(eds Olsen, R. W. & Venter, C. J.) (Liss, New York, 1986). 3. Schofield P. R. et al. Nature 328, 221-227 (1987). 4. Levitan, E. S. et al. Nature 335, 76-79 (1988). 5. Pritchett et al. Science 242,1306-1308 (1988). 6. Grenningloh, G. et al. Nature 328, 215-220 (1987). 7. Imoto, K. et al. Nature 335, 645-648 (1988). 8. Hopfield, J. F., Tank. D. W., Greengard, P. & Huganir, R. L. Nature 336, 677-680 (1988). 9. Noda, M. et al. Nature 302, 528-532 (1983).

10. Hunkeler, W. et al. Nature 290, 514-516 (1981). 11. Schofield, P. R. et al. FEBS Lett (in the press). 12. Sieghart, W. & Schuster, A. Biochem. Pharmac. 33, 4033-4038 (1984). 13. Hamill, 0. P., Marty, A., Neher, E.. Sakmann, B. & Sigworth, F. I. Pfiiigers Arch. Ges. Physiol. 391,

85-100 (1981). 14. Skeritt, J. H. & MacDonald, R. L. Eur. J. Pharmac. 101,135-141 (1984). 15. Young, W. S., Ill & Kuhar, M. J. Pharmaa Ther. 212, 337-346 (1980). 16. Richards, J. G. & Mohler, H. Neuropharmacology 23, 233-242 (1984). 17. Siegei, R. E. Neuron 1, 579-584 (1988). 18. S§quier, J. M. et al Proc. natn. Acad. Sci. U.S.A. 85, 7815-7819 (1988). 19. Braestrup, C. & Nielsen, M. J. J. Neurochem. 37, 333-341 (1981). 20. Cooper, S. J., Karkham, T. C. & Estall, L. B. Trends pharmac. Sci. 8 ,180-184 (1987). 21. Mohler. H. & Okada, T. Science 198, 849-851 (1977). 22. Casalotti, S. 0., Stephenson, F. A. & Barnard, E. A. J. blol. Chem. 261,15013-15016 (1986). 23. Fuchs, K., Mohler. H. & Sieghart. W. Neurosc. Lett 90 314-319 (1988). 24. von Heijne. G. Nucleic Acids Res. 14, 4683-4690 (1986). 25. Sanger, F.. Nicklen, 8. & Coulson, A. R., Proc. natn. Acad ScL U.S.A. 74, 5463-5467 (1977). 26. Vieira, J. & Messing, J. Meth. Enzym. 153, 3-11 (1987). 27. Eaton, D. L. et al. Biochemistry 25, 8343-8347 (1986). 28. Chen, C. & Okayama, H. Mol. cell. BioL 7, 2745-2751 (1987). 29. Melton. D. A. et a/. Nucleic Acids Res. 12, 7035-7056 (1984). 30. Shivers, B. D., Schachter, B. S. & Pfaff, D. W. Meth. Enzym. 124, 497-510 (1986). 31. Paxinos, 6. & Watson, 0. The Rat Brain in Stereotaxic Coordinates (Academic. Sidney, 1982).

ACKNOWLEDGEMENTS. We acknov^ledge Pia Werner who first isolated a y2 cDNA sequence from a bovine adrenal medulla library, Martin Kohler who helped in the cloning experiments and Dr Rolf Sprengel for the construction of the SP6 vector containing rat y2 cDNA. We thank Drs Bert Sakmann, Heinrich Betz, Andreas Draguhn and Hartmut Liiddens for helpful suggestions and Jutta Rami for secretarial help. This work was funded by the DFG and BMFT (P.H.S.).

Printed in Great Britain by Turnergraphic Limited, Basingstoke, Hampshire

^^Boifenc Modulation of Ammo Acid Receptors. TheTopeuiic Implications edited by E. A . Barnard and E. Costa Raven Press, Ltd . . N e w York © 1989

Molecular Biology of the G A B A A Receptor

A. Barnard, 'D. R. Burt, 'M. G. Darlison, 'N. Fujica, •E. S. Levitan, R. Schofield, " P . H. Seeburg,

'M. D. Squire and T . A. Stephenson 'MRC Molecular Neurobiobp Unit, Cambridge CB2 2QH. United Kingdom; and

"Laboratory of Molecular Neuroendocrinobgy, ZMBH, 0-6900 Heidelberg, Federal Republic of Germany

BINDING SITES OF THE GABAA RECEPTOR

R J he GABAA receptor is the major molecular site of the ubiquitous inhib-I itory activities of the brain, being present on the great majority of mam-

malian brain neurons (19). Electrophysiological studies and, especially, recent patch-clamp studies on cultured neurons (4,9) and adrenal chromaffin cells (8) have established that at these sites G A B A opens a chloride channel that is integrally associated with its receptor. Further, the GABAA receptor at brain syn-apses is known to be a site of action for several pharmacologically important classes of drugs. Pharmacological and ligand-binding studies (reviewed in ref. 36) have identified at least five types of binding site on this receptor: (a) the G A B A agonist/ antagonist site; (b) the benzodiazepine site, which itself is complex, having in-teractions with both anxiolytic agonists and anxiogenic "inverse agonists" (12); (c) the picrotoxin site, where agents such as picrotoxin (12) or t-butylbicyclo-phosphorothionate (51) block the GABA-activated channel; (d) the depressant site, recognizing the CNS-depressant barbiturates and certain other depressant drugs that at lower concentrations prolong the lifetime of the GABA-activated channel (4) and at higher concentrations can themselves open the channel (32) [this site also appears to be multiple, as certain steroids (32), the anesthetic pro-panidid (30) and avermectin B ^ (17,45) act similarly to depressants in some ways); and sites binding the channel-permeating anions (but not other ions) (9,36). Each of these types of ligand site can interact allosterically with one or more of the other types (36). From this network of interactions, it can be deduced that several of chese sites can be occupied by their respective ligands simultaneously and that each of the five or more types of site must be physically separate on the receptor structure.

Variations in some of these sites have been recognized in the form of GABAA receptor subtypes. Thus, two GABA/benzodiazepine receptor subtypes (type 1 and type II) have been defined by the sensitivity of the binding (11,48) or of the m ^ivo pharmacology (15,34) of benzodiazepines to certain P-carbolines, triazolo-

19

20 £. A. Barnard ei al.

pyridazines, or pyrazoloquinolines. Recently, a third such subtype was detected in brain with yet different binding properties at its benzodiazepine sites (see Guidotti et al., this volume). Multiple GABAA receptors, defined functionally, have also been deduced in electrophysiological and chloride flux studies (2,14).

T H E P U R I F I E D G A B A A R E C E P T O R P R O T E I N

The unitary protein structure postulated has, indeed, been identified. Purifi-cation of the receptor in an appropriate detergent was accomplished on a benzo-diazepine affinity column (42), and the single protein finally obtained binds ligands for all the sites listed above (43,44). The rank order of drug potencies in these classes is preserved from the in vivo state to the membrane to the purified protein state (44,49). The purified protein also shows the characteristic allosteric inter-actions between sites, although these are quantitatively lowered in detergent so-lution compared to the membrane state (43).

In the purified receptor from bovine (43,49), chick (35), or rat (38) brain, only two subunit types are detectable in standard polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate (SDS-PAGE), a (MW 53,000) and p (MW 57,000). In the receptor in the membrane the a subunit can be photoaffinity-labeled by flunitrazepam, and this reaction proceeds on the a subunit in the pure receptor protein (13). A photoaffinity-labeling by [^H]muscimol can also occur, albeit with low efficiency, and in the pure receptor it is the P subunit that is the site of such labeling by the GABAA agonist muscimol (13,16).

However, in each case the situation is too complex to allow a simple assignment of the receptor benzodiazepine site to an a subunit and of its GABA site to a P subunit. In brain membranes or in cultured neurons, the [^H]flunitrazepam pho-toaffinity reaction results in the labeling of polypeptides of several apparent Mr values, in the range 48,000 to 58,000 daltons: The size seen depends on the brain region, the age, and the species (18,23,40,41). This has been considered to denote that a number of different forms of the a subunit occur. For the p subunit, the affinity of [^H]GABA involved in the labeling is much higher than the affinity (which is in the micromolar or higher range) involved in the channel opening or the allosteric potentiation of benzodiazepine binding (4,36). Therefore, a high-affinity agonist binding site is located on the p subunit, but low-affinity agonist binding sites may also be present on the oligomeric receptor. The latter are difficult to measure by the equilibrium binding of a radiolabeled agonist. Further, the GABAA receptor shows very strong desensitization, which may be associated with a rapid increase in agonist affinity, as in the case of the nicotinic receptor (1,24,25). Whether this latter process involves only a pair of known high-affinity agonist sites (on the P subunits, for the GABAA receptor), or also other, low-affinity sites is not yet known for either receptor type.

Molecular Biology of G A B A A Recepcor 11

M O L E C U L A R C L O N I N G A N D E X P R E S S I O N

Cloning of the cDNAs Encoding the Receptor Subunits

The GABAA receptor protein from bovine cerebral cortex, purified by a modified sodium deoxycholate/CHAPS procedure (42,43), was used either in its entirety or after separating the a subunit by gel electroelution. Selective cleavage by cyanogen bromide or by trypsin, and high-pressure liquid chromatography separation of the peptides produced, were followed by their gas-phase microsequencing, and a set of oligonucleotide probes corresponding to various segments of the two subunits was thus prepared: These phases of the work were accomplished by H. Rodriguez and J. Ramachandran at Genentech, San Francisco. These probes were used to screen bovine brain cDNA libraries constructed in phage \gtlO, and by cloning plus primer extension where appropriate, full-length cDNAs encoding the a subunit and the P subunit were ultimately obtained (39). These were definitively identified by the chemically determined peptide sequences encoded within them.

The calculated molecular weights of the mature a and (3 subunits are 48,800 and 51.400 daltons, respectively. Values of 53,000 and 57,000 daltons were de-duced from protein gels, as noted above; the difference is ascribed partly to several thousand daltons of carbohydrate subsequently attached in vivo to each subunit (35) and partly to the known anomalous migration in denaturing gels of such membrane-bound polypeptides.

Expression in the Xenopus Oocyte System

Identification of the cDNAs as encoding subunits of the functional GABAA receptor channel structure has been obtained by translation of the corresponding RNAs in the Xenopus oocyte system. This system, as applied to the special case of receptor mRNA identification, was developed previously (5 ,6 ,46,50) . Earlier, -he correct assembly of the functional GABAA receptor in this system was shown, using total mRNA extracted from rat or chick brain (26,46,47) . Therefore, the cloned a and P cDNAs were used as templates to synthesize the corresponding RNAs in vitro, using the bacteriophage SP6 RNA polymerase/transcription system. The two pure RNAs were microinjected, together, into the Xenopus oocyte. The application of a low concentration of G A B A produced a large and immediate conductance change in the cell membrane. Control (noninjected or medium-only injected) oocytes never gave any response to G A B A .

The expressed receptor shows the characteristic sensitivities to bicuculline and picrotoxin and the strong potentiation by pentobarbital (39). The channel opened has been identified as a chloride channel of the same type as that (4,9) opened

22 E. A . Barnard t'[ al.

in the native GABAA receptor on neurons. The evidence for this is: ( I ) The

reversal potential in normal amphibian Ringer solution was found to be about

- 25 mV (at 20°C), which agrees closely with the equilibrium potential of chloride

in the Xenopus oocyte in these conditions. (2) When the external chloride con-

centration was varied (in further studies made by E. S. Levitan) the reversal po-

tential varied linearly, with a slope of - 58 mV for a lO-foId increase, as predicted

for chloride permeation. (3) The series of relative anion permeabilities, which has

been determined here for the expressed receptor channel, compares closely with

that reported (4,9) for the GABAA receptor channel on cultured neurons.

These results show that the cloned a and 3 cDNAs encode a structure that has

pharmacological properties of the GABAA receptor and its anion channel.

SEQUENCE COMPARISONS FOR THE GABAA AND RELATED RECEPTORS

To interpret the amino acid sequences of the GABAA and glycine receptor

polypeptides, hydropathy plots were generated from them by a suitable computer

program (Fig. I): These give a quantitative estimate of hydrophobic (a'bove the

line) or hydrophilic (around or below the line) character of the side-chains along

the sequence. Candidates for membrane-spanning segments are those where a clear

peak of hydrophobicity occurs over a region of the order of 20 amino acids, i.e.,

sufficient to form a helix that can span the bilayer. Four of these putative trans-

membrane regions are seen in each of the subunits, at the same relative positions,

designated M l , M2, M3, and M4. Since in each of the cDNAs an N-terminal

signal peptide sequence is encoded, the N-termini are predicted to be extracellular,

and the region from the mature N-terminus to the start of M l is presumed to lie

outside the membrane.

When aligned with the corresponding hydropathy profiles for any of the subunits

of the nicotinic acetylcholine receptor (nAChR) , the number and location of the

transmembrane hydrophobic segments assigned match fully those in the GABAA

receptor subunits. The segment of approximately 210 amino acids between the N-

terminus and the M l a-helix in the nAChR a subunit is known to contain the

high-affinity A C h binding sites (3,29,37).

The deduced amino acid sequences of the two subunits of the GABAA receptor

exhibit a significant degree of homology to those of the nAChR subunits (Table

1). Further, comparison of either GABAA receptor subunit amino acid sequence

with that recently obtained by H. Betz and co-workers (21) for one of the subunits

of the glycine receptor, the 48,000 Mr strychnine-binding subunit, reveals quite

striking homologies (Fig. 1, Table I).

We have concluded (39), therefore, that the GABAA receptor is part of a

superfamily of ligand-gated ion channel/receptors. Reasonable phylogenetic trees

Molecular Biolog^' of GABAa Receptor 23

100 200 300 400

100 200 300 400

100 200 300 400

FIG. 1. Hydropathy profiles of the GABAA and glycine receptor subunit sequences. Hydropathy profiles of the mature polypeptides (i.e., excluding the proposed signal sequences) were computed according to Kyte and Doolittle (31) using a window size of 17 residues and plotted with a one-residue interval. Solid bars indicate the positions of the proposed hydrophobic transmembrane domains.

can be constructed based on all the sequences, which are consistent with the idea that these receptors all evolved from a common ancestral receptor, presumably a homo-oligomer. Evidence for the existence of this family leads one to expect that the major excitatory neurotransmitter receptor type in the mammalian brain, the glutamate receptor series with an associated cation channel, would be more ho-mologous to the n A C h R rather than to the G A B A A or glycine receptors.

24 E. A. Barnard et al.

T A B L E I. Percentage amino acid sequence homology of receptor/ion chimnel subunits"

Identical residues

G A B A a G A B A p GLY

n A C h R a 19 15 15 G A B A a 35 34 G A B A p — — 39

Identical plus conservatively substituted residues

G A B A a G A B A p GLY

n A C h R a 38 32 3? G A B A a 57 56 G A B A P — — 59

" The sequences of the mature polypeptides of the bovine GABAA receptor, the rat glycine receptor, and the bovine muscle n A C h R a subunit have been compared (see ref. 7). Note that similar degrees of homology are found when the latter is replaced by any of the 20 or so subunit sequences known for electric organ, muscle, or brain nicotinic receptors. Optimized alignments of all the pairs were made usmg the com-puter program DIAGC)N (see ref. 39) . [From ref. 7 (where conservative substitution is defined).}

MULTIPLE TYPES OF GABAA RECEPTOR SUBUNITS

Further cloning has yielded several related but differing cDN As, and two of these again contained deduced peptide sequences identical to short sequences chemi-cally derived from the a subunit of the purified receptor. Full-length a I (the original a) , a2, and a 3 subunit sequences were thus identified. a2 shows 79% identity and a 3 72% identity with the a l sequence, and each only about 35% identity with p (33). They both show the same hydropathy profile as the a l polypeptide.

When the corresponding RNA was injected with P-subunit RNA into oocytes, each gave expression generally similar to that for ( a l + p), with the same effects of modulatory drugs as found there and noted above. Each pair gates a chloride channel of the same ion selectivity. However, quantitatively each pair functions differently (33). (a2 + P) is more sensitive to GABA and (a3 + p) is less sensitive than the (al + P) pair, as can be seen in Fig. 2. The half-maximal concentrations are:

a l + p 12 M-M GABA a2 + P 1.3 [JiM GABA a3 + p 42 |xM GABA

Molecular Biology of G A B A a Receptor 25

Further evidence chat these represent different subtypes of the receptor a subunit has come from their differing regional and temporal expressions in the brain (33) . By Northern blotting, it was found, for example, that the a l m R N A is highest in the bovine cerebellum and (x2 in the hippocampus. a 3 m R N A is prominent in the neonatal cerebellum, but declines with age. These findings have been confirmed by in situ hybridization with oligonucleotide probes specific for each a sequence: A differential distribution of the three a subtypes is thus revealed in recent studies at the tissue and cellular levels by W . Wisden, B. J . Morris, and S. P. Hunt at Cambridge and by B. Shivers and M. Kohler at Heidelberg.

1 / iM

G A B A

3 0 0 / i M G A B A

a1 + p

a2 + p

a3+ p

lOOnA

2 0 0 n A

lOOnA

FIG. 2 . Exchanging a subunics alters the sensitivity of the receptor to G A B A re-sponses in Xenopus oocytes to 1 |XM and to 3 0 0 (JUM, as shown, after the injection of ( a l + P) , (aZ + P) . or ( a 3 + P) RNAs. (From ref. 3 3 . )

26 h: A Barnard c( al

A surprising recent observdcion has been chac the a subunic alone can form a GABA-activated channel. When a 1 RN A was first injected (39) this phenomenon was not detected m the vokaj^e-clamp studies because of the low level of this single-subunit response, but more sensitive voltage-clamp and patch-clamp studies (7a) have repeatedly confirmed that a 1, a2 , o r a 3 expressed alone in the oocyte yields some channels that can be opened by C A B A . Their properties, including their pharma-cology, are similar to those of the (a + (3) channels. This is not unprecedented, as a single subunit of a neuronal nAChR, a 4 of Boulter et al. (10) , can produce with low efficiency, via its RNA in the oocyte, an ACh-gated channel.

A Model for the Structure in the Membrane

Figure 3 shows a model for the a and (3 subunits of the GABAA receptor in its neuronal membrane. A view cutting through the molecule is shown, with one of the a and one of the p subunits illustrated. The four hydrophobic segments in each are assumed to be transmembrane helices. The structure of the large N-terminal extracellular domain is shown in an arbitrary' manner. Two cytoplasmic loops are shown on each subunit, one long and one very short: On the (3 subunit only, the long loop contains a site for cyclic AMP-dependent phosphorylation, a potential regulatory feature.

A remarkable similarity of amino acid sequence is observed within the M2 do-mains in both the GABAA and glycine receptors, extending to 10 identical residues in a hydroxy-rich sequence of 12. Surely it is no coincidence that this domain has been implicated [by affinity labeling (20 ,22 ,27) or mutagenesis (28) resultsl as forming part of the lining of the ion channel in the nAChR. This stretch of 12 residues in the M2 sequence of the GABAA and glycine receptor subunits deter-mined to date does not occur in any of the 20 known n A C h R subunits (peripheral and neuronal). We suggest, therefore, that this stretch contributes to the ion selectivity of the channel.

T h e model (Fig. 3) also shows a high concentration and excess of positively charged amino acid side-chains in the immediate vicinity of the ends of the pro-posed transmembrane domains of the GABAA receptor A and (3 subunits. Ex-amination of the corresponding regions of the 48 ,000 Mr subunit of the glycine receptor reveals a similar distribution of positively charged residues. In contrast, most of these positions are occupied by neutral or negatively charged residues in the nAChRs. Supporting this marked clustering of cationic groups seen in the model, there is much documentation (36) for a strong positive influence of per-meating anions on the binding of the specific allosteric ligands for the G A B A A receptor mentioned earlier. Moreo\'er, channel conductance studies (9) give evi-dence for at least two anion binding sites associated with the channel. Clusters of arginines and lysines at the channel mouth, therefore, are proposed to act in both the G A B A A and glycine receptors as an ion-selective filter and to increase the driving force for anion flow on opening of the channel.

Molecular Biology of G ABA a Recepior 11

HON

o o O ^ ^

o o o o o 9 1

E ) a m : m u u \ R

O O © « O ^ O ^

0 0 0 0 9 0 0 0

o o

o

ITsTTRACELLUUSJ

a B F I G . 3 . A schemat ic model for the topology of t h e G A B A A receptor in the mem-brane (39) . Four membrane-spanning helices in each subuni t are shown as cylinders. T h e structure in the extracellular domain is d rawn in an arbitrary manner , but the presumed (i loop formed by the disulfide bond predic ted at cysteines 139 and 153 ( a I - s u b u n i t numbering) is shown (a). Potential extracel lular sites for N-glycosylation are indicated by triangles, and a possible site for c A M P - d e p e n d e n t serine phosphor-ylation, present only in t he P subunit, is deno ted by an encircled P. Those charged residues tha t are located wi th in or close to the ends of the membrane-spanning do-mains are shown as small circles with charge marked. It is proposed that four or five subunits of this general type are complexed in the receptor molecule so as to align the membrane-spanning domains, only some of wh ich will form the inner wall of a central ion channe l .

Z8 E. A. Barnard et al.

Finally, in che N-terminal domain a region adjacent to a pair of cysteines is particularly well conserved between all of the known subunits of ligand-gated re-ceptors. A constant aspartic acid is one of its general features. The invariant proline halfway along the loop is predicted to be in a hairpin p turn, and the whole loop can form a p structure (in a scale model), with one face completely hydrophobic. This is the only constant structural motif that can be invoked readily to explain a low-affinity site for the cationic agonist on all the subunits, should this be established.

We do not know at present how many subunits of which type form the rest of the GABAA receptor structure. Chemical evidence had suggested an azpz stmcture (35), but now that we know that a as well as P subunits carry low-affinity GABA sites, that issue has become uncertain. From human brain another type of subunit cDNA has recently been isolated (P. R. Schofield and P. H. Seeburg, unpublished observations), which is less strongly homologous to either a and p, and is termed X; this can confer benzodiazepine sensitivity on expressed units lacking this. Its polypeptide may not be resolvable from a in SDS-PAGE. At present we must leave open the question of the subunit stoichiometry of the receptor. The heterogeneity of benzodiazepine receptors noted in the introduction should become explicable in terms of the different subunit types, a l through a3, etc. The high-affinity G A B A sites, presumed to be located on the p subunit, may be involved primarily in desensitization, while the low-affinity ones appear to be on all other subunits and to be involved in opening the channel.

C O N C L U S I O N S

The mass of detailed information provided by DNA cloning has brought us to the truly molecular level of the G A B A receptor, but has thereby raised many complex questions hitherto scarcely perceived. We believe the further application of cloning and expression techniques will in time provide a satisfying answer to the still-open questions of how the GABAA receptor is constructed and how it functions.

R E F E R E N C E S

1. Adams, P. R. (1981): J. Membr. Biol, 58 :161 -174 . 2. Akaike. N. , Inoue, M.. and Krishtal, O. A. (1986): J. Physiol., 379 :171-185 . 3. Atassi, M. Z., Mulac-Jericevic, B.. Yokoi. T. , and Manshouri, T. (1987): Fed. Proc.,

46 :2538 -2547 . 4. Barker, J. L., and Owen, D. G. (1986) : In: Benzodiazepine/GABA Recepcors and Chlo-

ride Channels: Structural and Functional Properties, edited by R. W. Olsen and J. C. Venter, pp. 135-166 . Alan R. Liss, New York.

Molecular Biology of G A B A A Receptor 29

5. Barnard, E. A.. Miledi, R.. and Sumikawa. K. (1982): Proc. R. Soc. Lorvi. B., 215:241-246.

6. Barnard, E. A., Beeson, D., Bilbe, G., Brown, D. A., Constanti, A., Houamed, K., and Smart, T. G. (1984): ]. Recept. Res., 4 :681-704.

7. Barnard, E. A., Darlison, M. G., and Seeburg, P. (1987): Trends Neurosci., 10:502-509.

7a. Blair. L A. C , Levican, E. S., Marshall. J., Dionne, V. E., Barnard, E. A. (198X): (in press).

8. Bormann, J. , and Clapham, D. E. (1985): Proc. Nad. Acad. Sci. USA, 82:2168-2172. 9. Bormann, J . , Hammill, O. P., and Sakmann, B. (1987): J. Physiol., 385:243-286.

10. Boulter, J. , Connolly, J., Deneris, E., Goldman, D., Heinemann, S., and Patrick. J. (1987): Proc. Nad. Acad. Sci. USA 84:7763-7767.

11. Braestrup, C. . and Nielsen. M. J. (1981): J. Neurochem., 37:333-341. 12. Brown, C. L , and Martin, I. L. (1985): Eur. ]. Pharrruicol, 106:167-173. 13. Casalotti, S. O., Stephenson, F. A., and Barnard, E. A. (1986): J. Biol. Chem.,

261:15013-15016. 14. Cash, D. J. , and Subbarao, K. (1987): Biochemistry, 26:7562-7570. 15. Cooper, S. J . . Karkham, T. C., and Estall, L. B. (1987): Trends Pharmacol. Sci.,

8:180-184. 16. Deng, L., Ransom, R. W., and Olsen, R. W. (1986): Biochem. Biophys. Res. Commun.,

138:1308-1314. 17. Drexler, G., and Sieghart, W. (1984): Eur. J. Pharmacol, 99:269-277. 18. Eichinger, A., and Sieghart. W. (1986): J. Neurochem., 46:173-180. 19. Enna, S. J. (1983): In: The GABA Receptors, edited by S. J. Enna, pp. 1-23. Humana

Press, Clifton, New Jersey. 20. Giraudat, J . , Dennis, M., Heidmann, T. , Chang, J.-Y. and Changeux, j .-P. (1986):

Proc. Nad. Acad. Sci. USA, 83:2719-2723. 21. Grenningloh, G., Rienitz, A., Schmitt, B., Methfessel, C. , Zensen, M., Beyreuther,

K., Gundelfinger, E. D.. and Betz, H. (1987): Nature, 328:215-220. 22. Guy. R. H.. and Hucho, F. (1987): Trends Neurosci., 10:318-321. 23. Hebebrand, J . , Friedl, W., Unverzagt, B., and Propping, P. (1986): J. Neurochem.,

47:790-793. 24. Heidmann, T . , Bernhardt, ]., Neumann, E., and Changeux, J.-P. (1983): Biochem-

istry, 22:5452-5459. 25. Hess, G. P., Cash, D. J., and Aoshima, H. (1983): Ann. Rev. Biophys. Bioeng.,

12:443-473. 26. Houamed, K., Constanti, A., Smart, T. G. , Bilbe, G. , Brown, D. A., Barnard, E.

A., and Richards, B. M. (1984): Nature, 310:318-321. 27. Hucho, F., Oberthiir, W., and Lottspeich, F. (1986): FEBS Utt., 205:137-142. 28. Imoto, K., Methfessel, C.. Sakmann, B., Mishina, M.. Mori, Y., Konno, T . . Fukuda,

K., Kurasaki, M., Bujo, H., Fujita, Y., and Numa, S. (1986): Nature, 324:670-674. 29. Karlin, A., Kao, P. N., and DiPaola, M. (1986): Trends Pharmacol. Sci., 7:304-308. 30. Kirkness, E. F., and Turner, A. J. (1986): Biochem. J., 233:259-264. 31. Kyte, J . , and Doolittle, R. F. (1982): J. Mol. Biol., 157:105-132. 32. Umbert, J. J . . Peters, J. A., and Cottrell, G. A. (1987): Trends Pharmacol. Sci.,

8:224-227. 33. Levitan, E. S., Schofield, P. R.. Burt. D. R., Rhee, L. M., Wisden, W.. Kohler,

M.. Rodriguez, H., Stephenson, F. A., Darlison. M. G. , Barnard, E. A., and Seeburg, P. H. (1988): (in press).

30 E. A. Barnard et al.

34. Lippa, A. S., Garrett:, K. M., TabakofF, B., Beer, B., Wennogle, L. P., and Meyerson, L. R. (1985): Brain Res. Bull., 14:189-195.

35. Mamalaki, C , Stephenson, F. A., and Barnard, E. A. (1987): EMBO J., 6 :561-565.

36. Olsen. R. W., and Venter, J. C , (eds.) (1986): Benzodiazepine/GABA Receptors and Chloride Channels: Structural and Functional Properties. Alan R. Liss, New York.

37. Ralston, S., Sarin, V., Thanh, H. L , Rivier, J. , Fox, J. L., and Lindstrom, J. (1987); Biochemistry, 26:3261-3266.

38. Sato, T. N., and Neale, J. H. (1987): Biochem. Biophys. Res. Commun., 146:568-574.

39. Schofield, P. R., Darlison, M. G., Fujita, N., Burt, D. R.. Stephenson, F. A., Rodriguez, H., Rhee, L. M., Ramachandran, J., Reale, V., Glencorse, T. A., See-burg, P. H., and Barnard, E. A. (1987): Nature, 328:221-227.

40. Sieghart, W., and Karobath, M. (1980): Nature, 286:285-287. 41. Sieghart, W., Mayer, A., and Drexler, E. (1988): Eur. J. Pharmacol., 88:291-299. 42. Sigel, E., Mamalaki, C., and Barnard, E. A. (1982): FEBS Lett., 147:45-48. 43. Sigel, E., and Barnard, E. A. (1984): J. Biol Chem., 259:7219-7223. 44. Sigel, E., Stephenson, F. A., Mamalaki, C., and Barnard, E. A. (1984): J. Recept.

Res., 4:175-188. 45. Sigel, E., and Bauer, R. (1987): MoL Pharmacol., 32:749-752. 46. Smart, T. G., Constanti, A., Bilbe, G., Brown, D. A., and Barnard, E. A. (1983):

Neurosci. Lett., 40:55-59. 47. Smart, T. G., Houamed, K. M., Van Renterghem, C., and Constanti, A. (1986):

Biochem. Soc. Trans., 15:117-122. 48. Squires, R. F., Benson, D. I., Braestrup, C., Compet, J., Klepner, C. A., Myers,

v . , and Beer, B. (1979): Pharmacol. Biochem. Behav., 10:825-830. 49. Stephenson, F. A., Mamalaki, C., Casalotti, S. O., and Barnard, E. A. (1986):

Biochem. Soc. Symp., 52:33-40. 50. Sumikawa, K., Houghton, M., Emtage, J. S., Richards, B. M., and Barnard, E. A.

(1981): Nature, 292:862-864. 51. Van Renterghem, C., Bilbe, G., Moss, S., Smart, T. G., Constanti, A., Brown, D.

A., and Barnard, E. A. (1987): Mol. Brain Res., 2:21-31.

The EMBO Journal vol.8 no.6 pp. 1 6 6 5 - 1670 , 1 9 8 9

GABA;^ receptor (3 subunit heterogeneity: functional expression of cloned cDNAs

Sanie Ymer, Peter R.SchofieldV Andreas Draguhn^, Pia Werner, Martin Kohler and Peter H.Seeburg^

Laboratory of Molecular Neuroendocrinology, ZMBH, University of Heidelberg, INF 282, D-6900 Heidelberg, FRG 'Present address: Pacific Biotechnology Ltd, 7 2 - 7 6 McLachlan Avenue, Rushcutters Bay 2011, Australia ^Max-Planck-Institut for Biophysical Chemistry, Am Fassberg, D-3400 Gottingen, FRG ^Corresponding author Communicated by P.Seeburg Cloned cDNAs encoding two new /? subunits of the rat and bovine GABA^ receptor have been isolated using a degenerate oligonucleotide probe based on a highly conserved peptide sequence in the second transmembrane domain of GABA^ receptor subunits. The i82 and /33 subunits share -72% sequence identity with the previously characterized 131 polypeptide. Northern analysis showed that both and /33 mRNAs are more abundant in the brain than (31 mRNA. All three (3 subunit encoding cDNAs were also identified in a library constructed from adrenal medulla RNA. Each jS subunit, when co-expressed in Xenopus oocytes vdth an a subunit, forms functional G A B A A receptors. These results, together with the known a subunit heterogeneity, suggest that a variety of related but functionally distinct GABA^ receptor subtypes are generated by different subunit combinations. Key words: GABA^ receptor/jS subunit/receptor subtypes/ molecular cloning/oocyte expression

Introduction GAB A (7-aminobutyric acid), the major inhibitory neurotransmitter in the vertebrate brain, mediates neuronal inhibition by opening a chloride channel integral to the G A B A A receptor, which is also the target for a variety of therapeutically important drugs (reviewed in Olsen and Venter, 1986). Affmity-purified receptor is electrophor-etically resolved into two major bands {a, 4 8 - 5 3 kd and |3, 5 5 - 5 7 kd). The smaller band can be photoaffmity-labelled by benzodiazepine derivatives and the larger one by GABA agonists (reviewed by Stephenson, 1988). Molecular cloning of cDNAs encoding GABA^ receptor subunits was facilitated by peptide sequences derived from purified receptor (Schofield etal, 1987; Levitan etal, 1988). Analysis of these cDNAs established that the ' a band' is heterogeneous, consisting of several variants of the a subunit (Levitan et al., 1988) and of other subunits (Pritchett et al, 1989). This confirmed the molecular heterogeneity of G A B A A receptors, postulated on the basis of pharmaco-logical (Squires et al, 1979; Braestrup and Nielsen, 1981; Unnerstall etal, 1981; Cooper etal, 1987) and photo-©IRL Press

affinity labeUing (Mohler et al, 1980; Sieghart et al, 1983; Fuchs et al, 1988) studies. However, the '13 band' was regarded as being homogeneous (Haring etal, 1985; Mamalaki etal, 1987). This band is so far molecularly characterized by only one cloned cDNA that encodes a subunit (/31) with significant sequence similarity to the a subunits and, when co-expressed with these in Xenopus oocytes, produces functional G A B A A receptors (Schofield et al, 1987; Levitan et al, 1988). Sequence homology also extends, in part, to other ligand-gated ion channels, reflecting the existence of a receptor superfamily (Schofield et al, 1987; Grenningloh etal, 1987a).

Comparison of the polypeptide sequences of the three G A B A A receptor a subunits, the jSl subunit as well as the 48 kd subunit of the glycine receptor revealed that the highest sequence identity resides in the four putative transmembrane segments (Ml—M4). In particular, a contiguous sequence of eight amino acids rich in threonines is found in M2 of all these polypeptides (Grenningloh et al, 1987b) which, by analogy to M2 of nicotinic acetylcholine receptor (Imoto et al, 1988; Leonard et al, 1988), is thought to form part of the channel lumen. We have used a highly degenerate oligonucleotide probe encoding this peptide sequence to screen for additional G A B A A receptor subunits. We report the isolation of two new /3 subunit encoding cDNAs, and iS3, from both rat and bovine brain cDNA libraries and show that these new (3 subunits are more abundant in brain than the (31 subunit. Functional expression in Xenopus oocytes demonstrates that these (3 subunits are capable of combining with an a subunit to form G A B A A receptor chloride channels.

Results Cloning of and (33 subunit cDNAs A cDNA library constructed from bovine brain RNA was screened with the degenerate oligonucleotide probe and numerous hybridizing signals were obtained. Among these were cDNAs encoding the known G A B A A receptor subunits a l , Q;2, 0.3 and /31 (Schofield et al, 1987; Levitan et al, 1988), which were identified by subunit specific oligo-nucleotides. The remaining hybridizing clones were sequenced. Preliminary sequence data were obtained either directly from X DNA or from recombinant M13 DNA, using the degenerate oligonucleotide as a primer. Clones encoding new subunits were identified by homology of their deduced amino acid sequence with the third transmembrane region of previously characterized G A B A A receptor subunits.

Two C D N A clones were isolated that encoded different polypeptides showing high sequence identity to the G A B A A receptor /31 subunit. These polypeptides were designated as /32 and jS3 subunits. By comparison with the jSl polypep-tide, the encoded /32 subunit sequence lacked an initiation codon and the j83 subunit sequence lacked the N-terminal 45 amino acids of the complete polypeptide. Full-length

1665

S.Ymer et al.

cDNA clones were not identified upon rescreening, but a complete jSS subunit encoding clone was isolated from a bovine adrenal gland cDNA library (see below). We then screened a rat forebrain cDNA library with oligonucleotides based on the bovine 13 subunit coding sequences and isolated full-length cDNA clones for the rat |(31, (32 and 133 subunits. The nucleotide and deduced amino acid sequences of the rat subunits, including part of the 5' and 3' non-coding regions, are shown in Figure 1. All three cDNAs encode polypeptides of - 4 5 0 amino acid residues (M^ 52 kd) with a 25-residue signal peptide (Von Heijne, 1986). The predicted mature polypeptide sequences contain four putative membrane-spanning regions, three potential N-linked glycosylation sites and a 15-residue-long, cysteine-flanked region, characteristic

of all subunits of ligand-gated ion channels (Criado et al., 1986; Schofield et al, 1987). Notably, all three |8 subunits contain a cAMP-dependent phosphorylation site in a homologous position within their putative intracellular domain, suggesting involvement in the cellular control of receptor activity.

