Suppression of Tetherin-Restricting Activity upon Human Immunodeficiency Virus Type 1 Particle Release Correlates with Localization of Vpu in the trans-Golgi Network^{▿ †}

Cells, transfection, and siRNAs.

HeLa-TZM cells were obtained from the AIDS Research and Reference Program (NIH). HEK 293T and HeLa cells were obtained from ATCC. All cells were maintained as described previously (15). Unless specified, HEK 293T and HeLa cells were transfected by the calcium-phosphate method, and analyses were performed 48 h posttransfection. All specific small interfering RNAs (siRNAs) were synthesized by Dharmacon as 21-mers with UU overhangs. The siRNAs were transfected using Oligofectamine (Invitrogen) according to the manufacturer's instructions. Briefly, siRNAs were preincubated with 15 μl of Oligofectamine and overlaid on cells at 50% confluence. Both siRNAs specific to CLCa (GGAAAGUAAUGGUCCAACA) and CLCb (GGAACCAGCGCCAGAGUGA) were used for CLC depletion at a final concentration of 62.5 nM each. The siRNA specific to CHCs (GCAAUGAGCUGUUUGAAGA) or tetherin (Dharmacon Smartpool, catalogue no. L-011817) and the nontargeting siRNA Scrambled #2 from Dharmacon were used at a final concentration of 125 nM. The siRNA-transfected cells were then transfected with the appropriate proviral construct 72 h post-siRNA transfection using Lipofectamine 2000 (Invitrogen) and processed 24 h later. Empty plasmid DNA was added to each transfection to keep the amount of transfected DNA constant.

Plasmid constructs.

The mammalian expression plasmids SVCMV-vpu wt, SVCMV-vpu-, and SVCMV-CD4 have been described previously (2). HxBH10-vpu wt and HxBH10-vpu- are two isogenic infectious HIV-1 molecular clones that only differ in the expression of Vpu (35). HxBH10-vpu R30A,K31A, HxBH10-vpu Y29A, HxBH10-vpu I32A,L33A, SVCMV-vpu R30A,K31A, SVCMV-vpu I32A,L33A, and SVCMV-vpu Y29A were generated by PCR-based site-directed mutagenesis as described previously (38). The plasmids pEGFP-Rab7 wt, pEGFP-Rab7 N125I, and pEGFP-Rab7 Q67L were kindly provided by Robert Lodge (Université Laval, Quebec, Canada). To construct pEYFP-N1-Vpu wt, Vpu was amplified by PCR from the molecular clone HxBH10-vpu wt using the forward primer 5′-TCAAAGCAGTCTAGAGTACATGTA-3′ and the fusion primer 5′-TCCGCCGCCCCCCAGATCATCAACATC-3′. Enhanced yellow fluorescent protein (EYFP) was amplified by PCR from the mammalian expression plasmid pEYFP-N1 (Clontech) using the fusion primer 5′-GGGGGCGGCGGAGTGAGCAAGGGCGAG-3′ and reverse primer 5′-GATTATGATCTATAGTCGCGGCCGC-3′. The fusion of EYFP to the C-terminal end of Vpu was achieved by a second PCR on the two previously generated fragments using the forward and reverse primers. The resulting PCR product was inserted in pEYFP-N1 within the NheI/NotI restriction sites. The Vpu deletion mutants were generated using a similar strategy. The forward primer mentioned above was used to amplify VpuΔ9, -Δ13, -Δ14, -Δ18, and -Δ23 with the fusion primers 5′-CTCACTCCGCCGCCCCCGTGCCCCATCTCCACCCCCATC-3′, 5′-CTCACTCCGCCGCCCCCCACCCCCATCTCCACAAGTGC-3′, 5′-CTCACTCCGCCGCCCCCCCCCATCTCCACAATGCTGA-3′, 5′-CTCACTCCGCCGCCCCCAAGTGCTGATATTTCTCCTTC-3′, and 5′-CTCACTCCGCCGCCCCCTCCTTCACTCTCATTGCCACTG-3′, respectively. To amplify EYFP, the same reverse primer described above and the fusion primer 5′-GGGGGCGGCGGAGTGAGCAAG-3′ were used. The resulting PCR fragments were inserted in the pEYFP-N1 vector within the NheI/NotI restriction sites. All constructs were confirmed by automatic DNA sequencing.

Pulse-chase and radioimmunoprecipitation experiments.

