GLUT-1 AS A RECEPTOR FOR HTLV ENVELOPES AND ITS USES

Abstract

The present application relates to a method for diagnosing a glucose transporter type 1 (GLUT1) deficiency syndrome that utilizes polypeptides derived from the soluble part of the glycoprotein of the enveloped virus of primate T-cell leukemia virus (PTLV). The polypeptides, named receptor binding domain ligands (RBD), are selected for their ability to bind specifically to GLUT1. The method involves determining the level of GLUT 1 expression at the cell surface and comparing the level to a reference value.

Claims

1. A method for diagnosing a glucose transporter type 1 (GLUT1) deficiency syndrome, comprising: a) collecting sample from a subject, b) determining the level of GLUT1 expression at a cell surface using an isolated polypeptide, wherein said polypeptide is a soluble receptor binding domain (RBD) ligand derived from the soluble part of the glycoprotein of a primate T-lymphotropic virus binding to GLUT1 or a fragment thereof, and c) comparing said level to a reference value.

2. The method of claim 1, wherein GLUT1 comprises an amino acid sequence presenting a sequence identity of at least 70% with SEQ ID NO: 2.

3. The method of claim 1, wherein the RBD ligand binds to at least one of the following fragments of GLUT1: TABLE-US-00008 SEQ ID NO: 42 (IVGMCFQYVEQLC) SEQ ID NO: 35 (NAPQKVIEEFY) SEQ ID NO: 36 (NQTWVHRYGESILPTTLTTLWS) SEQ ID NO: 37 (KSFEMLILGR) SEQ ID NO: 38 (DSIMGNKDL) SEQ ID NO: 39 (YSTSIFEKAGVQQP) SEQ ID NO: 40 (EQLPWMSYLS) SEQ ID NO: 41 (QYVEQLC)

4. The method of claim 1, wherein the RBD ligand is selected from the group consisting of human T-cell leukemia virus (HTLV) 2.RBD, HTLV1.RBD, HTLV4.RBD, HTLV3.RBD, simian T-cell leukemia virus (STLV) 1.RBD, STLV2.RBD and STLV3.RBD.

5. The method of claim 4, wherein HTLV2.RBD comprises the amino acid sequence SEQ ID NO: 4 or SEQ ID NO: 5 or SEQ ID NO: 7 or SEQ ID NO: 43 or fragments or variants thereof.

6. The method of claim 4, wherein HTLV1.RBD comprises the amino acid sequence SEQ ID NO: 9 or SEQ ID NO: 10 or SEQ ID NO: 11 or SEQ ID NO: 13 or SEQ ID NO: 15 or SEQ ID NO: 17 or SEQ ID NO: 19 or SEQ ID NO: 21 or fragments or variants thereof.

7. The method of claim 4, wherein HTLV4.RBD comprises the amino acid sequence SEQ ID NO: 22 or SEQ ID NO: 23 or SEQ ID NO: 51 or fragments or variants thereof.

8. The method of claim 4, wherein HTLV3.RBD comprises the amino acid sequence SEQ ID NO: 53 or fragments or variants thereof.

9. The method of claim 4, wherein STLV1.RBD comprises the amino acid sequence SEQ ID NO: 25 or fragments or variants thereof.

10. The method of claim 4, wherein STLV2.RBD comprises the amino acid sequence SEQ ID NO: 27 or fragments or variants thereof.

11. The method of claim 4, wherein STLV3.RBD comprises the amino acid sequence SEQ ID NO: 29 or SEQ ID NO: 55 or fragments or variants thereof.

12. The method of claim 1, wherein the RBD ligand is fused to a tag, an antibody constant fragment or a fluorescent protein.

13. The method of claim 1, wherein the reference value consists of the level of GLUT1 expression at the cell surface determined in a sample from a substantially healthy subject.

14. The method of claim 1, wherein the reference value consists of the level of GLUT1 expression at the cell surface determined in samples from a reference population comprising at least 100 substantially healthy subjects.

15. The method of claim 1, wherein the reference value consists of the level of GLUT1 expression at the cell surface determined in a sample from a subject having a GLUT1 deficiency syndrome.

