Eur J Biochcm. 184, 455-464 (1989) (C FEBS 1989

Glucosephosphate isomerase from Trypanosoma brucei Cloning and characterization of the gene and analysis of the enzyme

Martine MARCHAND', Ulco KOOYSTRA', Rik K . WIERENGA', Anne-Marie LAMBETR', Jos Van BEEUMEN ', Fred R. OPPERDOES and Paul A. M. MICHELS'

' International Institute of Cellular and Molecular Pathology, Research Unit for Tropical Diseases, Brussels Laboratory of Chemical Physics, State University Groningen Laboratory of Microbiology and Microbial Gcnetics, State Univcrsity Ghcnt

(Received March 23/May 31, 1989) - EJB 89 0360

In Trypanosomu brucei the enzyme glucose-6-phosphate isomerase, like most other enzymes of the glycolytic pathway, resides in a microbody-like organelle, the glycosome. Here we report a detailed study of this enzyme, involving a determination of its kinetic properties and the cloning and sequence analysis of its gene. The gene codes for a polypeptide of 606 amino acids, with a calculated M, of 67280. The protein predicted from the gene sequence has 54 - 58% positional identity with its yeast and mammalian counterparts. Compared to those other glucose-6-phosphate isomerases the trypanosomal enzyme contains an additional 38 -49 amino acids in its N- terminal domain, as well as a number of small insertions and deletions, The additional amino acids are responsible for the 5-kDa-larger subunit mass of the T. brucei enzyme, as measured by gel electrophoresis. The glucose-6- phosphate isomerase of the trypanosome has no excess of positive residues and, consequently, no high isoelectric point, in contrast to the other glycolytic enzymes that are present in the glycosome. However, similar to other glycosomal proteins analyzed so far, specific clusters of positive residues can be recognized in the primary structure.

Comparison of the kinetic properties of the T. brucei glucose-6-phosphate isomerase with those of the yeast and rabbit muscle enzymes did not reveal major differences. The three enzymes have very similar pH profiles. The affinity for the substrate fructose 6-phosphate ( K , = 0.122 mM) and the inhibition constant for the competitive inhibitor gluconate 6-phosphate (Ki = 0.14 mM) are in the same range as those of the similar enzymes. The K, shows the same strong dependence on salt as the rabbit muscle enzyme, although somewhat less than the yeast glucose-6-phosphate isomerase. The trypanocidal drug suramin inhibits the T. brucei and yeast enzymes to the same extent (Ki = 0.29 and 0.36 mM, respectively), but it had no effect on the rabbit muscle enzyme. Agaricic acid, a potent inhibitor of various glycosomal enzymes of T. brucei, has also a strong, irreversible effect on glucose-&phosphate isomerase, while leaving the yeast and mammalian enzymes relatively unaffected.

Trypanosomes have a unique compartmentation of the glycolytic pathway. Most of their glycolytic enzymes are localized within cytoplasmic organelles which resemble mi- crobodies and which have been called glycosomes [I]. Such a compartmentation enables the trypanosome to carry out glycolysis at an extremely high rate [2]. Although no direct evidence exists for the presence of a well-ordered multi-en- zyine complex inside the glycosome, tunnelling of glycolytic metabolites through the glycosome has been demonstrated in the case of intact Trypunosonza brucei cells and it has been shown that the individual enzymes of glycolysis have a strong tendency to stick together 131. Even after removal of the sur- rounding membrane by detergent treatment, the constituent enzymes can only be solubilized in the presence of salt. Glucose-6-phosphate isomerase is an exception in that it is the only enzyme that is easily solubilized upon rupture of the glycosomal membrane.

Correspondence to P. A. M. Michels, ICPiTROP 74.39, Avenue Hippocrate 74, B-1200 Brussels, Belgium

Enzymes. Glucose-6-phosphate isomerase (EC 5.3.1.9); fructose- 1 ,&bisphosphate aldolase (EC 4.1.2.1 3); triosc-phosphate isomerase (EC 5.3.1.1); ~-glyceraldehyde-3-phosphate dehydrogenase (EC 1.2.1.12); phosphoglyceratc kinase (EC 2.7.2.3).

Note. These sequence data will appcar in the EMBL/GenBank/ DDBJ Nucleotidc Scquencc Databases under thc accession number x15540 PGI.

Almost all glycosomal enzymes involved in glycolysis have high isoelectric points (pl), ranging over 8.8 - 10.2 [4]. This is I - 4 higher than those of their mammalian counterparts and even 3 - 6 higher when compared with those of various other unicellular organisms. Glucose-6-phosphate isomerase, again, is the only exception to this rule, with a p l of 7.6. By analyzing the predicted amino acid sequences of four of the glycosomal enzymes (aldolase, triosephosphate isomerase, o-glyceralde- hyde-3-phosphate dehydrogenase and phosphoglycerate kinase) we were able to identify unique sequence elements, often associated with insertions specific for the trypanosome polypeptides, that are responsible for the observed high isoelectric points [5 - 81. In the case of three of these enzymes, Wierenga et al. [9] were able to localize these additional posi- tive charges into two distinct areas (hot spots) on the surface of the individual enzymes, approximately 4 nm apart. Similar positively charged areas could not be found in homologous enzymes from other organisms. Hart el al. [lo] have shown that the glycosomal enzymes are synthesized on free ribo- somes in the cytosol and are subsequently transported into the glycosome, without any form of processing or secondary modification. Therefore, it was hypothesized by Wierenga et al. 191 that such positive hot spots could serve as the topogenic signals responsible for import of these enzymes into the gl ycosome.

Recently a serious objection against the hot-spot hypoth- esis was raised by Swinkels et al. [ l l ] who demonstrated

456

that no charge difference exists between the glycosomal and cytosolic isoenzymes of phosphoglycerate kinase from Crithi- dia fasciculatu, a closely related member of the Trypanoso- ma tidue.

Alternative explanations for the presence of hot spots could be that they would function in keeping the complex of glycosomal enzymes together, or that they would be necessary for the proper function of the glycolytic enzymes inside the glycosomes : they could, for instance, neutralize the negative charges of the phosphorylated intermediates of glycolysis, which, in T. brucei, are present at high concentrations inside the glycosome [4].

To better understand why glucose-6-phosphate isomerase has properties different from all other glycosomal enzymes studied so far, and to provide an answer as to how glycosomal enzymes end up inside glycosomes, we have started a detailed study of this particular enzyme from T. brucei. Except for some fragmentary studies by Hanon and Parr [12], Misset et al. [4] and Widmer et al. [13], trypanosomal glucose-6- phosphate isomerase has not been the subject of any detailed investigation. The enzyme catalyses the first isomerization stcp in glycolysis : the interconversion of glucose 6-phosphate and fructose 6-phosphate. In trypanosomes, as in other organ- isms, the enzyme occurs as a dimer of approximately 130 kDa molecular mass [4, 141.

Amino acid sequences of affinity labelled, active-site peptide fragments from the enzymes of pig muscle [15, 161, rabbit muscle and human placenta [17] have been published. Recently, the complete nucleotide sequences of glucose-6- phosphate isomerase from Succhurornyces cerevisiue [18], pig [19] and a partial sequence from the mouse have become available [20]. Surprisingly, the sequence of the mouse enzyme was completely identical to the sequences of the mouse neurotrophic factor neuroleukin [20, 211. Moreover, a 90% sequence identity was demonstrated between mouse neuro- leukin and pig glucose-6-phosphate isomerase [19, 211.

A high-resolution three-dimensional crystal structure is not yet available for glucose-6-phosphate isomerase. Muir- head and coworkers have obtained a 0.35-nm X-ray diffrac- tion data set of the pig muscle enzyme, in which most of the main chain could be traced [22]. Later a 0.26-nm electron- density map was obtained [I 51.

Here we report the cloning and analysis of the T. hrucei gene for glucose-6-phosphate isomerase and compare the kin- etic properties of the trypanosome enzyme with its counter- parts from rabbit muscle and yeast.

