N-terminally truncated POM121C inhibits HIV-1 replication

Recent studies have identified host cell factors that regulate early stages of HIV-1 infection including viral cDNA synthesis and orientation of the HIV-1 capsid (CA) core toward the nuclear envelope, but it remains unclear how viral DNA is imported through the nuclear pore and guided to the host chromosomal DNA. Here, we demonstrate that N-terminally truncated POM121C, a component of the nuclear pore complex, blocks HIV-1 infection. This truncated protein is predominantly localized in the cytoplasm, does not bind to CA, does not affect viral cDNA synthesis, reduces the formation of 2-LTR and diminished the amount of integrated proviral DNA. Studies with an HIV-1-murine leukemia virus (MLV) chimeric virus carrying the MLV-derived Gag revealed that Gag is a determinant of this inhibition. Intriguingly, mutational studies have revealed that the blockade by N-terminally-truncated POM121C is closely linked to its binding to importin-β/karyopherin subunit beta 1 (KPNB1). These results indicate that N-terminally-truncated POM121C inhibits HIV-1 infection after completion of reverse transcription and before integration, and suggest an important role for KPNB1 in HIV-1 replication.

Nuclear import of host cell proteins is generally mediated by importin family members. Importin-α family proteins mainly recognize nuclear-localizing signals (NLSs) of cargo proteins and importin-βfamily proteins are usually necessary for docking to the NPC and translocation through the nuclear pore [11]. Importin-β/karyopherin subunit beta 1 (KPNB1) protein was shown to bind to and import the HIV-1 proteins Tat and Rev independently of Importinα [19]. Recent reports showed the TNPO3, a member of the importin-β protein family [20], prevents cleavage and polyadenylation specificity factor subunit 6 (CPSF6)-mediated capsid stabilization in the cytoplasm and contributes to efficient nuclear entry of the CPSF6-associated HIV-1 preintegration complex (PIC) [12,21]. There have been conflicting reports about the role of an importin-β family member importin-7 in HIV-1 nuclear import [22,23]. Thus, the roles of multiple host cell factors in HIV-1 nuclear import remain to be further studied.
In the present study, we employed cDNA expression screening to identify an N-terminally truncated form of POM121C which strongly inhibited HIV-1 replication. POM121C is a nucleoporin located in the middle of the nuclear pores which plays an essential role in the formation of NPC [24]. Here, we describe how this truncated protein interferes with HIV-1 replication.

Generation of cells stably expressing POM121C mutants
HEK293, HeLa, Jurkat and MT4C5 cells were transduced with retrovirus vectors for POM121C mutants and puromycin resistance and express POM121C-mutants. Cells stably expressing POM121C-mutants were obtained after selection with 4 μg/ml of puromycin for HEK293 and HeLa or 2 μg/ml of puromycin for Jurkat and MT4C5. PHA-stimulated PBMCs were transduced with lentivirus vectors for POM121C mutants and Blasticidin S resistance, and were cultured in the presence of 6 μg/ml of Blasticidin S.

Immunocytochemistry
HeLa cells (1 × 10 4 ) were cultured for 48 h at 37˚C in Nunc Lab-Tek II 8-well glass Chamber Slides (Thermo Fisher Scientific, Waltham, MA). Cells were then fixed with 4% paraformaldehyde for 10 min at room temperature followed by two washes in PBS(-). Fixed cells were permeabilized for 20 min at room temperature with 0.5% Triton X-100 in PBS(-) prior to incubation with antibodies. For staining, cells were incubated at 4˚C overnight with primary antibodies diluted appropriately in 20% ImmunoBlock (DS Pharma Biomedical Co., Ltd., Osaka, Japan)/PBS(-). Cells were washed twice with PBS(-) and incubated with Alexa Fluor 594-conjugated secondary antibodies (Thermo Fisher Scientific, Waltham, MA) and Hoechst33342 (Thermo Fisher Scientific, Waltham, MA) for 2 h at room temperature. Samples were then washed twice with PBS(-) and mounted with VECTASHIELD (Vector Laboratories, inc., Burlingame, CA

Quantification of viral cDNA
Measurement of reverse-transcribed viral cDNA in HIV-1-infected cells was performed as described previously [27]. The integrated viral DNA was measured as described previously [39]. Briefly, the first PCR was carried out with an Alu-sequence-specific sense primer (Alu-HIV) and an antisense primer recognizing the LTR/gag region of the HIV sequence (M661). The first-round PCR samples were diluted with deionized distilled water by a factor of 10, and 5 μL was used as the template for the real-time PCR assay for measuring R/U5 DNA using M667, AA55 and HIV-FAM. For standardization, β-globin cDNA was quantified as described previously [40]. Real-time PCR was carried out with the StepOnePlus Real-Time PCR system (Applied Biosystems, Carlsbad, CA). The ratios of viral cDNA toβ-globin cDNA levels were calculated.

