Regulation of CD4 Expression via Recycling by HRES-1/RAB4 Controls Susceptibility to HIV Infection*

A novel 2986-base transcript encoded by the antisense strand of the HRES-1 human endogenous retrovirus was isolated from peripheral blood lymphocytes. This transcript codes for a 218-amino acid protein, termed HRES-1/Rab4, based on homology to the Rab4 family of small GTPases. Antibody 13407 raised against recombinant HRES-1/Rab4 detected a native protein of identical molecular weight in human T cells. HRES-1 nucleotides 2151-1606, located upstream of HRES-1/Rab4 exon 1, have promoter activity when oriented in the direction of HRES-1/Rab4 transcription. The human immunodeficiency virus, type 1 (HIV-1), tat gene stimulates transcriptional activity of the HRES-1/Rab4 promoter via trans-activation of the HRES-1 long terminal repeat. Transfection of HIV-1 tat into HeLa cells or infection of H9 and Jurkat cells by HIV-1 increased HRES-1/Rab4 protein levels. Overexpression of HRES-1/Rab4 in Jurkat cells abrogated HIV infection, gag p24 production, and apoptosis, whereas dominant-negative HRES-1/Rab4S27N had the opposite effects. HRES-1/Rab4 inhibited surface expression of CD4 and targeted it for lysosomal degradation. HRES-1/Rab4S27N enhanced surface expression, recycling, and total cellular CD4 content. Infection by HIV elicited a coordinate down-regulation of CD4 and up-regulation of HRES-1/Rab4 in PBL. Moreover, overexpression of HRES-1/Rab4 reduced CD4 expression on peripheral blood CD4+ T cells. Stimulation by HIV-1 of HRES-1/Rab4 expression and its regulation of CD4 recycling reveal novel coordinate interactions between an infectious retrovirus and the human genome.

Endogenous retroviruses (ERV) 3 belong to the larger family of retrotransposable elements that make up as much as 40% of the human genome (1). Human ERV (HERV) have the basic structures of the integrated proviral form of infectious retroviruses with long terminal repeats (LTRs) flanking sequences homologous to gag, pol, and env genes (2). Human ERV have generally been found to be defective proviruses having accumulated deletions or stop codons in gag, pol, and/or env open reading frames (3). Human ERV are commonly designated as HERV followed by a single letter amino acid code corresponding to a tRNA. The 3Ј terminus of tRNA is predicted to initiate reverse transcription by annealing to an 18-nucleotide-long primerbinding site (PBS) at the 5Ј-LTR. ERV copy numbers vary from one to several hundred per haploid genome (4).
ERV may represent a key molecular link between the host genome and infectious viral particles. They constitute a large reservoir of viral genes that may be activated by mutations caused by radiation or chemicals or recombination with exogenous retroviruses. Although exogenous retroviruses are infectious, with a replication cycle that requires integration of proviral DNA into host cell DNA, ERV are transmitted genetically in a classical Mendelian fashion through the germ line as proviral DNA. Expression of ERV can influence the outcome of infections in different ways that are both beneficial and detrimental to the host (2). These include provision of genes for recombination with exogenous viruses, interference with virion assembly, modulation of immune responses to exogenous viruses, and blocking cellular receptors for viral entry (5).
Human immunodeficiency virus, type 1 (HIV-1), uses two receptors for cellular attachment and viral entry. Initial viral attachment occurs through the binding of the envelope protein gp120 to the CD4 molecule expressed on the surface of T lymphocytes and macrophages. Viral binding to CD4 is necessary but insufficient to mediate viral entry. Interaction between CD4 and gp120 increases the affinity of virions for coreceptor molecules CXCR4 and CCR5. Genetic polymorphisms or deletions within CCR5 diminish or abrogate viral binding to the receptor, which leads to a lower susceptibility to infection and slower disease progression in persons carrying these mutations (6). CXCR4-using viruses are generally more pathogenic, via depletion of CD4 T cells, than are CCR5-using viruses. CD4 appears to play a role in HIV entry distinct from merely serving as the attachment protein for the virus. CD4 undergoes endocytosis following T cell activation via the activation of protein kinase C and subsequent phosphorylation of CD4 (7,8). HIV binding was also reported to induce phosphorylation of CD4 via a protein kinase C-dependent pathway; however, internalization of HIV does not require endocytosis of CD4 (9,10).
