Mutagenesis of three surface-exposed loops of a Bacillus thuringiensis insecticidal toxin reveals...

Downloaded from www.microbiologyresearch.org by

IP: 54.224.135.207

On: Sat, 04 Jun 2016 11:16:27

Microbiology (1 996), 142, 161 7-1 624 Printed in Great Britain

Mutagenesis of three surface-exposed loops of a Bacillus thuringiensis insecticidal toxin reveals residues important for toxicity, receptor recognition and possibly membrane insertion

Damian P. Smedley and David J. Ellar

Author for correspondence: David J . Ellar. Tel: +44 1223 333651. Fax: +44 1223 333345. e-mail : D JEl @MOLE.BIO.CAM.AC.UK

Department of Biochemisl University of Cambridge, Tennis Court Road, Cambridge CB2 lQW, UK

:ry, Information on the molecular determinants of receptor recognition, membrane insertion and toxin pore-formation was sought by making 42 single and multiple substitutions of residues 312-314 (GYY), 367-370 (YRRP) and 438-441 (SGFS) in the Baci//us thuringiensis insecticidal CrylAc G-endotoxin by site-directed mutagenesis. These three regions correspond to three putative surface-exposed loops (loops 1,2 and 3, respectively) in domain II of the 6- endotoxin, forming the molecular apex of the structure. All except mutants GFY (loop l), YKRA, SRRA, YRKA (loop 2) and TGFS (loop 3) expressed 6- endotoxin protein at wild-type levels which was stable upon activation by Pieris brassicae gut extract or trypsin. Toxicity assays for all the fully stable mutants using Manduca sexta larvae showed that 6312, Y367, R368, R369,5438 and 6439 are important for activity. Wild-type toxin was then labelled in vivo with [35S]methionine and heterologous competition binding assays were carried out for all the mutants using brush border membrane vesicles prepared from Manduca sexta midgut. Most and least conservative mutations of G439 and least conservative substitutions of Y367, R368 and R369 reduced the ability of the toxin to bind competitively. The most conservative mutation, S441T, gave significantly increased binding. These results suggested that these four residues play a role in the initial receptor binding step in the toxin mechanism. As no significant effect on binding affinity was observed in relatively non-toxic mutants in which residues 6312 and 5438 were mutated, we suggest that these residues are involved in the subsequent steps of membrane insertion and pore-formation.

Keywords : Bucihs tburingiensis d-endotoxin, site-directed mutagenesis, competition binding assay

INTRODUCTION

Bacillw thritzgietzsis produces crystalline protein inclusions during sporulation which are variously toxic to Lepido- pteran, Dipteran and Coleopteran insect larvae. The inclusions (6endotoxins) are composed of one or more polypeptides (protoxins) which require solubilization and proteolysis in the insect gut to form an active toxin. It is

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Abbreviations: BBMV, brush border membrane vesicles; BSA, bovine serum albumin fraction V.

thought that the d-endotoxin mechanism involves two steps whereby the activated toxin first binds to specific, high-affinity receptors in the luminal plasma membrane of midgut epithelial cells (Hofmann e t a!., 1988a, b; Van Rie e t al., 1990) and then inserts into the membrane to create a 1-2 nm ‘pore’ leading to colloid osmotic lysis (Knowles & Ellar, 1987). The 8-endotoxins have been divided into five major classes (Cry1 to CryV) based upon their sequence similarity and insect specificity (Hofte & Whiteley, 1989; Tailor e t al., 1992). The lepidopteran- active CryI-type d-endotoxins (Hofte & Whiteley, 1989) consist of 130-135 kDa protoxins and are processed to a

0002-0516 0 1996 SGM 1617


IP: 54.224.135.207

On: Sat, 04 Jun 2016 11:16:27

D. P. SMEDLEY and D. J. ELLAR

1

Fig. 1. Crystal structure of CrylllA toxin (modified from Li et a/., 1991). A schematic ribbon representation of the three domains (1-111) of CrylllA toxin is shown, with the positions (arrowed) of the three surface-exposed domain II loops (1, 2 and 3) forming the molecular apex of the toxin.

protease-resistant active toxin of 60-65 kDa by insect gut proteases (Huber & Luthy, 1981). The structure of the coleop teran- specific Cry111 A 8-endo toxin determined by X-ray crystallography (Fig. 1 ; Li e t al., 1991) consists of three domains. Domain I comprises the N-terminal 250 amino acids folded into a bundle of seven amphipathic helices. This feature is highly conserved amongst the Cry toxins (Li e t al., 1991) and it has been proposed that it is involved in ‘ pore-formation ’ by analogy with the helical bundle pore-forming structures of colicin A toxin (Parker e t al., 1989) and diphtheria toxin (Choe e t al., 1992) and evidence from several workers that the central helix (as) is involved in the ‘ pore ’ formation (Ahmad & Ellar, 1990 ; Wu & Aronson, 1992; Gazit & Shai, 1993). Domain I1 consists of three anti-parallel p-sheets, each terminating in a loop (Fig. l), packed around a hydrophobic core and is thought to participate in receptor binding. Domain I11 is a p-sandwich of two anti-parallel sheets. Since the core of the molecule including all the domain interfaces is built up from five highly conserved blocks of amino acids among the d-endotoxins, it has been suggested that the other 6- endotoxins will have a similar structure (Li e t al., 1991). Ligand blotting has been used to identify toxin receptors for several insects (Garczynski e t al., 1991 ; Oddou e t a!.,

