Carboxyl Terminus of hVIP/mov34 Is Critical for HIV-1-Vpr Interaction and Glucocorticoid-mediated Signaling*

Human immunodeficiency virus, type 1 (HIV-1) vpr is a highly conserved gene among lentiviruses. The diverse functions of Vpr support interactions of this HIV accessory protein with host cell partners of important path-ways. hVIP/mov34 ( h uman V pr I nteracting P rotein) is one of these identified Vpr ligands. hVIP is a 34-kDa member of the eIF3 family that is vital for early embryonic development in transgenic mice and important in cell cycle regulation. Its interaction with Vpr, however, is not yet clearly defined. Therefore, we constructed a panel of deletion mutants of this cytoplasmic cellular ligand to map the protein domain that mediates its interaction with Vpr. We observed that the carboxyl-ter-minal region of hVIP is critical for its interaction with Vpr. In the absence of Vpr or HIV infection, full-length hVIP is expressed in the cytoplasm. The cytoplasmic localization pattern of full-length hVIP protein, how-ever, is shifted to a clear nuclear localization pattern in cells expressing both hVIP and Vpr. In contrast, Vpr did not alter the localization pattern of hVIP mutants, which have their carboxyl-terminal domain deleted. The movement of hVIP supported prior work that suggested that Vpr triggers activation of the GR receptor complex. In fact, we also observed that dexamethasone moves hVIP into the nucleus and that glucocorticoid antagonists inhibit

Human immunodeficiency virus, type 1 (HIV-1) vpr is a highly conserved gene among lentiviruses. The diverse functions of Vpr support interactions of this HIV accessory protein with host cell partners of important pathways. hVIP/mov34 (human Vpr Interacting Protein) is one of these identified Vpr ligands. hVIP is a 34-kDa member of the eIF3 family that is vital for early embryonic development in transgenic mice and important in cell cycle regulation. Its interaction with Vpr, however, is not yet clearly defined. Therefore, we constructed a panel of deletion mutants of this cytoplasmic cellular ligand to map the protein domain that mediates its interaction with Vpr. We observed that the carboxyl-terminal region of hVIP is critical for its interaction with Vpr. In the absence of Vpr or HIV infection, full-length hVIP is expressed in the cytoplasm. The cytoplasmic localization pattern of full-length hVIP protein, however, is shifted to a clear nuclear localization pattern in cells expressing both hVIP and Vpr. In contrast, Vpr did not alter the localization pattern of hVIP mutants, which have their carboxyl-terminal domain deleted. The movement of hVIP supported prior work that suggested that Vpr triggers activation of the GR receptor complex. In fact, we also observed that dexamethasone moves hVIP into the nucleus and that glucocorticoid antagonists inhibit this effect. Interestingly, the expression of an hVIP carboxyl-terminal mutant, which is not responsive to Vpr, is also not responsive to dexamethasone. These data illustrate that the carboxyl-terminal domain of hVIP is critical for mediating hVIP-Vpr interaction as well as for its glucocorticoid response. These results support the view that hVIP is a member of the complex array of nucleocytoplasmic shuttling proteins that are regulated by HIV infection and glucocorticoids.
HIV-1 1 (human immunodeficiency virus, type 1) encodes a 14-kDa accessory protein, Vpr (viral protein R) (1,2), that has been the subject of much investigation. HIV-1 vpr, although dispensable for viral replication in T-cell lines, is required for efficient replication in primary monocytes/macrophages (1,3). Vpr is packaged inside virions through its interaction with the Pr55gag precursor at a molar amount equivalent to that of Gag, suggesting an important role for Vpr in early infection (4 -6). Throughout the evolution of lentiviruses, the vpr gene has been highly conserved, suggesting its importance in viral pathogenesis (7)(8)(9).