The pulse-chase experiments were performed as described previously (2). Briefly, transfected HEK 293T cells were pulse labeled for 30 min with 800 μCi/ml of [³⁵S]methionine and [³⁵S]cysteine (Perkin Elmer) and chased in Dulbecco's modified Eagle's medium (DMEM) containing 5% fetal bovine serum. For some experiments, 10 μM BFA was added throughout the labeling and chase periods. At the indicated time periods, radiolabeled cells were lysed in radioimmunoprecipitation assay (RIPA-DOC) buffer (10 mM Tris [pH 7.2], 140 mM NaCl, 8 mM Na₂HPO₄, 2 mM NaH₂PO₄, 1% Nonidet P-40, 0.5% sodium dodecyl sulfate [SDS], 1.2 mM deoxycholate) supplemented with a cocktail of protease inhibitors (Protease Inhibitors Complete, Roche Diagnostics). Following lysis, the labeled proteins were immunoprecipitated with the indicated antibodies and analyzed by SDS-polyacrylamide gel electrophoresis (PAGE) and autoradiography (38).

Steady-state detection of proteins by Western blotting.

HeLa or HEK 293T cell transfectants were lysed in RIPA-DOC buffer. The proteins from the lysates were resolved on 10 to 12.5% SDS-PAGE Tricine gels and electroblotted as described elsewhere (2). Western blotting was performed as described previously (10).

Fluorescent and confocal microscopy.

Immunofluorescence procedures were described previously (10). Briefly, immunostaining was performed on HeLa cells transfected with either the HxBH10 proviral DNA constructs or the Vpu-expressing plasmids. Cells were fixed with 4% paraformaldehyde and permeabilized with 0.2% Triton X-100. Next, the cells were incubated with the first antibody and incubated with the appropriate secondary antibody, and for the samples analyzed by fluorescence microscopy, nuclei were stained with DAPI (4′,6-diamidino-2-phenylindole) for 5 min. To increase the EYFP signal, anti-GFP antibodies were used. The cells were examined by conventional epifluorescence micrographs on a Zeiss cell observer system (Zeiss) equipped with an Axiovert 200 M microscope using the 100× oil lens. Where specified, analyses were also performed with an Axiovert 100 M confocal microscope (Zeiss). The images were digitally deconvoluted with the AxioVision 3.1 software, using the nearest-neighbor deconvolution method. The quantitation of Vpu accumulation in the TGN was determined using the Zeiss LSM510 software. The percentage of Vpu accumulating in the TGN was calculated by evaluating the Vpu signal intensity in the TGN, as delimited by the TGN46 marker, relative to the total Vpu signal intensity in the cell. Statistical analysis was performed using an unpaired Student's t test, and values were considered statistically significant at P values of <0.05.

Transferrin uptake assay.

For the transferrin uptake experiments, siRNA-transfected HeLa cells were washed with phosphate-buffered saline (PBS) 96 h post-siRNA transfection. The cells were incubated at 4°C in serum-free DMEM. After 15 min, Alexa-Fluor-488-conjugated transferrin (Molecular Probes) was added and the cells were incubated for 30 min at 4°C. The cells were then washed with PBS and shifted to 37°C in serum-free DMEM for 10 min. The internalization process was blocked by shifting the cells at 4°C. Uninternalized transferrin on the cell membrane was washed away by washing the cells with 150 mM glycine (pH 2.5), and the uptake was determined by flow cytometry.

Viral release assays.

HeLa cells were seeded on 100-mm² dishes and transfected as described above. The transfectants were metabolically labeled with 200 μCi/ml [³⁵S]methionine-cysteine for 5 h. The virus particles from the supernatants were clarified by centrifugation, filtered through a 45-μm filter, and then pelleted by ultracentrifugation onto a 20% sucrose cushion in PBS for 90 min at 130,000 × g at 4°C. The cells and virions were subsequently lysed in RIPA-DOC buffer. Gag-related products were immunoprecipitated from the cell and virus lysates using a mixture of human anti-HIV (#153) serum and mouse anti-p24 monoclonal antibody. Additionally, Vpu was immunoprecipitated from the cell lysates using a polyclonal rabbit anti-Vpu serum. The radiolabeled proteins were resolved on 10% SDS-PAGE Tricine gels and analyzed by autoradiography. For the steady-state analysis, viral particles were pelleted from the supernatant of the transfected HeLa cells using the procedure described above. The proteins were analyzed by Western blotting using specific antibodies. To measure the infectious virus released from the transfected cells, HeLa-TZM indicator cells were inoculated with an aliquot of virus-containing supernatant. After 48 h, HeLa-TZM cells were lysed and luciferase activity was determined using the Promega luciferase assay system. A statistical analysis was performed using a paired Student's t test, and values were considered statistically significant at P values of <0.05.

Mutations in the hinge region of Vpu affect the subcellular localization of the protein.

