Molecular and Biochemical Parasitology 91 (1998) 195–205

Trypanosome glucose transporters

Michael P. Barrett b, Emmanuel Tetaud c, Andreas Seyfang d, Frederic Bringaud a,Theo Baltz a,*

a Laboratoire de Parasitologie Moleculaire, UPRESA CNRS 5016, Uni6ersite de Bordeaux II, 146 Rue Leo Saignat,33076 Bordeaux Cedex, France

b Di6ision of Infection and Immunity, Institute of Biomedical and Life Sciences, Joseph Black Building, Uni6ersity of Glasgow,Glasgow G12 8QQ, UK

c Department of Biochemistry, Medical Sciences Institute, Uni6ersity of Dundee, Dundee DD1 4HN, UKd Oregon Health Sciences Uni6ersity, Mail code L-220, Portland, OR 97201, USA

Keywords: Trypanosome; Glucose transporter; Glycosylation

1. Introduction

The genus Trypanosoma comprises a group ofparasitic protozoa which cause widespread diseasein man and animals. African trypansomes aretransmitted by tsetse flies, or other biting insects,or by coitus in the case of the equine parasite T.equiperdum. The South American trypanosome,infectious to humans, T. cruzi is transmitted fe-cally by reduviid bugs. The different trypanosomespecies have different life cycles. African try-panosomes live free in the blood, central nervoussystem and other tissue fluids of their mammalianhosts. T. cruzi, by contrast adopts an intracellularenvironment within their mammalian hosts. Inva-sion and passage through different anatomicallocations within insect vectors also distinguishesthe parasites.

All trypanosome species use glucose as a crucialsource of energy, and all have specific plasmamembrane transporters to facilitate the uptake ofthis molecule. Four different trypanosome glucosetransporter genes have been cloned, and theirfunction verified by expression in either Xenopusoocytes or Chinese hamster ovary (CHO) cells. Itappears that the transporters have adapted to bestsuit the needs of the parasites in the differentenvironments with which they are confronted. Forexample, availability of free glucose in serum dif-fers greatly from that in a mammalian cell’s cyto-plasm or within the midgut of insect vectors.

The trypanosome glucose transporters belongto the glucose transporter superfamily, exem-plified by the mammalian erythrocyte transporter,GLUT1 [1]. Nevertheless they have particularstructural and functional features which distin-guish them from their mammalian counterparts.These include: a conserved array of cysteineresidues in the first exofacial loop; a relative in-sensitivity to various pharmacological reagents

* Corresponding author. Tel.: +33 557571014; fax: +33557571015.

0166-6851/98/$19.00 © 1998 Francqui Foundation. Published by Elsevier Science B.V. All rights reserved.

PII S 0166 -6851 (97 )00192 -8

M.P. Barrett et al. / Molecular and Biochemical Parasitology 91 (1998) 195–205196

such as cytochalasin B; and a substrate recogni-tion profile which includes the ability to transportD-fructose. It is possible that the trypanosomeglucose transporters may represent eitherchemotherapeutic targets themselves, or gatewayswhich allow the targeting of other toxic moleculesto these parasites.

2. Cloning of the trypanosome glucose transportergenes

A gene cloned from a T. brucei EATRO 164cDNA library [2] was found to have high homol-ogy to a glucose transporter gene previously iso-lated from the related kinetoplastid parasiteLeishmania enriettii [3]. The gene was found to bea member of a clustered gene family in the T.brucei brucei genome. In fact, two separate iso-forms, called THT1 and THT2 (THT= try-panosome hexose transporter) were isolated [4],differing predominantly in the carboxy terminaldomain, and a cysteine rich first exofacial loop.Identical genes have been found in T. bruceirhodesiense, T. b. gambiense, T. equiperdum and T.e6ansi [5].

The T. brucei THT1 gene was used to probe alibrary of T. cruzi genomic DNA, and a relatedgene was identified [6], although only a singleisoform (TcrHT1), also in multiple copies, couldbe identified in T. cruzi.

Having identified genes encoding glucose trans-porters from T. brucei and T. cruzi, and given theavailability of related sequences from Leishmaniaparasites, regions of high conservation were iden-tified. Oligonucleotides based on these motifswere designed and used to amplify related se-quences from Trypanosoma 6i6ax using the poly-merase chain reaction [7]. Sequencing of threetandemly arranged TvHT1 copies present in theY481 strain of T. 6i6ax showed that there is onlya single TvHT isoform (Bringaud et al., unpub-lished data).