1991 ; Vadlamudi e t al., 1993) and the 120 kDa CryIAc toxin receptor from Mandma sexta brush border membrane vesicles (BBMV) has recently been cloned and identified as aminopeptidase N (Knight e t al., 1994,1995; Sangadala e t al., 1994). A previous study has shown that N-acetylgalactosamine is a likely component of this receptor (Knowles e t al., 1991). Attempts to identify the receptor-binding regions of the toxins have been carried out by several groups. Ge e t al. (1989) managed to transfer high Bombyx mori toxicity to CryIAc toxin by exchanging amino acids 332-450 in domain I1 with the equivalent segment of CryIAa and later this transfer in activity was correlated with a change in binding properties towards B. mori BBMV (Lee e t al., 1992). Further segment swapping between these same two toxins showed residues 335-450 in domain I1 of CryIAc are associated with Tricboplt/sia ni activity and residues 335-61 5 (domains 11-111) with Helioth virescens activity (Ge e t al., 1991). Independent segment-swapping studies on the same toxins showed that segments 332-428 (domain II), 429-447 (domain 11) and 448-722 (domain 11-111) of CryIAc contributed additively towards H. virescens activity whilst only a swap of the 429-447 segment affected M. sexta activity, with a 27-50 fold reduction in CryIAc activity observed (Schnepf e t al., 1990). Site- directed mutagenesis approaches have also been used. The ability of the CryIAa toxin to bind competitively to B. mori was decreased by deleting residues 365-371 in the predicted loop 2 of CryIAa (Lu e t al., 1994). Recently it was found that individual alanine substitutions of residues 370-375 in the predicted loop 2 of CryIAb toxin do not alter initial (reversible) binding to M. sexta BBMV but do affect the ability to bind irreversibly (Rajamohan e t al., 1995). The possibility of regions outside domain I1 being involved in receptor binding was raised by a mutagenesis study in which residues S503 and S504 in a domain I11 loop of CryIAc were reported to be involved in toxicity and binding to M. sexta and H. virescens (Aronson e t al., 1995). Chemical modification of CryIAc residues showed that modifying 4 of the 43 arginine residues or 7 of the 27 tyrosines resulted in greatly reduced toxicity and binding to M. s o d a BBMV (Cummings & Ellar, 1994). Finally, mutational analysis of predicted residues in loops 1 and 2 of CryIC toxin showed their importance in specificity determination for Aedes aegypti and a Spodoptera frugiPerda (SB) cell line (Smith & Ellar, 1994). In this study we have investigated the role in toxicity and binding of residues in three predicted domain I1 loops of the CryIAc d-endotoxin. These loops (loop 1, 312GYY314; loop 2, 3B7YRRP37,; loop 3, 438SGFS4,1) correspond to the three loops forming the molecular apex and putative receptor binding site in the CryIIIA crystal structure (Fig. 1 ; Li e t al., 1991).

METHODS

Sitedirected mutagenesis. The plasmid pSV J WKa contains the entire cr~yIAc gene cloned from B. thuringiensis subsp. AurstuAi HD-73 (supplied by the late Dr H. Dulmage) in the vector pSV2 (both plasmids constructed by Dr N. Crickmore, School of Biological Sciences, University of Sussex, UK). Mutations in

