Vpu Enhances HIV-1 Virus Release in the Absence of Bst-2 Cell Surface Down-Modulation and Intracellular Depletion

Vpu enhances HIV-1 virus release in the absence of Bst-2 cell surface down-modulation and intracellular depletion

Eri Miyagi1, Amy J. Andrew1, Sandra Kao, and Klaus Strebel2

Laboratory of Molecular Microbiology, Viral Biochemistry Section, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD 20892-0460

Communicated by Malcolm A. Martin, National Institutes of Health, Bethesda, MD, December 24, 2008 (received for review November 24, 2008) HIV-1 Vpu enhances the release of virions from infected cells. type II transmembrane protein consisting of 180 aa (27). The Recent work identified Bst-2/CD317/tetherin as a host factor protein has an unusual topology in that it has both an N-terminal whose inhibitory activity on viral release is counteracted by Vpu. transmembrane domain and a C-terminal glycosylphosphatidyl- A current working model proposes that Bst-2 inhibits virus release inositol (GPI) anchor (28). Bst-2 associates with lipid rafts at the by tethering viral particles to the cell surface. Here, we analyzed cell surface and was identified on internal membranes, presum- endogenous Bst-2 with respect to its effect on virus release from ably the trans-Golgi network (28). However, the precise function HeLa cells, T cells, and macrophages. We noted significant cell of Bst-2 remains unknown. Bst-2 mRNA was constitutively expressed in cell types such as HeLa, Jurkat, and at low levels in type-dependent variation in Bst-2 expression. Vpu caused a reduc- ϩ tion in Bst-2 expression in transfected HeLa cells and long-term CD4 T cells but not 293T or HT1080 cells (29) and thus infected macrophages. However, Vpu expression did not result in correlated with cell types known to depend on Vpu for efficient cell surface down-modulation of Bst-2 or a reduction in intracel- virus release. Bst-2 expression was inducible by IFN treatment in 293T or HT1080 cells, consistent with the previous observation lular Bst-2 expression in CEMx174 or H9 cells, yet virus replication that cell lines that did not normally require Vpu for efficient in these cells was Vpu-responsive. Surprisingly, Bst-2 was unde- virus release became Vpu-dependent upon IFN treatment (22). tectable in cell-free virions that were recovered from the surface of Although Vpu had no overt effects on ectopically expressed HeLa cells by physical shearing, suggesting that a tethering model Bst-2 levels and cellular distribution (29), Vpu clearly down- may not explain all of the functional properties of Bst-2. Taken regulated cell surface expression of Bst-2 in 293T cells (23) and together we conclude that enhancement of virus release by Vpu caused a reduction of endogenous Bst-2 expression in HeLa cells does not, at least in CEMx174 and H9 cells, require cell surface (30). Randomization of the Vpu TM domain abolished the down-modulation or intracellular depletion of Bst-2, nor does it ability of Vpu to overcome the inhibitory effects of Bst-2 on virus entail exclusion of Bst-2 from viral particles. release (23), consistent with our previous data on the importance of the Vpu TM domain for this effect (17). CD317 ͉ host restriction ͉ tetherin ͉ virus assembly Here, we analyzed endogenous Bst-2 with respect to its effect on virus release from HeLa cells, T cells, and macrophages. We pu is an 81-aa type 1 integral membrane protein (1, 2) that noted significant cell type-dependent variation of total intracell- Vhas been shown to cause proteasomal degradation of CD4 ular Bst-2 expression. However, with the exception of Jurkat (3, 4), enhance the release of virions from infected cells (5), and cells, all Vpu-dependent cell types analyzed exhibited surface prevent endocytosis of nascent viral particles (6–11). These expression of Bst-2. Interestingly, whereas Vpu caused a reduction in Bst-2 expression in transfected HeLa cells and long-term latter biological activities of Vpu are mechanistically distinct infected macrophages, HIV-1 replication in CEMx174 and H9 from CD4 degradation and involve different structural domains cells did not result in cell surface down-modulation of Bst-2 or in Vpu. Vpu does not affect assembly of viral particles but a reduction of intracellular Bst-2 levels; yet virus release was facilitates the detachment of virions from the cell surface (5, 6). Vpu-responsive. Surprisingly, Bst-2 was undetectable in virions Several mechanisms of Vpu-mediated virus release have been that had been removed from the surface of cells by physical proposed. First, Vpu has the ability to assemble into a cation- agitation, an observation that is difficult to explain by the specific ion channel (12–16). Randomization of the Vpu trans- tethering model of Bst-2. We conclude that enhancement of membrane (TM) domain did not affect membrane association or virus release by Vpu does not require cell surface down- the ability to induce CD4 degradation but did inhibit Vpu ion modulation or intracellular depletion of Bst-2, nor does it involve channel activity and, at the same time, impaired its ability to exclusion of Bst-2 from viral particles. regulate virus release (14, 17). This suggested a functional correlation between Vpu ion channel activity and virus release. Results However, it remained unclear exactly how a Vpu ion channel Expression of Bst-2 in Primary Cells and Cell Lines Is IFN-Inducible. To activity might trigger increased detachment of particles from the characterize endogenous expression of Bst-2 protein, we com- cell surface. It also became clear that Vpu dependence of virus release was tissue-specific, suggesting the involvement of host factors in Vpu function (18, 19). Several host factors have since Author contributions: E.M., A.J.A., and K.S. designed research; E.M. and A.J.A. performed research; S.K. contributed new reagents/analytic tools; E.M., A.J.A., and K.S. analyzed data; been implicated in the Vpu-sensitive inhibition of virus release, and K.S. wrote the paper. including Task-1 and UBP (20, 21). More recently, two addi- The authors declare no conflict of interest. tional host factors, Bst-2/CD317/tetherin/HM1.24 (22, 23) and 1 calcium-modulating cyclophilin ligand (CAML) (24) have been E.M. and A.J.A. contributed equally to this work. correlated with Vpu-sensitive restriction of HIV-1 virus release. 2To whom correspondence should be addressed at: Viral Biochemistry Section, Laboratory of Molecular Microbiology, National Institute of Allergy and Infectious Diseases, National Our current work focuses on the characterization of Bst-2/ Institutes of Health, Building 4, Room 310, 4 Center Drive MSC 0460, Bethesda, MD CD317/tetherin/HM1.24 (for simplicity referred to as Bst-2 for 20892-0460. E-mail: [email protected]. the remainder of this work). Bst-2 was originally identified as a This article contains supporting information online at www.pnas.org/cgi/content/full/ membrane protein in terminally differentiated human B cells of 0813223106/DCSupplemental. patients with multiple myeloma (25, 26). Bst-2 is a 30- to 36-kDa © 2009 by The National Academy of Sciences of the USA