The finding that Vpu is localized in the TGN and is associated to some extent with late endosomal compartments suggests that the protein is trafficking through the endosomal network and, as such, might contain motifs and/or domains governing trafficking or the retention of the protein to specific subcellular compartments. Structurally, Vpu consists of two major domains, an N-terminal transmembrane (TM) domain that anchors Vpu in cellular membranes (29, 34, 37) and a cytoplasmic tail consisting of two putative α-helices (α-helix 1, amino acids [aa] 37 to 51, and α-helix 2, aa 57 to 72) separated by a conserved DSGΦXS β-TrCP recognition site phosphorylated by casein kinase II (Fig. ​(Fig.2A)2A) (9, 30, 31). Upon analysis of our Vpu mutant collection, we identified a mutant of Vpu that was more stable than the Vpu wild type (wt). This mutant, Vpu R30A,K31A, harbored two alanine substitutions at the well-conserved arginine (Arg) and lysine (Lys) basic residues located at positions 30 and 31 in the hinge region between the TM and the cytoplasmic domains of the protein (Fig. ​(Fig.2A).2A). Interestingly, these two positively charged residues are part of a putative tyrosine-based, YXXΦ sorting signal (Y corresponds to Tyr, the Xs are residues that are highly variable, and Φ corresponds to residues with bulky hydrophobic side chains) and an overlapping dileucine-based motif, (D/E)XXXL(L/I) (D/E corresponds to Asp or Glu, the Xs are residues that are highly variable, and L/I corresponds to Leu or Ile) (Fig. ​(Fig.2A).2A). In contrast to the Vpu wt, Vpu R30A,K31A levels remained relatively similar throughout a 5-h pulse-chase experiment and were not affected when the protein was prevented from trafficking out of the ER by treatment with BFA, a fungal metabolite that blocks protein sorting from the ER to the Golgi complex (8) (see Fig. S1 in the supplemental material). These results suggested that these highly conserved positively charged amino acid residues (96% conservation throughout the Vpu subtypes) located in the hinge region of Vpu may affect Vpu trafficking and/or stability. In fact, single R30A or K31A mutations were also found to be sufficient to stabilize Vpu (data not shown).

Open in a separate window

FIG. 2.

Mutations in the hinge region between the TM and cytoplasmic domains of Vpu affect its subcellular localization. (A) Schematic representation of the Vpu structural domains with the amino acid sequence of the hinge region derived from several strains of HIV-1 group M. Representative consensus amino acid sequences of the Vpu hinge region for each HIV-1 subtype are indicated. The red rectangle highlights the overlapping tyrosine-based (YXXΦ) and dileucine-based ([D/E]XXXL[L/I]) sorting motifs. Among sequences of all HIV-1 strains from the HIV databases of the National Institute of Allergy and Infectious Diseases (NIAID) of the National Institutes of Health (NIH; www.hiv.lanl.gov/content/index), 65% of the sequences show RK residues at positions 30 and 31 (red letters), while 96% of the sequences show a conservation of positively charged residues (RR, KR, or KK) at those positions within the hinge region. Yellow circled amino acids indicate phosphoacceptor sites at serines 52 and 56 (bold). −, amino acid identical to Vpu from HxBH10; αH, α-helix. (B) HeLa cells expressing HxBH10-vpu R30A,K31A were costained with the anti-Vpu serum (red) and anti-TGN46, anti-CD63, anti-LAMP1, or anti-calreticulin antibodies (green). Nuclei were counterstained with DAPI (blue). Pictures show representative examples of Vpu R30A,K31A-expressing cells. Enlarged pictures are shown beside the panels. White arrows indicate noticeable examples of punctate colocalization. White bars represent a distance of 10 μm. (C) Quantitation of Vpu wt and Vpu R30A,K31A accumulation in the TGN as determined by the Vpu signal measured in the TGN (region 1) relative to the total Vpu signal (region 2) in the cell. Error bars indicate the standard deviations of the means from the quantitative analysis of at least 25 distinct Vpu-expressing cells.

To investigate the effect of mutations in the hinge region on Vpu trafficking, we compared the subcellular localization of native Vpu wt and Vpu R30A,K31A in HIV-1-producing HeLa cells. As shown in Fig. ​Fig.2,2, the R30A,K31A mutations decreased very significantly the accumulation of native Vpu in the TGN (∼40% of the total Vpu R30A,K31A is detected in the TGN compared to ∼80% for the Vpu wt [Fig. 2B and C]) while increasing the association of the protein with the late endosomal CD63 marker (Fig. ​(Fig.2B).2B). Furthermore, in contrast to wt Vpu, we observed a strong colocalization of Vpu R30A,K31A with LAMP1 (compare Fig. ​Fig.2B2B to Fig. ​Fig.1B).1B). Importantly, no significant colocalization was observed between Vpu R30A,K31A and calreticulin, thus ruling out the possibility that the decreased accumulation of the mutant protein in the TGN resulted from a sequestration in the ER.

Overall, these findings suggest that mutations at Arg30 and Lys31 within the hinge region of Vpu affect both the accumulation of the protein in the TGN as well as its efficient delivery to late endosomal degradative compartments, indicating that Vpu might traffic between the TGN and the endosomal system.

Vpu traffics through the late endosomal network and is degraded in lysosomes.

