Cyclophilin A Interacts with HIV-1 Vpr and Is Required for Its Functional Expression*

Viral protein R (Vpr) of human immunodeficiency virus, type 1 (HIV-1) is the major virion-associated accessory protein that affects a number of biological functions in the retroviral life cycle, including promotion of the transport of the preintegration complex into the nucleus and the induction of G2 host cell cycle arrest. Our recent investigation of the conformational heterogeneity of the proline residues in the N terminus of Vpr suggested a functional interaction between Vpr and a host peptidylprolyl cis/trans isomerase (PPIase) that might regulate the cis/trans interconversion of the imidic bond within the conserved proline residues of Vpr in vivo. Using surface plasmon resonance spectroscopy, Far Western blot, and pulldown experiments a physical interaction of Vpr with the major host PPIase cyclophilin A (CypA) is now demonstrated. The interaction domain involves the N-terminal region of Vpr including an essential role for proline in position 35. The CypA inhibitor cyclosporin A and non-immunosuppressive PPIase inhibitors such as NIM811 and sanglifehrin A block expression of Vpr without affecting pre- or post-translational events such as transcription, intracellular transport, or virus incorporation of Vpr. Similarly to CypA inhibition, Vpr expression is also reduced in HIV-1 infected CypA–/– knock-out T cells. This study thus shows that in addition to the interaction between CypA and HIV-1 capsid occurring during early steps in virus replication, CypA is also important for the de novo synthesis of Vpr and that in the absence of CypA activity, the Vpr-mediated cell cycle arrest is completely lost in HIV-1-infected T cells.

In addition to the genes encoding the structural and enzymatic proteins common to all retroviruses, the human immu-nodeficiency virus, type 1 (HIV-1) 1 genome contains four accessory genes that serve to accelerate viral replication. One of these gene products, the highly conserved 96-amino acid viral protein R (Vpr) has received considerable attention, and a number of biological functions have been attributed to its presence in various cellular and extracellular compartments. The most intensively investigated biological functions of Vpr are those affecting the translocation of the preintegration complex of the incoming virus from the cytoplasm to the nucleus and the arrest in the G 2 phase of the cell cycle (1)(2)(3). The nuclear targeting function of Vpr has been associated with HIV-1 infection of terminally differentiated macrophages. Regarding the second function of Vpr, which leads to G 2 cell cycle arrest in HIV-1-infected and/or Vpr-transfected human cells, it was implicated that this activity provides an intracellular milieu conductive for enhanced viral replication by increasing HIV LTRdriven gene expression (4). This response has been linked recently (5) with the capacity of Vpr to alter the structure of the nuclear lamina, leading to transient, DNA-containing herniations of the nuclear envelope that intermittently rupture. Other studies (6 -8) suggest that the prolonged G 2 arrest induced by Vpr ultimately leads to apoptosis of the infected cell. Conversely, early anti-apoptotic effects of Vpr have also been described that are superceded later by its pro-apoptotic effects (9). These pro-apoptotic effects of Vpr may result from either effects on the integrity of the nuclear envelope or direct mitochondrial membrane permeabilization (10), perhaps involving Vpr-mediated formation of ion channels in cellular membranes (11).
As the biologically relevant activity of the multifunctional Vpr has not yet been clarified, the molecular bases for many of its effects remain elusive. To date structural studies of the full-length molecule have been hampered by the fact that this protein does not crystallize, and the use of NMR techniques is complicated by the strong tendency for Vpr to undergo selfassociation. In addition, the structure of Vpr depends critically on the solution conditions, and an unresolved apparent heterogeneity in the composition of the polypeptide has been observed (12,13). However, a somewhat limited model is slowly emerging from structural studies of fragments of Vpr (14 -19) and, more recently, of full-length synthetic forms of Vpr (12,13). These studies were performed in aqueous solution that required the presence of either detergents or micelles to detect the most structured forms of Vpr. The secondary structures in Vpr emerging from these analyses (summarized in the accompanying paper (19)) suggest the presence of an ␣-helix-turn-␣helix motif between residues 17 and 48 and an amphipathic ␣-helix between amino acids 53-55 and 78 -83 (12,15). These helices most likely play a key role in self-association and the interaction of Vpr with heterologous proteins (13,20). Both helical domains are highly conserved and contribute to the formation of a novel nuclear import signal (21). The distal helix also contains a leucine-rich nuclear export signal whose function is inhibited by leptomycin B (22). The C terminus of Vpr contains a basic amino acid-rich segment between residues 73 and 96 for which little definite structure has been assigned (15). However, this region is conserved and influences the stability and, potentially, the structure of the entire protein (8). It contains a bipartite, arginine-rich nuclear import signal that can promote nuclear uptake of heterologous proteins via the nuclear pore complex (22,23). Various investigations have shown that the C-terminal domain also participates in a number of specific protein-protein and protein-nucleic acid interactions such as with NCp7 (24), Tat (25), RNA (26), and the adenine nucleotide translocator of the mitochondrial pore (27).
In contrast to the C-terminal domain, few functional properties have been attributed to the N-terminal domain of Vpr, and its participation as an interaction surface with other molecules has received less attention. Many previous structural characterizations of Vpr (for instance see Refs. 12,13,and 20) have shown the important influence of environmental factors and indicate that in vivo folding of Vpr will probably require the presence of structure-stabilizing interacting factors (for review see Ref. 3). Furthermore, we have recently focused our attention on the N-terminal domain of Vpr (19) and found that two of the four conserved prolines at residues 14 and 35 in this domain exhibit an unusually high content of cis-conformations of the imidic bond. Based on the cis/trans phenomenon that could be important for the folding of Vpr, it was projected that a potential interaction between Vpr and a host peptidyl-prolyl isomerase (PPIase) might regulate the interconversion of the imidic bond of these N-terminal proline residues of Vpr in vivo.
As for most other intracellular parasites, the replication of HIV depends on the interaction with host cell factors, and some of these are incorporated specifically into progeny virions. Among the most abundant cellular proteins and the first ever found in HIV-1 virions is cyclophilin A (CypA). This protein is specifically incorporated into HIV-1 virions, but not into virions of other lentiviruses, through an interaction with a proline-rich region of the HIV-1 capsid (CA) protein. In particular, Pro-222, which is conserved in the CA region of all HIV-1 Gag polyproteins, appears to be important for the interaction of HIV-1 Gag with CypA (Refs. 28 and 29; for review see Refs. 30 and 31). Earlier studies using two-hybrid screens found that HIV-1 Gag binds to most of the known members of the family of CyPs (32). CyPs were originally described as binding partners of cyclosporin A (CsA), an immunosuppressive cyclic undecapeptide used clinically to prevent allocraft rejection (33). CsA binds with nanomolar affinity to CypA, and this complex inhibits calcineurin, a calcium-dependent phosphatase that regulates the expression of various cytokine genes in activated T cells (for review see Ref. 34). CyPs belong to the PPIases, a group of enzymes found in organisms ranging from prokaryotes to humans, that catalyze the otherwise relatively slow cis/trans isomerization of peptidyl-prolyl bonds in vitro. As CyPs have been proposed to regulate protein folding in vivo (34) it was conceivable that the family of CyPs, and CypA in particular, are possible candidates for providing a PPIase activity that regulates cis/trans interconversion in Vpr.