2868–2873 ͉ PNAS ͉ February 24, 2009 ͉ vol. 106 ͉ no. 8 www.pnas.org͞cgi͞doi͞10.1073͞pnas.0813223106 Downloaded by guest on September 25, 2021 Fig. 1. Expression of Bst-2 in primary cells and cell lines is IFN-inducible. (A) HeLa and 293T cells were cultured in the absence (lanes 1 and 3) or presence (lanes 2 and 4) of IFN␣ (1 ng/mL) for 24 h. Whole-cell lysates were subjected to immunoblotting by using a Bst-2-speciﬁc polyclonal rabbit antibody. The same blot was subsequently probed with a tubulin-speciﬁc antibody (tub) as internal reference for sample loading. (B) HeLa cells, MDM, and PBMC were cultured in the absence of IFN␣ (lanes 1, 3, and 7) or were treated for 24 h with 0.1 ng/mL (lane 4), 1 ng/mL (lanes 2, 5, and 8), or 10 ng/mL (lane 6) IFN␣. For comparison, PBMC were stimulated with CD3/CD28 antibodies as described in Materials and Methods (lane 9). Whole-cell lysates were processed for immunoblotting as described in A except that actin was used as internal reference for sample loading. (C) CEMx174, A3.01, H9, and Jurkat cells were analyzed with or without prior IFN␣ treatment (1 ng/mL) as in A.