MATERIALS AND METHODS

Trypanosomes

rats and harvested as described by Opperdoes et al. [23].

Molecular biology

Trypanosomal DNA was prepared as described by Bernards et al. [24]. A genomic library of T. hrucei DNA was constructed in Escherichiu coli NM539 using EcoRI-digested DNA, ligated into the phage vector /1 EMBL4 [25]. Ligation to EMBL4 arms, in vitro packaging and infection of bacteria were performed, following the instructions of Promega (USA). Recombinant clones containing the trypanosomal glucose-6-phosphate isomerase gene were selected by screen- ing of this library with a yeast gene probe [26] kindly donated by Prof. F. Zimnierman and Dr. A. Aguilera (Technische

Bloodstream forms of T. brucei stock 427 were grown in

Hochschule, Darmstadt, FRG). Colony hybridization was performed at a stringency of 3 x NaCl/Cit (0.15 M NaCl, 15 mM sodium citrate, pH 7.0), 6 0 T , in the presence of 10Yu dextran sulphate. Post-hybridization washes were carried out for 2 h at 60°C with 5 x NaCl/Cit, 0.1 % SDS, followed by a 1 h incubation with 3 x NaCl/Cit, 0.1% SDS. High-titre phage lysates were obtained and DNA was purified from the phages using a combination of methods described by Maniatis et al. [27] and by Amersham International PLC (UK), with several modifications. The inserts of the recombinant phage clones were liberated by digestion with EcoRI and subcloned into the plasmid pUC9. Bacterial hosts for recombinant plasmids were the E. coli strains JM 109 [28] or XLI-Blue (Stratagene, USA), which were made competent and transformed accord- ing to the procedure of Hanahan [29].

Growth of bacterial cells containing recombinant plasmids and the isolation of plasmid molecules by an alkaline lysis procedure were done according to Maniatis et al. [27].

Nuclear DNA and cloned recombinant DNA were digest- ed by restriction enzymes, following the instructions of the manufacturer (Promega, USA). The digested products were size-fractionated by electrophoresis in 89 mM Tris/89 mM boric acid/2 mM EDTA, pH 8.3, (Tris/borate/EDTA) buffer, in 0.5 - 1.0% (mass/vol.) agarose gels and blotted onto nitro- cellulose or nylon (Hybond) filters (Amersham, UK) after acid depurination [30].

T. brucei chromosomes were separated by pulsed-field gradient electrophoresis as described by Johnson and Borst [31]. Samples for electrophoresis were prepared by embedding and lysing trypanosomes in agarose blocks as described by Van der Ploeg et al. [32]. Southern blots of size-fractionated chromosomes were made as described above.

Hybridizations were performed as described [24], using DNA probes that were radioactively labelled by nick trans- lation [33]. DNA fragments were isolated from recombinant plasmids by cleavage with the appropriate restriction endo- nuclease, followed by size fractionation in low-melting-point agarose gels and, after melting the agarose, purification by DEAE-cellulose column chromatography [34].

For sequence analysis DNA templates, either single- or double-stranded, could be prepared after subcloning of ap- propriate genomic fragments into the phagemid vector pTZ18R. Overlapping fragments for sequence analysis were obtained by progressive shortening of the subcloned frag- ments using the Erase-a-Base system (Promega, USA). Single- stranded DNA was produced following the instructions of Pharmacia LKB (Sweden). The double- and single-stranded DNAs were used as templates in the sequencing reactions with dideoxynucleoside triphosphates as chain-stoppers [35, 361, [E-~’P]~CTP or [LX-~’S]~ATP (Amersham, UK) as label and universal oligonucleotides as primers [37]. Occasionally new 12 - 16-nucleotide primers were used (Eurogentec, Belgium) corresponding to a part of the glucose-6-phosphate isomerase gene sequence that had already been determined. The sequencing reaction mixtures were electrophoresed on pre- heated 6 - 8% polyacrylamide gels in the presence of 50% urea. Gels were subsequently treated with 10% acetic acid, 10% methanol, dried and autoradiographed. All autoradio- graphy was performed with Hyperfilm-MP (Amersham, UK).

Enzymology

Glycosomes were purified from T. brucei bloodstream forms and glucose-6-phosphate isomerase was isolated essen-

457

Fig. 1 . Phj)sic.al map ojthe glucose-6-phosphate isomerase gene in T. brucei. The solid block indicates the protein-coding sequence. Part of the map is enlarged to illustrate the sequencing strategy. Two DNA fragments (SalI- SalI and Hind111 -BamHI) were subcloned in both orientations in the piasmid pTZ18R. These recombinant clones were subjected to unidirectional digestion with exonuclease 111, generating a set of subclones with progressive deletions over several thousand base pairs and which formed the basis of our DNA sequencing strategy. The direction and extent of the sequence determination are indicated by arrows. All sequence determinations were performed more than once on independently isolated subcloned fragments. €3, BarnHI; E, EcoRI; H, HindIIl; K, KpnI; P, Pstl; S, SaA; Bg, BglII; Kb, kilobasc

tially as described by Misset et al. [4]. Various preparations were used of which the specific activity ranged over 360 - 570 unit/mg. Purified enzyme preparations showed a single band after SDS/polyacrylamide gel electrophoresis.

Glucose-6-phosphate isomerase activity was measured spectrophotometrically by coupling the formation of glucose 6-phosphate from fructose 6-phosphate to the reduction of NADP with glucose 6-phosphate dehydrogenase [38]. The standard assay was performed at 25°C in a volume of 1 ml, containing 0.1 M triethanolamine (pH 7.6), 1.3 mM fructose 6-phosphate, 0.4 mM NADP, 7 mM MgClz and 1 p1 glucose- 6-phosphate dehydrogenase ( 5 mgiml). The reaction was started by adding 10 - 50 pl diluted glucose-6-phosphate isomerase solution to the assay mix.

Rabbit muscle and yeast glucose-6-phosphate isomerase were obtained from Boehringer Mannheim (FRG). They had specific activities of 565 unit/mg and 690 unit/mg. The enzyme from Bacillus stearothermophilus was purchased from Sigma Chemical Company (USA) and had a specific activity of 350 unit/mg. All enzymes were free of ammonium sulphate before assay. Enzyme units are defined as 1 pmol substrate con- verted/min.

Fructose 6-phosphate (disodium salt), NADP (disodium salt) and glucose-6-phosphate dehydrogenase and 6-phospho- gluconate were purchased from Boehringer Mannheim (FRG). Agaricic acid (n-hexadecylcitric acid) was from Fluka (Switzerland). Suramin and melarsen oxide were gifts from Bayer (FRG) and May and Baker Ltd (UK), respectively.

Michaelis-Menten constants were determined from direct plots of initial rates as a function of substrate concentration. Experimental points were fitted to a rectangular hyperbola with the aid of a computer program.

Variation of kinetic constants with ionic strength were determined in a 20 mM triethanolamine buffer at pH 7.6. Ionic strength was adjusted with a concentrated NaCl solution by comparing the conductivity of the solution with that of a calibration curve of known NaCl concentrations.

Peptides from the native enzyme were obtained by partial hydrolysis with 2% formic acid at 106°C during 4 h. The peptides were separated by HPLC on a 4.5 n m x 100 nm pBondapak/C, column (Waters, USA). Elution was per- formed with a gradient of the following solvents: the initial solvent A (0.1 % trifluoroacetic acid in water) was gradually replaced by solvents B (0.1 YO trifluoroacetic acid in acetonitrile) and C (0.3 % trifluoroacetic acid in isopropanol) as follows : 0-40% B in 70 min, 40-60% B and 0-20% C

in 30 min, 20-40% C with B constant at 60% in 10 min, and 60-100% B simultaneous with 40-0% C in 10 min. The acetonitrile and the isopropanol were HPLC grade from Rathburn (UK).

The sequence of two of the peptides was determined on a 477A pulsed-liquid sequencer (Applied Biosystems, Inc., USA) with on-line analysis of the amino acids on a 120A phenylthiohydantoin analyzer.