In this study, we investigated potential interactions between a newly identified gene product of the human T-cell leukemia virus type I-related endogenous retroviral sequence, HRES-1, and HIV-1. HRES-1 was previously isolated based on homology to the long terminal repeat and gag regions of human T cell leukemia virus type I (11). Hybridization analysis with genomic DNA samples of selected phylogenetic stages revealed that HRES-1 was confined to the primate lineage (11). HRES-1 is a single copy sequence in the haploid genome that has been mapped to 1q42 at the long arm of chromosome 1 (12). A 684-bp flanking region 5Ј to p28 contains a TATA box, a polyadenylation signal, a potential histidyl tRNA primer-binding site (PBS), and characteristic inverted repeat sequences at locations that are typical of a retroviral LTR (2,11). HRES-1 is one of the few human ERV that remain transcriptionally active (3,11).
By utilizing HRES-1-derived probes, a novel 2986-base-long transcript encoded by the antisense strand of the HRES-1 locus was isolated from human peripheral blood lymphocytes (PBL). The sequence of this cDNA showed considerable homology to the 735-base-long Rab4a gene and is thus termed HRES-1/ Rab4. Antibody raised against recombinant HRES-1/Rab4 detected a native protein of identical molecular weight in PBL and Jurkat and H9 T cell lines. The first coding exon of HRES-1/Rab4 is embedded in the HRES-1 endogenous retroviral sequence. HRES-1 nucleotides 2151-1606, located upstream from HRES-1/Rab4 exon 1, showed strong promoter activity when oriented in the direction of HRES-1/Rab4 transcription. The tat gene of HIV-1 stimulates transcriptional activity of the HRES-1/Rab4 promoter via trans-activation of the HRES-1 LTR and increases HRES-1/Rab4 protein levels. In turn, HRES-1/Rab4 regulates recycling and surface expression of CD4 and thus controls susceptibility to infection by HIV-1.

Screening of Expression Library and
Cloning and Sequencing of cDNA-A human lymphocyte cDNA library prepared in gt11 phage (Stratagene, La Jolla, CA) was screened with HRES-1 probes under high stringency conditions, as described earlier (14). Positive clones were transferred into the EcoRI site of Bluescript KSϩ plasmid (Stratagene) and sequenced in both strands. Nucleotide sequence of the 2986-base-long HRES-1/Rab4 cDNA was submitted to GenBank TM (accession number AY585832).
Prokaryotic Expression of Recombinant Protein-Full-length HRES-1/Rab4 protein was expressed as a fusion protein with glutathione S-transferase (GST), as described earlier (15). BamHI and XhoI sites were generated by PCR-mediated mutagenesis immediately 5Ј of the first methionine codon and 3Ј of the stop codon of HRES1/Rab4 and cloned into BamHI-and XhoIcleaved pGEX-6P1 plasmid vector (Amersham Biosciences). Optimum stimulation of expression of the recombinant fusion protein was obtained with 1 mM isopropyl thio-␤-galactoside after 2 h. HRES1/ Rab4-GST fusion protein was affinity-purified through binding of GST to glutathione-coated agarose beads (Sigma). HRES-1/Rab4 protein was cleaved from GST by precision protease (Amersham Biosciences) and was separated from the agarose bead-bound GST by centrifugation.

HRES-1/Rab4 Regulates CD4 Expression
Jur-TA cells are neomycin (G418)-resistant Jurkat cells transfected with rtTA-containing pUHD172-1neo and thus produce the reverse trans-acting factor (rtTA) capable of interacting with promoters harboring tetracycline operator sequences only in the presence of tetracycline analogs. Jur-TA cells were electroporated with GFP-producing pTR5-DC/GFP*TK/hygro control (construct 4480), wild-type (construct 6678), and dominant-negative mutant HRES-1/Rab4 S27N -containing vector (construct 9035); and doubly transfected cells were selected in 400 g/ml G418 and 200 g/ml hygromycin. Percentage of GFP-positive cells in an aliquot of the transfected cultures was periodically checked by flow cytometry. After 24 h of incubation with 1 g/ml doxycycline, which did not affect cell viability, the most brightly fluorescent 1% of cells was sorted using a flow cytometer and subsequently maintained in the presence of 200 g/ml G418 and 100 g/ml hygromycin. Bulk-sorted aliquots of 4480, 6678, and 9035 construct-transfected cells were utilized in all experiments. Transfected cells were not cloned to avoid artifacts related to variations in sites of vector integration. Control and HIV-1 tat gene-transfected HeLa and Jurkat cells (18) were obtained from the NIH AIDS Research and Reference Program.