Previous studies demonstrate that Vpr can exert several independent functions on host cells. These include growth arrest and alteration in host cell transcription (10), transactivation of the HIV-1 long terminal repeat (7,11), efficient viral replication in mononuclear phagocytes (12), cell cycle arrest at the G2/M check point (13)(14)(15), translocation of pre-integration complex into the nucleus of non-dividing cells (16), tumor growth arrest (17), stimulation of apoptosis (18 -21), suppression of immune activation (18,22), enhanced viral expression (23,24), and disruption in nuclear pore architecture (25). Several cellular proteins have been identified as interacting with Vpr (11, 26 -32). Their role in the function of Vpr, however, is still being investigated. Using a yeast two-hybrid system, we cloned the cDNA of a Vpr-interacting cellular factor, termed human Vpr Interacting Protein (hVIP/mov34) (32), which has complete homology with a reported member of the eIF3 complex (33). eIF3 is a large multimeric complex that regulates transcriptional events and is essential for G1/S and G2/M phase progression through the cell cycle. This cytoplasmically localized protein (34) has been linked to the cell cycle arrest function of Vpr. However, the additional biology of hVIP-Vpr interaction has not yet been determined. Here we demonstrate that Vpr interactions with hVIP are completely dependent on the carboxyl-terminal region of mov34. We further find that hVIP is a GR-responsive protein. Experimental results strongly suggest that hVIP is associated with the activated glucocorticoid receptor complex. The carboxyl-terminal region of this ligand appears to be critical in governing its interaction with Vpr as well as in mediating the glucocorticoid-induced effect observed on hVIP.

EXPERIMENTAL PROCEDURES
Cells and Reagents-HeLa and CV-1 cell lines were obtained from ATCC (Manassas, VA). Culture media and other standard tissue culture reagents were obtained from Invitrogen. The stock of recombinant vaccinia (vFT7-3) virus containing the T7 RNA polymerase gene was also received from ATCC. Polyclonal antisera for immunological detection of Vpr were described earlier (35). Anti-His-tagged (carboxyl terminus) mouse monoclonal antibody and rabbit polyclonal anti-glucocorticoid receptors were purchased from Clontech and Santa Cruz Biotechnology, Inc (Santa Cruz, CA), respectively. Dexamethasone and mifepristone (RU486) were purchased from Biomol Research Laboratories, Inc. (Plymouth Meeting, PA).
Plasmids and Cloning Strategies-The pchVIP expression plasmid that encodes a chimeric hVIP with a polyhistidine tag at its carboxyl * This work was supported by a National Institutes of Health grant (to D. B. W.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
‡ To whom correspondence should be addressed: Dept. of Pathology and Laboratory Medicine, University of Pennsylvania School of Medicine, 422 Curie Blvd., 505 Stellar-Chance Laboratories, Philadelphia, PA 19104. Tel.: 215-662-2352; Fax: 215-573-9436; E-mail: dbweiner@ mail.med.upenn.edu. 1 The abbreviations used are: HIV-1, Human immunodeficiency virus, type 1; GR, glucocorticoid receptor; GRE, glucocorticoid-responsive element; DAPI, 4,6-Diamidino-2-phenylindole; Vpr, viral protein R; hVPR, human VPR-interacting protein; FITC, fluorescein isothiocyanate; eIF, eukaryotic initiation factor. terminus was previously described (32,34). pcVpr vector for expression of Vpr was constructed using the PCR-amplified Vpr sequence of the pNL43 allele of HIV-1, which contains an open reading frame encoding the sequence of the wild-type Vpr. The resulting PCR fragment was appropriately digested and cloned into HindIII-XhoI sites of the pcDNA3.1 vector (Invitrogen). Similarly, a vector expressing hGR␣, pcGR, was constructed by sub-cloning a PCR-amplified GR coding region from the plasmid, pRShGR␣, that was purchased from ATCC. The pool of PCR fragments corresponding to the GR-coding region was digested appropriately and inserted into KpnI-XbaI-digested pcDNA3.1 expression vector. The integrity of these constructs was confirmed by automated sequencing.