1618


IP: 54.224.135.207

On: Sat, 04 Jun 2016 11:16:27

Mutagenesis of a B a c i h thwingiensis toxin

cylAc were performed with the template pMSV.CryIAc. This was created by introducing a HincII site downstream of the cyIAc gene and subcloning the 4 kb HincII fragment containing the entire gene into SmaI-treated pMSV1. pMSVl (also constructed by Dr N. Crickmore) is a mutagenic shuttle vector based on PALTER (Promega) but with the tetracycline resistance gene replaced with a chloramphenicol resistance gene and a B. tburingiensis origin of replication. The site-directed mutagenesis procedure (Altered-Sites in vitro Mutagenesis System ; Promega) was carried out following the manufacturer's instructions and using oligonucleotides synthesized by the Oligonucleotide Synthesis Facility, Department of Biochem- istry, University of Cambridge, UK. Double-stranded sequencing was performed following the method of Hsaio (1991) to denature the DNA and the method of Sanger e t al. (1977) for sequencing using the Sequenase kit (United States Biochemical) as recommended by the manufacturer. All restriction enzymes were purchased from Pharmacia and chemicals purchased from standard commercial sources unless stated otherwise. B. thuringiensis electrotransformation and purification of 6- endotoxin inclusions. DNA was transformed into the acrystal- liferous strain, B. tburingiensis subsp. isruelensis IPS-78/11, by electroporation (Bone & Ellar, 1989). Growth and crystal purification were as previously described (Thomas & Ellar, 1983). The yield of crystal was determined by the Lowry method, using bovine serum albumin fraction V (BSA) as a standard. Solubilization and trypsin treatment. The expressed proteins were solubilized by incubation in 50 mM Na2C03/10 mM dithiothreitol (DTT), pH 10.5, at 37 OC for 60 min. Concen- trations of solubilized samples were determined using the Bio- Rad protein microassay following the manufacturer's instructions, with BSA as a standard. Proteolytic stability of the solubilized samples was assessed either by digestion with 1 : 5 (w/w) trypsin (Sigma):toxin at 37 OC for 1 h followed by addition of a further 1 : 10 (w/w) trypsin: toxin and 10 % (v/v) 2 M Tris pH 7 and further incubation at 37 "C overnight, or by digestion with 5% (v/v) Pieris brassicae gut extract, pH 10.5 (prepared as described by Knowles e t al., 1984), at 37 OC for 1 h. All samples were analysed by SDS-PAGE (13%, w/v, acryl- amide) as described by Laemmli & Favre (1973). 35S-labelling of S-endotoxin. In vivo 35S-labelling of CryIAc was carried out based on the method of Koller e t al. (1992). Fifty millilitres of PWYE medium (Ellar & Posgate, 1974) in a 250 ml flask was inoculated with 5 ml of pre-grown B. tburingiensis PWYE culture (6 h growth) and incubated at 30 OC, 200 r.p.m., overnight. The harvested pellet was washed once in 20 ml GSM medium (Koller e t al., 1992) and then resuspended in 20 ml GSM medium containing 5 mCi (185 MBq) [35S]methionine (Amersham ; in vivo cell-labelling grade). The culture was then grown until > 95% lysis occurred (approximately 24 h) and the harvested pellet washed twice in deionized H20 , twice in 2 M NaCl and twice more in deionized H20. The pellet was then solubilized and activated by treatment with trypsin as described above and the activated toxin desalted on a PD-10 column (Pharmacia) pre-equilibrated with phosphate- buffered saline (PBS; 8 mM NaHP0,/2 mM KH2P0,/ 150 mM NaC1, pH 7-4). The specific activity obtained was 5.2 x lo4 c.p.m. pmol-l. Insect bioassays. Toxicity assays were performed with neonate M a n d m sexta larvae reared from eggs supplied by Dr S. Reynolds, School of Biological Sciences, University of Bath, UK. Solubilized samples of toxin were diluted in PBS and 20 p1 of each dilution layered onto 1 cm2 discs of artificial diet (Poitout & Bues, 1972), allowed to dry and a single larva placed on each disc using tweezers. Mortality was scored after 7 d. LC,,

values and 95% confidence intervals were determined by the program of Lieberman (1983) for probit analysis (Finney, 1952). Brush border membrane vesicle preparation. BBMV were prepared by the Mg2+ precipitation method (Wolfersberger e t al., 1987) using final-instar M. sexta larvae reared on an artificial diet (Poitout & Bues, 1972). The final pellet was resuspended at approximately 5 mg protein ml-' in PBS and the precise protein concentration determined using the Bio-Rad protein microassay following the manufacturer's instructions and using BSA as a standard. Competitive binding assay. BBMV (10 pg) were incubated with 1 nM activated 35S-CryIAc toxin in the presence of increasing concentrations (0-200 nM) of activated unlabelled toxin and PBS/O*l% BSA buffer in a total volume of 100 p1 at room temperature for 30 min in a 1.5 ml microcentrifuge tube. The BBMV pellet containing bound toxin was collected by centrifugation at 13500 g for 15 min at room temperature using an LKB microcentrifuge, washed three times in PBS/O.l% BSA buffer and resuspended in 4 ml liquid scintillation fluid (Optisafe) by shaking vigorously before measuring the radio- activity with a Beckman LS 3801 counter. Binding data were analysed by the LIGAND computer program (Munson & Rodbard, 1980).

RESULTS

Mutagenesis of the loop regions

Amino acid alignments with the known secondary structure elements of CryIIIA toxin (Li e t a/., 1991) and a computer structure prediction program (PEPTIDESTRUC-

(a) Loop 1

Predicted loop sequence: G Y Y Most conservative change: A F F Least conservative change: v s s Mutagenic oligonucleotide:

GATGCTCATAGGGGTTATTATTATTGGTCCAGGGC C T T T C C

(b) Loop2

Predicted loopsequence: Y R R P Mostconservativechange: F K K S Leastconservativechange: S I I A Mutagenic oligonucleotide:

TAGAACATTATCGTCGACTTTATAGAAGACCTTTTAATATAATAGGGAT T A A T C T T G

(c) Loop 3

Predicted loopsequence: S G F S Mostconservativechange: T A Y T Leastconservativechange: I V I Mutagenic oligonucleotide:

TGTTTCAATGTTTCGAAGTGGCTTTAGTAATAGTAGTGTAAGT C C A C T T C T

............................................................................. . ............................................................................ Fig. 2. Choice of amino acid substitutions in (a) loop 1, (b) loop 2 and (c) loop 3, and sequences of the oligonucleotides designed to introduce these changes to the coding region. The residue exchange matrices of Bordo & Argos (1991) were used as a guide in choosing the most and least conservative mutations at each position in loops 1, 2 and 3 of the CrylAc toxin.