Site-directed Mutagenesis-Dominant-negative HRES-1/Rab4 was created by replacing serine 27 with asparagine (S27N) using the Quick-Change site-directed mutagenesis kit (Stratagene, La Jolla, California). Briefly, 25 ng of the wild-type HRES-1/Rab4 cDNA plasmid (template) was incubated with 125 ng of sense (5Ј-GCAGGAACTGGCAAAAAT-TGCCTTACTTCATCAG-3Ј) and antisense primers (5Ј-CTGATGAA-GTAAGCAATTTTTGCCAGTTCCTGC-3Ј with mutagenic residues underlined) and dNTPs and subjected to 18 PCR cycles with Pfu Turbo DNA polymerase with denaturation at 95°C for 1 min, annealing at 55°C for 1 min, and extension at 68°C for 12 min. DpnI was used to digest the parental supercoiled doublestranded methylated DNA for 1 h at 37°C. Transformations were performed in Escherichia coli XL-1 Blue cells using DpnI-treated DNA. Mutagenesis was confirmed by sequencing of the resultant plasmids.
Transient Transfections and Reporter Gene Assays-HeLa cells were transfected with 1.6 g of pBLCAT3-based HRES-1 promoter chloramphenicol acetyltransferase (CAT) reporter gene constructs (19) at 80% confluency in 9.5-cm 2 wells using 20 l of PLUS reagent and 4.8 l of Lipofectamine reagent (Invitrogen). To normalize transfection efficiency, each plasmid was cotransfected with 1.6 g of the pRSV␤-gal plasmid (␤-galactosidase reporter gene driven by the Rous sarcoma virus promoter) (20). The HRES-1/Rab4-containing plasmids did not affect promoter activity of pRSV␤-gal, as determined by the amount of transfected cell lysate used for CAT assays. For each experiment, the promoterless pBLCAT3 vector was used as a negative control, and pRSV-CAT was used as positive control. After 4 h of exposure to the DNA/Lipofectamine complex in serum-and antibiotic-free media, the DNA complex was removed, and cells were further cultured for an additional 36 h in complete growth media. Cells were harvested in 150 l of 250 mM Tris, pH 7.8, and solubilized by three rounds of freezing and thawing. For the ␤-galactosidase assay, 30 l of lysate was incubated with 270 l of reaction mixture (1 mM MgCl 2 , 50 mM ␤-mercaptoethanol, 3 mM o-nitrophenyl ␤-D-galactopyranosidase, and 0.1 M NaP i , pH 7.5, at 37°C) and terminated with addition of 500 l of 1 M Na 2 CO 3 once the reaction turned yellow. Absorbance values were measured at 420 nm and used  to calculate the quantity of each transfected cell lysate to be used in the CAT assay. Cell lysates were heated at 65°C for 10 min to inactivate acetylases, and CAT assays were performed at 37°C in 50 l of reaction mixture of 250 mM Tris, pH 7.8, normalized cell lysate, 0.4 mM acetyl-coenzyme A, 0.025 Ci of [ 14 C]chloramphenicol. Acetylated chloramphenicol was extracted with ethyl acetate, dried in a vacuum centrifuge, resuspended in ethyl acetate, spotted on a silica gel (Analtech, Newark, DE), and resolved in an equilibrated chromatography tank containing 19:1 chloroform/methanol. A 445 SI Phospho-rImager with ImageQuant software (Amersham Biosciences) was used to determine the ratio of acetylated [ 14 C]chloramphenicol. All assays were conducted within the range of linearity of CAT activities with respect to incubation time, based on the ␤-galactosidase assay (21).