Comparison of the three /5 polypeptides predicted from either rat or bovine cDNAs (Figure 2) shows that 72 % of the residues are invariant, reflecting an identity similar to that seen between different a subunits (Levitan et al., 1988). As for the a variants, regions of highest homology include the membrane-spanning domains and large extracellular region. Low sequence similarity is seen in the signal peptides and in the intracellular domain located between M3 and M4.

GAATTACTGCACTGGGMGACTAAGTTGGATCTCCTCICTTCAOTOMTCCCTCAATCCMrcAAAAACTAASOCCJTGTCMOAGTCCG

GAAAAGGGOCTACTTTGOGATtTCGTCATTTCCCTTAATAATCCCCGCTOTCIGTGCTCAGAGTGTCAATGACCCTAGTAATATGTCGCT K R G r F G I U S F P L I I A A V C A W S V I I D P S M H S L

GGITAAAGAGACGGTGGACAGACTGTTGAAAGGCTAIGACATTCGICTGAGACCACATITCGGAGGICCCCCTGICGCAGTAGGAATGAA V K E T V D R L L I C G y O I R L R P O F G G P P V A V G H N

CATTGATATCCCCAGCATCGATATGGTTTCTGAAGTCAAIATGGACTACACCTIGACCATGIATTICCAGCAAGCCTGGAGAGATAAGAG l O I A S I O H V S E V K H D T I L I H l t F O O A U R D K R

ACTGTCCTACAATGTAATCCCITIAAACTTGACTTTCGACAATCGAGTGGCAGACCAGCTCTCGGTGCCTGACACCIACTTCCTGAATCA L S Y M V I P l K L T L O H R V A D O L U V P O T t F L N O

TAAGAAGTCATTTGTACATGGAGTGACTGTCAAAAACCGTATGATTCGACtGCATCCAGATGGTACTGICCTGTATGGCCTCAGAATCAC K K S F V H G V T V K N R H I R L H P D G T V L Y G L I I I T

AACTACAGCTGCCTGCATGAIGGACCTAAGGCGGTATCCACTGGAIGAACAAAACTGCACGTTGGAGATCGAAAGCTATGGCTATACAAC T T A A G M H D L R . T . . t . . 9 . , 9 T I E I E S Y G Y T I

TGATGACATTGAGTTTTACTGGCGTGGCGATGACAATGCAGTCACGGGAGIGACAAAGATTGAGCTTCCTCAGTICICCATTGTAGATTA D D I E F Y U 8 G D D N A V T G V T I C I E 1 . P 0 F S I V D Y

lAAACTCATCACCAAGAAAGTTGTTTTCICCACAGGITCITAICCCAGATTGTCCCTAAGCTTTAAGCTGAAAAGAAACATTGGCTACTT K L I I K I C V V F S T G S Y P R L S L S F K L I C R N I G ](_£

CATCCTGCAGACAIACATGCCATCCATTCTGATIACCATCCICTCCTGGGTCTCCTTTTGGATCAACTAIGATGCtTCTGCTGCACGEGT H P S 1 L I S » V S F S A A

10B1 CCAGCGCCAAAAGAAAGCAGCTGAGAAAOCTGCYAATGCCAACAACGAGAAGAYGCGCCTGGATGTCAACAAGAYCGACCCACATGAGAA 1170 R 0 K K A A H A H N E IC

ACTCATACGACAGCGCCAGCATCCAGTACCGCAAAGCACTGAGCAGCCGTGAGGGGTYCGGACGAGGGCTGGACAGGCACGGGGTGCCCG S Y D S A S i a Y R K P L S S R E G F G R G L D R H G V P G

GCAAGGGTCGCATCCGCAGACGTGCGTCTCAGCTCAAAGTGAAGATCCCTGACTTGACTGACGTGAACTCCATAGACAAGTGGTCCCGAA S R H «30

1531 CTTTCCTCCTCGCTTGtTTTTTAACCCCACAGATITCCAACAGTGGTAACTGCTATGGTTTTIGAGGTAAGAGAITCCGCCTTCCAGnG 1621 GITTTCGATCTTGTGTGYCAOTTTTTATTArCACACCArGITTACTICAACAACAACAACAAAAOGIAYTTttTCCTGICrATTGTGGTC 1711 CAAGGTACCAGCTCTGTAAGAGTTCCATCAATTACTTTGTTTACAAACACAAAGAGAGGTAGGTGGGTGTTAGACA1TGTTGGCAGTCAT laOl TCTTAGTIGCCCIGGAICTTACTGGATTATTTTAAAATGAAAATGTACAGTGGACCCTGCCGICCACCIGCACTGTGITCTGGGTAAACT la91 GTAACCATCTCATGCTGCCAAACAA1TAAATAGAAAATTTAAAAAAAAAAAAAG 19A4

1620 1710 1800 1890

GTATGATGCCTCCAGCATCCAGTATCGGAAAGCTGGCTTGCCTAGGCArAGTTTTGGCCGCAACGCCCIGGAACGACATGIGGCACAAAA y o a s s i q y r k a c l p r h s f g r b a l e r h v a q k

GAAAAGTCGCCTGAGGAGACGIGCCTCCCAACTGAAAATCACCAICCCCGACTTGACTGATGTGAACGCCATIGATCGGTGGICCCGCAI 402 I C S R L R R R A O L K I T I P O L T D V N A I O S

1441 TITCITCCCTCTGGTGT1TTCCTTCTTCAACATCGTCTAT1GGC1TTACTATGTGAACTAAACTCCAGCCTCCCATGAGAAGCAAGGACI 1530 F P V F 5 F F J t_V

AGATTCCTCTTCAAACAGITGTACAGCTTGATGCCTGATGTAGGACTTGGAAAACACATCAATCCAGGACAAAAGCGATGTTAAAATACC TTAGTTGCTGGCCTAICCTGTGGTCCATTTCAIAACAITTGGGITGCIICTACIAAGTAATGACTACACIAAGGtCCTCGTGGTITICCA GTTACAAIGCAAGTGATTTGTACACATGCCGGCAAGAICTTTGTCAGITTTAGGACACAGCCTACICAGAGGGTTAATIACCTAGAIICT AGAAGACACTGGAAAGCTCAATGGCATGGGCAGTCAAGTCTCTGAAACAGTCATTTCCAAATGCTCTCCATCGTTGTTCCIAAGGTTGGC ACGCAGTTGGGACAGCACTGIGCtTATAAAACATTATCCGCAAIAATCGAAACATGCTACTICAATATGGGClTIGAGGTCIAAGCCAGA TGATGATGATGGTGGTGATG 2000

1620 1710 1800 1890 1980

1 GCAGCACCCCGCCTCGGGGTCGCGACGGCGGCGGGGCGCCCCCTCCCCCGTGCCGGGGCOCGGCGAAGGGATGTGGGGCTTIGCGGGAOG 90 H W G F A G G -19

1GAACAIGTCCIT1GT 180 N H S F V 12

GAAGGAGACGGICGACAAGCYGriOAAAGGCrACGACAYTCGCCTGAGACCGOACTTCGGGGGTCCCCCAGTCTGCGTGGGCAtGAACAT 270 K E T V O K L L I C G Y D I R L R P D F G G P P V C V G K I I I 42

CGACAYCGCCAGCATCGACATGGITTCTGAAGICAACATGGATTATACCITAACIATGTATTICCAACAATAITGGAGAGATAAAAGGCT 360 D I A S l D H V S E V H K D Y T L T H Y F O O Y U R O I t R l 72

COCCTACTCTGCOAICCCTCTCAACCTCACOCTTGACAATCGAGTGGCYGACCAGCTCTGGGTGCCTGACACATATTTCITAAATGACAA 450 A Y S G I P L W L T l O H R V A O O L U V P D I Y F L I I D I t 102

AAAGTCATTTGTGCACGGAGTGACAGTGAAAAACCGCATGATCCGCCTCCACCCTGATGGAACAGTGCIGIACGGGCTCAGGATCACCAC 540 K S F V H G V T V K N R M I R L H P O G I V L Y G L R I T T 132

CACAGCAGCTTGCATGATGGACCTCAGAAGATACCCACIGGAIGAGCAAAACTGCACCCTGGAAATIGAAAGCIATGGATACACCACGGA 630 T A A C..N.,1,9..L.,R,,«..Y P .L 9 E 0 II C I L E I E S Y G Y T I D 162

TGACATIGAATTTIACTGGCGTGGCGGGGACAAGGCTGTTACTGGCGIGGAAAGGATCGAGCTCCCACACTTCTCCATTGTGGAGCACCG 720 d i e f y w r o o d i : a v t o v e r i e i . p o f s i v e h r 192

IClGGTCICCAGGAATGTTGTCTICGCCACAGGTGCCTACCCTCGACTCTCATTGAGTtTTCGGTIGAAGAGAAACATTGGGTACTICAT 810 V V F A I G P R S L S F I G

B11 ACTTCAGACGTATAIGCCCTCAAICATGATCACAATCCTCTCATGGGTGTCCTTCTGGATCAAITATGATGCATCTGCTGCICGAGTTGC 900 223 L 0 T Y M P S 1 H I T I I S U V S F U 1 N Y 0 A S A A R V A 252

901 CCTAGGGATTACCACCGIGCTCACCAIGACAACCAICAACACICACCTICGAGAGACICTACCCAAAAICCCCTACGTCAAAGCCATCCA 990 253 L G 1 f T V L T H T T 1 11 I H L R E I L P K 1 P Y V K A J_D 282

991 CATGIACCTGATGGGTTGCTTCGTCTTTGTATTCCTGGCACTtCTGGAGTATGCCITIGTCAACTAIATTTTCTTTGGACGAGGTCCCCA 1080 Y I- C F F V F I A L L F V N G P U

ACGGCAGAAGAAGCTTGCGGAGAAGACAGCCAAGGCCAAGAATGATCGATCCAAGAGTGAAATCAACCGGGTGGAIGCTCACGGGAATAT R O K K L A E K T A I C A K I I D R S K S E I H R V D A K G N I

CCYACTAGCACCGATGGATGtTCACAATGAAATGAATGAGGTTGCAGGCAGCGTTGGTGACACCAGGAAtTCAGCAATAYCCYTTGACAA L L A P H O V U H E M I I E V A G S V G D I R H S A I S F D I I

CTCAGGAAYCCAGTATAGGAAACAGAGCATGCCCAAGGAAGGGCAIGGGCGGTACAIGGGAGACAGAAGGATCCCGCACAAGAAGACGCA S G I O Y R I C a S H P K E G H G R Y M G D R S I P U I C K T H

CCTACGGAGGAGOTCYTCGCAGCICAAAATCAAAATCCCTGATCTAACCGATGTGAATGCCATAGACAGATGGTCCCGGATCGTGTITCC 403 L R R R S S L K I K I P D I A ! D R Si_ F P

1441 ATTCACCTTTrcrCICYTCAACTTAGTTTACTGGCTGTACTATOTTAACTGAGTGACTGIACIIGATIliTTCAAAGACITCATITAACA 1530 433 F T F S L Y Y V

GTGAGTGAAAIATTACCCTGCCTGICAAGTTTTTAIACCAGTATACACACACACACACAGAAACAAACGCACACACAAACACACACAIAC ACAAACACACACACACACACACATACACAGACACACACAATTGIATATAIATGTGAACTTTCICAGCATATAlATAAAATACACGIGtAT ATGAGAATGTATGTGTATATGTTTATGTACACTGGTGAGAGTGCCTGTGIATGIAAAACAAATACGCATACAIACCATACATTTTGCAAC TATGGACAATTTAACACAGGATGCATAITAAAGAAAGTCAIAGCTTTTTITTTCTITITTAAITGAAAGGGACAAGTATC1AAATATIAT GCCTCAAGAAIGAGGGCGAGAAACACAGTCAICCCAAAGTGYGICTTrTATTAICATAAGTIITTOCIIAAGAAICAAAACGGAAYTCIT AGTTAAICTIIGGGCACICC 2000

1620 1710 1800 1890 1980

Fig. 1. Nucleotide and deduced amino acid sequences of rat GABAa receptor (31 (a), |82 (b) and (33 (c) subunits. Nucleotide and amino acid positions are indicated. Negative numbers refer to signal peptide residues. Potential N-linked glycosylation sites carry asterisks, the /3-structural loop flanked by cysteines is indicated by a broken line and the four transmembrane regions are underlined. Putative regulatory sites in the intracellular region of the |S subunits are denoted by small circles. Underlined residues in |32 (positions 3 3 6 - 3 5 2 ) correspond to a chemically determined peptide sequence of the homologous bovine subunit.

1666

GABAft receptor (3 subunits

Importantly, in both the rat and bovine 132 subunits this domain contains one of the peptide sequences (MXPHENI-LLSTLEIKNE) chemically determined from cyanogen bromide-cleaved affmity-purified bovine GABA^ receptor (Schofield et al., 1987). The /32 subunit sequence is the only one in which a methionine residue (the site of cyanogen bromide cleavage) precedes this peptide sequence (Figures 1 and 2) and residues P3, E5 and K15 uniquely distinguish the 132 from the (3\ and sequences. This demonstrates that the 132 polypeptide is a component of the natural GABA^ receptor complex. Interspecies homologies for /3 subunits are extremely high since on average only 10 residues are

V K E T

V K E T

M K L R P D F G G

M R 1 . K P D F G C

I I R L R P D F G C

V G M

V C M , A S 1 D M V S E V 1

A S I D M V S E V i

R A n i i : T A 2

R A ' l " f t E ' l ' A 3

T L T M V F Q Q r 1 . T M V F Q Q

T 1 . T M Y F Q Q T 1 . T M y F Q Q T L T H Y F Q Q TLTHYFQQ

1 P L N L T L D N R

. n Q L W V P D

R A T I i E T A I

R A T l i K r A 2

1 . U P D G T V L V G L R l T T T L I I P D G T V L Y G L R I T T T L M P D G T V L Y G L R l T T T [ . I I P D G T V L Y G L R I T T T L l l P D G T V l . Y G L R l T T T L H P D G T V l . Y G L R l T T T

B O V B E ' i A 2

B O V B F . T A J

1 . D E 0 N C T S Y G Y T T D n 1 S Y G Y T T 1 > D ] S Y C Y T T r > n 1

I E L P Q F S

1 ) Y K 1 . 1 T K K

substituted between the rat and bovine homologues. A similarly high degree of sequence conservation has been observed for the bovine (Schofield et al., 1987) and human (Schofield etal., 1989) al and /31 subunits. /3 subunit localization The extent of j3 subunit expression in the brain was investigated by Northern blot analysis. RNA samples were prepared from rat and calf total brain as well as from the cortex, hippocampus and cerebellum. Northern blots of these RNAs were hybridized with ^^P-labelled oligonucleotides complementary to DNA sequences encoding the divergent intracellular domain of the three (3 subunits. The autoradio-graphs (Figure 3) document that in the rat and bovine brain both the (32 and subunits are considerably more abundant than the /51 subunit. The relative abundance of the /3 subunit mRNAs parallels that of the respective cDNA clones in the brain libraries.

In rat brain, the hippocampus and cerebellum contain the highest levels of (3 subunit mRNAs while the cortex shows only small amounts of /3-specific RNA. ^2 subunit mRNA is altogether more abundant than /33 mRNA. The three 13 subunit encoding mRNAs differ in size, with jSl mRNA being the largest ( ~ 12 kb) followed by (32 mRNA ( ~ 8 kb) and |S3 mRNA, which is represented by two size forms

pi P2 |)3 in |i2 |!3

() 5 M P R t

M N Y 1 > A S A A 1

R A T B F . T A I B t 1 V B i ; ' I " A 1 R A r B F ; i A 2 B O V i l K T A 2 R A I K F T A S

I . M G C F V F

T R N S A R S

5 Q I . S Q L S Q 1 . S Q L

R A T B E T A l B d V I H V I A l R A I ' B i : r A 2 B I J V H E I A 2

R A T B F T A I H O V B i : l A ! R A r i U ' . T A 2 H ( l V f i r . ! A 2 R A T B E T A 3 B O V I i r . T A !

Fig. 2. Comparison of three rat and bovine GABAa receptor /3 subunits. Numbering and symbols are as in Figure 1.

Ill |i; 113 pi |13 ill [12 fl3 111

Fig. 3. Regional brain distribution of (3 subunit mRNAs. Northern blots of poly (A) + RNA from rat (A) and calf (B) cortex (Cx), hippocampus (Hi), cerebellum (Cb) and total brain (Br) were probed using iSl, |82 and j33 subunit-specific oligonucleotides. Size markers indicated on the left represent 9.5, 7.5, 4.4 and 2.4 kb.

IpMGABA 3|iM GABA ItiM GABA 3liM GABA 1 tJM GABA 3mM GABA

Fig. 4. Expression of GABA^ receptor (3 subunits in Xenopus oocytes. Clamp potential was - 7 0 mV. Downward deflections reflect inward currents (see calibration bar). GABA-application (1 and 3 fiM for each subunit pair) is indicated by horizontal bars.

1667

S.Ymer et al.

( ~ 6 kb and ~ 2.5 kb), the smaller transcript being the major form in the rat cerebellum.

As observed for the a subunit mRNAs (M.Kohler and P.H.Seeburg, unpublished), the bovine brain contains substantially higher amounts of jS subunit mRNA than the rat brain. In particular, the cortex and cerebellum display high levels of i32 and jSS mRNAs. Two 131 mRNAs (13 and 4 kb) are observed which are of equal abundance in calf cortex, whereas the smaller transcript appears to be the major form in cerebellum; (32 mRNA is represented by a single 8 kb transcript, while (33 mRNA exists as 8 kb (major form) and 7 kb transcripts in cortex and cerebellum. Several transcript sizes originating from one gene have been observed previously and usually reflect the use of different polyadenylation sites to generate different-sized 3' flanking regions (Setzer etal, 1980; Tosi etal, 1981).

Three P subunits in the adrenal medulla Determination of the in vivo subunit composition of a single type of G A B A A receptor may help to resolve the role of multiple a and jS subunit polypeptides. The adrenal medullary chromaffin cells express electrophysiologically (Bormann and Clapham, 1985) and pharmacologically (Kataoka et al., 1984) well-characterized GABA^ receptors that might constitute a single population of receptors. To analyse these receptors in molecular terms we constructed a bovine adrenal medulla cDNA library and screened it at low stringency using as probes ^^P-labelled al (Levitan et al., 1988) and /SI (Schofield et al., 1987) subunit cDNAs as well as the degenerate oligonucleotide based on the conserved octameric peptide sequence in M2 (see above). Clones encoding a l , |S1, /32 and j83 subunits were obtained, but no al or a 3 subunit cDNAs were found. While these results do not prove the absence of a variants, they suggest that these subunits represent at best very minor species in the adrenal medulla. This is substantiated by the absence of al and a'i subunit specific hybridization signals on Northern blots of adrenal medullary RNA (not shown). However, the presence of three /3 subunits may indicate that chromaffin G A B A a receptors constitute a heterogeneous population.

All adrenal cDNA sequences were colinear with those of brain-derived cDNAs. Occasionally observed third-base substitutions in codons are probably a consequence of different RNA sources used in cDNA library construction. These results show that many of the genes encoding G A B A a receptor subunits are expressed both centrally and peripherally without the use of alternative splicing. Hence, alternate exon usage is not a major mechanism for generating diversity of GABA^ receptors. Expression in Xenopus oocytes We investigated whether the novel jS subunits could contribute to the formation of functional GABA^ receptors by expressing each rat ^ subunit in combination with the rat a l subunit (unpublished) in Xenopus oocytes. Expres-sion was achieved by nuclear injection (Voellmy and Rungger, 1982) of pairs of recombinant CDM8 vectors (Seed, 1987) in which the cloned cDNAs are under the transcriptional control of the cytomegalovirus promoter. This mode of expression is sensitive to small amounts of DNA and circumvents the need for the in vitro synthesis of RNA (Ballivet etal, 1988). Electrophysiological recordings (Figure 4) showed that either the /32 or polypeptide can be substituted for the jSl subunit to yield dose-dependent

GABA-evoked currents. These were always inhibited by bicuculline, enhanced by the barbiturate pentobarbital and had reversal potentials (Er) close to the chloride equilibrium potential (E^ —23 mV) of Xenopus oocytes (Dascal et al, 1984), indicating that each receptor forms GABA-gated chloride channels (not shown). Thus, substitution of either the /32 or ^3 subunit for the j31 subunit does not appear to change the qualitative properties of the expressed GABA^ receptors.

Discussion

Molecular cloning has revealed heterogeneity of the G A B A A receptor jS subunit. The assignment of the two novel subunits and |33 as true components of the G A B A a receptor is based on the observations that (i) (31, (31 and /33 subunits share a high degree of sequence identity, (ii) the (S2 subunit contains a peptide sequence obtained by chemical means from affinity-purified G A B A A receptor complex and (iii) all three subunits form GABA-responsive chloride channels when co-expressed with the a l subunit. The jS subunits are highly sequence-conserved and contain a larger intracellular domain than other subunits (Levitan et al., 1988; Pritchett et al, 1989). This domain in the P subunits contains a consensus site for cAMP-dependent phosphorylation by protein kinase A (Feramisco et al, 1980). The presence of this site strengthens our earlier hypothesis (Schofield et al., 1987) that the jS subunit provides a target for the cellular regulation of G A B A A receptor activity.

While molecular heterogeneity of the a subunit was anticipated, the existence of j8 subunit variants is unexpected. Northern analysis indicates that the novel (3 subunits far exceed the jSl subunit in abundance. This result provides an explanation for the previously noted imbalance of in situ hybridization signals for a and (3 subunit mRNA in regions of rat and bovine brain (Sequier et al., 1988; Siegel, 1988). In fact, evidence from this laboratory indicates that the different jS subunits are indeed expressed in distinct, possibly overlapping neuronal populations (B.D.Shivers, unpublished).

The increasing number of variants of G A B A A receptor a and (3 subunits poses the problem of determining the subunit composition of natur^ G A B A A receptors. As the analysis of a homogeneous G A B A A receptor population may provide one way of dissecting the subunit composition, we investigated the receptor in adrenal medullary chromaffin cells by molecular cloning. Besides finding only one a subunit we were surprised by the presence of all three jS subunits. Thus, either all (3 variants are part of the same receptor or the adrenal chromaffin G A B A A receptors are composed of different subtypes. Of these alternatives, the first is less likely considering the different expression levels of the three (3 subunit mRNAs and their distinct localization in the brain (unpublished). While GABAAreceptor diversity in chromaffin cells has yet to be substantiated by electro-physiology or pharmacology, the presence of several conductance states in the G A B A A receptor of adrenal chromaffin cells (Bormann and Clapham, 1985) might indicate such a heterogeneity. It will be of interest to examine other peripheral G A B A A receptor populations, such as those involved in the control of pituitary hormone release (Grandison and Guidotti, 1979; Racagni etal., 1979).

The role of the JS subunits in generating different G A B A A

1668

GABA;^ receptor /3 subuni ts

receptor subtypes remains to be established. Combinations of different o; and (3 subunits may define receptors that can be distinguished by pharmacology and channel properties, generating a greater diversity of GABA responses than would be achieved with fewer receptor subtypes. While the true extent of a and (3 subunit heterogeneity is unknown, recent cDNA cloning experiments in this laboratory have provided evidence for die existence of additional GABAA receptor a subunits and of novel subunits not of the a or ^ type (Pritchett et al, 1989). However, no additional jS subunit variants were identified.

Materials and methods Isolation of cDNA clones A XgtlO bovine brain cDNA library (Schofield et al., 1987) was screened with a 96-fold degenerate ^^P-labelled 23mer oligonucleotide encoding a conserved octameric peptide sequence in M2 of GABA^n receptor subunits, 5' AC(A,C)AC(A,T)GT(G,T)CT(A,C,G)AC(A,C)ATGAC(A,C)AC 3'. Only indicated third position choices were included. Known subunits were identified using a l , a2 , aS and /3 subunit-specific oligonucleotides (Levitan et al., 1988; and see below). cDNAs hybridizing to the 23mer but not to the subunit-specific oligonucleotides were sequenced in XglO or after sub-

: cloning into M13 vectors (Vieira and Messing, 1987) by the chain termination method (Sanger etal., 1977). Sequencing reactions using the 23mer oligonucleotide were performed with 0.5 /iM primer and reactions were at 55°C when recombinant X DNA was used as template. A rat forebrain cDNA library was screened with the following ^^P-labelled subunit-specific oligonucleotides complementary to sequences encoding part of the large intracellular domain of the three subunits: /SI; 75mer (5' TCCCACGCCCGTGAGCACTTCAGAGCCGCTCGTCTCGTTCCT-GATCTCCAGGGTACTGAGGAGAATGTTGCCGTG 3'); |S2, 60mer (5' TTTCCGATACTGGATGCTGGAGGCATCATAGGCCAGCATTGT-GCTCCTTGGGTCTCCAAG 3'); |83, 60mer (5' TCTTGCTGAATTC-CGGGTATCACCAACGCCGCCGGCAACCTCGTTCATCTCATTG-TGAAC 3'). The longest cDNA clones were completely sequenced. Furthermore, a bovine adrenal medulla cDNA library was constructed in XgtlO by standard methods (Huynh et al., 1985) and screened using as probes both the degenerate 23mer oligonucleotide and two internally labelled EcoRI fragments of cloned bovine jSl and a2 subunit encoding cDNAs. The /31 cDNA fragment comprised nucleotides 1 - 7 2 6 (Schofield et al., 1987) and the a2 cDNA fragment contained nucleotides 133 — 1755 (Levitan et al., 1988). Sequence analysis was as described above.

Northern blot analysis RNA was isolated by published methods (Chomczynski and Sacchi, 1987) from three brain regions of 8-month-old calf and young adult rats (200 g). Poly(A)+ RNA was prepared using oligo(dT)-cellulose chromatography. For Northern analysis, RNA (3 fig) was electrophoresed in 1.2% form-aldehyde-containing agarose gels and blotted onto nitrocellulose. These blots were hybridized to subunit-specific ^^P-labelled (sp. act. 10^ c.p.m./pmol) oligonucleotides in 40% formamide at 42''C, washed in 2 X SSC, 0.1% SDS at 55 °C and exposed to X-ray film, using an intensifying screen at - 8 0 ° C for 5 days (bovine) or 14 days (rat). For the bovine Northerns the oligonucleotide sequences are listed above. For the rat experiment, the /St-and /32-specific oligonucleotides were 5' GTAAGAGAGAAGCCCCAA-ACTCACTTAGTCTGTCTGCGATTTTGTACTGTC 3' and 5' AGAGA-GGAGATCCACCCAGTGCAGTAATTC 3', while the same /33 probe was used.

Expression in Xenopus oocytes The rat a l (unpublished), jSl, /32 and /33 subunit-encoding cDNAs were cloned into the CDM8 vector (Invitrogen, San Diego, CA). Either )Orol or fcoRI fragments containing the entire coding sequences were converted to blunt ends and ligated with adaptor sequences to generate BstXl cohesive termini. The adaptor sequences were 5' CGAATTCAGAGAACA 3' and 5' CTCTGAATTCG 3'. The terminally modified cDNAs were then used to replace the stuffer fragment in CDM8 (Seed, 1987). Orientations of subcloned cDNAs relative to the vector-carried cytomegalovirus promoter were determined by restriction analysis. The nuclei of oocytes were injected (Voellmy and Rungger, 1982) with these constructs (10 nl, 350 pg of each expression plasmid). After incubation of injected oocytes at 19°C for 3 - 6 days, currents were recorded in a conventional two-electrode voltage clamp in normal frog Ringer solution (115 mM NaCl, 2.5 mM KCl, 1.8 mM

CaCl2, 10 mM Hepes, pH 7.2) before and during superfiision with frog Ringer containing different concentrations of GABA. Bicuculline was used at 10 /^M and pentobarbital at 5 /iM.

Acknowledgements S.Y. and P.R.S. are equal contributors. We are indebted to Dr Brenda Shivers for her expert help in dissecting the brains and preparing the animal tissues used in this study and thank Hildegard Kluding for RNA isolation. A.D. acknowledges the expert training in nuclear oocyte injection technique by Dr Duri Rungger. We grateftjlly acknowledge the active interest of Dr Bert Sakmann and his kind support of this project. We further thank Jutta Rami for her efficient help in preparing this manuscript. This work was supported by grants from the DFG and BMFT to P.H.S.

References Ballivet,M., Nef,P., Couturier,S., Rungger,D., Bader.C.B., Bertrand,D.

and Cooper,E. (1988) Neuron, 1, 847-852 . Bormann,J. and Clapham,D.E. (1985) Proc. Natl. Acad. Sci. USA, 82,

2168-2172. Braestrup,C. and Nielsen,M.J. (1981) J. Neurochem., 37, 333-341 . Chomczynski,P. and Sacchi,N. (1987) Anal. Biochem., 162, 156-159. Cooper,S.J., Karkham,T.C. and Estall,L.B. (1987) Trends Pharmacol. Sci.,

8, 180-184. Criado,M., Sarin,V., Fox,J.L. and Lindstrom,J. (1986) Biochemistry,

25, 2839-2846. Dascal,N., Landau,E.M .and Lass,Y. (1984)7. Physiol, 352, 551-574 . Feramisco,J.R., Glass,D.B. and Krebs,E.G. (1980)7. Biol. Chem., 255,

4240-4245. Fuchs,K., M6hler,H. and Sieghart,W. Neurosci. Lett., 90, 314-319. Grandison,L. and Guidotti,A. (1979) Endocrinology, 105, 754-759 . Grenningloh,G., Rienitz,A., Schmitt,B., Methfessel,C,. Zensen,M.,

Beyreuther,K., Gundelfinger,E.D. and Betz,H. (1987a) Nature, 328, 215-220 .

Grenningloh,G., Gundelfinger,E., Schmitt,B., Betz,H., Darlison,M.G., Bamard,E.A., Schofield,P.R. and Seeburg,P.H. {\9?nh) Nature, 330, 2 5 - 2 6 .

Haring,P., Stahli,C., Schoch,P., Takacs,B., Staehelin,T. and M6hler,H. (1985) Proc. Natl. Acad. Sci. USA, 82, 4837-4841.

Huynh,T.V,. Young,R.R. and Davis,R.W. (1985) In Glover,D.M. (ed.), DNA Cloning Techniques: A Practical Approach. IRL Press, Oxford, Vol. 1, pp. 4 9 - 7 8 .

Imoto,K., Methfessel,C., Sakmann,B., Mishina,M., Mori,Y., Konno,T., Fukuda,K., Kurasaki,M., Bujo,H., Fujita,Y. and Numa,S. {19%?,) Nature, 324, 670 - 674.

Kataoka,Y., Gutman,Y., Guidotti,A., Panula,P., Wroblewski,J. and Cosenza-Murphy,D. (1984) Proc. Natl. Acad. Sci. USA, 81, 3218-3222.

Leonard,R.J., Labarca,C.G., Chamet,P., Davidson,N. and Lester,H.H. (1988) Science, 242, 1578-1581.

Levitan,E., Schofield,P.R., Burt,D.R., Rhee.L.M., Wisden,W., K6hler,M., Fujita,N., Rodriguez,H.F., Stephenson,A., Darlison,M.G., Barnard,E.A. and Seeburg,P.H. (1988) Nature, 335, 7 6 - 7 9 .

Mamalaki,C., Stephenson,F.A. and Bamard,E.A. (1987) EMBOJ., 6, 561-565 .

M6hler,H., Battersby,M.K. and Richards,J.G. (1980) Proc. Natl. Acad. Sci. USA, 77, 1666-1670.

01sen,R.W. and Venter,C.J. (eds) (1986) Benzodiazepine/GABA Receptors and Chloride Channels: Structural and Functional Properties. Alan R.Liss, New York.

Pritchett,D., Sontheimer,H., Shivers,B.D., Ymer,S., Kettenmann,H., Schofield,P.R. and Seeburg,P.H. (1989) Nature, 338, 582-585.

Racagni,G., Apud,J.A., Locatelli,V., Cocchi,D., Nistico,G., Di Giorgio,R.M. and Muller,E.E. (1979) Nature, 281, 575-578 .

Sanger,F., Nicklen,S. and Coulson,A.R. (1977) Proc. Natl. Acad. Sci. USA, 74, 5463-5467.

Schofield,P.R,. Darlison,M.G., Fujita,N., Burt,D.R., Stephenson,F.A., Rodriguez,H., Rhee,L.M., Ramachandran,J., Reale,V., Glencorse,T.A, Seeburg,P.H. and Bamard,E.A. (1987) Nature, 328, 221-227 .

Schofield,P.R., Pritchett,D.B., Sontheimer,H., Kettenmann.H. and Seeburg,P.H. (1989) FEBS Lett., 244, 361-364 .

Seed,B. (1987) Nature, 329, 840-842. Sequier,J.M., Richards,J.G., Malherbe,P., Price,G.W., Mathews,S. and

Mohler.H. (1988) Proc. Natl. Acad. Sci. USA, 85, 7815-7819. Setzer,D.R., McGrogan,M., Nunberg,J.H. and Schimke,R.T. (1980) Cell,

22, 361-370.

1 6 6 9

S.Ymer et al.

Siegel,R.E. (1988) Neuron, 1, 5 7 9 - 5 8 4 . Sieghart.W., Mayer,A. and Dexler,G. (1983) Eur. J. Pharmacol., 88,

2 9 1 - 2 9 9 . Squires,R.F., Benson,D.I., Braestrup,C., Coupet,J., Klepner,C.A., Myers,V.

and Beer,B. (1979) Pharmacol. Biochem. Behav., 10, 8 2 5 - 8 3 0 . Stephenson,F.A. (1988) Biochem. J., 249, 2 1 - 3 2 . Tosi,M., Young,R.A., Hagenbuchle,0. and Schibler,U. (1981) Nucleic

Acids Res.,9,13n-2m. Unnerstall,J.R., Kuhar,M.J., Niehoff,D.L. and Palacios,J.M. (1981) J.

Pharmacol. Exp. Ther., 218, 797 - 804. Vieira,J. and Messing,!. (1987) Methods Enzymol, 153, 3 - 1 1 . Voellmy,R. and Rungger,D. (1982) Proc. Natl Acad. Sci. USA, 79,

1776-1780.

Von Heijne,G. (1986) Nucleic Acids Res., 14, 4683-4690 .

Received on March 3, 1989

1670

Neuron, Vol. 2, 1491-1497, M jy , 1989, Copyright © 1989 by Cell Press

Functional Chloride Channels by Mammalianr^ Cell Expression of Rat Glycine Receptor Subunit

H. Sontheimer,»+ C.-M. Becker,+ * D. B. Pritchett,* P. R. Schofield,*§ G. Grenningloh,* H. Kettenmann,* H. Betz,^ P. H. Seeburg* * Institute for Neurobiology Im Neuenheimer Feld 364 •Center for Molecular Biology im Neuenheimer Feld 282 University of Heidelberg D-6900 Heidelberg 1 Federal Republic of Germany

Summary

Cultured human cells were transfected with cloned rat glycine receptor (GlyR) 48 kd subunit cDNA. In these cells glycine elicited large chloride currents (up to 1.5 nA), which were blocked by nanomolar concentrations of strychnine. However, no corresponding high-affinity binding of [^HJstrychnine was detected in membrane preparations of the transfected cells. Analysis by mono-clonal antibodies specific for the 48 kd subunit revealed high expression levels of this membrane protein. After solubilization, the 48 kd subunit behaved as a macro-molecular complex when analyzed by sucrose density centrifugation. Approximately 50% of the solubilized complex bound specifically to a 2-aminostrychnine af-finity column, indicating the existence of low-affinity antagonist binding sites on most of the expressed GlyR protein. Thus, the 48 kd strychnine binding subunit ef-ficiently assembles into high molecular weight com-plexes, resembling the native spinal cord GlyR. How-ever, formation of functional receptor channels of high affinity for strychnine occurs with low efficiency.