To further examine whether Vpu is targeted for degradation in late endosomal structures, we treated Vpu-expressing HEK 293T and HeLa cells with CQ, a lysosomotropic agent that inhibits lysosomal protein degradation by neutralizing the acidic environment of endocytic vesicles (7). As shown in Fig. ​Fig.3A,3A, the treatment of transfected cells with CQ led to a dose-dependent accumulation of Vpu. Similar results were obtained when Vpu-expressing HEK 293T cells were treated with ammonium chloride, another lysosomotropic weak base that acts as a lysosomal protein degradation inhibitor (data not shown). Furthermore, to provide direct evidence that Vpu is targeted for degradation in lysosomes, we analyzed Vpu degradation and subcellular localization in conditions where we interfered with the function of Rab7, a small GTPase that is found in late endosomes (26) and is implicated in the biogenesis and maintenance of the perinuclear lysosomal compartment (5). For that purpose, we took advantage of constructs encoding enhanced GFP (EGFP)-Rab7 fusion proteins that do not perturb the delivery of cargo to lysosomes (the Rab7 wt and the constitutively active Rab7 Q67L) or, inversely, that interfere with lysosomal targeting and degradation (Rab7 N125I) (5, 26). While the overexpression of EGFP-Rab7 wt or EGFP-Rab7 Q67L did not alter the level of Vpu at steady state, the DN EGFP-Rab7 N125I drastically increased the steady-state levels of Vpu (Fig. ​(Fig.3B,3B, compare lanes 8 to 10 to lanes 2 to 7). Interestingly, expression of DN EGFP-Rab7 N125I in Vpu+ HIV-1-expressing HeLa cells led to an accumulation of Vpu in Rab7-positive vesicular structures while expression of EGFP-Rab7 wt (Fig. ​(Fig.3C)3C) or EGFP-Rab7 Q67L did not (data not shown). Importantly, accumulation of Vpu in the TGN was not significantly affected by the overexpression of the DN Rab7 N125I mutant, suggesting that endosome-to-TGN trafficking was not perturbed under these conditions (Fig. ​(Fig.3D).3D). Altogether, these data indicate that native Vpu traffics through the late endosomal network and undergoes degradation in lysosomes.

Open in a separate window

FIG. 3.

Abrogation of lysosomal function or targeting increases Vpu stability. (A) Vpu wt-expressing HEK 293T (top panels; SVCMV-vpu+) or HeLa cells (bottom panels; HxBH10-vpu wt) were treated for 20 h with the indicated increasing concentrations of CQ prior to lysis. Vpu was subsequently resolved on SDS-PAGE and detected by Western blotting using anti-Vpu antibodies. Actin served as a loading control. (B) HEK 293T cells were cotransfected with SVCMV-vpu wt and the pEGFP-Rab7 plasmids as indicated. Proteins were resolved on SDS-PAGE and detected by Western blotting using anti-Vpu, anti-actin, and anti-GFP antibodies. Note that the blots in panels A and B are underexposed to clearly show Vpu stabilization upon CQ treatment or pEGFP-Rab7 N125I coexpression. (C) HeLa cells were cotransfected with the HxBH10-vpu wt proviral construct and either the pEGFP-Rab7 wt or pEGFP-Rab7 N125I plasmids. Transfected cells were stained with anti-Vpu rabbit serum (red). Nuclei were counterstained with DAPI (blue). White arrows indicate noticeable colocalization of Vpu with EGFP-Rab7 N125I in vesicular structures. The white bar represents a distance of 20 μm. (D) Quantitation of Vpu accumulation in the TGN was performed as described in the legend to Fig. ​Fig.22.

Mutations in the hinge region of Vpu interfere with Vpu-mediated enhancement of HIV-1 particle release.