1619


IP: 54.224.135.207

On: Sat, 04 Jun 2016 11:16:27

D. P. S M E D L E Y a n d D. J. ELLAR

Table 1. CrylAc loop mutants produced by site-directed mutagenesis with degenerate mutagenic oligonucleotides

~~~

Wild-type Single Double Triple mutants mutants mutants

(a) Loop 1 GYY GYF GFF VSF

GFY GFS GSY GSF GY S ASY AYY VSY

VYF (b) Loop 2

YRRP Y RIP SKRP YKKA YRRA YRKA FKRA YTRP YKRA YRKP SRRA

FIRP FRRA FKRP

(c) Loop 3 SGFS SGSS TGY S TASS

TGFS IVFS NGYI SGY S SVFI SAFS SAYS SVFS SASS SGFT IVFS SDFS IGFI IGFS

TURE on the Genetics Computer Group package, Uni- versity of Wisconsin) were used to predict the three or four most likely residues in the three domain I1 loops forming the molecular apex of the CryIAc toxin. The residues chosen were 312GYY314 (loop l), 367YRRP370 (loop 2) and 438SGFS441 (loop 3). The residue-exchange matrices of Bordo & Argos (1991) were used as a guide in choosing the most and least conservative mutations to introduce. The mutations chosen and the degenerate mutagenic oligonucleotides are shown in Fig. 2. Using degenerate oligonucleotides produces a range of possible mutants involving single or multiple substitutions. Mutants were identified by sequencing and the range identified is shown in Table 1. Note that a few substitutions were identified that were unexpected from the oligonucleotide design (YTRP in loop 2; NGYI and SDFS in loop 3). The reason for this is unknown but as the surrounding sequence was not altered these mutants were still studied.

Expression, solubilization and activation of crystal proteins

The mutants were expressed and crystal protein purified. Mutants GFY (loop l), YKRA (loop 2) and TGFS (loop 3) were expressed at very low levels, suggesting they had

folded badly or that new protease sites had been created; therefore they were not studied further. Loop 2 mutants YRKA and SRRA expressed a stable 40 kDa product rather than the full-length 133.1 kDa protoxin. N-terminal sequencing (data not shown) showed that this was due to cleavage at the mutated loop region during expression. SDS-PAGE analysis of the remainder showed that they expressed the 133.1 kDa protoxin at wild-type levels and yielded 60-65 kDa stable products upon activation by Pieris brassicae gut extract or trypsin. Representative samples of this analysis are shown in Fig. 3.

Toxicity assays

Initial bioassay of the wild-type toxin gave an LC,, of 11.3 ng cm-2 (95 YO confidence interval 7-5-18.9 ng cmW2) for M. sexta neonates. Subsequent bioassays on the wild- type and mutants were carried out at a concentration of 40 ng cm-2, which gives approximately 99 YO kill for the wild-type toxin. All assays were carried out with eight larvae, repeated twice; the results are shown in Table 2. This type of bioassay was chosen instead of determining LC,, values for each of the mutants because all the mutants from one loop can be assayed with a single batch of larvae. This reduces variation in the data due to differing larval susceptibilities to the toxin from batch to batch.

Competitive binding studies

Representative traces of the heterologous competition assays are shown in Fig. 4. Each assay was carried out twice and the mean values used in the analysis. The results of analysis of the binding curves to give Kd values (Table 2) clearly show that residues in loops 2 and 3 but not loop 1 appear to play an important role in reversible binding of CryIAc to M. sexta midgut membrane vesicles. Note that the Kd values obtained are not the true K d values for the mutants as these can only be obtained from homologous competition assays involving labelled mutant toxins and the equivalent unlabelled mutants. The values reported may well match those that would be obtained from homologous competition assays but as it stands they are merely a means of comparing the reversible binding affinity of the wild-type and mutants.

DISCUSSION

All the mutants except GFY (loop l), YKRA, SRRA, YRKA (loop 2) and TGFS (loop 3) expressed 133.1 kDa d-endotoxin at wild-type levels, suggesting that they had folded correctly, as misfolding or protease sensitivity can correlate with poor expression (Almond & Dean, 1993). The fact that the stably expressed mutants showed the same solubility characteristics as the wild-type and gave stable products when activated by proteases also suggested that they had folded normally. Therefore the effects on toxicity and binding seen for these mutants are not likely to be due to changes in folding. Results from the loop 1 mutagenesis (Table 2a) were especially interesting. The non-conservative G312V

-

1620


IP: 54.224.135.207

On: Sat, 04 Jun 2016 11:16:27

Mutagenesis of a B a c i l h thzrringiensis toxin