The different trypanosome glucose transporteramino acid sequences are all highly homologousto one another (52–80% identity). The geneswhich flank the glucose transporter gene clustersfrom all three species of trypanosome are con-

served and found in the same positions relative tothe glucose transporter genes revealing a highconservation of genome organisation within thegenus Trypanosoma (Bringaud et al., unpublisheddata).

All of the trypanosome glucose transportergenes conform to the widely held view of mem-brane topology of the glucose transporter super-family in that hydrophobicity plots predict 12transmembrane helices (Fig. 1). The sequences arealigned in Fig. 2.

2.1. T. brucei contains two glucose transporterisoforms: T. 6i6ax and T. cruzi contain only one

The glucose transporters from all of the try-panosome species studied exist as multiple tandemcopies. In T. brucei two different isoforms existwhile T. cruzi and T. 6i6ax contain just one. Thecopy number varies greatly between individualisolates in the case of T. brucei [8].

Northern blot analysis (Fig. 3) revealed that inT. brucei THT1 is the most abundantly expressedisoform in bloodstream form parasites, whileTHT2 is the most abundant in procyclics. THT2levels are similar throughout the life-cycle whileTHT1 transcripts are some 40-fold more abun-dant than THT2 transcripts in bloodstream formsand absent or present at very low levels in pro-cyclic forms [4]. The 3% untranslated region of theTHT1 gene is responsible for its differential ex-pression, rendering transcripts stable in blood-stream but not procyclic cells [9]. The expressionof two glucose transporter isoforms in T. brucei isreflected by the presence of different measurablekinetic properties of transport in bloodstreamform and procyclic organisms. Bloodstream formcells possess a low affinity, high capacity glucoseuptake system [10–13], while procyclics possess ahigher affinity lower capacity system [14–17](Table 1).

Sequence information alone was insufficient toassign functions to the putative transporters,hence all of the trypanosome glucose transportergenes have been expressed in heterologus systems,including CHO cells and Xenopus oocytes to testfunction. Expression of the THT2 isoform inCHO cells [16] (in which the endogenous trans-

M.P. Barrett et al. / Molecular and Biochemical Parasitology 91 (1998) 195–205 197

Fig. 1. Schematic representation of the mammalian GLUT1 glucose transporter and the trypanosome transporters. The 12transmembrane hydrophobic helices are numbered. A ‘Y’ represents the glycoslylation site of the first exofacial loop of themammalian transporter. Shaded circles in the trypanosomal first exofacial loop represent the cysteine residues. A particularly longintracellular loop between helices 6 and 7 of GLUT1 is not present in the trypanosome transporters.

porter was inhibited by cytochalasin B) revealed itto be kinetically and pharmacologically similar tothe transporter identified in procyclic cells. THT1cRNA, expressed in Xenopus oocytes [4] had asubstrate recognition profile and pharmacologysimilar to the transporter measured in blood-stream form trypomastigotes.

The parasites therefore have two transportergenes that are expressed in a fashion which allowsmaximal exploitation of the host’s extracellularenvironment. Bloodstream forms express predom-inantly THT1, a high capacity low affinity trans-porter to exploit the high concentration (�5mM) of glucose in mammalian serum. Procyclicforms express the higher affinity transporter,THT2, in the insect midgut where glucose is rela-tively scarce, and amino acids such as prolinebecome the major energy source [18].

T. cruzi, by contrast, has a single glucose trans-porter gene isoform [6], which is expressed atsimilar levels in epimastigotes, which dwell in theinsect midgut, and trypomastigotes which livetransiently in the bloodstream of the mammalbefore invading cells (particularly neurons andmyocytes) where they transform to cytoplasmicamastigotes. Glucose transport has not been mea-sured in amastigotes.

The metabolism of glucose by T. cruzi differssignificantly from that of T. brucei bloodstreamforms [19,20] which lack a functional Krebs’ cycleor a full mitochondrial respiratory chain and gen-erate, under aerobic conditions just two moles ofATP per mole of glucose consumed via the gly-colytic flux which ends at pyruvate. All life-stagesof T. cruzi have a less profligate use of glucose,using functional mitochondrial systems to gener-ate more ATP per mole of glucose. (Procyclic

M.P. Barrett et al. / Molecular and Biochemical Parasitology 91 (1998) 195–205198

Fig

.2.