Having shown that the Vpu R30A,K31A mutant displayed defects in lysosomal delivery and degradation as well as in accumulation in the TGN, we examined the implication of these defects on the Vpu-mediated enhancement of viral particle release. Equal DNA amounts of Vpu-defective proviral construct (HxBH10-vpu-) or proviral constructs encoding the Vpu wt (HxBH10-vpu wt) or Vpu R30A,K31A (HxBH10-vpu R30A,K31A) were transfected in restrictive HeLa cells. Forty-eight hours posttransfection, HIV-1-producing cell cultures were metabolically labeled for 5 h and both cells and extracellular supernatants were harvested. Virus particles were further isolated from the supernatants by clarification, filtration, and ultracentrifugation as described in Materials and Methods. Gag protein levels were then analyzed in cell and viral particle lysates by immunoprecipitation using a serum from an HIV-1-positive individual combined with anti-24 polyclonal antibodies. Figure 4A and B reveal that while newly synthesized Vpu wt viral particles were released more efficiently than Vpu-defective particles, the Vpu R30A,K31A mutant virus exhibited a significant (P < 0.001) ∼40% attenuation of viral particle release efficiency compared to that of the Vpu wt virus control. Interestingly, this attenuation was accentuated to ∼80% when viral particle release was evaluated in conditions where the Vpu wt and Vpu R30A,K31A protein levels were comparable (Fig. 4C and D, compare lane 5 to lane 4). In this context, the mislocalization of Vpu R30A,K31A from the TGN was slightly enhanced (data not shown). Importantly, the Vpu R30A,K31A mutant was found to mediate CD4 degradation as efficiently as the Vpu wt (see Fig. S2 in the supplemental material), indicating that these mutations in the hinge region did not affect the overall conformation of the protein but rather specifically interfered with the protein's ability to promote viral particle release. These results suggest that the mislocalization of Vpu outside of the TGN and/or the inefficient delivery of the protein to lysosomal compartments leads to the impairment of Vpu-mediated enhancement of viral particle release. To discriminate between these two possibilities, we examined the effect of disrupting Rab7-mediated transport on the release of HIV-1 particles in HeLa cells. The Vpu-defective proviral construct was cotransfected with Vpu-expressing or control plasmids and the pEGFP-Rab7 wt or pEGFP-Rab7 N125I plasmids. As shown in Fig. 4E and F, the expression of the DN Rab7 N125I mutant stabilized Vpu, as expected, but had no significant effect on the enhancement of viral particle release. Similarly, the abrogation of endocytic vesicle acidification and lysosomal degradation by CQ treatment did not affect the release of Vpu wt HIV-1 particles (data not shown). These findings indicate that the delivery and degradation of Vpu in lysosomes do not represent essential steps in the process underlying the Vpu-mediated enhancement of viral particle release and, consequently, point toward the accumulation of Vpu in the TGN as a possible important requirement for the protein's ability to promote HIV-1 particle release.

Open in a separate window

FIG. 4.

Mutations in the hinge region of Vpu interfere with the Vpu-mediated enhancement of HIV-1 particle release. (A) HeLa cells were mock transfected (lane 1) or transfected with either the HxBH10-vpu- (lane 2), HxBH10-vpu wt (lane 3), or HxBH10-vpu R30A,K31A (lane 4) proviral plasmids as indicated. Radiolabeled cells and supernatant-containing viral particles were harvested 48 h posttransfection and analyzed for the presence of Vpu and Gag proteins by immunoprecipitation as described in Materials and Methods. (B) Densitometric quantitation of viral release efficiency. Bands corresponding to Gag products in cells and viral particles were scanned by laser densitometry. The release efficiency of HxBH10-vpu wt was arbitrarily set at 100%. Error bars indicate the standard deviations of the means of the results from four separate experiments. (C) HeLa cells were transfected with HxBH10-vpu- alone (lane 1) or with increasing quantities of HxBH10-vpu wt (lanes 2 to 4) or HxBH10-vpu R30A,K31A (lanes 5 to 7). The total amount of proviral DNA transfected in each condition was identical. Cells and supernatant-containing viral particles were harvested 48 h posttransfection, and Gag proteins in cell and virus lysates were analyzed by Western blotting using anti-p24 antibodies. Vpu was detected by Western blotting using anti-Vpu antibodies. The asterisk indicates a nonspecific band. (D) Same as described for panel C except that infectious virions in culture supernatants were measured using HeLa-TZM indicator cells and a chemiluminescence assay in relative light units (RLU), as described in Materials and Methods. The maximal RLU value obtained with HxBH10-vpu wt-expressing cells (lane 4) was arbitrarily set at 100%. Note that the data shown in lanes 4 and 5 correspond to the activities obtained with comparable levels of Vpu wt and Vpu R30A,K31A (panel C). (E) HeLa cells were cotransfected with the HxBH10-vpu- proviral construct, the indicated pEGFP-Rab7 plasmids, and the SVCMV-vpu wt or control constructs. Cells and virus particles were processed and analyzed as described for panel C. (F) Densitometric quantitation of the results shown in panel E. Viral particle release efficiency obtained in cells coexpressing HxBH10-vpu- and SVCMV-vpu wt was arbitrarily set at 100%. Error bars indicate the standard deviations of the means of the results from three separate experiments.

Disruption of Vpu trafficking from the TGN to the endosomal system stimulates HIV-1 particle release.