Introduction

The inhibitory glycine receptor (GlyR) from spinal cord has been extensively characterized by electrophysiol-ogy, biochemical methods, and pharmacology (reviewed by Betz, 1987; Betz and Becker, 1988). In this receptor glycine gates an intrinsic chloride channel whose activ-ity stabilizes the resting potential of the neuronal mem-brane. Upon affinity purification, the GlyR contains two types of subunits (48 kd and 58 kd) in addition to a 93 kd receptor-associated peripheral membrane protein (Pfeiffer et al., 1982; Graham et al., 1985; Becker et al., 1986; Schmitt et al., 1987). Recently, cloned cDNA en-coding the 48 kd subunit was characterized, and the en-coded protein was predicted to contain both the glycine and the strychnine binding sites of the receptor (Gren-ningloh etal., 1987a). Strychnine antagonizes glycine re-sponses with high affinity (Curtis et a!., 1968; Snyder and

' Both authors have contributed equally to this work. 5Present address: Pacific Biotechnology Ltd., 74 McLachlart Avi -nue, Rushcutters Bay, N S W 2011, Australia.

Bennett, 1976) and is incorporated into the 48 kd sub-unit upon UV illumination (Graham et al., 1981, 1983).

Sequence analysis of the cloned cDNA revealed a close homology to subunits of the GABAA receptor, which is ubiquitously expressed in brain and contains, like the GlyR, an intrinsic chloride channel (Grenningloh et al., 1987a, 1987b; Schofield et al., 1987). In addition, the 48 kd subunit shows distant homology to nicotinic acetylcholine receptor proteins. Based on structural considerations, it has been proposed that all ligand-gated ion channels are members of a protein family (Grenningloh et al., 1987a; Schofield et al., 1987) that form neurotransmitter-gated channels by the assembly of several subunits. Each subunit is thought to cross the lipid bilayer four times and to contribute a homologous stretch of transmembrane region to the formation of an ion channel that opens in response to neurotransmitter binding.

The assembly of neurotransmitter receptor subunits into gated ion channels is only poorly understood. The nicotinic acetylcholine receptor of the muscle endplate has a pentameric structure (a2|3Y5) containing four different subunits (Hucho, 1986). Recently, cross-linking studies have indicated a similar pentameric arrange-ment (a3P2) for the channel of the GlyR containing the 48 kd and 58 kd subunits (Langosch et al., 1988). How-ever, a tetrameric structure (a2P2) has been proposed for the GABAA receptor (Mamalaki et al., 1987). To complicate matters, recent expression studies using cloned receptor subunit cDNAs have suggested that cat-ionic and anionic channels can be formed by self-assembly of single subunits (Boulter et al., 1987; Pritch-ettetal., 1988; Blairetal., 1988; Schmieden etal., 1989). To investigate the assembly of the 48 kd subunit into the GlyR, we engineered the cloned rat cDNA for expres-sion in cultured mammalian cells and now report the formation of homomeric glycine-gated chloride chan-nels. Biochemical analysis reveals that the assembly of functional 48 kd subunit GlyR is an inefficient process, suggesting that formation of heterooligomeric com-plexes may be the preferred pathway of channel assem-bly in vivo.

Results

Expression of the 48 kd GlyR Subunit cDNA To express the strychnine binding subunit of the rat GlyR, a full-length cDNA was constructed using two cDNA clones, GRl-6-1 and GRl-6 (Grenningloh et al., 1987a), that together encode the entire mature protein as well as 6 amino acid residues of the signal peptide (Figure 1, top). To ensure correct expression of the en-coded polypeptide, signal peptide encoding and Kozak consensus initiation sequences was added using syn-thetic DNA. The semisynthetic, full-length cDNA was then subcloned into the eukaryotic expression vector pClS2 (Pritchett et al., 1988). The resultant construct was

Neuron 1492

a) cDNA clones

E I H E GRl-6-1

b) 4-w8y ligation

pSPM SP6 H P E E| H H

c) Insertion of signal peptide and initiation lequences

Hindlll PstI Ncol EcoRI " G K S P G L L D Y L H A W T L F E F P S K E A O A ^ . .

5 AAGCTTGGGCTGCA GCCGCCACCATGGGCAAAAGTCCGGGTCTTCTAGACTACCTTTGGGCCTGGACCCTCTTTG AATTCCCTTCCAAAGAGGCTGACGCTGCCCGC 3" 3 TTCGAACCCG ACGTCGGCGGTGGTACCCGTTTTCAGGCCCAGAAGATCTGATGGAAACCCGGACCTGGGAGAAACTTAA GGGAAGGTTTCTCCGACTGCGACGGGCG 5"

d) Subcloning into eukaryotic expression vector

pCIS2 CHV PiE Hp (P) N E J H i I I I I

Pv Hp pA*

s c TC NTC

•A8K

Figure 1. (Top) Construction of an expressabie glycine receptor cDNA. (a) The rat glycine receptor 48 kd subunit cDNA clones CR1-6-1 and CR1-6 (Grenningloh et a!., 1987) were used, (b) A four part ligation was performed in which the 67 bp EcoRI-Hindlll, 450 bp Hindlll (both from GRl-6-1), and 950 bp Hindlll-EcoRI (from GR1-6) fragments were iigated into cleaved plasmid pSP64 DNA (Melton et al., 1984). Correct order and orientation of the cloned DNA fragments were determined by restriction mapping and confirmed by DNA sequence analysis. This construct contained the entire cDNA sequences encoding the mature poly'peptide as well as 6 amino acid residues of the signal peptide. Three additional amino acids (Clu-Phe-Pro) were encoded by the EcoR! linker sequence used in cDNA li-brary construction, (c) For construction of the synthetic signal peptide sequence, the plas-mid was partially digested with EcoRI. The linear plasmid was electroeluted and recut with Pstl. A synthetic DNA fragment containing Pstl and EcoRI sites was Iigated into the EcoRI-and Pstl-cleaved recombinant plasmid. The synthetic DNA contained a signal peptide-

encoding sequence based on that observed in the al subunit of the GABAa receptor (Schofield et al., 1987). Additionally, the oligonucleo-tide sequence included the consensus eukaryotic initiation sequence 5'-CCACCATGG-3' (Kozak, 1986). Restriction sites are as follows: E, EcoRI; H, Hindlll; S, Sad; P, Pstl; Pv, Pvuil; N, Ncol. The asterisk indicates the stop codon; SP6, the position of the SP6 polymerase promoter; and the arrow, the signal peptide cleavage site. The synthetic DNA sequences are shown, with their coding potential indicated above. The gaps within the Pstl and EcoRI restriction sites show the 5' and 3' ends of the oligonucleotide sequences used. The Pstl site indicated in brackets did not form as predicted; a single nucleotide error was revealed by sequence analysis. (Bottom) Western blot analysis of membranes from transfected cells. Crude membranes from rat spinal cord (80 ^g of protein; lane 1) and from 48 kd subunit cDNA transfected (lane 2) and nontransfected (lane 3) cells (20 ^g of protein each) were electrophoresed on a 10% SDS-polyacrylamide gel, blotted onto nitrocellulose, and immunostained using MAb 4a.

92.5 K — 66.2 K —

^5.0 K —

31.0 K —

21.5 K —

used to transfect h u m a n embryon i c k idney cells. H igh levels of transient transfection w e r e ind icated by West-ern blot analysis (F igure 1, bottona) using a m o n o c l o n a l ant ibody, M A b 4a, that recognizes the 48 kd G l y R subunit (Pfeiffer et al., 1984). N o cor respond ing ant igen was de tec ted in untransfected cells.

Transfected cel ls w e r e ana lyzed for glycine-el ic i ted c o n d u c t a n c e changes using the whole-ce l l , patch-c l a m p m o d e . in 14 out of 25 transfected cel ls tested, ap-pl icat ion of g l yc ine p roduced a current response (Figure 2a). N o such response was seen in nontransfected cel ls

or in cel ls t ransfected w i th vector a lone . S imi lar ly , no re-sponse to g l y c i ne w a s detec ted w h e n the expression vec to r c o n t a i n e d c D N A e n c o d i n g G A B A receptor sub-units; a response to G A B A was seen (data not shown ) . A t a ho ld ing potent ia l of - 6 0 mV, cu r ren t responses w e r e e l ic i ted by g l yc ine concent ra t ions as l o w as 25 \iM (Figure 2a). M a x i m a l responses (up to 1.5 nA) w e r e seen at 1 m M g lyc ine , and half-maximal cur rents w e r e ob-served at g l y c i ne concen t ra t ions of 100 ^iM. Desensit iza-tion was seen at g l yc ine concen t ra t ions a b o v e 100 i i M , and appa ren t saturat ion v^as r eached at approx imate l y

Homomeric Glycine Receptors in Mammalian Cells 1493

A

500 pA

V

25

50

75

100

250

500

1000 GLY (pMl

I

IpAl

1000

100

10

B

10 25 50 100 2S0 1000

GLYlpMl

Figure 2. Analysis of Glycine-lnduced Cur-rents in Transfected Cells

(a) At a holding membrane potential of - 6 0 mV, membrane currents were recorded from a cell transfected with 48 kd subunit cDNA using the patch-clamp technique in the whole-cell recording configuration (Hamill et al., 1981). Glycine was applied, as indi-cated by the bar, at the given concentrations. A dose-dependent inward current was ob-served.

(b) Dose-response curve of glycine-induced peak currents (1) for the experiment shown in (a), log-log scale. The slope of the curve is approximately 2, suggesting at least 2 bind-ing sites for^glycine. (c) Currents (I) elicited by 100 mM glycine were determined at different holding poten-tials as described in Experimental Proce-dures and plotted as a function of membrane potential (Vm). Currents reversed close to the cloride equilibrium potential (0 mV), quite distinct from the sodium {>80 mV) and potas-sium { < - 8 0 mV) equilibrium potential.

250 m M (Figure 2b). Analysis of the dose-response curve revealed cooperativity in glycine-induced channel activation. At low nondesensitizing doses of glycine, a Hill coefficient of about n = 2 was obtained, a value close to n = 2.7 reported for spinal cord glycine recep-tors (Barker, 1985). Also, current reversal occurred at 0 mV (Figure 2c), which corresponds to the chloride equi-l ibrium potential in these experiments (Bormahn et al., 1987).

Pharmacology of the Glycine Response Glycine-gated currents in transfected cells were blocked completely by strychnine, a selective GlyR antagonist (Figure 3a). This effect was rapidly reversed upon washout of the alkaloid. Strychnine competi t ion of the glycine response over a range of antagonist concentra-tions (1 nM to 1 tiM) showed half-maximal inhibit ion around 20 nM (Figure 3b). From these data and the gly-cine response curve shown in Figure 2b, an apparent Kj value of 10 nM was calculated; this is in excellent agree-ment with the Kd of pH]strychnine binding to spinal cord GlyR (Young and Snyder, 1973; Pfeiffer et al., 1982; Betz and Becker, 1988). Similarly, picrotoxin reversibly blocked glycine-gated currents almost completely, and this inhibit ion was reversed upon washout (Figure 3a). Picrotoxin is thought to bind within the gated chloride channel and is commonly used to block GABA^ recep-tors (Olsen, 1981). No contr ibut ion from GABA-gated channels was.detectable here, since bicuculline, a spe-cific G A B A a receptor antagonist, failed to affect the glycine-induced currents. These results indicate that ex-pression of the 48 kd subunit in a human ceil line produces glycine-gated chloride channels that display the characteristic pharmacology of the spinal cord GlyR.

The 48 kd Subunit Assembles into High Molecular Weight Complexes Western blot analysis of membrane fractions of trans-fected cells using MAb 4a revealed an antigen of the same apparent molecular weight as the 48 kd GlyR subunit in membrane preparations from rat spinal cord (see Figure 1). By quantitative dot receptor immunoas-say (DORA; see Becker et al., 1989), the concentration of the 4'8 kd antigen in m'embrane preparations from transfected cells was found to be similar or even higher than that in spinal cord membranes (Table 1). Only min-ute amounts of GlyR antigen were detectable in cyto-solic cell fractions (data not shown). Thus, the 48 kd subunit expressed upon transfection behaves similarly to the GlyR found in spinal cord homogenates (Becker et al., 1989).

To determine whether the 48 kd subunit expressed in vitro assembles into macromolecular complexes, deter-gent extracts of membranes from both transfected cells and spinal cord were subjected to sucrose density gra-dient centrifugation. Fractions were analyzed by DORA using MAbs la, 2b, and 4a, which recognize the 48 kd subunit (Pfeiffer et al., 1984). In accord with earlier ob-servations on spinal cord preparations (Becker et al., 1989), solubilized GlyR migrated as a single immuno-reactive peak wi th an apparent sedimentation constant of 7.4 S (Figure 4). Most of the immunoreactive material found in transfected cell extracts also sedimented wi th a similar S value (7.0 S). However, some heterogene-ously distributed immunoreactivity was also observed (Figure 4b). In particular, large amounts of antigen sedi-mented to the bottom fractions of the gradient, probably as a result of aggregate formation. These observations in-dicate that the expressed 48 kd subunit assembles into

Neuron 1494

\

•Slry V

V •Pic

V-• Bk:

\ •

g Inhibi t ion |%

100

5 10 100

Stry lnMl

1000

Figure 3. Pharmacology of the Clyc ine- lnduced Current in Trans-fected Cells (a) Appl icat ion of glycine (100 | iM) to a cell expressing the 48 kd subunit c D N A , voltage-clamped at - 6 0 mV, gave rise to an inward current (left traces). Appl icat ion bar denotes 30 s. These currents were compared w i th those elicited in the presence of strychnine (upper middle trace 100 | iM), picrotoxin (center trace, 100 ^ M ) , and bicucul l ine (lower middle trace, 50 nM). After a 5 min washout, all agonist responses were similar to control currents (right traces). (b) The effect of strychnine on glycine-induced currents was tested as described in (a), and the reduction in current (in percent of con-trol) was plotted versus the strychnine (Stry) concentrat ion.

high molecular weight complexes of probably variable subunit stoichiorrietry.

Another significant difference emerged from a quan-titative comparison of MAb binding to membrane prep-arations from both spinal cord and transfected cells. The ratio of immunoreactivities for MAbs la and 4a in trans-fected cells exceeded that obtained with spinal cord membranes by a factor of 2.0 (Table 1). As both antibod-ies recognize the 48 kd subunit (Pfeiffer et al., 1984;

Becker et al., 1988), the stoichiometry or accessibility of antigenic epitopes in receptors of transfected cells must be different from that in the spinal cord ClyR.

Antagonist Binding of the 48 kd Subunit GlyR In contrast to the high content of 48 kd protein revealed by antibody binding, no high-affinity [^H]strychnine binding sites were detectable in membrane fractions from transfected cells (Table 1), even for ligand concen-trations up to 100 nM (data not shown). Assuming a spe-cific binding capacity as observed in spinal cord mem-branes, the sensitivity of our assay should have allowed the detection of <5% of the bindingMctivity anticipated from the cellular antigen contents. Thus, only a very mi-nor fraction of the 48 kd subunit was present in a high-affinity ligand binding conformation.

Recently, a GiyR subtype of low affinity for strychnine containing a ligand binding subunit of 49 kd has been found in spinal cord of neonatal rats (Becker et al., 1988). This neonatal ClyR is not detectable by standard [^H]strychnine binding assays, but binds to highly sub-stituted 2-aminostrychnine agarose and can be specifi-cally eluted by the competing agonist glycine (Becker et al., 1988). To determine whether the expressed 48 kd subunit receptor displays similar antagonist binding properties, detergent extracts of membrane fractions from transfected cells were subjected to affinity chroma-tography on a 2-aminostrychnine agarose column. The 48 kd subunit in flow through, wash fractions, and gly-cine eluate was then quantified by DORA using MAb 4a. Under these conditions, approximately 50% of the re-covered antigen was detected in the glycine eluate (data not shown). Thus, a major fraction of the 48 kd protein in the transfected cells exhibited glycine-displaceable, low-affinity strychnine binding. Apparently most of the cell-expressed 48 kd subunit GlyR does not acquire a high-affinity antagonist binding conformation, even though binding sites for strychnine and glycine were de-tected by affinity chromatography.

Discussion

Functional glycine-gated chloride channels were gener-ated by mammalian cell expression of an engineered cDNA encoding a single subunit of the rat ClyR. These channels closely resemble those of spinal cord GlyR in both chloride permeability and pharmacological prop-

Table 1. (^HjStrychnine and M A b Binding to Membranes f rom Spinal Cord and Cul tured Cells

Membrane Preparation

[^HJStrychnine Bound (pmol/mg)

Immunoreact iv i ty w i t h

M A b l a (U/ j ig) M A b 4a (U/t ig) Ratio of M A b Bind ing

Spinal cord Transfected cells Nontransfected cells

2.37 ± 0.04 <0.10 <0.04

13.1 ± 2.9 66.6 ± 10.6 35.8 ± 1.9 88.4 ± 4.6 <2 .2 <0 .9

0.2 0.4

Binding of [^H]strychnine (24 nM) to crude membr. ine fractions f rom rat spinal cord and f rom transfected and nontransfected cells was determined by f i l tration assay. Reactivity of membranes w i i h MAbs la and 4a was quant i f ied by DORA and is displayed as arbi t rary units per ^ g of protein. The ratio (MAb l a : M A b 4a) of ant ibody b ind ing is der ived f rom these reactivit ies. The exper iment was repealed tw ice wi th similar results, except for some variai ion in ClyR antigen contents resulting; f rom di f ferent t ransfect ion eff ic iencies.


5 1 0 1 5 2 0 2 5 3 0 f r a c t i o n n u m b e r

Figure 4. Sucrose Density Gradient Centrifugation of Detergent Ex-tracts from Rat Spinal Cord and CellsTransfected with 48 kd Subunit cDNA

(A) Extract from rat spinal cord. (8) Extract for cells transfected with 48 kd subunit cDNA. Sucrose gradients ( 5 % - 2 0 7 o ) underlayered with a 6 0 7 o sucrose cushion were fractionated, and fractions were analyzed by DORA using the 48 kd subunit-specific MAbs la, 2b, and 4a. Immunoreactivities were normalized to peak values at 7.0-7.4 S. Results are only shown for MAb 4 (filled squares); very similar profiles were obtained with the other MAbs. Sucrose con-centrations determined refractometrically are indicated for both gradients (open squares). The position of marker enzyme activities is indicated (filled triangles) for 3-galactosidase (15.9 S), catalase (11.3 S), aldolase (7.4 S), lactate dehydrogenase (7.0 S), malate de-hydrogenase (4.3 S), and cytochrome C (2,1 S) (from bottom, left, to top of gradient).

erties (Bormann et al., 1987; Bormann, 1988). In particu-lar, glycine responses in transfected ceils were sensitive to the antagonistic alkaloids strychnine and picrotoxin at concentrations that also block GlyR channel activity in spinal cord and RNA-injected oocytes (Schmieden et al., 1989). For strychnine, half-maximal inhibition was ob-served at about 20 nM, a value that is in excellent agree-ment with the K d determined in [^H]strychnine binding studies using membrane fractions and purified receptor from mammalian spinal cord (reviewed by Betz and Becker, 1988). In other words, the functional GlyR chan-nels detected in the transfected cells possess high-affinity antagonist binding properties indistinguishable from those characterizing the spinal cord receptor.

The quarternary structure of the GlyR channels de-tected in our electrophysiological experiments is not known, but probably corresponds to oligomeric assem-blies of the 48 kd subunit. Different observations sup-port this interpretation. First, our sucrose centrifugation experiments showed that a considerable portion of the 48 kd antigen displayed a sedimentation behavior com-parable to that of the pentameric "core" structure of the spinal cord GlyR (Langosch et al., 1988). Second, GlyR channels of comparable char-icteristics are also seen in oocytes injected with 48 kd subunit RNA (Schmieden et

al., 1989), making contributions of cellular polypeptides to channel formation unlikely. Third, functional ligand-gated channels have been demonstrated upon expres-sion of single subunits of other neurotransmitter recep-tors, e.g., the a subunit of a neuronal nicotinic AChR (Boulter et al., 1987) and the a and P subunits of the GABAA receptor (Pritchett et al., 1988; Blair et al., 1988). The capacity to form homomeric receptor chan-nels therefore represents a newly emergent general property of the ligand binding subunits of such receptor proteins. This property can be exploited to classify newly cloned cDNAs encoding receptor subunits and may re-flect the proposed evo!uti6nary origin of gated ion chan-nels by single polypeptide assembly.

Assembly of functional channels from single subunits occurs with different efficiencies for the receptor sys-tems analyzed to date, possibly depending on their de-gree of complexity. Consonant with this notion, in trans-fected cells the 48 kd subunit GlyR showed current responses more than 10-fold higher than those observed with homomeric GABA^ receptors (Pritchett et al., 1988). Similarly, oocyte injection of GlyR 48 kd subunit RNA resulted in agonist-elicited currents exceeding those reported for single nicotinic AChR and GABA^ receptor subunit expression by a factor of >100.(Boulter et al., 1987; Blair et al., 1988; Schmieden et al., 1989). GlyRs are thought to precede the GABAA receptor in evolution (Betz and Becker, 1988). It could therefore be argued from the high responses of 48 kd subunit GlyR that this type of receptor forms in vivo. However, homo-meric assembly is highly inefficient as evidenced by bio-chemical analysis of the transfected cells. Although Western blots and quantitative immunoassays revealed high expression levels of the 48 kd subunit and high mo-lecular weight complexes were seen on sucrose gra-dients, no high-affinity [^Hlstrychnine binding could be detected in membrane preparations of the transfected cells. These data contrast with the presence of strych-nine-sensitive, glycine-gated chloride channels on cell membranes and argue that >95% of the 48 kd subunit is not assembled correctly to form functional channels. In view of the prominent glycine responses seen in the transfected cells, this conclusion may seem surprising. However, assuming single-channel conductance values as reported for GlyR in spinal cord cultures (approxi-mately 28 pS; see Bormann et al., 1987), <1000 open channels could account for the largest currents ob-served. This number is certainly lower than that of func-tional receptor channels due to parameters such as open probabilities and desensitization, but its magni-tude illustrates the congruency between the elec-trophysiological and biochemical data.

Although only a minor portion of the expressed 48 kd subunit was recovered in functional cell surface GlyR, most of the antigen was assembled into rapidly sedi-menting material and displayed low-affinity antagonist binding as revealed by 2-aminostrychnine affinity chro-matography. This is reminiscent of the properties of the neonatal GlyR isoform, which, however, is a different protein with distinct immunological characteristics and

Neuron 1496

sedimentation behavior (Becker et al., 1988). Further-more, the GlyR antigens in transfected cells and spinal cord membranes differ in relative MAb binding, indicat-ing conformational alterations in the expressed receptor protein. Thus, posttranslationai events required for as-sembly of functional channels may not occur efficiently in the transfected cells. Alternatively, cell surface incor-poration may be delayed compared with neurons, lead-ing to an accumulation of oligomers in intracellular membranes. Indeed, immunocytochemical experiments on transfected cells indicate that GlyR antigen is present both intrac^llularly and on the cell surface (B. D. Shivers, H. Betz j0''hd R H. Seeburg, unpablished data). Another possibility is that the comparatively low efficiency of homomeric GlyR channel formation reflects the lack of other required subunits, e.g., the 58 kd polypeptide, in fact, wherever investigated (i.e., for nicotinic AChR and GABAa receptor subunits), heteropolymeric receptors are formed much more efficiently in both Xenopus oo-cytes and mammalian cells (Mishina et al., 1984; Deneris et al., 1988; Levitan et al., 1988; Pritchett et al., 1988). Our data are therefore consistent with the in-terpretation that assembly of receptors from different subunits corresponds to the favored pathway of channel biogenesis in vivo.

Experimental Procedures

Transient Expression in Mammalian Cells Human embryonic kidney cells (293) were transfected at high effi-ciencies with recombinant expression vector DN A using a modified CaP04 precipitation technique (Chen and 01<ayama, 1987), as de-scribed earlier (Pritchett et al., 1988). The functional features of the vector are described by Eaton et al. (1986). Cells were used 48 hr after transfection for biochemical analysis or electrophysiology.

Electrophysiology For recording membrane currents, cells on coverslips were trans-ferred to the stage of an inverted microscope and maintained at about 25°C in a recording chamber that was continuously per-fused, permitting rapid application of transmitters and drugs. The standard salt solution contained 116.0 mM NaCI, 11.1 mM glucose, 26.2 mM NaHCOa, 5-4 mM KCl, 1.8 mM CaClj, and 0.8 mM MgClj, buffered at pH 7.2 with 5 mM HEPES. Glycine, strychnine, bicucullin, and picrotoxin were added to the standard salt solution at the concentrations indicated. All voltage-clamp recordings were made with an EPC-7 patch-clamp amplifier (List Electronics, Darm-stadt, FRG) using the tight-seat, patch-clamp technique in the whole-cell recording configuration (Hamill etal., 1981). Recording pipettes with resistances ranging from 2 to 5 MQ were filled with a solution containing 130 mM CsCl, 1 mM MgCU, 0.5 mM CaC^, 5 mM EGTA, 10 mM HEPES. Calcium activity was calculated to be approximately 11 nM. To determine the reversal potential of glycine-induced currents, cells were clamped briefly (200 ms) to -120, -90, -30, 0, 30, and 60 mV, in the presence and absence of the transmitter as described by Sontheimer et al. (1988). The transmitter-induced current was obtained as the difference of peak currents in the absence and presence of glycine and plotted against the clamp potentials.

Preparation of Membranes and Detergent Extracts Membrane fractions and detergent extracts from rat spinal cord and transfected cells were prepared as described (Becker et al., 1989). Briefly, tissue or cells were homogenized in 50 mM Tris-HCI (pH 7.4) containing a cocktail of protease inhibitors and washed twice by centrifugation at 50,000 x g for 30 min. Membranes were finally suspended in 25 mM potassium phosphate (pH 7.4) containing 200

mM KCl. Membrane proteins were solubilized in the presence of 2% (w/v) Triton X-100, 1 M KCl (Pfeiffer et al., 1982).

Immunological Methods and Sucrose Density Gradient Centrifugation The DORA was performed as described (Becker et al.. 1989). Mem-brane proteins solubilized in 0.5% (w/v) deoxycholate in Tris-buffered saline containing 20% (v/v) methanol were adsorbed onto nitrocellulose filters. These sheets were subsequently reacted with MAbs as indicated and anti-mouse immunoglobulin coupled to horseradish peroxidase. Enzyme activity was quantified pho-tometrically. Western blots were performed according to Schmitt et al. (1987) using alkaline phosphatase-coupled second antibody. GlyR antigen-containing detergent extracts of membranes from

• transfected cells and spinal cord tissue were layered onto 5%-20% sucrose gradients on top of a 60% sucrose cushion and centrifuged at 100,000 X g for 24 hr. Gradient fractions were analyzed by DORA. Marker proteins included in the gradients were detected by their enzymatic activities (see Becker et al., 1989).

[^H]Strychnine Binding Assay and Affinity Chromatography Glycine-displaceable [^H]strychnine binding to crude membrane fractions from spinal cord and transfected cells was determined by filtration assay (Becker et al., 1986). Detergent extracts from both spinal cord and transfected cells were cycled overnight over a 2-aminostrychnine column (1.5 ml). The column was washed, and GlyR was eluted from the affinity resin by 200 mM glycine (Pfeiffer et al., 1982). Flow-through fractions, wash, and column eluates were analyzed by DORA.

Acknowledgments

We thank I. Wolters for expert technical assistance and J. Rami for skillful help during the preparation of the manuscript. This work was supported by grants from the Bundesministerium fur Forschung und Technologie, Deutsche Forschungsgemeinschaft (SFB 317), and Fonds der Chemischen Industrie.

Received January 1, 1989; revised February 23, 1989.

References

Barker, J. L (1985). GABA and glycine: ion channel mechanisms. In Neurotransmitter actions in the Vertebrate Nervous System, M. A. Koganski and J. L. Barker, eds. (New York: Plenum Publishing Corp.), pp. 71-100. Becker, C. -M., Herrhans-Borgmeyer, I., Schmitt, 8., and Betz, H. (1986). The glycine receptor deficiency of the mutant mouse spas-tic: evidence for normal glycine receptor structure and localization. ]. Neurosci. 6, 1358-1364. Becker, C.-M., Hoch, W., and Betz, H. (1988). Glycine receptor het-erogeneity in rat spinal cord during postnatal development. EMBO J. 7, 3717-3726. Becker, C.-M., Hoch, W., and Betz, H. (1989). Sensitive immunoas-say shows selective association of peripheral and integral mem-brane proteins of the inhibitory glycine receptor complex. J. Neu-rochem., in press. Betz, H. (1987). Biology and structure of the mammalian glycine receptor. Trends Neurosci. 70, 113-117. Betz, H., and Becker, C.-M. (1988). The mammalian glycine recep-tor; biology and structure of a neuronal chloride channel protein. Neurochem. Int. 73, 137-146. Blair, L. A. C., Levitan, E., Marshall, J., Dionne, V. E., and Barnard, E. A. (1988). Single subunits of the GABA^ receptor form ion chan-nels with properties of the native receptor. Science 242, 577-579. Bormann, J. (1988). Electrophysiology of GABA^ and GABAg receptor subtypes. Trends Neurosci. 11, 112-116. Bormann, ]., Hamill, O. P., and Sakmann, B. (1987). Mechanism of anion permeation through channels gated by glycine and y-amino-butyric acid in mouse cultured spinal neurones. J. Physiol. 385, 234-286.


Boulter, J., Conoliy, J., Deneris, E. , Goldman, D., Heinemann, S., and Patrick, J. (1987). Functional expression of two neuronal nico-tinic acetylcholine receptors from c D N A clones identifies a gene family. Proc. Natl. Acad. Sci. USA 84, 7763-7767. Chen , C . , and Okayama, H . (1987). High efficiency transformation of mammalian cells by plasmid DNA . Mol. Cell . Biol. 7, 2745-2751. Curtis, D. R. , Hosli, L , Johnston, C . A. R., and Johnston, I. H . (1968). The hyperpolarization of spinal motoneurones by glycine and related amino acids. Exp. Brain Res. 5, 235-258. Deneris, E. S., Connolly, J. , Boulter, J. , Wada, E. , Wada, K., Swan-son, L. W. , Patrick, J., and Heinemann, S. (1988). Primary structure and expression of 32: a novel subunit of neuronal nicotinic acetyl-chol ine receptors. Neuron I , 45-54 .

Eaton, D. L., Wood, W- U Eaton, D. , Hass, P. E., Hollingshead, R, Wion , K. , Mather, J., U w n , R. M. , Vehar, C . A . , and Gorman, C . (1986). Construction and characterization of the active factor V l l l variant lacking the central one-third of the molecule. Biochemistry 25, 8343-8347.

Graham, D., Pfeiffer, F., and Betz, H . (1981). UV light-induced cross- • linking of strychnine to the glycine receptor of rat spina! cord mem-branes. Biochem. Biophys. Res. Commun. 702, 1330-1335. Graham, D. , Pfeiffer, F., and Betz, H. (1983). Photoaffinity-Iabelling of the glycine receptor of rat spinal cord. Eur. J. Biochem. 131, 519-525.

Graham, D., Pfeiffer, F., Simler, R. , and Betz, H. (1985). Purification of the glycine receptor of pig spinal cord. Biochemistry 24, 990-994 .

Grenningloh, G . , Rienitz, A . , Schmitt, B. , Methfessel, C . , Zensen, M. , Beyreuther, K., Gundelfinger, E. D., and Betz, H. (1987a). The strychnine-binding subunit of the glycine receptor shows homol-ogy with nicotinic acetylcholine receptors. Nature 328, 215-220. Grenningloh, G . , Gundelfinger, E.. Schmitt, B., Betz, H. , Darlison, M. G . , Barnard, E. A . , Schofield, P. R. , and Seeburg, P. H. (1987b). G lyc ine vs GABA receptors. Nature 330, 25-26. Hamil l , O. P., Marty, A . , Neher, E., Sakmann, B., and Sigworth, F. J. (1981). Improved patch-clamp techniques for high-resolution cur-rent recording from cells and cell-free membrane patches. Pflugers Arch . 391, 85-100.

Hucho, F. (1986). The nicotinic acetylcholine receptor and its ion channel . Eur. J. Biochem. 158, 211-226.

Kozak, M. (1986). Point mutations define a sequence flanking the A U G initiator codon that modulates translation by eukaryotic ribo-somes. Cell 44, 283-292.

Langosch, D. , Thomas, L. , and Betz, H . (1988). Conserved quarter-nary structure of ligand-gated ion channels: the postsynaptic gly-cine receptor is a pentamer. Proc. Natl. Acad. Sci. USA 85, 7394-7398.

Levitan, E., Schofield, R R. , Burt, D. R. , Rhee, L. M. , Wisden, W. , Kohler, M. , Fujita, N. , Rodriguez, H . F., Stephenson, A . , Darlison, M . G . , Barnard, E. A . , and Seeburg, R H. (1988). Structural and functional basis for GABA^ receptor heterogeneity. Nature 335, 76-79.

Mamalaki , C . , Stephenson, F. A . , and Barnard, E. A . (1987). The GABAA/benzodiazepine receptor is a heterotetramer of homolo-gous a and P subunits. E M B O J. 6, 561-565. Melton, D. A . , Krieg, R A . , Rebagliati, M. R., Maniatis, T , Z inn, K. , and Green, M. R. (1984). Efficient in vitro synthesis of biologically active RNA and RNA hybridization probes from plasmids contain-ing a bacteriophage SP6 promoter. Nucl . Acids Res. 12, 7035-7056.

Mishina, M. , Kurosaki, T., Tobimatsu, T , Morimoto, Y., Noda, M . , Yamamoto, T., Terao, M. , Lindstrom, J., Takahashi, T , Kuno, M. , and Numa, S. (1984). Expression of functional acetylcholine receptor from cloned cDNAs . Nature 307, 604-608.

Olsen, R. W. (1981). GABA-benzodiazepine-barbiturate receptor interactions. J. Neurochem. 37, 1-13.

Pfeiffer, F., Graham, D., and Betz, H. (1982). Purification by affinity chromatography of the glycine receptor of rat spinal cord. J. Biol.

Pfeiffer, F., Simler, R. , Grenningloh, G . / a n d Betz, H. (1984). Mono-clonal antibodies and peptide mapping reveal structural similarities between the subunits of the glycin^ receptor of rat spinal cord. Proc. Natl. Acad. Sci. USA 81, 7224-7227. Pritchett, D. B., Sontheimer, H . , Gorman , C . M. , Kettenmann, H . , Seeburg, R H . , and Schofield, R R. (1988). Transient expression shows ligand gating and allosteric potentiation of GABA^ receptor subunits. Science 242, 1306-1308.

Schmieden, V., Grenningloh, G . , Schofield, R, and Betz, H. (1989). Functional expression inXenopus oocytes of the strychnine binding 48 kd subunit of the glycine receptor. EMBO J. 8, 695-700. Schmitt, B., Knaus, R, Becker, C .-M. , and Betz, H . (1987). The Mr 93000 polypeptide of the postsynaptic glycine receptor complex is a peripheral membrane protein. Biochemistry 26, 805-811. Schofield, R R., Darl ison, M . G^, Fujita, N. , Burt, D. R., Stephen-son, F. A . , Rodriguez, H . , Rhee, L. M. , Ramachandran, ) . , Reale, V., Glencorse, T. A . , Seeburg, R H . , and Barnard, E. A . (1987). Se-quence and functional expression of the GABAA receptor shows a ligand-gated receptor super-family. Nature 328, 221-227. Sontheimer, H . , Kettenmann, H . , Backus, K. H . , and Schachner, M. (1988). Glutamate opens Na+/K^ channels in cultured astrocytes. G l ia I, 328-336.

Snyder, S. H . , and Bennett, ). R (1976). Neurotransmitter recejDtors in the brain: biochemical identification. Annu . Rev. Physiol. 38, 153-175.

Young, A. B., and Snyder, S. H . (1973). Strychnine binding as-sociated with glycine receptors of the central nervous system. Proc. Natl. Acad. Sci. USA 70, 2832-2836.