Construction of Mutants-The deletion mutants were generated by PCR-based mutagenesis. Briefly, selected portions of the hVIP open reading frame were amplified by Taq polymerase (Roche Molecular Biochemicals) from pchVIP plasmid as template using appropriate primers. Fig. 2 reveals the schematic plan for the construction of the mutants. Primers for the construction of the hVIP-NM mutant are 5Ј-CTCCAGATGGAGGTGGATG-3Ј and 5Ј-GCATAGGCCACCCGGCA-GGAT-3Ј; primers for making the hVIP-MC mutant are 5Ј-CCACCTG-ACCCCTCGGACAT-3Ј and 5Ј-GCATAGGCCACCC-GGCAGGAT-3Ј; primers for amplifying the carboxyl-terminal region alone to make the hVIP-C mutant are 5Ј-ATGACAGCAA-CAGGCAGTGG-3Ј and 5Ј-AGC-CCGCGCATTCTCCTGC-3Ј. To construct an hVIP mutant (hVIP-NC) that has a deletion of its central core domain, an overlap-PCR approach was used. Briefly, in the first set of PCR, the amino-and carboxylterminal regions were amplified in separate reactions. During this step, a reverse primer for the amino-terminal amplification was designed to have overlapping bases corresponding to the 3Ј-end of the carboxyl-terminal region. In the second set of reactions, the aminoand carboxyl-terminal-amplified fragments from the first set of PCR were used as templates to amplify the fusion fragment, using the forward primer of the amino-terminal region and the reverse primer of the carboxyl-terminal domain of hVIP. Because the 5Ј-end of the amino-terminal had overlapping bases corresponding to the 3Ј-end of the carboxyl terminus, the second set of PCR reactions resulted in the fusion of the amino-and carboxyl-terminal fragments in-frame. The amplified DNA fragments were purified from agarose gel and inserted into eukaryotic TOPO-cloning expression vector pCDNA3.1/V5/His to generate fusion hVIP proteins containing polyhistidine tags at their carboxyl termini, according to the instructions provided by the supplier. Transformants were screened from LB-agar plates containing ampicillin (50 g/ml) plates, and all the mutant FIG. 1. Co-expression of HIV-1 Vpr and its ligand in HeLa cells. HeLa cells were transfected with either pchVIP or pcVpr expression constructs individually or together and subjected to indirect fluorescent assay. Perinuclear pattern of Vpr localization (A and B) and the cytoplasmic localization of its ligand factor, hVIP, by themselves are shown (E and F). During co-expression of both the proteins together (I-L), the localization pattern of hVIP is altered from the cytoplasmic region completely to the nucleus (J and L). The transfected cells were fixed and permeabilized appropriately and treated with polyclonal anti-Vpr antibody and anti-His tag antibody; Vpr was stained with rhodamine-conjugated secondary antibody and hVIP was stained with FITCconjugated secondary antibody. As a negative control, the cells were transfected with pcDNA3.1 vector and stained individually with anti-Vpr antibody (C and D) as well as with anti-His tag antibodies (G and H), followed by incubation with FITC-and Texas Red-conjugated secondary antibodies, respectively. Neither of these antibodies resulted in any significant staining pattern in the mock-transfected cells. The cells were also treated with DAPI to counterstain the nuclei with a typical bright blue color (A, C, E, G, I).
The cells were visualized under digital fluorescent microscopy.

FIG. 2. Schematic panel of hVIP mutants and their localization patterns.
An outline of the hVIP mutants and the exact localization pattern of these mutants with or without co-expression of HIV-1 Vpr is listed. Polyhistidine (6 ϫ his) epitope is fused in-frame to the carboxyl-terminal end of the mutants. Except for one mutant that is lacking the specific carboxyl-terminal region, all the hVIP proteins are interacted with and translocated into the nucleus by Vpr. constructs were sequenced by automated sequencing to verify the integrity of the mutations.