Retroviral Proteins and Antibodies-Retroviral reagents were obtained from the AIDS Research and Reference Program, National Institutes of Health. Infectious stock of the strain HIV-1 IIIB was harvested from 24-h supernatants of freshly infected H9 cells (ATCC CRL-8543), and infectious titer was determined by an in situ infectivity (MAGI) assay (22). Supernatants with titers of 2.1 ϫ 10 5 infectious units (IU)/ml were filtered through a 0.45-m filter, concentrated by ultracentrifugation, and aliquots were stored at Ϫ70°C. Infections were performed with cell-free virus supernatants containing 100 ng of p24 core protein measured by an enzyme-linked immunosorbent assay following the manufacturer's recommendations (NEK-060, PerkinElmer Life Sciences), corresponding to 2.1 ϫ 10 5 IU per 5 ϫ 10 6 cells or a multiplicity of infection of 0.04. Thus, 1 ng of p24 core protein corresponded to 2000 infectious virions, in accordance with earlier findings (23). Virus infection was carried out in 1 ml of serum-free RPMI 1640 medium for 2 h in the presence of 10 g/ml Polybrene (Sigma). Subsequently, cells were washed in serum-free RPMI and resuspended in 10 ml of fresh RPMI 1640 medium containing 20% fetal calf serum, 100 units/ml penicillin, 100 g/ml streptomycin, 10 g/ml amphotericin B, and 2 mM L-glutamine. HIV-1 SF2 p25/245 gag contained the gag 24 protein (24). HIV-1/IIIB Gag4 contained the p17 C terminus, beginning at amino acid position 146, all of p24, and the p15 N terminus (Repligen, Cambridge, MA). To monitor production of viral proteins, HIV-1 gag p24-specific polyclonal sheep antibody (25) and tatspecific rabbit antibody 705 (AIDS Research and Reference Program, National Institutes of Health) were utilized.
Separation of CD4ϩ T Cells from Human Peripheral Blood-PBMC were separated from peripheral blood (26), and 10 7 cells were incubated with 17.5 l of anti-CD4 beads to isolate CD4ϩ T cells (catalog number 111.45, Dynal, Lake Success, NY). After stimulation with 5 g/ml concanavalin A and 200 units/ml interleukin-2 (Sigma) for 3 days, beads were removed by washing in PBS with 2% fetal calf serum over a Dynal magnet, and 10 7 CD4ϩ T cells (Ն98% pure) were infected with HIV-1, as described above.
Receptor Recycling-Internalization of surface receptors was induced by treatment with 100 nM phorbol 12,13-dibutyrate (PDBu) for 1 h at 37°C (7,8). Subsequently, cells were washed twice and kept at 4°C to assess internalization or returned to 37°C to allow receptor recycling in the absence or presence of doxycycline. After termination of recycling, cells were kept on ice until analysis of surface expression by flow cytometry.
Permeabilization of Cells for Detection of Intracellular Antigen by Flow Cytometry-10 6 Jurkat cells were centrifuged, resuspended in Hanks' balanced salt solution (HBSS, Cellgro) containing 4% paraformaldehyde, and incubated for 10 min at room temperature. After centrifugation, cells were resuspended in 500 l of HBSS with 0.1% saponin (28), 10 mM HEPES, pH 7.4, and 1% fetal calf serum and incubated for 10 min at room temperature. After centrifugation, cells were resuspended in 100 l of HBSS with 0.1% saponin, 10 mM HEPES, pH 7.4, 1% bovine serum albumin, and PE-conjugated IgG1 monoclonal antibody KC57 RD1 for detection of HIV-1 gag p24 (Beckman Instruments, Fullerton, CA) or isotype control antibody and incubated for 30 min at room temperature. Cells were washed in 900 l of HBSS with 0.1% saponin and 10 mM HEPES, pH 7.4, and resuspended in 500 l of HBSS with 1% paraformaldehyde and kept on ice up to 24 h prior to analysis. For detection of intracellular HRES-1/Rab4, cells were stained with primary rabbit antibodies SC312 (Santa Cruz Biotechnology, Santa Cruz, CA) or 13407, followed by washing and incubation of secondary PE-conjugated donkey anti-rabbit IgG.