TiPS-June 1989 [118] Volume 10, No. 6

Trends in

Sciences mcluding TOXICOLOGICAL SCIENCES

Molecular pharmacology and drug action: structural information casts light on ligand binding HOW CAN AGONISTS and antagonists be discriminated at a molecular level? This question formed the basis of discussion at a recent Titisee conference'^. To even begin to answ^er this ques-tion, the molecular functions of drug receptors must be under-stood, and the recent cloning of the genes for numerous drug receptors is throwing a light on this problem. It is now clear that receptors belong to structurally related families such as G protein coupled receptors or ligand-gated ion channels. Also, receptor subtype diversity, as revealed by molecular biol-ogy, has been shown to be very large, larger than ever anticipated.

C5n the basis of this

Strongylo-oentrotug Bel QC QC

Clinical pharmacology curricula in Europe and North America

. . . s e e p a g e 2 2 4

Bivalent iigands and the message-address concept in the desip of selective opioid receptor antagonists

. . .page 230

Intenetikin 2: protot^e for a new generation of immnnoaetire pharmaceuticals

^rcfem imi,ase G and iong^ter^. poteiitiatiert in me hiBDocarmj^

Fig. 1. Topographical representation of adenylyl cyclase from bovine brain, and four different guanylyl cyclases (GC) from various sources. Areas in red denote homologous sequences which are hy-pothesized to specify nucleotide-binding domains. Figure courtesy of A. Gilman.

knowledge, participants grappled with questions concerning the nature of drug binding sites, how effector molecules are coupled and the role of receptor heterogeneity. In the course of the three-day meeting, other advances in molecular pharmacology were presented, including the first sequence data for a mammalian adenylyl cyclase.

Adenylyl cyclase The elusive adenylyl cyclase has finally yielded to cloning technology. A1 Gilman (University of Texas, Dallas) presented his group's recently acquired primary amino acid sequence data on bovine brain adenylyl cyclase. Somewhat surprisingly, its modelled topography places it in the general family of transporter or channel proteins.

The sequence contains two similar domains of around 250 amino acids; comparison of these domains with other protein sequences indicates that they may specify nucleotide binding sites. Thus they have been shown to be similar to a cytoplasmic domain present in the four guanylyl cyclases that have so far Continued on p. 208

I'fHq, Elsevier Scionce Publishers Ltd. (UK) OlhS - feHy/Sf/Sn^.On

208

continued from front page been cloned (see Fig. 1). Although the regulatory m e c h a n i s m s of these enzymes m a y differ from adenylyl cyclase, their catalytic mechanisms m a y be a s s u m e d to be the same. Furthermore, w h e n c o m -pared with protein data banks, the nine highest h o m o l o g y scores were observed with other nucleo-tide binding proteins.

Hydropathy data indicate the protein to have a pair of six m e m -brane-spanning segments separ-ated by a large (43kDa) cyto-plasmic loop and one major extra-cellular loop containing one of the four possible glycosylation sites in the sequence. There is a short cyto-plasmic amino-terminal sequence and a long (36kDa) cytoplasmic tail. Thus most of the protein in the Gilman model is located on the cytoplasmic side of the m e m b r a n e . The proposed t r a n s m e m b r a n e

_ spans have been modelled as a -helices and several of these are amphipathic .

The structure that Gilman has hypothesized for adenylyl cyclase bears a surprising resemblance to proteins of markedly different function. Transporters and chan-nels such as the dihydropyr idine-sensitive Ca^"^ channel and the drug-efflux p u m p - the P-glyco-protein whose synthesis is en-hanced in mult idrug-resistant cells^ - share the feature of large cytoplasmic d o m a i n s separating sets of six t r a n s m e m b r a n e spans. P-glycoprotein, like adenylyl cyclase, is hypothesized to have a short amino-terminal cytoplasmic domain, two sets of six t r a n s m e m -brane spans separated by a large cytoplasmic d o m a i n and a large carboxy-terminal cytoplasmic tail; each of the large cytoplasmic do-mains contains an A T P binding site. These similar topographies do not reflect similar pr imary amino acid s e q u e n c e s - b u t do they reflect similar functions? Although ap-parently unlikely, there m a y be some evolutionary logic to this question: the cellular slime mold Dictyostelium discoideum exports cAMP as an extracellular signal for chemotaxis.

G protein-coupled receptor superfamily

The G protein-coupled ' family of receptors is characterized by a

'Drug Action at the Molecular l.ein-l: Dif-ferences Between Agonists and Ant:i<;;onists. Titisee. rRC. 12-16 April 19S9.

c o m m o n topology with seven t ransmembrane regions and by varying degrees of p r i m a r y se-quence similarity^. W o r k reported at the conference indicates that agonists and antagonists actually bind within the t r a n s m e m b r a n e domains . More than 20 receptors from this family have n o w been cloned. H o w e v e r , ' the list', c o m -piled by Lutz B i r n b a u m e r (Baylor College, Houston) , includes at least 70 different receptors . This n u m b e r is certain to g r o w as n e w receptor subtypes c o n t i n u e to be identified. Within the adreno-ceptor family, two s u b t y p e s of the a 2 - receptor have b e e n cloned (Marc Caron, Duke Universi ty , Durham). They c o r r e s p o n d to the platelet and kidney a a - a d r e n o -ceptors and, while they both have a 2 -adrenergic p h a r m a c o l o g y , they can be dis t inguished by their affinities for epinephr ine , nor-epinephrine and oxymetazol ine .

Proteolytic removal of the large extracellular and intracellular d o m a i n s from purified Pi -adreno-ceptors (Elliot Ross, Univers i ty of Texas, Dallas) s h o w e d that l igands bound within the r e m a i n i n g trans-m e m b r a n e domains . Moreover , a t ryptophan residue in the seventh t ransmembrane s e g m e n t was identified as a port ion of the ligand binding site by photolabelling. In another set of e x p e r i m e n t s described by Marc C a r o n , chimeric a2-P2-adrenoceptors w e r e con-structed. Switching of t r a n s m e m -brane domains (TM) indicated that TM7 of these receptors appears to be particularly important in deter-mining ligand binding specificity.

Structure-funct ion s tudies indi-cate that negatively c h a r g e d aspar-tic acid residues within the second and third t r a n s m e m b r a n e do-mains are crucial for l igand bind-ing''-^. These d o m a i n s h a v e also been identified by var ious photo-labelling experiments^. T h e role of these negatively c h a r g e d residues appears to be in binding (or chan-nelling) positively charged ligands (e.g. adrenergic , muscar inic ) into the centre of the receptor . Nigel Birdsall (National Institute of Medical Research, L o n d o n ) exam-ined the pH d e p e n d e n c e of antag-onist binding to m2 (cardiac) mus-carinic receptors. Most antagonists possess a single posit ive charge and can act with either of two receptor-borne carboxyl groups (presumably two of the conserved aspartate residues in T M 2 and

TiPS-fune 1989 (Vol. 101

TM3). It is also possible for two antagonists to bind simul-taneously to the receptor with the positive charge of each antagonist interacting with a separate car-boxylate residue. In the case of methoctramine, an antagonist with multiple positive charges, its s imultaneous interaction with both carboxylate residues was demonstrated. It appears that cardioselective antagonists of differing selectivities interact predominantly with one particular aspartate. This choice is manifest in the magni tude of the apparent pK estimates for protonation of the receptor w h e n the different antagonists bind.

The role of the interaction of positively charged residues on the cytoplasmic surface of the receptor with G proteins was suggested by Elliot Ross. He described the photolabelling of G protein a - s u b -units by the wasp v e n u m masto-paran. This helical peptide pre-sents a face of positively charged residues which can activate the G protein. This face m a y be akin to a similar motif present in the puta-tive G protein binding domain of this receptor family. G protein-coupling sites on the receptors in-clude three domains on the cyto-plasmic side of TM5, TM6 and TM7 (Marc Caron) .

Together , these results indicate that the t r a n s m e m b r a n e domains fold around to make a pocket in which the ligand binds. The cyto-plasmic side of this structure has a n u m b e r of contact points for G proteins, thus providing a means of signal transduction.

A novel covalent modification of the h u m a n (32-adrenoceptor in which palmitate is attached by a thioester linkage to a cytoplas-mically located cysteine residue (C-terminal to TM7) was also de-scribed by Marc Caron. This cys-teine residue is conserved be-tween various m e m b e r s of the superfamily. Mutating this residue to a glycine results in no incor-poration of palmitic acid and also produces receptors that are essen-tially uncoupled from G protein interaction.

G proteins Nine different genes have been

so far found to code for G proteins, and the family is expected to grow (Gilman, University of Texas). The CYPy-trimer is necessary for inter-action with receptor, but disso-

TiPS-lune 1989 (Vol. 70/ 209

n'i A Fflt^f^L

PfifLryAL

^ f — ' y f t ^ A S O h J i S T ^

Meeting highlights

MV M/frff tt l i n^

^a m i !

1 2 5

n Arguments about the acceptability of the term 'partial antagonist' continued on in the ba r . . .

Ky IS B€TT€K THAhl VOVK P€reR€€NT!

. . . while the arguments about the G protein subunit that is the active mediator of hormone action moved towards resolution.

The newly sequenced adenylyl cyclase was revealed to have topographical but not sequence similarity with other known structures including the Ca^^ channel, the drug efflux pump and the Loch Ness monster.

The group photo shows the participants enjoying clement weather

in the Black Forest location of the meeting.

210

ciation of the subunits is required for activation. In the continuing debate about which subunit is the active mediator of hornr^onal signal to effector (the variant a, or the invariant py). all participants at the meeting declared themselves 'a-chauvinists' (a term coined by Henry Bourne in a recent issue of Nature'^). The differing results from the Clapham and Birnbaumer labs on activating the K" channel with PY" and a-subunits respectively had tentatively been ascribed to contamination of subunit prepara-tions®. However, the differences can now be put down to levels and type of detergent used in the assays (CHAPS vs Lubrol, concen-tration cut-off point 1[xm); Lutz Birnbaumer maintains his belief that the major cellular role of Py is as a 'noise suppressor', the dimer acting from a general pool to atten-uate actions of all G„s acting in a normal membrane environment.

Pete Downes (Smith Klein and French) reported experiments where the relationship of Gp, a putative pertussis toxin-insen-sitive G protein assumed to regu-late phospholipase C, to Gg was investigated. Py-Subunits from other (non-Gp) G proteins were added at varying concentrations to turkey erythrocytes in which phospholipase C had been acti-vated by fluoride. Inhibition of both phospholipase C and aden-ylyl cyclase was observed over the same concentration range of Py, indicating that py has similar affin-ities for Gp and Gg.

Various fine-tuning mechan-isms of the regulatory G pro-tein cycle were suggested. Alex Levitzki (Hebrew University of Jerusalem) suggested that experi-mental data gathered to date do not preclude the possible per-manent association of G^s with adenylyl cyclase. Birnbaumer pointed out that it was not yet possible to determine which step was controlled by the hormone-receptor-G protein complex - the binding of GTP or the activation of the whole complex.

Ligand-gated ion channel receptors

The ligand-gated ion channel receptors are multimeric protein complexes. In the prototypical example, the neuromuscular nico-tinic acetylcholine receptor, the receptor consists of a pentameric

structure in which five structurally related subunits are assembled to generate an integral ion channel. Each subunit is believed to contain four transmembrane domains, of which one, TM2, forms the ion channel lining^. Thus, unlike the G protein-coupled receptors, the transmembrane domains of the ligand-gated ion channel receptors do not appear to form the ligand binding domain.

The coral-derived toxin, lopho-toxin, photolabels the a-subunit of the Torpedo nicotinic acetylcholine receptor and Tyrl90 is labelled in a competitive manner (Palmer Taylor, University of California, San Diego). This, and other work (such as that by Changeux and co-workers^®), has demonstrated that the acetylcholine binding site is very close to two cysteine residues located at positions 192 and 193, in the extracellular region immedi-ately adjacent to the transmem-brane domains of the ion channel. An analogous site in the glycine receptor, where a putative di-sulphide-bonded loop also exists, was proposed by Heinrich Betz (Centre for Molecular Biology, Heidelberg) to be the competitive binding site for strychnine. Inter-estingly, no equivalent disulphide bond is found in any of the G AB AA receptor subunits.

Receptor subtypes have also emerged as one of the distin-guishing features of studies on the nature of drug action. Ralf Schoepfer (Salk Institute, San Diego) outlined the progress that has been made in linking neuronal nicotinic acetylcholine receptor proteins to their cognate genes. In the chick brain, two receptor sub-types have been identified that share a common structural subunit (designated P or non-a) but differ in acetylcholine-binding subunits (a). Both receptor subtypes are insensitive to a-bungarotoxin, despite the presence of bungaro-toxin binding sites in the brain. Schoepfer described two new receptor subunits that, although members of this receptor super-family, are only distantly related to both a- and non-a-subunits. Subunit-specific antibodies are capable of immunoprecipitating a-bungarotoxin labelled receptors.

Peter Schofield (Pacific Biotech-nology, Sydney), describing ex-periments undertaken at the Uni-versity of Heidelberg, reported

TiPS-]une 1989 [Vol. 101

that 13 different GABA^ receptor subunits had been cloned. These represented five different families, three of which contained highly related subtypes. Changing a-sub-unit combinations appeared to alter the affinity of receptors for ligand, while the presence of the y2-subunit was required to ob-serve the full benzodiazepine pharmacology, including allosteric potentiation of GABA responses and the effects of inverse agonists. By altering the receptor subunit combinations, receptor subtypes are obtained that have pharma-cological profiles that match the type I and type 11 benzodiazepine receptors. In-situ hybridization shows that mRNA encoding the 6-subunit is found in various inter-neurons, suggesting a role in local signal modulation, whereas the y2-subunit mRNA for example is located in a totally different set of neurons. Thus, even these two subunits define distinct GABAA receptor subtypes. All known GABAA receptor heterogeneity derives from separate genes en-coding the various subunits; how-ever, novel variants generated by alternative splicing were reported for the glycine receptor (Betz). Betz has also isolated a cDNA encoding an embryonic form of the strych-nine-binding subunit of the gly-cine receptor. The developmental change in levels of adult and embryonic subunit expression correlates with a juvenile insen-sitivity to strychnine.

At the functional level, David Colquhoun (University College, London) described work per-formed by Claire Newland in which extraordinary GABAA receptor heterogeneity was seen in single channel recordings of gang-lionic membrane patches. Either of two interpretations would raise many questions: namely, is this heterogeneity caused by a single channel with great variability or by many different channels?

Each of the known GABAA receptor subunits (Schofield) and the strychnine-binding subunit of the glycine receptor (Betz) can form homomeric ligand-gated ion channels. In the latter case. Hill coefficients of 1.5-3.0 are seen, indicating positive cooperativity. Both glycine (Betz) and nicotinic acetylcholine receptor (Bob Stroud, University of California, San Francisco) transmembrane

TiPS-juue 1989 IVol. 10} 211

peptides will also form channels when reconstituted in lipid bilayers. Unfortunately, the chan-nels typically have long open times; the rest of the receptor must therefore be important in provid-ing a means of channel gating.

The stoichiometry of receptors in vivo continues to be controversial. Whole receptor iodination com-bined with individual subunit analysis suggests that the neuronal nicotinic acetylcholine receptor may have a tetrameric structure (Schoepfer). This is in contrast to the pentameric structures of the neuromuscular nicotinic acetyl-choline receptor (Stroud) and the glycine receptor (Betz).

Many of the issues concerning receptor structure and function will be clarified when crystal struc-tures are resolved. Stroud has obtained nicotinic acetylcholine receptor crystals. However, they refract with poor resolution. Hartmut Michel (Max Planck Insti-tute for Biochemistry, Frankfurt), who outlined the most elegant work on the crystal structure of the photoreaction centre, considered that many membrane proteins would not yield usable crystals. Despite this, a single structure for any member of a receptor super-family would provide information applicable to other members.

Jeff Watkins (University of Bristol) described the rapidly growing pharmacology of the exci-tatory amino acid (glutamate) receptors, of which multiple sub-types have already been defined (NMDA, kainate and quisqualate). Dozens of labs are attempting to purify and/or clone these recep-tors. Indeed, some dinner-table conversations planned the content of the first glutamate receptor paper - the feeling being that we already know much about this important receptor class.

EGF receptor The EGF receptor is a single,

membrane-spanning polypeptide. Following ligand binding at the extracellular domain, receptors dimerize and the intrinsic protein tyrosine kinase activity within the intracellular domain is activated (see Ref. 11). The transmembrane segment appears to serve only an anchor role. Several hundred mutant receptors (Jossi Schles-singer, Rorer Pharmaceuticals Inc., King of Prussia) and chimeric

receptors (Axel Ullrich, Max-Planck-Institut fur Biochemie, Martinsried) have been syn-thesized to analyse in more detail the molecular requirements for signalling through this receptor system. For example, chimeric EGF-insulin receptors can be syn-thesized and transported normally in cells. Experiments with these chimeras demonstrate that a ligand for the extracellular domain can activate the intracellular do-main of a different receptor, but the source of the intracellular domain dictates whether or not a receptor will be degraded (EGF) or recycled (insulin).

Single point mutations can lead to oncogenic transformation of EGF receptors which demonstrate l igand-independent constitutive tyrosine kinase activity. In this regard, the new family of selective tyrosine kinase inhibitors intro-duced by Levitzki has obvious potential. From the lead com-pound erbstatin (IC50 ~ 14JI,M) found in Streptomyces by Umezawa and colleagues in 1986, Levitski's group has developed a series of compounds, some of which they have termed tyrphostins, with improved affinity and selectivity for EGF over insulin receptors and other tyrosine kinases.

Steroid hormone receptors The question as to how agonists

and antagonists differ at the mol-ecular level was approached most directly during discussion of steroid hormone receptors. Unlike membrane-bound peptide recep-tors, steroid receptors are localized in the soluble fraction of the cell. Following steroid binding, the complex is activated, binds to specific sequences in the chrom-atin and initiates transcription and protein synthesis. Receptors for all steroid hormones are organized into three major domains: a vari-able amino-terminal domain; a relatively well-conserved carboxy-terminal domain which contains the hormone-binding region; and a well-conserved, cysteine-rich central domain^^. The carboxy-terminal domain is also respon-sible for interaction with the 90kDa heat-shock protein (hsp90) which dissociates on hormone binding, thereby having a possible indirect effect on nuclear trans-location (movement of the complex to the chromatin). The central

Quotes from the meeting

'It does not matter what you find, you must give it a name.' A. Levitzki

'When the news broke that adrenoceptors had a structure similar to rhodopsin it struck the receptor community like a bomb.'

Pierre de Meytes summarized the high-powered meeting.

212

domain, with its two Zn^"^-stabi-lized 'fingers' and additional hormone-dependent translocation signals, is most likely responsible for DNA binding of the receptor. In the absence of hormone, the DNA binding site of the receptor may be hidden within its tertiary structure; following agonist bind-ing, conformational changes ex-pose this region, allowing binding of the steroid receptor to the hor-mone-responsive element (HRE), an enhancer region upstream of the genes that are regulated. HREs are palindromic or near palin-dromic consensus sequences which bind steroid receptors. Since sometimes more than one receptor can bind to a consensus sequence, steroid specificity of gene activation must also involve relative titre of receptors and dif-ferences in frans-activation.

John Katzenellenbogen (Univer-sity of Illinois, Urbana) reported several differences between agon-ists and antagonists. In cell-free preparations, differences are apparent in ligand dissociation rates, sensitivities to proteases and thiol reagents, aggregation with hsp90, monoclonal antibody bind-ing and elution from ion-exchange columns; in nuclear preparations, antagonist-receptor complexes have a greater tendency to form

TiPS-june 19S9 (Vol. lOJ

dimers and are less rapidly pro-cessed; and in chromatin binding, antagonist complexes show dif-ferent binding kinetics. It is not clear, however, at what level these differences are critical, nor whether in-vitro chromatin bind-ing and transfection systems are providing fully accurate models for steroid antagonist action in vivo.

Ullrich Gehring (Institute for Biological Chemistry, Heidelberg) presented evidence that the gluco-corticoid receptor from S49 lym-phoid cells has differential stab-ility when bound to agonist (triamcinolone) or antagonist (RU-38486). Following agonist binding, activation by high salt (300 mM KCl) or warming causes the hetero-tetrameric glucocorticoid receptor to dissociate, and a low molecular weight form binds to DNA. Recep-tor-antagonist complexes acti-vated by salt at low temperatures follow a similar pattern of acti-vation and the low molecular weight form can be shown by foot-printing techniques to bind to the same specific sequence in the HRE. However, activation by warming to 28°C or 30°C (closer to physiological reality) caused dis-sociation of the antagonist from the receptor. Once the ligand has left the activated receptor which is in its 'unfolded' form, it is rapidly

degraded. Thus in this lymphoid cell system, molecular events fol-lowing either antagonist or agonist binding coincide to the point of nuclear association, but very little antagonist-receptor complex is available to bind DNA.

PETER R. SCHOFIELD AND ALISON ABBOTT*

Pacific Biotechnology Ltd. 74 McLachlan Avenue, Rushcutters Bay 2011. NSW Australia, and ^Trends in Pharmacological Sciences.

References 1 Gottesman, M. M. and Pastan, I. (1988)

Trends Pharmacol. Sci. 9, 54-58 2 Gilman, A. G. (1987) Annu. Rev. Biochem

56, 615-650 3 Dohlman, H. G., Caron, M. G. and

Lefkowitz, R. J. (1987) Biochemistry 26 2657-2664

4 Strader, C. D. et al. (1988) J. Biol. Chem 263. 10267-10271

5 Fraser, C. M., Chung, F-Z., Wang, C-D. and Venter, J. C. (1988) Proc. Natl Acad. Sci. USA 85, 5478-5482

6 Strader, C. D. et al. (1987) Proc. Natl Acad. Sci. USA 84, 4384-^388

7 Bourne, H. R. (1989) Nature 337, 504-505 8 Bimbaumer, L. (1987) Trends Pharmacol.

Sci. 8, 209-211 9 McCarthy, M. P., Earnest, J. P., Young,

E. F., Choe, S. and Stroud, R. M. (1986) Annu. Rev. Neurosci. 9, 383--413

10 Dennis, M. et al. (1988) Biochemistry 27, 2346-2357

11 Ramachandran, J. and Ullrich, A. (1987) Trends Pharmacol. Sci. 8, 2&-31

12 Beato, M. (1989) Cell 56, 335-344

TiPS will be publishing a short review on tyrphostins later in the year.

Neuron, Vol. 3, 327-337, September, 1989, Copyright © 1989 by Cell Press

Two Novel G A B A a Receptor Subunits Exist in Distinct Neuronal Subpopulations

Brenda D. Shivers* Iris Killisch* Rolf Sprengel* Harald Sontheimer,+ Martin Kohler,* Peter R. Schofield** and Peter H. Seeburg* * Center for Molecular Biology University of Heidelberg + Institute for Neurobiology University of Heidelberg D6900 Heidelberg Federal Republic of Germany

Summary

Two cDNAs encoding novel G A B A A receptor subunits were isolated from a rat brain library. These subunits, y2 and 6, share approximately 35% sequence identity with a and 3 subunits and form functional GABA-gated chloride channels when expressed alone in vitro. The y2 subunit is the rat homolog of the human y2 subunit recently shown to be important for benzodiazepine pharmacology. Cellular localization of the mRNAs en-coding the y2 and 6 subunits in rat brain revealed that largely distinct neuronal subpopulations express the two subunits. The 6 subunit distribution resembles that of the high affinity G A B A A receptor labeled with [3H]muscimol; the y2 subunit distribution resembles that of GABAA/benzodiazepine receptors labeled with [^HJflunitrazepam. These findings have implications for the composition of two different G A B A A receptor subtypes and for information processing in networks using G A B A for signaling.

Introduction

Neurotransmission effected by y-arninobutyric acid (GA-BA) is mediated mainly by a chloride channel-receptor complex (GABAA receptor), present in approximately one-third of all brain synapses (Stephenson, 1988). Neu-rotransmitter-elicited channel activity can be alloste-rically modulated by a variety of compounds, with barbiturates and benzodiazepines (BZ) representing well-known, clinically used examples (Olsen and Venter, 1986). The GABAA receptor consists of subunits that are structurally related to those of other ligand-gated ion channels (Schofield et al., 1987; Grenningloh et al., 1987). Molecular cloning of cDNAs encoding GABAA receptor subunits is shaping our view of the molecular properties of this receptor. Several variants of both the a and the 3 subunits have been characterized (Levitan et al., 1988; Ymer et al., 1989). Members within these subunit classes display approximately 75% overall se-quence identity and confer different ligand affinities on receptors formed by them (Levitan et al., 1988). Cloned

• Present address: Pacific Biotechnology Limited, 72 McLachlan Av-enue, Rushcutters Bay 2011, Australia.

cDNAs encoding two members of a third subunit class that share 35% homology with the a and 3 subunits were isolated. One of these subunits, y2, is found in a large neuronal subpopulation and represents an impor-tant constituent of BZ-responsive GABAA receptors (Pritchett et al., 1989). The other subunit, y1, is ex-pressed in glial cells (unpublished data). Here, we report the characterization of a fourth subunit class, termed 5, which also shares 35% sequence similarity with the a, 3, and Y subunits. The 6 subunit seems to be a unique member of this class, as no 5 variants were found.

The emergence of this diversity of subunits, unex-pected from classic GABAA receptor pharmacology and biochemistry, calls for ways of exploring which subunit combinations form naturally occurring GABAA receptor subtypes. In an attempt to characterize these subtypes on the basis of regional expression of receptor subunits in rat brain, we noticed that the y2 and 6 subunit mRNAs exist in largely distinct neuronal subpopulations. Hence, these subunits are presumably part of functionally and pharmacologically distinct GABAA receptors. In this pa-per, we show evidence that both of these novel subunits belong to the GABAA receptor family and describe their expression pattern in rat brain.

Results

Cloning of 8 Subunit cDNA Screening cDNA libraries with a degenerate oligonucle-otide probe encoding a conserved octapeptide sequence in the second transmembrane region (M2) of GABAA receptor subunits led to the identification of several new subunit-encoding cDNAs (Schofield et al., 1987; Gren-ningloh et al., 1987; Levitan et al., 1988). In particular, two new 3 subunits were found in bovine and rat (Ymer et al., 1989), and a human member of a third subunit class, Y, important for BZ was also identified (Pritchett et al., 1989). Using this approach, cDNAs encoding the rat homolog of y2 as well as a member of a fourth subunit class, 6, were also isolated. The nucleotide and encoded amino acid sequences of these cDNAs are shown in Fig-ure 1. The ratY2 sequence is closely related to its human homolog and thus displays the same features (Pritchett et al., 1989). The rat 5 cDNA encodes a 450 residue polypeptide whose primary structure predicts an N-ter-minal signal peptide, a disulfide-bonded, 3-structural loop, two adjacent N-linked glycosylation sites in the putative extracellular domain, and four transmembrane segments. Three membrane-spanning regions are in close proximity in the middle of the molecule, and one is present at the C-terminus. These features are the hallmarks of all subunits of ligand-gated ion channels (Schofield et al., 1987; Grenningloh et al., 1987).

The transmembrane regions of the 6 polypeptide are closely sequence-related to those found in other GABAA receptor subunits and in the 48 kd subunit of the glycine receptor. This is illustrated in Figure 2, which compares

Neuron 328

1 - 1 6

CCGGAACGTC CCGCGCACAG CCCGCAAGCC ATG Met

GAC GTT CTG GGC TGG CTG CTG CTG CCG Asp Val Leu Gly Trp Leu Leu Leu Pro

61 -6

CTG Leu

CTT Leu

GTG Leu

GTG Leu

TGG Cys

AGG CAG ThrTGln

CGG Pro

GAC His

GAT His

GGG Gly

GGC Ala

AGA Arg

GCA ATG AAT GAC ATT GGG GAC Ala Met Asn Asp H e Gly Asp

121 15

TAC Tyr

GTG Val

GGC Gly

TCC Ser

AAC Asn

CTG Leu

GAG Glu

ATA H e

TCC Ser

TGG Trp

CTG Leu

CCC Pro

AAC Asn

CTG GAT GGA GTA ATG GAG GGC Leu Asp Gly Leu Met Glu Gly

181 35

TAC Tyr

GCC Ala

CGA Arg

AAC Asn

TTC Phe

CGA Arg

CGA Pro

GGC Gly

ATT H e

GGA Gly

GGT Gly

CCT Pro

CCA Pro

GTG AAT GTG GGG CTT GCC GTA Val Asn Val Ala Leu Ala Leu

241 55

GAG Glu

GTG Val

GCC Ala

AGC Ser

ATT H e

GAC Asp

CAC His

ATC H e

TCA Ser

GAA Glu

GCA Ala

AAT Asn

ATG Met

GAA TAG AGC ATG ACA GTG TTC Glu Tyr Thr Met Thr Val Phe

301 75

GTG Leu

CAC His

AGA Arg

GCT Ala

TGG Trp

CGA Arg

GAC Asp

AGC Ser

AGG Arg

GTG Leu

TCC Ser

TACrAACnCAT ACCrAACiGAG ACC CTG GGC Tyr Asn His Thr Asn Glu Thr Leu Gly

361 95

CTG Leu

GAT Asp

AGC Ser

GGC Arg

TTC Phe

GTG Val

GAC Asp

AAG Lys

CTG Leu

TGG Trp

CTC Leu

CCT Pro

GAG Asp

ACC TTG ATT GTG AAT GCC AAA Thr Phe H e Val Asn Ala Lys

421 115

GTC Val

TGG Cys

CT5 Leu

GTT Val

CAT His

GAT Asp

GTG Val

ACC Thr

GTG Val

GAA Glu

AAC Asn

AAG Lys

CTT Leu

ATG GGC GTA CAG CGG GAG GGT H e Arg Leu Gin Pro Asp Gly

481 135

GTG Val

ATT H e

TTA Leu

TAC Tyr

AGC Ser

ATC H e

GGC Arg

ATC H e

AGC Thr

TGC Ser

ACA Thr

GTG Val

GGC Ala

TGT GAC ATG GAC CTT GCC AAG Cys Asp Met Asp Leu Ala Lys

541 155

TAG Ty

CGG Pro

ATG Met

GAC Asp

GA*; Glu

CAG Gin

GAG Glu

TGC CVS

ATG Met

GTG Leu

GAC Asp

CTG Leu

GAG Glu

AGC TAT GGC TAC TCT TCT GAG Ser Tyr Gly Tyr Ser Ser Glu

601 175

661 195

721 215

781 235

841 255

901 275

961 295

1021 315

1081 335

1141 355

1201 375

1261 395

1321 415

1381 1451 1521 1591 1661 1731

GAC ATT GTC TAT TAT TGG TCA GAA AAC CAG GAG Asp lie Val Tyr Tyr Trp Ser Glu Asn Gin Glu

CTG GCC CAG TTC ACT ATG ACC AGT TAC CGC TTC Leu Ala Gin Phe Thr H e Thr Ser Tyr Arg Phe

GCT GGC CAG TTC CCT CGA CTC AGC TTA CAC TTC Ala Gly Gin Phe Pro Arg Leu Ser Leu His Phe

ATC ATC CAG TCT TAC ATG CCC TCT GTC CTC CTG H e H e Gin Ser Tyr Met Pro Ser Val Leu Leu

CAG ATC CAC GGG CTG GAC AGG CTG CAA Gin H e His Gly Leu Asp Arg Leu Gin

ACC ACG GAG CTG ATG AAC TTC AAA TCA Thr Thr Glu Leu Met Asn Phe Lys Ser

CAG CTT GGG AGG AAC GGG GGT GTC TAC Gin Leu Arg Arg Asn Arg Gly Val Tyr

GTT GCC ATG TCC TGG GTC TCC TTC TGG Val Ala Met Ser Trp Val Ser Phe Trp

ATT AGC CAA GCA GCA GTG CCT GCC AGA GTA TCT H e Ser Gin Ala Ala Val Pro Ala Arg Val Ser

GTA GGC ATC ACC ACT GTG GTG AGA ATG Leu Gly H e Thr Thr Val Leu Thr Met

ACC ACA CTC ATG GTT AGT GGG CGC TGG TCC CTC Thr Thr Leu Met Val Ser Ala Arg Ser Ser Leu

GAT GTG TAT TTC TGG ATC TGG TAT GTC TTC GTG Asp Val Tyr Phe Trp H e Cys Tyr Val Phe Val

GGG CGG GCT TCT GCT ATC AAG GCT GTG Pro Arg Ala Ser Ala H e Lys Ala Leu

TTT GCT GCC GTG GTG GAG TAT GCA TTT Phe Ala Ala Leu Val Glu Tyr Ala Phe

GCC GAC TTC AAT GGT GAC TAG AGG AAG AAA CGG Ala His Phe Asn Ala Asp Tyr Arg Lys Lys Arg

AGG GCA GAG ATG GAC GTG AGG AAC GCC ATT GTC Arg Ala Glu Met Asp Val Arg Asn Ala H e Val

AGG CAG GAG TTG GCT ATC TCC CGG GGT CAA GGC Ser Gin Glu Leu Ala H e Ser Arg Arg Gin Gly

TAT AGG TCT GTA GAA GTG GAG GCA AAG AAG GAG Tyr Arg Ser Val Glu Val Glu Ala Lys Lys Glu

GGG ATC GGT TCC AGA CTC AAA GCC ATC GAT GCA Gly H e Arg Ser Arg Leu Lys Pro H e Asp Ala

GTG TTC CCG GCA GGG TTT GCA GCA GTC AAC ATC Val Phe Pro Ala Ala Phe Ala Ala Val Asn H e

AAA GGG AAG GTC AAG GTC AGG AAG CCA Lys Ala Lys Val Lys Val Thr Lys Pro

CTC TTC TCC CTG TCT GCT GCC GGG GTC Leu Phe Ser Leu Ser Ala Ala Gly Val

CGG GTG CCT GGG AAC CTC ATG GGT TCC Arg Val Pro Gly Asn Leu Met Gly Ser

GGG GGG GTC CCG CCA GGG GGC CCA GGA Gly Gly Val Pro Pro Gly Gly Pro Gly

GAC ACC ATC GAG ATC TAT GCC GGC GCT Asp Thr H e Asp H e Tyr Ala Arg Ala

ATC TAC TGG GCA GGG TAT AGC ATG TGA H e Tyr Trp Ala Ala Tyr Thr Met

CGGCAGTGCG TAGACCACAT GAGGGCTTTA CATGTACCAC ACAGCCCTGT CCGTGGATGG CAAAGTGGGA GAGAGAGGAG GGTCTACCAG CCTGCACTTG GTGTAGATGG AGCAGGATGG CCCTGAAGCT GAGGCTGGCA GTAAGGTGGC CTTGGGAGCT TCTGGGCCTT TGCTCTGTGG GATGAGGATC AGAGAGAAAG GAACTGGAGG AGAAGGGCAT TATGAGGCCT GTTTGGTCCA

GTGCGGTGGT TGCCACAAAG TCCTGGAGGA GCTCCGCTCC CTGGGCTTTA CCTCAATTTT GCTTTGCACA GAGGGCCATT AGTTGGCTGT GTGAAGTTTA GACTCAGGAT GGGCCTGATT AGAGTAGGGG TGGATGACCA CTCTCATGGT GCATGAATAA AGCCTTGGCC TGGCAAAAA

the primary structures of one member of each of the four classes of GABAA receptor subunits. The longest con-served sequence Is in M2 (Schofieid et al., 1987; Gren-ningioh et al., 1987; Levitan et al., 1988) and encom-passes the 8 residues used for designing the degenerate oligonucleotide probe. Although only a few other posi-tions are conserved, all four subunits display a common design, including positions of transmembrane regions and distribution of positively charged amino acid resi-dues in regions corresponding to the outer and inner channel mouth (Schofieid et al., 1987; Levitan et al., 1988; Pritchett et al., 1989; Ymer et al., 1989). These fea-tures represent the structural correlate of the functional signature common to all of these subunits, namely to participate in the formation of GABA-gated chloride channels. Particularly low sequence conservation char-acterizes the extended putative intracellular loop of all subunits. This region Is of unknown cell physiological significance, although the presence of consensus se-quences for phosphorylation suggests an involvement in

the regulation of channel activity (Huganir and Green-gard, 1987).