In Vitro Binding Assay-Using the TNT-coupled in vitro transcription/translation system (Promega Corp., Madison, WI), 35 S-labeled protein products were generated using plasmids containing hVIP constructs and the vpr gene as templates. The reaction mix was prepared according to the instructions supplied by the manufacturer, and the reaction was carried out at 30°C for 1 h. Equal amounts of in vitrotranslated 35 S-labeled hVIP and HIV-1 Vpr were mixed and incubated for 60 min at 4°C in a binding buffer containing 25 mM Hepes (pH 7.9), 150 mM KCl, 0.1% Nonidet P-40, 5% glycerol, 0.5 mM dithiothreitol, and 0.4 mM phenylmethylsulfonyl fluoride. Respective antibodies were added to each tube with 150 l of binding buffer and incubated for 90 min at 4°C. Approximately 5 mg of protein A-Sepharose beads were added to each immunoprecipitation reaction from a freshly prepared stock solution (100 mg/ml), and the samples were incubated at 4°C for 90 min in a rotating shaker. The beads were washed three times with the binding buffer and finally suspended in 2ϫ SDS sample buffer. The immunoprecipitated protein complexes were eluted from the Sepharose beads by being briefly boiled and then resolved using SDS/PAGE (15%) gels. The gel was fixed, treated with a 1 M sodium salicylate solution, and dried in a gel drier (Bio-Rad). The dried gel was exposed overnight to x-ray film and developed using an automated developer (Kodak). Radiolabeled GR protein sample was also prepared as described above. In vitro binding assay was performed to examine the physical interaction between hVIP and GR.
Indirect Immunofluorescent Assay-The localization patterns of hVIP proteins in cells with or without the coexpression of Vpr protein were monitored by an indirect immunofluorescent assay approach. HeLa cells grown in slide chambers (BD PharMingen) were transfected with appropriate plasmids as outlined above. For 1 h prior to the transfection, the cells were infected with a recombinant vaccinia virus (vFT7-3) that expresses recombinant RNA polymerase and then transfected overnight with either hVIP or Vpr expression vectors alone or in combination. The cells were fixed on slides using 4% paraformaldehyde and incubated with appropriate primary antibodies for 45 min. Rabbit anti-HIV-1 Vpr and mouse anti-His tag (1:100 dilution) were used as primary antibodies. Following incubation with primary antibodies, the slides were rinsed with phosphate-buffered saline and incubated with fluorescein isothiocyanate (FITC) or rhodamine-conjugated secondary antibodies for 45 min. DAPI (Sigma) was used to stain the nuclei with characteristic blue color. The slides were mounted with mounting media containing anti-fading reagent (Molecular Probes, Inc, Eugene, OR) for fluorescent microscopy, and the pictures were taken with appropriate filters.

Co-expression of HIV-Vpr and Its Ligand in Transfected
Cells-In vitro synthesis of 35 S-labeled protein samples prepared from pchVIP and pcVpr were examined by gel electrophoresis and found to correspond to the appropriate molecular mass of the predicted proteins, thereby confirming the integrity of the constructs. Specifically, Vpr generated a 14-kDa protein, whereas hVIP moves at the expected 34,000. Staining pchVIPtransfected HeLa cells with a monoclonal anti-His tag antibody reveals a cytoplasmic localization (Fig. 1, C and D) for the full-length protein as previously described (34). In contrast, the expression of vpr in the transfected cells reveals typical nuclear rim staining (Fig. 1, A and B), as previously reported (32,35). Interestingly, co-expression of Vpr with hVIP significantly alters the localization pattern of this Vpr ligand. The cytoplasmic localization pattern of full-length hVIP protein was shifted to a clear nuclear localization pattern in cells expressing both hVIP or Vpr constructs overlapping with Vpr, while Vpr still maintains its perinuclear staining pattern (Fig. 1, E-H). This result supports that Vpr interferes with the localization pattern of hVIP.