pared 293T cells, where Bst-2 is not constitutively expressed, lane 8), although at a relatively lower level than in the other cell MEDICAL SCIENCES with HeLa cells, where Bst-2 expression is constitutive (29). It types examined. We conclude that Bst-2 expression is low to has been reported that upon treatment with IFN␣, 293T cells undetectable in primary T cells and macrophages but can be express Bst-2 mRNA (29). To investigate the effect of IFN␣ on induced by IFN␣ in all cell types examined, including HeLa cells, the expression of endogenous Bst-2 protein, we cultured 293T which already constitutively express high levels of Bst-2. How- and HeLa cells in the presence or absence of IFN␣ for 24 h and ever, activation of PBMC with CD3/CD28 (Fig. 1B, lane 9) did examined expression of Bst-2 by immunoblotting. As expected, not stimulate Bst-2 expression, suggesting that Bst-2 levels Bst-2 was expressed in HeLa cells in the absence of IFN␣ (Fig. remain undetectable in these cells even after T cell activation. 1A, lane 3) but was below the level of detection in 293T cells (Fig. Analysis of Bst-2 expression in various T cell lines revealed that 1A, lane 1). Upon treatment with IFN␣, Bst-2 expression was CEMx174 cells express high levels without IFN␣ stimulation (Fig. induced in 293T cells (Fig. 1A, lane 2). Surprisingly IFN␣ also 1C, lane 1) whereas expression in H9 cells was low but IFN␣- augmented intracellular Bst-2 levels in HeLa cells (Fig. 1A, lane inducible (Fig. 1C, lanes 5 and 6). Bst-2 expression in A3.01 (lanes 4). Of note, Bst-2 appears as multiple bands on the gel with 3 and 4) and Jurkat cells (lanes 7 and 8) was near or below the limit estimated sizes of 25–35 kDa presumably because of glycosyla- of detection even after IFN␣ stimulation. This result is surprising tion as reported (30). FACS analysis revealed low but detectable because both A3.01 and Jurkat T cell lines were found to require surface expression of Bst-2 in untreated 293T cells that was Vpu for efficient virus release (1, 10). FACS analyses confirmed the strongly up-regulated after IFN␣ treatment [supporting infor- absence of cell surface Bst-2 in Jurkat cells whereas H9 cells were mation (SI) Fig. S1A]. IFN␣ treatment also led to increased positive for surface Bst-2 (Fig. S1B). Despite our inability to detect Bst-2 surface expression in HeLa cells, consistent with the Bst-2 in A3.01 cells by immunoblotting, the cells were Bst-2-positive increased intracellular protein levels (Fig. S1A). based on FACS analysis. However, the assay does not allow direct We next examined the level of Bst-2 expression in monocyte- comparison of Bst-2 surface levels across cell types, and it is possible derived macrophages (MDM) and peripheral blood mononu- that surface expression of Bst-2 is low in A3.01 cells despite mean clear cells (PBMC), where we had observed varying levels of Vpu fluorescence intensity similar to that of CEMx174 cells (Fig. S1B). dependence for efficient virus release (31, 32). As shown in Fig. 1B, MDM expressed only low levels of Bst-2 in the absence of Vpu Reduces Bst-2 Levels. Bartee et al. (30) had noted a reduction IFN␣ treatment compared with HeLa cells (Fig. 1B, compare of cellular Bst-2 levels after expression of Vpu from an adeno- lanes 1 and 3). However, Bst-2 expression was induced by IFN␣ virus vector. To explore the possible effects of Vpu on Bst-2 treatment (Fig. 1B, lanes 4–6). Low levels of IFN␣ (0.1 ng/mL) expression or stability, HeLa cells were transfected with pcDNA- were sufficient to induce near-maximal expression of Bst-2 in Vphu, a codon-optimized vector for the expression of NL4-3 MDM (Fig. 1B, lane 4), and increasing amounts of IFN␣ to 1 Vpu (33), together with pCMV-GFP as tracer. Control cells were ng/mL (Fig. 1B, lane 5) or 10 ng/mL (Fig. 1B, lane 6) had little transfected with the GFP vector only. Whole-cell extracts were additional benefit for Bst-2 expression. Of note, even at the prepared 24 h after transfection from unsorted (Fig. 2A, total) highest concentration of IFN␣, Bst-2 levels in MDM remained or GFP-positive [Fig. 2A, GFP(ϩ)] cells and analyzed by im- well below those observed in unstimulated HeLa cells when munoblotting for expression of Bst-2 (Upper)orVpu(Lower). A normalized for actin levels (Fig. 1B, compare lanes 1 and 6). small reduction in Bst-2 steady-state levels was apparent in Also, the overall Bst-2 modification profile seems to differ for unsorted cells (Fig. 2A, lanes 1 and 2) but was more pronounced HeLa cells, MDM, and PBMC. Our analysis of PBMC revealed in the GFPϩ population, which was enriched for Vpu-expressing that levels of Bst-2 in unstimulated cells were below the limit cells (lanes 3 and 4). Transfected cells from Fig. 2A were also of detection (Fig. 1B, lane 7). Upon treatment with IFN␣ used for FACS analysis to study the effect of Vpu on cell surface (1 ng/mL), Bst-2 protein could be detected in PBMC (Fig. 1B, expression of Bst-2 in HeLa cells (Fig. 2B). Signals were gated

Miyagi et al. PNAS ͉ February 24, 2009 ͉ vol. 106 ͉ no. 8 ͉ 2869 Downloaded by guest on September 25, 2021 Fig. 2. Vpu reduces Bst-2 levels in transfected HeLa cells and in HIV-infected macrophages. (A) HeLa cells were transfected with 1 ␮g of pEGFP-N1 (Clontech) with (ϩ) or without (Ϫ)1␮g of pcDNA-Vphu. Total DNA was adjusted to 5 ␮g with empty-vector DNA. Cells were harvested 24 h after transfection and split into two fractions. One fraction was used to prepare whole-cell lysates (total); the other fraction was sorted for GFP-positive cells [GFP(ϩ)] before cell lysis with a FACS Aria (BD Biosciences) as reported in ref. 42. Cell lysates were analyzed by immunoblotting with antibodies to Bst-2 (Upper) or Vpu (Lower). Proteins are identiﬁed on the right. (B) HeLa cells were transfected as in A, stained with Bst-2 antibodies as described in Materials and Methods, and analyzed on a FACS Calibur. Transfected cells were gated for GFP-positive cells. As a control, untransfected HeLa cells were stained with preimmune serum (control). (C) Human MDM were infected with HIV-1 Ada as described in ref. 34 (lane 2). Uninfected macrophages were cultured in parallel as a control (lane 1). Cells were harvested 24 days after infection, and whole-cell lysates were analyzed by immunoblotting for Bst-2, viral capsid protein (CA), or Vpu. The blot was subsequently reprobed with tubulin-speciﬁc antibody to control for sample loading.