16. The method of claim 1, wherein the reference value consists of the level of GLUT1 expression at the cell surface determined in samples from a reference population comprising at least 10 subjects having a GLUT1 deficiency syndrome.

17. The method of claim 1, wherein the GLUT1 deficiency syndrome is characterized by an encephalopathy marked by childhood epilepsy that is refractory to treatment, deceleration of cranial growth leading to microcephaly, psychomotor retardation, spasticity, ataxia, dysarthria or other paroxysmal neurological phenomena often occurring before meals.

18. The method of claim 1, wherein the GLUT1 deficiency syndrome is associated with de novo or inherited mutations in the SLC2A1 gene.

19. The method of claim 1, wherein the GLUT1 deficiency syndrome is associated with low glucose level and low lactate concentration in the cerebrospinal fluid (CSF) in the absence of hypoglycemia.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0257] FIGS. 1a-1d Expression of the HTLV receptor-binding domain alters cellular metabolism.

[0258] FIG. 1a, Medium acidification and syncytia formation in 293T cells one day post-transfection with control DNA or Env expression vectors, including syncytial wild-type HTLV-1 Env and HTLV-2 Env, a non-syncytial chimeric H.sub.183FEnv, and syncytial A-MLV ΔR Env.

[0259] FIG. 1b, Extracellular lactate and glucose in the culture medium of 293T cells were measured two days following transfection with an irrelevant DNA (control), F-MLV Env, H.sub.183FEnv, HTLV-1 RBD (H1.sub.RBD) or amphotropic MLV RBD (A.sub.RBD) expression vectors. Lactate and glucose concentrations were normalized to cellular protein content.

[0260] FIG. 1c, 2-deoxyglucose and fructose uptake following transfection of 293T with an irrelevant DNA (control), H1.sub.RBD, H2.sub.RBD or A.sub.RBD expression vectors. Control cells were also incubated with glucose transporter inhibitors cytochalasin and phloretin. Data are the means of triplicate measures and are representative of two to three independent experiments.

[0261] FIG. 1d, Expression of the HTLV and amphotropic-MLV receptors on 293T (1) and Jurkat T (2) cells cultured overnight in the presence or absence of glucose was monitored by binding of H1.sub.RBD and A.sub.RBD, respectively.

[0262] FIGS. 2a and 2b HTLV receptor properties correlate with GLUT1 properties.

[0263] FIG. 2a, Expression of the HTLV and amphotropic-MLV receptors at the surface of human and murine erythrocytes, as well as human primary hepatocytes.

[0264] FIG. 2b, H1.sub.RBD and A.sub.RBD binding to Jurkat cells in the absence or presence of the Glut-1 inhibitor cytochalasin B.

[0265] FIGS. 3a-3c HTLV receptor-binding correlates with altered lactate metabolism.

[0266] FIG. 3a, Expression of H1.sub.RBD and the derived mutants D106A and Y114A was monitored by Western blot analysis of the supernatants of 293T cells following transfection with the various expression plasmids.

[0267] FIG. 3b, Binding of H1.sub.RBD and the D106A and Y114A mutants to the HTLV receptor on HeLa cells.

[0268] FIG. 3c, Extracellular lactate in the medium of 293T cells one day post transfection with an irrelevant DNA (control), H1.sub.RBD or the H1.sub.RBD D106A and Y114A mutants. Data are representative of three independent experiments.

[0269] FIGS. 4a-4c GLUT-1 is a receptor for HTLV envelopes.

[0270] FIG. 4a, Binding of H1.sub.RBD, H2.sub.RBD, H2.sub.RBD D102A mutant, and A.sub.RBD to control 293T cells or 293T cells overexpressing either GLUT-1 or PiT2.

[0271] FIG. 4b, Binding of H2.sub.RBD-EGFP to cells overexpressing GLUT-1-HA or GLUT-3-HA, and corresponding immunoblots using an anti-HA antibody.