Expression and Localization of hVIP Mutants-In Fig. 2, a schematic plan on the construction of specific deletion mutants of hVIP is illustrated. During their construction, a polyhistidine tag was fused to the carboxyl terminus of the hVIP molecules for characterization studies. The expression of hVIP mutant constructs was initially confirmed by SDS gel analysis hVIP/mov34 and Glucocorticoid Signaling of 35 S-labeled in vitro-translated protein products. The mobility of the individual mutant molecules corresponds to their gene sequence. Monoclonal anti-His tag and polyclonal Vpr antibodies were used to monitor the expression of the hVIP constructs and the vpr gene, respectively. Interestingly, deletion mutagenesis induced significant changes in the localization pattern of some of the mutant molecules. Although the full-length hVIP is localized in the cytoplasm (Fig. 3A, hVIP-WT), hVIP with a deleted carboxyl-terminal portion (Fig. 3F, hVIP-NM) as well as the mutant that has the central core domain deleted also localized in the cytoplasm (Fig. 3K, hVIP-NC). Deletion of the amino-terminal domain resulted in nuclear localization (Fig. 3P), and the expression of the carboxyl-terminal portion alone (Fig. 3U, hVIP-C) revealed a cytoplasmic localization.
Interestingly, mutants of the hVIP molecule differ in terms of their localization with the presence of Vpr protein. In Fig. 3, the full-length hVIP moves to the nucleus upon co-transfection with Vpr-expressing plasmid (Fig. 3, B-E). The mutant (hVIP-NM), which has a deletion of the carboxyl-terminal region, revealed a cytoplasmic localization when expressed alone; this localization pattern was not altered when this mutant was expressed in the presence of Vpr (Fig. 3, G-J). Thus, unlike the full-length hVIP, expression of vpr does not result in the translocation of this mutant completely from the cytoplasm into the nucleus. Together with the expression of Vpr, the cytoplasmic localization of the hVIP mutant that has its middle core domain deleted is altered to localize in the nuclear region (Fig. 3, L-O). The amino-terminal-deleted mutant (hVIP-MC) was expressed clearly in the nucleus, regardless of the co-expression of vpr; hence, the presence of Vpr did not alter the localization pattern of this mutant (Fig. 3, Q-T). Interestingly, the carboxyl-terminal region is localized in the cytoplasm and is moved into the nucleus by the expression of Vpr (Fig. 3, V-Y).
Interaction between hVIP Mutants and Vpr in Vitro-To confirm these relationships, a direct in vitro binding assay was performed. The interaction between hVIP molecules and Vpr was determined by immunoprecipitation followed by SDS-PAGE. Vpr was coimmunoprecipitated with the full-length hVIP molecule as previously reported. Additionally, four hVIP mutants also bound directly to Vpr. Only the mutant lacking the carboxyl-terminal region failed to interact with Vpr (Fig. 4). As a repeat control, anti-Vpr sera also was used to immunoprecipitate the complexes. Again the precipitation pattern for hVIP was consistent with the pattern revealed by the anti-His tag studies (data not shown). This analysis and the immunofluorescence analysis demonstrate that interaction of Vpr is distinct with hVIP and appears to be mediated through the carboxyl-terminal domain of hVIP.
Mifepristone Blocks the Vpr-induced Nuclear Entry of hVIP-We have reported hVIP as a potential Vpr ligand and demonstrated its role in cell cycle regulation as antisense of this gene-induced cell cycle arrest at the G2/M phase (32), the stage where the expression of Vpr induces growth arrest (13)(14)(15). To gain further insight into the biology of hVIP-Vpr interaction, we investigated the effect of Vpr antagonists on hVIP function. Glucocorticoids have been demonstrated to mimic the effects of Vpr; mifepristone has been shown to revert these effects of Vpr (18,22,36,37). With this background, we examined whether mifepristone could block the nuclear translocation of hVIP induced by Vpr in cells. HeLa cells were cotransfected with pchVIP and pcVpr expression plasmids, and the localization pattern of hVIP was monitored under the influence of mifepristone. As witnessed previously, in the untreated cells hVIP was localized in the nucleus of the cells co-expressing both hVIP and Vpr. (Fig. 5, A-D). Interestingly, in the mifepristone-treated cells, hVIP was observed in the cytoplasm despite the co-expression of Vpr (Fig. 5, E-H). This result clearly demonstrates that mifepristone inhibits the translocation of hVIP induced by the expression of Vpr.