Functional Expression The sequence features of the y2 and 5 polypeptides mark them as GABAA receptor subunits. In fact, the hu-man as well as the rat y l subunit, when coexpressed, forms receptors with al and p i subunits which display high affinity BZ binding (Pritchett et al., 1989). In these receptors, substitution of the 8 subunit for y l abolishes the expression of BZ binding sites (D. B. Pritchett and R H. Seeburg, unpublished data). As a and P subunits show a propensity to self-assemble and form homo-meric GABA-gated chloride channels, we determined whether the y l and 5 subunits also display these func-tional characteristics typical of GABAA receptor sub-units. Human embryonic kidney 293 cells were transfected with the cloned y l or 5 subunit cDNAs engineered for expression and were analyzed using the whole-cell, patch-clamp technique (Hamill et al., 1981). As shown

Novel CABAA Receptor Subunits 329

B 1

- 38 TC TTCTGCAACC AGAGGCGAGA GGCGAGAGGA AAAAAAAGCG ATG

Het AGT Ser

TCG Ser

CCA Pro

AAT ACA Asn Thr

61 - 3 2

TGG Trp

AGC ACT Ser Thr

GGA AGC ACA Gly Ser Thr

GTC Val

TAC Tyr

TCT Ser

CCT Pro

GTA Vat

TTT TCA Phe Ser

CAG G in

AAA Lys

ATG Met

ACG Thr

CTG Leu

TGG ATT Trp H e

121 -12

CTG Leu

CTC CTG Leu Leu

CTA TCG CTC Leu Ser Leu

TAC Tyr

CCA Pro

GGC Gly

TTC Phe

ACT Thr

AGC CAA S e r t G l n

AAG Lys

TCA Ser

GAT Asp

GAT Asp

GAC Asp

TAT GAA Tyr Glu

181 9

GAT Asp

TAT GCT Tyr A l a

TCT rA AT-, AAA Ser A sn Lys

ACA Thr

TGG Trp

GTG Val

TTG Leu

ACT Thr

CCA Pro

AAA Lys

GTT Val

CCC Pro

GAG Glu

GGT Gly

GAT Asp

GTC ACT Val Thr

241 29

GTC Val

ATC TTA H e Leu

AAC Asn

AAC CTT Asn Leu

CTG Leu

GAA Glu

GGG Gly

TAC Tyr

GAC Asp

AAC Asn

AAA Lys

CTT Leu

CGG Arg

CCC Pro

GAC Asp

ATA H e

GGA GTG Gly Va l

301 49

AAA Ly s

CCA ACA P ro Thr

TTA Leu

ATT CAT H e H i s

ACA Thr

GAT Asp

ATG Met

TAC Tyr

GTG Val

AAC Asn

AGC Ser

ATT H e

GGT Gly

CCA Pro

GTG Val

AAT Asn

GCT ATC A l a H e

361 69

AAT A sn

ATG GAA Het G lu

TAC Tyr

ACA ATT Thr H e

GAT Asp

ATT H e

TTT Phe

TTT Phe

GCC A l a

CAA Gin

ACC Thr

TGG Trp

TAT Tyr

GAC Asp

AGA Arg

CGT Arg

TTG AAA Leu Lys

421 89

TTTrAACnAGT Phe Asn Ser

ACC Thr

ATT AAA H e Lys

GTT Val

CTC Leu

CGA Arg

TTG Leu

AAT Asn

AGC Ser

AAT Asn

ATG Met

GTG Val

GGG Gly

AAA Lys

ATC H e

TGG ATT Trp H e

481 109

CCA Pro

GAC ACT Asp Thr

TTC Phe

TTC AGG Phe A rg

AAC Asn

TCC Ser

AAA Lys

AAA Lys

GCG A l a

GAT Asp

GCT A l a

CAC H i s

TGG Trp

ATC H e

ACG Thr

ACT Thr

CCC AAC Pro Asn

541 129

AGG Arg

ATG CTG Met Leu

AGA Arg

ATT TGG H e Trp

AAT Asn

GAC Asp

GGT G ly

CGA Arg

GTT Val

CTC Leu

TAC Tyr

ACC Thr

TTA Leu

AGG Arg

CTA Leu

ACA Thr

ATT GAT H e Asp

601 149

GCC A l a

GAG TGC Glu Cys

CAG G in

TTG CAA Leu Gin

TTA Leu

CAC H i s

AAC Asn

TTC Phe

CCA Pro

ATG Met

GAT Asp

GAA Glu

CAC H i s

TCC Ser

TGC Cys

CCC Pro

CTG GAG Leu Glu

661 169

721 189

781 209

841 229

901 249

961 269

1021 289

1081 309

1141 329

1201 349

1261 369

1321 389

1381 409

1441 1511 1581 1651 1721 1791

TTC TCC AGT Phe Ser Ser

GAA GTG GGA Glu Val G l y

ACC ACT GAA Thr Thr Gtu

CTG AGC AGA Leu Ser Arg

GTT CTG TCC Vat Leu Ser

TAT GGT Tyr G ly

GAC ACA Asp Thr

GTA GTG Val Va l

AGA ATG Arg Het

TGG GTG Trp Va l

TAT CCT Tyr Pro

AGG TCA A rg Ser

AAG ACA Ly s Thr

GGG TAG G ly T ^

TCC TTC Ser Phe

CGT GAA A rg G lu

TGG AGG Trp Arg

ACT TCT Thr Ser

TTT ACC Phe Thr

GAA ATT Glu H e

CTG TAT Leu Tyr

GGT GAC G ly Asp

ATC CAG H e G in

GTT TAT CAA Val Tyr G in

CAG TTT TCC G in Phe Ser

TAT GTG GTT Tyr Val Val

ACC TAC ATT Thr Tyr H e

TGG AAG Trp Lys

TTT GTT Phe Va l

ATG TCC Met Ser

CCC TGC Pro Cys

CGC AGT Arg Ser

GGA TTG G ly Leu

GTG TAC Val Tyr

ACA CTC Thr Leu

TCT GTT Ser Val

AGG tAATT A r g l A s n l

TTT GAT Phe Asp

ATT GTG l i e Val

TGG ATC Trp l i e

GGA ATC ACG G ly l i e Thr

ACT GTG Thr Va l

CTG ACG Leu Thr

ATG ACC Het Thr

AAT AAG Asn Lys

ACT CTC Thr Leu

GAT GCT GTC Asp A l a Val

CCT GCA Pro A l a

AGA ACA Arg Thr

TCT TTA Ser Leu

AGC ACC ATA Ser Thr l i e

AAG GTC TCC Lys Val Ser

TCA GCT TTG Ser A l a Leu

TAT GTC Tyr Val

GTG GAG Val G lu

ACA GCA Thr A l a

TAT GGT Tyr G ly

ATG GAT Met Asp

CTC TTC Leu Phe

GTC TCT GTT Val Ser Va l

GCC CGG A l a Arg

TGC TTC Cys Phe

AAG TCT Lys Ser

ATC TTT H e Phe

CTG CCC Leu Pro

GTG TTT Vat Phe

ACC CTG Thr Leu

GAT AAA GAC Asp Lys Asp

ACG ATC CAA Thr H e G in

TGT TTG GAT Cys Leu Asp

GCC TGG AGA A l a Trp Arg

TTC TTC CCT Phe Phe Pro

AAA AAG Ly s L y s

ATG AAC Met Asn

GGC AAG Gly Lys

CAC GGG H i s G l y

ACC GCC Thr A l a

AAG AAA Ly s Ly s

AAT GCC Asn A l a

GAC TGT Asp Cys

AGG ATA A rg H e

TTC TGC Phe Cys

AAC CCT Asn Pro

ACC CAC Thr H i s

GCC AGT A l a Ser

CAC ATT H i s H e

TTG TTC Leu Phe

CAC TAT H i s Tyr

GCC CCT A l a Pro

CTT CAA Leu G in

TTC TTT Phe Phe

CGC ATT Arg H e

AAT CTT Asn Leu

TTT GTG AGC Phe Va l Ser

ACC ATT GAT Thr H e Asp

GAG AGG GAT Glu Arg Asp

TGC TGT TTT Cys Cys Phe

GCC AAA ATG A l a Lys Met

GTT TAC TGG Vat Tyr Trp

AAC CGG Asn Arg

ATC CGT H e Arg

GAA GAA Glu Glu

GAA GAC Glu Asp

GAC TCC Asp Ser

GTC TCC Val Ser

AAA CCA Lys Pro

CCC AGA Pro Arg

TAT GGC Tyr G l y

TGC CGA Cys Arg

TAT GCT Tyr A l a

AGC AAG Ser Ly s

TCA GCA Ser A l a

TAT GAG Tyr G lu

ACA GGA Thr G ly

CGG ATC Arg H e

TAT CTT Tyr Leu

TAT CTG Tyr Leu

TGAGGAGGTT TGGGTTTTAT CGATAGGGGT CTTATTCGCT GAATCTCACG AAGAAAAAAT GTCCTTTCTA AGTCCAACGA TCTAATACCC TATGTGGGTC ACTGAATGTG TTTCTCCGTC TCACCTAGTA ATACATAGGT CCTACGATCC GTGCCCAACT CTCCTTTGGT TATTGTAATT TGAATTTTAG CGTGCACTGT CTTTCAAACC TGCAAGGCAA GTTAAGTTAG AACAAGAGCT GTTAGACTGA GCAAGATACC TTTGAGCAAA AGCACTGCAA ACTGAAATCT GTGAGTGTTT AAAGGGGCAT CATTAGCCTT TGATTCACGT ATAACTTAGT ACAAAAGGTA TACAGCTGTG TGCTACGTTC GCTTGGGGTA CGTTGTGTAA ATTGAA

Figure 1. Nucleotide and Predicted Amino Acid Sequences of Cloned cDNAs for Rat GABAA Receptor 6 and y l Subunits (A) h subunit; (B) Y2 subunit. Nucleotides are numbered from the start of the cloned cDNA. Amino acid residues that are part of the signal sequence carry negative num-bers. A small arrows denotes the proposed sig-nal peptide cleavage site (von Heijne, 1986). N-linked glycosylation sites are boxed, the disulfide-bonded, (J-structural loop sequence is denoted by a dotted line, and the four transmembrane segments are underlined.

in Figure 3, the expression of either subunit generated chloride channels gated by high GABA concentrations. These were always bicuculline- and picrotoxin-sensitive (data not shown), and their activity was potentiated by the barbiturate pentobarbital. However, as seen with homomeric channels formed from a or |3 subunits, no BZ-mediated potentiation could be observed. More-over, glycine failed to gate homomeric channels formed from a, (3, y, or 5 subunits. In contrast, the glycine recep-tor 48 kd subunit (Grenningloh et al., 1989), whose se-quence homology to a and P subunits is as low as that of the 6 subunit, forms homomeric glycine-gated chlo-ride channels that are strychnine-sensitive and not po-tentiated by barbiturates (Sontheimer et al., 1989). These data strongly indicate that the 5 subunit is a constituent of G A B A a receptors, as istheY2 subunit (Pritchett et al., 1989).

Neuronal Subpopulations Expressing 5 mRNA Inspection of film autoradiograms of rat brain sections hybridized with a cDNA probe for 5 mRNA revealed that the most prominently expressing region for this mRNA is the cerebellar cortical gray matter (Figure 4). This was consistent with its relative abundance shown first by Northern blot analysis of different brain regions (M. Kohler, unpublished data). Other regions having 8 mRNA hybridization densities detectable on film au-toradiograms include thalamic nuclei, dentate gyrus, olfactory bulb, olfactory tubercle, cerebral cortex, and nucleus accumbens. In younger animals (day 20) mod-erately high hybridization densities were observed in caudate-putamen (Figure 5B). 6 mRNA is a rare tran-script in hypothalamic, midbrain, and brainstem re-gions. Observation of emulsion autoradiograms of these sections showed that 5 mRNA is expressed in a subset

Neuron 330

1 QPHHGARAMNDIGDYVG...SNLEISWLPNLDGLMEGYARNFRPGIGGPPVNVALALEVASIDHISEANMEYTMTVFLHRAWRDSRLSYNHTNETLGLDSRFVDKLWLP 5

1 QKSDDDYEDYASNKTWVLTPKVPEGDVTVILNNLLEGYDNKLRPDIGVKPTLIHTDMYVNSIGPWAINMEYTIDIFFAQTWYDRRLKFNSTIKVLRLNSNMVGKIWIP 72

1 QPSQDELKDNTTVFTR ILDRLLDGYDNRLRPGLGERVTEVKTDIFVTSFGPVSDHDMEYTIDVFFRQSWKDERLKFKGPMTVLRLNNLMASKIWTP a l

1 HSSNEPSNMSYVKET VDRLLKGYDIRLRPDFGGPPVDVGMRIDVASIDMVSEVNMDYTLTMYFQQSWKDKRLSYSGIPLNLTLDNRVADQLWVP p i

L G Y R P G V S M Y T W D R L L L W P

1 0 7 DTFIVNAKVCLVHDVTVENKLIRLQPDGVILYSIRITSTVACDMDLAKYPMDEQECMLDLESYGYSSEDIVYYWSENQEQ..IHGLDRLQLAQFTITSYRFTTELMNFK

1 1 0 DTFFRNSKKADAHWITTPNRMLRIWNDGRVLYTLRLTIDAECQLQLHNFPMDEHSCPLEFSSYGYPREEIVYQWKRSSVE..VGDTRSWRLYQFSFVGLRNTTEWKTT

97 DTFFHNGKKSVAHNMTMPNKLLRITEDGTLLYTMRLTVRAECPMHLEDFPMDAHACPLKFGSYAYTRAEWYEWTREPARSWVAEDGSRLNQYDLLGQTVDSGIVQSS

95 DTYFLNDKKSFVHGVTVKNRMIRLHPDGTVLYGLRITTTAACMMDLRRYPLDEQNCTLEIESYGYTTDDIEFYVfNGGEGA..VTGVNKIELPQFSIVDYKMVSKKVEFT

D T N K H T N R D G L Y R T C L P D C L S Y Y W L Q

TM 1 TM2 TM 3

5 72 a l

p i

214 SAGQFPRLSLHFQLRRNRG VYIIQSYMPSVLLVAMSWVSFWI SQA AVPARVSLGITTVLTMTTLMVSA RSSLPRASAIK ALDVYFWICYVFVFAALVEYAFA H 5 216 S. GDYWMSVYFDLSRRMG YFTIQTYIPCTLIWLSWVSFWI NKD AVPARTSLGITTVLTMTTLSTIA RKSLPKVSYVT AMDLFVSVCFIFVFSALVEYGTL H 72 205 T. GEYWMTTHFHLKRKIG YFVIQTYLPCIMTVILSQVSFWL NRE SVPARTVFGVTTVLTMTTLSISA RNSLPKVAYAT AMDWFIAVCYAFVFSALlEFATV N al 201 T.GAYPRLSLSFRLKRNIG YFILQTYMPSTLITILSWVSFWI NYD ASAARVALGITTVLTMTTISTHL RETLPKIPYVK AIDIYLMGCFVFVFLALLEYAFV N pl

G F L R G Q Y P S VSFW AR G TTVLTMTT R LP A D C FVF AL E

317 FNAD YRKKRKAKVKVTK.PRAEMDVRNAIVLFSLSAAGVSQELAISRRQGRVPGNLMGSYRSV EVEAKKEGGVPPGGPGGIRSRLKPI. 5 319 YFVSN RKPSKDKDKKKKNPAPTIDIRPRSATIQMNNATHLQERDEEYGYECLDGKDCASFF CCFEDCRTGAWRHGRIHIRIAK 72 306 YF. .TKRGYAWDGKSWPEKPKKVKDPLIKKNNTYAPTATSYTPNLARGDPGLATIAKSATIEP KEVKPETKPPEPKKT al 304 YIFFGKGPQKKGASKQDQSANEKNKLEMNKVQVDAHGNILLSTLEIRNETSGSEVLTGVSDPKSTMYSYDSASIQYRKPLSSREGYGRGLDRHGVPGKGRIRRRASQLK Pl

K K K

404 DADTIDI 402 MDS 385 FNSVSKIDR 413 VKIPDLTDVNSIDK

D

TM 4 YARAVFPAAFAAVNIIYWAAYTM YARIFFPTAFCLFNLVYWVSYLY LSRIAFPLLFGIFNLVYWATYLN WSRMFFPITFSLFNWYWLYYVH

R FP F N YW Y

REPQLKAPTPHQ 72 a l pl

Figure 2. Amino Acid Sequence Comparisons of Rat GABAA Receptor 5, y l , al, and pi Subunits Amino acid numbers start with the proposed mature N-termini. Spaces are inserted to maximize alignment. The proposed disulfide-bonded loop region isoverlined, and the four transmembrane regions are boxed. Residues identical in all four polypeptides are listed below the respec-tive sequences.

®

GABA(100) GABA(10) GABA( IO) + PB

100pA

GABA( IO) 30 sec

®

50 pA

GABA(100) GABA(100) + PB

GABA(IOO) 30 sec

Figure 3. Homooligomeric Receptors Containing Either 5 or y2 Subunits Show Ligand-Gating and Pentobarbital Potentiation (A) Current recording obtained from a 293 cell expressing the 6 subunit. Application of GABA induced an inward current that was dose-dependent and strongly potentiated by pentobarbital (50 nM). The observed potentiation was reversible upon washout. The numbers in paren-theses correspond to the GABA concentrations (in \LM) applied. The time intervals between subsequent applications were 5 min. (B) Current recording of a 293 cell expressing the y2 subunit. GABA (100 m,M) induced an inward current that was reversibly potentiated about 2-fold by pentobarbital (50 ^M).

Novel C A B A A Receptor Subunits 331

m f

OB •

Cx

LS Md

'm Acb Tg

Pn

Figure 4. Film Autoradiogram of a Sagittal Rat Brain Section Hybridized with PSJONA Complementary to mRNA Encoding the 5 Subunit of the C A B A A Receptor

Subunit expression was highest in the granule cell layer of the cerebellar cortex (Cb). 5 subunit was also expressed in the olfactory bulb (OB) by periglomerular neurons (*) and by cells in the granule cell layer (g). Neurons in the olfactory tubercule (Tu), nucleus accumbens (Acb), and thalamus (Th) expressed the 5 subunit. In the neocortex (Cx), 6 subunit was most prominently localized in layers 11-111. In the hippocampal region, neurons in the granule cell layer of the dentate gyrus (dg), as well as a subpopulation of nonpyramidal cells in the pyramidal cell layer, most frequent in the CA1 field, expressed 5 subunit. This subunit was rare in the lateral septum (LS) and in the hypothalamic, pontine (Pn), and medullary (Md) regions. Structures were identified by reference to the brain atlas of Paxinos and Watson (1982). Exposure time, 5 days.

of neurons. Selected regions are described in more de-tail below, and all regions examined are summarized in Table 1.

In the olfactory bulb, the granule cell layer was well labeled as were, to a lesser extent, a subset of periglo-merular cells. The olfactory tubercle cells were also well labeled, with an absence of signal in the islands of Calleja. Expression in the cerebral cortex showed a lami-nar pattern with no obvious regional differences. The cells containing 5 mRNA were heterogeneous in size and shape and were present in all layers, but were most frequent in layers 11-111 followed by layer V (Figure 5A). In most thalamic nuclei, a large and very numerous cell type highly expresses 5 mRNA, which is likely to repre-sent relay neurons (Figure 5B). The caudate-putamen, unlike the globus pallidus, which was not labeled, con-tained numerous 5 mRNA-expressing cells, perhaps representing the medium spiny projection neurons (Fig-ure 5B). Cells in this region were lightly labeled in adult rats and moderately labeled in a 20-day-old rat, suggest-ing developmental regulation of 5 mRNA. In the dentate gyrus it was not possible to determine whether the gran-ule cells or interneurons were labeled. However, in the hippocampus it was obvious from morphology, frequen-cy, and localization that a nonpyramidal cell type was la-beled (Figures 5D and 5E). This cell type decreased in frequency from CAT to CA4. FHypothalamic cells rarely

expressed 6 mRNA, although magnocellular neurons in the supraoptic nuclei were lightly labeled. An infrequent cell type was lightly labeled in the superficial gray layer of the superior colliculus. In the inferior colliculus, a lightly labeled cell type was found in the outer layers of gray matter. The substantia nigra was unlabeled, and the deep cerebellar nuclei and a few brainstem nuclei were lightly labeled. In the cerebellar cortex, the granule cell layer was intensely labeled (Figures 5C and 5F); no signal was seen in Purkinje cells or in any cell type in the mo-lecular layer (Figure 5F).

Neuronal Subpopulations Expressing y2 m R N A Both 6 and y2 mRNA were expressed in neurons. FHow-ever, y2 subunit mRNA was more widely and abun-dantly expressed than mRNA encoding the 5 subunit (Table 1). The Y2 mRNA was most abundant in the olfac-tory bulb, olfactory tubercle, ventral forebrain, neocor-tex, globus pallidus, hippocampal formation, inferior colliculus, cerebellar cortex, and deep cerebellar and brainstem nuclei. Importantly, in several brain regions, different neuronal subpopulations expressed y2 mRNA as compared with 6 mRNA. For example, in the olfac-tory bulb, the mitral cells and tufted neurons promi-nently expressed yl mRNA (Figure 6A), in contrast to 5 mRNA, which was not expressed by this cell type. The periglomerular cells expressed y2 mRNA, as was true for

Neuron 332

alv

Figure 5. Dark-Field and Bright-Field Autoradiograms of Sagittal Rat Brain Sections Hybridized with [^^SJDNA Compiennentary to mRNA En-coding the 5 Subunit of the GABAA Receptor Dorsal is toward the top and rostral is to the left. In (A), the 6 subunit, expressed in all layers of the neocortex, was most highly expressed by neurons in layers 11-111. In (B), neurons of both caudate-putaman (CPu) and thalamus (Th) expressed the 5 subunit (day 20 rat). In (C) and (F), only granule neurons located in the granular cell layer (gel) of the cerebellar cortex express 6 subunit. Purkinje neurons (F, arrowheads) showed little or no 5 subunit expression. In (D) and (E), a subpopulation of nonpyramidal cells most frequent in the pyramidal cell layer (pyr) of the CAT field expressed the subunit. Exposure time, 21 days. Bar in (C), 200 |jm (A-C); bar in (F), 50 lm. Other abbreviations: alv, alveus; fi, fimbria; LV, lateral ventricle; ml, molecular layer; or, oriens layer; rad, radiatum layer; wm, white matter.

6 mRNA; however, as this cell population is heteroge-neous, it was not possible to determine whether the same population was expressing both mRNAs. In the forebrain, cells in the anterior olfactory nuclei and olfac-tory tubercles (excluding islands of Calleja) expressed y2 mRNA. In this region, the expression patterns of y2 and 8 mRNA resembled each other except that 5 mRNA-

expressing cells were less well labeled relative to 5 mRNA-expressing cells elsewhere. In the neocortex, cells heterogeneous in size and shape, present in all six layers, expressed y2 mRNA. No obvious areal or laminar pattern in neocortex was observed, unlike 5 mRNA ex-pression in neocortex (Figure 5A). In the hippocampus, the pyramidal neurons as well as nonpyramidal neurons

Novel GABAA Receptor Subunits 333

Table 1. Summary of Regional and Cellular Locations of mRNAs Encoding the 5 and y l Subunits of the GABAA Receptor in the Rat Brain

Subunit

Regional/Cell Type 6 Y2

Neocortex + 4-+ (ll-lll) + (IV) + + (V-VI)

+ + + (all layers)

Olfactory bulb Periglomerular + + + Tufted 0 + + + + Mitral 0 + + + + Internal Granule + -n- ( + )

Anterior olfactory nucleus + + + + Olfactory tubercule + + + + Accumbens + -1- -1-Caudate-putamen + + Globus pallidus 0 -1- + Hippocampal formation

Subiculum -1- + -1-Hippocampus

Pyramidal 0 + + + Nonpyramidal -h(CA1>2>3>4) -h -1- -1-Dentate gyrus + + + +

Thalamus + + + + + Preoptic area 0 + + Superoptic nucleus + + + + Substantia nigra 0 + + Superior colliculus + + Inferior colliculus + + -1- -1-Cerebellar cortex

Purkinje 0 + -1- -1- + Basket/stellate 0 + Granule + + -1- + ( + ) Golgi M 0 0

Deep cerebellar nuclei + + + + Pons/medulla -1- -1- -1- + + + + +, very high; + + +, high; + +, moderate; +, low; 0, unde-tectable expression; (+), possible expression. The 8 scale is not com-parable to the Y2 scale. Regions not listed should not be considered as nonexpressing.

were labeled. While the y l and 5 patterns were distinct for hippocampus, it was not possible in the dentate gy-rus to determine whether granule cells and/or interneu-rons were labeled for either probe. In the globus palli-dus, y2 mRNA-expressing cells were numerous and well labeled (Figure 6B), whereas little or no 5 mRNA expres-sion was seen.

In the basal forebrain, a subset of cells in the diagonal bands of Broca and in the preoptic area, especially the supraoptic nuclei (Figure 6C), were well labeled. In tha-lamic nuclei, y l mRNA was expressed presumably by the same cell population (data not shown) that contains 6 transcripts. However, the relative abundance of y l mRNA was low in contrast to 8 mRNA, which was very highly expressed in this region (Figure 4). In the substan-tia nigra, cells in both pars compacta and reticulata con-tained y2 mRNA (Figure 6D). In the inferior colliculus, a numerous cell type distributed apparently randomly throughout the gray matter was labeled. Deep cerebel-lar (Figure 6E) and brainstem nuclei, including vestibu-lar, inferior olive, and trigeminal nuclei, prominently ex-

pressed Y2 mRNA. In the cerebellar cortex, the Purkinje cells were very well labeled, and a population of cells in the molecular layer, presumably basket/stellate neu-rons, were labeled to a lesser extent (Figure 6F). This is in strong contrast to 5 mRNA, which occurs only in the granule cell layer (Figures 5C and 5F).

Discussion

The Novel Subunits Define Two GABAA Receptor Subtypes Neurons utilizing GABA as a neurotransmitter are ex-tremely numerous and exist at every level of the neu-roaxis (Mugnaini and Oertel, 1985). Accordingly, the receptors targeted by this transmitter (GABAA and GABAB) are ubiquitously distributed. By analogy to other transmitter receptor systems and evidenced by molecu-lar cloning of GABAA receptor subunits, GABA recep-tors display a subtype heterogeneity of unknown com-plexity and functional diversity. Evidence that the novel subunits, y l and 5, are members of the GABAA recep-tor family is shown by their structural similarity (Figure 2) and functional signature (Figure 3). These subunits seem to define two major GABAA receptor subtypes, as judged from their different expression patterns in rat brain (Table 1). In discussing their relation to previously proposed classifications and distributions of GABAA receptor subtypes, we have relied primarily on maps de-rived from in vitro receptor autoradiography (Young and Kuhar, 1979; Unnerstall et al., 1981; Schoch etal., 1985). We acknowledge the wealth of data obtained by immu-nocytochemistry (Schoch et al., 1985; Richards et al., 1987; deBlas et al., 1988), but recognize that no charac-terized antibodies against the y2 or 5 subunits exist. Hence, no distinction of receptor subtypes containing these subunits has been achieved by this technique.

The y2 Subunit-Containing Subtype The y2 subunit-containing subtype is abundant and present in many regions as judged from cellular yl mRNA content and distribution. The expression pattern of yl mRNA resembles that of the al subunit mRNA (Sequler et al., 1988; B. D. Shivers, unpublished data). Furthermore, our own observations show that mRNA encoding a p subunit variant (p2; Ymer et al., 1989) codistributes with many of these same neurons (e.g., mitral, hippocampal pyramidal, and cerebellar Purkinje neurons). Thus, one version of the GABAA receptor may contain al, P2, and yl subunits of unknown stoichiometry.

As described recently, the yl subunit is important for BZ binding and responsiveness (Pritchett et al., 1989). In concordance with this result, yl mRNA is located in brain regions shown by in vitro autoradiography to con-tain prominent BZ binding (Young and Kuhar, 1979; Un-nerstall et al., 1981; Schoch et al., 1985). These regions include the olfactory bulb, anterior olfactory nuclei, neo-cortex, hippoccampal formation, preoptic area, globus pallidus, inferior colliculus, substantia nigra, cerebel-lum, and several brainstem nuclei (Table 1; Figure 6). We

Neuron 334

Figure 6. Dark-Field Autoradiograms of Rat Brain Sections Hybridized with p^SJcRNA Complementary mRNA Encoding the y2 Subunit of the GABAA Receptor (A) A horizontal section; (B-F) sagittal sections with dorsal to the top and rostral to the right. In (A), mitral neurons in the mitral cell layer (M) and tufted neurons in the external plexiform layer (EPL) as well as neurons in the periglomerular region (Gl) express this subunit. In (B), the y l subunit was highly expressed in many neurons of the globus pallidus (GP) and, to a lesser extent, in neurons of the caudate-putamen (CPu). in (C), the y l subunit was highly expressed by the magnocellular neurons of the supraoptic nucleus (SO) and by many neurons in the lateral preoptic area (LPO). In (D), neurons of both subdivisions of the substantia nigra, i.e., compacta (SNC) and reticulata (SNR), ex-pressed the yl subunit. In (E), the neurons of the nucleus interpositus (INt), a deep cerebellar nuclear region, highly express the yl subunit. In (F), the y2 subunit is most highly expressed by Purkinje neurons in the Purkinje cell layer (pel) and are, to a lesser extent, expressed by presumptive basket/stellate neurons located in the molecular layer (ml). Exposure time, 8 days. Bar, 200 nm. Other abbreviations: gel, granule cell layer; opt, optic tract; wm, white matter.

especially wish to emphasize the congruence between the previously observed BZ binding sites in the cerebel-lar molecular layer (presumably on Purkinje neuron dendrites) and our observation of high y2 expression in Purkinje neurons (Figure 6F).

Our current hypothesis sees the yl subunit as part of the G A B A A / B Z receptor. This receptor has been shown to consist oftwo subtypes termed B Z ! and B Z I I by classic pharmacology (reviewed in Sieghart, 1985). However, our experiments substituting different a subunits suggest


an even wider variety of Y2-containing G A B A A / B Z receptors (D. B. Pritchett et al., unpublished data). Re-garding the role of the y2 subunit-containing subtype in information processing, it is noteworthy that it is most highly expressed in principal neurons, in which this sub-type may contribute importantly to regulating the over-all excitability of these output neurons.

The 5 Subunit-Containing Subtype Little is known about the properties of the subtype to which the 5 subunit belongs. Its distribution, especially in the cerebellar cortex (Figure 5C), resembles that of the high affinity G A B A A receptor sites labeled by pH]mus-cimol and lacking BZ binding (Palacios et al., 1980,1981; Unnerstall et al., 1981). Allowing for different cellular lo-cations of receptor protein and mRNA, other regional congruities would include the thalamus, the dentate gy-rus, the CA1 region of the hippocampus, and the olfac-tory bulb. Consistent with the notion that the 5 subunit participates in the formation of G A B A A receptors that lack BZ binding, this subunit cannot replace y2 in the in vitro generation of high affinity BZ binding sites (D. B. Pritchett and R H. Seeburg, unpublished data). How-ever, we cannot exclude the possibility that 5-contain-ing G A B A A receptors represent previously unrecognized subtypes.

Regarding possible subunit companions for 8, none of the known subunits show a matching distribution, leav-ing open the possibility that 5-compatible subunits have not been isolated. Alternatively, 5-containing subtypes may be assembled from subsets of known G A B A A receptor subunits. For example, the al subunit has been localized in the cerebellar granule cell population in rat (Sequier et al., 1988; B. D. Shivers, unpublished data) and cow (Siegel, 1988; Wisden et al., 1988). Of the p subunits, pi has been localized in bovine cerebellar granule cells (Siegel, 1988) and we have also localized p2 in this region of the rat brain (B. D. Shivers, unpub-lished data).

Regarding the role of the 6-containing G A B A A recep-tors in information processing, we note that 5 mRNA is most prominently expressed in neurons whose dendritic processes exist in complex synaptic arrangements with other neurons, called glomeruli or rosettes. Specifically, we refer to the periglomerular cells of the olfactory bulb, the thalamo-cortical relay neurons, and the granule cells of the cerebellar cortex (Figure 5). The effect of G A B A release in these networks is to limit the spread of excita-tory impulses to neighboring cells, hence creating the surround inhibition property first observed in the ret-ina (Eccles et al., 1984; Jones, 1985; Nieuwenhuys, 1985).

In conclusion, the characterization of new G A B A A receptor subunit classes has led to the description of two distinct subtype families whose special contributions to brain function remain to be addressed. Our current ef-forts are directed at determining the extent of subtype diversity and mapping the developmental and spatial ex-pression patterns of these receptors on neurons and giia (Bormann and Kettenmann, 1988). Combined with the analysis by electrophysiology and pharmacology of re-

combinantly expressed subtypes, such an effort should provide a molecular description of the diverse roles that G A B A A receptors subserve in brain function.


Molecular Cloning A cDNA library (10^ recombinant phage) constructed in XgtlO fronn polyadenylated rat forebrain RNA was screened with a 96-fold degenerate ^^P-labeled 23-mer oligonucleotide encoding a con-served octameric peptide sequence in M2 of GABAA receptor subunits, 5'-AC(A,C)AC(A,T)GT(G,T)CT(A,C,C)AC(A,C)ATGAC(A,C)-AC-3'. Only indicated third position choices were included. Known subunits were Identified using a and 3 subunit-specific oligonucle-otides (Levitan et al., 1988; Ymer et al., 1989). cDNAs hybridizing to the 23-mer but not to the subunit-specific oligonucleotides were sequenced in X,gt10 or after subcloning into M13 vectors (Vieira and Messing, 1987) by the chain termination method (Sanger et a!., 1977). Sequencing reactions using the 23-mer oligonucleotide were perfomed with 0.5 (iM primer, and reactions were at 55°C when recombinant X DNA was used as template. One of the clones iso-lated in this manner was the 6 cDNA. The rat y2 cDNA was isolated by screening the same library with 32p.|abeled cloned cDNA en-coding the human y l subunit (Pritchett et al., 1989).

Expression and Electrophysiology Human embryonic kidney cells (293) grown on coverslips were transfected at high efficiencies with recombinant expression vector DNA, using a CaP04 precipitation technique (Chen and Okaya-ma, 1987), as described earlier (Pritchett et al., 1988). The func-tional features of the vector are described by Eaton et al. (1986). For electrophysiology, cells were used 48 hr after transfection. For recording of rnembrane currents, cells on coverslips were trans-ferred to the stage of an inverted microscope, where they were maintained at about 25°C in a continuously perfused recording chamber. The standard salt solution contained 116 mM NaCI, 11.1 mM glucose, 26.2 mM NaHCOj, 5.4 mM KCI, 1.8 mM CaCb, and 0.8 mM MgCl2, buffered at pH 7.2 with 5 mM HEPES. GABA and pentobarbital were added to the standard salt solution. All voltage-clamp recordings were made with an EPC-7 patch-clamp amplifier (List Electronics, Darmstadt, FRG) using the tight-seal, patch-clamp technique in its whole-cell recording configuration (Hamill et al., 1981). Recording pipettes were fabricated from borosilicate capil-laries (Hilgenberg, Malsfeld, FRG) with resistances ranging from 2 to 5 Mf2 and filled with a solution containing 130 mM CsCI, 1 mM MgCl2, 0.5 mM CaClj, 5 mM EGTA, and 10 mM HEPES.

In Situ Hybridization To localize 6 mRNA, a Hindlll-BamHI fragment (nucleotides 456-1465) was excised from the cloned 5 cDNA and labeled with [a-35S]cJATP (1000 Ci/mmol; Amersham) to a specific activity of 10 cpm/ng using random primers according to the method of Feinberg and Vogelstein (1983). To localize y2 mRNA, RNA probes were tran-scribed from linearized Bluescript plasmid DNA containing the coding region of rat y l cDNA on a 1500 bp EcoRI fragment. An-tisense or sense RNA was synthesized using [a-"S]CTP (1000 Ci/ mmol, Amersham) to a specific activity of 3 x 10® cpm/jig accord-ing to the method of Melton et al. (1984). Transcripts were hydro-lyzed to an average length of 100 nucleotides.