A large body of evidence supports a function of CypA in formation of infectious HIV-1 viruses (for review see Refs. 30 and 31). These studies were based either on mutation within gag or the use of competitive CypA inhibitors such as CsA that interfere with the CypA-Gag interaction (35)(36)(37)(38)(39). A more conclusive proof that among the other 14 known members of mammalian CyPs only CypA plays a functional role in supporting HIV-1 replication was provided by selective genetic inactivation of the gene encoding CypA in human CD4 ϩ T cells (40). In addition, the competitive inhibitor CsA exhibits high affinity for the hydrophobic pocket of almost all CyPs and thus competitively inhibits the interaction of CypA with HIV-1 Gag (28,29,32,35,37,41,42). The immunosuppressive activity of CsA is not correlated with anti-HIV activity, since the nonimmunosuppressive derivative NIM811 ([methyl-Ile 4 ]-cyclosporin) is an even more potent inhibitor of CypA-mediated HIV-1 replication than the parental CsA (29,36,39,41,42). Thus, CsA and related nonimmunosuppressive derivatives form an interesting class of drugs that can modulate the interaction of CypA with other proteins.
In the present work, we demonstrate an unexpected interplay of Vpr with CypA that depends of the N-terminal domain of Vpr containing conserved proline residues. Most importantly, we report that CypA regulates the expression of Vpr and is required for the Vpr-mediated G 2 cell cycle arrest in HIV-1infected T cells.

MATERIALS AND METHODS
BIAcore Spectroscopy-Surface plasmon resonance measurements were performed at 25°C using a BIACORE 2000 (BIAcore AB, Uppsala, Sweden) equipped with a CM5 research-grade sensor chip. Recombinant CypA (Sigma) was immobilized at a concentration ranging from 4200 to 11200 response units using standard amine-coupling chemistry in three flow cells, and a further flow cell without CypA was used as a control. The peptides Vpr 1-40 , Vpr 1-20 , Vpr [21][22][23][24][25][26][27][28][29][30][31][32][33][34][35][36][37][38][39][40] , and respective proline mutants were dissolved at concentrations ranging from 1 to 250 M in a running buffer (10 mM Hepes, 150 mM NaCl, 50 M EDTA, 0.005% Tween 20, pH 7.4) and were injected over the flow cells at a flow rate of 5 l/min. Data were collected at a rate of 2.5 Hz during the 120-s association and dissociation phase. Results were corrected for the response unit values of the reference cell (without CypA) to exclude unspecific binding of Vpr to the chip matrix. Experiments were repeated at least three times, each with two different charges of Vpr peptides and CypA, and afforded reproducible data.
Northern Blot Analysis-HeLa cells were transfected with pCMV-FLAG-Vpr, and 5 h post-transfection cells were treated with CsA (50 g/ml), NIM811 (10 g/ml), SFA (10 g/ml), or no inhibitor (aliquot of solvent Me 2 SO). Following 0, 2, or 8 h of treatment, cells were harvested, washed once in ice-cold phosphate-buffered saline, and frozen immediately at Ϫ80°C. Total cell RNA was isolated using Trizol (Invitrogen), and 10 g of each RNA sample was separated in 1.2% denaturing agarose gel containing 17.8% formaldehyde. Following blotting and UV cross-linking to the nylon membrane (Hybond-N TM ; Amersham Biosciences) the membrane was pre-hybridized for 5 h at 65°C in Denhardt's solutions containing salmon sperm DNA and poly(A) RNA according to modification of a standard procedure described previously (43). For hybridization oligonucleotides (FLAG probe 5Ј-cttgtcgtcatcgtctttgtatgccat-3Ј, Vpr probe 5Ј-agtaacgcctattctgctatgtcgacacccaattctgaaatg-3Ј) were radioactively labeled with [␣-32 P]dCTP using terminal transferase (220 582; Roche Applied Science) and added to the hybridization solution for 12 h at 65°C. After washing, blots were analyzed by autoradiography. For internal control blots were stripped for 30 min in 0.1% SDS at 70°C and re-hybridized with a glyceraldehyde-3phosphate dehydrogenase probe (5Ј-ccatggtggtgaagacgccagtggactcc-3).
Transfection, Infections, Immunostaining, Flow Cytometry, and DNA Content Analysis-Cells were transfected with LipofectAMINE 2000 TM or calcium phosphate for precipitation of DNA. T cell cultures were infected with isogenic strains of HIV-1 NL4 -3 , differing only by the presence or absence of vpr (44,45). Generally, 10 7 cells were incubated with virus stocks (200 ng of p24 Gag ) in 1 ml of medium for 2 h prior to washing and subsequent culturing at 1 ϫ 10 6 cells per ml for the indicated times. HIV-infected cells were fixed and permeabilized with a solution of 1% paraformaldehyde, 1 mg/ml human IgG (Gemini Bio-Products, Inc.), and 0.1% Tween 20 in FACS buffer for at least 1 h. 10 6 cells were immunostained in 50 l of FACS buffer with 1:50 dilution of the monoclonal anti-p24 antibody KC57 (Lot 13; Coulter), conjugated to fluorescein isothiocyanate. Cellular DNA content was assessed by additional staining with 0.01 mM To-Pro-3 iodide (Molecular Probes) in the presence of 1 mg/ml RNaseA followed by analysis using a FACS-can™ flow cytometer acquiring linear fluorescence in the FL4 channel. DNA profiles were analyzed with FlowJo software (Treestar).
Cell Culture, Pulse-Chase, and Western Blot Analyses-CD4 ϩ human T lymphoma Jurkat T cell lines and the knock-out line PPIA Ϫ/Ϫ were cultured in RPMI 1640, and HeLa cells (ATCC CCL2) were propagated in Dulbecco's modified Eagle's medium. Cells were metabolically labeled with [ 35 S]methionine (2 mCi/ml), and pulse-chase experiments and immunoprecipitation were performed as described (46). For Vpr detection in virions cell, culture supernatants were tested with p24 enzyme-linked immunosorbent assay to assess peak virus replication during a 7-to 10-day period. Virus pellets were normalized for CA content by Western blot and enzyme-linked immunosorbent assay, and aliquots were separated by SDS-PAGE and probed by Western blot with rabbit antibodies specific for CA and Vpr followed by enhanced chemiluminescence staining. For Western blot kinetic analyses, HeLa cells were transfected with pCMV-FLAG-Vpr, pCMV-3HA-Vpr wild type, pCMV-3HA-VprP5,10,14A, or pCMV-3HA-VprP35N, and after the optimum time post-transfection necessary to initiate homologous gene expression that was determined for each construct, cells were harvested by scraping, aliquoted in RPMI, and treated with CsA (50 g/ml), NIM811 (10 g/ml), SFA (10 g/ml), or no inhibitors (aliquot of solvent Me 2 SO). Samples were taken 0, 1, 2, 4, and 8 h after beginning of drug treatment, and cell lysates were separated in 14% SDS-PAA gels. Proteins were transferred onto nitrocellulose membranes and probed with specific antibodies followed by enhanced chemiluminescence detection. For internal controls blots were stripped for 20 min in 2% mercaptoethanol and 1% SDS at 65°C and a re-incubated with antibodies specific for actin and CypA.