Am

ino-

acid

sequ

ence

alig

nmen

tof

the

tryp

anos

ome

gluc

ose

tran

spor

ters

.Id

enti

cal

amin

oac

ids

are

unde

rsco

red

wit

han

aste

risk

.T

hecy

stei

nes

ofth

efir

stex

ofac

ial

loop

are

mar

ked

wit

han

arro

wab

ove

the

first

sequ

ence

.T

rans

mem

bran

e,hy

drop

hobi

cdo

mai

nsar

ein

dica

ted.

Pot

enti

alN

-lin

ked

glyc

osyl

atio

nsi

tes

are

unde

rlin

ed.

M.P. Barrett et al. / Molecular and Biochemical Parasitology 91 (1998) 195–205 199

Fig. 3. Expression of the THT genes in the T. brucei EATRO-164 bloodstream forms (BF) and T. brucei Stib-247 procyclicforms (PF). The Northern blot was successively hybridizedwith different 32P-labeled probes common to THT1 and THT2genes (common probe), specific to THT1 genes (THT1 probe)and specific to THT2 genes (THT2 probe).

T. cruzi bloodstream form trypomastigotes aretransiently bathed in the high glucose concentra-tions of mammalian serum, however, intracellularforms are exposed to relatively low free glucose(most cellular glucose is locked up by the host ina phosphorylated or polymeric form). Presumablyfor this reason they need only a single, relativelyhigh affinity transporter (Table 1), and sincebloodstream form trypomastigotes do not divide,and as their duration in the bloodstream is onlytransient, they have not evolved a low affinity,high capacity transporter to exploit this environ-ment.

The metabolism of T. 6i6ax is unlike that ofbloodstream form T. brucei in that a partialKrebs’ cycle is present [21,22] and glucose utilisa-tion is more efficient in terms of ATP production.Nevertheless, it is exposed to high glucose concen-trations in the mammalian bloodstream and ex-presses a correspondingly low affinity transporter[7] (Table 1). Intriguingly, the parasite has only asingle isoform glucose transporter gene, whichmight reflect the fact that its passage through thetsetse fly, unlike T. brucei, does not take it beyondthe proboscis and is relatively fast (several days asopposed to several weeks).

form T. brucei have a similar metabolism to T.cruzi.)

Table 1Kinetic parameters and substrate specificities of the trypanosome glucose transporters

Transport of D-fructoseGenesOrganisms Type ofVmax (nmol min−1Km D-glucose ortransportper mg protein)2-DOG (mM) or inhibition

0.49–0.9a,b,c,d �250a,b,c,d +b FacilitatedaT. brucei blood-stream forms

4–10e,f,g +e,f Facilitated/0.045–0.080e,f,gT. brucei procyclicforms Activee,f

FacilitatedhT. brucei +hNDNDTHT1 (Xeno-pus)

0.151 (2-DOG)fTHT2 (CHO) +f Facilitatedf4.3f

T. cruzi trypo- ND Facilitated0.294 (2-DOG) 1.4mastigote forms

FacilitatedT. cruzi epimastig- 0.312 (2-DOG) 3.65 +ote formsi,j

T. cruzi 0.315 (2-DOG) Facilitated+TcrHT1 (CHO) 7.7

+88.50.585 FacilitatedT. 6i6ax blood-stream formsk

4.30.545 (2-DOG) FacilitatedTvHT1 (CHO) +T. 6i6ax

Values are from: a [10]; b [11]; c [13]; d [31]; e [14]; f [16]; g [17]; h [4]; i [6]; j [26]; k [7].