In situ hybridization was performed essentially as described by Shivers et al. (1986). Frozen, parasagittal rat brain sections (10 jim) were thaw-mounted onto organosilianized slides, fixed in 3% neutral-buffered paraformaldehyde containing 0.02% diethylpyro-carbonate, rinsed in PBS, dehydrated in alcohol, air dried, and stored frozen at -70°C. Radiolabeled probe (3000 cpm/^l of ^ S-labeled 5 DNA or of ^^S-labeled y l cRNA) was dissolved in hybrid-ization buffer, 50% formamide containing 14 mM P-mercapto-ethanol (for details see Shivers et al., 1986). Fifty microliters was applied to each of the sections, which were then covered with Parafilm for hybridization in sealed, humidified chambers at 42°C for 3 days. The sections were washed twice for 10 min each in 2x SSC containing 0.05% inorganic sodium pyrophosphate and 14 mM 3-mercaptoethanol at room temperature. Next, the sections were

Neuron 336

washed twice for 15 min each in 0.1 x SSC containing 0.05% inor-ganic sodium pyrophosphate and 14 nnM 3-iTiercaptoethanol at 55°C for DNA, or at 65°C for cRNA, and dehydrated in 70%, 90%, and absolute ethanol. Following air drying, the sections were ex-posed to P-max film (Amersham) and/or dipped in NTB2 emulsion (Kodak, diluted 1:1 with water). After development of emulsion au-toradiograms, the sections were stained in 2% fast green and 0.5% cresyl violet. The 5 subunit localization was confirmed using a cRNA hybridizing to the 3'untranslated region of the mRNA, which produced the same pattern of hybridization as the DNA hybridizing to the coding portion of the mRNA. The y2 subunit localization was confirmed using a sense probe that produced no hybridization pat-tern and with a cRNA hybridizing to the 3' untranslated region of the mRNA that produced the same hybridization pattern as the cRNA hybridizing to the coding portion of the mRNA.

Acknowledgments

We appreciate the technical assistance of G. Muncke, M. Stein, and A. Herb and thank J. Rami for typing the manuscript. We acknowl-edge the contributions of Dr. H. Kettenmann, in whose laboratory the electrophysiological recordings were made, and Dr. D. B. Pritchett, who provided the transfected 293 cells. Received May 2, 1989; revised June 26, 1989.

References

Bormann, J., and Kettenmann, H. (1988). Patch-clamp study of y-aminobutyric acid receptor CI" channels in cultured astrocytes. Proc. Natl. Acad. Sci. USA 85, 9336-9340. Chen, C., and Okayama, H. (1987). High efficiency transformation of mammalian cells by plasmid DNA. Mol. Cell. Biol. 7, 2745-2751. deBlas, A. L., Vitorica, J., and Friedrich, R (1988). Localization of the CABAA receptor in the rat brain with a monoclonal antibody to the 57000 M^ peptide of the GABAA receptor/benzodiaze-pine/CI" channel complex. J. Neurosci. 8, 602-614. Eaton, D. L , Wood, W. !., Eaton, D., Hass, P. E., Hollingshead, R, Wion, K., Mather, J., Lawn, R. M., Vehar, G. A., and Gorman, C. (1986). Construction and characterization of an active factor III variant lacking the central one-third of the molecule. Biochemistry 25, 8343-8347. Eccles, J. C , Ito, M., and Szentagothai, J. (1964). The Cerebellum as a Neuronal Machine (Berlin: Springer Verlag), p. 223. Feinberg, A. R, and Vogelstein, B. (1983). A technique for radio-labeling DNA restriction endonuciease fragments to high specific activity. Anal. Biochem. 132, 6-13. Grenningloh, G., Rienitz, A., Schmitt, B., Methfessel, C., Zensen, M., Beyreuther, K., Gundelfinger, E. D., and Betz, H. (1987). The strychnine-binding subunit of the glycine receptor shows homol-ogy with nicotinic acetylcholine receptors. Nature 328, 215-220. Hamill, O. R, Marty, A., Neher, E., Sakmann, B., and Sigworth, F. J. (1981). Improved patch-clamp techniques for high resolution cur-rent recording from cells and cell-free membrane patches. PflOger's Arch. Ges. Physiol. 397, 85-100. Huganir, R. L., and Greengard, R (1987). Regulation of receptor function by protein phosphorylation. Trends Neurosci. 8, 472-477. Jones, E. G. (1985). The Thalamus (New York: Plenum Publishing Corp.), p. 176. Levitan, E. S., Schofield, R R., Burt, D. R., Rhee, L. M., Wisden, W., Kohler, M., Fujita, N., Rodriguez, H., Stephenson, E A., Darlison, M. G., Barnard, E. A., and Seeburg, R H. (1988). Structural and functional basis for GABAA receptor heterogeneity. Nature 335, 76-79. Melton, D. A., Krieg, R A., Rebagliati, M. R., Maniatis, T, Zinn, K., and Green, M. R. (1984). Efficient in vitro synthesis of biologically active RNA and RNA hybridization probes from plasmids contain-ing a bacteriophage SP6 promoter. Nucl. Acids Res. 7, 7035-7056. Mugnaini, E., and Oertel, W. H. (1985). An atlas of the distribution of GABAergic neurons and terminals in the rat CNS as revealed by GAD immunohistochemistry. In Handbook of Chemical Neuro-

anatomy, Vol. 4: GABA and Neuropeptides in the CNS, Part 1, A. Bjorklund and T. Hokfelt, eds. (Amsterdam: Elsevier), pp. 436-608. Nieuwenhuys, R. (1985). Chemoarchitecture of the Brain. (Berlin: Springer Verlag), p. 167. Olsen, R. W., and Venter, C. J. (1986). Benzodiazepine/GABA Receptors and Chloride Channels: Structural and Functional Prop-erties (New York: Alan R. Liss). Palacios, J. M, Young, W. S., III., and Kuhar, M. J. (1980). Autoradio-graphic localization of y-aminobutyric acid (GABA) receptors in rat cerebellum. Proc. Natl. Acad. Sci. USA 77, 670-674. Palacios, J. M., Wamsley, J. K., and Kuhar, M. J. (1981). High-affinity GABA receptors: autoradiographic localization. Brain Res. 222, 285-307. Paxinos, G., and Watson, C. (1982). The Rat Brain in Stereotaxic Coordinates (Sidney: Academic Press, Inc.). Pritchett, D. B., Sontheimer, H., Gorman, C. M., Kettenmann, H., Seeburg, R H., and Schofield, R R. (1988). Transient expression shows ligand gating and allosteric potentiation of GABAA receptor subunits. Science 242, 1306-1308. Pritchett, D. B., Sontheimer, H., Shivers, B. D., Ymer, S., Ketten-mann, H., Schofield, R R., and Seeburg, R H. (1989). Importance of a novel GABAA receptor subunit for benzodiazepine pharma-cology. Nature 338, 582-585. Richards, J. G., Schoch, R, Haring, R, Takacs, B., and Mohler, H. (1987). Resolving GABAA/benzodiazepine receptors: cellular and subcellular localization in the CNS with monoclonal antibodies. J. Neurosci. 7, 1866-1886. Sanger, F, Nicklen, S., and Coulson, A. R. (1977). DNA sequencing with chain-terminating inhibitors. Proc. Natl. Acad". Sci. USA 74, 5463-5467. Schoch, R, Richards, J. G., Haring, R, Takacs, B., Stahli, C., Staehe-lin, T, Haefely, W., and Mohler, H. (1985). Colocalization of GABAA receptors and benzodiazepine receptors in the brain shown by monoclonal antibodies. Nature 374, 168-171. Schofield, R R., Darlison, M. G., Fujita, N., Burt, D. R., Stephen-son, F. A., Rodriguez, H., Rhee, L. M., Ramachandran, J., Reale, V., Glencorse, T A., Seeburg, R H., and Barnard, E. A. (1987). Se-quence and functional expression of the GABAA receptor shows a ligand-gated receptor super-family. Nature 328, 221-227. Sequier, J. M., Richards, J. G., Malherbe, R, Price, G. W , Mathews, S., and Mohler, H. (1988). Mapping of brain areas containing RNA homologous to cDNAs encoding a and P subunits of the rat GABAA y-aminobutyrate receptor. Proc. Natl. Acad. Sci. USA 85, 7815-7819. Shivers, B. D., Schachter, B. S., and Pfaff, D. W. (1986). In situ hy-bridization for the study of gene expression in the brain. Meth. En-zymol. 124, 497-510. Siegel, R. E. (1988). The mRNAs encoding GABAA/benzodiaze-pine receptor subunits are localized in different cell populations of the bovine cerebellum. Neuron 1, 579-584.

Sieghart, W. (1985). Benzodiazepine receptors: multiple receptors or multiple conformations. J. Neural Transmission 63, 191-208.

Sontheimer, H., Becker, C.-M., Pritchett, D. B., Schofield, R R., Grenningloh, G., Kettenmann, H., Betz, H., and Seeburg, R H. (1989). Functional chloride channels by mammalian cell expression of rat glycine receptor subunit. Neuron 2, 1491-1497.

Stephenson, F. A. (1988). Understanding the GABAA receptor: a chemically gated ion channel. Biochem. J. 249, 21-32. Unnerstall, J. R., Kuhar, M. J., Niehoff, D. L., and Palacios, J. M. (1981). Benzodiazepine receptors are coupled to a subpopulation of y-aminobutyric acid (GABA) receptors: evidence from a quan-titative autoradiographic study. J. Pharmacol. Exp. Therap. 218, 797-804.

Vieira, J., and Messing, J. (1987). Production of single-stranded plas-mid DNA. Meth. Enzymol. 153, 3-11. von Heijne, G. (1986). A new method for predicting signal se-quence cleavage sites. Nucl. Acids Res. 14, 4683-4690.


Wisden, W., Morris, B. J., Darlison, M. G. , Hunt, S. P., and Barnard, E. A. (7988). Distinct GABA^ receptor a subunit mRNAs show differential patterns of expression in bovine brain. Neuron 7, 937-947.

Ymer, S., Schofield, P. R., Draguhn, A., Werner, P., Kohler, M., and Seeburg, P. H. (1989). GABA^ receptor p subunit heterogeneity: functional expression of cloned cDNAs. EMBO J. 8, 1665-1670. Young, W. S., ill, and Kuhar, M. J. (1979). Autoradiographic local-ization of benzodiazepine receptors in the brains of humans and animals. Nature 280, 393-395.

Reprinted f romTrends in Pharmacological Sciences

elsevier publications

CAMBRIDGE

476

The GABAA receptor: molecular biology reveals a complex picture It has become textbook knowledge that the major inhibitory neuro-transmitter receptor, the GABAA receptor, is a multisubunit recep-tor-channel complex which can be allosterically modulated by two important classes of drug, the benzodiazepines and the barbitu-rates. The primary structures of the GABAA (Ref. 1) and glycine^ receptors were defined in 1987, showing that these two receptors were members of a ligand-gated ion channel receptor superfamily. Since then, there have been major efforts to determine the number, stoichiometry and function of the GABAA receptor subunits. How-ever, instead of clarifying further the 'classic' pharmacological pic-ture, this recent molecular bio-logical onslaught has revealed unexpected complexities.

Early work^ using affinity puri-fied GABAA receptor preparations suggested the receptor complex to be composed of only two sub-units, the 50-53 kDa a subunit and the 55-57 kDa (3 subunit. However, recent discoveries by DNA homology cloning of novel receptor subunits (y, 6 and e) (Refs 4 and 5; P. H. Seeburg et al, unpublished) and subtypes of subunits (a l -a4 , |31-p3, yl, y2 and others) (Refs 6-9, P. H. Seeburg et al, unpublished) provides many combinatorial possibilities for the formation of functional receptors.

How near has this taken us towards defining GABAA receptor in-vivo stoichiometry, and under-standing modulatory mechanisms at the molecular level? Reconsti-tution studies with the cloned a and P subunits have not yielded receptors that mimic in-vivo

pharmacology; indeed, some studies have demonstrated agonist responses elicited by benzodiaz-epine antagonists and inverse agonists! However, recent work has demonstrated that coexpres-sion of a and |3 subunits with the y2 subunit yields a functional receptor which can be modulated in a predictable way by benzodiaz-epines. A combination of such expression and localization studies is starting to unravel the functions of individual GABAA receptor subunits.

Homomeric channel formation Electrophysiological and/or

pharmacological analyses of ex-pressed receptor subunits in either Xenopus oocytes^® or cul-tured mammalian cells^^ have shown that functional homomeric, or single subunit, receptors can form. All GABAA receptor sub-units examined to date form homomeric channels ' °' ^ each of which is activated by GABA and potentiated by barbiturates such as pentobarbital. These results are surprising since photoaffinity labelling studies had earlier indi-cated that GABA analogues such as [^HJmuscimol bound the (3-subunit only^ " ; however, they are consistent with suggestions that the C\~ channel itself is the site of action of the barbiturates.

Homomeric channels are rela-tively inefficient in that only small whole cell currents are ob-served^®' ^ (and high concentra-tions of GABA are required for gating, see below). These currents increase in magnitude as addi-tional subunits are included in the expression experiments. Homo-

meric channel formation has also been demonstrated for the neur-onal nicotinic acetylcholine receptor^^ and the strychnine binding (48 kDa) subunit of the glycine receptor^ ' . In the latter case, relatively efficient channel formation was observed.

Since homomeric GABAA re-ceptors require higher concen-trations of GABA for channel

and none of the sub-units show differential GABA activation, it is likely that more than one receptor subunit plays a role in channel activation. In-vivo receptors show a positive co-operativity of GABA activation (Hill coefficients >1.0) which may reflect the ability of several dif-ferent subunits to respond to the neurotransmitter. To date, expres-sion of cloned receptor subunits has not revealed this property, although positive cooperativity has been seen with expressed gly-cine receptors^^.

Single channel patch clamp analysis of homomeric a l , oc2, q:3 or pi GABAA receptor subunits also reveals that many of the single channel properties of these receptors are qualitatively the same as or similar to those seen in normal heterooligomeric recep-tors^®. Multiple subconductance states, with preferred conduc-tances of 19 pS and 28 pS for all three a-subunits and 18 pS and 27 pS for the (3-subunit have been observed. These values compare favourably with the GABAA re-ceptor subconductance states seen in spinal cord neurons^®.

Benzodiazepine binding sites Despite evidence that the a-

subunit forms the benzodiazepine binding site ' ' , detailed analyses of the benzodiazepine respon-siveness of both human^^ and

© 1989, Elsevier Science Publishers Ltd. (UK)- 0165 - 6147/89/$02.00

TiPS - December 1989 [Vol. 10] 477

recombinant a - and P-subunit receptors has failed to demonstrate any benzodiazepine potentiation of GABA-activated C r flux. However, these results contrast with those obtained using rat a l - and pi-subunit cDNAs where Malherbe et al}^ demonstrated benzodiazepine potentiation of expressed recep-tors, although in only two out of three GABA-responsive oocytes. It is not clear why this should be so since GABAA subunits from rat, bovine and human brain share - 9 9 . 5 % (a-subunit) and - 9 7 . 5 % (P-subunit) primary sequence identity^'^^'^^. In both cases, there are only two amino acid differ-ences in the extracellular domain, the presumed site of benzodiaz-epine action. A second unique phenomenon has been observed with rat cDNAs: benzodiazepine antagonists and inverse agonists (which block or reverse GABA potentiation in vivo, respectively) potentiated GABA-induced re-sponses^^. This observed reversal of known effects of benzodiaz-epine antagonists and inverse agonists suggests that the recep-

tors are not fully functional. Never-theless, there are some precedents for this type of pharmacology. Chick brain mRNA, expressed in oocytes, forms GABAA receptors in which the benzodiazepine in-verse agonist DMCM potentiates rather than reduces GABA-acti-vated Cl~ flux^^. Similarly, in primary cultures of neonatal rat astrocytes, benzodiazepine agon-ists potentiate GABA responses in a manner reversible by benzodiaz-epine antagonists^^. However, two benzodiazepine inverse agonists (DMCM and Ro 22-7497) were seen to potentiate, rather than depress GABA-induced responses^^. The subunits responsible for these atypical inverse agonist responses in glial cells have not yet been identified. Comparison of natural (neonatal rat cortical neurones) and reconstituted (human) GABAA receptors has shown similar inhibition of GABA-elicited CI" currents by both DMCM and 4'-chlorodiazepam^®. The specific profile of inverse agonist activities of 4'-chloro-diazepam has revealed yet an-other modulatory site of GABAA

receptors.

Novel subunits, new functions The cDNA cloning of novel

GABAA receptor subunits has provided several putative ligand-gated Cl~-conducting ion channel receptor subunits e.g. a4, yl , 8 (Refs 4-9). Expression of these cDNAs in transfected mammalian cells has allowed both electro-physiological and pharmacological analyses to be used in determin-ing their function. Homomeric y l or 8 subunits form GABA- but not glycine-activated CI" channels, thus defining two new GABAA receptor subunits^'^. Coexpression of a l ^ l y l subunits yields a chan-nel with high affinity benzodiaz-epine binding sites which either potentiate (agonists), reduce (in-verse agonists) or do not affect (antagonists) the GABA-induced c r currents^. Homomeric yl sub-units, or even pairwise Ck:lY2 or ^ l y l combinations were not potentiated by benzodiazepines, indicating a requirement for all three subunits in the formation of functional benzodiazepine recep-tors. In-situ hybridization shows that the y2 subunit is encoded within various neuronal subsets which are known to contain high

478

affinity benzodiazepine binding sites and to express a- and f3-subunit mRNAs '2^28 similar analysis of 6 subunit mRNA localization shows a good corre-lation with the distribution of high affinity muscimol binding sites^.

i Types I and II Benzodiazepine receptors

Photoaffinity labelling with [^H]flunitrazepam has indicated that t h g ^ ^ e p t o r a subtmit forms the benzodiazepine bind-ing site 'i ' . The photolabelled a subunit subtypes correlate well with the pharmacologically de-fined type I and type II benzo-diazepine subtypes^ ' . Thus the presence of specific a-subunit subt)^es has been suggested to generate the observed benzodiaz-epine receptor subtypes. Com-pared to type II benzodiazepine receptors, type I receptors show a 5-10-fold greater affinity for certain benzodiazepine-related drugs such as 2-oxoquazepam, the p-carbolines (e.g. p-CCM) and the triazolopyridazine CL-218872 (Ref. 19). Type I benzodiazepine receptors have a specific neuronal localization, being found most abundantly in the cerebellum. Type I receptors are also found in most brain regions, while type 11 benzodiazepine receptors, which are less widely distributed, represent about 50% of the benzodiazepine binding sites of the hippo campus

Pritchett et al?^ have extended the functional analysis of the Y2 subunit and examined its pharmacological properties when expressed with the various a-sub-unit subtypes (al, a l and a3). The three possible a^y receptor com-binations all display high-affinity benzodiazepine • binding sites, although with Hill coefficients of unity, none of these combinations is truly representative of the in-vivo receptor. None the less, the al(3lY2 combination preferentially shows a temperature-modulated selectivity for the benzodiazepine-related compounds CL-218872, 2-oxoquazepam and ^-CCM. This is the pharmacological profile of type I benzodiazepine receptors, implicated as sites of anxiolytic action with reduced ataxic and depressant effects. The a2piY2 and a3piY2 GABAA receptors have properties more typical of

type II benzodiazepine receptors. The localization of a-subunit

mRNAs '2^2® is also in good agreement with the localization of the type I and t3rpe II benzodiaz-epine receptors, as determined by ligand binding analysis and by [^Hjflunitrazepam ^ photoaffinity labelling. Four putative a-sub--units of the GABAA receptor have been identified by photolabelling, of molectilar mass 51 kDa, 53 kDa, 55kDa and 59kDa^ . The 51 kDa protein has been shown to be a component of type I benzodiaz-epine receptors^% while the type II receptor probably contains the 55 kDa protein^ ' . Indeed, sub-unit-specific peptide antibodies have recently been used to con-firm that the a3 subunit corre-sponds to the 59 kDa [^H]flunitraz-epam-labelled protein^ . Based on colocalization of specific mRNAs and photolabelled proteins, and using the predicted molecular weights of the various subunit glycoproteins, the a l subunit probably encodes the 51 kDa pro-tein and the a2 subunit encodes either the 53 kDa or 55 kDa pro-tein. Thus, even type H benzodiaz-epine receptors are heterogeneoui/T being represented by receptor subtypes containing either a2 or a3 subunits.

• • • The importance of GABAergic

neurotransmission as the major inhibitory pathway is reflected in the multitude of receptor subunits and subtypes that have been iden-tified to date. The combined use of expression and localization studies has enabled the identi-fication and molecular description of several of the physiologically relevant properties of GABAA receptors. As the functional prop-erties of further GABAA receptor subtypes are defined, it should be possible to use this knowledge in the rational screening or design of new therapeutically useful sub-type-specific drugs.

PETER R. SCHOFIELD

Pacific Biotechnology Ltd, 74 McLachlan Ave, Rushcutters Bay 2011, Australia.

References 1 Schofield, P. R. et al. (1987) Nature 328,

221-227 2 Grenningloh, G. et al. (1987) Nature 328,

215-220 3 Sigel, E. and Barnard, E. A. (1984)). Biol.

Chem. 259, 7219-7223

TiPS - December 1989 [Vol. 10J

4 Pritchett, D. B. et al. (1989) Nature 338, 582-585

5 Shivers, B. D. et al. (1989) Neuron 3, 327-337

6 Levitan, E. S. et al. (1988) Nature 335, 76-79

7 Ymer, S., Schofield, P. R., Draguhn, A., Werner, P., Koehler, M. and Seeburg,

; P. H. (1989) EMBO J. 8,1665-1670 8 Ymer, S., Draguhn, A., Koehler, M.,

' Schofield, P. R. and Seebiug, P. H.JEBS Lett, (in press)

9 Lolait, S. J., O'CarroU, A., Kusano, K., Muller, J., Brownstein, M. J. and Mahan, L. C. (1989) FEBS Lett. 246,143-14S

10 Blair, L. A. C , Levitan, E. S., Marshal, J., Dionne, V. E. and Barnard, E. A. (1988) Science 242, 577-579

11 Pritchett, D. B., Sontheimer, H., Gorman, C M., Kettenoann. IL, Seeburg, P. H. and Schofield, P. R. (1988) Science 242,1306-130S

12 Cavalla, D. and Neff, N. H. (19S5) J. Neurochem. 44, 916-921

13 Deng, L., Ransom, R. W. and Olsen, R. W. (1986) Biochem. Bioph^s. Res. Commun. 138,1308-1314

14 Casalotti, S. O., Stephenson, F. A. and Barnard, E. A. (1986) J. BioL Chem. 261, 15013-15016

15 Boulter, J., Connolly, J., Deneris, E., Goldman, D., Heinemann, S. and Patrick, J. (1987) Proc. Natl Acai. Sd. USA 84, 7763-7767

16 Schmieden, V., Grenningloh, G., Schofield, P. R. and Betz, H. (1989) EMBO! 8,695-700

17 Sontheimer, H. et al (1989) Neuron 2, 1491-1497

18 Borman, J., Hamill, O. P. and Sakmann, B. (1987) J. Physiol London 3S5, 243-2S6

19 Sieghart, W. (1989) Trends Pharmacol Sci. 10, 407-411

20 Levitan, E. S., Blair, L. A. C , Dionne, V. E. and Barnard, E. A. (19SS) Netirm 1, 779-791

21 Malherbe, P., Draguhn, A., Multhaup, G., Beyreuther, K., Sakmann, B. and Mohler, H. Neuron (in press)

22 Schofield, P. R., Pritchett. D. B.. Sontheimer, H., Kettenmaim, H. and Seeburg, P. H. (1989) FEBS Lett. 244. 361-364

23 Sigel, E. and Baur. R. (1988) / . Neurmd. 8, 289-295

24 Backus, K. H., Kettenmann, H. and Schachner, M. (19SS) G/ia 1,132-140

25 Puia, G., Santi, M. R.. Vicini, S.. Pritchett, D. B., Seebxirg, P. H. and Costa, E. (1989) Proc. Natl Ac^tl Sd. USA 86, 7275-7279

26 Sigel, R. E. (19SS) Neuron 1. 579-5S4 27 Sequier, J. M.. Richards. J. G..

Malherbe, P., Price, G. W.. Matthews, S. and Mohler, H. {19SS) Proc. Nat! Acsd. Sci. USA 85, 7S15-7819

28 Wisden, W., Morris, B. J., DaxHson. M. G., Hunt, S. P. and Barnard, E, A. (1988) Neuron 1, 937-947

29 Sieghart, W. and Karobath, M. (1980) Nature 285, 2S5-2S7

30 Pritchett, D. B., Luddens, H. and Seeburg, P. H. Science (in press)

31 Stephenson, F, A., Duggan, M, J, and Casalotti, S. O. {1939}''FEBS Lett 243, 358-362

CL-218872: 3-methyl-6-[3-trinuorome{hyl-phenyl]-l,2,4-tria2olo[43l'lpyrida2ine P-CCM: p-carboline-S-carboxylate methyl ester DMCM: methyl-6,7-dimelhoxy-4-elhyl-P-carboIine-3-carbox\'late

\ ( ) l I \H K. \l \lHh R ^ l|>( H I ) " s- : . 4f.n ilus>Ji ISSX ((."""'siiU J l \F: 1989

S p e c i a l Issue: I S R K S h o r t ( o n i n i u i i i c a t i i

352 International Symposium on Receptors and Ion Channels

Huganir, R. L.. Delcour, A. H - Grecngard, P., ami Hess. G. P. (1986). Nature 321, 774-776. Imoto, K., Mcthfcsjcl , C.. Sakmann, B., Mishina, M., Mori, J.. Konno, T., Fukuda, K., Kurasaki. M.,

Bujo, H.. Fujita. J., and Numa, S. (1986). Nature 324, 670-674. Imoto. K.. Busch, C.. Sakmann, B., Mishina, M., Konno, T., Nakai. J.. Bujo. H., Mori, Y., Fukuda. K.,

and Numa, S. (1988). Nature 335, 645-648. Kao, P. N., and Karlin, A. (1986). J. Biol. Chem. 261, 8085-8088. Uv i t an , E. S., Blair, L. A., Dionne. V. E., and Barnard, E. A. (1988a). Neuron I, 779-781. Uv i t an , E. S., Schoficid, P. R., Burt, D. R.. Rhcc, L. M., Wysdcn, W.. Kochlcr, M., Fujita, N., Rodriguez,

H. F., Stephenson, A.. Darlison, M. G., Barnard, E. A., and Seeburg, P. H. (1988b). Nature 335,76-79. Malherbe, P., Draguhn, A., and Mohier, H. (1989). (in preparation). Mamaiaki, C., Stephenson, F. A., and Barnard, E. (1987). EMBOJ. 6, 561-565. Mohier, H., Battersby, M. K., and Richards, J. G. (1980). Proc. Natl. Acad. Set. USA 77, 1666-1670. Mohier, H., and Richards, J. G. (1988). Eur. J. Anaesthesial. 2(suppl.). 15-24. Noda, M., Furutani, Y., Takahashi, H., Toyosato, M., Tanabc, T., Shimizu, S., Kikyotani, S., Kayano,

T., Hirose, T., Inayami, S., and Numa, S. (1983). Nature 305, 818-823. Oberthiir, W., Muhn, P., Baumann, H., Lottspeich, F., Wittmann-Lieboid. B., and Hucho, F. (1986).

E M B O / 5 , 1815-1819. Pritchett, D. B., Sontheimer, H., Gorman . C. M., Keltenmann. H., Seeburg, P. H., and Schofield. P. R.

(1988). Science 242, 1306-1308. Richards. J. G.. Mohier. H., and Haefely, W. (1986). In Panula, Paivarinta, and Soinita (eds.). Neurology

and Neurobiology, Vol. 16, Neurohistochemistry, Modern Methods and Applications (Alan R. Liss. New York), p. 629.

Richards. J. G., Schoch, P.. Haring, P., Takacs, B., and Mohier. H. (1987). J. Neurosci. 7, 1866-1886. Richards, J. G., Malherbe. P., and Mohier, H. (1989). (in preparation). Sakmann, B., Hamill, O. P., and Bormann, J. (1S83). J. Neural Transm. [SuppL] 16, 83-95. Schoch, P.. Haring, P., Takacs, B., Stahli, C., and Mohier, H. (1984). J. Receptor Res. 4. 189-200. Schoch, P., Richards, J. G.. Haring, P., Takacs, fl., Slahli, C , Staehelin, T., Haefely. W., and Mohier,

H. (1985). Nature 314, 168-171. Schofield. P. R.. Darlison. M. G.. Fuji ta. N.. Burt, D. R., Stephenson, F. A., Rodriguez. H., Rhcc, L. M..

Ramachandran, V. R., Glcncorsc, T. A., Seeburg, P. H.. and Barnard. E. A. (1987). Nature 328. 221-227.

Scquier. J. M.. Richards. J. H. G., Malherbe. P.. Price, G. W.. Mathews, S., and Mohlcr. H. (1988). Proc. Natl. Acad. Sci. USA 85. 7815-7819.

Sigel, E., and Barnard, E. A. (1983). J. Biol. Chem. 258, 6965-6971. Sigel, E., and Bauer, R. (1988). J. Neurosci. 8, 289-295. Stelzcr, A., Kay, A. R., and Wong R. K. S. (1988). Science 241, 339-341.

18. S . Ymer,' P. R. Schof ie ld , ' B. D . Shivers,' D . B. Pritchett, ' H. Liiddens,' M . Kbhier,' P. Werner, ' H . Sontheimer,^ H. Kettenmann,^ and P. H. Seeburg. ' Molecu lar Studies of the G A B A ^ Receptor. ( ' L a b o r a t o r y o f M o l e c u l a r N e u r o e n d o c r i n o l o g y , ^ D e p a r t m e n t of N e u r o b i o l o g y , Im N e u e n h e i m e r Fe ld 364, Un ive r s i t a t H e i d e l b e r g , H e i d e l b e r g D - 6 9 0 0 , F R G )

G A B A ( y - a m i n o b u t y r i c a c i d ) - e r g i c n e u r o t r a n s m i s s i o n is p r e d o m i n a n t l y m e d i a t e d b y a g a t e d c h l o r i d e c h a n n e l i n t r in s i c to t h e G A B A ^ r e c e p t o r . T h i s h e t e r o o l i g o m e r i c r e c e p t o r [ r e v i e w e d in S t e p h e n s o n ( 1 9 8 8 ) ] exis ts a t t h e m a j o r i t y o f i n h i b i t o r y s y n a p s e s in t h e v e r t e b r a t e c e n t r a l n e r v o u s s y s t e m ( C N S ) , a n d c a n b e r e g u l a t e d by c l in ica l ly i m p o r t a n t c o m p o u n d s ; f o r e x a m p l e , b e n z o d i a z e p i n e s ( B Z ) a n d b a r b i t u r a t e s ( r e v i e w e d in O l s o n a n d Ven te r ( 1 9 8 6 ) ] . Aff in i ty p u r i f i c a t i o n o f t h e r e c e p t o r i d e n t i f i e d t w o s u b u n i t s , a n a s u b u n i t of 53 k D , o n w h i c h the BZ

Short Communications 353

binding site is thought to be locsted, and a /3 subunit of 57 IcD which can be photoaffinity-labeled by GABA agonists (Sigei et al, 1983; Mohler et al., 1980; Cavalla and Neff, 1985; Deng et al., 1986; Casalotti et al, 1986). Peptide sequence was obtained which facilitated isolation of bovine cDN A clones encoding both subunits (Schofield et al., 1987). The sequences predicted from cDNAs showed about 35% amino acid homology to each other and were homologous to other ligand-gated receptor subunits , namely the nicotinic acetylcholine receptor (nAchR) and glycine receptor 48 kD subunit . Human and rat c D N A clones have also been isolated, and across species there is greater than 95% amino acid homology.

As a consequence several common structural features have emerged for the ligand-gated ion channel super family. All have four putative transmembrane (M1-M4) regions. The presence of a signal peptide predicts the N-terminal region to be extracellular, with a /3-structural loop formed by the disulphide bonding of two conserved cysteine residues. The Ml contains an invariant proline residue which is speculated to introduce a protusion into the channel lumen that might keep the channel closed in the absence of neurotransmitter. There is an abundance of hydroxyl side chains in M2 which may be important in the formation of a hydrophilic lining of the channel, conducive to ion flow. A clustering of positively charged side-chains occurs at both ends of the postulated membrane-spanning domains of the G A B A A and glycine receptors which are presumed to form the channel mouth. Most of these clustered basic residues are absent or exchanged for acidic residues in the corresponding nAchR locations. Between the M3 and M4 there is a large intracellular loop region where the greatest sequence divergence occurs. This may be a site for intracellular control mechan-isms of channel activity involved in, for example, desensitization. In fact the p subunit has here a putative control site for serine phosphorylation by protein kinase A.

Expression of GABA^ receptors has been achieved after injection of poly A* m R N A into Xenopus oocytes, allowing electrophysiological analysis using the voltage-clamp tech-nique (Smart et al., 1983; Houamed et al., 1984). This system was used to confirm the cloned a and /3 subunit cDNAs as GABA^ receptor forming. Properties observed for this receptor include competitive inhibition by bicuculline, potentiation of GABA response by the barbitur-ate pentobarbital , inhibition by the channel blocker picrotoxin, anion selectivity, and voltage-dependent gating. The recombinant receptors are, however, not identical with native GABA/v, receptors. Hill coefficients derived from the data were approximately 1, rather than 1.4 or 2 as expected, and BZ potentiation was not reliably observed. It was thought that additional subunit subtypes may confer these properties and contribute to G A B A A receptor heterogeneity, as suggested for the BZ binding site by electropharmacological and autoradio-graphic studies (Olson and Venter, 1986).

Further screening of c D N A libraries was carried out using a variety of probes. One oligodeoxyribonucleotide probe was designed against the possible nucleotide sequences encoding a conserved eight amino acid sequence in the M2 region of GABA and glycine subunits. By analogy with the nAchR, this sequence was proposed to form part of the channel lumen. Two additional a subunit (02 and Oj) clones were isolated (Levitan et al, 1988). The three a subunits have about 70% amino acid homology to each other and 35% to the /3 subunit. Expression of the corresponding cDNAs in oocytes showed that either the a j or the "3 polypeptide could be substituted for a^ to yield GABA-evoked currents with the expected pharmocology; however, an apparent 30-fold variance in sensitivity between these GABA^ receptors was observed. Neither co-expression of all four known subunits nor the use of a range of BZs gave receptor potentiation. Although the {3 subunit has not been previously implicated as heterogenous, and /3j clones have been isolated, sharing 70% homology to /3,. Receptors with differences, such as ligand affinites, channel life times, allosteric modulation sites, and intracellular controls, could increase information handling and perhaps even contribute to neural plasticity. The existence of multiple GABA^ subunit subtypes and the

354 Iiilernalional Symposium on Receptors and Ion Channels

dillerent ligand sensitivities displayed by ihe three a subunits provides a structural and functional basis for receptpr subtypes.

Surprisingly, new subunits were also d.scovered. The y, and >-2 subunits share 35-40% amino acid homology to the a and P subunits and have similar structural features. In the variable intracellular loop domain these polypeptides contain a consensus sequence for tyrosine-specific protein phosphorylation, suggesting regulation via tyrosine kinase. The y^ subunit , along with a , and /3,, was characterized in a mammalian expression system (Pritchett et al., 1988, 1989). The cDN As were cloned into a vector containing the human cytomegalovirus promoter /enhanccr and transfected into human embryonic kidney (293) cells. After 48 hr, cells were harvested for binding or electrophysiological studies using whole cell patch-clamp. In filler binding assays, cells transfected with a , + cDNAs possessed sites for both ^H-muscimol (GABA analogue) and '*S-TBPS (GABAy^ receptor channel-specific cage con-vulsant). These were not observed with a , or alone. Binding for *H-Rol5-1788 (BZ antagonist) and ' 'H-fluniirazepam was only found in cells transfected with a , , /3, plus y , . This could be blocked by diazepam or the inverse agonist D M C M . Receptors formed from o , + Pi displayed similar electrophysiological properties to those expressed in oocytes. Only with the combination of a , . /3, plus y^ was there two-fold potentiation by BZ agonists, as expected from in vivo studies. DMCM decieased GABA-induced flow by 50% and all these effects were blocked by Rol5-1788. An unexpected result was the GABA-induced current flow observed in cells expressing a , , or y2 alone. The whole cell currents of these homomeric channels were about I07o of those seen with the heteromeric receptors ( a , + /3,) which may reflect reduced efficiencies of receptor assembly or a change in single channel characteristics. Pharmacologically, these receptors display GABA receptor profiles similar to Of, + /3, receptors.