Clathrin-coated vesicles are major carriers for endocytic cargo and mediate important intracellular trafficking events at the TGN and endosomes. Whereas CHCs provide the structural backbone of the clathrin coat and are critical for all clathrin-mediated trafficking processes recent evidence suggests that CLCs are not required for clathrin-mediated endocytosis but are critical for clathrin-mediated trafficking between the TGN and the endosomal system (25). For instance, CLC depletion by siRNA did not have any influence on the clathrin-mediated endocytosis of transferrin but caused the CI-MPR to cluster near the TGN (25). To obtain direct evidence that the localization of Vpu in the TGN is important for Vpu function on HIV-1 particle release, we assessed the effect of disrupting clathrin-mediated trafficking from the TGN to the endosomal system on the Vpu-mediated enhancement of HIV-1 particle release using siRNA specific for CLCs or CHCs. Equal DNA amounts of proviral constructs defective for Vpu or encoding wt Vpu or the Vpu R30A,K31A mutant were transfected in HeLa cells that were depleted for CLCs beforehand. Viral particle release efficiency was then evaluated 96 h post-siRNA transfection by measuring the levels of Gag associated with cell or viral particle lysates (Fig. ​(Fig.5A)5A) as well as by evaluating the relative infectivity of the released viral particles using the HeLa-TZM indicator cell line (Fig. ​(Fig.5B).5B). HeLa cells showed reduced levels of CLCs (approximately a 60% decrease) after the transfection of siRNA specific for CLCs compared to those of the control HeLa cells transfected with a nonspecific scrambled siRNA (Fig. ​(Fig.5A,5A, compare lanes 4 to 6 to lanes 1 to 3). As previously reported (25), the depletion of CLCs led to a clustering of CI-MPR near the TGN and did not affect the clathrin-mediated endocytosis of transferrin from the cell surface (see Fig. S3A and 3B in the supplemental material). Although the depletion of CLCs did not influence the release of Vpu-defective HIV-1 particles, it had a stimulatory effect on the release of Vpu⁺ HIV-1 particles. Indeed, a significant ∼30% increase (P = 0.035) in viral particle release efficiency was observed in Vpu-expressing cells transfected with CLC siRNA compared to that of the control (Fig. 5A and B). This stimulation correlated with a stabilization of the Vpu wt (Fig. ​(Fig.5A,5A, compare lane 2 to lane 5), suggesting that the depletion of CLCs is indeed efficiently retaining Vpu within the TGN and, as such, preventing its trafficking to lysosomes. Interestingly, the defect in viral particle release observed with the Vpu R30A,K31A mutant was completely abrogated under conditions where CLC levels were reduced. Indeed, under these conditions, the release efficiency of the Vpu R30A,K31A mutant was comparable to that of the Vpu wt. Importantly, this abrogation of the viral particle release defect correlated with an increased accumulation of the Vpu R30A,K31A mutant in compartments that costained with the TGN marker TGN46 (Fig. 5C and D). As expected, Vpu R30A,K31A was not increasingly stabilized upon CLC siRNA treatment (Fig. ​(Fig.5A,5A, compare lane 3 with lane 6) since this mutant is already impaired in its ability to be targeted for degradation in lysosomes. Overall, these findings suggest that this mutant is not impaired in its ability to overcome the restricting activity of tetherin but is rather inefficient at localizing at the TGN. As expected, similar results were obtained when viral particle release was analyzed in HeLa cells that were transfected with CHC siRNA. Under these conditions, the release efficiencies of the Vpu wt- and Vpu R30A,K31A-encoded HIV-1 particles were equivalent (see Fig. S4 in the supplemental material). Taken together, these results further suggest that native Vpu is trafficking between the TGN and the endosomal system. Additionally, they provide strong evidence that the proper distribution of Vpu in the TGN at steady state is a key requirement to overcome the restricting activity of tetherin on HIV-1 particle release.

Open in a separate window

FIG. 5.

Effect of the depletion of CLCs on the subcellular localization of Vpu and its viral particle release enhancing function. (A) HeLa cells were transfected either with control siRNA (Scrambled) (lanes 1 to 3) or specific siRNA against CLCa and CLCb (CLCs) (lanes 4 to 6). Seventy-two hours post-siRNA transfection, cells were transfected with similar amounts of the HxBH10-vpu- (lanes 1 and 4), HxBH10-vpu wt (lanes 2 and 5), or HxBH10-vpu R30A,K31A (lanes 3 and 6) proviral constructs as indicated. Cells and supernatant-containing viral particles were harvested 96 h post-siRNA transfection. Gag proteins in cell and virus lysates were analyzed by Western blotting using anti-p24 antibodies. Vpu and CLC levels were determined using specific antibodies. Actin served as a loading control. (B) Same as described for panel A except that infectious virions in culture supernatants were measured using HeLa-TZM indicator cells and a chemiluminescence assay in relative light units (RLU) as described in Materials and Methods. The RLU value obtained in HxBH10-vpu wt-expressing cells transfected with the nontargeting control siRNA was arbitrarily set at 100%. Error bars indicate the standard deviations of the means of the results from four independent experiments. (C) HeLa cells were transfected with either the scrambled siRNA or with CLC siRNA. Cells were transfected with the HxBH10-vpu wt or HxBH10-vpu R30A,K31A proviral plasmid 72 h later. Twenty-four hours later, cells were costained for Vpu (red) and TGN46 (green). Nuclei were counterstained with DAPI (blue). The white bar represents a distance of 10 μm. (D) Quantitation of Vpu accumulation in the TGN was performed as described in the legend to Fig. ​Fig.22.