Plasmids and Antibodies-The pCMV-FLAG-Vpr DNA plasmid directs the expression of HIV-1 NL4 -3 Vpr with the FLAG epitope at the N terminus (47). The pCMV-3HA-Vpr vector was used to express Vpr with an N-terminal hemagglutin (HA) tag of HIV-1 NL4 -3 (21). Mutations at proline residues 5, 10, 14, and 35 were introduced by designing PCR primers containing the indicated changes and cloning into the plasmid directly. The plasmid pNLenv1 represents an env-deleted version of HIV-1 NL4 -3 (46). Antibodies specific for FLAG, actin, and HA were obtained from Sigma, the CypA antibody was from Calbiochem, and the Vpr antibody was as described (13). The peroxidase-coupled anti-mouse and anti-rabbit antibodies were obtained from Dianova.
In Vitro and in Vivo Complex Formation of Vpr and CypA-1 g of each, recombinant CypA (Calbiochem) and sVpr, were incubated in 1 ml of phosphate-buffered saline, pH 7.2, at 22°C for 12 h. Aliquots of the mixture were denatured in SDS-PAGE sample buffer (without reducing agents), incubated at room temperature for 30 min, separated in a 12.5% PAA gel, and analyzed by Western blotting using rabbit anti-CypA (Calbiochem) and anti-Vpr (R-96) antibodies. For co-immunoprecipitation, 3 ϫ 10 7 HeLa cells transfected with pCMV-FLAG-Vpr were lysed in lysis buffer (140 mM NaCl, 10 mM Tris/HCl, pH 7.4, 1 mM EDTA, protease inhibitor mixture (Roche Applied Science), 1% (w/v) digitonin (Wako)), and the cytosolic fraction was isolated by centrifugation (4°C, 15 min, 14,000 ϫ g) and incubated with anti-FLAG antibodies for 12 h at 4°C. Immune complexes were recovered on protein G-Sepharose, washed five times in wash buffer (lysis buffer with 0.1% digitonin), separated in 14% SDS-PAA gel, and probed in Western blot with anti-CypA antibodies and also with anti-M2 antibodies following stripping and reprobing. As a control, the same amounts of mock transfected HeLa cells were treated identically. For quantification, serial dilution of recombinant CypA was analyzed in parallel. For Far Western blot technique recombinant purified CypA was separated in a preparative 14% SDS-PAA gel and transferred onto nitrocellulose, and membrane strips were blocked by incubation in 3% bovine serum albumin in Tris-buffered saline. Individual strips of the membrane were incubated with 1 g/ml of either sVpr 1-96 or recombinant CA (48) for 4 h at 4°C. Membrane strips were washed three times in 0.05% Tween 20 in Tris-buffered saline, and bound antigen was detected by staining either with anti-Vpr or anti-CA polyclonal rabbit sera followed by anti-rabbit peroxidase conjugate and enhanced chemiluminescence reaction. For mock controls, CypA-containing strips were incubated with isolated proteins like lysozyme or whole cell lysates, instead of sVpr  or p24 CA , followed by staining with antibodies specific for CA or Vpr, as well as rabbit pre-immune serum. Under no circumstances was crossreactivity of CA or Vpr-specific antibodies with immobilized CypA observed.

Biochemical Evidence for CypA-Vpr Interaction in Vitro and
in Vivo-Our recent finding (19) of unusually large amounts of cis proline conformations about the imidic bond for two of the proline residues in the N terminus of Vpr suggested that Vpr folding may require a PPIase enzyme activity. Since both Vpr and CypA are present in HIV-1 virions, and CypA binds to HIV-1 CA and promotes virus infectivity, we investigated whether Vpr and CypA also interact. Initially, we utilized a Far Western blotting technique in which purified CypA bound to nitrocellulose was incubated in ambient buffer with sVpr. As a control, recombinant HIV-1 CA was tested in parallel. Proteins bound to CypA were visualized with specific antibodies (Fig.  1A). For both CA and sVpr, a clear interaction with immobilized CypA was detected. This signal was specific, because it was not detected when either control antibodies or control proteins were incubated with CypA-containing membranes ( Fig. 1A and data not shown).
To investigate the interaction between Vpr and CypA in their native conformations in solution, we analyzed the potential of isolated CypA and Vpr to engage in disulfide-bound formation. Recombinant CypA and sVpr were incubated in ambient buffer, followed by separation by SDS-PAGE under non-reducing conditions and Western blot analysis using CypA-specific antibodies. As demonstrated in Fig. 1B, in addition to the monomeric 18-kDa CypA and a dimeric 36-kDa disulfide-linked dimer of CypA, a novel ϳ29-kDa product was detected in the presence of sVpr that was reactive with antibodies specific for both CypA (Fig. 1B) and Vpr (not shown). This product could not result from disulfide-linked Vpr dimerization shown previously for sVpr (13) since the occurrence of the ϳ29-kDa product was strictly dependent on the presence of CypA, and no crossreactivity of the anti-CypA antibody with Vpr was observed (Fig. 1B). Furthermore, the CypA dimer, as well as the proposed 29-kDa product, was not observed in the presence of reducing agents (not shown), indicating the formation of disulfide-linked CypA-Vpr heterodimers in solution. Direct evidence for CypA-Vpr complex formation in vivo was obtained in immunoprecipitation employing HeLa cells transfected with the plasmid pCMV-FLAG-Vpr that directs the expression of an N-terminally epitope-tagged version of Vpr. Potential interaction of endogenous CypA with Vpr was analyzed by lysis of these cells in digitonin-containing buffer, immunoprecipitation with anti-FLAG antibodies, and immunoblotting of the immunoprecipitates with antibodies specific for CypA (Fig. 1C). Based on a standard dilution of recombinant CypA that was analyzed in parallel, ϳ1 to 5 ng of CypA was recovered from 10 6 FLAG-Vpr-expressing cells. No CypA was recovered when vprdeficient mock transfected HeLa cells were subjected to the same immunoprecipitation procedure indicating that anti-FLAG antibodies did not bind unspecifically to CypA. These findings reveal a specific interaction between Vpr and CypA, both of which are nonstructural components of mature HIV-1 virions (30).