Protein concentrations were determined by the fluorescamine method [39] with bovine serum albumin as standard.

RESULTS

Isolation and characterization of the glucose-6-phosphate isomerase gene of T. brucei

A genomic library of trypanosomal DNA in the vector 1. EMBL4 (10 x genome size) was screened with a cloned yeast gene as probe [26], at reduced stringency (3 x NaCl/Cit, 60°C). In this manner, ten identical clones were obtained with an insert size of 11 kb. After subcloning and sequence analysis those clones could be unambiguously identified as coding for glucose-6-phosphate isomerase : the predicted amino acid sequence showed a high level of similarity with conserved regions of glucose-6-phosphate isomerase of other organisms (see below). Moreover, 100% identity was found with two peptide fragments from the purified protein, each fifteen amino acids long (residues 214- 229 and 598 - 613 in Fig. 4). An internal and a C-terminal peptide were sequenced, because the N-terminus of the protein appeared to be chemically blocked.

A physical map of the genomic clone was constructed by hybridization analysis of Southern blots, made with DNA digested by various restriction enzymes, using the yeast gene as probe (Fig. 1). Our restriction analysis of the clones and of uncloned DNA has indicated that T. brucei contains only one glucose-6-phosphate isomerase locus with a sequence of appreciable similarity to the yeast gene.

When chromosome-sized molecules were separated by pulse-field gradient electrophoresis, followed by blotting and hybridization analysis with a glucose-6-phosphate isomerase clone, we observed equal hybridization with two different size classes: with material that remained trapped in the slot, whatever the pulse time, and with resolved molecules in the megabase pair area (Fig. 2) . In contrast, hybridization with

458

Fig. 2. Chromosomal location of'the glucose-6-phosphate isomerase gene in T. brucei. The chromosomes of two different variants (1 18a and 1 .X b) were separated by pulsed-field gradient electrophoresis, using the system described by Johnson and Borst [31]. Electrophoresis was performed in 0.5% agarose gels, with Tris/borate/EDTA, at 14"C, applying a voltage of 190 V and a pulse time of 260 s, during 25 h. Thc gel was stained with ethidium bromide, photographed by ultraviolet transillumination (A) and then blotted. The blot was hybridized twice; first with the gene for glucosc-6-phosphate isomerase (PGI; B) and subsequently, after removal of the probe, with the gene for ~-glyceraldehydc-3-phosphate dehydrogenase (GAPDH; C). After the hybridizations the blot was washed to 0.1 x NaCl/Cit at 65°C and autoradiographed

an aldolase probe was predominantly with DNA in the slot (not shown), whereas with a probe for ~-glyceraldehyde-3- phosphate dehydrogenase hybridization was observed with material present in the slot and with large chromosomes. The results for aldolase and ~-glyceraldehyde-3-phosphate dehydrogenase are in agreement with those previously ob- served [8, 40, 411. Essentially similar results were obtained when electrophoresis was performed in a homogeneous elec- trical field using the Pulsaphor system (Pharmacia LKB, Sweden; not shown).

Sequence analysis of'the glucose-6-phosphate isomerase gene

To sequence the glucose-6-phosphate isomerase gene, two restriction fragments of the genomic clone were subcloned in a phagemid vector (pTZ18R). The recombinant plasmids were subjected to unidirectional digestion with exonuclease 111, thus generating a set of progressively shortened molecules which formed the basis of our sequencing strategy. This is indicated in Fig. 1, beneath the physical map. The nucleotide sequence and the predicted amino acid sequence of the trypanosomal glucose-6-phosphate isomerase are presented in Fig. 3. The sequence analysis revealed an open reading frame coding for a polypeptide of 606 amino acids, with a calculated M , of 67280 (not counting the initiator methion- ine). This is somewhat larger than the value of 64 kDa, deter- mined by SDSjPAGE for the purified enzyme (see below). The optimal alignment between the amino acid sequence of the T. brucei glucose-6-phosphate isomerase and the sequence of the corresponding protein of other organisms, as well as the mouse and human neuroleukin sequences is shown in Fig. 4. The most striking features of the T. brucei enzyme, as revealed by this comparison with the yeast and mammalian proteins, are the presence of 38 - 49 additional residues at the N-terminus and a five-amino-acid insertion rich in arginine. The amino acid sequence of the T. brucei enzyme has 54- 58% similarity with its mammalian and yeast counterparts, as shown in Table 1. The charge of the different glucose-

6-phosphate isomerases, as calculated from the amino acid composition, assuming neutral pH, is presented in Table 2. The calculated net charge for the trypanosomal protein is -2, compatible with the measured PI of 7.5 [4]. This is clearly different from the other four glycosomal proteins of which the sequence has been determined and which all have in com- mon an excess of basic residues and consequently a high PI [4]. However, clusters of positive residues, specific for the T. brucei amino acid sequence, can be recognized. They are indicated by asterisks in the sequence of Fig. 4. These clusters may be analogous to the positively charged hot spots observed in four other glycosomal proteins [9].

A determination of the hydrophobic character of the glu- cose-6-phosphate isomerase according to Kyte and Doolittle [42] indicates that the trypanosomal enzyme is less hydrophilic than its counterparts from other organisms (Table 3) which may be compatible with its glycosomal localization. It is, however, significantly more hydrophilic than the other glycosomal proteins of which the sequence is known, with the exception of aldolase (Table 3).

Properties of the enzyme

To identify differences between the trypanosome enzyme and those from other organisms properties of purified glu- cose-6-phosphate isomerase from T. brucei were compared with those of similar commercially available forms from rab- bit muscle, yeast and B. stearothermophilus.

Fig. 5 shows the relative mobilities of the T. brucei, the yeast and the rabbit muscle enzyme on a SDS/polyacrylamide gel. In agreement with the sequence data presented in Fig. 4, the T. brucei enzyme (64 kDa) is 5 kDa larger than the rabbit muscle and yeast enzymes (59 kDa). The pH activity profiles for glucose-6-phosphate isomerase from T. brucei, rabbit muscle and yeast are shown in Fig. 6. All three enzymes displayed a similar but broad pH optimum with a maximum between pH 7 and 9.5. A semilogarithmic plot (not shown) of activity against pH revealed that the pH activity curve for the

459

I61 f w G v l T M T T A C T m G A ~ w ~ l l C C G G M C G Met Str Ser Tyr Leu kp k p

ATG AGC AGT TAC TTG GAT GAT 233

Leu Ar$ I l c Asp Leu A l a A l a Ser Pro A l a Scr Glr 611 Ser A l a br l l e A l a Val Gly TTG CGC ATC GAT CTG GCG GCA AGT CCT GCG TCC GGG GGT 113 GCA AGT ATT GCC GTC GGT 293

br Phe Asn IIc Pro Trr G l u Val Thr Ar Ar Leu Ly5 G l y Val G l y A l a Asp A l a Asp TCA TTC PAT ATA CCT TAT GAG GTC A M CG! AG! CTA PAt GGG GTG GGA GCI) GAT GCC GAC 353

Thr Thr Lru Thr L r C r 5 A l a Ser Trp Thr G l n Leu G l n Lys Leu Tyr Glu G l n Trr G l y ACA ACA CTT ACT TCT TGC GCT TCT TGG ACT CAG TTA CAG PAG CTG TAT 646 cp.A TAT GGC 413

Asp G l u Pro IIe LYS L r s H i s Phc Glu A l a Asp Scr Glu Ar Gly Gln Ar Tyr Ser Val WC Gcrc CCT A l C M G MG CAT TTC GM GCT GAC AGT GAG AG! GGT cp.A Ad TAT TCT GTA 473

LYS Val S i r LPU G I Y Ser LYS As Glu A5n Phc Leu Phe Leu Asp T r r 5cr L r s S i r His MG GTG AGT TTG GGG AGC M GA! GM MC T l l TTA Xl TTG GAT TAC TCC PAG TCG M C 533