The membrane proximal tyrosine- and dileucine-based sorting motifs of subtype B Vpu are not functionally active.

Having obtained evidence that Vpu traffics between the TGN and the endosomal system and that mutations at Arg30 and Lys31 in the Vpu hinge region affect this process, we next evaluated whether these residues are part of bona fide tyrosine-based (YXXΦ) or dileucine-based ([D/E]XXXL[L/I]) sorting signals, which might regulate Vpu trafficking between the TGN and late endosomal/lysosomal compartments. Toward this goal, we generated expression plasmids encoding Vpu mutants that harbored substitution mutations at the Tyr residue located at position 29 (SVCMV-vpu Y29A) or at the Ile and Leu residues located at position 32 and 33 (SVCMV-vpu I32A,L33A) (Fig. ​(Fig.2A)2A) since residues at these positions have been previously shown to be essential for the function of tyrosine- and dileucine-based trafficking signals (3). An analysis of Vpu Y29A and Vpu I32A,L33A subcellular localization did not reveal any marked redistribution of the mutant proteins compared to that of wt Vpu (Fig. 6A and B). Furthermore, mutations at Tyr29 or at Ile32 and Leu33 did not reveal any significant change in the ability of Vpu to enhance HIV-1 particle release (Fig. 6C and D). Moreover, in contrast to Vpu R30A,K31A, the Vpu Y29A and Vpu I32A,L33A mutants did not display any increase in their steady-state levels compared to that of the Vpu wt (Fig. ​(Fig.6C,6C, compare lanes 4 and 5 with lane 2). These findings indicate that the tyrosine- and the dileucine-based trafficking motifs do not appear to actively regulate Vpu trafficking from the TGN to the endosomal system and, additionally, are not essential for the viral particle release enhancing function of the protein at least in the context of subtype B Vpu. Consequently, the phenotypes observed when the Arg30 and Lys31 residues are mutated might result from a limited activation of these putative trafficking motifs.

Open in a separate window

FIG. 6.

Mutations in the putative overlapping tyrosine- and dileucine-based trafficking signals do not interfere with Vpu subcellular localization or Vpu-mediated enhancement of HIV-1 particle release. (A) HeLa cells transfected with the SVCMV-vpu wt, SVCMV-vpu Y29A or SVCMV-vpu I32A,L33A plasmid were costained with the anti-Vpu serum (red) and anti-TGN46 antibodies (green) 48 h posttransfection and observed by confocal microscopy. Pictures show representative examples of Vpu localization patterns. The white bar represents a distance of 10 μm. (B) Quantitation of Vpu accumulation in the TGN as determined by the Vpu signal measured in the TGN relative to the total Vpu signal in the cell. Error bars indicate the standard deviations of the means from the quantitative analysis of at least 25 distinct Vpu-expressing cells. (C) HeLa cells were cotransfected with the specified HxBH10 proviral constructs. Cells and supernatant-containing viral particles were analyzed as described in the legend to Fig. ​Fig.4C.4C. (D) Same as described for panel C except that infectious virions in culture supernatants were measured using HeLa-TZM indicator cells and a chemiluminescence assay. The RLU value obtained in Vpu wt-expressing cells (lane 2) was arbitrarily set at 100%. Error bars indicate the standard deviations of the means of the results from four independent experiments.

Deletion analysis of the Vpu cytosolic tail identifies a determinant important for Vpu retention/localization in the TGN and efficient HIV-1 viral particle release.

To further delineate the region of the cytoplasmic tail that might be important for the subcellular localization of Vpu, we used a Vpu-EYFP reporter system to assess the intracellular localization of Vpu C-terminal deletion mutants by fluorescence microscopy. This approach was previously used by Pacyniak et al. (23) to identify a region within the cytoplasmic domain of subtype B Vpu that might be responsible for the retention of the protein in the TGN. The Vpu-EYFP fusion protein expressed in restrictive HeLa cells demonstrated a subcellular distribution pattern similar to that observed with native Vpu expressed from an infectious molecular clone (compare Fig. ​Fig.7B7B to Fig. ​Fig.1A).1A). Notably, ∼87% of the Vpu fluorescence staining was found to localize in a perinuclear region that overlapped with the Golgi marker TGN46 (Fig. 7B and C). No accumulation of wt Vpu-EYFP was observed at the plasma membrane. We next examined the subcellular localization of various Vpu-EYFP deletion mutants (Fig. ​(Fig.7A).7A). As reported by Pacyniak et al. (23), removal of 9 (VpuΔ9-EYFP) or 13 (VpuΔ13-EYFP) aa from the carboxyl terminus of Vpu had, overall, marginal effects on the localization of the protein (Fig. 7B and C). In contrast, the deletion of the C-terminal 14, 18, and 23 amino acids, which encompass the second helical structure of the Vpu cytoplasmic tail, decreased significantly the accumulation of Vpu-EYFP in the TGN while increasing its detection within cytosolic vesicular structures, which costained in part with the early endosome marker Rab5 and with CD63 (Fig. 7B to D). Interestingly, VpuΔ14, -Δ18, and -Δ23 could also be detected at the plasma membrane in some cells (data not shown).