Surface Plasmon Resonance Measurement of CypA-Vpr Interaction in Vitro-To further verify the apparent interaction of Vpr and CypA, we employed surface plasmon biosensor analysis (spectroscopy; BIAcore), which allows measurement of protein binding in real time (Fig. 2). The technique has the potential of detecting transient complexes and of locating binding sites using suitable peptide fragments. In initial experiments we were unable to obtain reliable BIAcore data using fulllength sVpr under a variety of experimental conditions, presumably because of the intrinsic tendency of Vpr to self-aggregate (13,20,49). Hence, in subsequent BIAcore studies we focused on the analysis of the N-terminal domain of Vpr, exemplified by the peptide Vpr 1-40 that exists as a monomer in solution and contains two proline residues in positions 14 and 35 with a high content of the cis conformation of the imidic bond (19).
For the first BIAcore experiment, increasing concentrations up to 250 M Vpr 1-40 were injected over the flow cells containing immobilized CypA. Results were corrected by the response unit values of the reference cell (without CypA) to exclude nonspecific binding of Vpr to the chip matrix. Dose-dependent binding of Vpr 1-40 to CypA rapidly reached equilibrium ( Fig.  2A). Binding responses returned to baseline levels within 30 s after washing with buffer, indicating that the complex undergoes rapid dissociation. The responses in Fig. 2 are such that quantitative equilibrium binding data could not be evaluated because baseline shifts and mass transfer rates were too small to allow meaningful calculations. Accordingly, we next used specific peptide domains of Vpr to explore their interaction with CypA. First, we examined the potential role of individual proline residues in Vpr 1-40 for CypA binding. were purified to homogeneity and tested for binding to immobilized recombinant CypA in this biosensor system. Although wild type Vpr 1-40 exhibited a concentration-dependent binding to CypA ( Fig. 2A), mutation of Pro-35 to Asn completely abrogated the binding of Vpr 1-40 P35N to CypA (Fig. 2C). Surprisingly, mutation of prolines 5, 10, and 14 did not alter dose-dependent binding of the mutant Vpr 1-40 P5,10,14N peptide to CypA (Fig.  2B), although removal of all proline residues completely prevented CypA binding (Fig. 2D). Thus, in the context of Vpr 1-40 , these Biacore analyses indicate that among the prolines at positions 5, 10, 14, and 35 only Pro-35 is essential for the binding of CypA.
Consistent with these findings obtained with the Vpr 1-40 peptide, we also observed that an N-terminal fragment, Vpr 1-20 lacking Pro-35 failed to bind to CypA and further that mutation of proline residues in positions 5, 10, and 14 of Vpr 1-20 did not support any interaction with CypA (data not shown). In contrast, the Pro-35-containing fragment Vpr 21-40 exhibited dosedependent binding similar to that observed for Vpr   (Fig.  2E). Again, mutation of Pro-35 to Asn resulted in an almost complete loss of binding of the mutant Vpr 21-40 P35N to CypA (Fig. 2F). As an additional control, we tested a Vpr peptide from the C-terminal region spanning amino acids 47-96 of Vpr from HIV-1 NL4 -3 and detected no interaction with CypA (not shown). In summary, these BIAcore analyses provide direct evidence for a specific interaction between the N-terminal domain of Vpr and CypA and highlight an essential role of Pro-35 in this interaction.
CypA Inhibitors Block Vpr Expression-In view of the interaction of the N-terminal region of Vpr with CypA, we next examined whether CsA, an inhibitor of CypA activity, altered the synthesis of Vpr in HIV-1-expressing cells. For these studies, we performed pulse-chase radiolabeling experiments in HeLa cells transfected with the env-deleted HIV-1 NL4 -3 subgenomic expression vector pNLenv1 (Fig. 3). This system allows the expression of relatively high levels of Vpr in the context of other HIV-1 proteins, by possibly bypassing the cytotoxic effects often observed when HIV Env glycoproteins are expressed at high levels (data not shown). Parallel cultures of HeLa cells transfected with pNLenv1 were treated or not treated with 50 g/ml of CsA 30 min prior to a 15 min pulse labeling of the cells with [ 35 S]methionine. CsA was maintained in the culture throughout the experiment. Cells were chased for up to 8 h, and Gag (Fig. 3A, anti-CA) or Vpr (Fig. 3A, anti-Vpr) protein was recovered from cell lysates by immunoprecipitation with specific antibodies, separated by SDS-PAGE, and analyzed by fluorography. Consistent with previous reports studying the effects of CsA treatment (28,29), Gag mutation (50), or CypA gene targeting (40) on HIV-1 replication, CypA activity was not required for efficient expression and processing of HIV-1 Gag polyproteins (Fig. 3A). Although the kinetics of the conversion of Pr55 into CA were not affected, CsA treatment did induce a relatively minor ϳ25% reduction in Pr55 expression (Fig. 3A) that was similarly observed when other control proteins such as green fluorescent protein (not shown) were co-expressed. Thus, these small changes can probably be attributed to the nonspecific inhibitory effect of CsA on protein translation (51).
Strikingly, in the presence of CsA, no Vpr was detected in HIV-1-expressing cells, although Vpr was detected readily in the untreated cultures (Fig. 3A). The relatively weak signal of Vpr recovered from the untreated culture reflects most likely the inefficient metabolic labeling of the N-terminal Met-1 of Vpr during the short pulse. To unambiguously confirm this result, in a further experiment the inhibitory effect of CsA on Vpr expression was demonstrated using the CMV-driven expression vector, pCMV-FLAG-Vpr, which mediates high level synthesis of FLAG-tagged HIV-1 Vpr that also affords strong signals after [ 35  fected with pCMV-FLAG-Vpr were analyzed by pulse-chase labeling in the presence or absence of CsA. Again, we observed a marked 90% reduction in the level of Vpr recovered and almost complete disappearance of Vpr after 8 h of chase (Fig.  3A). This detrimental effect of CsA on Vpr expression is not a consequence of a nonspecific effect on gene expression, because the expression of other control proteins, such as HIV-1 Gag proteins (Fig. 3B), green fluorescent protein (data not shown), or Vpr mutants (see Fig. 5), was not severely compromised by CypA inhibition. These findings demonstrate that inhibition of CypA function with CsA is associated with a marked reduction of intracellular Vpr expression.
Two possible mechanisms could account for the loss of Vpr (Fig. 3): either CsA causes rapid degradation of Vpr that occurs during the 15-min pulse labeling period, or it induces a direct effect on the expression of Vpr. To study the kinetics and dose response of CsA-induced Vpr loss more carefully, we conducted short term pulse-chase experiments. When cells were treated with graded doses of CsA for 30 min and then pulse-labeled for 5 min a marked reduction in Vpr expression was again observed (Fig. 4). This response was dose-related since an approximately 3-fold loss of Vpr was observed in cells treated with 10 g/ml CsA whereas a more than 10-fold loss of Vpr occurred in cells treated with 50 g/ml CsA (Fig. 4A). Noteworthy, even under these short term kinetics, a consistent loss of Vpr was evident at the conclusion of the initial labeling period, and no further decay was observed throughout the chase. A similar loss of Vpr was also detected when Vpr was isolated from the nuclear fraction, whereas the relative ratios of Vpr distribution in cytosol and nucleus remained constant under all conditions, indicating that the shuttling of newly synthesized Vpr in and out the nucleus is not affected by CsA. This was further demonstrated by immune fluorescence studies that showed that CsA had no obvious effect on the subcellular localization of Vpr in HIV-1-expressing cells (data not shown). These findings suggest that CypA acts at a very early step during Vpr synthesis to stabilize its expression. These findings further argue against the loss of Vpr occurring through altered intracellular transport of Vpr.