Preparation of naked HIV-1 capsid core
Envelope-stripped HIV-1 virions (naked capsid cores) were prepared as described previously [41]. Briefly, HIV-1-containing culture supernatants were prepared by transiently transfecting HeLa cells with pNL4-3 using LipofectAMINE LTX PLUS (Invitrogen Corp., Carlsbad, CA). Two ml of 20% sucrose solution was placed at the bottom of SW55 centrifuge tubes and overlaid with 3 ml of the HIV-1-containing culture supernatant described above. Samples were then centrifuged for 60 min at 35,000 rpm at 4˚C. Particulate HIV-1 was resuspended with PBS (-) containing a protease inhibitor cocktail (Nacalai Tesque, Inc., Kyoto, Japan). This suspension was loaded onto the top of a discontinuous sucrose density gradient composed of 1.0 ml 30% sucrose solution at the bottom of SW55 centrifuge tubes covered or not covered by 1.0 ml 0.1% Triton X-100 in 10% sucrose solution and then centrifuged in an SW55Ti rotor for 120 min at 35,000 rpm at 4˚C. Particulate CA proteins were subjected to Western blotting with anti-HIV-1 env (gp120) or HIV-1-positive pooled serum from infected individuals (subtype B) to verify removal of HIV-1 env and MA proteins.

TOF-mass spectrometry analysis to identify proteins interacting with POM121C mutants
Proteins were identified by liquid chromatography nano-electron spray ionization tandem mass spectrometry (LC-nESI-MS/MS) (QTRAP 5500 LC/MS/MS System, AB SCIEX, Concord, ON, Canada). Proteins excised from the gels were analyzed by LC-nESIMS/MS after separation on a HiQ sil C18W-3P column (0.1 mmF × 100 mm, KYA TECH Corporation, Tokyo, Japan). The proteins were eluted at a flow rate of 300 nL/min using water containing 0.1% (v/v) formic acid as the eluent A and acetonitrile containing 0.1% (v/v) formic acid as the eluent B with a linear gradient from 5% B to 45% B in 70 min. The data were analyzed using ProteinPilot™ software (Version 3, Applied Biosystems, Warrington, UK). Protein candidates were selected from those that had more than one distinct peptides with at least 95% confidence.

Statistical analysis
All data were obtained from at least three independent experiments. The average values are presented with error bars indicating the standard deviation (SD) and the statistical significance was analyzed using one-way analysis of variance (ANOVA) with Dunnett's multiple comparison test, or Student's t-test. All the statistical analyses were performed using Prism 6 software (GraphPad Software, Inc