11, Asn Asp G l u I I e Lys C y s A l a Leu Leu Ar Leu A l a G l u Glu Ar Gly I l e Ar G l n ATT AAC GAT GM AT4 MG TGC GCT CTT CTG AG8 CTG GCC 646 GM CG! GGC ATC CG? CAG

h e Va l G l n Ser Val Phe Arq 6 1 y Glu 4,- Val Asn Thr Thr Glu Asn Arg Pro Va l LPU

H i s I l c A l a Leu Arq Asn Ar Ser Asn Arg Pro 11e Tyr Va l Asp Gly Ly5 Asp Va l Met CAC ATT GCT CTG CGA PAT CG? TCC MT CGA CCT AT1 TAT GTT GAC GGA MG GAT GTA ATG

Pro A l a Val Asn LYS Val Leu Asp Gln Hct Arq Scr Phc Br Glu L r s Va l Arg Thr Glr CCG GCT GTA M C MG GTT C l T GAC w14 ATG AGA AGC TTT TCC GAG P A G GTG AGG ACT GGG

G l u Trp L r s GIY H i s Thr G l y LYS A l a l i e Arq H i s Val Val Asn l l e G l y I I e Glr Glr GAG TGG MG GGC CAT ACC GGC MG GCC ATC C W CAC GTT GTA M C ATT GGC ATC GGA GGC

593

TTT GTG WI TCC GTT TTC FGT GAG AG! GTA mc ACG ACA GAG MC CGC CCA GTC m 653

713

773

833 Scr Asp Leu G I Y Pro Va l Net A l a Thr Glu A l a Leu LYS Pro Phe Scr Gln Arq Asp Leu AGT GAC TTG GGA CCC GTT ATG GLA ACG GAG GCA CTG P A G CCG TTT TCT CAG AGG GAT CTT 893

Scr Leu H i s Phe Val Scr Asn Val Asp Glr Thr H i s Ilc Ala G l u Val Leu Ly5 Scr I I P

As I I e Glu A l a Thr Leu Phe I l e Val A l a Scr LYS Thr Phe Thr Thr G l n GIu Thr I I e GA! ATT GPA GCA ACT CTT TTC ATC GTC GCG AGC &% ACG lll ACG ACC CAII GM ACG ATC 1013 Thr Asn A l a Lru Scr A l a Ar Ar A l a Leu Lcu Asp Tyr Leu Ar Scr Ar 611 I I c Acp PCA PAC GCC CTC TCC GCC AG! CG! GCG TTG TTG GAC TAC CTG CG8 TCG CG8 GGC ATC GAT 1073 Glu LYS G l r Scr Val A l a LYS H i s Phc Val A l a Leu Ser Thr Asn Asn Gln Lys Val Lys GAG MG GGT TCC GTT GCA MG CAC TTC GTT GCA CTG TCG ACA M T PAT CAG M GTC MG 1133 G l u Phc GIY Ile As G l u Glu Asn Met Phe Gln Phc lrp Asp T r p Val G l y Gly Ar Tyr

1193 S i r Net T r p Scr A l a 11s G l y Leu Pro 11, Het IIr B r l l e G l y l y r Glu Asn Phc Va l TCT ATG TGG TCA GCT ATT GGC TTG CCC ATT ATG A T 1 TCA ATC GGA TAC GAG MC TTT GTG 1253 G l u Leu Leu Thr G l v A l a H i s Val Ile Asp Glu H i s Phc A l a Asn A l a P r o Pro G l u Gln GM CTC TTA ACT GGG GCC CAT GTC ATC GAT GW CAC TTT GCC M T GCA CCC CCT GPA CAG 1313 Asn Val Pro Leu LPU Leu Ala Leu Val G l y Val Trp Tyr I l r Asn Phc Phe G l y A l a Va l MT GTT CCT CTG CTT CTT C U TTG GTG GGG GTT TGG TAC ATC MT TTC TTC GGC GCT GTG 1373 Thr H i s A l a 11s Lru Pro Tyr Asp G l n Tyr Leu Trp Ar Leu Pro A l a T r r Leu G l n G l n ACC CAT GCC A l l CTG CCG TAT W C cp.A TAC TTG TGG AG8 TTA CCG GCG TAC CTT CAG C M 1433 Leu Asp h t G l u Str Asn Gly Lrs Tyr Val Thr Arq S i r G l y L r s Thr Va l Ser Thr Leu CTT GAT ATG GV4 AGT MT GGT W TAC GTT ACG AGG AGC GGT M ACA GTT TCT ACC CTT I493 Thr Glr Pro I I e I l e Phe G l y Glu A l a Glr Thr Asn G l y G l n His A I a Phr Trr G l n Leu ACT GGC CCC A T 1 ATT TTl GGG GPA GCT GGG ACG M C GGC CPA CAT GCA TTT TAT CAG CTT 1553 IIr H i s Gln Gly Thr Asn Lru 11c Pro C r s Asp Phe I I e G l y A l a IIe G l n Ser G l n Asn ATC CAT 0% GGG ACT MT CTC AT1 CCG TGC GAT TTC ATC GGT GCG ATT CAG AGT W MT 1613 LYS I I e Glr Asp H i s H is LYS IIr Phi h t Srr Asn Phe Phe A l a G l n l h r G l u A l a Leu M ATC (jTiT GAC CAC CAC MG A T 1 TTT ATG AGC PAC TTC TTC GCC CM ACT GM GCG CTG 1673 k t I l e Gly L r s S r r Pro S i r G l u Val Ar Ar G l u Leu G l u A I a A l a Glr Glu k Str ATG AT4 GGC MG TCT CCC lKiT GAG GTT C G ! CG! Gcrc TTG Gcuj GI3 GCG GW W M8 TCT I733 A l a Glu LYS l l e Asn A l a Leu Lru RD H i s L r s Thr Phc IIe G l y Gly 4- Pro S i r Asn GCT GPA M AT4 M T GCA CTT CTC CCT UIC W ACT TTT ATT GM GGC Cd CCA AGC MC 1793 Thr Leu Leu IIe Lys Ser Leu l h r Pro Arq A l a Leu G l y A l a I I e IIe A l a h t l r r G l u ACT TTG CTG A l l pplc TCC CTA AC6 CCG CGG GCT TTG G N GCA ATC A l C GCT ATG TAC GPA I rnq

TCC CTG CAC m GTT TCC mc GTC GAT GGA ACT CAC ATC GCC GAG GTT CTA M TCG ATT 953

GAG TTC GGT aic GA! GAG w PAC ATG m CAG TTT TGG wc TGG m GGT GGT c d T A c

Table 1. Amino acid similarity of glucose-6-phosphate isomerases fiom various organisms The percentage of similarity was calculated from the number of iden- tical residues between the aligned sequences in Fig. 4; insertions were not counted. NLK, neurolcukin

Source of Glc6P isomerase T. hrucei Yeast Mouse Pig Human

Similarity to Glc6P isomerase from

( N W "LK)

Yo

T. hrucei - Yeast 54 Mouse (NLK) 57 59 - Pig 58 59 89 -

Human (NLK) 57 58 89 93

-

-

Table 2. Predicted charge of the glucose-6-phosphate isomerasesfrom various organisms The calculations are based on the sequences given in Fig. 4. The charges of Asp and Glu are taken as - 1 ; Lys and Arg as + 1 ; all other amino acids as 0. NLK, neuroleukin

Source of Positive charge Negative charge Net Glc6 P isomerase Lys Arg total Asp Glu total

charge ______

number/molecule

T. brucei 31 32 63 29 36 65 -2 Ycast 40 10 50 25 35 60 -10 Mouse(NLK) 42 20 61 25 35 60 + 1 Pig 38 24 62 25 36 61 + l Huinan(NLK) 35 27 62 27 32 59 f 3

Table 3. Average hydropathy of the glucose-6-phosphate isomeruses from diiferent orxanisms and the glycosomal enzymes of T. brucei The hydropathy values were calculated for the various polypeptides using the amino-acid hydropathy indices proposed by Kyte and Doolittle 1421. Negative values correspond with a hydrophilic charac- ter. NLK, neuroleukin