The fact that the rapid loss of Vpr occurs even in the short term pulse-chase experiment and that there is no apparent subsequent loss during the chase period mitigates against a CypA-mediated proteolytic degradation of Vpr. Nevertheless, if proteolysis is involved in this phenomenon, it must occur in a very short time period, probably seconds. The 26 S proteasome is the major multienzyme complex that digests damaged or unwanted proteins at a high turnover rate (52). To test for a potential role of the proteasome in CsA-induced Vpr loss, pulse-chase experiments were conducted in cells with inactivated proteasomes. Proteasome inhibitors, however, did not reverse the CsA-mediated loss of Vpr in the cytosolic (Fig. 4B) or the nuclear fraction (data not shown). Thus, the proteasome does not appear to be involved in this phenomenon, and even under the shortest kinetic conditions employed, we were unable to detect any evidence for CsA-induced degradation of Vpr.
CypA Inhibitors Specifically Block de Novo Synthesis of Vpr-From the rapid effect of CsA on Vpr expression, it seemed likely that CypA regulates the de novo synthesis of Vpr rather than influencing its stability. With this in mind, we initially investigated the potential effects of CypA inhibition on transcription and splicing of vpr mRNA by Northern blot analysis. In these studies, we also included CypA inhibitors that, unlike CsA, block the PPIase activity of CypA without affecting the calcineurin pathway. For example, the CsA analog NIM811 binds to CypA and blocks its PPIase activity in vitro but does FIG. 4. Vpr expression is dependent on CypA but not on proteasome activity. A, parallel cultures of HeLa cells transfected with pCMV-FLAG-Vpr were treated with increasing concentration of CsA and subjected to a 5-min pulse followed by an up to 30-min chase. Vpr proteins from the cytosol or nucleus fraction were immunoprecipitated, separated by SDS-PAGE, and detected by fluorography. B, in a similar pulse-chase experiment cells were incubated in the absence or presence of 10 M proteasome inhibitors zLLL and LC, in addition to CsA, and Vpr from the cytosolic fraction was analyzed.

FIG. 3. CypA inhibitor CsA induces specific and rapid loss of Vpr.
A, parallel cultures of HeLa cells were transfected with the env-deleted HIV-1 NL4 -3 subviral clone pNLenv1 or the CMVdriven expression vector pCMV-FLAG-Vpr, and pulse-chase analyses were performed in the presence or absence of 50 g/ml CsA. Gag and Vpr were immunoprecipitated from cell lysates using anti-CA, anti-Vpr, or anti-FLAG antibodies, respectively, separated in 16% SDS-PAA gels, and analyzed by fluorography. B, relative amounts of FLAG-Vpr were plotted as a function of time. Pfu, arbitrary phosphorimager units. not exhibit immunosuppressive activity in vivo since the drug-CypA complex does not bind to calcineurin (53). Similarly, SFA, a macrolide from actinomycetes, also binds with high affinity to CypA and inhibits its PPIase activity but does not affect the phosphatase activity of calcineurin when complexed with CypA (54). HeLa cells transfected with pCMV-FLAG-Vpr were treated with these different CypA inhibitors, and total poly(A) mRNA was isolated and hybridized with [ 32 P]-labeled DNA probes specific for either Vpr or, as internal control, glyceraldehyde-3-phosphate dehydrogenase (Fig. 5). Clearly, none of the CypA inhibitors affected either the level of transcription or splicing of vpr mRNAs. Thus, the loss of Vpr cannot be ascribed to an effect at the level of mRNA integrity.
Although we cannot entirely rule out the possibility that CsA induces degradation of Vpr, pulse-chase (Fig. 3), steady-state Western blot (not shown), and mRNA analysis (Fig. 5) point to a specific requirement of CypA activity for the de novo synthesis of Vpr. To test this hypothesis, we employed Western blot kinetic analyses to measure the effect of CypA inhibition on the levels of Vpr that accumulate after induction of Vpr expression. HeLa cells were transfected with pCMV-FLAG-Vpr, and the kinetic analyses was initiated 5 h post-transfection, the earliest time when Vpr expression was detectable in this system. The culture was divided up, and upon addition of CsA, NIM811, or SFA aliquots of the cultures were removed during an 8-h treatment period, and total cell lysates were analyzed in parallel by Western blotting using Vpr-specific antibodies (Fig. 6A). For control, the Western blots were probed with antibodies specific for CypA or actin. Although CsA did not affect the steady-state level of these cellular proteins, the de novo synthesis of Vpr was stalled immediately after the addition of any of the CypA inhibitors (Fig. 6A). To quantitate this effect, the Vpr signal was measured by densitometry, and the value of Vpr detected in the absence of CsA at the beginning of the harvest period was set arbitrarily to 100%. Accordingly, the steady-state level of Vpr increased up to 45-fold during the 8-h harvest period in the control culture, whereas in the presence of inhibitors Vpr levels remained steady and increased only slightly after 4 h of treatment with CsA (Fig. 6B). Inhibition of the de novo synthesis of Vpr was also observed when parallel cultures were treated with CypA inhibitors NIM811 or SFA (Fig. 6, A and B), supporting the conclusion that for efficient translation of Vpr the PPIase activity of CypA is selectively required, rather than the calcineurin pathway.
To control for the possibility that these inhibitors nonspecifically block the de novo synthesis of Vpr, we tested the effect of the CypA inhibitors on the synthesis of HIV-1 Gag in parallel. HeLa cells were transfected with pNLenv1, and 8 h post-transfection when Gag expression was first detectable, aliquots of cells were harvested after 2, 4, and 8 h of incubation in the presence and absence of the inhibitors and analyzed for expression and processing of Gag (Fig. 6C). None of the CypA inhibitors affected the accumulation of newly synthesized Gag proteins. Consistent with the pulse-chase data (Fig. 3), the conversion of Pr55 into CA remained unaffected.

FIG. 6. CypA activity is required for de novo synthesis of Vpr.
A, parallel cultures of HeLa cells transfected with pCMV-FLAG-Vpr were treated with 50 g/ml CsA, 10 g/ml NIM811, or 10 g/ml SFA starting 5 h post-transfection. Aliquots of cell lysates harvested up to 8 h upon initiation of treatment were analyzed by Western blot using anti-FLAG, anti-CypA, or anti-actin antibodies, respectively. B, relative intensities of Vpr signals were calculated by densitometric scanning, and -fold increase of Vpr expression (100% ϭ 0 h, no inhibitors) was plotted as a function of time. C, for control, HeLa cells were transfected with pNLenv1, and samples from cultures treated as in A were analyzed by Western blot with anti-CA antibodies.