[0272] FIG. 4c, Immunoprecipitation of GLUT-1-HA from 293T cells transfected with either an irrelevant construct, GLUT-1 alone, H1RBD alone, H1RBD Y114A alone, GLUT-1 with H1.sub.RBD or GLUT-1 with H1.sub.RBD Y114A expression vectors. Immunoprecipitation was performed using anti-rabbit-Pc beads and probed with an anti-HA antibody. Total cell extracts were blotted using an anti-rabbit Fc or an anti-HA antibody.

[0273] FIG. 5 GLUT-1 is an entry receptor for HTLV. Infections titer of MLV particles pseudotypes with HTLV-2 or A-MLV envelopes on 293T cells following transfection of an irrelevant or interfering H2.sub.RBD expression vectors alone or in addition to GLUT-1, GLUT-3 or Pit2 expression vectors.

[0274] FIG. 6 represents a schematic diagram of the HTLV-1 envelope glycoprotein (Env). Mature Env is constituted of two subunits formed after cleavage of the amino terminal signal peptide (SP) and cleavage of the Env polyprotein precursor into the extracellular SU and the membrane-anchored TM. SU comprises three distinct subdomains: an amino terminal receptor-binding domain (RBD), a central proline-rich region (PRR) and a carboxy terminal domain (C-term).

EXAMPLES

[0275] The present invention is further illustrated by the following examples.

Example 1

HTLV Envelopes Alter Lactate Metabolism

[0276] Cell proliferation in standard culture media is accompanied by acidification of the milieu that translates into a color change from red to yellow tones in the presence of the phenol-red pH indicator. Upon transfection of either highly syncytial HTLV-1 and HTLV-2 envelopes, or a non-syncytial chimeric envelope that harbors the HTLV-1 RBD in a MLV Env backbone (H.sub.183FEnv), culture medium did not readily acidify, and harbored red tones for several days post-transfection (FIG. 1a). Moreover, expression of truncated soluble HTLV RBD proteins fused with either GFP, -HA, or -rFc tags also inhibited medium acidification. In contrast, no envelope construct that lacked HTLV RBD, including different MLV group envelopes, feline, porcine, lentiviral and Jaagsiekte retroviral Envs, as well as VSV-G and Ebola glycoproteins, had this effect. The lack of acidification associated with HTLV-1 or HTLV-2 Env expression was not an indirect consequence of their syncytial activity, since (i) medium acidification was observed in cells expressing a syncytial amphotropic-MLV Env (A-MLV devoid of the R peptide) (FIG. 1a) and (ii) medium acidification was blocked when HTLV Env was expressed in cells that are resistant to HTLV-Env mediated syncytia formation (NIH3T3 TK.sup.− cells)[Kim, 2003].

[0277] Decrease of pH in cell culture is primarily due to extracellular accumulation of lactate [Warburg, 1956]. Lactate is the major byproduct of anaerobic glycolysis in vitro and its excretion is mediated by an H+/lactate symporter [Halestrap, 1999], We monitored lactate content in culture supernatants following transfection of various retroviral envelopes and RBD. Lactate accumulation was consistently 3-fold lower in H.sub.183FEnv- and HTLV RBD-transfected cells than in control- or MLV Env-transfected cells (FIG. 1b). This decrease in extracellular glucose and fructose accumulation after HTLV RBD transfection was DNA dose-dependent. Moreover, we found that the decrease in glucose and fructose accumulation following transfection of HTLV RBD was apparent as early as 4 hours after the addition of fresh media (FIG. 1e).

Example 2

Receptor Binding and Lactate Metabolism

[0278] To examine whether a direct relationship exists between binding of the HTLV envelope receptor and diminished extracellular acidification and lactate accumulation, we attempted to generate HTLV-1 RBD (H1.sub.RBD) mutants with impaired receptor binding capacities. To this end, mutations resulting in single alanine substitutions were introduced at two different positions in H1.sub.RBD, D106 and Y114 which are highly conserved among primate T-lymphotropic viruses. Although both D106A and Y114A RBD mutants were expressed and secreted as efficiently as the wild-type H1.sub.RBD (FIG. 3a), they exhibited significantly reduced (D106A) or non-detectable (Y114A) binding to the HTLV receptor as detected by FACS analysis (FIG. 3b). Moreover, perturbations in lactate metabolism correlated with binding to the HTLV receptor: lactate accumulation was not reduced in cells expressing the non-binding Y114A RBD mutant and was minimally reduced in cells harboring the D106 RBD (FIG. 3c). Similar results were obtained with H2.sub.RBD harboring the same allelic mutations. These data favor a direct association between lactate-related metabolic alterations and HTLV Env receptor binding.