for GFPϩ cells. Vpu induced a strong down-regulation of cell of Bst-2 exhibited slight variations, there were no changes that surface Bst-2 within 24 h of transfection consistent with recent could be attributed to Vpu function (third from Top). Further- data (23). Thus, when overexpressed from a codon-optimized more, cell surface levels of Bst-2 were virtually unchanged even vector, Vpu reduced both surface expression and steady-state though Ϸ90% of the cells were HIV-infected (Fig. 3C). Only the levels of endogenous Bst-2. To analyze the effects of Vpu on WT NL4-3-infected culture revealed a modest reduction of Bst-2 Bst-2 expression under more physiological conditions, we com- in Ϸ5% of the p24-positive cell population. Similar results were pared Bst-2 levels of uninfected MDM with those infected for 24 observed in H9 cells, which express lower levels of Bst-2 than days with HIV-1 Ada as described in ref. 34. Consistent with the CEMx174 (Fig. S2B). Accumulation of cell-free viruses was less results from HeLa cells, we noted a reduction of endogenous Vpu-responsive in H9 cells than in CEMx174 cells (Fig. S2A), Bst-2 levels (Fig. 2C). Thus, Vpu has the ability to reduce which is consistent with the lower levels of Bst-2 in these cells steady-state expression of Bst-2 in transiently transfected HeLa (see Fig. 1). FACS data are shown for days 6 and 8 (Fig. S2B). cells and in HIV-infected macrophages. Thus, unlike transfected HeLa cells and infected macrophages, HIV-1-infected CEMx174 and H9 cells did not exhibit Vpu- Replication of HIV-1 in CEMx174 and H9 Cells Is Vpu-Responsive but induced down-regulation of Bst-2. Does Not Necessitate Bst-2 Down-Regulation. To investigate the effect of Vpu on Bst-2 during acute infection of T cells, we first Packaging of Bst-2 into HIV-1 Virions. Bst-2 was proposed to chose the CEMx174 cell line because these cells were highly function as a ‘‘tetherin,’’ physically linking virus to the plasma susceptible to HIV infection and had the highest steady-state membrane and linking viral particles to each other (29). At least level of Bst-2 among the cell lines tested (Fig. 1C). We also the latter function would require Bst-2 to be present in viral wanted to examine further our previous conclusions on the particles. To address this issue, we investigated the possible importance of the Vpu TM domain, which we found to be encapsidation of Bst-2 into WT and vpu-deficient NL4-3 virions important for enhanced virus release, and two phosphoserine produced from Bst-2-expressing HeLa cells (Fig. 4). Mock- residues in the cytoplasmic domain, which we had found to be infected HeLa cells were analyzed in parallel. Cells were de- critical for CD4 degradation but negligible for the virus release tached 24 h later by scraping and pelleted. Virus-containing function of Vpu (17). CEMx174 cells were infected with wild- type NL4-3 (WT), its Vpu-deficient variant NL4-3/Udel (Udel) supernatants were removed (Fig. 4, virus 1). Cells were then (5), or viruses containing a scrambled Vpu TM domain (Urd), suspended in an equal volume of complete culture medium and or having serine-to-asparagine mutations of the two cytoplasmic agitated for 10 s by vortexing as described in ref. 5. Cells were phosphoserine residues S52 and S56 (U26) (17). Virus release then pelleted, and supernatants were removed (Fig. 4, virus 2). was monitored by determining the viral reverse transcriptase Cells and virus-containing supernatants were then processed for activity in the culture supernatants over time (Fig. 3A). As immunoblotting as described in Materials and Methods. Consis- observed previously, WT NL4-3 and NL4-3/U26 released similar tent with our previous data (5), physical shearing of cells levels of RT activity into the culture supernatant, whereas removed more vpu-defective particles than WT virions. Inter- release of NL4-3/Udel and NL4-3/Urd virions was impaired. estingly, however, neither WT nor vpu-deficient viruses removed Peak virus production for all cultures was on day 4, at which time by physical shearing contained detectable amounts of Bst-2 (Fig. cells were removed from the culture and used for immunoblot 4, lanes 8 and 9). There were trace amounts of Bst-2 in the viral analysis to assess Bst-2 steady-state levels (Fig. 3B) and for FACS fraction 1 (Fig. 4, lane 6); however, similar levels of Bst-2 were analysis to determine surface expression of Bst-2 (Fig. 3C). As also found in the mock sample (Fig. 4, lane 4). We conclude that seen in Fig. 3B, all cultures produced similar amounts of viral Bst-2, despite its presence at the cell surface, is excluded from Gag proteins (Top). Also, similar levels of WT Vpu, Vpurd, and WT and vpu-deficient HIV-1 viral particles. The fact that Bst-2 Vpu26 were found in the day 4 cell lysates (second panel from is absent from WT and vpu-deficient particles suggests that Bst-2 Top, lanes 2, 4, and 5). Constant levels of tubulin (Bottom) is excluded from HIV-1 virions through a mechanism that is verified equal sample loading. Interestingly, even though levels unrelated to cell surface down-modulation by Vpu.