M.P. Barrett et al. / Molecular and Biochemical Parasitology 91 (1998) 195–205200

2.2. Facilitati6e 6ersus secondary acti6e transport

The uptake of nutrients into cells may occur bya variety of mechanisms [23]. Large polarmolecules, such as glucose, and indeed most bio-chemical substrates are too large and polar tosimply diffuse across the lipid bilayer, and thusrequire specific transport proteins to catalyse up-take. Cells may accumulate nutrients without en-ergy expenditure, allowing the thermodynamicenergy in a diffusion gradient to propel substratesinto cells. Alternatively, if a diffusion gradientcannot be relied upon (i.e. if substrate is scarce ora cell needs to concentrate it beyond the extracel-lular level) energy is expended. Such active trans-port can be direct (the transporter uses eitherATP or pyrophosphate to accumulate substrate),or indirect. Indirect active transport involves theco-transport of two substrates (either in the same,symport, or opposite, antiport, direction). Theco-transport of glucose with sodium ions in themammalian intestine, and lactose with protons inEscherichia coli serve as classic examples of sec-ondary active transport [23]. The energy to drivethese systems comes from the diffusion gradientsof co-transported ions which are maintained bysecondary energy dependent pumping systems.

The different types of transport may be distin-guished experimentally. Active transporters accu-mulate substrates against a concentration gradientand comparison of intracellular and extracellularconcentrations of substrate can show when a sub-strate has been actively concentrated. If a secondsubstrate (usually an ion) is required, transport inthe presence and absence of that ion will bemarkedly different. Finally pharmacologicalreagents which either dissipate the ion gradientsrequired to drive secondary active transport, oragents which inhibit energy production within thecell can also inhibit secondary active transport.

It is generally agreed that bloodstream forms ofT. brucei depend on a facilitiative diffusionmolecule to accumulate glucose [10–13]. Giventhese organisms have evolved in a glucose richenvironment, and that glucose is metabolised vir-tually instantaneously, ensuring a perpetual con-centration gradient, energy expenditure onglucose accumulation would be unnecessary. One

study did suggest a dependence on sodium ions atlow external glucose concentration [24].

The case of the procyclic trypanosome is morecontroversial. Procyclics are exposed to low levelsof glucose in the tsetse midgut, however, after atsetse feed they are exposed, transiently, to highlevels of the sugar in mammalian serum. The firststudy on these organisms revealed a relativelyhigh affinity transporter, which according topharmacological investigation was dependent onthe co-transport of protons [14]. Both the respira-tory inhibitor KCN and the protonophore FCCPreduced transport. Two further studies, both em-ploying high concentrations of substrate failed toshow either an accumulation of glucose beyondthe external levels [15], or inhibition by KCN orFCCP [17]. In fact, an inhibitory effect by bothKCN and FCCP depends on the concentration ofsubstrate [16]. It appears that these reagents in-hibit transport at low external glucose concentra-tion, but exert progressively less impact as thesubstrate concentration rises to 1 mM. This phe-nomenon is characteristic of ‘slippage’ [25]whereby the dependence on the counter-ion isdiminished as substrate concentration rises. In thecase of the procyclic trypanosome, such a mecha-nism makes physiological sense given that nor-mally the parasites are exposed to low externalglucose and might expend energy upon its acquisi-tion. During the transient explosion in glucoseconcentration following a tsetse meal, however,dependence on energy to acquire the abundantsubstrate would be futile. One difficulty with the‘slippage’ hypothesis is that transport via THT2,the procyclic transporter, expressed in CHO cells,had no dependency on protons or energy regard-less of substrate concentration [16]. The situationin procyclic cells might be more complicated,involving other phenomena including putativeauxiliary factors or membrane potential whichwould be effected by the same pharmacologicalreagents used to assess proton dependence.

All available evidence suggests that T. cruzi[6,26] and T. 6i6ax [7] have straight forward facil-itative transporters, although controversy alsosurrounds the mechanism employed by leishma-nial transporters [27,28].

M.P. Barrett et al. / Molecular and Biochemical Parasitology 91 (1998) 195–205 201

Fig. 4. Alignment of the cysteine rich loop of THT1 with domains from two T. brucei variant surface glycoproteins. The amino acidsequence of the first exofacial loop, when used to search the protein data bases identified homology with small conserved regionsof T. brucei variant surface glycoproteins. Alignment with the regions of the AnTaT1.8 and ILTaT1.24 sequences is shown here. Thecysteine residues are over-scored with an arrow. An asterisk below the aligned sequences highlights identical residues.