Organism Enzyme Average hydropathy

i iyi-Lrs Val Leu Val Gln G I Y A l a IIe l r p M Y I I e Asp b r Tyr Asp G l n l r p Glr Val CAC MG GTT CTC GTT CAG GGT GCC ATA TGG GGT AT1 MC AGT TAT GAT CAG TGG GGT GTT IPI 3 GIi-Leu G l r L r s Val Leu & l a LYS b r I I e Leu Pro Gln Leu Arq Pro Glr h t k q Val GAG CTG GGG PAG GTG CTA LXC PAG TCT ATT CTT CCT CAG TTG AGG CCA GGT ATG MI3 GTG 1973 Asn Asn His Asp b r Srr Thr Asn G I Y Leu I I e Asn h t Phe Asn G l u Leu Slr His Leu PAT PAT CAC GAC TCA TCC ACC PAC GGT CTA ATC PAT A l G TTT MT W TTG TCG CAC CTT 2033 001 ~ C & X L W I C I C A T ~ ? A T G ~ T G A C T M T i 3

Fig. 3. The complete nucleotide sequence and predicted amino acid - . - . sequence of the glucose-6-phosphate isomerase gene in T. brucei

Yeast Glu6P + glucose-6-phosphate

Mouse (NLK) Glu6P + glucose-6-phosphate

Pig Glu6P + glucose-6-phosphate

Human (NLK) Glu6P + glucose-6-phosphate

T. hrucei Glu6P --t glucose-6-phosphate

isomerase

isomerase

isomerase

isomerase

isomerase aldolase Trip + triose phosphate isomerase GraP --f glyceraldehyde-phosphate deh ydrogenase GriP --+ phosphoglycerate kinase

- 0.229

-0.295

-0.341

-0.337

-0.180 -0.301 +0.058

-0.132 -0.109

460

1 2

3

4

5

1

2

3

4

5

1

2

3

4

5

1

2

3

4

5

1 2

3

4 5

1

2 3 4

5

1 2

3

4

5

20 30 40 50 60 70 80 90 100

TLSV.QE.QK.AK.FEK LN.TFT NY.G

uw* 10

MSSYLDDLRIDLAASPASGGSASIAYGSFNIPYEVTRRLKGVGADADTTLTSCASWTQLQKLYEqYGD EPIKKHFEADSERGQRYSVKVSLGSKOE MSNNSFTNFKLA.ELPA.SK ... I . .SQ.K

MAALTRNP.F ... L.WHRANSANLKLREL .... P..FN NFSLNLNTNH

MAHLTQNP.FK..QTW.HEHRSDLNLRRL..G.KD.FN HFSLNLNTNH

MAALTRDP.F ... QQW.REHRSELNLRRL.D.NKD.FN HFSLTLNTNH

110 120 130 140 150 160 170 180 190 200 NFLFLDYSKSHINDEIKCALLRLAEERGIRQFVQSVFRGERVNTTENRPVLHIALRNRSNRPIYVDGKDVMPAVNKVLDQMRSFSEKVRTGEWKGHTGKA S K I L F .... NLV .... I A .. I E .. K.ANVTGLRDAM.K .. H1.S .. D.A.Y.V ..... A.K.M .... VN.A.E.08 .. KH.KE ... 4 .. S . . . . . Y . . . K

GHILV .... NLV.K.VMQM.VE .. KS .. VEAARDNM.S.SK1.V .. 0.A ... V ....... T..K... . . . . .E..R...K.K..CQR..S.D...Y... S

G R I L ..... NLVTEAVMQM.VD .. KS .. VEAARERM.N .. K 1 . F .. D.A . .. V ....... T..L... . . . . .E..R..EK.K..CKR..S... . .YS.. 8

G H I L V .... NLVTEDVMRM.VD .. KS .. VEAARERM.N .. K1.Y . . G.A . .. V ....... T..L... . . . . .E... . . .K.K..CQR..S.D...Y... T

210 220 2 30 240 250 260 270 280 290 300 I R H V V N I G I G G S O L G P V M A T E A L K P F S Q R D L S L H F V S N V O G ~ H I A E V L K ~ I O I ~ A T L F I V ~ S K T ~ T T Q ~ T I ~ ~ A L S A ~ R * L L * Y L ~ S ~ I D E K G S V A K H F

.TD ............... V.. . . . HYAG V.DV . .... I ....... T..VV.P.T...LI.......A......NT.KNW F.S

. T D I I ........... L.V ...... Y.KGGPRVW .... I......KT.A.LSP.TS...I..............ET.KEW F.E

. T D . I ........... L.V . . . . . . Y.AEGPRVW .... I......KT.ATLNP.SS...I..............ET.KEW F.Q

. T D . I .... V......L.V......Y.SGGPRVWY...l......KT.AQLNP.SS...l..............ET.KEW F.Q

KTGNDPSHI . . .. AAKDPSA ..... SAKDPSA ..... AAKDPSA .....

310 320 330 340 350 360 370 380 390 400 VALSTNNQKVKEFGIOEENMF~FWDWVGGRYSMWSAIGLPIMlSIGYENFVELLTGAHVlDEHFANAPPEQNVPLLLALVGVWYlNFFGAVTHAlLPYDQ A ..... ETE.AK . ... TK ... G.ES ....... V......SVALY...D..EAF.K..EAV.N..TQT.L.D.I...GG.LS...N.....Q..LVA.F..

PQ ... E .......... L......S.ALHV.FDH.EQ..S...WM.Q..LKT.L.K.A.V....L.I....CY.CE...L..... ...... L ...... S.ALHV.FD . . EQ . . S...WM.Q..RTT.L.K.A.V... .L. l . . . . . . .CE...M... . .

PQ.. .E.. ....... .L.. . . . .S.ALHV.FD. .EO.. 8 . . .WM.O. .RTT.L.K.A. V.. . . L . I.. . .C. .CE.. .M.. ...

410 420 430 440 450 460 4 70 480 490 500 YLWRLPAYLQQLDMESNGKYVTRSGKTVSTLTGPI lFGEAGTNGQHAFY~LIHQGTNLIPCDFIGAlQSQNKI GDHHKIFMSNFFAQTEALMIGKSPS

... S ... GNVFTDYS .. 8.L . .. PA . . A..S.F..V... .K...S...L.A..H.P.ENKL.Q.MLA.... . .A... .V.. DEE

.... 1.K .. AR.DHQ .... VW .. P ................ KM L1Pv.T.HP.RK.L ....LLA..L.......K.. L.E

.... 1.K .. TR.DHQ .... VW .. P ................ KM L1PV.T.HP.RK.L .. . .LLA..L... . . . .K... TE .. H.FA .. F .. G.... . . . . I .K..TR.DHQ....VW..P... . . . . . . . . . . . . .KM.... .LIPV.T.HP.RK.L... .LLA..L... . . . .R... T E

590 600 EVRRELEAAGERSAEKINALLPHKTFlGGRPSNTLLIKSLTPRALGAIlAMYEHKVLVQGAI~GIDSYDQWGVELGKVLAKSILPQLRPGMRVNNHDSST

Q.KA.GATG. .V ... V.8.N .. T T S I . A Q K I . . AT ... L..Y...VTFTE....N.N.F.............V.GKE.DNSSTIST..A..

.A.K..Q... K.P.DFEK ..... V.E.N .. T . S I V F T K ... F I ... L ....... I F ... V..D.N.F.........Q...K.E.E.DGSSP.TS.....

510 520 530 540 550 560 570 580

......... ..... .A.K..Q... K.P.DLEK ..... V.E.N .. T . S I V F T K . . . F I ... L ....... I F . . . I M . D . N . F Q...K.E.E.EGSSA.TS

.A.K..Q... K.P.DLER ..... V.E.N .. T . S I V F T K . . . FM ... LV ...... I F . . . I . . D . N . F ......... Q...K.E.E.DGSAQ.TS..A..