These results show the versatility of the expression system where binding and electrophy-siological studies can be carried out on the same cells. The sensitivity is indicated by the measurement of homomeric channels. This channel property questions a previous concept, derived from the nAchR model, that the neurotransmitter site involved in channel gating is located on one subunit type only. Our results suggest sites conserved in all these subunits, possibly of low apparent affinity, responsible for agonist gating of the channel and allosteric effects.

The existence of the y subunits were not anticipated from biochemical studies as subunits of affinity-purified receptor are electrophoretically resolved into two major bands (Stephenson, 1988). However, these bands are heterogeneous, consisting of various a and 0 subunits, and a mature y polypeptide may comigrate with the a subunit and escape detection. In siiu hybridization of rat brain sections localized the y2 subunit in neuronal subsets consistent with BZ binding sites (Olson and Venter, 1986), further confirming y j participation in BZ pharmacology.

The y, subunit was localized in glial cells. Electropharmacological studies have revealed the presence of GABAY^ receptors on glia, the only defined difference to neural GABAA receptors being an inconsistent effect of inverse agonists. The role of the GABA^ receptor on glial cells in unknown but is postulated to be involved in restoring the neuronal CI" gradient during synaptic activity.

Finally, another new subunit has been designated 5. It has 35-40% amino acid homology to other GABA subunits. It is most highly expressed in granule cells of the cerebellar cortex with localization similar to the reported high affinity GABA binding site (Olson and Venter, 1986). This neuronal subset in which 5 is expressed is largely separate f rom that of a , and y j . When R c D N A is transiently expressed alone, a CI channel is observed after GABA application. This subunit along with fty, and y, are currently undergoing further elec-tropharmacological analysis.

Short Communications

Our studies have revealed that the GARA^ receplor is Tar more complex than was envisioned from classical pharmacology. The isolation oTihe many subunits poses the problem of defining the composition and stochiomeiry of the naturally occurring GABA^ receptor subtypes. The single channel characteristics of various subunit combinations transiently expressed in mammalian cells are being compared to patched single channels on rat brain slices. In parallel, in situ hybridization studies are focusing on certain brain regions having well-characlerized neural circuits. For example, in the cerebellar Purkinje neurons we have established that or,, /Sj, and y^ subunits are expressed. Finally, we are investigating subunits expressed in homogeneous tissues or clonal neuronal cell lines which may contain a single subtype of GABAy^ receptor.

References Casalotti. S. O.. .Slephenson, P. A., and Barmird. K. A. (1^X6). J. Binl. ( hem. 261. 15()I.V Cavaila. IX, and NclT. N. H. (IQS.S). J. Neurnchem. 44. VI6. Deng, L., Ransom. R. W.. and Olsen. R. W. Himhem. Hiophvs. Res. Cnmmun. 1.18. I.108. Houamcd. K. M., el al. (I9«4). Nature 310, 31X Lcvitan. E. S.. and Schoficld. P. R.. ei al. (I9KX). Noiure 335. 76. Mohler, H.. Battcrshy, M. K.. and Richards. J. G (19X0). Prnc. Nail. Acad. Sci. USA 77. 1666. Olsen. R. W.. and Venter. C. J. (1986). Benzodiazepine/CAHA Recepion and Chloride Channels: Slniciural

and Functional Properties (Allan Lis.s. New York I. Pritchclt. D. B., et al. (1988). Science 242. 1.106 Prirchcti, D. B., el al. (1989). Submitted. Schoficld. P. R., et al. (1987). Nature 32)1. 221. Sigel, E., Stephenson, F. A.. Mamalaki. C.. and Barnard. E. A. (1983). ./ Biol. Chem. 258. 6965. Smart. T. G., Constani. A., Bilbe. G., Brown. D. A., and Barnard. F.. A. (1983). Neurnsci. Lett. 40. 55. Stephenson. F. A. (1988). Biochem. J. 249. 21.

19. L. N. Zimmermann, H. H. Schneider, and D. N. Stephens. Binding Characteristics Reveal Partial GABA Agonist Activities of SR 95531. (Re.search Laboratories of Schering AG, West Berlin and Bergkamen, FRG)

The-pyridazinyl-GABA derivative SR 95.^31 has been identified as a competitive ligand at GABAy^ binding sites with a 200-fold higher displacing affinity than the classical GABA antagonist bicuculline (Heaulme ei al., 1986).

In electrophysiological and behavioral experiments, this compound has been character-ized as an antagonist, but in these experiments, eliects occur in the same dose range as that of bicuculline (Hammann et al, 1988).

GABA agonists in vitro stimulate the binding of benzodiazepine receptor agonists (Tallman et al., 1977) and inhibit the binding of inverse agonists (Honore et al., 1983). We examined the eflfect of SR 95531 on the binding of *H-lormetazepam, a benzodiazepine receptor agonist, and "^H-DMCM, an inverse agonist. Additionally, we investigated the effect of SR 95531 on the binding of *'^S-TBPS. a putative CI channel ligand sensitive to modulation by GABAergic compounds (Squires and Saederup. 1987). Radioligand binding experiments were performed by filtration assays using repeatedly frozen-chawed, washed, and dialyzed membrane preparations from rat whole forebrain. The binding of 'H-lormetazepam and 'H-DMCM was assayed at a final concentration of 0.65 nM and 0.45 nM. respectively, with

Vuiiime 258, n u m b e r I, 119-122 F E B 07842 November 1989

Sequence and expression of a novel G A B A a receptor a subunit

Sanie Ymer*, Andreas Draguhn"", Martin Kohler, Peter R. Schofield** and Peter H. Seeburg

Laboratory of Molecular Neuroendocrinology. Center for Molecular Biology (ZMBH). Im Neuenheimer Feld 282. D-6900 Heidelberg, a n d ^ Dept. of Celt Physiology, Max Planck Institut fiir Medizinische Forschung. Johns tr. 29 D-6900 Heidelberg, FRG

Received 9 August 1989; revised version received 3 October 1989

Cloned cDNA encoding the bovine subunit of the GABA^ receptor has been isolated. The predicted 521 amino acid long mature protein contains an exceptionally long intracellular domain and shares 53-56% sequence similarity to the previously characterized a,, Oj and Oj subunits. Co-expres-sion of and |i^ in Xenopus oocytes resulted in the formation of GABA-gated chloride channels with expected pharmacology, although no benzodiazepine potentiation was observed. Northern analysis indicates that a 4 kb mRNA is expressed in the calf cerebellum, cortex and hippo-

campus but is barely detectable in the rat brain.

^-Aminobutyric acid A receptor; Ligand-gated ion channel; Receptor subtype; Voltage clamp recording; cDNA cloning

1. INTRODUCTION

Fast synaptic inhibition in the central nervous system is primarily mediated by G A B A (y-aminobutyric acid). This neurotransmitter opens the intrinsic chloride chan-nel of the GABAA receptor, the activity of which can be modulated by a variety of drugs, notably ben-zodiazepines (BZs) and barbiturates [1,2]. Peptide se-quences derived from affinity-purified material have facilitated the isolation of c D N A s encoding GABAA receptor Q and ^subunits [3]. As a consequence, several structural features have emerged common to other ligand-gated receptor subunits [3,4]. These features in-clude 4 transmembrane domains (M1-M4) and a disulphide loop formed by two cysteine residues located extracellularly. Another feature is the highly divergent intracellular loop domain between M3 and M4. Within the M2 of GABAA and glycine receptors there is a con-served 8 amino acid sequence [5] thought to form part of the channel lumen [6,7]. Degenerate oligonucleotide probes encoding these amino acids have been used to isolate further GABAA receptor subunit encoding cDNA clones. Currently 4 classes of subunits have been

Correspondence address: P .H . Seeburg. Laboratory of Molecular Neuroendocrinology. University of Heidelberg. ZMBH. INF 282, D-6900 Heidelberg. FRG

*Preseni address: Division of Biochemistry and Molecular Biology, John Curtin School of Medical Research, The Australian University, GPO Box 334, Canberra City. 2601 Australia

**Preseni address: Pacific Biotechnology Ldt. 72-76 McLachlan Avenue, Rushcutlers Bay, 2011 Australia 'he nucleotide sequence(s) presented here has (have) been submitted

the EMBL/Gen Bank database under the accession number no. Y07515

^^blished by Elsevier Science Publishers B. V. (Biomedical Division) '^145793/89/$3.50 s) 1989 Federation of European Biochemical Societies

identified and designated a [8], J3 [9,10], y [11] and S [12]. All classes share 35-45% sequence similarity and within each class (with the exception of S) variants exist which display in excess of 70% identity. These studies have confirmed and extended the functional hetero-geneity of the GABAA receptor suggested by phar-macology [13-16] and photoaffinity labelling [17-19]. Using the same approach in this study, cloned c D N A encoding a novel A subunit (OA) was isolated from a bovine brain c D N A library. Functional expression in Xenopus oocytes demonstrates that this a subunit is capable of combining with a P subunit [3,9] to form GABA-gated chloride channels.

2. MATERIALS A N D M E T H O D S A bovine brain cDNA library in AgtlO [3] was screened using a

96-fold degenerate "P-labelled 23-mer oligonucleotide pool [9.12] en-coding the conserved 8 amino acids in M2 of GABAA and glycine receptor subunits. Known subunits were identified using a pool of a u az, a i and/?i subunit-specific oligonucleotides [9]. cDNAs hybridiz-ing only to the 23-mer were sequenced [21] in AgtlO [9,12] or after subcloning into M13 vectors [20]. The longest a^ cDNAs were com-pletely sequenced with the aid of internal primers (5' GTG TCC TAT GCA ACT GCC 3 ' , 5 ' CAC AAT GAG ACT CAC CAT 3' and 5 ' ATT TCA GCT GCT CCA GTG CTG 3 ' ) .

A 2.0 kb EcoRI fragment encoding the entire bovine a^ subunit was subcloned into pGEM-2 (Promega, Madison. WI) and the resulting construct linearized with ///Vidlll. Capped bovine a* and/?i [3] RNA transcripts were synthesized using Sp6 RNA polymerase and m^G(5')ppp{5')G according to Promega. Approximately 50 nl of subunit-specific RNAs ( 1 i " H2O) were co-injected into defolliculated Xenopus oocytes and after incubation at 19°C for 2-6 days, induced currents were recorded using a conventional two-electrode voltage clamp [22].

RNA was isolated [23] from whole bovine brain, from brains of young, sexually mature rats, and from the cerebellum, cortex and hip-pocampus of an 8-month-old calf. RNA was enriched for polyCA)"^ RNA by oligo(dT)-cellulose chromatography. Northern analysis was carried out [9] using an 0-4 subunit-specific ^^P-end labelled 60-mer

119

Volume 258, number 1 FEBS LETTERS November 1939

oligonucleotide (5' AGC AGA GGG AGT AGT AGT GGC TGA TAA CTT CCC CGA AGT CCC TAT GCT ATT A A C TGT GGT 3 ' ) as the probe. Other probes used were oligonucleotides specific for

az and m subunits [8].

3. RESULTS A N D D I S C U S S I O N

Numerous hybridization signals resulted from screen-ing a bovine brain AgtlO cDNA library with the degenerate M2 oligonucleotide probe. Those clones which did not also hybridize to the known subunit-specific oligonucleotide pool ( o n , 02,0:3 and/S*!) were in-itially characterized by sequencing using the degenerate M2-encoding oligonucleotide as a primer. Cloned cDNAs encoding new subunits were identified by homology of their deduced amino acid sequence with previously characterized G A B A a receptor subunits. As a result, a cDNA clone was identified that contained an open reading frame of 1701 bp displaying significant similarity to the previously characterized G A B A a

receptor a subunit sequences [8]. This cDNA encodes a polypeptide (designated 0-4) of 556 amino acids, in-cluding a 35-residue signal peptide [24]. The cDNA se-quence was not generated by alternative splicing of any other a subunit transcript.

A comparison of the predicted 0-4 polypeptide se-quence with that of the a i and 0-3 subunits [8] shows that, as for other G A B A a receptor subunits, regions of highest similarity include the putative transmembrane domains, the most conserved being M2 (fig.l). Here, a i , 0-2 and 0-3 are identical and share 96% identity with 0-4 in which a valine residue substitutes for an isoleucine (position 258). The extracellular domain is conserved to about 73% betweenai,a2, Q'3 while0-4 shares 60% with each. This domain contains putative N-glycosylation sites (fig.l). Overall, 0 4 shares only 56% invariant amino acid residues with a i , az and 53% with a^. The relationship of 0-4 to subunits of other classes is as follows: /?i, 30%; 72, 40%; S, 29%; glycine 48 kDa, 33%.

The intracellular loop domain of the a4 polypeptide is extremely long, making a4 the largest G A B A a recep-tor subunit to date with a predicted molecular mass of 64 kDa for the unglycosylated mature protein. This do-main displays the greatest amino acid sequence diversity between different subunits and may contain sites for in-tracellular regulation of channel activity [3]. In fact, both;^ [9] and 7 [II] subunits have consensus sequences for phosphorylation located here. No such sites have been found in a4 (or any a) but other unidentified regulatory features may be present.

To study functional expression the novel a4 subunit was co-expressed with the bovine /Si subunit [3,9] in X e n o p u s oocytes (fig.2). Following injection of in vitro synthesized RNA, 89% of oocytes (a? = 131) expressed GABA-induced inward currents that were dose-dependent between O.OI and 1 0 0 W h e n measured, the slope of the log dose versus log response curve

ijpnti

I c.?.e:N O'N I I ' V

;ODi CCLS ruH. D O I r O'.N I T' f-j

LlOCrDNtLkF LLDCroNtLar tLOCrONKLHrC LLOGTONRlll

' r 1 r cr• • T » p c r V J o | h ' T 1 f C P V , 0 T

T i r e » V J 0 T

N C t K S N C < X t N C ( ( S N C K » S V TY

I T E D I Q 0 O L V OCH

L r T H > I T L T T H K I T L T T H K I T L T T UK L T

c r L t r c s T . CPLIPCI T, C P i i r c s r . P K S T » A T T S

K C P E "A.'..* K^N i t. L G Y-IJ

17} s y E 7 P K E S S s L Q T 0 I 1 C q T V f I E T • K , T C E T 1 V M r m 5 V V V A E D C s R I N g T D L 1. c QTVDICI vq s s T C E r V V M T m 0 S V Q V A P D C s % L N q T 0 L P c q S 1 S K E T 1 K 1 i T S E r T V MI m K S V E V A Q D C s » 1- N q T 0 1. L c H V V C T E 1 1 A s i T C I T * V MT

>l> P H L "1 „ GYP M 1 q j T , P C 1 M T * I L s q V s p « I N E « y P A » T Jl» PHI K t t 1 OTP V 1 q T r L P C 1 M T ¥ 1 I s q V s p « L N R E S V P A A T iu PHI fH A ( 1 OTP V 1 q T r L P C 1 M T V 1 L s q V 1 p « L N A E I y P A t T PHI A X 1 c r P I q j T L P C 1 M T V 1 L s q y J p * L « E J y P A « T

N T P T N T P T N r P T

ItCTAVDCKS VVPt - -IP1K-VED KIICVAVDCKS VVNO - - - ( K K - I I t KUt •AWtOKHVPtAlEHHUKTPAV

PtKHPETPlQNTNANLS HI ES DVCS UTOVCII

H t S K t S T ' 'QCISEATPQiTlASSP

H A I P t A l P S r P S I KCDPCIATI AKS. K-OPVLSTI SKS. K'OTEPSAISKCi

T c r v p n g v p v i T- l E P l t V . . TTPEPNH---- A P J T » J T P r I

PFStANAAETI SAA TTAPTATS »TPNIA NNAYAVAVANYAPNtl STTFNI VCITYPI NLA

SASTQHVFCJIILQ*!*

CTSCtLSATTT UPETK UPENK CVQDl

t A P P P S C S C T PEPHHTFN5V AEAKHTFNSV Tt-TATTNSV

ISKOrMCKSESLM LNAEPQLlAPTPHg LNRLPViCV^P VNKES AI HCMl RKg

Fig . l . Comparison of the predicted amino acid sequence of the bovine GABAa receptor a^ subunit with bovine a i , 0-2 and 0-3 subuiuts [8]. The sequence alignment contains the following corrections totlK previously published a i , a i and a i sequences [8]: deletions of T » position 368 (a , ) . T at position 365 (ffz). and KGA at positions 393-395 (q-j). These residues had been included in [81 as a result ofa faulty alignment program and had escaped proof-reading. Anaio^ acid sequence numbering starts at the proposed mature N-terniin residue [24], the presumptive signal sequences being negative^ numbered. Invariant residues are enclosed in solid boxes and tW putative N-linked glycosylation sites in dotted boxes. Postulate membrane-spanning hydrophobic sequences M1-M4 are marked J solid lines and the cysteine residues forming the disulphide loop I dots [3]. Dashes have been introduced to improve sequence alig ment. The cDNA sequence of the a4 subunit has been deposited int

EMBL database under accession no. Y07515.

120

V o l u m e 2 5 8 , n u m b e r FEBS LETTERS November 1989

IOmM g a b a lOuM GABA lOuM Pb

10»iM GABA lOpM GABA lOjiM MID

VJ 1 OOnA

1 m i n

Fig.2. Funct ional expression of bovine G A B A a receptor 0-4 a n d subunits in Xenopus oocytes . M e m b r a n e potent ia l of the oocytes was held at - 7 0 mV in a convent iona l 2-microe lec l rode vol tage c l amp . Downward def lec t ions represent inward C l ~ cur ren t s fo l lowing ap-plication of 10 / /M G A B A a lone or together with indicated d rugs . Each appl ica t ion was fol lowed by washing with no rma l f rog Ringer

solut ion [13,23] for at least 3 min .

H i

W*' f j y

(0.3 ± 0.07, determined between 0.1 and 1 /^M G A B A ; n = 4) was significantly below the expected value of at least one for other recombinant cv 4-/5'receptors [8]. This shallow slope could reflect a particularly rapid desen-sitization of q'4+/?i GABA a receptors.

The current response to G A B A (10/^M) was blocked by the antagonist , bicuculline (10 /iM), to SO /o of initial amplitude (not shown) and potentiated two-fold upon application of the barbi turate pentobarbi ta l (10/^M), indicating the expression of a barbiturate-sensit ive site. However, no potentiat ion by the BZ-receptor agonists, diazepam (2-5 ^M, n = 20) and midazolam (10

was observed at G A B A concentrat ions ranging from 1 to 40 /^M (fig.2). Thus the pharmacology displayed here differs f rom neuronal G A B A a receptors. Recent results [11] suggest that a third subunit (72) may be required in order to form channels displaying physiological responses to BZs [1,2].

The extent of 04 expression was investigated by Nor-thern blot analysis (fig.3). RNA samples were prepared from rat and bovine total brain as well as calf cerebel-lum, cortex and hippocampus. Nor thern blots of these RNAs were hybridized with a ^^P-end labelled oligonucleotide (60-mer) complementary to D N A se-quence encoding part of the distinct intracellular do-

p a i n of the 0-4 subunit . In all 3 regions of the calf brain investigated, a single 4.0 kb m R N A was observed, be-ing about equally abundant in the cortex and cerebellum and about one-f i f th as abundan t in the hip-pocampus (fig.3A). Probing with a subunit variant-specific oligonucleotides indicated the order of abun-dance in the bovine brain to be a i , a s , 0 4 then a i (fig.SB). In the rat brain, q!4 expression was hardly detectable (not shown). No cDNA clones were iden-tified in a rat forebrain c D N A library but a clone en-coding an incomplete human a4 subunit was isolated from a human brain cDNA library (unpublished). Hence, the a4 subunit may be extremely rare in the ^dult rat brain but developmental regulation of this subunit needs investigating. Our results may also reflect the general observation that GABA a receptor subunits

9 - 5 7 - 5

4 - 4

2 - 4 -

1 - 4

a 1 a 2 a 3 a 4

Fig.3. Express ion of a^ subunit m R N A in the bovine bra in . (A) N o r t h e r n blots of polyCA)"" R N A f r o m ' c a l f cerebel lum (Ce), cortex (Cx) a n d h i p p o c a m p u s (Hi) p robed with an a4 subunit-specif ic o l igonuc leo t ide . (B) Nor the rn blots of poly(A)^ R N A f r o m bovine to ta l b ra in p r o b e d with a i . a z , a j and 0-4 subuni t -specif ic oligonucleo-

t ides. Size marke r s (kb) are indicated on the lef t .

have a lower cellular expression in the rat brain than the bovine brain [9].

In conclusion, the GABAa receptor 0:4 subunit can be classified as such because it shares greatest sequence similarity to the a subunits and can induce GABA responses when co-expressed with a /? subunit . It will be impor tan t to elucidate the physiological role of the a4 subunit in G A B A a receptors of central and peripheral neural tissue.

Acknowledgements: W e wish to thank Dr Brenda Shivers and Pro-fessor Bert S a k m a n n for their interest and suppor t . We also thank Dr Brenda Shivers for providing the rat and bovine tissue samples . Hi ldegard Kluding for R N A prepara t ion , Dr Michael Nassal for o l igonucleot ide synthesis and Ju t t a Rami for secretarial assistance. This work was suppor ted in part by the Deutsche Forschungsge-me inscha f t , SFB 317 Gran t B /9 and B M F T Gran t BCT 0381/5 to P . H . S .

121

V o l u m e 258, n u m b e r FEBS LETTERS November

R E F E R E N C E S

[1]

[2]

[3]

[4]

15]

[6]

17]

[8]

[9]

[10]

Olsen, R.W. and Venter, J .C. (1986) Benzodiazepine/GABA Receptors and Chloride Channels: Structural and Functional Properties, A.R. Liss, New York. Stephenson. F.A. (1988) Biochem. J. 249, 21-32. Schofield, P.R. , Darlison, M.G., Fujita, N., Burt, D.R., Stephenson, F.A., Rodriguez, H., Rhee, L.M., Ramachandran, J . , Reale, V., Glencorse, T.A. , Seeburg, P .H. and Barnard. E.A. (1987) Nature 328, 221-227. Grenningloh, G., Rienitz, A., Schmitt, B., Methfessel, C., Zensen, M., Beyreuther, K., Gundelfinger, E.D. and Betz, H. (1987) Nature 328, 215-220. Grenningloh, G. , Gundelfinger, E., Schmitt, B., Betz. H. , Darlison. M.G., Barnard. E.A., Schofield, P.R. and Seeburg, P .H. (1987) Nature 330, 25-26. Imoto, K., Methfessel, C., Sakmann, B., Mishina, M., Mori, Y., Konno, T. , Fukuda, Y., Kurasaki, M., Bujo, H. , Fujita, Y. and Numa, S. (1986) Nature 324, 670-674. Leonard. R.J . , Labarca. C.G. . Charnet. P. . Davidson, N. and Lester, H .A. (1988) Science 242, 1578-1581. Levitan. E.S., Schofield, P .R. , Burt. D.R., Rhee, L.M., Wisden, W.. Kohler, M. , Fujita, N.. Rodriguez, H.F . , Stephenson, A., Darlison, M.G. . Barnard, E.A. and Seeburg, P .H. (1988) Nature 335, 76-79. Ymer, S., Schofield, P .R. , Draguhn, A., Werner, P. , Kohler. M. and Seeburg, P .H . (1989) EMBO J. 8, 1665-1670. Lolait, S.J. , O'Carroll , A. , Kusano, K., Muller, J . , Brownstein, M.J . and Mahan, L.C. (1989) FEBS Lett. 246, 145-148.

[11) Pritcheit, D.B.. Sontheimer, H. . Shivers, B.D.. Yrner, 3 Kettenmann, H., Schofield, P.R. and Seeburg, P.H. (193( 1 Nature 338, 582-585. Shivers, B.D., Killitsch, 1., Sprengel, R.. Sontheimer, tj Kohler, M.. Schofield, P.R. and Seeburg. P.H. (1989) Neurc^ in press. Squires, R.F., Benson, D.I. , Braestrup, C.. Coupet, j Klepner, C.A., Myers, V. and Beer, B. (1979) Pharmacol Biochem. Behav. 10, 825-830. Braestrup, C. and Nielsen, M.J. (1981) J. Neurochem. 3-1 333-341. ' '

[15] Unnerstall, J .R. , Kuhar, M.J . , Niehoff, D.L. and Palacio^ J .M. (1981) J. Pharmacol. Exp. Ther. 218, 797-804. Cooper, S.J. , Kirkham, T .C. and Estall, L.B. (1987) Trends Pharmacol. Sci. 8, 180-184. Mohler, H., Battersby,.M.K. and Richards, J .G. (1980) Proc Natl. Acad. Sci. USA 77, 1666-1670. Sieghart, W., Mayer, A. and Drexler, G. (1983) Eur. j Pharmacol. 88, 291-299. Fuchs, K., Mohler, H. and Sieghart, W. (1988) Neurosci. Len 90, 314-319. Vieira, J . and Messing, J . (1987) Methods Enzymol. 153, 3-11 Sanger, F., Nicklen, S. and Coulson, A.R. (1977) Proc. Nai: Acad. Sci. USA 74, 5463-5467. Methfessel, C., Witzemann, V., Takahashi, T., Mishina, M. Numa, S. and Sakmann, B. (1986) Pfluger's Arch. 407, 577-588. Chomczynski, P. and Sacchi, N. (1987) Anal. Biochem. 162, 156-159.

124] von Heijne, G. (1986) Nucleic Acids Res. 14, 4683-4690.

[12]

[13]

[14]

[16]

[17]

[18]

[19]

[20]

[21]

[22]

[23]

122

Neuron. Vo l 647-b54. November . 1989. Co(>yrighl © 19H9 by Cel l Press

Chromosomal Localization of GABAA Receptor Subunit Genes: Relationship to Human Genetic Disease Veronica Buckle,* Norihisa Fujita,*' Allan S. Ryder-Cook/ Jonathan M. J. Derry,< Pene J. Barnard,* Roger V. Lebo,5 Peter R, Schofield,!** Peter H. Seeburg,! Alan N. Bateson,* Mark G. Darlison,* and Eric A. Barnard* • Nuffield Department of Clinical Medicine University of Oxford Oxford OX3 9DU England •MRC Molecular Neurobiology Unit-MRC Centre Cambridge CB2 2QH England § Howard Hughes Medical Institute University of California San Francisco, California 94143 • Laboratory of Molecular Neuroendocrinoiogy ZMBH University of Heidelberg D6900 Heidelberg Federal Republic of Germany

Summary ^

Hybridization of GABA* receptor probes to human chromosomes in situ and to DNA from sorted human chromosomes has localized the genes encoding a (J subunit and three isoforms of the a subunit. The a2 and P genes are both located on chromosome 4 in bands p12-pl3 and may be adjacent. The a l gene is on chromosome 5 (bands q34-q35) and the a3 gene is on the X chromosome. The a3 locus was mapped also on the mouse X chromosome using genetic break-point analysis in an interspecies pedigree. The combined results locate the human a3 gene within band Xq28. in a location that makes it a candidate gene for the X-linked form of manic depression.

Introduction

The GABA;^ receptor is the major inhibitory neurotrans-mitter receptor of the vertebrate brain. It is known to contain a and P subunits, and three isoforms of the a subunit and one of the p subunit have been identified by cDNA cloning from the bovine brain (Schofield el al.. 1987; Levitan el al., 1988). The cDNA sequences show that the a subunits are homologous but derive from different genes. Furthermore, the corresponding mRNAs have different distributions between brain regions and between cell types in those regions (Wisden et al., 1988. 1989). It can be expected that the loss of any of these

' P r e sen t address C i b a - C c i g y ( | a p j n ) Ltd.. Takara^uk.i 065. lap . in ••Present address; Paci f ic B i o t e c h n o l o g y Ltd.. 72-76 M c L a c h l a n .\venue. Rushcut ters B a y 2011. Austra l ia

subunits as .i functional entity due to mutation could be deleterious in some brain locations. Hence, these genes could be candidates for neurologic and psychiatric in-herited disorders in certain cases: possibilities where this should be considered include those conditions benefit-ting from treatment with valproate or benzodiazepines (which potentiate the action of GABA).

It is therefore of both clinical and fundamental interest to know on which chromosomes these GABA^ receptor genes are located and whether some are associated on a single chromosome. From a clinical point of view it is necessary to make this determination in the human ge-nome. but it IS also of value to do so in the mouse, in which genetic mapping and breeding of mutants with genetic defects of the central nervous system are more advanced than in other mammals and the homology of certain chromosomal regions of linked genes with those found in man (linkage analysis) can give further defini-tion to gene locations on human chromosomes. We re-port here on hoth these s[)ecies and find a suggestive as-sociation (ji ,1 GABA.v rece|)tor gene with an inherited disease, meriting detailed investigation.

Results

Either the cloned bovine GABA.x receptor subunit cDNAs (Schofield el al.. 1987: Levitan et al., 1988) encoding three forms of the a subunit, al, a2. and a3, and the P subunit. or corresponding human brain cDNAs (Schofield et al.. 1989) were used to generate probes (see Experi-mental Procedures). The four human brain subunits are almost identical in amino acid sequence to the corre-sponding bovine subunits (approximately 98% identity: Schofield et al.. 1989; P. R. Schofield and R H. Seeburg. unpublished data), justifying the cross-hybridization of bovine probes to human chromosomes used in some cases. Similarly, the a3 subunits of bovine and rat are ap-proximately 99% identical in amino acid sequence (P. R. Schofield and R H. Seeburg. unpublished data), so the bovine a3 probe was likewise used directly to hybridize to mouse DNA.

Chromosome Assignment Using DNAs from Sorted Human Chromosomes W e have used a set of nitrocellulose filters (Lebo et al.. 1984) containing dual laser-sorted human chromo-somes (DNA spot-blots) to determine the locations of tour GABA.V receptor subunit genes. These were sequen-tially hybridized with the corresponding bovine cDNA probes (see Experimental Procedures). An example of these results, for the P subunit gene, is shown in Figure 1. A strong hybridization signal is seen with DNA from chromosome 4 only; all other chromosomes are neg-ative.

When hybridized with the al probe, strong positive signals were detected with chromosome 4 and with chromosome .S (Table 1). The a2 probe hybridized most

Neuron 648

4 1 2 6 6 8

13

9 X -12

21 19 20

Figure 1 Hybridization of CABA^ Receptor Gene Probes to Spot-Blot Filters Containing a Complete Set oi Sorted Human Chromosomes An example is shown of an autoradiograph obtained with the P subunit probe. Numbers refer to chromosomes. Note the strong hybridization to the filter containing chromosome 4.

Strongly to chromosome 4. whereas the a3 probe gave strong signals with the X chromosome and with chro-mosome 8. From these results we conclude that the GABA; receptor a2 and* P subunit genes are both lo-cated on chromosome 4. These data also tentatively sug-gest that the al subunit gene is located on chromosome 5 and that the hybridization seen with chromosome 4 {Table 1) is probably due to cross-hybridization of the al probe to the a2 gene on the latter chromosome. Using these filters, we were unable to assign unambiguously the human a3 subunit gene, although the results suggest that it is located either on chromosome 8 or on the X chromosome.

In Situ Hybridization to Human Lymphocyte Chromosomes In this technique, metaphase chromosomes, prepared from human blood lymphocytes, are stained in order to reveal replication banding. Appropriate radioactively labeled cDNA probes are hybridized to these chromo-

Table 1. Chromosomal Locations of GABA^ Receptor Genes Using DNA Spot-Blots

Subunit Gene Chromosomes Hybridizing Strongly

a l

a2

a3

4. 5

4

8. X

D N A spot-blots were hybridized with the four G A B A ^ receptor subunit CDNA probes in turn. In addition to the strong signals, as shown, there were usually several other chromosomes giving much weaker signals: these were ignored. An example of such an experi-ment is shown in Figure t.

some spreads, and after autoradiography, the grain dis-tribution over all of the human chromosomes is ana-lyzed (Buckle and Craig, 1986).

The distribution of silver grains over the entire karyo-type was examined in 30 or more cells, for each of the four hybridizations with the G A B A A receptor al, a2, a3, and p subunit cDNA probes. An example is shown in Figure 2A. These genes were localized thus to chro-mosomes 5, 4, X, and 4, respectively, and in each case there was no evidence of cross-hybridization, since no other chromosome showed significant labeling. The dis-tribution of silver grains along the length of the relevant chromosome was examined in detail to yield higher resolution localization (Figures 2B-2E).

Localization of the Human al Subunit Gene One or both chromosomes 5 were labeled in 70% of the 30 cells scored. Of the 50 grains counted over the whole karyotype, 40% were localized to the region 5q32-q35, which represents roughly 1% in length of the haploid ge-nome. More specifically, 76% of all grains scored over chromosome 5 were localized to bands 5q34-q35 (Fig-ure 28). We designate this gene GABRal.

Localization of the Human a2 Subunit Gene One or both chromosomes 4 were labeled in 50% of the 30 cells scored. Of the 72 grains counted over the whole karyotype, 19% were localized to the region 4cen-p15, which represents 1% in length of the haploid genome. The distribution of silver grains over chromosome 4 was analyzed in more detail in 27 cells (Figure 2C): 67% of all grains scored over chromosome 4 were localized to bands 4p12-p13. We designate this gene GABRa2.

Localization of the Human a3 Subunit Gene The X chromosome was labeled in 37% of the 35 cells

c n f l i i M fi" L M u r x j f c a a ; n j j i L J i j _ i u r K l i _ u a i i . a i i : l K * « » a a r n o r t c

c i i j i i i A i L c i E i i n f i K d a c n i i • im. iw;—iLfc i : j o c w r i b x a ? H C b a z i i i j a H

I) k IS j j o a J i f c j c a i n c VTwr rM- b t J t x i 3 1 - 1 T n a c L U G c n ( i n x n a w i u x ^

16 11 w « » 7' J: X

B D ::: •

H i

m

% X

C "

HFi

E

Figure 2. Chromosomal Localization of Four CABA^ Receptor Genes by In Situ Hybridization ^ ^ (A) An example of the grain distr ibution seen over the whole karyotype. In this case 30 cells were ^ ^ " d . z e d J ^ ^ ^ ^ ^ j T o ' b e ; (D) subunit probe. (B-E) The grain distr ibution over the labeled chromosomes: (B) chromosome 5. a l probe; (C) chromosome 4. a2 prot>e. X chromosome. a3 probe; (El chromosome 4, (1 probe.

Noiiron 650

scored. Of the 68 grains counted over the whole karyo-type, 19% were localized to the region Xq27-q28. This represents 38% of grains scored per haploid genome, since the cells analyzed have only one X chromosome. The region Xq27-q28 represents 0.5% in length of the haploid genome. More specifically, 79% of all grains scored over the X chromosome were localized to the band Xq28 (Figure 2D). W e designate this gene GABRa3.

Localization of the Human P Subunit Gene One or both chromosomes 4 were labeled in 47% of the 32 cells scored. Of the 61 grains counted over the whole karyotype. 26% were localized to the region 4cen-pl5.1. The grain distribution over chromosome 4 was analyzed in more detail in 19 cells (Figure 2E). Of all grains scored over chromosome 4, 70% were localized to bands 4pl 2-pl3. We designate this gene GABRpi.

There was no inconsistency between the locations found by the in situ method and those found using spot-blots of sorted human chromosomes. However, the extension to the in situ method was required, due to the ambiguities remaining in the spot-blot results and the resolution to bands provided by the in situ hybridiza-tion method.

Localization of the Mouse a3 Subunit Gene Accurate positioning of lo^i on the mouse X chromo-some is possible using interspecies mouse pedigrees to increase the frequency of DNA polymorphisms (Avner et al., 1988), which occurs due to evolutionary diver-gence of the genomes, thus facilitating the detection of recombination between gene loci. The evolutionary con-servation of X-linked genes in placental mammals (Oh-no. 1969) and extensive mapping studies of the mouse X chromosome (Amar et al., 1988; Ryder-Cook et al.. 1988; Searle et al., 1987) have allowed identification of groups of genes in which the order has been preserved between mammals (linkage analysis). One of these groups that includes loci from the region q26-qter in the human X chromosome has been particularly well mapped in the mouse, and the following order of equivalent loci has been established:

Mouse: cen - Hprt - DXPas6 - Rsvp -Cf-8/Cdx - C6pd - tel Human: cen - HPRT- DXS144 - RCP- F8C/GDX- G6PD - lei

The loci referred to, with their accepted abbreviations for the two species, are (in the order shown above): hypoxanthine phosphoribosyl transferase; an equivalent pair of random DNA markers (DXPasS, mouse; DXS144. human), both.revealed by the human DNA probe G i l ; the red-sensitive visual pigment gene in the mouse and the human equivalent, the red cone pigment gene (RCP); coagulation factor VIIIC (Cf-S. mouse; F8C, human); Gc/x (mouse) or GDX (human), a gene known to encode a small ubiquitin-like housekeeping protein. This gene can serve as a very close marker for the glucose-6-phosphate dehydrogenase gene (C6pd, mous^; G6PD, human), since the human GDX is less than 45 kb 3' to the last exon of the G6PD gene (Toniolo et al.. 1988).