In summary, inhibition of CypA activity produces a specific and sudden loss of de novo synthesized Vpr. Since this effect was detectable to the same extend, both by Western blot (Fig.  6) and with immunoprecipitation using various epitope-different anti-Vpr antibodies (see Figs. 3 and 4), misfolding of Vpr does not appear to be induced by the CypA inhibitors. All evidence gathered so far indicates that CypA activity regulates the expression of Vpr either co-translationally or very shortly after its synthesis.
Proline Residues in Vpr Confer Sensitivity to CypA Inhibition-As noted above, the observation of the isomerization of imidic bonds associated with proline residues of Vpr (19) and the presence of conserved proline residues resembling a potential CypA binding domain (Fig. 2, top) prompted us to analyze the impact of mutating these proline residues on Vpr expression. This was explored by expression analyses using a CMVdriven expression system, pCMV-3HA-Vpr, that directs the expression of HA-tagged Vpr (Fig. 7). Western blot kinetic analyses similar to that described in Fig. 6A were conducted, and Vpr proteins accumulating after treatment were analyzed with anti-HA antibodies. Upon CsA addition, the typical arrest of Vpr synthesis was observed for the wild type protein (Fig.  7A). In contrast, when proline in position 35 was exchanged for asparagine (VprP35N), the accumulation of de novo synthesized mutant protein occurred at near wild type levels, and further expression of this mutant protein was not inhibited by any of the CypA inhibitors including CsA, NIM811, or SFA (Fig. 7B). Thus, Pro-35, which was also important for binding of Vpr to CypA in vitro (Fig. 2), appears to be required for the negative regulation of Vpr synthesis by CypA inhibitors. In contrast, mutation of prolines 5, 10, and 14 (VprP5,10,14A) forming a proposed CypA binding domain (Fig. 2, top) to alanines resulted in a dramatic decrease in Vpr expression even in the absence of CypA inhibitors, and further addition of CsA (Fig. 7C) or other CypA inhibitors (not shown) had no detectable effect on these low Vpr levels. The low expression level of the mutant VprP5,10,14A that was barely detectable in all experiments did not increase during the 8 h of Western blot kinetics. Thus, in both mutants, VprP5,10,14A and VprP35N, elimination of specific proline residues negates the impact of CypA inhibitors on Vpr expression. Further, whereas Pro-35 is dispensable for Vpr expression yet confers sensitivity to CypA inhibitors, prolines 5, 10, and 14 are necessary for efficient synthesis of Vpr in the absence of CypA inhibitors. Clearly, more detailed mutational analyses are necessary to fully unravel the functional contribution of each of the four proline residues in the N terminus of Vpr to the CypA-mediated regulation in Vpr synthesis.
CypA Activity Regulates Expression and Biological Function of Vpr in HIV-1-infected T Cells-Besides blocking PPIase activity, CypA inhibitors, particular CsA, exert other less specific effects on cell metabolism. To further test the specific requirement for CypA for the observed Vpr effects in the absence of CypA inhibitors, we took advantage of a CD4 ϩ CypA knock-out Jurkat cell line, PPIA Ϫ/Ϫ , that was created previously by inactivating both CypA alleles (40). Both parental and the PPIA Ϫ/Ϫ cells were infected with HIV-1 NL4 -3 , and at peak virus replication each culture was subjected to pulse-chase radio labeling similar to that described in Fig. 3A. To compare relative levels and kinetics of intracellular Vpr expression, both Vpr and Gag proteins were recovered by immunoprecipitation and analyzed by fluorography (Fig. 8A). Compared with the parental cells, Vpr expression was compromised relative to the CA signal in the PPIA Ϫ/Ϫ cells. Similar to the effects observed with CypA inhibitors (see Figs. 3, 6, and 7), the deficiency of CypA activity did not influence the half-life of Vpr during the 8-h chase period, and the decline in Vpr expression was observed immediately after the pulse labeling period. Thus, genetic inactivation of CypA resulted in effects on Vpr expression similar to those obtained by treating wild type cells with CypA inhibitors, indicating that this enzyme is directly involved in the regulation of Vpr expression.
Since Vpr is incorporated into HIV-1 virions (49), the CypAdependent decrease in Vpr expression could lead to lower levels of intravirion Vpr. To test this possibility, parallel cultures of parental or PPIA Ϫ/Ϫ Jurkat cells were infected with either wild type HIV-1 NL4 -3 or the vpr-deficient virus HIV-1 NL4 -3⌬Vpr . At peak virus replication, virions present in the supernatants were pelleted and analyzed by Western blotting. Samples were normalized for CA protein, because it has been shown that inactivation of CypA does not affect the content and processing of Gag in HIV-1 virions (40). A 5-fold reduction in the amount of Vpr incorporated into virions produced from the PPIA Ϫ/Ϫ Jurkat cells was observed compared with parental Jurkat cells (Fig. 8B), which is consistent with the loss of Vpr expression in the absence of CypA activity (Fig.  8A). As yet we have no evidence that CypA activity regulates interaction of Vpr with the C-terminal p6 Gag region of Pr55, which mediates incorporation of Vpr into virions (data not shown). This supports the hypothesis that CypA selectively affects only the de novo synthesis of Vpr and that reduction of intracellular Vpr in the absence of CypA activity leads to a loss of Vpr in released virions.
As reported previously (40), gene targeting of CypA has the potential to select for changes in the expression profile of other enzymes with PPIase activity that might compensate for the loss of CypA. Therefore, we tested the effect of CypA inhibitors on the incorporation of Vpr into virions produced from the parental Jurkat cells. Parallel cultures of cells infected with wild type HIV-1 NL4 -3 were treated with increasing concentration of CsA or the non-immunosuppressive CypA inhibitor SFA, and virions pelleted from aliquots of cell culture supernatants were analyzed by Western blotting (Fig. 8B). The ratio of the signal intensities for CA and Vpr detected in the blot revealed a dose dependent reduction of virus-associated Vpr reaching ϳ3-fold reduction at 3 g/ml CsA. Treatment of Jur- kat cells with SFA caused a similar effect, reaching ϳ3-fold decline of virion-associated Vpr at 5 g/ml of the drug (Fig. 8C). In summary, these results demonstrate that CypA activity is required for efficient Vpr expression in HIV-1-infected T cells.