610 NGLINMFNELSHL

..... Q.K.WM

.... SFIKQQRDTKLE

..... FIKQEREARSQ

..... FIKQQREARVQ

1. T . Bruce i

2. Yeast

3. Mouse ( N L K ) 4 . P i g

5. Human ( N L K )

Fig. 4. A compurison of the T. brucei glucose-6phosphute isomerusr amino acid sequence with the sequence of the protein offi)ur other orgunisins. Amino acid sequences are aligned to obtain maximal homology. Dots represent amino acids that are identical to those on the corresponding positions in the T. hruceienzyme. Blank spaces indicate the absence of amino acids at corresponding positions. The amino acids in the T. brircei scquencc that may constitute the positively charged signal for import into the glycosome are indicated by asterisks. The amino acid sequences are: (1) T. brucei; (2) yeast 18; (3) mouse neuroleukin (NLK) [21]; pig muscle (391; ( 5 ) human neuroleukin [57]. Neuroleukin is a neurotrophic factor that has been shown to be identical to glucose-6-phosphate isomerase [I 9, 201

T. hrucei enzyme resulted from two ionizable groups with pK values of 6.8 and 30. These values are in good agreement with those reported for the enzyme from bovine mammary gland (6.82 and 9.83) [43] and those for yeast, (6.8 and 10.1) [44].

Affinity constants (K,,,) were determined with fructose 6- phosphate as the substrate for the enzymes from T. brucei, yeast, rabbit muscle and B. stearotlzrrmophilus. The results are shown in Table 4. Thc K, value of 0.122 mM for the T.

hructi enzyme is in good agreement with the value of 0.10 mM as determined by Hannon and Parr 1121, although this value was measured at pH 8 instead of 7.6. The values for the rabbit muscle and yeast enzymes, as measured here, are also in good agreement with published data 1141.

6-Phosphogluconate is a strong competitive inhibitor of glucose-6-phosphate isomerase 1141. Inhibition constants for this compound were determined as the inverse of the slope of

461

Fig. 5. Sodium dodecyl sulphate/polyacrylarnide gel electrophoresis of glucose-6-phosphate isomerases. Lanes 1 and 5 : molecular mass markers: phosphorylase b (94 kDa), bovine serum albumin (67 kDa) and ovalbumin (43 kDa). Lanes 2 - 4, glucose-6-phosphate isomerase from T. hrucei (T.b.), rabbit muscle (RM) and yeast (Y) , respectively

100

- 80 8

‘5 .- 60

- h c

c 0 m a,

m a, U

.5 40 c -

20

0

A Y A

0 A a a 0 a

n A 0

%Q a B I I 1 I

5 6 7 8 9 10 11

PH

Fig. 6. Activity of’glucose-6-phosphate isomerase. The relative activity of glucose-6-phosphate isomerase from T. hrucei (01, yeast (0) and rabbit muscle ( A ) is plotted as a function of the pH. Buffer concen- trations were 50 mM and ionic strength was adjusted to 0.1 M with NaCI. The following buffers were used: ( 2 ) (Mes, pH 5.1 -6.9); (2) (triethanolamine/HCl, pH 6.8-8.7); ( 3 ) (Hepes, pH 7.1 -8.2); (4) ( 3 - cyclohcxylamineethanesulphonic, pH 8.4- 10.4); (5) (3-cyclohexyl- aminopropane-3-sulphonic acid, pH 9.5 - 11.0)

Table 4. Kinetic cunstants for glucose-6-phosphate isomerase from vari- ous sources measured under standard conditions The number of determinations is given in brackets

Enzyme source Fru6P Gluconate6P K* Ki

mM

T. brucei 0.122 0.045 (17) 0.14 0.03 (6) Yeast 0.167 2 0.048 (8) 0.48 0.12 (4) Rabbit muscle 0.119 & 0.028 (10) 0.12 f 0.03 (4) B. stearo-

thermophilus 0.048 (1) 0.16 & 0.008 (2)

.4

.3 - 5:

E E .2 Y

.1

0

( 0 ) : T. brucei (0) : Yeast / (A) : Rabbit

I I I l

0 .1 .2 .3 .4 .5

Ionic strength (M)

Fig. 7. Dependence of the K,,, f o r fructose 6-phosphate ofglucose-6- phosphate isomeruse on ionic strength. Constants were measured in triethanolamine/HCI buffer a t pH 7.6 as described in Materials and Methods. Ionic strength was adjusted with NaCI. (0 ) T. brucei; (0) yeast; ( A ) rabbit muscle

K, (apparent)/K, versus inhibitor concentration plots and the results are summarized in Table 4. The values for the T. brucei and rabbit muscle enzyme correspond well with previously published data. The value for the yeast enzyme, however, is more than 30 times higher that that reported by Noltman [14]. Lack of information on the experimental conditions used by this author does not allow us to identify the reason for this discrepancy.

The maximum velocity of the trypanosomal enzyme was hardly influenced by the ionic strength of the assay mixture (not shown). The K,, however, showed a strong dependence on the salt concentration, an effect which was not unexpected since charged residues were reported to be involved in the enzyme reaction [15 - 17, 43, 45 -471. The variations of K , with ionic strength for the enzymes from T. brucei, yeast and rabbit muscle are shown in Fig. 7. Although, at high ionic strength, there is a significant difference between the yeast enzyme on the one hand and the rabbit muscle and T. hrucei enzymes on the other hand, such a difference does not exist at low ionic strength. Thus the three enzymes apparently have identical affinities for fructose 6-phosphate and differences in measured K, are mainly due to a different dependence on ionic strength, which may reflect sequence differences at some distance of the active site.

Inhibition of trypanosomal glucose-6-phosphate iso- merase by the trypanocidal drug suramin was of a mixed type. The inhibitor had a significant effect on the K, for fructose 6-phosphate while the effect on V,,, was only slight. The T. hrucei enzyme had a Ki for suramin of 0.29 mM, while the yeast enzyme, which also displayed similar mixed-type inhi- bition, had a Ki of 0.36 mM. Suramin had no effect on the rabbit muscle enzyme over the concentration range tested (0 - 150 pM).

Agaricic acid has been reported as an inhibitor of the NAD +-dependent glycerol-3-phosphate dehydrogenase of C. fasciculuta [4X] and has recently been found to be also a potent inhibitor of a number of other glycosomal enzymes of T. brucei (D. Kuntz & F. Opperdoes, unpublished results). This drug is a potent irreversible inhibitor of the trypanosome glucose-6-phosphate isomerase with 50% inhibition at 10 pM. The yeast and rabbit enzymes were hardly affected at the highest concentrations used (less than 20% inhibition at 60 pM, not shown).

Melarsen oxide, a trivalent arsenic compound and the active principle of the trypanocidal drug melarsoprol (MelB,

Arsobal) had only a slight inhibitory effect on the trypano- some glucose-6-phosphate isomerase (50% inhibition at 100 pM, not shown).

DlSCUSSION

We have cloned and characterized a gene that codes for glucose-6-phosphate isomerase in T. brucei. Mapping of re- striction enzyme sites indicated the presence of only one gen- etic locus for this enzyme in our laboratory strain (427). By searching for allelic restriction-enzyme-site polymorphisms in the DNA of a large number of different trypanosome stocks (Gibson et al. [40]) have previously shown that T. hrucei is diploid for some genes of glycolytic enzymes. A similar result has been obtained for the gene of glucose-6-phosphate iso- merase (W. Gibson, personal communication). Furthermore, using a glucose-6-phosphate isomerase probe on blots of chromosome-sized molecules that were fractionated in pulsed- field gradient gels, we consistently found an equal hybridiza- tion with two distinct DNA fractions : material that remained in the slot and chromosomes in the megabase pair area. In- terestingly, the length of the migrating chromosome that con- tains a gene for glucose-6-phosphate isomerase is different in the two antigenic variants examined, i.e. variants 118a and 1.8 b. It has been shown that chromosomes of variant 1.8 b have undergone considerable size changes during the antigenic switch [49, SO]. Since it is very unlikely that the same drastic shortening has occurred simultaneously in two chromosomes, one can exclude the possibility that the migrating band con- tains two similar chromosomes. Therefore, the hybridizing DNA in the slot and the hybridizing molecule in the megabase pair area presumably represent chromosomes that are hom- ologous, but have considerable differences in length or other physical characteristics.