2870 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0813223106 Miyagi et al. Downloaded by guest on September 25, 2021 Fig. 3. Vpu does not reduce Bst-2 levels and does not induce cell surface down-regulation of Bst-2 during replication in CEMx174 cells. (A) CEMx174 cells were infected with equal reverse transcriptase units of WT NL4-3, NL4-3/Udel, NL4-3/Urd, and NL4-3/U26 virus stocks produced in 293T cells. Virus replication was

monitored by measuring the virus-associated reverse transcriptase activity in the culture supernatants. Error bars reﬂect mean error from duplicate independent MEDICAL SCIENCES infections. (B and C) On day 4 after infection, Ϸ40% of the cells were removed from the infected cultures and divided into two aliquots. One aliquot of each sample was used for preparation of whole-cell extracts, followed by immunoblotting (B); the other part of the cells was used for Bst-2 cell surface staining followed by staining for intracellular p24 as described in Materials and Methods (C).

Discussion spurred new interest in the mechanism of Vpu-mediated virus The recent identification of Bst-2 as an IFN-inducible host factor release. The Vpu TM domain plays a critical function in the whose expression renders cells restrictive for vpu-defective vi- process of virus release although its exact function remains a ruses (29) and whose intracellular expression and surface pre- mystery. Two known functions of the Vpu TM domain are the sentation are reduced by the HIV-1 Vpu protein (23, 30) has assembly into homooligomeric complexes (35) and the formation of ion-conductive membrane pores (reviewed in ref. 36). However, it is unclear whether the interference with Bst-2 function requires Vpu oligomerization. Vpu does traffic to the cell surface (37), and it is conceivable that it regulates virus release from the cell surface although the site of Vpu function has yet to be established. The model proposed by Neil et al. (29) suggesting that Bst-2 functions as a tether to prevent the release of vpu-deficient virions from the cell surface has a lot of appeal. Our observation of extensive cell type-specific variations in intracellular Bst-2 levels versus relatively constant surface expression in most cell types tested is consistent with a model that involves cell surface Bst-2 in inhibiting virus release. One possible mechanism of Vpu would then be to induce surface down-modulation of Bst-2. Indeed, we confirmed data first reported by Van Damme et al. (23) that given enough time as is the case in long-term infected macrophages or when overexpressed in HeLa cells, Vpu down-modulates surface Bst-2. We are currently investigating the possibility that this mechanism is operative in primary CD4ϩ cells, which express very low levels of Bst-2. However, the observation that Bst-2 surface expression was little if at all altered in the course of a rapidly spreading Fig. 4. Packaging of endogenous Bst-2 into HIV-1 virions. HeLa cells were infection of CEMx174 and H9 cells suggests that cell surface transfected with 5 ␮g each of WT pNL4-3 (lanes 2, 5, and 8) or NL4-3/Udel (lanes down-modulation of Bst-2 may not be the only function of Vpu. 3, 6, and 9). A mock-transfected culture was analyzed in parallel (lanes 1, 4, and Our inability to detect Bst-2 in virus particles even when Bst-2 7). Viral fractions (virus 1 and virus 2) were collected as described in the text. was not down-modulated by Vpu (e.g., in vpu-deficient virions) Cell lysates and concentrated viral supernatants were subjected to immunoblotting with antibodies to Bst-2 (Top). Viral capsid protein (CA) was identiﬁed and when particles were removed from the cell surface by with an HIV-positive patient serum (Middle). The blot was reprobed with an physical shearing argues against a simple tethering function of antibody to tubulin (tub) as internal reference (Bottom). Proteins are identi- Bst-2. Also, Bst-2 identified in the “virus-1” fraction in Fig. 4 did ﬁed on the right. not copurify with the viral fraction in linear sucrose gradients,