Glucocorticoids and hVIP Expression-The effects of Vpr on viral replication and host cell biology are mediated in part through the GR complex (22,36). The relationship of hVIP to the GR pathway, if any, has not been previously explained. 35 Slabeled in vitro-translated hVIP mutants and HIV-1 Vpr were mixed and incubated on ice in incubation buffer for 1 h. The protein complexes were immunoprecipitated with anti-His tag antibody and then subjected to SDS-PAGE on 15% gel. Fig. 4 shows the interaction of the mutants with Vpr. Vpr coimmunoprecipitates with hVIP mutants, except that the mutant has its carboxyl terminus deleted, revealing the lack of interaction between Vpr and this hVIP mutant (NM).

FIG. 5. Effect of mifepristone on the Vpr-induced movement of hVIP.
Mifepristone is a strong GR antagonist and inhibits glucocorticoid-induced signals. Because HIV-1 Vpr was earlier shown to mimic glucocorticoid activity, in this study mifepristone was examined for its effect on the Vpr-induced movement of hVIP. Without mifepristone, hVIP moves into the nucleus in cells expressing both hVIP and Vpr (A-D). Mifepristone inhibited the nuclear movement of hVIP induced by Vpr (E-H). DAPI was used to stain the nuclei of cells available in the field (A, E). The FITC-staining reveals the expression of hVIP (B, F) and the rhodamine-staining was used for Vpr (C, G). Images in panels D and H are overlaid images from the corresponding FITC-and rhodamine-stained cells. The cells that coexpress both hVIP and Vpr are marked with arrows.

hVIP/mov34 and Glucocorticoid Signaling
Accordingly, we looked for a direct effect of a glucocorticoid compound, dexamethasone, on the hVIP localization pattern in vivo. For this study, cells were transfected with hVIP or hVIP mutants and treated with dexamethasone. We examined the expression pattern of hVIP by immunofluorescent analysis 36 h later. Fig. 6, C and D clearly illustrates the alteration in the localization pattern of hVIP in cells treated with dexametha-sone compared with the untreated cells. The effect of the glucocorticoid in moving hVIPmov34 to the nucleus was dosedependent. We analyzed four different concentrations of dexamethasone (10, 25, 50, 100 nM). At the lowest concentration tested, only about 10% of the cells did not show a cytoplasmic localization pattern. Interestingly some cells (5%) appeared to show a confused stage where hVIPmov34 was initiating movement into the nucleus. Hence the protein stained both the cytoplasm and the nucleus (data not shown), whereas at the highest concentration tested, 50 nm, every cell that expressed hVIPmov34 showed typical nuclear staining of this protein. To further confirm this effect, we next sought to interfere with the activity of the glucocorticoid. Mifepristone has been reputed to have an antagonistic effect on glucocorticoid activities. We examined whether mifepristone would inhibit the dexamethasone-induced movement of hVIP into the nucleus. Dexamethasone failed to move hVIP under the influence of mifepristone (Fig. 6, E and F), as these cells displayed a clear cytoplasmic localization of this protein in the presence of both dexamethasone and mifepristone, which is distinct to FIG. 7. hVIP expression in CV-1 cells. CV-1 cells are known to be negatively responsive to glucocorticoid treatments because they lack the functional GR receptor complex. These cells were transfected with hVIP plasmids to study the expression pattern of hVIP under the influence of dexamethasone. Dexamethasone failed to move hVIP into the nucleus, as revealed by the cytoplasmic localization of hVIP in both untreated (A, B) and dexamethasone-treated (C, D) cells. hVIP expressing cells are marked.