Next, we investigated whether CypA activity was required for functional effects of Vpr. First, the effect of CsA on Vprmediated G 2 cell cycle arrest was studied in 293T cells transiently producing relatively high levels of Vpr using the pCMV-3HA-Vpr expression system. Cells were co-transfected with pCMV-green fluorescent protein to select for Vpr-expressing cells. Treatment of Vpr-expressing cells with 50 g/ml CsA completely abrogated the Vpr-mediated arrest of the cell cycle in G 2 phase (data not shown). Further, the importance of CypA activity for Vpr-mediated G 2 arrest was analyzed in HIV-1infected PPIA Ϫ/Ϫ cells. Both parental and PPIA Ϫ/Ϫ Jurkat cells were infected with wild type HIV-1 NL4 -3 or the mutant HIV-1 NL4 -3⌬Vpr , and at peak viral replication DNA content of CA-expressing cells was analyzed (Fig. 9). As reported previously in other cell systems (2, 4, 6, 7, 13), infection of the wild type Jurkat T cells resulted in a typical Vpr-dependent G 2 cell cycle arrest as demonstrated by an inversion of the G 2 /M to G 0 /G 1 ratio compared with cells productively infected with HIV-1 NL4 -3⌬Vpr (Fig. 9, left panel). However, in the infected PPIA Ϫ/Ϫ cells, no such arrest was demonstrated as the G 2 /M to G 0 /G 1 ratio was not markedly altered following infection with wild type HIV-1 NL4 -3 (Fig. 9, right panel). These results support the notion that CypA activity is important for expression of biological active Vpr.

DISCUSSION
In our recent structural studies of full-length Vpr and its N-terminal domain we drew attention to the heterogeneity of the molecular species present in apparently homogeneously synthesized Vpr peptides (13,19). We found that this phenomenon occurs as a consequence of cis/trans isomerization of imidic bonds associated with the four highly conserved proline residues in the N terminus of the molecule. Specifically, residues Pro-14 and Pro-35 of HIV-1 Vpr show abnormally high content of cis-isomers that could imply their relevance for the biological activity of Vpr as the interconversion of such isomers should require the presence of a PPIase activity for efficient protein folding. These circumstantial data led us to explore the possibility of a mutual interconnection between Vpr and CypA, an abundant host factor that interacts functionally with viral structure proteins Gag and Env and is specifically incorporated into HIV-1 virions. In terms of the CypA-Gag interaction, it was proposed that CypA facilitates disassembly of the viral RNA containing core following virus entry and thus supports the efficient reverse transcription of the HIV-1 genome (for review see Refs. 30 and 31). In addition to these post-entry steps, it has been proposed that CypA enhances virus attachment to the host cell membrane through interactions with heparans (for review see Ref. 55) and after membrane fusion through interaction with CD147 (56). We now report that, in addition to these early steps in viral replication, CypA is also important for the de novo synthesis of Vpr and that in the absence of CypA, Vpr incorporation into virions is diminished, and Vpr-mediated cell cycle arrest is sharply impaired in HIV-1-infected T cells. Vpr immunostaining was then compared between samples, and densitometry was used to determine the ratio of Vpr to p24 CA . C, virions released from HIV-1 NL4 -3 -infected wild type Jurkat cells treated with increasing concentration of SFA or CsA were analyzed by Western blot. Samples were normalized to equimolar input of p24 CA , allowing for the densitometric analysis and calculation of the Vpr to p24 CA protein expression ratio as indicated. C, virions released from acutely infected wild type Jurkat cells treated with increasing concentrations of SFA or CsA were analyzed by Western blot, and the ratios of p24 CA and Vpr proteins are indicated. Direct biochemical evidence for CypA-Vpr interactions was obtained by in vitro binding studies and by pulldown experiments with cell extracts. Surface plasmon resonance data for a number of Vpr peptide fragments further indicated that this interaction involves the N-terminal domain of Vpr including the requisite presence of Pro-35. The form of the resonance spectra implies a rapid association and dissociation indicative of the formation of a low affinity complex between CypA and Vpr, which would be typical for an enzyme-to-substrate contact. The relatively weak association between Vpr and CypA was further supported by the pulldown experiments, where CypA-Vpr complexes could be detected only in the presence of low stringency detergents or in the absence of reducing agents allowing disulfide complex formation between CypA and Vpr. Unequivocally, the importance of Pro-35 for CypA-Vpr interaction and the responsiveness of Vpr expression to CypA inhibitors is consistent with our recent NMR studies (19) positioning Pro-35 in a flexible region between two helices that are present in the most structured conformations of Vpr (12,16) and thus critical for the folding in the N-terminal region.
From the results gathered so far, we cannot differentiate between the two potential functions of CypA in Vpr synthesis, namely whether it subserves a chaperone-like role simply by binding to Vpr or whether its enzymatic function is required involving the isomerization of proline residues in Vpr. A similar situation holds for the CA-CypA interaction, where the prevalent view is that CypA most likely acts as the sole binding partner rather than functioning as a PPIase in CA folding. Even though direct evidence of the latter has been presented recently (57), the incorporation of a catalytically inactive form of CypA into HIV-1 virions that was able to support viral infectivity argues against an isomerase function of CypA in Gag folding (58). In contrast to the CA-CypA interaction, the unusual high content of cis conformers in Vpr argues for a specific requirement of the isomerase function of CypA for Vpr expression. Further, the CypA-Vpr interaction appears to be different from that of CA with CypA. This is highlighted by the differential sequences of the primary binding sites of CypA in both HIV-1 proteins; the CA-CypA interaction has been shown to be mediated exclusively by an exposed loop that spans Pro 85 to Pro 93 and incorporates the critical Gly-Pro 90 recognition motif in mature CA (59,60) as opposed to the sequence environment of the critical residue Pro 35 in Vpr. In addition, the Gly-Pro 90 motif within the CypA binding site in CA is in a trans-proline conformation, whereas the Phe-Pro 35 motif in Vpr would favor a cis-proline conformation according to previous reports on the effect of neighboring side chains on extent of cis/trans isomerization (61). Consistent with this prediction, our NMR studies on Vpr revealed an unusually high cis-content for Pro 35 (19). Thus, these data provide circumstantial evidence that favors an isomerization function of CypA necessary for efficient expression of Vpr. According to this model, CypA binds primarily to Pro 35 , leading to additional isomerization effects on neighboring Pro residues, most importantly Pro 14 that like Pro 35 exhibits unusual high content of cis conformers. Under the conditions used for plasmon resonance spectroscopy, we were unable to demonstrate an interaction of CypA with N-terminal Vpr peptides containing the potential CypA binding motif consensus P 5 XXXGPXRXP 14 , whereas binding to the Pro 35 surrounding sequence was readily detected. In contrast, mutational analyses suggested a role for these proline residues in the responsiveness of Vpr expression to CypA inhibitors. Further, in the context of the full-length molecule, the CypA-Vpr interaction might also involve other domains of Vpr, particularly Cys-76 for disulfide bond for-mation, which might occur temporarily during CypA-Vpr engagement.