These loci were |)rc'vic)usly mapped in an interspecies mouse pedigree in which iheir order had been deter-mined by recombin<\lion frequency analysis of back-cross progeny and the individual genotype of each ani-mal had been determined by Southern blot analysis of genomic DNA (Ryder-Cook el al., 1988). New genes can now be added to the X chromosome map, as probes be-come available, by resampling the existing DNAs from the recombinant animals used above and comparing the genotype with that detected with the gene markers pre-viously used. W e call this procedure pedigree break-point analysis, and it has the advantage that only a few animals, already known to be recombinant between given loci, need to be selected for mapping a new gene. We have used this method to map the position of the GABA,^ receptor a3 subunit gene in our interspecies mouse pedigree.

We mapped the mouse a3 subunit gene using a TaqI polymorphism in 62 backcross animals in our pedigree that showed recombination between the loci listed in Ta-ble 2 and have found no recombination with Rsvp (Figure 3: Table 2). One animal m the pedigree (N13) recom-bined between Rs\p and Ci-H/Cdx. The results for this animal allow us to position the a3 subunit gene of the G A B A a receptor (which we shall designate CabraJ in the mouse) centromeric to Ci-8 and Cdx on the X chro-mosome of the mouse Eighteen animals recombined between Hprl and Rs\p. including 7 that recombined between DXPjsb and R.svp. e.g.. animal B22, and in all ot these the genotype detected with the GABRa J probe was identical to that at Rsvp. Thus, the CabraJ gene maps telomeric to DXPas6. Therefore we have estab-lished that the a3 gene lies between DXPasS and Ci-8/Cdx and is close to Rsvp.

These mapping data in the mouse confirm the posi-tion of the a3 subunit gene found in the human by in situ hybridization. Furthermore, they provide the high resolution mapping required to position GABRa i rela-tive to other genetic loci in this region. The human equivalent of Rsvp (RCP) has been assigned (Vollrath et al.. 1988) to band Xq28. A very low frequency of recom-bination is seen in the human between RCP and the F8C/G6PD pair of genes (Gitschier et al., 1985; Baron et al.. 1987) and likewise in the mouse between the equiva-lent genes Rsvp and Cf-8/Gdx (Ryder-Cook et al., 1988). Thus, the mapping data on the human and mouse X chromosomes are consistent, and the order of loci on the mouse X chromosome can be used to infer an order within this region of the human X chromosome, namely:

q27 q28 - H PRT- DXSl44-GABRa 3/RCP- F8C/G DX - G6PD - tel

Discussion

The initial cloning of cDNAs for the mammalian GABAA receptor revealed sequences encoding three isoforms of the a subunit and one of the p subunit (Schofield et al., 1987; Levitan et al., 1988). W e now see that the genes

Chromosomal Locations of CABA^ Receplor Cones O J I

< N CO — «N -CQ z tn

CN CO Q o

CV

m

(D Tt ca

CD CNJ in — - O

M S

S M w

ID

u .

- 2 3 - 1

- 9 - 4

- 6 6

- 4 - 4

- 2 - 3

- 2 0

F.8urc 3. Segregation of Cahrai in Selected Recombfnani Animals from the M . muscu lus/M. spretus Pedigree

Taq I-digested genomic D N A s o f 62 backcross progeny trom the interspecies mouse pedi gree were separated by agarose gel electro phoresis, transferred to nylon filters, hybrid ized under moderately stringent conditions and washed (see Experimental Procedures) X-specitic bands of musculus (M) origin (4.6 and 9 .0 kb) and of spretus (S) origin (5.0 and 6.7 kb) are indicated in this sample of I I backcross progeny known to recombme be tween other X chromosomal loci. Additional bands ot 2.2. 3.0, 3.2. and 6.4 kb are also de-tected. Those at 2.2 and 6.4 kb appear com mon to the genomic D N A of all animals. The bands of 3.0 and 3.2 kb may be polymor-phic. but since they are weak in intensity they were not used for the determination of genotype. In total. 62 animals previously mapped with Cybb, Hprt. DXPasb. Cdx. Ci-8. Kivp. mdx. or Pgk-I probes (see Table 2) were mapped with the CABA^ a3 siibunit CDNA probe. The positions of X.-Hindlll size mark-ers are indicated in kb. The deduced geno-tyjx's of each animal on the blot are as fol-lows: 822 . M ; N I 3 . S; B4I . S; D32. MS. N22. S; C 4 7 . MS: 842. M ; 846. MS; 156. MS: C I 2 . S: F26. M ( M - musculus; S - spretus; MS = heterozygote)

for one a (a2) and one P (Pl) subunit are both located on chromosome 4 and map to the same interval; the genes for the two other a subunits (a1 and a3) are lo-cated on two other chromosomes, 5 and X, respectively. A somewhat similar situation has been found for the ace-tylcholine receptor of muscle, the only other hetero-oligomeric synaptic receptor to be analyzed wi th re-spect to chromosomal localizations. In the mouse it was found (Heidmann et al., 1986) that the a subunit gene is on chromosome 17, P is on chromosome 11, and y and 5 are closely l inked on chromosome 1. Hence, when more than one gene is needed for assembly of one of these receptor molecules, there is no requirement that these genes be physically close on a single chromosome. On the other hand, the cases of the GABA^ receptor a2 and P gene pair and of the nicotinic y and 5 gene pair suggest that a tandem arrangement of two such genes is sometimes, maintained.

For the a2 and P subunit genes, the present evidence shows only that they are in the same chromosomal band; however, the nicotinic y and 6 subunit genes have been proven by DNA sequencing to be truly adjacent in the chicken (Nef et al., 1984) and in the mouse (Gardner et al., 1987). In such cases it is inferred that a gene dupli-cation event has led to the differentiation of one subunit

type into two, without subsequent gene dispersion. This may also be associated wi th a coordinated transcription of the two genes, i.e., it may indicate that the a2 and P subunits occur in the same receptor oligomer.

Our mouse mapping study of the a3 gene locus has refined the localization that we obtained by in situ hy-bridization to the human X chromosome so that we can order the a3 locus against other sequences within the q28 band. Thus, on the basis of the mouse order we would predict that the GABRa3 gene is proximal to both the coagulation factor VIIIC and G6PD genes and very close to the red cone pigment gene, RCR

Considering possible relationships to genetic defects, it is interesting to note that the position of the a3 subunit gene on the mouse X chromosome corresponds with the position of trembly (Ty). an X-linked mouse mutant with a severe neurological defect (Taylor et al., 1978). In this spontaneous mutant, affected males exhibited tremors at 2 weeks of age and, later, seizures. They did not sur-vive beyond weaning. Some presumptive heterozygotic females showed mi ld tremors and poor coordination. Histological examination of the brains of affected males revealed no gross abnormalities. Thus, both the pheno-type and the genetic mapping data suggest that the mu-tation could have occurred in this GABAA receptor

Neuron 652

Table >. Bredk-Pomi Analvsi> oi iho GABAv Kctfpior a i Subunil Gene

Backcross Mouse

Mouse Locus D?2 B2> Nl .} FJ(> Gene Order

Cvhb Hprt

WM \tM

S S

ND S

ND MM 1

C\bb Hprt

O.VPd.s6 MS S S MM » i

l)\f'j>h

MS M S MM 1 Ks\. p/( '.,i/)fa i

Cl-8/Cdx M M MM t CiHHUh

mdx Pgkl

MS MS

M M

M M

MS MS

t i

nuh Pak 1

Cabrai MS M S MM tei

Examples oi miormative backcross progeny showing recombina-tion break-fK>ints between loci (indicated by the dashesl that per-mit ordering ot the CabraJ locus, as shown in the righthand column. The genotypes ot the animals at each locus are indicated: M. mus-culus (M). M. spretus (Si. heteroiygotes (MS). ND. not determined. Mouse loci are as defined in the text, except tor Cybb. the mousi-locus equivalent to the human cytochrome 8-245 beta polypep-tide (chronic granulomatous disease) locus (CYBB). and Pfik l. the mouse locus equivalent to the human phosphoglycerate kinase (PGK) gene.

gene. Unfortunately, this hypothesis cannot be tested, as the Ty mutation has recently been lost in an attempt to maintain it in an inbred strain (L. Mobraaten, personal communication).

Other known neurologic mutants in the mouse could have defects in the other known GABA^ receptor genes (i.e., a1, a2, and pi) . Detailed comparison of the mouse genome with the human genome has revealed many areas of preserved gene order (linkage analysis) on the autosomes (Searle et al., 1987). Genes with localizations that encompass the pl2-pl3 bands on human chromo-some 4 map on mouse chromosome 5, in a region that contains a neurologic mutant, tilted (tU). Thus, the possi-bilities should be examined that tit arises by mutation of either the a2 or P subunit genes.

The short arm of human chromosome 4, where the a2 and p genes are located, is well known as the region where the Huntington's disease gene occurs (Gusella et al., 1985). but the latter is in fact located extremely close to the telomere of 4p (Gilliam et al.. 1987) and is clearly well separated from the 4pl2-pl3 region, where the a2 and P genes lie. Hence, there is no association with the Huntington's disease locus. The positions of the a l . a2, and P subunit genes do not correspond with the posi-tions of any inherited human neurologic diseases known at present (McKusick, 1988). However, evidence has been found tor a susceptibility locus for schizophrenia (Sherrington et al.. 1988) on the 5q arm. i.e.. the arm on which the al subunit gene is located, and it will be im-portant to exclude the 5q34 region of the latter in such pedigrees.

The {)osition of the a3 subunil gene, in band Xq28, is more interesting in this respect. One form of heritable manic depressive illness (MAFD2) is X-linked and has been mapped, using linkage analysis in five large Israeli kindred (Baron et al., 1987) and in two Sardinian pedi-grees (Del Zompo et al., 1984), close to the color-blind-ness and C6PD genetic markers, i.e., in the same region as the GABA;^ receptor a3 gene. Another study (Mend-levvicz et al., 1987) has mapped X-linked manic depres-sion in ten Belgian pedigrees and found linkage to the coagulation factor IX gene in Xq27 (the only marker used) with a recombination fraction of 0.11. which could be consistent with a location in the proximal part ot Xq28.

Inherited manic depression appears lo arise also from several other genetic loci. One form in Old Order Amish pedigrees has been associated with chromosome 11 (Egeland et al., 1987). There is evidence for another form linked neither to 11 nor to X (Hodgkinson et al., 1987; Detera-Wadleigh el al., 1987). Nevertheless, since the X-lmked form has been recognized in independent stud-ies of geographically separated pedigrees and since those studies agree on a position that is so far indistinguishable from that of ihe GABA^ receptor a3 subunit gene, we conclude that ihe possibility thai the GABA.^ receptor a 3 subunit is involved in one particular form of inherited manic depression now merits further investigation.

The a3 subunit of the mammalian GABA^ receptor is a relatively minor species, and its mRNA is found at lev-els lower than those of a 1 mRNA; the main known loca-tions of a3 mRNA are in layers V and VI of some regions of the cortex and in the neonatal cerebellum, but not in the adult cerebellum (Levitan et al.. 1988; Wisden el al.. 1988). At the protein level, the a3 subunit is separable in denaturing gel electrophoresis from the major band of a subunits and is seen there as a minor component in the adult bovine or porcine brain (Duggan et al.. 1988, Biochem. Soc., abstract; Kirkness and Turner, 1988, Bio-chem. Soc., abstract). These results indicate that the a3 subunit may be present in GABA^ receptors that occur on only a small minority of neurons, and hence it is nol implausible that a defect in it would be pathogenic but nol lethal.


ONA Probes cDNA probes for DNA spot-blot hybridizations were as tollows; al. an ~2.4 kb EcoRI fragment from pbGRasense (Schotleld et al.. 19871 encoding the entire bovine a l subunit and including some 5" and 3" untranslated sequences; a2. two EcoRI fragments ot 772 and 844 bp from pbGRa2sense • K + 5" ut (Levitan et al.. 1988) encoding, respectively, irom amino acid 17 to 273 and from amino acid 274 to the C-terminus of the mature bovine a2 subunit (the latter fragment also contains some 3' untranslated sequence): a3. an ~l .3 kb EcoRI fragment from pbCRa3 (Levitan et al., 1988) en-coding from the start of the putative signal peptide lo amino acid 384 of the mature bovine a3 subunit and containing some 5' un-translated sequence: P. an ~720 bp EcoRI fragment from pt)GR3 sense (Schofield et al.. 1987) encoding from the start of the signal peptide to amino acid 199 of the mature bovine (J subunit and con-taining some 5' untranslated sequence. These were labeled to high sjKKific activity (~I0" cpm/^g) using |*-PJdCTP (Amersham) and

p romosoma l Locations of C A B A ^ Receplor Genes 653

the random hexanucleotide method of Feinberg and VogeUtein (1983).

cDNA probes for in situ hybridization of chromosomes were as follows: al. whole plasmid. pBS + , containing an ~1.7 kb EcoRI in-sert (Schofield et al.. 1989) that encodes the entire human al subunit and includes some 5' and 3' untranslated sequences; a2. the probe used for DNA spot-blots (seeabove); ai. whole plasmid. pbGRa3 (Levitan et al., 1988), i.e., a probe that encodes the entire bovine a3 subunit and contains 5' and 3" untranslated sequences: PI, whole plasmid. pBS + . containing an ~1.9 kb EcoRI insert (Schofield et al., 1989) that encodes the entire human pi subunit and includes some 5' and 3' untranslated sequences. These were labeled by nick translation (Rigby el al.. 1977) using ('H]dCTP (Amersham) to a specific activity of dpm/ng.

For mapping in the interspecies mouse pedigree, the bovine a 3 subunit cONA probe was the same as that used for DNA spot -blots (see above). This was labeled using the random hcxanucleoiido method (Feinberg and Vogelstein, 1983).

ONA Spot-Blot Hybridizations Nitrocellulose filters containing sorted human chromosomes (U-Imj et al.. 1984) w«re prehybridized in 6x SSC ( Ix SSC - 0 !.=> M mi dium chloride. 0.015 M sodium citrate). 0.5% (w/v) SDS. 5*. Den hardfs solution. 100 ug/ml denatured salmon sperm DNA at 55-C for 4-6 hr. Hybridization was performed overnight in the same so-lution at 55°C. Filters were then washed in 0.2x SSC at 55°C tor 45 min and exposed to X-ray film at -70°C with an intensiiymij screen.

In Situ Hybridization of Chromosomes Chromosome preparations were obtained from normal male lym-phocytes. Bromodeoxyuridine was incorporated into the chro-mosomal DNA diKing the first half of S phase in order to provide a replication banding pattern (Dutrillaux and Viegas-Pequignot 1981). The in situ hybridization techniques used have been de-scribed in detail elsewhere (Buckle and Craig. 1986). Briefly, slides vv-ere treated with RNAase A (100 jig/ml in 2x SSC) at 37®C for 1 hr and denatured at 65°C for 4 min in 70% (v/v) formamide in 2x SSC. Labeled probe was lyophilized. resuspended at 0.3-0.6 ng/m' in hybridization buffer (50% (v/v) formamide. 5x SSPE (pH 7.2). 5x Denhardfs solution. 10% (w/vj dextran sulphate, 1(X) Mg/ml salmon sperm DNA). and denatured by boiling ( Ix SSPE - 0.18 M sodium chloride, 0.01 M sodium dihydrogen orthophosphate. 0.001 M EDTA). Probes were hybridized overnight, at 10-20 ng of DNA per slide, under a coverslip at 43°C. The slides were washed extensively in 2x SSC at room temperature, followed by 0.1x-0.5x SSC at 60''C-63''C, and then dehydrated. The slides were dipped in Ilford L4 liquid emulsion, left to expose at 4®C for 10-23 days, developed in D19 for 5 min at 20°C, stained with Hoechst 33258 (10 in 2x SSC) for 25 min. exposed to long-wave UV light for 30 min under 2X SSC, and stained in 10% (v/v) Ciemsa (pH 6.8) for 15 min. Only grains actually touching the chromosomes were scored. Clusters of grains were noted, but scored as a single hybridization event on the ideograms illustrating grain distribution.

Southern Blot Analysis of Mouse Pedigree DNA An interspecies mouse pedigree, segregating for a number of X chromosome loci, was used as described in detail elsewhere (Ryder-Cook et al., 1988). In brief, an inbred female M. musculus mouse carrying the X-linked mutation mdx, which is located within the mouse dystrophin gene (Sicinski et al.. 1989), was crossed to a normal male M. spretus mouse. Five female F, heterozygote progeny were backcrossed to an inbred male M. musculus mouse carrying the mdx mutation, producing over 350 progeny; the proportion of these progeny with a musculus, spretus, or heterozy-gote genotype was close to the predicted backcross ratio of 2:i: i .

Genomic mouse DNA from the backcross progeny in the inter-specific pedigree was isolated (Crosschedl et al., 1984). digested with Taql. electrophoresed in a 0.8% (w/v) agarose gel, and trans-ferred to a nylon membrane (Hybond N; Amersham). Hybridiza-tion with the probe was at 42"C in 50% (v/v) formamide, 3x SSC, 2% (w/v) SOS. 5x Denhardt's solution, ICX) jig/ml denatured salmon sperm DNA, 5% (w/v) dextran sulphate for 18 hr. Filters were

washed in 2x SSC, 0,1% (w/v) SDS at 65°C for 30 min prior to ex-posuro to X-ray film at - 70°C with an intensifying screen.

Rcct'ivcd February 7. 1989; revised August 1, 1989.

References

Amar. L. C.. Dandolo. L.. Hanauer, A.. Ryder-Cook, A.. Arnaud. D.. Mandcl. ).-L.. and Avner.R (1988). Conservation and reorganization of loci on the mammalian X chromosome: a molecular framework tor the identification of homologous subchromosomal regions in man and mouse. Genomics 2. 220-230. Avner. P.. Amar. L.. Dandolo. L.. and Guenet, J.-L. (1988). Genetic analysis of the mouse using interspecific crosses. Trends Genet 4 18-23. Baron. M.. Risch. N., Hamburger. R., Mandel, B.. Kushner. S.. New-man. M.. Drumer. D.. and Belmaker, R. H. (1987). Genetic linkage between X-chromosome markers and bipolar affective illness. Na-ture J26. 289-292. Buckle, v. I . and Craig, 1. VV. (1986). In siw hybridisation. In Hu-man Genetic Diseases. K. E. Davies. ed. (Oxford: IRL Press), pp 8S-1(X). Del Zompo. M.. Bocchetta, A.. Goldin. L. R., and Corsini. G. U (1984). Linkage between X-chromosome markers and manic-de-pre>4ive illness: two Sardinian pedigrees. Acta Psych. Scand. 70. J82-287. Uetera-Wadleigh. S. D . Berrettini. W. H., Goldin, L. R.. Boorman. D.. Anderson. S.. and Gershon. E. S. (1987). Close linkage of c-Har-vey-ra.v-l and the insulin gene to affective disorder is ruled out m three North American pedigrees. Nature 325, 806-808. Dutrillaux. B., and Viegas-Pequignot, E. (1981). High resolution R- and G-banding on the same preparation. Hum. Genet. 57. 93-95. Egeland. |. A.. Gerhard. D. S.. Pauls. D. L., Sussex. ). N.. Kidd. K K.. Allen. C. R.. Hostetter. A. M., and Housman, D. E. (1987). Bipolar affective disorders linked to DNA markers on chromosome n . Nature 325. 783-787. Feinberg, A. P. and Vogelstein. B. (1983). A technique for radio-labeling DNA restriction endonuclease fragments to high specific activity. Anal. Biochem. 132. 6-13. Gardner, P., Heinemann, S . and Patrick, J. (1987). Transcriptional regulation of nicotinic acetylcholine receptor genes: identification of control elements of a y-subunit gene. Mol. Brain Res. 3, 69-76. Gilliam, T. C.. Tanzi. R. E.. Haines, |. L., Bonner, T. I., Faryniarz, A. G., Hobbs. W. MacDonald. M. E., Cheng. S. V , Folstein, S. E., Conneally, P. M., Wexler, N. S., and Gusella. J. F. (1987). Localiza-tion of the Huntington's disease gene to a small segment of chromo-some 4 flanked by D4S10 and the telomere. Cell 50. 565-571. Gitschier, Drayna, D., Tuddenham, E. G. D.. White, R. L., and Lawn. R. M. (1985). Genetic mapping and diagnosis of haemophilia A achieved through a Bell polymorphism in the factor VIII gene. Na-ture 314. 738-740. Grosschedl, R.. Weaver, D., Baltimore, D.. and Costantini, F. (1984). Introduction of a n immunoglobulin gene into the mouse germ line: sp>ecific expression in lymphoid cells and synthesis of functional an-tibody. Cell 38. 647-658.

Gusella. |. F.. Tanzi. R. E., Bader. R I.. Phelan, M. C , Stevenson, R., Hayden. M. R.. Hofman. K. |., Faryniarz, A . G.. and Gibbons, K. (1985). Deletion of Huntington's disease-linked C8 (D4S10) locus in Wolf-Hirschhorn syndrome. Nature 318. 75-78. Heidmann. O., Buonanno, A., Geoffroy, B., Robert, B.. Guenet, J.-L.. Merlie, |. P, and Changeux, |.-P. (1986). Chromosomal local-ization of muscle nicotinic acetylcholine receptor genes in the mouse. Science 234, 866-868. Hodgkinson. S.. Sherrington, R.. Curling, H., Marchbanks. R.. Reeders, S.. Mallet, Mclnnis. M., Petursson, H.. and Brynjolfsson, I. (1987). Molecular genetic evidence for heterogeneity in manic depression. Nature 325. 805-806. Lebo. R. v.. Gorin. F.. Fletterick. R. |.. Kao, F.-T., Cheung. M.-C..

Neuron 654

Brucc. B. D.. j n d Kan. Y. W. (1984). H.gh-resolul ion chromosome sorting and DNA spot -b lo t analysis assign McArdle's syndrome to chromosome 11. Science 225. 57-59.

Levit jn. E S . Scholield. P. R.. Burt. O. R.. Rhee. L. M. , Wisden. W.. Kohler. M.. Fujita. N.. Rodr iguei . H . F., Stephenson, A.. Darl ison. M C. . Barnard. E. A., and Seeburg. P. H. (1988). Structural and lunct ional basis tor CABA;^ receptor heterogeneity. Nature J35. 76-79.

McKusick. V. A. (1988). Mende l ian Inheri tance in Man; Catalogs o i Autosomal Dominant. Autosomal Recessive, and X-Linked Pheno-types. 8th edit ion (Baltimore: The lohns Hopkins University Press) Mendlewicz. Simon, P., Sevy. S.. Charon. F.. Brocas, H. . Legros S.. and Vassart. C . (1987). Polymorphic D N A marker on X chromo-some and manic depression. Lancet i. 1230-1232. Net. R. Mauron. A.. Stalder. R., A l l iod, C.. and Ballivet. M . (1984). Structure, linkage, and sequence of the two genes encoding the & and y subunits of the nicot inic acety lchol ine receptor. Proc. Nat l Acad Sci. USA 81. 7975-7979.

Ohno. S. (1969). Evolution of sex chromosomes in mammals. Annu . Rev. Genet i . 495-524.

Rigby. R W. I . Dieckmann. M. . Rhodes. C.. and Berg, R (1977) Labeling deoxyribonucleic acid to high specific activity in vitro by nick translation wi th D N A polymerase 1.1. Mo l . Biol. 113. 237-251 Ryder-Cook. A. S.. Sicinski, R, Thomas. K.. Davies. K. E., Wor ton . R C.. Barnard. E. A.. Darl ison. M . G. . and Barnard. P. |. (1988). Lo-calization of the mdx mutat ion w i th in the mouse dystrophin gene EMBO I. 7. 3017-3021.

Scholield. R R.. Darlison, M . C . , Fujita. N. . Burt. D. R.. Stephen-son. F. A.. Rodriguez. H.. Rhee. L. M . , Ramachandran, Reale. V.. Glencorse, T. A.. Seeburg. R H. . and Barnard, E. A. (1987). Se-quence and functional expression of the GABA^ receptor shows a ligand-gated receptor super-family. Nature 328. 221-227. Schotield. R R.. Pritchett, D. 8., Soniheimer. H. , Kettenmann. H. . and Seeburg, R H. (1989). Sequence and expression of human CABAa receptor a l and Pl subunits. FEBS Lett. 244. 361-364. Searle, A. C., Peters. Lyon. M. F.. Evans. E. R. Edwards. |. H. . and Buckle, V. I. (1987). Chromosome maps of man and mouse. III. Genomics 1. 3-18.

Sherrington. R., Brynjolfsson. Petursson. H.. Potter. M. , Dud-leston, K.. Barraclough. B.. Wasmuth. Dobbs. M. . and Gur l ing. H. (1988). Localization of a susceptibi l i ty locus for schizophrenia on chromosome 5. Nature 336. 164-167.

Sicinski, R, Ceng, Y.. Ryder-Cook, A. S.. Barnard. E. A., Darl ison. M. G., and Barnard. P. J. (1989). The molecular basis of muscular dystrophy in the mdx mouse: a point mutat ion. Science 244. 1578-1580.

Taylor, B. A.. Meier. H., MacPike, A., and Wil l iams. M . (1978). New mutants and biochemical variants: t rembly (ty). Mouse Nevi/s Lett 59. 25.

Toniolo, D., Ptrsico. M . G. . and Alcalay, M . A. (1988). Housekeep-ing gene on the X chromosome encodes a prote in similar to ub iqu i -t in. Proc. Natl. Acad. Sci. USA 55, 851-855. Vollrath, D.. Nathans. and Davis. R. W. (1988). Tandem array of human visual pigment genes at Xq28. Science 240. 1669-1672. Wisden. W.. Morris. B. J., Darl ison, M . G. , Hunt , S. R. and Barnard. E. A. (1988). Distinct GABA^ receptor a subunit mRNAs show differential patterns of expression in bovine brain. Neuron I . 937-947.

Wisden. W.. Morris. B. Darl ison. M . G. . Hunt , S. R. and Barnard. E. A. (1989). Localization of GABAA receptor a-subuni t mRNAs in relation to receptor subtypes. M o l . Brain Res. 5. 305-310.

Note Added in Proof

The P subunit gene localized herein should now be referred to as p i . since the existence of t w o addi t ional GABA^ receptor P subunit genes has recently been revealed by c D N A c lon ing (Ymer et a!.. 1989. EMBO |. 8. 1665-1670).

In addi t ion, the work cited ai, B iochem. Soc. abstracts has been publ ished: Dugganet al.. 1989. B iochem. Soc. Trans. 17. 769-770; Kirknoss and Turner. 1989. Biochem. Soc. Trans. 17. 767-768.

The following articles have been removed from the digital copy of this thesis. Please see the print copy of the thesis for a complete manuscript.

Title: The role of receptor subtype diversity In the CNS Authors: Peter R. Schofield, Brenda D. Shivers and Peter H. Seeburg Journal: Perspectives Title: Alpha subunit variants of the human glycine receptor: primary structures, functional expression and chromosomal localization of the corresponding genes Authors: Gabriele Grenningloh, Volker Schmieden, Peter R.Schofield, Peter H.Seeburg, Teepu Siddique, Thuluvancheri K.Mohandas, Cord-Michael Becker and Heinrich Betz Journal: The EMBO Journal Title: Structural and functional characterization of the y^ subunit of GABA a/benzodiazepine receptors Authors: Sanie Ymer, Andreas Draguhn, William Wisden, Pia Werner, Kari Keinanen, Peter R.Schofield, Rolf Sprengel, Dolan B.Pritchett and Peter H.Seeburg Journal: The EMBO Journal Title: Amino-terminal leucine-rich repeats in gonadotropin receptors determine hormone selectivity Authors: T.Braun, P.R.Schofield and R.Sprengel. Journal: The EMBO Journal. Title: Biological Activities of Recombinant Human Follicle Stimulating Homione Authors: Peter R Schofield, Anna Cerpa-Pouak, Mark F. Albrecht, Margaret C Stuart, and Yvonne J. Hort. Journal: Follicle Stimulating Hormone Regulation of Secretion and Molecular Mechanisms of Action Title: Purification and Characterization of Recombinant Human Follicle Stimulating Hormone Authors: Glenn M Smith, Leonora A. Bishop, Robert DeKroon, Greg Wright, Anna Cerpa-Pouak, and Peter R Schofield. Journal: Follicle Stimulating Hormone Regulation of Secretion and Molecular Mechanisms of Action Title: Antagonism of ligand-gated ion channel receptors: Two domains of the glycine receptor a subunit form the strychnine-binding site. Authors: Robert J. Vandenberg, Chris R. French, Peter H Barry, John Shine and Peter R Schofield Journal: Proceedings of the National Academy of Sciences of the United States of America Title: Distinct Agonist- and Antagonist-Binding Sites on the Glycine Receptor Authors: Robert J. Vandenberg, Cheryl A. Handford, and Peter R. Schofield Journal: Neuron. Title: Isoelectric Charge of Recombinant Human Follicle-Stimulating Hormone Isoforms Determines Receptor Affinity and in Vitro Bioactivity Authors: Anna Cerpa-Pouak, Leonora A. Bishop, Yvonne J. Hort, Catherine K. H. Chin, Robert De Kroon, Stephen M. Mahler, Gleen M. Smith, Margaret C. Stuart and Peter. R. Scholfield Journal: Endocrinology Title: The Importance of Being Inhibited: Brain GABA A and Glycine Authors: Robert J. Vandenberg & Peter R. Schofield Journal: Today’s Life Science

https://www.google.com.au/url?sa=t&rct=j&q=&esrc=s&source=web&cd=&cad=rja&uact=8&ved=2ahUKEwiCw9WK7arpAhXIxTgGHQMyC7kQ6F56BAgGEAI&url=https%3A%2F%2Ftwitter.com%2FgregwrightYP%3Fref_src%3Dtwsrc%255Egoogle%257Ctwcamp%255Eserp%257Ctwgr%255Eauthor&usg=AOvVaw3jRsGV_cv4x-s8gtoz8NJN

Title: The Extracellular Disulfide Loop Motif of the Inhibitory Glycine Receptor Does Not Form the Agonist Binding Site Authors: Robert J. Vandenberg, Sundran Rajendra, Chris R. French, Peter H Barry, Peter R Schofield Journal: Molecular Pharmacology Title: Inhibitory Ligand-Gated Ion Channel Receptors: Molecular Biology and Pharmacology of GABA A and Glycine Receptors Authors: Robert J. Vandenberg and Peter R Schofield Journal: Handbook of Membrane Channels: Molecular and Cellular Physiology Title: A Threonine Residue in the Seventh Transmembrane Domain of the Human Ai Adenosine Receptor Mediates Specific Agonist Binding Authors: Andrea Townsend-Nicholson and Peter R. Schofield Journal: The Journal of Biological Chemistry Title: Localization of the Glycine Receptor ofi Subunit Gene (GLRA1) to Chromosome 5q32 by FISH Authors: Elizabeth Baker, Grant R. Sutherland, and Peter R. Schofield. Journal: Genomics Title: The Functional Structure of the human Inhibitory Glycine Receptor Authors: Sundran Rajendra, Joseph W. Lynch, Robert J. Vandenberg, Kerrie D. Pierce, Chris R. French, Peter H. Barry and Peter R. Schofield. Journal: Studies in Honour of Karl Julius Ullrich: An Australian Symposium Title: Exclusion of close linkage of bipolar disorder to the G^-a subunit gene in nine Australian pedigrees Authors: F. Le, P. Mitchell, C. Vivero, B. Waters, J. Donald, L.A. Selbie, J. Shine, P. Schofield. Journal: Journal of Affective Disorders Title: Startle Disease Mutations Reduce the Agonist Sensitivity of the Human Inhibitory Glycine Receptor Authors: Sundran Rajendrat, Joseph W. Lynch, Kerrie D. Pierce, Chris R. French, Peter H. Barry, and Peter R. Schofield Journal: The Journal of Biological Chemistry Title: Specific Roles for the Asparagine- Linked Carbohydrate Residues of Recombinant Human Follicle Stimulating Hormone in Receptor Binding and Signal Transduction Authors: Leonora A. Bishop, David M. Robertson, Nicholas Cahir, and Peter R. Schofield Journal: Journal of Molecular Endocrinology Title: A missense mutation in the gene encoding the a^ subunit of the inhibitory glycine receptor in the spasmodic mouse Authors: Stephen G. Ryan, Marion S. Buckwalter, Joseph W. Lynch, Cheryl A. Handford, Lillian Segura, Rita Shiang, John J. Wasmuth, Sally A. Camper, Peter Schofield, & Peter O'Connel Journal: Nature Genetics Title: Mutation of an Arginine Residue in the Human Glycine Receptor Transforms p-Alanine and Taurine from Agonists into Competitive Antagonists Authors: Sundran Rajendra, Joseph W. Lynch, Kerrie D. Pierce, Chris R. French, Peter H. Barry, and Peter R. Schofield Journal: Neuron

Title: Both of the B-Subunit Carbohydrate Residues of Follicle- Stimulating Hormone Determine the Metabolic Clearance Rate and in Vivo Potency Authors: Leonora A. Bishop, Tuan V. Nguyen, Peter R. Schofield Journal: Endocrinology Title: Localization of the Adenosine A2b Receptor Subtype Gene (AD0RA2B) to Chromosome 17p11.2-p12 by FISH and PGR Screening of Somatic Cell Hybrids Authors: Andrea Townsend-Nicholson, Elizabeth Baker, Grant R. Sutherland, and Peter R. Schofield Journal: Brief Reports Title: Mutations Affecting the Glycine Receptor Agonist Transduction Mechanism Convert the Competitive Antagonist, Picrotoxin, into an Allosteric Potentiator Authors: Joseph W. Lynch, Sundran Rajendra, Peter H. Barry, and Peter R. Schofield Journal: The Journal of Biological Chemistry Title: Molecular mechanisms of inherited startle syndromes Authors: Sundran Rajendra and Peter R. Schofield Journal: Trends in Neurosciences Title: Localization of the Adenosine A1 Receptor Subtype Gene (AD0RA1) to Chromosome 1q32.1 Authors: Andrea Townsend-Nicholson, Elizabeth Baker, Peter R. Schofield, and Grant R. Sutherland Journal: Genomics Title: The unique extracellular disulfide loop of the glycine receptor is a principal ligand binding element Authors: Sundran Rajendra, Robert J.Vandenberg, Kerrie D.Pierce, Anne M.Cunningham, Peter W.French, Peter H.Barry and Peter R.Schofield Journal: The EMBO Journal Title: Expression of Human Thyrotropin in Cell Lines with Different Glycosylation Patterns Combined with Mutagenesis of Specific Glycosylation Sites Authors: Mathis Grossmann, Mariusz W. Szkudlinski, Joseph E. Tropea, Leonora A. Bishop, N. Rao Thotakura, Peter R. Schofield, and Bruce D. Weintraub Journal: The Journal of Biological Chemistry Title: A mutation in codon 111 of the amyloid precursor protein gene in an Australian family with Alzheimer's disease Authors: William S. Brooks, Ralph N. Martins, Joke De Voechts, Garth A. Nicholson, Peter R. Schofield, John B.J. Kwok, Christopher Fisher, Leone U. Yeung, Christine Van Broeckhoven Journal: Neuroscience Letters Title: The human glycine receptor B subunit: primary structure, functional characterisation and chromosomal localisation of the human and murine genes Authors: Cheryl A. Handford, Joseph W. Lynch, Elizabeth Baker, Graham C. Webb, Judith H. Ford, Grant R. Sutherland, Peter R. Schofield Journal: Molecular Brain Research

Publications on Membrane Receptors and Their Ligands by ...

Documents

Transcript of Publications on Membrane Receptors and Their Ligands by ...