Open in a separate window

FIG. 7.

Effect of deletions of the Vpu cytosolic tail on Vpu subcellular localization. (A) Schematic representation of the Vpu-EYFP deletion constructs. Yellow circled amino acids indicate phosphoacceptor sites at serines 52 and 56 (bold). αH, α-helix. (B) HeLa cells transfected with plasmids encoding the indicated Vpu-EYFP deletion mutant were costained with the anti-GFP (green) and anti-TGN46 (red) antibodies. Cells were observed by confocal microscopy. Pictures show representative examples of the localization pattern observed for each deletion mutant. (C) Quantitation of Vpu accumulation in the TGN was performed as described in the legend to Fig. ​Fig.2.2. (D) HeLa cells expressing Vpu-EYFP or VpuΔ23-EYFP were costained with anti-GFP (green) and anti-CD63 or anti-Rab5 (red) antibodies. Cells were observed by confocal microscopy. Enlarged pictures are shown beside the panels. White arrows indicate noticeable examples of punctate colocalization. All the white bars represent a distance of 10 μm.

We next assessed the effect of Vpu cytoplasmic tail deletion on the ability of the protein to promote the release of the HIV-1 particle. Vpu-EYFP or Vpu-EYFP deletion mutants were cotransfected with the Vpu-defective proviral construct HxBH10-vpu- in HeLa cells, and viral particle release was determined 48 h posttransfection by measuring Gag protein levels in cell and viral particle lysates by Western blotting (Fig. ​(Fig.8A).8A). In addition, we also determined the viral particle release efficiency by evaluating the relative infectivity of the released viral particles (Fig. ​(Fig.8B).8B). The results revealed that the VpuΔ9-EYFP and VpuΔ13-EYFP mutants, which were minimally affected in their localization to the TGN, promoted the release of HIV-1 viral particles almost as efficiently as wt Vpu-EYFP (Fig. ​(Fig.8A8A [compare lanes 3 and 4 to lane 8] and B). In contrast, the VpuΔ14-EYFP, VpuΔ18-EYFP, and VpuΔ23-EYFP deletion mutants, which displayed a significant decrease in the accumulation of the protein in the TGN, showed an attenuation of their viral particle release-enhancing function (Fig. ​(Fig.8A8A [compare lanes 5 to 7 to lane 8] and B). The removal of the C-terminal 23 amino acids of Vpu abrogated the capacity of the protein to enhance particle release to a level comparable to that of the Vpu-defective HIV-1 virus control (Fig. ​(Fig.8A,8A, compare lane 7 to lane 2). Overall, these findings suggest that the C-terminal 23 amino acids of Vpu, which contain the second α-helical domain of the Vpu cytoplasmic domain, might be important for the localization/retention of the protein in the TGN. Furthermore, these results provide additional evidence that the proper distribution of Vpu in the TGN is important for the efficient release of viral particles from the cell surface.

Open in a separate window

FIG. 8.

Effect of deletions of the Vpu cytosolic tail on Vpu-mediated enhancement of viral HIV-1 release. (A) HeLa cells were mock transfected (lane 1) or transfected with HxBH10-vpu- and plasmids encoding EYFP (lane 2) or Vpu-EYFP (lane 8) or the specified Vpu-EYFP deletion mutants (lanes 3 to 7). Cells and supernatants were analyzed as described in the legend to Fig. ​Fig.4C.4C. EYFP or Vpu-EYFP derivatives were detected using anti-GFP antibodies. (B) Same as described for panel A except that infectious virions in culture supernatants were measured using HeLa-TZM indicator cells and a chemiluminescence assay as described in Materials and Methods. The RLU measured with the Vpu-EYFP wt was arbitrarily set at 100%. Error bars indicate the standard deviations of the means of the results from four independent experiments.

We thank Robert Lodge for the kind gift of the pEGFP-Rab7 plasmids used in this study as well as René-Pierre Lorgeoux and Alexandre Orthwein for experimental help and for sharing reagents and unpublished data. We also thank present and former members of the Cohen group for helpful discussions. HeLa-TZM cells were obtained from the AIDS Research and Reference Reagent Program (NIH).

This work was supported by grants from the Canadian Institutes of Health Research (CIHR) (MOP-14228) and from the Fonds de la Recherche en Santé du Québec (FRSQ) to E.A.C. M.D. and P.G.-G. are the recipients of studentships from the CIHR strategic training program in cancer research, while J.B. is the recipient of a studentship from the FRSQ. E.A.C. holds the Canada research chair in human retrovirology.