Previous experimcnts from this laboratory (Misset, O., unpublished rcsults) have never provided evidence for the presence of more than one distinct isoenzyme for glucose-6- phosphate isomerase in T. hrucei with respect to native and subunit molecular mass and isoelectric point. Although some soluble activity is consistently found upon subcellular frac- tionation of T . brucei bloodstream forms, this can easily be explained by leakage from damaged glycosomes. Also an analysis by Tait [51], in which cell extracts prepared from 17 T. hrucei stocks were subjected to starch gel electrophoresis, did not reveal any heterogeneity in this protein. This may not be true for other representatives of the Trypanosomatidue. Hart and Opperdoes [52] have reported a large cytosolic con- tribution to the total glucose-6-phosphate isomerase activity for several species of Leishmania and a similar situation was found in the case of Crithidia luciliae [53]. A true cystolic isoenzyme should play a role in the hexose-monophosphate pathway which is functional in Leishmania and Crithidia sp. [52, 531 but which has not been demonstrated in T. brucei bloodstream forms. Miles and coworkers [13] have analyzed glucose-6-phosphate isomerase in Trypanosoma cruzi. They have also found various, stable isoenzymes which were iden- tical in size, but different in pl. It was hypothesized that those different isoenzymes, for which the precise function and localization is not known, are either due to the presence of two slightly different genes, or are the result of posttranslational modification. The latter possibility has been invoked to ex- plain the different isoenzymes in rabbits; in that case, however, the different forms are presumably artificial, because they could be interconverted by altering the oxidative condition of

the sample [54]. Whether or not there exists in other represen- tatives of the Trypanosoinatidac more than one genetic locus for glucose-6-phosphate isomerase, remains to be investi- gated.

Previously we have argued that in trypanosomes a corre- lation may exist between the amount of a protein in the cell and the number of genes, usually arranged as tandem repeats, coding for that protein [6,8]. In this respect it is not surprising that for glucose-6-phosphate isomerase only one gene/haploid genome is found because the protein is not very abundant (less than 1% of the total) [4].

The comparison of the purified T . brucei enzyme with those of other organisms shows that, despite its different localization within the cell and its 39-amino-acid, N-terminal extension, the enzyme has remarkably similar kinetic proper- ties. Affinities for the substrate fructose 6-phosphate and in- hibitor gluconate 6-phosphate, as well as specific activity, are not significantly different from those of other glucose-6- phosphate isomerases. Our kinetic data also provide indirect evidence for the involvement of histidine (pK, 5.5- 7.0) and lysine ( p K , 9.5-10.5) residues in the catalytic site, as was previously inferred by others for the mammalian and yeast enzyme [IS-17, 43, 44, 551. More direct evidence has been obtained for the mammalian enzyme by affinity labelling. Using pyridoxal 5’-phosphate, two lysines could be specifi- cally labelled in the active site of the pig muscle enzyme [15, 161. After labelling and partial cleavage of the enzyme the labelled peptide fragments were purified and sequenced, to obtain the primary sequence of the active site [15, 161. Comparing those data with the collected, complete sequences presented in this article, allowed us to identify the two labelled lysine rcsidues as KZG3 and K5,, (numbering of Fig. 4). Other labelling studies using the yeast [56], human placenta and rabbit muscle enzymes [17] suggested the presence of a glutamate and a histidine residue in the active site. However, we have been unable to locate the peptides containing these residues in any of the primary structures given in Fig. 4. Cysteines are apparently not important for catalysis, because in the T. hrucei enzyme, as for its pig homologue [15], two out of three can be modified withp-mercuribenzoate, without loss of enzyme activity (data not shown). The third cysteine was inaccessible at conditions that did not lead to precipitation of the enzyme. The limited importance of cysteines is further indicated by their absence in the yeast enzyme [18].

The trypanosome enzyme seems to be more sensitive to the action of thc inhibitors suramin and agaricic acid, compounds that are almost without effect on homologous enzymes from other sourccs. Inhibition by agaricic acid is irreversible (not shown) and this may be due to the detergent-like nature of this compound. The mixed-type inhibition of the trypanosome enzyme by suramin suggest a simultaneous interaction of the drug, both at the active site and elsewhere on the protein’s surface. A Ki of 290 pM for the trypanosome enzyme is relatively high as compared to those measured for several other glycosomal enzymes and is well above the physiologi- cally obtainable concentration of the drug. Inhibition of glucose-6-phosphate isomerase is therefore, most probably, not the primary mode of action of suramin on trypanosomes.

The overall degree of similarity between the T. hrucei and the other glucose-6-phosphate isomerases varies between 54% and 58%. A comparable value has been found previously for four other glycolytic proteins, indicating the distant relation- ship between the Trypanosomatidue and all other organisms studied [5 - 81. The residues of the glucose-6-phosphate isomerase that are the most conserved are most likely those

463

involved in the catalytic process and the subunit contacts. Amongst the unique sequence elements in the trypanosomal enzyme one should find those responsible for its entry into the glycosome and its functioning within the organelle. In this respect it may be significant that, despite the neutral PI, specific clusters of positively charged residues can be re- cognized within the T. hrucei glucose-6-phosphate isomerase sequence, as in the sequences of other glycosomal proteins, which do have a high PI [4,9]. However, the configuration of those positive oligopeptides and their localization within the three-dimensional structure remain to be determined. An- other prominent feature of the T. hrucei enzyme is the long N-terminal extension of which the function is not yet clear. Most glycosomal proteins have a higher subunit molecular mass than their cytosolic counterparts 141, but glucose-6-phos- phate isomerase is, so far, the only enzyme in which an N- terminal extension is responsible. In the other glycosomal proteins insertions and/or a C-terminal extension contribute to the higher molecular mass [5, 81.

Glycosoinal proteins have physicochemical properties that are different from those of glycolytic enzymes of other sources. They tend to aggregate under conditions of low ionic strength. Disociation is promoted by the addition of salt 131. However, the glycosomal glucose-&phosphate isomerase is different be- cause it is easily solubilized upon rupture of the glycosomal membrane. This ease to be liberated may be due to its more neutral and/or hydrophilic character.

We are grateful to Prof. F. Zimmerman and Dr. A. Aguilera (Technische Hachschule, Darmstadt, FRG) for the gift of the glucose- 6-phosphate isomerase clone from yeast; to Dr. P. Olson (Chiron Corporation, USA) for making the sequence of the yeast gene avail- able to us prior to publication; to Dr. Johan De Jonckheere (Institute of Hygiene and Epidemiology, Brussels) and Mr. Joost Zomerdijk (The Netherlands Cancer Institutc, Amsterdam) for providing us with thc fxilitics to perform the pulsed-field gradient electrophoresis experiments; to Dr. Wendy Gibson (University of Bristol, UK) for sharing her unpublished rcsults with us; to Dr. Douglas Kuntz for critical reading of the manuscript; to Mr. Joris Van Roy for the supply of trypanosomes and glycosomes and to Ms Franqoise Van de Calseyde-Mylle for assistance in the preparation of the manuscript. This investigation receivcd financial support from the UNDP/World Bank/WHO Special Programme for Research and Training in Trop- ical Diseases to P.A.M.M. and to F.R.O. and from the Belgian National Fund for Joint Basic Research and the Belgian National Incentive Programme on Fundamental Research on Life Science, contract Bioi22 to J.V.B.

REFERENCES 1. Opperdoes, F. R. & Borst, P. (1977) FEBS Lett. 80, 360-364. 2. Opperdoes, F. R. (1987) Annu. Rev. Microhiol. 41, 127- 151. 3. Misset, 0. & Opperdoes, F. R. (1984) Eur. J . Biochem. 144,475-

483. 4. Misset, O., Bos, 0. J. M. & Opperdoes, F. R. (1986) Eur. J .