2.3. The first extracellular loop of trypanosomeglucose transporters contains a conser6edarrangement of cysteine residues

All of the trypanosome glucose transporterspossess numerous, spatially conserved cysteineresidues in their first exofacial loop. Extensiveanalysis of the sequences of THT1 and THT2reveal the loop to have been inserted indepen-dently [5], probably to replace a different type offirst exofacial loop in an ancestral gene. Intrigu-ingly, when the amino-acid sequence of the loopwas used to scan protein sequence data-bases itidentified a similar conserved arrangement of cys-teine residues in variant surface glycoproteinsfrom African trypansomes (Fig. 4). In fact, manytrypanosomatid surface proteins reveal this ar-rangement of cysteines [29]. The motif has beenconserved suggesting it performs a critical role inthe architecture of kinetoplastid surface mem-branes.

3. Glycosylation of trypanosomal glucosetransporters

The mammalian GLUT1 transporter, and vari-ous other eukaryotic hexose transporters are gly-cosylated, usually in the first exofacial loop [30]which corresponds to the cysteine rich loop of thetrypanosome transporters described above. West-ern blot analysis of THT1 after proteinase Kdigest of intact parasites, using site-specific anti-bodies against the amino terminus, carboxy termi-nus or two predicted extracellular loopsconnecting helices 5/6 and 7/8, has revealed that

THT1 is oriented in the plasma membrane ac-cording to the predicted model based on GLUT1(Seyfang et al., unpublished data).

Amino acid analysis of the various trypanoso-mal glucose transporters revealed significant dif-ferences in the occurrence of potential N-linkedglycosylation sites (Fig. 2). Asn-69 in T. bruceiTHT2, Asn-81 in T. cruzi TcrHT1 and Asn-90,Asn-91 in T. 6i6ax TvHT1 are the only potentialexofacial N-glycosylation sites in these trans-porters, all located within the first extracellularloop.

The T. brucei bloodstream form transporter,THT1, lacks potential N-linked glycosylation siteson any of the predicted extracellular loops (onlythe loops which end up outside the cell are topo-logically available for glycoslylation within theendoplasmic reticulum). THT1 does contain oneN-glycosylation consensus sequence (Asn-7), al-though this is located on the amino terminal tailand is predicted to face the cytoplasm. THT2(Asn-7) and TvHT1 (Asn-10) also have additionalN-glycosylation consensus sequences at similarpositions.

De-glycosylation experiments with THT1-en-riched plasma-membrane fractions [31] (using N-glycosidase, O-glycosidase and sialidasetreatment) in combination with Western blotanalysis showed no band shift (reduced molecularweight) for the THT1 protein [32]. In contrast, theGLUT1 protein purified from human erythrocytesand de-glycosylated had a reduction in apparentmolecular weight of about 3 kDa after N-glycosi-dase F treatment. THT1-enriched plasma-mem-brane fractions were also analysed forglycoproteins by lectin blot analysis. A glyco-

M.P. Barrett et al. / Molecular and Biochemical Parasitology 91 (1998) 195–205202

protein of 45 kDa (gp45) which co-migrated onreducing SDS-PAGE gels with THT1 [32] wasdetected. However, in non-denaturing gels, THT1showed a lower electrophoretic mobility (shiftingfrom 45 to 51 kDa) and under these conditions noglycoproteins co-migrated with THT1. These cu-mulative data demonstrated that THT1 is notglycosylated, in agreement with the predictionsdrawn from amino acid sequence data and mem-brane topology predictions.

It is believed that glycoslylation is critical incorrectly orienting mammalian transporters in theplasma membrane [33]. Clearly in the case ofbloodstream form T. brucei the transporter can beoriented without glycoslylation. Possibly thedense packaging of the highly glycosylated variantsurface glycoproteins of bloodstream form T. bru-cei might influence the need for glycosylation ontransporter proteins in the plasma membrane. In-terestingly, expression of THT2, TvHT1 andTcrHT1 was successful in CHO cells, while THT1could not be functionally expressed (Tetaud et al.,unpublished data). Possibly it is the lack of glyco-sylation to THT1 which hinders expression inmammalian cells.