Confocal Microscopy-For TFR staining, 10 6 cells were incubated for 30 min in serum-free RPMI medium at 37°C, washed once, and incubated in uptake medium (RPMI containing 20 mM HEPES, 0.5% bovine serum albumin, pH 7.4) with 50 g/ml Alexa 647-conjugated transferrin (Molecular Probes, Eugene, OR) for 30 min at 37°C, and then pipetted onto poly-L-lysine-coated (0.1 mg/ml poly-L-lysine; Sigma) coverslips for permeabilization and staining with Rab4 antibody. For CD4, CD5, fusin, and CD3⑀ staining, 10 6 cells were stained with Alexa Fluor 647-or PE/Cy5-conjugated antibodies on ice for 20 min in complete medium, and internalization was induced with 100 nM PDBu for 1 h at 37°C. Cells were then washed and kept on ice or allowed to recycle at 37°C. After internalization and recycling, cells were adhered to coverslips precoated with 0.1 mg/ml poly-L-lysine for 10 min at room temperature and fixed in 4% paraformaldehyde in Hanks' balanced salt solution (HBSS). Cells were permeabilized with 0.1% saponin, 0.01 M HEPES buffer, and 1% fetal bovine serum in HBSS and stained with 4 g/ml rabbit anti-Rab4a antibody in 0.1% saponin, 0.01 M HEPES buffer, and 1% bovine serum albumin in HBSS for 30 min. After three washes the cells were incubated in 12.5 g/ml Texas Red-conjugated donkey anti-rabbit secondary antibody (Jackson ImmunoResearch) for 30 min, washed four times, and the coverslips mounted onto glass slides in ProLong Gold (Molecular Probes). Imaging was performed on a Bio-Rad MRC-1024 ES confocal microscope equipped with a krypton/argon laser capable of delivering excitation at 488, 568, and 647 nm using the Bio-Rad Lasersharp 2000 software. All images were recorded with a 60ϫ 1.4 numerical aperture oil immersion objective with appropriate discriminatory filter sets. To minimize the noise and to keep a low photobleaching rate, we selected an acquisition time of 1 s per scan and averaged 10 scans to produce each 512 ϫ 512-pixel image.
The images registered from the same confocal plane for the green, red, and blue signals were saved and superimposed. For every pixel in the original image, the system plots a corresponding point related to the intensities of the green versus red, green versus blue, and red versus blue channels. Colocalization was quantified by correlation calculation in each pixel of paired images of Ն15 cells exhibiting green, red, and blue color fluorescence using the Simple PCI software version 5.3.0.1102 (Compix, Sewickly, PA).

HRES-1/Rab4 Regulates CD4 Expression
To test for expression of native HRES-1/Rab4, an antibody was raised in rabbit against full-length recombinant protein (rHRES-1/Rab4) expressed as a fusion protein with GST in E. coli. The affinity-purified and cleaved rHRES-1/Rab4 was used to raise rabbit antibody Ab 13407 (Fig. 3C). This antibody recognized the rHRES-1/Rab4 protein produced in E. coli or Jurkat cells and detected the native protein of identical molecular weight in both PBL and Jurkat cells. Antibody SC312, raised against a 20-amino acid C-terminal segment of Rab4a, recognized HRES-1/Rab4 and Rab4a of slightly lower molecular mass. The predicted molecular weight of HRES-1/Rab4 is 24,389, whereas that of Rab4a is 23,901. Because the HRES-1 is a single copy sequence in the haploid genome (12), the previously identified Rab4 may originate from another chromosomal locus or correspond to an alternative translation product of the polymorphic HRES-1/Rab4 genomic locus.