.... ................................................................ . .................................... .. ....... ..... .......................... * ...... ....*........ .... . ..................................................................................................... . ........ . ........... ..... ....... . ............ . .... ... Fig. 3. Coomassie blue-stained SDS-13 % (w/v) polyacrylamide gel showing (a) the protoxins and P. brassicae gut extract- activated toxins and (b) the trypsin-activated toxins. (a) Protoxins and activated toxins were run consecutively: lanes 1 and 2, CrylAc; lanes 3 and 4, GYF (loop 1 mutant); lanes 5 and 6, WF (loop 1 mutant); lanes 7 and 8, YRlP (loop 2 mutant); lanes 9 and 10, FKRP (loop 2 mutant); lanes 11 and 12, SGSS (loop 3 mutant); lanes 13 and 14, TASS (loop 3 mutant). (b) Trypsin-activated samples of CrylAc and the above six mutants run in the same order. Each lane contained 3 pg of toxin except lanes 1, 2, 5 and 6 in (a), which contained 5 pg. Molecular mass marker positions are indicated to the left.

change produced a significant decrease in toxicity (compare GYY, GSF and GYF with VSF and VYF) but this was not accompanied by a significant decrease in reversible binding affinity. One possibility to explain this result is that this residue is important in pre- or post- binding events in the toxin mechanism such as long-term stability of the activated toxin to gut proteases, or membrane insertion or pore-formation. In future experiments toxin stability can be checked by prolonged exposure to midgut conditions, and membrane insertion and pore formation can be assayed by the methods of Rajamohan e t al. (1995) or Carroll & Ellar (1993). Table 2(a) shows that the other residues in loop 1 (Y313, Y314) accommodate the most or least conservative substitutions without large changes in toxicity or binding, provided that residue 312 is the wild-type glycine or its most conservative replacement, alanine.

In loop 2 the data (Table 2b) show that Y367 could be replaced by phenylalanine without significant effects on toxicity or binding provided arginine or lysine was present at position 368. The activity and binding of this type of mutant was also unaffected when they additionally contained alanine instead of P370. Replacement of Y367 with serine produced a marked drop in toxicity and some reduction in binding (compare YRKP and SKRP). The role of tyrosine in binding and toxicity has been high- lighted in a previous study, where chemical modification of seven or more of the 27 tyrosines in the CryIAc toxin completely abolished toxicity and decreased binding to M. sexta BBMV (Cummings & Ellar, 1994). Replacement of either R368 or R369 with the slightly shorter lysine side-chain was without effect provided Y367 was un- changed. When both arginines were replaced by lysine there was a dramatic reduction in toxicity but not in binding (compare YRRA and YKKA). Replacement of

R368 by isoleucine or threonine reduced binding significantly but the effect on toxicity was slight. However when R369 was replaced by isoleucine both toxicity and binding were dramatically lower. The importance of arginine residues is again mirrored in the results obtained by Cummings & Ellar (1994), where chemical modification of four or more arginines in the CryIAc toxin was enough to abolish toxicity and dramatically reduce binding to M. sexta BBMV. The results for loop 2 match the findings of Lu e t al. (1994), where deletion or alanine substitution of the equivalent segment of CryIAa (,,,LYYRIIL,,,) greatly decreased the toxicity and reversible binding affinity to B. mori, suggesting that the same region is involved in binding for both toxins and insect receptors.

In loop 3 (Table 2c) S438 is important for toxicity and cannot be replaced by the conservative substitution of threonine (compare SGYS and TGYS). Replacement of S438 by isoleucine (IGFS) produced a mutant similar to VSF and VYF in loop 1 which has lowered toxicity but no change in binding. Once again as with these loop 1 mutants this phenotype suggests that the post-binding events of membrane insertion and/or pore-formation may be hindered by this change. G439 is similarly important and even the conservative change to alanine (SAFS) abolishes toxicity and decreases the binding affinity more than tenfold. F340 appears not to be critical (compare SGFS and SGSS) and S441 can be replaced conservatively by threonine without any change in toxicity. The identification of S438 and G439 as being important for M. sexta activity is in agreement with the segment-swapping studies of Schnepf e t al. (1990), in which the region containing these residues was shown to be critical for CryIAc toxicity towards M . sexta. Glycine is a residue often associated with turn formation and the