4. Therapeutic targets and therapeutic targeting

Any aspect of a parasite’s biochemical make-upmay be vulnerable to therapeutic attack, particu-larly where significant differences between hostand parasite can be identified. In the case oftrypanosome glucose transporters significant dif-ferences can be identified in terms of both thepharmacology and substrate recognition profileswhen compared to the mammalian GLUT1 trans-porter. For example, all of the trypanosome trans-porters also carry D-fructose [7,16,26,34] whileGLUT1 does not. GLUT1 is very much moresusceptible to cytochalasin B and phloretin thanthe trypanosome transporters. An extensive studyof other reagents would undoubtedly identifysome which react more potently with the try-panosome transporters. None of the drugs used inthe chemotherapy of human African trypanoso-miasis inhibit the glucose transporter of blood-stream form trypomastigotes when used at

concentrations usually toxic to the cells, suggest-ing that none of these drugs exert their activity atthe level of this transporter (Barrett et al., unpub-lished observations).

Detailed analysis of substrate analogues, substi-tuting the various hydroxyl groups has identifiedthe key features involved in the structure-activityrelationship of the various trypanosome glucosetransporters (Fig. 5). Bloodstream form T. bruceidepends exclusively on glucose metabolism togenerate energy, hence inhibition of glucose up-take would certainly be expected to exert a lethaleffect. So far, however, specific inhibitors of theT. brucei glucose transporter have not been iden-tified. Armed with the substrate recognition infor-mation, and given the high capacity of the T.brucei bloodstream from glucose transporter tofulfill the prodigious appetite these cells have forthis substrate, it might be possible to subvert thetransporter, to carry toxic analogues of glucoseinto the cell.

A similar situation exists in the case of themelaminophenyl based arsenical and diamidineclasses of drug which enter trypanosomes via anunusual amino-purine transporter, termed P2 [35].The drugs share a chemical motif with the normalamino-purine substrates which is recognised bythe transporter. Loss of the transporter rendersthe parasites resistant to these drugs.

Since the T. brucei bloodstream form trans-porter does not make hydrogen bonds with thehydroxyls at positions 2 and 6 of the glucose ring[11], these sites were considered available for theattachment of other chemical constituents, whichwould not interfere with recognition by the trans-porter. In fact, in the case of the C-6 position, astrict limit on the size of substituent groups hasbeen noted (Barrett et al., unpublished data),whereas, provided substituent groups were neithercharged nor aromatic, relatively large replace-ments could be added at position C-2. At leastone compound containing a substituent group atposition C-2 has been developed, which is toxic tobloodstream form parasites grown in vitro (Bar-rett et al., unpublished data). Fructose analogueshave also been developed [36], and toxic examplesare known. Further work is underway to assessthe chemotherapeutic value of these compounds,

M.P. Barrett et al. / Molecular and Biochemical Parasitology 91 (1998) 195–205 203

Fig. 5. Substrate recognition of the T. brucei bloodstream form, and procyclic form glucose transporters, compared with mammalianGLUT1. Hydrogen bonds predicted to occur between the transporters and D-glucose are represented by jagged lines. The differentcarbon positions are numbered 1–6 (data from [11,16,39]).

and to distinguish whether the products actuallyinhibit the transporter itself or hit intracellulartargets once taken into the cell via the transporter.A potential drawback with the glucose transporterrelates to its low affinity for substrate, and thehigh abundance of glucose within serum, whichwould out-compete analogues for the transporterif such analogues were prescribed in quantitieswithin the range of currently used antimicrobialdrugs.

5. Discussion

Glucose transporter genes have been clonedfrom several parasitic protozoa of the genus Try-panosoma. The function of the genes has been

verified by expression in both Xenopus oocytesand CHO cells.

T. brucei contains two separate glucose trans-porter gene isoforms, which account for the dif-ferent kinetic properties of the transportersmeasured in the different life cycle forms of theparasite. T. cruzi, and T. 6i6ax contain only asingle isoform. Substrate affinity and possiblymechanistic differences distinguish the trans-porters, in a manner which may reflect the differ-ent extracellular environments to which theparasites are exposed.

All of the trypanosome glucose transportersalso recognise D-fructose which distinguishesthem from the main mammalian glucose trans-porter, GLUT1. However, there are at least fourother mammalian plasma membrane hexose

M.P. Barrett et al. / Molecular and Biochemical Parasitology 91 (1998) 195–205204

transporters expressed in different tissues, with arange of substrate specificities, including two iso-forms (GLUT2-liver and GLUT5-intestine)[37,38] which also recognise D-fructose, highlight-ing the difficulties in pin-pointing unique featuresof the trypanosome transporters which might hin-der their utility as chemotherapeutic targets.