N-terminally truncated POM121C blocks HIV-1 infection
To identify proteins that interfere with HIV-1 infection, we employed a functional screen using a cDNA expression library. The details of the method of screening were essentially the same as those previously reported [27] (Fig 1A and 1B). Stable expression of POM121C (614-987) in HEK293 cells with an influenza virus hemagglutinin (HA) tag, resulted in potent suppression of single-round infection of these cells with VSVG-pseudotyped NL4-3luc reporter virus (Fig 1C, left panel). Similar inhibition of HIV-1 infection was observed in the Jurkat cell line stably expressing POM121C (614-987) (Fig 1C, middle panel). Transduction of phytohemagglutinin/interleukin-2-stimulated peripheral blood mononuclear cells (PBMCs) with a lentivirus vector expressing HA-tagged POM121C (614-987) also resulted in a marked resistance to VSV-G-pseudotyped NL4-3luc infection (Fig 1C, right  panel). Stable expression of POM121C (614-987) in MT4C5 cells markedly reduced the infectivity of VSV-G-pseudotyped HIV-1 (Fig 1D, left panel), replication-competent HIV-1 reporter virus, NL4-3luc-RC (Fig 1D, middle panel) and NL4-3 in spreading infection assays (Fig 1D, right panel). Neither the surface expression of CD4 and CXCR4 nor cell growth was affected by stable POM121C (614-987) expression (data not shown). However, stable expression of HA-tagged full-length POM121C in HeLa cells did not significantly influence VSV-G-pseudotyped HIV-1 infection (Fig 1E, right panel). A weaker expression of full-length POM121C (Fig 1E, left panel) may partly explain this lack of inhibition, but immuno-  (Fig 2A, middle panel). POM121C (614-987) also moderately reduced production of the 2-LTR form that accumulates exclusively in the nucleus and is therefore recognized as a marker of nuclear import of viral cDNA (Fig 2A, bottom panel). The moderate and marked reduction in the 2-LTR and integrated provirus forms, respectively, in the presence of POM121C (614-987) implies possible blockade by POM121C (614-987) at the nuclear import, nuclear trafficking, or integration stage. Recent studies indicate that HIV-1 nuclear entry is regulated by viral capsid (CA) [14,42]. To investigate potential interactions between POMC121C (614-987) and the viral core in HIV-1-infected cells, we performed HIV-1 CA core binding assays with envelope-stripped HIV-1 virions. The integrity of the prepared naked capsid core was verified biochemically by Western blotting (Fig 2B) showing loss of the envelope and matrix proteins after stripping, and confirmed morphologically by electron microscopy (Fig 2C) showing the cone-shaped viral core. Pull-down assays with naked HIV-1 cores and purified OSF-tagged proteins revealed that POM121C (614-987) did not interact significantly with the capsid core (Fig 2D, panel pull down). In contrast, the association of the HIV-1 core with Cyclophilin A (CYPA), known to be an HIV-1 core-binding protein [43], was clearly demonstrated (Fig 2D, panel pull down). These results indicate that POM121C (614-987) inhibits HIV-1 infection after reverse transcription without binding to HIV-1 core.

POM121C (614-987) inhibits nuclear import and integration
To gain more insight into blockade of infectivity after completion of RT, HEK293 cells expressing POM121C (614-987) or mCPSF6-358, which had been reported to block nuclear  import of viral cDNA [15], were infected with VSV-G-pseudotyped NL4-3luc carrying the N74D mutation in the CA or the D116G mutation in the integrase (IN). The N74D mutation significantly reduces the binding affinity of mCPSF6-358 to HIV-1 CA, resulting in viral escape from restriction by mCPSF6-358 [15], while the D116G mutation disables HIV-1 for integration within host chromatin without affecting RT or nuclear entry [28]. We stably expressed POM121C (614-987) or mCPSF6-358 in HEK293 cells for single-round reporter virus infection assays. Expression of POM121C (614-987) or mCPSF6-358 was verified by Western blotting (Fig 3A). As shown in Fig 3B, mCPSF6-358 weakly affected N74D HIV-1 infection, consistent with a previous report [15]. POM121C (614-987) potently or moderately inhibited infection with N74D or D116G mutant viruses, respectively (Fig 3B, panels CA  N74D and IN D116G). Kinetic studies of viral cDNAs following infection of cells expressing POM121C (614-987) with the N74D or D116G viruses showed that the late RT form was produced at levels similar to those of control cells (Fig 3C, upper panels), while accumulation of the 2-LTR form was decreased (Fig 3C, lower panels). This suggests that the action of POM121C (614-987) does not depend on the route of nuclear import of HIV-1 cDNA. The moderate inhibition of D116G virus infection by POM121C (614-987) suggests that POM121C (614-987) not only blocks nuclear import but also affects integration of HIV-1 providing that the D116G virus lacks the integrase activity. In this case, the overall inhibitory effects of POM121C (614-987) on this mutant virus may be proportionally alleviated. Additionally, overexpression of POM121C (614-987) did not inhibit the infectivity of the VSV-Gpseudotyped murine leukemia virus (MLV)-based vector, suggesting that POM121C (614-987) expressed in the cytoplasm has no effect on gamma retrovirus infection (Fig 3D, middle  panel). A previous study had shown that CA is a pivotal determinant of retrovirus infection in non-dividing cells, using an HIV-1-based chimeric virus MHIV-mMA12CA, in which the entire Matrix (MA) and CA proteins were replaced by those of MLV [14]. Having shown that POM121C (614-987) does not inhibit MLV infection in dividing cells, we next investigated whether it inhibits infection with MHIV-mMA12CA, which should not encounter inhibition of nuclear entry in dividing cells. Indeed, POM121C (614-987) did not significantly inhibit MHIV-mMA12CA infection in dividing cells (Fig 3D, right panel). These results indicate that Gag is a genetic determinant of POM121C (614-987)-mediated HIV-1 restriction although POM121C (614-987) did not seem to interact with the capsid core (Fig 2D).
Prevention of HIV-1 infection by POM121C (614-987) requires the Cterminal α-helix structure POM121C (614-987) contains FG repeat motifs like other nucleoporins and a putative short α-helix motif at the extreme C-terminus. To determine the functional importance of these motifs, a series of truncation mutants was prepared and tested for their anti-HIV-1 activity in HEK293 cells (Fig 4A). Single-round infection studies revealed that mutants containing the α-     (Fig 4B and 4C). To assess the contribution of the C-terminal α-helix structure in POM121C (614-987) to the prevention of HIV-1 infection, we tested mutants with a small truncation or deletion of the α-helix motif of POM121C (614-987), POM121C (614-973) and POM121C (614-987: Δhelix). The deletion of the α-helix motif did not essentially alter subcellular localization (data not shown). Both αhelix motif mutants were expressed at similar levels to POM121C (614-987), but had lost inhibitory activity (Fig 4D and 4E). These results indicate that the α-helix motif plays an essential role in POM121C (614-987)-mediated HIV-1 restriction.   (Fig 5A and 5B). This fusion protein efficiently inhibited HIV-1 infection in HEK293 cells (Fig 5B, lower panel), did not affect the synthesis of the late RT product and reduced the amount of the 2-LTR form of viral cDNA in the same manner as POM121C (614-987) (Fig 5C). GST pull-down experiments using HEK293 cells stably expressing GST-POM121C (801-987) reproducibly demonstrated that several proteins specifically co-precipitated with GST-POM121C (801-987) (Fig 5D). Analyses of the isolated protein bands by mass spectrometry revealed that the most prominent, band 3, contains the karyopherin, importin subunit beta-1 (KPNB1) ( Table 2).