HRES-1/Rab4 Inhibits HIV-induced Gag p24
Production and Apoptosis-Because HRES-1/Rab4 expression is up-regulated by HIV-1 tat, its potential role in viral pathogenesis was investigated. Jurkat cells overexpressing HRES-1/Rab were generated using a doxycycline-inducible GFP-encoding bi-cistronic expression vector system (34). Jurkat cells containing pUHD172-1neo (Jur-TA) stably transfected with the pTR5-DC/GFP*TK/hygro vector alone (4480) produced GFP in the presence of 1 g/ml doxycycline and were utilized as negative control. Jur-TA cells stably transfected with the pTR5-DC/ HRES-1Rab4*GFP*TK/hygro vector (6678 cells) produced 97.06 Ϯ 3.2-fold ( p Ͻ 0.0001) increased amounts of HRES-1/ Rab4 in the presence of 1 g/ml doxycycline. However, HRES-1/Rab4 expression was also increased in 6678 Jur-Ta cells in the absence of doxycycline by 3.01 Ϯ 0.2-fold ( p Ͻ 0.0001). Representative Western blot of Jurkat cells overexpressing HRES-1/ Rab4 is shown in Fig. 5A. HRES-1/Rab4 profoundly reduced HIV gag p24 production detected by Western blot (Fig. 5B) and the percentage of HIV-1 infected cells determined by flow cytometry of intracellular gag p24 staining (Fig. 5C) and diminished HIV-induced apoptosis by 51% ( p ϭ 0.009) and 57% ( p ϭ 0.0057) in the absence or presence of doxycycline, respectively (Fig. 5D). Low levels of HIV gag p24 were detected by Western blot (Fig. 5B), but none by flow cytometry of 6678-transfected cells (Fig. 5C). To further substantiate specificity of HRES-1/ Rab4-induced changes on infection by HIV-1, Jurkat cells overexpressing a dominant-negative form of HRES-1/Rab4 were generated. Substitutions in the GTP-binding domains of all rab proteins and other members of the ras superfamily, such as S22N mutant Rab4 (35), were potent trans-dominant inhibitors of vesicular transport (36). Therefore, S27N substitution was created in the HRES-1/Rab4-producing pTR5-DC/GFP*TK/ hygro vector by site-directed mutagenesis. The mutated plasmid was transfected into Jur-TA cells. Following doxycycline induction, GFP-positive cells were sorted by flow cytometry. Expression of HRES-1/Rab4 S27N was monitored by Western blot analysis and automated densitometry with respect to ␤-actin internal control (Fig. 5A). In comparison to native HRES-1/ Rab4 protein levels of 4480 control cells, HRES-1/Rab4 S27N was expressed in 9035 cells at 5.05 Ϯ 0.6-fold elevated levels in the absence of doxycycline ( p Ͻ 0.0001) and 49.96 Ϯ 2.2-fold elevated levels in the presence of doxycycline ( p Ͻ 0.0001; based on four independent experiments). To avoid artifacts related to variations in sites of vector integration, transfected cells were not cloned, but bulk-sorted aliquots of cells overexpressing HRES-1/Rab4 or HRES-1/Rab4 S27N or GFP alone were utilized in all experiments. Following pretreatment with doxycycline for 24 h, each cell type was infected with HIV-1. HRES-1/Rab4 abrogated production of HIV-1 gag p24 as determined by Western blot analysis (Fig. 5B). Along the same line, HRES-1/ Rab4 reduced the percentage of HIV-1-infected cells determined by flow cytometry of intracellular gag p24 staining (Fig.  5C) and diminished HIV-induced apoptosis (Fig. 5D). By contrast, HRES-1/Rab4 S27N enhanced gag p24 production and HIV-induced apoptosis, and the latter was markedly enhanced in 9035 cells treated with doxycycline ( p ϭ 0.0003; Fig. 5D).

HRES-1/Rab4 Regulates Surface Expression of CD4 via
Recycling-Rab proteins belong to the family of small GTPases that regulate endosome recycling. Rab4a has been shown to regulate recycling of early endosomes carrying the TFR in epithelial cells (37) or GLUT4 in adipocytes (38). Therefore, we examined whether the impact of HRES-1/Rab4 on HIV infection was mediated via recycling and expression of surface receptors. TFR expression was not affected in the absence of doxycycline; however, it was reduced by HRES-1/Rab4 in the presence of doxycycline ( Fig. 6A; p Ͻ 0.0034). Similar results were obtained by using Alexa 647-conjugated transferrin and PE-Cy5-conjugated CD71 monoclonal antibody for detection of TFR (data not shown). Strikingly, expression of CD4 was reduced on the surface of 6678 cells as compared with 4480 cells, even in the absence of doxycycline (Ϫ62 Ϯ 2.9%, p Ͻ 0.0001). Addition of doxycycline further reduced expression of CD4 on 6678 cells (Ϫ72 Ϯ 0.9%, p ϭ 0.022). By contrast, surface expression of CD4 was enhanced on HRES-1/Rab4 S27N -producing 9035 cells (Fig. 6A). CD4 expression was not affected by increased production of GFP in control 4480 cells (Fig. 6A). Expression of HIV coreceptor fusin/CXCR4 (Fig. 6A) and CD45RO (data not shown) was not influenced by HRES-1/ Rab4. Thus, coordinate suppression by HRES-1/Rab4 and upregulation by HRES-1/Rab4 S27N of CD4 expression indicate a specific role for HRES-1/Rab4 in regulation of CD4 expression.