[0279] Extracellular lactate accumulates in cell cultures following its transport across cellular membranes by the MCT1 monocarboxylate transporter [Garcia, 1994]. Because HTLV and MLV share a common organization of the extracellular envelope [Kim, 2000]and the receptors for MLV Env are multispanning metabolite transporters [Overbaugh, 2001], we assessed whether the HTLV RBD bound to MCT1. Moreover, similar to our previous data concerning expression of the HTLV receptor on T cells [Manel, 2003], expression of MCT1 chaperone CD147 [Kirk, 2000] increases during T cell activation [Kasinrerk, 1992]. However, separate and combined overexpression of MCT1 and CD147 did not result in increased H1.sub.RBD binding, arguing against a role for these molecules as receptors for HTLV Env.

Example 3

HTLV Receptor and Glucose Metabolism

[0280] In addition to a decrease in extracellular lactate accumulation, expression of the HTLV RBD also led to decreased intracellular lactate content, indicative of metabolic alterations upstream of lactate transport. In cell cultures, lactate accumulation results from the degradation of glucose during anaerobic glycolysis. Therefore, we assessed whether the decreased accumulation of lactate observed upon expression of HTLV RBD was linked to glucose metabolism. We measured glucose consumption as normalized to cellular protein content. Glucose consumption of cells expressing an HTLV RBD within the context of the H.sub.183FEnv entire envelope or the H1.sub.RBD was significantly decreased as compared to control cells (FIG. 1b) and this defect was detectable as early as 8 hours post transfection. To determine if this decrease in glucose consumption corresponded to a decrease in glucose transport across cellular membrane, we measured 2-deoxyglucose and fructose uptake in control cells and cells expressing HTLV RBD (FIG. 1c). We observed that expression of either HTLV-1 or HTLV-2 RBD induced an approximatively 4-fold decrease in 2-deoxyglucose uptake, while A-MLV RBD had only a minor effect. Inhibitors of glucose uptake, cytochalasin B and phloterin, also inhibited glucose uptake. These results were also true for 3-O-methylglucose transport. Fructose uptake in the same cells was not altered by the presence of HTLV-1 nor HTLV-2 RBD however A-MLV RBD induced a slight decreased. We next evaluated the effect of glucose deprivation on the availability of the HTLV receptor in both adherent human 293T cells and suspension Jurkat T cells. After overnight culture of cells in the absence of glucose, binding of H1.sub.RBD was consistently increased by 2-fold in both cell types (FIG. 1d). This effect of glucose deprivation was specific to HTLV as amphotropic MLV RBD (A.sub.RBD) binding was only marginally affected (FIG. 1d). This phenomenon is reminiscent of a general metabolite transport feedback loop, whereby transporter availability at the cell surface increases upon substrate starvation [Martineau, 1972].