POM121C (801-987) binding to KPNB1 correlates with inhibition of HIV-1 infection
Immunoblotting studies revealed that KPNB1 specifically interacts with GST-POM121C (801-987) (Fig 6A, panel GST pull-down). To investigate the functional consequences of POM121C (801-987) binding to KPNB1, truncation or deletion mutants of GST-POM121C (614-987) were stably expressed in HEK293 cells (Fig 6B and 6C) and tested for HIV-1 inhibition and KPNB1 binding (Fig 6D and 6E). As expected from the mutational studies of POM121C (614-987), two GST fusion proteins lacking the C-terminal α-helix motif were unable to inhibit HIV-1 infection, whereas those containing the C-terminal part of the FG repeat motifs and the α-helix motif did block HIV-1 in the single-round infection assay ( Fig  6D). The binding of these mutants to KPNB1 correlated with the inhibition of HIV-1 infection (Fig 6E, panel GST pull-down). GST proteins linked to the extreme C-terminus of POM121C containing the α-helix motif but not the FG repeat motifs or simply to the α-helix motif failed to inhibit HIV-1 infection or to interact with KPNB1 (Fig 6F, 6G and 6H), suggesting that the α-helix motif alone is not sufficient for HIV-1 blockade or binding to KPNB1. These results suggest that POM121C (614-987)-mediated HIV-1 inhibition is closely linked to its binding to KPNB1. Transient transfection with the proviral plasmid pNL4-3luc allows for bypassing the replication processes up to the integration stage and mimics the expression of viral genes from the integrated HIV-1 provirus. EV-control-or POM121C (614-987)-expressing HEK293 cells were transfected with the NL4-3luc provirus plasmid which has the luciferase gene in the nef position. This resulted in similar levels of reporter gene activity and newly-synthesized viral protein (Fig 7A and 7B). Likewise, POM121C (614-987) did not affect the release of HIV-1 CA protein from cells transfected with the pNL4-3 provirus plasmid (Fig 7C). These results suggest that POM121C (814-987) does not inhibit the late phase of the HIV-1 replicative cycle.