CD4 undergoes protein kinase C-mediated endocytosis following T cell activation (7). Thus, CD4 internalization was induced by treatment with 100 nM PDBu for 1 h at 37°C. Subsequently, PDBu was removed by washing twice, and cells were kept at 4°C to assess internalization or returned to 37°C to allow recycling. Recycling of TFR/CD71, CD4, fusin/CXCR4, CD45RO, and CD3⑀ to the cell surface was monitored in Jurkat cells overproducing wild-type HRES-1/Rab4 or HRES-1/ Rab4 S27N with respect to control cells. Base-line expression of TFR was reduced after induction of HRES-1/Rab4 by doxycycline for 24 h (Fig. 6A), whereas the rate of TFR recycling after serum deprivation was not affected by HRES-1/Rab4 or HRES-1/Rab4 S27N (Fig. 6B). Surface expression of CD4 was profoundly reduced in cells overexpressing HRES-1/Rab4, thus preventing further reduction by PDBu and evaluation of recy-cling (Fig. 6B). Although base-line expression of CD4 was markedly enhanced by HRES-1/Rab4 S27N , after internalization and recycling for 120 min, surface expression of CD4 exceeded base-line levels by 36% ( p ϭ 0.0117) in the presence of doxycycline (Fig. 6B, 9035D cells). After internalization and recycling, TFR levels remained below base-line levels by 25% on cells overproducing HRES-1/Rab4 S27N in the presence of doxycycline (Fig. 6B, 9035D cells). Thus, recycling of CD4, but not TFR, was selectively enhanced in cells overproducing HRES-1/Rab4 S27N .

HRES-1/Rab4 Regulates CD4 Expression
Coordinate Regulation of CD4 and HRES-1/Rab4 Expression in Peripheral Blood Lymphocytes-To investigate the effect of HIV-1 on expression of HRES-1/Rab4 in peripheral blood lymphocytes, PBMC and affinity-purified CD4 T cells were infected by HIV-1. Western blot analysis showed a coordinate down-regulation of CD4 and up-regulation of HRES-1/Rab4 in HIV-infected cells (Fig. 9A). Infection of CD4 T cells and PBMC with HRES-1/Rab4-producing AAV resulted in down-regulation of CD4 expression with respect to uninfected cells or cells infected with control AAV or HRES-1/Rab4 S27N -producing AAV (Fig. 9, B and C).

DISCUSSION
ERV constitute a large reservoir of viral genes that can influence the outcome of infections by exogenous retroviruses via provision of genes for recombination with proviral DNA, interference with virion assembly, blocking cellular receptors for viral entry, and modulation of immune responses to exogenous viruses (2,5). Although most human ERV have truncated and mutated proviral DNA, few of them are transcriptionally active (3). The present data provide evidence that both the sense and antisense strands of the human ERV HRES-1 can be transcribed in human cells. Bidirectional transcription has been documented at several genomic loci (40,41), including another ERV, HERV-H (6), and the 1q42 locus harboring HRES-1 (42). Although the sense strand of HRES-1 encodes a 28-kDa gag-like protein, HRES-1/p28, expressed in T lymphocytes and salivary gland epithelial cells (5,11,29), the antisense strand provides exon 1 and transcriptional regulatory elements of the HRES-1/ Rab4 protein. The tat gene of HIV-1 stimulates expression of HRES-1/Rab4 protein via trans-activation of the HRES-1 LTR. In turn, HRES-1/Rab4 modulates surface expression of CD4 and, thus, infection of T cells by HIV-1.