1621


IP: 54.224.135.207

On: Sat, 04 Jun 2016 11:16:27

D. P. SMEDLEY and D. J. ELLAR

I I I I

00 --* FRRA

---*--- FKRP 80

60

40

I 1 I I

L + W(SGFS)

100

80

60

40

20 I I I

0.1 1 10 100 1000 Competitor concn (nM)

Fig, 4. Representative binding curves from heterologous competition assays involving the loop mutants: (a) loop 1, (b) loop 2 and (c) loop 3. M. sexta BBMV (1Opg) were incubated with 1.0 nM 35S-labelled CrylAc toxin in the presence of various concentrations of unlabelled CrylAc or loop mutants in a total volume of lOOp1. Binding is expressed as a percentage of the amount bound with no competitor added. These amounts were 821, 785, 777, 850, 843, 797, 802, 765, 794 and 823c.p.m. for wild-type CrylAc, GFF, AYY, VSF, FRRA, FKRP, YRIP, SGYS, IVFS and SASS, respectively.

large decrease in binding affinity seen when this residue was mutated suggests that the conformation of loop 3 may be critical for the receptor-binding interaction. Several of the mutants showed slightly lower Kd values than the wild-type7 suggesting that they bound with greater affinity to the M. sexta BBMV. Whilst for most cases this increase in binding is probably within the errors of the assay system and hence not genuine, for those fully toxic mutants with evidence of increased binding, it is possible that a true increase in binding and toxicity has occurred. For example, in the case of the fully toxic loop 3 mutant, SGFT, the Kd was reduced by more than ninefold. Bioassays at 20 ng cm-2 allow any increases in toxicity relative to the wild-type to be observed as only 60-70% mortality is observed for CryIAc against M.

Table 2. Toxicity and binding affinities of the CrylAc loop 1 ,2 and 3 mutants toward M. sexta

Toxin Toxicity (%) Kd (nM) f S E M t

f SEM*

CryIAc (GYY) GFF GSF GFS GSY ASY AYY GYF GY S VSY VYF VSF

CryIAc (YRRP) FRRA FKRA YRKP FKRP YRRA YTRP FIRP SKRP YKKA Y RIP

CryIAc (SGFS) SGY S SGFT SGSS IGFS TGYS IGFI NGYI SASS SAYS TASS SAFS SVFS SVFI IVFS SDFS

(a) Loop 1 mutants 100 f 0 1OOfO 9 7 f 3 97+3 9 7 f 3 9 7 f 3 9 4 f 0 9 4 f 0 84+3 81 f 0 1 6 f 3 1 3 f 6

(b) Loop 2 mutants 1OOfO 100 f 0 100 f 0 100 f 0 9 4 f 6 91 f 9 8 8 f 6 84f 16 2 8 f 2 1 3 f 6 9 f 3

(c) Loop 3 mutants 1 O O f O 100 f 0 l O O f 0 9 7 f 3 3 8 f 0 38f13 1 6 f 9 9 f 3 9 + 9 3 f 3 3 f 3 3 f 3 3 f 3 o + o Of0 Of0

4.8 f 1.1 4 1 f2.0 3.2 f 0.9 5.8 f 0.9 5.0f 1.6

10-4f05 2.2 f 0 3 1.4 f 0.4 1.8f0.2 1.2 f 0.4 6.6 f 2.8 6.3 f 1.0

4.8 f 1.1 2.4 f 0.5 6.2 f 0.7 2.7 f 0.4 4 8 f 0.5

10.1 f 3.1 27.9 f 6.3 31-4 f 7.3 11.5 f 2.3 10.3 f 0.8 35.3 f 8.5

4 8 f 1.1 7.4f 1-6 0-5 f 0.3 3-8 f 1-1 4.8 f 0.3 9.0 f 1.6

16.5 f 4.0 13.6 f 1.3 27.2 f 6.6 17.0 f 4.6 23.5 f 2.9 59.2 f 18.9 62.0 f 10.1 39.4 -f 6.7 68.7 f 19.8 70.9 f 5-3

* Toxicity at 40 ng cm-' concentration. Each value is the mean of two experiments with eight larvae. t Kd, dissociation constant determined from heterologous competition binding assays (except for the wild-type toxin which is from an homologous assay). Each value is the mean of two experiments.

sexta. A duplicate bioassay on the loop 3 mutant, SGFT, at this concentration gave 71.9 % f 9.4 % (mean f SEM) toxicity compared to 6843 % f 6.3 YO for the wild-type,

1622


IP: 54.224.135.207

On: Sat, 04 Jun 2016 11:16:27

Mutagenesis of a Bacillus thringiensis toxin

revealing that the toxicity towards M. sexta was not significantly increased, despite the change in binding affinity. Further investigation and mutagenesis of this and the other mutants with apparently raised binding affinity should prove interesting.

The demonstrated involvement of predicted loop 2 and loop 3 residues in CryIAc &endotoxin binding does not exclude the possibility that other regions are also involved, such as the domain I11 loop identified by Aronson e t al. (1995). The VMO-1 protein in the vitelline membrane of the hen’s egg is composed entirely of a structure which is remarkably similar tc the bendotoxin domain I1 (Shimizu e t a/., 1994). The fact that VMO-1 binds oligosaccharides and that N-acetylgalactosamine has been shown to be a likely component of the CryIAc receptor (Knowles e t al., 1991) is strong indirect evidence for the role of domain I1 in insect receptor recognition.

ACKNOWLEDGEMENTS

T h s work was supported by the Medical Research Council. We acknowledge the Cambridge Centre for Molecular Recognition for oligonucleotide synthesis and the Wellcome Trust for providing these facilities. We thank D r Neil Crickmore, Dr Joe Carroll, Munah Abdul Rauf and Trevor Sawyer for technical help and Richard Summers and Kim Rowsell for photographic assistance.