FIG. 8. Physical interaction of hVIP/mov34 with GR.
In vitrotranslated 35 S-labeled protein samples generated from pchVIP and pcGR were mixed and immunoprecipitated using either anti-His that targets His tag-fused hVIP or anti-GR antibodies. Either of the antibodies could precipitate both hVIP and GR. As negative control, radiolabeled hVIP and GR were incubated with anti-GR and anti-His tag antibodies, respectively. Anti-GR antibody failed to precipitate hVIP, and anti-His antibody does not precipitate the GR protein.
FIG. 6. hVIP/mov34 is glucocorticoid-responsive. The expression of hVIP in HeLa cells was studied under the influence of a glucocorticoid, dexamethasone. The untreated cells revealed the typical cytoplasmic localization of hVIP (A, B). In dexamethasone-treated cells, hVIP is translocated into the nucleus (C, D). The addition of mifepristone to dexamethasone-treated cells inhibited the nuclear movement of hVIP (E, F). The cytoplasmic localization (G, H) of hVIP mutant that has its carboxyl-terminal deleted (NM) is not affected by dexamethasone treatment (I, J), and hence this mutant is not glucocorticoid-responsive. FITC-conjugated secondary antibody was used to study hVIP localization (right column), and DAPI was used to counterstain the nuclei with bright blue color (left column). hVIP expressing cells are marked by arrows. hVIP/mov34 and Glucocorticoid Signaling cells treated with dexamethasone alone. Additionally the African green monkey cell line, CV-1, has been characterized as unable to respond to the effects of glucocorticoids (36 -38) because it lacks a functional GR receptor complex. As an additional confirmation, we were interested in examining the dexamethasone effect on hVIP localization in this cell line. CV-1 cells were transfected with hVIP expression plasmids and either treated or mock-treated with dexamethasone. As would be expected in this glucocorticoid-negative cell line, both untreated and dexamethasone-treated cells displayed a similar cytoplasmic localization pattern for hVIP expression (Fig. 7); typical solid nuclear staining was not witnessed in these cells. Therefore, in CV-1 cells dexamethasone was not effective in altering the localization pattern of hVIP. Collectively, these data support the interaction of hVIP with the GR pathway.
In the next set of experiments, we determined whether the nuclear import of hVIP in the glucocorticoid-treated cells is mediated through direct interaction between hVIP and GR. We examined the ability of these two proteins to interact with each other using in vitro radiolabeled protein samples. In vitrotranslated 35 S-labeled samples of GR and hVIP were mixed together and immunoprecipitated with either anti-GR or anti-His tag antibodies. We found that 35 S-labeled hVIP was pre-cipitated with GR by anti-GR antibody and GR was coimmunoprecipitated along with hVIP while anti-His tag was used. To determine cross-reactivity, we also checked the precipitation of hVIP by anti-GR antibody and vice versa. Anti-His tag and anti-GR antibodies failed to precipitate radiolabeled GR and hVIP proteins, respectively (Fig. 8). These data permit the conclusion that these two proteins interact with each other in an in vitro system also. DISCUSSION Several features of Vpr activity have suggested that this protein interacts with a host cell ligand that participates in its transit from the host cytoplasm to the host nucleus. hVIP appears to be a candidate for part of this activity. Prior studies have shown that hVIP can interact with Vpr directly and that hVIP is important in its own right for cell cycle arrest (32).
Here we sought to characterize the interaction between hVIP and Vpr.
Experimental results presented here clearly indicate that the carboxyl-terminal region of hVIP mediates its interaction with Vpr in vitro and in vivo (Figs. 3 and 4). Mutants of hVIP behave differently in terms of their localization patterns in cells transfected with these mutant constructs. The differential localization pattern supports the conclusion that the amino terminus of hVIP is important for its cytoplasmic localization, whereas the carboxyl terminus is more nucleophilic. These features likely may play a role in the shuttling of this protein between the nucleocytoplasmic compartments and likely have a crucial role in the natural function of this protein. By blast search, hVIP has homology with some of proteins involved in the initiation of the eukaryotic translation (33) as well as in proteasome formation; these protein families have crucial roles during embryonic development in eukaryotes. The knockout of this gene in transgenic mice results in a lethal phenotype.