Miyagi et al. PNAS ͉ February 24, 2009 ͉ vol. 106 ͉ no. 8 ͉ 2871 Downloaded by guest on September 25, 2021 suggesting that this Bst-2 was nonspecifically released from the Tissue Culture and Transfections. Cell maintenance and transfections were cells. The question remains of course how many Bst-2 molecules performed by using standard techniques as detailed in SI Materials and would be required to physically tether a mature virion to the cell Methods. surface. We were able to detect nonspecifically secreted Bst-2 in Preparation of MDM and PBMC and Stimulation with IFN␣. PBMC were isolated our experiment (Fig. 4, lanes 4 and 6). Therefore, we believe that from the leukophoresed blood of HIV-seronegative donors after gradient the sensitivity and specificity of our antibody are excellent; separation with lymphocyte separation medium (Organon Teknika–Cappel). however, we cannot rule out the presence of miniscule amounts Cells (2 ϫ 106 each) were cultured for 24 h in 1.5 mL of RPMI medium 1640 (10% of Bst-2 in viral particles that are below the limit of detection. FCS, 4 mM L-glutamine, 2 mM sodium pyruvate, 50 ␮g/mL gentamicin, and 50 Until these questions can be answered, alternative mechanisms ␮M 2-mercaptoethanol) either alone or in the presence of 1 ng/mL IFN␣-2c or 10 ng/mL anti-CD3 plus 5 ␮g/mL anti-CD28 (both BD Biosciences PharMingen) of Bst-2 and Vpu function may need to be considered in addition for 24 h. Cells were harvested 24 h later and lysed in 300 ␮L of sample buffer. to the tetherin model. It is possible that Bst-2 functions in Boiled samples (50 ␮L each) were processed for immunoblotting. concert with other host factor(s) (e.g., CAML). This would Monocytes from normal human donors were prepared as described in ref. explain why silencing of either Bst-2 or CAML in HeLa cells 40 to allow differentiation into MDM as reported in ref. 34. IFN␣-2c was added resulted in Vpu-independent release of particles (24, 29). It will as indicated in the text and detailed in SI Materials and Methods. therefore be interesting to investigate the potential functional interplay of Bst-2 and CAML. We confirmed that Vpu has the Virus Preparation. Virus stocks for infection were prepared by transfecting 293T cells with appropriate plasmid DNAs as described in ref. 41 and SI ability to reduce intracellular Bst-2 in HeLa cells. However, it Materials and Methods. remains to be shown whether the reduction of Bst-2 levels in transfected HeLa cells and infected macrophages is caused by Immunoblotting. For immunoblot analysis of intracellular proteins, whole-cell protein degradation or is an indirect consequence of Vpu lysates were prepared as described in ref. 41 and detailed in the SI Materials dysregulation of NF-␬B, both of which require functional phos- and Methods. phoserine motifs in Vpu (38). It is also important to note that reduced secretion of vpu-deficient viruses cannot be equated FACS Analysis. Cells were washed twice with ice-cold 20 mM EDTA-PBS, followed by two washes in ice-cold 1% BSA-PBS. Cells were blocked for 10 min with reduced virus spread. In fact, replication kinetics of WT and with 50 ␮g of mouse IgG (Millipore). Cells were incubated with primary vpu-deficient NL4-3 are generally quite similar irrespective of antibody (␣-Bst-2) for 30 min at room temperature, washed twice with ice-cold the cell type used (Fig. 3A and Fig. S2A), and there is strong 1% BSA-PBS followed by the addition of allophycocyanin-conjugated anti- evidence that vpu-deficient viruses are capable of spreading rabbit IgG secondary antibody (Jackson ImmunoResearch Laboratories) in 1% cell-to-cell despite reduced levels of cell-free virus. Until the true BSA-PBS. Incubation was for 30 min at room temperature in the dark. Cells significance of Vpu-enhanced virus release is fully understood were then washed twice with ice-cold 1% BSA-PBS and fixed with 1% para- formaldehyde in PBS. Finally, cells were analyzed on a FACS Calibur (BD for in vivo dissemination of HIV-1, the question of whether Bst-2 Biosciences Immunocytometry Systems). Data analysis was performed by using should be added to the ever-growing list of antiviral host factors CellQuest Pro (BD Biosciences) and Flow Jo (Tree Star). For staining of intra- remains open. cellular p24, cells were stained with primary and secondary antibodies as described. Before the fixation step, cells were permeabilized for 10 min at Materials and Methods room temperature in the dark by using 1ϫ FACS permeabilization buffer Plasmids. All plasmids have been described: pNL4-3 (39), pNL4-3/Udel (5), (Becton Dickinson). Cells were then washed twice with ice-cold 1% BSA-PBS ␣ pNL4-3/Urd and pNL4-3/U26 (17), and pcDNA-Vphu (33). and stained for 30 min at room temperature in the dark with -p24-Gag antibody (KC57 RD1; Beckman Coulter). Cells were washed twice with ice-cold 1% BSA-PBS, fixed, and analyzed as above. Antisera. Anti-Bst-2 antiserum was elicited in rabbits by using a bacterially expressed MS2-Bst-2 fusion protein composed of amino acids 1–91 of the MS2 ACKNOWLEDGMENTS. We thank Mohammad Khan, Ritu Goila-Gaur, Robert replicase and amino acids 41–162 of Bst-2, generating a polyclonal antibody Walker, and Ronald Willey for helpful discussion and critical comments. We against the extracellular portion of Bst-2. Polyclonal anti-Vpu serum (rabbit), thank Jason Brenchley for help with cell sorting, Bernard Lafont for help with directed against the hydrophilic C-terminal cytoplasmic domain of Vpu FACS, and Kathleen Clouse and Linda Tiffany (Center for Drug Evaluation and expressed in Escherichia coli (35), was used for detection of Vpu. Serum from Research, Beltsville, MD) for providing monocyte-derived macrophages. ␣ an HIV-positive patient was used to detect HIV-1-specific capsid (CA) and IFN -2c was a generous gift from Kathryn C. Zoon (National Institute of gag Allergy and Infectious Diseases, Bethesda, MD). This work was supported by a Gag precursor (Pr55 ) proteins. Tubulin and actin were identified by using grant from the National Institutes of Health Intramural AIDS Targeted Anti- monoclonal antibodies to ␣-tubulin and actin, respectively (both from viral Program (to K.S.) and by the Intramural Research Program of the National Sigma–Aldrich). Institute of Allergy and Infectious Diseases, National Institutes of Health.