The trypanosome transporters all contain a cys-teine-rich first exofacial loop with a structureconserved in other trypanosomatid membraneproteins, which may play a critical role in themembrane architecture of these parasites. Thecomparative situation of the loop in THT1 andTHT2 suggests it was inserted independently intoancestral genes. The presence of a similar struc-ture in other trypanosomatid membrane proteins,might suggest that the insertion of ‘cassettes’which donate particular functions to proteins, hasbeen involved in the evolution of trypanosomemembrane proteins.

Structure activity relationship studies betweenthe transporters and their substrates has allowedthe development of analogues which are stillrecognised by the transporter, and may herald anew pathway in the development of novel anti-trypanosomal reagents, which exploit the glucosetransporter as a specific gateway to carry toxicreagents into the parasites.

Acknowledgements

MPB was funded by a Royal Society EuropeanExchange fellowship. This work was funded bythe CNRS, the Commission of the EuropeanCommunity, the Ministere de l’Enseignement Su-perieur et de la Recherche and the GDR CNRS/DGA-DSP.

References

[1] Gould GW, Holman GD. The glucose transporter fam-ily: structure, function and tissue-specific expression.Biochem J 1993;295:329–41.

[2] Bringaud F, Baltz T. A potential hexose transportergene expressed predominantly in the bloodstream formof Trypanosoma brucei. Mol Biochem Parasitol1992;52:111–22.

[3] Cairns BR, Collard MW, Landfear SM. Developmen-tally regulated gene from Leishmania encodes a putativemembrane transport protein. Proc Natl Acad Sci USA1989;86:7682–6.

[4] Bringaud F, Baltz T. Differential regulation of twodistinct families of glucose transporter genes in Try-panosoma brucei. Mol Cell Biol 1993;13:1146–54.

[5] Bringaud F, Baltz T. African trypanosome glucosetransporter genes: organisation and evolution of a multi-gene family. Mol Biol Evol 1994;11:220–30.

[6] Tetaud E, Bringaud F, Chabas S, Barrett MP, Baltz T.Characterization of glucose transport and cloning of ahexose transporter gene in Trypanosoma cruzi. Proc NatlAcad Sci USA 1994;91:8278–82.

[7] Waitumbi JN, Tetaud E, Baltz T. Glucose uptake inTrypanosoma 6i6ax and molecular characterization of itstransporter gene. Eur J Biochem 1996;237:234–9.

[8] Barrett MP, Bringaud F, Doua F, Melville SE, Baltz T.Hypervariability in gene copy number for the glucosetransporter genes in trypanosomes. J Euk Microbiol1996;43:244–9.

[9] Hotz H-R, Lorenz P, Fischer R, Krieger S, Clayton C.Role of 3%-untranslated regions in the regulation ofhexose transporter mRNAs in Trypanosoma brucei. MolBiochem Parasitol 1995;75:1–14.

[10] Gruenberg J, Sharma PR, Deshusses J. D-Glucose trans-port in Trypanosoma brucei. Eur J Biochem1978;89:461–9.

[11] Eisenthal R, Game S, Holman GD. Specificity andkinetics of hexose transport in Trypanosoma brucei.Biochim Biophys Acta 1989;985:81–9.

[12] Seyfang A, Duszenko M. Specificity of glucose transportin Trypanosoma brucei : effective inhibition by phloretinand cytochalasin B. Eur J Biochem 1991;202:191–6.

[13] Ter Kuile BH, Opperdoes FR. Glucose uptake by Try-panosoma brucei : rate-limiting steps in glycolysis andregulation of the glycolytic flux. J Biol Chem1991;266:857–62.

[14] Parsons M, Nielsen B. Active transport of 2-deoxy-D-glucose in Trypanosoma brucei procyclic forms. MolBiochem Parasitol 1990;42:197–204.

[15] Ter Kuile BH, Opperdoes FR. Mutual adjustment ofglucose uptake and metabolism in Trypanosoma bruceigrown in a chemostat. J Bacteriol 1992;174:1273–9.

[16] Barrett MP, Tetaud E, Seyfang A, Bringaud F, Baltz T.Functional expression and characterization of the Try-panosoma brucei procyclic glucose transporter, THT2.Biochem J 1995;321:687–91.