Example 4

HTLV Envelopes Bind Glucose Transporter-1

[0281] A simple model whereby the HTLV envelope inhibits glucose consumption via direct binding to a glucose transporter can explain the metabolic effects described above. Upon evaluation of the different glucose transporter candidates, GLUT-1 appears to be the only one encompassing all the known properties of the HTLV receptor. Indeed, GLUT-1 expression is increased upon glucose deprivation and is transports glucose in all vertebrate cells [Mueckler, 1985], while fructose is transported by GLUT-5. Furthermore, GLUT-1 is not expressed on resting primary T cells and its expression is induced upon T cell activation [Rathmell, 2000; Chakrabarti, 1994] with kinetics that are strikingly similar to what we have reported for the HTLV receptor [Manel, 2003]. Since human but not murine erythrocytes have been described to be the cells exhibiting the highest concentration of GLUT-1 [Mueckler, 1994], we evaluated HTLV receptor availability on freshly isolated red blood cells. Binding of H1.sub.RBD on human erythrocytes was strikingly efficient, reaching levels higher than those observed on any other tested cell type, whereas A.sub.RBD binding to erythrocytes was minimal (FIG. 2a). On murine erythrocytes however, no significant H1.sub.RBD binding could be detected, despite a similar A.sub.RBD binding on murine and human erythrocytes. Furthermore, primary human hepatocytes do not express GLUT-1. Accordingly, we were unable to detect H1.sub.RBD binding to human primary hepatocytes, while A.sub.RBD binding could be readily detected.

[0282] In order to directly test the ability of HTLV envelopes to bind GLUT-1, we derived a tagged GLUT-1 expression vector and overexpressed this protein in HeLa cells. Both H1.sub.RBD and H2.sub.RBD binding was dramatically increased upon GLUT-1 overexpression (FIG. 4a). This interaction was specific as the HTLV-2 binding-defective mutant, D102A, as well as its HTLV-1 counterpart, D106A, did not bind GLUT-1 (FIG. 4a). Furthermore, H1.sub.RBD and H2.sub.RBD binding remained at background levels upon overexpression of the amphotropic MLV envelope receptor, the inorganic phosphate transporter PiT2 [Miller, 1994]. Conversely, binding of A.sub.RBD was not increased after GLUT-1 overexpression but as expected, this interaction was increased upon transfection of PiT2 (FIG. 4b). GLUT-3 is the closest isoform to GLUT-1, and transports glucose with kinetics similar to that of GLUT-1. Thus, we derived a tagged. GLUT-3 expression vector. Albeit similar overexpression levels of GLUT-1 and GLUT-3 in 293T cells, GLUT-3 did not induce any increase in H1.sub.RBD binding (FIG. 4b), suggesting that increase H1.sub.RBD binding in cells overexpressing GLUT-1 is not an indirect consequence of increased glucose uptake. To determine if GLUT-1 transfected cells were directly responsible for the observed increased in H1.sub.RBD binding, we derived fluorescent tagged GLUT-1 and GLUT-3 to uniquevocally identity GLUT-overexpressing cells in the course of our FACS analysis. In this context, only cells overexpressing GLUT-1-DsRed2 displayed a significant increase in H1.sub.RBD binding, while overexpressing GLUT-3-DsRed2 had no effect on H1.sub.RBD binding. Consequently, we tested if HTLV glycoproteins directly interact with GLUT-1 proteins. To this end, we evaluated the ability of H1.sub.RBD to immunoprecipitate GLUT-1. As shown on FIG. 4e, GLUT-1 could be readily detected upon immunoprecipitation with anti-rabbit-Fc-beads when it was co-expressed with H1.sub.RBD, but could not be detected when expressed alone or with the H1.sub.RBD Y114A mutant. Moreover, a GFP-tagged HTLV-2 RBD colocalized with GLUT-1 but not with PiT2 as assessed by fluorescence microscopy. Therefore, the GLUT-1 glucose transporter is an essential component of the HTLV envelope receptor.

[0283] Interaction of GLUT-1 with its ligand cytochalasin B inhibits glucose transport [Kasahara, 1977]. Since we showed that binding of HTLV envelopes to GLUT-1 inhibits glucose consumption and uptake, we tested whether cytochalasin B would abrogate HTLV RBD binding. Indeed, cytochalasin B treatment of Jurkat T cells dramatically inhibited binding of H1.sub.RBD, whereas binding of A.sub.RBD was not affected (FIG. 5). Thus, GLUT-1 directed glucose transport as well as binding of HTLV envelopes to GLUT-1 are similarly inhibited by the cytochalasin B ligand. Altogether, these data demonstrate that GLUT-1 is a receptor for HTLV envelopes.