Discussion
Based on similar late RT product synthesis 24 h post infection, we have shown here that POM121C (614-987) does not bind to the capsid (CA) core nor interfere with reverse transcription (Fig 2). After completing reverse transcription, the viral pre-integration complex (PIC) enters the nucleus by passing through the nuclear pore and approaches the host genome. One intriguing finding in the current study was that POM121C (614-987) markedly reduces the amount of the integrated provirus form, but the 2-LTR form is reduced to a lesser extent (Fig 2A). This is reminiscent of the difference in 2-LTR production levels and integrated forms previously shown in Nup153-depleted cells, in which HIV-1 nuclear entry was blocked [18]. Perhaps POM121C (614-987) localized in the cytoplasm primarily impairs nuclear entry of HIV-1, and this may indirectly affect the proper intra-nuclear migration process and eventually limit integration. HIV-1 CA plays a pivotal role in nuclear import of viral DNA and this may potentially be a primary cause of POM121C (614-987)mediated inhibition. Indeed, POM121C (614-987) inhibited HIV-1 but not MHIV-mMA12CA in dividing cells, suggesting that viral capsid defines the mode of nuclear entry and therefore determines whether POM121C (614-987) blocks infection or not. Accumulating evidence indicates that HIV-1 utilizes a particular pathway for nuclear entry that depends on CYPA, cleavage and polyadenylation-specific factor 6 (CPSF6) and NUP358, thereby evading innate immune sensors in the cytoplasm [42]. The HIV-1 capsid mutant N74D, which is not targeted by CPSF6, can use a different pathway for nuclear entry that does not involve CPSF6 nor NUP358, but is recognized by cytoplasmic innate immune sensors [44]. The present results that POM121C (614-987) inhibited infection with wild-type or N74D HIV-1 in a similar manner indicate that POM121C (614-987) suppresses both NUP358-dependent and -independent pathways (Fig 3B). The production of similar amounts of the late RT product but marked decrease of the 2-LTR form of viral DNA in cells infected with the D116G mutant HIV-1 further support the idea that blockade by POM121C (614-987) occurs before integration. Taking these results together, we conclude that POM121C (614-987) inhibits HIV-1 replication after completion of reverse transcription and before the integration event.
To investigate how POM121C (614-987) blocks HIV-1 replication during nuclear import and intra-nuclear trafficking of HIV-1 DNA, we sought proteins binding POM121C (614-987) involved in HIV-1 inhibition. Binding of POM121C to viral components remains controversial. Fouchier et al. [45] reported yeast two-hybrid interaction of POM121 with Vpr, while Le Rouzic et al. [46] did not detect this interaction. Using immunoprecipitation and GST pulldown assays, we investigated whether POM121C interacts with Vpr but failed to detect any positive signal. Attention was therefore directed towards host cell factors that are targeted by POM121C (614-987). Because POM121C (801-987) proved to be sufficient to inhibit HIV-1 infection when fused with GST (Fig 4C), this fusion protein was used as bait. In this way, we have identified a number of candidates by MS at a good confidence level (Table 2). We selected KPNB1 for further study because this protein was found in the most prominent band 3 and had the highest number of matches. Importantly, the KPNB1 binding properties of POM121C mutants closely correlated with their level of HIV-1 blockade (Fig 6). The binding of POM121C mutants to KPNB1 is, however, not completely unexpected given that several previous studies demonstrated interactions between FG sequence repeats found in many nucleoporins and transport factors including KPNB1 [10,47]. POM121C may belong to those nucleoporins that sequentially dock and undock with KPNB1/PIC complexes when HIV-1 PIC transits through nuclear pore complexes. Our attempts to determine whether endogenous full-length POM121C interacts with KPNB1 by means of immunoprecipitation experiments with two different antibodies specifically recognizing POM121C or KPNB1 were unsuccessful due in part to the limited expression of full-length POM121C in the nucleus. Thus, the possibility remains that endogenous POM121C physically interacts with KPNB1 under physiological conditions. To confirm the importance of endogenous POM121C or KPNB1 for HIV-1 replication, we had repeatedly attempted to knock-down POM121C or KPNB1 using shRNA, but unfortunately, depletion of these essential protein seriously affected the viability of cells, which made it impossible to assess HIV-1 infection. In future studies, it will be important to determine the exact role of POM121C and KPNB1 in the nuclear import and integration of HIV-1 DNA.

Conclusions
Accumulating evidence indicates that the HIV-1 capsid is an essential determinant of nuclear translocation of viral cDNA and that many host cell factors including karyopherins and nuclear pore complex components mediate this process. We have shown in this article that an N-terminally truncated form of POM121C potently inhibits HIV-1 infection after completion of reverse transcription and before integration. This inhibition closely correlates with its binding to KPNB1, suggesting an important role for this karyopherin in HIV-1 replication. Takeuchi