1. Strebel K, Klimkait T, Martin MA (1988) A novel gene of HIV-1, vpu, and its 16- 11. Van Damme N, Guatelli J (2008) HIV-1 Vpu inhibits accumulation of the envelope kilodalton product. Science 241:1221–1223. glycoprotein within clathrin-coated, Gag-containing endosomes. Cell Microbiol 2. Cohen EA, Terwilliger EF, Sodroski JG, Haseltine WA (1988) Identification of a protein 10:1040–1057. encoded by the vpu gene of HIV-1. Nature 334:532–534. 12. Piller SC, Ewart GD, Premkumar A, Cox GB, Gage PW (1996) Vpr protein of human 3. Willey RL, Maldarelli F, Martin MA, Strebel K (1992) Human immunodeficiency virus immunodeficiency virus type 1 forms cation-selective channels in planar lipid bilayers. type 1 Vpu protein induces rapid degradation of CD4. J Virol 66:7193–7200. Proc Natl Acad Sci USA 93:111–115. 4. Willey RL, Maldarelli F, Martin MA, Strebel K (1992) Human immunodeficiency virus 13. Ewart GD, Sutherland T, Gage PW, Cox GB (1996) The Vpu protein of human type 1 Vpu protein regulates the formation of intracellular gp160–CD4 complexes. immunodeficiency virus type 1 forms cation- selective ion channels. J Virol 70:7108– J Virol 66:226–234. 7115. 5. Klimkait T, Strebel K, Hoggan MD, Martin MA, Orenstein JM (1990) The human 14. Schubert U, et al. (1996) Identification of an ion channel activity of the Vpu transmem- immunodeficiency virus type 1-specific protein Vpu is required for efficient virus brane domain and its involvement in the regulation of virus release from HIV-1- maturation and release. J Virol 64:621–629. infected cells. FEBS Lett 398:12–18. 6. Neil SJ, Eastman SW, Jouvenet N, Bieniasz PD (2006) HIV-1 Vpu promotes release and 15. Marassi FM, et al. (1999) Correlation of the structural and functional domains in the prevents endocytosis of nascent retrovirus particles from the plasma membrane. PLoS membrane protein Vpu from HIV-1. Proc Natl Acad Sci USA 96:14336–14341. Pathog 2:e39. 16. Ma C, et al. (2002) Expression, purification, and activities of full-length and truncated 7. Harila K, et al. (2006) Vpu and Tsg101 regulate intracellular targeting of the human versions of the integral membrane protein Vpu from HIV-1. Protein Sci 11:546–557. immunodeficiency virus type 1 core protein precursor Pr55gag. J Virol 80:3765–3772. 17. Schubert U, et al. (1996) The two biological activities of human immunodeficiency virus 8. Harila K, Salminen A, Prior I, Hinkula J, Suomalainen M (2007) The Vpu-regulated type 1 Vpu protein involve two separable structural domains. J Virol 70:809–819. endocytosis of HIV-1 Gag is clathrin-independent. Virology 369:299–308. 18. Varthakavi V, Smith RM, Bour SP, Strebel K, Spearman P (2003) Viral protein U 9. Strebel K, Klimkait T, Maldarelli F, Martin MA (1989) Molecular and biochemical counteracts a human host cell restriction that inhibits HIV-1 particle production. Proc analyses of human immunodeficiency virus type 1 Vpu protein. J Virol 63:3784–3791. Natl Acad Sci USA 100:15154–15159. 10. Terwilliger EF, Cohen EA, Lu YC, Sodroski JG, Haseltine WA (1989) Functional role of 19. Sakai H, Tokunaga K, Kawamura M, Adachi A (1995) Function of human immunode- human immunodeficiency virus type 1 Vpu. Proc Natl Acad Sci USA 86:5163–5167. ficiency virus type 1 Vpu protein in various cell types. J Gen Virol 76:2717–2722.