[0284] Viral receptor permits entry and thus infection. No cellular system currently exists that lacks GLUT-1 expression. Thus, we developed a system in which HTLV infection is specifically inhibited at the level of envelope-receptor interaction. In this system, over-expression of HTLV-2 RBD interferes with infecting incoming HTLV particles and specifically decreases HTLV titers by at least 2 logs, while no effect is detected on control A-MLV titers. To determine if GLUT-1 is an entry receptor for HTLV, we overexpressed GLUT-1, GLUT-3 or Pit2 in addition to the interfering H2.sub.RBD. While Pit2 and GLUT-3 had no effect on HTLV titers, GLUT-1 completely alleviated the interference to infection induced by H2.sub.RBD (FIG. 5). Interestingly, both GLUT-1 and GLUT-3, but not Pit2, alleviated the alteration of glucose metabolism induced by the HTLV RBD. Thus, GLUT-1 is an entry receptor for HTLV.

Discussion

[0285] Here we show that HTLV-1 and -2 envelopes interact with GLUT-1 through their receptor binding domains. This interaction strongly inhibits glucose consumption and glucose uptake, leading to decreased lactate production and a block in extracellular milieu acidification. Mutations that specifically altered receptor binding of both HTLV-1 and 2 envelopes released the block in glucose consumption, indicative of a direct correlation between receptor binding determinants in the HTLV envelopes and glucose transport. Glucose starvation was rapidly followed by increased binding of HTLV envelopes, highlighting a nutrient-sensing negative feedback loop between glucose availability and cell surface HTLV receptor expression. Further evidence converged to identify GLUT-1 as the receptor, including increased binding of HTLV RBD upon overexpression of GLUT-1 but not GLUT-3, immunoprecipitation of GLUT-1 by H1.sub.RBD but not the receptor-binding mutant H1.sub.RBD Y114A, uppermost binding of HTLV RBD on human erythrocytes, where GLUT-1 is the major glucose transporter isoform, and no binding of HTLV RBD on human primary hepatocytes and murine erythrocytes, where GLUT-1 is minimally expressed. Finally, GLUT-1 could specifically alleviate interference to infection induced by HTLV RBD. GLUT-1 fits all other known properties of the HTLV receptor. Indeed, as previously demonstrated for the HTLV receptor [Manel, 2003], GLUT-1, but not the GLUT 2-4 isoforms, is not expressed on resting T lymphocytes [Chakrabarti, 1994; Korgun, 2002] and is induced upon immunological [Frauwirth, 2002; Yu, 2003] or pharmacological [Chakrabarti, 1994] activation. Moreover, GLUT-1 orthologues are highly conserved among vertebrates, but are highly divergent between vertebrates and insects [Escher, 1999].

[0286] GLUT-1 is thus a new member of the multimembrane spanning metabolite transporters that serve as receptors for retroviral envelopes. Interestingly, until now, all envelopes that recognize these receptors have been encoded by retroviruses that have a so-called simple genetic organization, such as MLV, feline leukemia viruses, porcine endogenous retrovirus and the gibbon ape leukemia virus [Overbaugh, 2001], whereas HTLV belongs to the so-called complex retroviruses which code for several additional regulatory proteins. However, we have shown that in contrast to the wide phylogenetic divergence of their genomic RNA, the envelopes of HTLV and MLV share a similar modular organization with some highly conserved amino acid motifs in their respective receptor binding domains [Kim, 2000].

[0287] Cell-to-cell contact appears to be required for HTLV transmission, and the cytoskeleton appears to play a major role in this process [Igakura, 2003]. Indeed, we observed that the HTLV receptor, despite pancellular expression, is specifically concentrated to mobile membrane regions and cell-to-cell contact areas. It should therefore be expected that the HTLV envelope receptor is associated to the cytoskeleton. Importantly, a cytoplasmic-binding partner of GLUT-1, GLUT1CBP, which encodes a PDZ domain, has been reported to link GLUT-1 to the cytoskeleton [Bunn, 1999]. It will therefore be interesting to evaluate the respective roles of the HTLV envelope, its cytoskeleton-associated cellular partners, such as GLUT-1, GLUT1CBP and their immediate interacting cell components.