Biochem. 157,441 -453. 5. Osinga, K . A., Swinkels, B. W., Gibson, W. C., Borst, P.,

Veeneman. G. H., Van Boom, J. H., Michels, P. A. M. & Opperdocs, F. R. (1985) EMBO J. 4, 3811-3817.

6. Michels, P. A. M., Poliszczak, A,, Osinga, K. A,, Misset, O., Van Beeumen, J., Wierenga, R. K., Borst, P. & Opperdoes, F. R.

7. Swinkels, B. W., Gibson, W. C., Osinga, K. A., Kramer, R., Veeneman: G. H., Van Boom, J. H. & Borst, P. (1986) EMBO

8. Marchand, M., Poliszcxak, A., Gibson, W. C., Wierenga, R. K., Opperdocs, F. R. & Michels, P. A. M. (1988) Mol. Biochrrn. Parasitol. 29, 65 - 76.

(1986) EMBO J. 5, 1049-1056.

J . 5 . 1291 -1298.

9. Wierenga, R . K., Swinkels, B., Michels, P. A. M., Osinga, K., Missct, O., Van Beeumen. J., Gibson, W. C., Postma, J. P. M., Borst, P., Opperdoes, F. R. & Hol, W. G. J. (1987) EMBO J . 6, 215-222.

10. Hart, D. T., Baudhuin, P., Opperdoes, F. R. & de Duve, C. (1987)

11. Swinkels, B. W.,Evers,R. & B0rst.P. (1988) E M B O J . 7, 1159-

12. Hannon, R. H. & Parr, C. W. (1978) Comp. Biochem. Physiol.

13. Widmer, R., Dvorak, J . A. & Miles, M. A. (1986) Biocheni. Biophys. Acta 873, 119 - 126.

14. Noltman, E. A. (3972) in The enzymes (Boyer, P. D., ed.) vol. 6, pp. 272 - 354, Academic Press, New York.

15. Achari, A,, Marshall, S. E., Muirhead, H., Palmieri, R. H. & Nollman, E. A. (1981) Phil. Trans. R . Soc. Lond. B293, 145- 157.

16. Palmieri, R. H., Gee, D. M. & Nokman, E. A. (1982) J . Biol. Chem. 257,7965-7968.

17. Gibson, D. R., Gracy, R. W. & Hartman, F. C. (1980) J . B id . Chem. 255,9369 - 9374.

18. Tekamp-Olson, P., Najarian, R. & Burke, R. L. (1988) Gene 73,

19. Chaput, M., Claes, V., Portctelle, D., Cludts, I., Cravador, A., Burny, A,, Gras, H. &Tartar, A. (1988) Nature 332,454-455.

20. Faik, P., Walker, J. I . H., Redmill, A. A. M. & Morgan, M. J. (1988) Nature 332, 455-457.

21. Gurney, M. E., Heinrich, S. P., Lee, M. R., Yin, H. S. (1986)

22. Shaw, P. J . & Muirhead, H. (1977) 1. Mol. Bid . 109,475-485. 23. Opperdocs, F. R., Aarscn, P. N., Van der Meer, R. & Borst, P.

(1976) Exp. Parusitol. 40, 198-206. 24. Bernards, A., Van der Ploeg, L. H. T., Frasch, A. C. C., Borst,

P., Boothroyd, J. C., Coleman, S. & Cross, G. A. M. (1981) Cell 27, 497 - 505.

25. Frischauf, A., Lehrach, H., Poustka, A. & Murray, N. (1983) J . Mol. Bid . 170, 827 - 842.

26. Aguilera, A. & Zimmerman, F. K . (1986) Mol. Gen. Genet. 202,

27. Maniatis, T., Fritsch, E. F. & Sambrook, J. (1982) Molecular cloning: a laboratory manual, p. 545, Cold Spring Harbor Lab- oratory Press, New York.

28. Yanisch-Perron, C., Viera, J. & Messing, J. (1985) Gene 33, 103 - 119.

29. Hanahan, D. (1983) J . Mol. Biol. 166, 557-580. 30. Wahl, G . M., Stern, M. & Stark, G. R. (1979) Proc. Nut1 Acad.

31. Johnson, P. J . & Borst, P. (1986) Gene 43, 213-220. 32. Van dcr Ploeg, L. H. T., Schwartz, D. C., Cantor, C. R. & Borst,

33. Rigby, P. W. J., Dieckniann, M., Rhodes, C. & Berg, P. (1977)

34. Smith, H. 0. (1980) Methods Enzyrnol. 65, 371 -380. 35. Sanger, F., Nicklen, S. & Coulson, A. R. (1977) Proc. Nut1 Acad.

Sci. USA 74, 5463 - 5467. 36. Sanger. F., Coulson, A. R., Barrcll, B. C. , Smith, A. J. H. & Roe,

B. A. (1980)J. Mol. Biol. 143, 161-178. 37. Messing, .I. (1983) Methods Enzymol. 101, 20-78. 38. Gracy, R. W. & Tilley, B. E. (1973) Methods Enzymol. 41, 392-

39. Stcin, S., Bohlen, P., Stone, J., Dairman, W. & Udcnfriend, S.

40. Gibson, W. C., Osinga, K. A., Michels, P. A. M. & Borst, P.

41. Gibson, W. C. & Borst, P. (1986) Mol. Biochem. Parasitol. 18,

42. Kytc, J . & Doolittle, R. I;. (1982) J . Mol. Biol. 157, 105-132. 43. Hines, M. C. & Wolfc, R. G. (1963) Biochemistry 2, 770-775. 44. Schray, K. J . & Howell, E. E. (1979) Arch. Biochem. Biophys. 192,

45. Lu, H. S. & Gracy, R. W. (1981) Arch. Biochern. Biophys. 212,

E M B O J . 6 , 1403-1411.

1165.

GOB, 177-181.

153- 161.

Science 234, 566 - 574.

83 - 89.

Sci. USA 76, 3683 - 3687.

P. (1984) Cell37, 77-84.

J . Mol. Biol. 113, 237-251.

400.

(1973) Arch. Biochem. Biophys. 155, 203-212.

(1985) Mol. Biochern. Parasitol. 16, 231 -242.

127-140.

241 - 249.

347 - 359.

464

46. Lu, H. S . , Talent, J . M. & Gracy, R. W. (1981) J . Bid. C h ~ m .

47. Pullan, L. M., Igarashi, P. & Noltmdn, E. A. (1983) Arch. H i m

48. Bacchi, C. J., Ciaccio, E. I. & Koren, L. E. (1969) J . Bacterial.

49. Michels, P. A. M., Van der Ploeg, L. H. T., Liu, A. Y. C. & Borst,

50. Van der Ploeg, L. H. T., Cornelissen, A. W. C. A,, Michels, P. A.

51. Tait, A. (1980) Nature 287, 536 - 538.

256,785 - 793.

chenz. Bi0Phy.r. 221, 489-498.

98, 23 - 28.

P. (1984) EMBO J . 3, 1345-1351.

M. & Borst, P. (1984) Cell39, 213-221.

52. Hart, D. T. & Opperdoes, F. R . (1984) Mol. Biochem. Purasitol.

53. Oppcrdoes, F. R. (1981) Pursitoloxy 82, 10- 12. 54. Blackburn, M. N., Chirgwin, J. M., James, G. T., Kempe, T. D.,

Parsons, T. F., Register, A. M., Schnackerz, K. D. & Noltmdn,

55. Dyson, J. E. D. & Noltman, E. A. (1968) J . Biol. Chem. 243,

56. O’Connell, E. L. & Rose, I. A. (1973) J . Biol. Chem. 248, 2225-

57. Gurney, M. E. (1987) Genbank, release 57, accession no. K03515.

13, 159-172.

E. A. (1 972) J . B i ~ l . Chem. 247, 1 170 - 11 79.

1401 - 1414.

2231.