REFERENCES

Ahmad, W. & Ellar, D. J. (1990). Directed mutagenesis of selected regions of a Bacillus tburingiensis entomocidal protein. FEMS Microbiol Lett 68, 97-104.

Almond, B. D. & Dean, D. H. (1993). Structural stability of Bacillus tburingiensis &endotoxin homolog-scanning mutants determined by susceptibility to protease. Appl Environ Microbiol 59, 2442-2448.

Aronson, A. I., Wu, D. & Zhang, C. (1995). Mutagenesis of specificity and toxicity regions of a Bacillus tburingiensis protoxin gene. J Bacteriol 177, 4059-4065.

Bone, E. J. & Ellar, D. J. (1989). Transformation of Bacillus tburingiensis by electroporation. FEMS Microbiol Lett 58, 171-178.

Bordo, D. & Argos, P. (1991). Suggestions for ‘safe’ residue substitutions in site-directed mutagenesis. J Mol Biol217, 721-729.

Carroll, J. & Ellar, D. J. (1993). An analysis of Bacillus tburingiensis 6- endotoxin action on insect-midgut-membrane permeability using a light-scattering assay. Eur J Biocbem 214, 771-778.

Choe, S., Bennett, M. J., Fujii, G., Curmi, P. M. G., Kantardjieff, K. A., Collier, R. J. & Eisenberg, D. (1992). The crystal structure of diphtheria toxin. Nature 357, 21 6-222.

Cummings, C. E. & Ellar, D. J. (1994). Chemical modification of Bacillus tburingiensis activated &endotoxin and its effect on toxicity and binding to Manduca sexta midgut membranes. Microbiology 140,

Ellar, D. J. & Posgate, J. A. (1974). Characterisation of forespores isolated from Bacillus megaterium at different stages of development into mature spores. In Spore Research 1973, pp. 21-40. Edited by A. N. Barker, G. W. Gould & J. Wolf. London: Academic Press. Finney, D. J. (1952). Probit AnaEysis, 2nd edn. Edited by D. J. Finney. Cambridge : Cambridge University Press. Garuynski, L. F., Crim, J. W. &Adang, M. J. (1991). Identification

2737-2747.

of putative insect brush border membrane-binding molecules specific to Bacillus tburingiensis d-endotoxin by protein blot analysis. Appl Environ Microbiol57, 281 6-2820.

Gazit, E. & Shai, Y. (1993). Structural and functional characterisation of the a5 segment of Bacillus tburingiensis bendotoxin. Biocbemisty 32, 3429-3436.

Ge, A. Z., Shivarova, N. 1. & Dean, D. H. (1989). Location of the Bombyx mori specificity domain on a Bacillus tburingiensis delta- endotoxin protein. Proc Natl Acad Sci U S A 86, 4037-4041.

Ge, A. Z., Rivers, D., Milne, R. & Dean, D. H. (1991). Functional domains of Bacillus tburingiensis insecticidal crystal proteins : refine- ment of Heliotbis virescens and Tricboplusia ni specificity domains on CryIAc. J Biol Cbem 266, 17954-1 7958.

Hofmann, C., LUthy, P., Hutter, R. & Pliska, V. (1988a). Binding of the delta-endotoxin from Bacillus tburingiensis to brush-border membrane vesicles of the cabbage butterfly (Pieris brassicae). Eur J Biocbem 173, 85-91.

Hofmann, C., Vanderbruggen, H., Hofte, H., Van Rie, J., Jansens, 5. & Van Mellaert, H. (1988b). Specificity of Bacillus tburingiensis 6- endotoxins is correlated with the presence of high-affinity binding sites in the brush border membrane of target insect midguts. Proc Natl Acad Sci U S A 85, 7844-7848.

Hofte, H. & Whiteley, H. R. (1989). Insecticidal crystal proteins of Bacillus tburingiensis. Microbiol Rev 53, 242-255.

Hsaio, K. C. (1991). A fast and simple procedure for sequencing double stranded DNA with Sequenase. Nucleic Acids Res 19, 2797.

Huber, H. E. & LUthy, P. (1 981). Bacillus tburingiensis delta endotoxin : composition and activation. In Pathogenesis of Invertebrate Microbial Diseases, pp. 209-234. Edited by E. D. Davidson. Totowa, N J : Allanheld, Osmum & Co. Knight, P. J. K., Crickmore, N. & Ellar, D. J. (1994). The receptor for Bacillus tburingiensis CryIA(c) delta-endotoxin in the brush border membrane of the lepidopteran M. sexta is aminopeptidase N. Mol Microbiol 11, 429-436.

Knight, P. 1. K., Knowles, B. H. & Ellar, D. J. (1995). Molecular cloning of an insect aminopeptidase N that serves as a receptor for Bacillus tburingiensis CryIA(c) toxin. J Biol Cbem 270, 17765-1 7770.

Knowles, B. H. & Ellar, D. J. (1987). Colloid-osmotic lysis is a general feature of the mechanism of action of Bacillus tburingiensis 6- endotoxins with different specificity. Biocbim Biopbs Acta 924,

Knowles, B. H., Thomas, W. E. & Ellar, D. J. (1984). Lectin-like binding of Bacillus tburingiensis var. hrstaki lepidopteran-specific toxin is an initial step in insecticidal action. FEBS Lett 168,

Knowles, B. H., Knight, P. J. K. & Ellar, D. J. (1991). N- Acetylgalactosamine is part of the receptor in insect gut epithelia that recognises an insecticidal protein from Bacillus tburingiensis. Proc R Soc Lond B 245, 31-35.