[0288] Because expression of the HTLV receptor is induced upon glucose starvation, transmission of HTLV may be more efficient in cells that are locally starved for glucose, such as lymphocytes in lymph nodes [Yu, 2003]. Furthermore, the ability of circulating erythrocytes to dock HTLV, as shown here, might provide a means to distribute HTLV to such tissues.

[0289] The identification of GLUT-1 as a receptor for HTLV envelopes provides additional clues as to the ubiquitous in vitro expression of the receptor on cell lines and the paradoxical restriction of HTLV tropism to T lymphocytes in vivo. Rapid and dramatic metabolic alterations associated with the blockade of glucose consumption are likely to take place upon expression of the HTLV envelope in vivo, early after infection. Therefore, we propose that in vivo, HTLV infection initially spreads with a large tropism, however early after infection the vast majority of cells that are highly dependent on GLUT-1 activity are rapidly eliminated. In contrast, resting T lymphocytes that have an extremely low metabolic rate and as such are much less dependent on glucose uptake, can tolerate this effect and are therefore maintained in vivo. Furthermore, local imbalances in the access to glucose following HTLV infection may lead to specific physiological alterations [Akaoka, 2001]. In this regard, it will be of interest to study the potential relationship between HTLV-associated neuropathologies and the specific dependence of neurons on GLUT-1 mediated glucose consumption [Siegel, 1998].

Materials and Methods

[0290] Cell culture, 293T human embryonic kidney and HeLa cervical carcinoma cells were grown in Dulbecco's modified Eagle medium (DMEM) with high glucose (4.5 g/l) and Jurkat T-cells were grown in RPMI supplemented with 10% fetal bovine serum (FBS) at 37° C. in a 5% CO2-95% air atmosphere. For glucose starvation experiments, cells were grown in either glucose-free DMEM (Life Technologies) or glucose-free RPMI (Dutscher) with 10% dialyzed FBS (Life Technologies) and glucose (1 g/l) was supplemented when indicated.

[0291] Expression vectors. Full length envelope expression vectors for HTLV-1 (pCEL/2[Denesvre, 1995]) and Friend ecotropic MLV (pCEL/F [Denesvre, 1995]), have been previously described. For the HTLV-2 envelope, a fragment from pHTE2 [Rosenberg, 1998] encompassing the tax, rex and env genes and the 3′ LTR was inserted in the pCSI [Battini, 1999] vector (pCSIX.H2). Full length envelope expression vectors for amphotropic MLV (pCSI.A), or devoid of its R peptide (pCSI.AΔR), and H.sub.183FEnv that contains the N-terminal 183 amino acids of the HTLV-1 receptor-binding domain in the F-MLV envelope background, as well as truncated envelope expression vectors, derived from pCSI and encoding either of the first 215 residues of HTLV-1 SU (H1.sub.RBD), the first 178 residues of HTLV2-SU (H2.sub.RBD) or the first 397 residues of the amphotropic marine leukemia virus (MLV) SU (A.sub.RBD), fused to a C-terminal rabbit IgG Fc tag (rFc) or to EGFP (H2 All point mutations introduced in HTLV-1 and -2 RBD constructs were generated using the quickchange site-directed mutagenesis method and mutations were verified by sequencing. Human Glut-1 and Glut-3 cDNA were amplified by PCR from the pLib HeLa cDNA library (Clontech), and inserted into pCHIX, a modified version of the pCSI vector that contains a cassette comprising a factor Xa cleavage site, two copies of the hemagglutinin (HA) tag, and a histidine tag. The resulting construct (pCHIX.hGLUT1) encodes a GLUT-1 protein with a HA-His tag at the C-terminal end. GLUT-1 and GLUT-3 were also inserted in a modified pCSI vector containing a DsRed2 C-terminal tag. Similarly, human CD147 was amplified from 293T total RNA by RT-PCR and inserted into the pCHIX backbone in frame with the HA-His tag (pCHIX.hCD147).