Glucocorticoids regulate diverse functions and are important to maintain central nervous system, cardiovascular, metabolic, and immune homeostasis. They also exert anti-inflammatory and immunosuppressive effects, which have made them invaluable therapeutic agents in numerous diseases (39). The actions of these hormones are mediated by their specific intracellular receptors, such as the GR. Several host co-activators of the GR have been described that directly interact with GR and compo- FIG. 9. Glucocorticoid-responsive and Vpr-interacting domain of hVIP/mov34. The predicted amino acid sequence of hVIP/mov34 with leucine/iso-leucine motifs, which are important in protein-protein interactions in biological functions, are underlined. The domain representing the carboxyl-terminal region of hVIP that has been identified as Vpr-interacting and the glucocorticoid-responsive domain is indicated in the box. hVIP/mov34 and Glucocorticoid Signaling nents of the transcription initiation complex to enhance the glucocorticoid signal to the transcription machinery (40).
The GR is the prototypic member of the translocating class of steroid receptors that are ubiquitously expressed in almost all human tissues and organs. Unliganded GR is found in the cytoplasm and moves rapidly into the nucleus in response to hormone stimulation (41,42). Potential steroid hormone receptor antagonists such as mifepristone prevent the GR from moving into the nucleus in response to appropriate stimulation. In the present study, both HIV-1 Vpr and dexamethasone induce hVIP to move into the nucleus. Interestingly, mifepristone prevents the movement of hVIP into the nucleus in response to the treatment of the cells with a glucocorticoid as well as in response to the co-expression of Vpr plasmid from the vpr-expressing construct. Thus both Vpr and glucocorticoid exert a similar effect on hVIP. This observation assumes significance because it is consistent with earlier reports that Vpr mimics glucocorticoid effects on apoptosis and immune suppression that are inhibited by mifepristone (18,26). Interestingly, the hVIP mutant that has the carboxyl-terminal deletion exhibited resistance to translocation into the nucleus even when co-expressed with Vpr, unlike the full-length hVIP molecule. This mutant fails to bind Vpr. Dexamethasone also failed to effect the nuclear movement of this mutant. These data suggest that the carboxyl domain of hVIP may interact with a shuttling protein that is part of the GR complex and that this interaction is necessary for the movement of hVIP into the nucleus. The deletion of the carboxyl-terminal domain renders this hVIP mutant unable to bind the shuttle protein and thus the nuclear translocation of this mutant is impaired (Fig. 9). Clearly, identification of this hVIP ligand will give important clues into Vpr and GR biology.
Although convincing data advance a theory that Vpr mimics some glucocorticoid effects, it is also clear that Vpr interacts directly with members of the GR complex (26,36). These findings suggest a clustered protein complex present in the cytoplasm that forms the cellular target to receive and transduce the signals triggered by Vpr as well as glucocorticoids (Fig. 10). GR interacts in the cytoplasm with a complex array of chaperone proteins, including HSP90 and HSP70, and ligand-dependent displacement of these proteins is thought to be intimately involved in the translocation process (43,44). Both GR and hVIP are known Vpr ligands. Further, in vitro binding analysis reveals a direct interaction between hVIP and GR. (Fig. 8). By responding to glucocorticoid activation, hVIP appears to have additional cellular roles besides its vital role in cell cycle regulation that was described earlier. Therefore, the further study of the biological relevance of hVIP-Vpr interaction has important implications for HIV biology and our understanding of the important host cell glucocorticoid receptor functions. The identification of hVIP as a target for GR interactions has important implications for our understanding of HIV biology.