The recognition site for CypA in HIV-1 CA harbors the sequence motif PXXXGPXXP 93 (see above), whereas a similar, albeit not identical, sequence, PXXXGPXRXP 14 , is highly conserved at the N terminus of HIV-1 Vpr (Fig. 2). In contrast, no such conserved motif exists in known Vpr sequences derived from HIV-2 or SIV isolates. The consensus PPEDEGPQREP 15 is present in Vpr derived from SIV subtype D isolates such as SIVsm, SIVmac, and SIVstm. Furthermore, the following sequences are present in the N terminus of Vpr derived from various SIV isolates: PPEDEGPPREP 22 in SIV sun sun , PPED-FGPPREP 23 in SIV lhoest lhoest , and EDQGPPREP 19 in SIV mnd gb1 . Thus, although the PXXXGPXRXP 14 motif is not 100% conserved in Vpr from HIV-2 and SIV, proline-rich sequences are present in most isolates, and some closely resemble those of Vpr from HIV-1. Beside a conserved N-terminal PXXXXPPG motif, proline-rich sequences are also conserved in the C-terminal domain of SIV Vpx (62). Notably, in HIV-2 and SIV only Vpx contributes to nuclear import of the preintegration complex. However, the question of whether CypA also plays a key role in stabilizing the expression of Vpr and Vpx from HIV-2 or SIV, respectively, clearly merits additional studies.
Our data on G 2 arrest are the first to demonstrate direct evidence for the functional relevance of a CypA-Vpr interaction. In contrast to the CA-CypA contact, whose significance was demonstrated by reduction of viral replication through genetic or pharmaceutical interference with CypA function (30,40,63), we have no data as yet that show the relevance of the CypA-Vpr interaction for viral replication. There are several reasons why such data will be difficult to acquire. First, the Vpr phenotype reflected upon viral replication is generally only measurable in tissue culture of primary lymphocytes or preferentially primary monocytes/macrophages, cell systems for which CypA inhibitors are particularly toxic. Second, the CypA-mediated effect on Vpr expression will require relatively high concentrations of CsA, NIM811, or SFA when compared with the effective concentration of these inhibitors necessary to block CA-CypA interaction. Noteworthy is also our observation that CypA Ϫ/Ϫ Jurkat cells are completely resistant to Vprinduced cell cycle arrest when infected with wild type HIV-1. Although this phenomenon most likely stems from the deficiency of Vpr expression, it does not completely rule out the possibility that G 2 arrest requires CypA activity in a manner occurring independently of its impact on synthesis of Vpr.
Regardless of the mechanism(s), the complete loss of G 2 arrest in CypA Ϫ/Ϫ Jurkat cells raises the possibility that CypA inhibitors could be used to block Vpr function in vivo. It has been shown previously (35,36) that CypA inhibitors interfere with import of the preintegration complex into the nucleus. In these studies it was also reported that cyclosporins, in particular the non-immunosuppressive derivative NIM811, exhibited potent anti-retroviral activity in vitro. NIM811 interferes at two stages of the viral replication cycle, namely the production of infectious virus particles and the nuclear localization of the preintegration complex. This latter observation could potentially arise from the inhibitory effect of this drug on Vpr expression and incorporation of Vpr into virions.
The inhibitory effect of CypA blockers such as CsA or NIM811 on virus replication were studied in T-cell cultures infected with vpr-deficient viruses such as HIV-1 HXB2 (29) and HIV-1 IIIB (41) or with Vpr-expressing viruses such as HIV-1 NL4 -3 , HIV-1 ELI , (28,64) or chimeric SIV mac 239-HIV-1CA viruses (63). In these studies, the responsiveness of viral replication to CypA inhibitors was strictly dependent on the presence of a CypA binding motif in CA. Further, the replication of CsA-resistant CA mutants of HIV-1 NL4 -3 in Jurkat T cells was not affected by treatment with CsA up to 2.5 M (64). However, these studies were not designed to analyze the importance of CypA-Vpr interaction for HIV-1 replication for the following reasons. First, the Vpr phenotype is almost undetectable in rapidly dividing T cells. Second, the concentration necessary to interfere with the CypA-Vpr interaction is about five times higher than that required to interfere with the CA-CypA interaction. Third, the importance of virus-associated CypA for the biological role of Vpr so far is still unknown. The effect of CsA or other CypA inhibitors on HIV-1 replication has never been studied in cultured macrophages, where effect of the Vpr phenotype on viral replication is most evident. Thus, our ongoing studies are exploring whether CypA inhibitors or mutation of Pro-35 in Vpr alters HIV infection of primary cells.
As yet we have little evidence as to how, where, and when the interaction of CypA with Vpr leads to diminished Vpr expression. Thus far, our findings support an effect of CypA on the de novo synthesis of Vpr. This effect could occur either co-translationally or shortly after translation as mature Vpr protein is not affected nor is nucleocytoplasmic shuttling or virion incorporation of Vpr altered. One possible mechanism of action involves the ribosome tunnel hypothesis (65). CypA inhibitors might block or re-route the nascent Vpr peptide chain as soon as the cis/trans labile N terminus emerges from the ribosome complex. If true, such a mechanism would be the first evidence for an important role of PPIase activity in protein translation. Clearly, further studies are necessary to validate this hypothesis. In this context it will be also necessary to explain why the CypA inhibitors do not completely prevent Vpr synthesis. It is conceivable that other proteins with PPIase activity, perhaps from the 15 known human CyPs, have the capacity to regulate Vpr expression and thus compensate to some extent for the inactivated or missing CypA function. This seems particularly likely because residual Vpr expression is detected even in CypA Ϫ/Ϫ Jurkat cells.
Since Vpr supports virus replication, particularly in nondividing cells, and has been reported to sustain pathogenesis in SIV-infected monkeys (66), this accessory protein could form a target for the development of new anti-retroviral therapies. Several approaches have been reported that interfere with Vpr function. These include expression of dominant negative mutants of Vpr (25), the use of antagonists of the glucocorticoid receptors (47), or peptide inhibitors (67). In contrast to these experimental in vitro approaches, the potential anti-retroviral effect of CsA has already been investigated. The immunosuppressive drug CsA blocks activation of gene expression for interleukin-2, interleukin-4, and the interleukin-2 receptor in T cells (68) and has been suggested to be able to block HIV-1 replication in vivo in two ways (for review see Ref. 69), indirectly by interfering with T cell activation and more directly by binding to Gag polyproteins and thus reducing infectivity of released HIV-1 virions. However, the first clinical studies using monotherapy with CsA in patients with chronic HIV-1 infection or at advanced stages of AIDS revealed only subtle beneficial effects of long term CsA treatment on preventing disease progression (70). In contrast, CsA treatment of rhesus monkeys acutely infected with SIV showed advantageous effects of CsA on antigenemia and loss of CD4 ϩ T cells (66). Further, it is well established that primary HIV infection is associated with massive immune activation, and it was shown recently (71) that this very critical period for initiation of highly active antiretroviral therapy can benefit from combining highly active antiretroviral therapy with CsA treatment. Although it was concluded in these studies that CsA-mediated block in T-cell activation during the early phase of primary HIV-1 infection was the underlying cause for the long lasting beneficial impact of the combination therapy with CsA on disease development, one could also speculate that the inhibitory effect of CsA on Vpr function may have contributed to the enhancing effect of CsA on the antiviral efficiency of highly active antiretroviral therapy. In conclusion, the availability of non-immunosuppressive CypA inhibitors that are less toxic than CsA might hold promise as novel anti-retroviral drugs by impairing Vpr production. These drugs should be re-evaluated clinically in the light of these new findings.