[0292] Envelope expression and metabolic measurements. 293T cells were transfected with the various envelope expression vectors using a modified version of the calcium phosphate method. After an overnight transfection, cells were washed in phosphate-buffered saline (PBS) and fresh medium was added. Media were harvested at the indicated time points, filtered through a 0.45 μm pore-size filter, and lactate and glucose were measured with enzymatic diagnostic kits (Sigma). Values were normalized to cellular protein content using the Bradford assay (Sigma) after solubilization of cells in lysis buffer (50 mM Tris-HCl pH 8.0, 150 mM NaCl, 0.1% sodium dodecyl sulfate, 1.0% Nonidet P-40, 0.5% deoxycholate) and clarification by centrifugation,

[0293] Assay of hexose uptake. 2-deoxy-D[1-.sup.3H]glucose, D[U-.sup.14C] fructose and 3-O-[.sup.14C]methyl-D-glucose were obtained from Amersham. Hexose uptake assay were adapted from Harrison et al. 1991). After transfection, approximatively 250,000 were seeded/well in 24-well plates. The next day, cells were washed two times in PBS, incubated in serum-free DMEM, washed one time in serum-free glucose-free DMEM, and incubated for 20′ in 500 μl serum-free glucose-free DMEM modulo inhibitors (20 μM cytochalasin B, 300 μM phloretin; SIGMA). Uptake was initiated by adding labeled hexoses to a final concentration of 0.1 mM (2 μCi/ml for 2-2-deoxy-D[1-.sup.3H]glucose and 0.2 μCi/ml for D[U-.sup.14C] fructose and 3-O-[.sup.14C]methyl-D-glucose) and cells were incubated for 5′ additional minutes. Cells were then resuspended in 500 μl cold serum-free glucose-free DMEM, wash one time in serum-free glucose-free DMEM, and solubilized in 400 μl of 0.1% SDS. 3 μl was used for Bradford normalization, while the rest was used for detection of either .sup.3H or .sup.14C by liquid scintillation in a Beckman counter.

[0294] Western blots. Culture media (10 μl) from 293T cells expressing wild type or mutant HTLV-1 RBDs, and/or GLUT-1 for GLUT-3 expression vector. were subjected to electrophoresis on SDS-15% acrylamide gels, transferred onto nitrocellulose (Protran; Schleicher & Schuell), blocked in PBS containing 5% powdered milk and 0.5% Tween 20, probed with either a 1:5000 dilution of horseradish peroxidase-conjugated anti-rabbit immunoglobulin or 1:2000 dilution of anti-HA 12CA5 (Roche) monoclonal antibody followed by a 1:5000 dilution of horseradish peroxidase-conjugated anti-mouse immunoglobulin, and visualized using an enhanced chemiluminescence kit (Amersham).

[0295] Binding assays. Binding assays were carried out as previously described [Manel, 2003]. Briefly, 5×10.sup.5 cells (293T, HeLa, Jurkat or freshly isolated human erythrocytes) were incubated with 500 μl of H2.sub.RBD, H2.sub.RBD or A.sub.RBD supernatants for 30 min at 37° C., washed with PBA (1% BSA, 0.1% sodium azide in PBS), and incubated with a sheep anti-rabbit IgG antibody conjugated to fluorescein isothiocyanate (Sigma). When indicated, cytochalasin. B (20 μM; Sigma) was added to cells for 1 hour prior to binding analyses. Binding was analyzed on a FACSCalibur (Becton Dickinson) and data analysis was performed using CellQuest (Becton Dickinson) and WinMDI (Scripps) softwares.

[0296] Infections. 293T cells were transfected in 6-wells plate, and one day after transfection, medium was replaced by high glucose DMEM supplemented with fructose (5 g/l) and non-essential amino acids. The next day, infection was initiated by adding supernatants containing MLV particles pseudotyped with either HTLV-2 or A-MLV envelopes. The following day, fresh medium was added, and 24 hours later cells were fixed and stained for alkaline phosphatase activity and dark focus of infection were counted. Viral particles were obtained by transfecting 293T cells with pLAPSN, pGagPol and either pCSIX.H2 or pCSI.A, and harvesting the 0.45 μm-filtered supernatants 24 hours later.