The present data identify HRES-1/Rab4 as a regulator of CD4 recycling. Rab proteins belong to the family of small GTPases that regulate receptor endosome recycling (43). In particular, Rab4a has been shown to influence recycling of the TFR in epithelial cells (37) or GLUT4 in adipocytes (38). Unlike fusin, CD5, and CD45RO, surface expression of CD4 and TFR was influenced by HRES-1/Rab4. Although HRES-1/Rab4 colocalized with both TFR and CD4, surface expression of TFR was only reduced in doxycycline-stimulated cells overproducing HRES-1/Rab4 up to100-fold. By contrast, CD4 levels were coordinately down-regulated by HRES-1/Rab4 and up-regulated by HRES-1/Rab4 S27N . CD4 levels were markedly diminished by 3-fold overexpression of HRES-1/Rab4. Similar increase of HRES-1/Rab4 expression was elicited by HIV infection or transfection of HIV-1-tat, suggesting that HIV-1 may utilize HRES-1/Rab4 to regulate CD4 expression. Along this line, production of HIV-1 gag p24 and apoptosis of HIV-infected cells were reduced by HRES-1/Rab4 and enhanced by HRES-1/Rab4 S27N . Surface expression or recycling of CXCR4/ fusin was not influenced by HRES-1/Rab4, suggesting that infection by HIV-1 was selectively modulated via CD4 expression. CD4 and TFR mRNA were not affected by HRES-1/Rab4 or HRES-1/Rab4 S27N (data not shown), suggesting that changes in surface expression were solely mediated on the protein level.
CD4 appears to play a role in HIV entry beyond merely serving as the attachment protein for the virus. CD4 undergoes endocytosis following T cell activation (7), and HIV entry may depend on the internalization of CD4. In turn, CD4 internalization is dependent on the activation of protein kinase C and subsequent phosphorylation of CD4 (8). Accordingly, treatment with PDBu, an activator of protein kinase C, markedly reduced surface expression of CD4. Protein kinase C-mediated down-regulation of CD4 involves altered endosomal sorting (44). CD4 is constitutively internalized into early endosomes and recycled to the cell surface. In the presence of phorbol ester, CD4 is diverted from the constitutive recycling early endosome pathway to the late endosome/lysosome pathway, and as a consequence, there is a reduction in the recycling of internalized CD4. Indeed, after removal of PDBu, CD4 expression increased. Re-appearance of CD4 on the cell surface was not influenced by inhibition of protein synthesis with cycloheximide (data not shown), suggesting that internalized CD4 was not synthesized de novo but recycled back to the cell surface. Overexpression of HRES-1/Rab4 markedly reduced surface expression of CD4 and targeted it for lysosomal degradation. CD4 levels suppressed by overexpression of HRES-1/Rab4 were normalized by lysosomal but not proteasomal inhibitors. In contrast, surface expression, recycling, and total cellular CD4 content was enhanced by HRES-1/Rab4 S27N . Thus, HRES-1/ Rab4 plays a dominant role in CD4 expression in T cells by regulating its traffic between the recycling and late endosome/ lysosome pathway.
The present data show that HIV stimulates expression of HRES-1/Rab4 which, in turn, abrogates recycling of CD4 to the cell surface. Indeed, profound down-regulation of CD4 expression on the surface of HIV-infected cells may be mediated, at least in part, via inhibition of CD4 recycling by HRES-1/Rab4. Previously, HIV-induced CD4 down-regulation has been attributed to nef-initiated internalization and retention (45), lysosomal degradation (46,47), and vpu-initiated degradation by the proteasome (48 -50). HIV nef also influences apoptosis signal processing, T-cell receptor expression (51), and formation of the immunological synapse (52). This study reveals a role for HRES-1/Rab4 in lysosomal degradation of CD4. HRES- 1/Rab4 belongs to the family of small GTPases that regulate receptor endosome recycling (43) and may be involved in Nefinduced lysosomal degradation of CD4 (46,47). Wild-type HRES-1/Rab4 inhibited whereas HRES-1/Rab4 S27N enhanced HIV infection and virion production, suggesting that regulation of HRES-1/Rab4 expression may play an active role in the life cycle of HIV-1. This notion is supported by coordinate up-regulation of HRES-1/Rab4 and down-regulation of CD4 expression in HIV-infected CD4 T cells and PBMC. HRES-1/Rab4 promoter activities and protein levels are increased in cells infected by HIV-1 or transfected by HIV-tat. In turn, enhanced expression of HRES-1/Rab4 may contribute to down-regulation of CD4 recycling to the cell surface, thus preventing reinfection by HIV-1, allowing for increased virion production, and protecting virus-infected cells against killing by cytotoxic T cells (53,54). Thus, stimulation of HRES-1/Rab4 expression by HIV-1 and regulation of HIV coreceptor CD4 recycling by HRES-1/Rab4 represent novel mechanisms of coordinate interaction between infectious viral particles and ERV of the human genome.