2872 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0813223106 Miyagi et al. Downloaded by guest on September 25, 2021 20. Hsu K, Seharaseyon J, Dong P, Bour S, Marban E (2004) Mutual functional destruction 32. Schubert U, Bour S, Willey RL, Strebel K (1999) Regulation of virus release by the of HIV-1 Vpu and host TASK-1 channel. Mol Cell 14:259–267. macrophage-tropic human immunodeficiency virus type 1 AD8 isolate is redundant 21. Callahan MA, et al. (1998) Functional interaction of human immunodeficiency virus and can be controlled by either Vpu or Env. J Virol 73:887–896. type 1 Vpu and Gag with a novel member of the tetratricopeptide repeat protein 33. Nguyen KL, et al. (2004) Codon optimization of the HIV-1 vpu and vif genes stabilizes family. J Virol 72:5189–5197. their mRNA and allows for highly efficient Rev-independent expression. Virology 22. Neil SJ, Sandrin V, Sundquist WI, Bieniasz PD (2007) An interferon-␣-induced tethering 319:163–175. mechanism inhibits HIV-1 and Ebola virus particle release but is counteracted by the 34. Miyagi E, et al. (2008) APOBEC3G-independent reduction in virion infectivity during HIV-1 Vpu protein. Cell Host Microbe 2:193–203. long-term HIV-1 replication in terminally differentiated macrophages. Virology 23. Van Damme N, et al. (2008) The interferon-induced protein BST-2 restricts HIV-1 release 379:266–274. and is down-regulated from the cell surface by the viral Vpu protein. Cell Host Microbe 35. Maldarelli F, Chen MY, Willey RL, Strebel K (1993) Human immunodeficiency virus 3:245–252. type 1 Vpu protein is an oligomeric type I integral membrane protein. J Virol 24. Varthakavi V, et al. (2008) Identification of calcium-modulating cyclophilin ligand as a 67:5056–5061. human host restriction to HIV-1 release overcome by Vpu. Nat Med 14:641–647. 36. Strebel K (2007) HIV accessory genes vif and vpu. Adv Pharmacol 55:199–232. 25. Goto T, et al. (1994) A novel membrane antigen selectively expressed on terminally 37. Bour S, Perrin C, Strebel K (1999) Cell surface CD4 inhibits HIV-1 particle release by differentiated human B cells. Blood 84:1922–1930. interfering with Vpu activity. J Biol Chem 274:33800–33806. 26. Ohtomo T, et al. (1999) Molecular cloning and characterization of a surface antigen 38. Akari H, Bour S, Kao S, Adachi A, Strebel K (2001) The human immunodeficiency virus preferentially overexpressed on multiple myeloma cells. Biochem Biophys Res Com- type 1 accessory protein Vpu induces apoptosis by suppressing the nuclear factor mun 258:583–591. ␬B-dependent expression of antiapoptotic factors. J Exp Med 194:1299–1311. 27. Ishikawa J, et al. (1995) Molecular cloning and chromosomal mapping of a bone 39. Adachi A, et al. (1986) Production of acquired immunodeficiency syndrome-associated marrow stromal cell surface gene, BST2, that may be involved in pre-B-cell growth. retrovirus in human and nonhuman cells transfected with an infectious molecular Genomics 26:527–534. clone. J Virol 59:284–291. 28. Kupzig S, et al. (2003) Bst-2/HM1.24 is a raft-associated apical membrane protein with 40. Gruber MF, Weih KA, Boone EJ, Smith PD, Clouse KA (1995) Endogenous macrophage an unusual topology. Traffic 4:694–709. CSF production is associated with viral replication in HIV-1-infected human monocyte- 29. Neil SJ, Zang T, Bieniasz PD (2008) Tetherin inhibits retrovirus release and is antago- derived macrophages. J Immunol 154:5528–5535. nized by HIV-1 Vpu. Nature 451:425–430. 41. Kao S, et al. (2003) The human immunodeficiency virus type 1 Vif protein reduces 30. Bartee E, McCormack A, Fruh K (2006) Quantitative membrane proteomics reveals new intracellular expression and inhibits packaging of APOBEC3G (CEM15), a cellular cellular targets of viral immune modulators. PLoS Pathog 2:e107. inhibitor of virus infectivity. J Virol 77:11398–11407. 31. Schubert U, Clouse KA, Strebel K (1995) Augmentation of virus secretion by the human 42. Goila-Gaur R, Khan MA, Miyagi E, Strebel K (2009) Differential sensitivity of ‘‘old’’ immunodeficiency virus type 1 Vpu protein is cell type-independent and occurs in versus ‘‘new’’ APOBEC3G to human immunodeficiency virus type 1 Vif. J Virol 83:1156– cultured human primary macrophages and lymphocytes. J Virol 69:7699–7711. 1160. MEDICAL SCIENCES

Miyagi et al. PNAS ͉ February 24, 2009 ͉ vol. 106 ͉ no. 8 ͉ 2873 Downloaded by guest on September 25, 2021