[17] Wille U, Seyfang A, Duszenko M. Glucose uptake oc-curs by facilitated diffusion in procyclic forms of Try-panosoma brucei. Eur J Biochem 1996;236:228–33.

[18] Cazzulo JJ. Aerobic fermentation of glucose by try-panosomatids. FASEB J 1992;6:3153–61.

[19] Fairlamb AH, Opperdoes FR. Carbohydratemetabolism in African trypanosomes, with special refer-ence to the glycosome. In: Morgan MJ, editors. Carbo-hydrate Metabolism in Cultured Cells. New York:Plenum, 1986:183–219.

M.P. Barrett et al. / Molecular and Biochemical Parasitology 91 (1998) 195–205 205

[20] Opperdoes FR. Compartmentation of carbohydratemetabolism in trypanosomes. Ann Rev Microbiol1987;41:127–51.

[21] Gardiner PR. Recent studies of the biology of Try-panosoma 6i6ax. Adv Parasitol 1989;28:229–317.

[22] Bowman IBR, Flynn IW. Oxidative metabolism of try-panosomes. In: Lumsden WHR, Evans DA, editors.Biology of the Kinetoplastida, vol. 1. New York: Aca-demic, 1976:435–476.

[23] Stein WD. Channels, Carriers, and Pumps: An Introduc-tion to Membrane Transport. San Diego: Academic,1990.

[24] Munoz-Antonia T, Richards FF, Ullu E. Differences inglucose transport between bloodstream and procyclicforms of Trypanosoma brucei rhodesiense. Mol BiochemParasitol 1991;47:73–82.

[25] Eddy AA. Slip and leak models of gradient-coupledsolute transport. Biochem Soc Trans 1980;8:271–3.

[26] Tetaud E, Chabas S, Giroud C, Barrett MP, Baltz T.Hexose uptake in Trypanosoma cruzi : structure-activityrelationship between substrate and transporter. BiochemJ 1996;317:353–9.

[27] Zilberstein D, Dwyer DM. Protonmotive force-drivenactive transport of D-glucose and L-proline in the proto-zoan parasite Leishmania dono6ani. Proc Natl Acad SciUSA 1985;82:1716–20.

[28] Ter Kuile BH, Opperdoes FR. Uptake and turnover ofglucose in Leishmania dono6ani. Mol Biochem Parasitol1993;60:313–22.

[29] Carrington M, Boothroyd J. Implications of conservedstructural motifs in disparate trypanosome surface-proteins. Mol Biochem Parasitol 1996;81:119–26.

[30] Mueckler M. Facilitative glucose transporters. Eur JBiochem 1994;219:713–25.

[31] Seyfang A, Duszenko M. Functional reconstitution ofthe Trypanosoma brucei plasma-membrane D-glucosetransporter. Eur J Biochem 1993;214:593–7.

[32] Seyfang A. Characterization and isolation of the glucosetransporter from Trypanosoma brucei. PhD thesis. Uni-versity of Tubingen, 1993.

[33] Hresko RC, Kruse M, Strube M, Mueckler M. Topol-ogy of the GLUT1 glucose transporter deduced fromglycosylation scanning mutagenesis. J Biol Chem1994;269:20482–8.

[34] Fry AJ, Towner P, Holman GD, Eisenthal R. Transportof D-fructose and its analogues by Trypanosoma brucei.Mol Biochem Parasitol 1993;60:9–18.

[35] Carter NS, Fairlamb AH. Arsenical resistant try-panosome lack an unusual adenosine transporter. Na-ture 1993;361:173–5.

[36] Page P, Blonski C, Perie J. An improved chemical andenzymatic synthesis of new fructose derivatives for im-port studies by the glucose transporter in parasites.Tetrahedron 1996;52:1557–72.

[37] Colville CA, Seatter MJ, Jess TJ, Gould GW, ThomasHM. Kinetic analysis of the liver-type (GLUT2) andbrain-type (GLUT3) glucose transporters in Xenopusoocytes: substrate specificities and effects of transportinhibitors. Biochem J 1993;290:701–6.

[38] Burant CF, Takeda J, Brot-Laroche E, Bell GI, David-son NO. Fructose transporter in human spermatozoaand small intestine is GLUT5. J Biol Chem1992;267:14523–6.

[39] Silverman M. Structure and function of hexose trans-porters. Annu Rev Biochem 1991;60:757–94.

.