Koller, C. N., Bauer, L. 5. & Hollingworth, R. M. (1992). Charac- terisation of the pH-mediated solubility of Bacillus tburingiensis var. san diego native d-endotoxin crystals. Biocbem Biopbys Res Commun

Laemmli, U. K. & Favre, M. (1973). Maturation of the head of bacteriophage T4. J Mol Biol 80, 575-599.

Lee, M. K., Milne, R. E., Ge, A. 2. & Dean, D. H. (1992). Location of Bombyx mori receptor binding region of a Bacillus tburingiensis delta- endotoxin. J Biol Cbem 267,3115-3121.

Li, J., Carroll, J. & El!ar, D. J. (1991). Crystal structure of insecticidal bendotoxin at 2.4 A resolution. Nature 353, 815-821.

Lieberman, H. R. (1983). Estimating LD,, using the probit

509-51 8.

197-202.

184, 692-699.

1623


IP: 54.224.135.207

On: Sat, 04 Jun 2016 11:16:27

D. P. S M E D L E Y and D. J. ELLAR

technique : a basic computer program. Drug Chem Toxicoi 6 ,

Lu, H., Rajamohan, F. & Dean, D. H. (1994). Identification of amino acid residues of Bacillus thuringiensis d-endotoxin CryIAa associated with membrane binding and toxicity to Bomlyx mori. J Bucterioll76, 5554-5559. Munson, P. J. & Rodbard, D. (1980). Data analysis and curve fitting for ligand binding experiments. And Biochem 107, 220-239. Oddou, P., Hartmann, H. & Geiser, M. (1991). Identification and characterisation of Heiiothis virescens midgut membrane proteins binding Bacillus thuringiensis b-endotoxins. Eur J Biochem 202,

Parker, M. W., Pattus, F., Tucker, A. D. &Tsernoglou, D. (1989). Structure of the membrane-pore-forming fragment of colicin A. Nature 337, 93-96. Poitout, 5. & Bues, R. (1972). Nutrition des insects. C R AcadSciSer

Rajamohan, F., Alcantara, E., Lee, M. K., Chen, X. J., Curtiss, A. & Dean, D. H. (1995). Single amino acid changes in domain I1 of Bacillus thuringiensis CryIAb b-endotoxin affect irreversible binding to Munduca sextu midgut membrane vesicles. J Bucterioi 177, 2276-2282. Sangadala, S., Walters, F. S., English, L. H. & Adang, M. J. (1994). A mixture of Munducu sexta aminopeptidase and phosphatase enhances Bacillus thuringiensis insecticidal CryIAc toxin binding and *‘Rb+-K+ efflux in vitro. J Biol Chem 269, 10088-10092. Sanger, F., Nicklen, 5. & Coulson, A. R. (1977). DNA sequencing with chain-terminating inhibitors. Proc Nutl Acud Sci U S A 7 4 ,

Schnepf, H. E., Tomczak, K., Ortega, J. P. & Whiteley, H. R. (1990). Specificity-determining regions of a lepidopteran-specific insecticidal protein produced by Buciiius thuringiensis. J Bioi Chem 265, 20923-20930.

11 1-116.

673-680,

D 274, 3113-3115.

5463-5467.

Shimizu, T., Vassylyev, D. G., Kido, S., Doi, Y. & Morikawa, K. (1 994). Crystal structure of vitelline membrane outer layer protein I (VMO-I) : a folding motif with homologous Greek key structures related by an internal three-fold symmetry. EMBO J 13,1003-1010. Smith, G. P. & Ellar, D. 1. (1994). Mutagenesis of two surface- exposed loops of the Buciiius thuringiensis CryIC b-endotoxin affects insecticidal specificity. Biochem J 302, 61 1-61 6. Tailor, R., Tippett, J., Gibb, G., Pells, S., Pike, D., Jordan, L. & Ely, 5. (1 992). Identification and characterisation of a novel Buciiius thuringiensis S-endotoxin entomocidal to coleopteran and lepidopteran larvae. Moi Microbioi 6 , 1211-1217. Thomas, W. E. & Ellar, D. J. (1983). Buciiius thuringiensis var. isruelensis crystal b-endotoxin : effects on insect and mammalian cells in vitro and in vivo. J Cell Sci 60, 181-197. Vadlamudi, R. K., Ji, T. H. & Bulla, L. A., Jr (1993). A specific binding protein from Munduca sexta for the insecticidal toxin of Bacillus thuringiensis subsp. beriiner. J Bioi Chem 286, 12334-1 2340. Van Rie, J., Jansens, S., Hofie, H., Degheele, D. &Van Mellaert, H. (1990). Receptors on the brush border membrane of the insect midgut as determinants of the specificity of Buciiius thuringiensis delta-endotoxins. Appl Environ Microbiol56, 1378-1 385. Wolfersberger, M., LUthy, P., Maurer, A., Parenti, P., Sacchi, V. F., Goirdana, B. & Hanozet, G. M. (1987). Preparation and partial characterisation of amino acid transporting brush border membrane vesicles from the larval midgut of the Cabbage butterfly (Pieris brassicue). Comp Biochem Pbysiol86A, 301-308. Wu, D. & Aronson, A. 1. (1992). Localised mutagenesis defines regions of the Bacillus thuringiensis S-endotoxin involved in toxicity and specificity. J Biol Chem 267, 231 1-2317.

Received 7 November 1995; revised 15 January 1996; accepted 8 February 1996.

1624

Mutagenesis of three surface-exposed loops of a Bacillus thuringiensis insecticidal toxin reveals...

Documents

Transcript of Mutagenesis of three surface-exposed loops of a Bacillus thuringiensis insecticidal toxin reveals...