Host cell species-specific effect of cyclosporine A on simian immunodeficiency virus replication

Background An understanding of host cell factors that affect viral replication contributes to elucidation of the mechanism for determination of viral tropism. Cyclophilin A (CypA), a peptidyl-prolyl cis-trans isomerase (PPIase), is a host factor essential for efficient replication of human immunodeficiency virus type 1 (HIV-1) in human cells. However, the role of cyclophilins in simian immunodeficiency virus (SIV) replication has not been determined. In the present study, we examined the effect of cyclosporine A (CsA), a PPIase inhibitor, on SIV replication. Results SIV replication in human CEM-SS T cells was not inhibited but rather enhanced by treatment with CsA, which inhibited HIV-1 replication. CsA treatment of target human cells enhanced an early step of SIV replication. CypA overexpression enhanced the early phase of HIV-1 but not SIV replication, while CypA knock-down resulted in suppression of HIV-1 but not SIV replication in CEM-SS cells, partially explaining different sensitivities of HIV-1 and SIV replication to CsA treatment. In contrast, CsA treatment inhibited SIV replication in macaque T cells; CsA treatment of either virus producer or target cells resulted in suppression of SIV replication. SIV infection was enhanced by CypA overexpression in macaque target cells. Conclusions CsA treatment enhanced SIV replication in human T cells but abrogated SIV replication in macaque T cells, implying a host cell species-specific effect of CsA on SIV replication. Further analyses indicated a positive effect of CypA on SIV infection into macaque but not into human T cells. These results suggest possible contribution of CypA to the determination of SIV tropism.

Background Viral replication is modulated by host cell factors, with the species specificity of these factors affecting viral tropism. Some of these host factors can restrict viral replication. The anti-viral systems mediated by such host restriction factors, termed intrinsic immunity, play an important role in determining species-specific barriers against viral infection. For instance, Fv-1 in mice is known to restrict replication of a murine leukemia virus [1][2][3]; and tripartite interaction motif 5α (TRIM5α) recently has been found to be responsible for restricting human immunodeficiency virus type 1 (HIV-1), but not simian immunodeficiency virus (SIV) infection in old world monkey (OWM) cells [4][5][6][7][8][9]. Restriction of retroviral replication by these host cell factors takes place after viral entry, but before the integration step; and the viral determinants for this type of restriction have been mapped to the capsid (CA) protein [1,3,[10][11][12]. Understanding how host cell factors affect viral replication, positively or negatively, would contribute to elucidating the molecular mechanism that determines viral tropism.
In contrast, the effect of CypA on SIV replication has not been well-documented, although a possible interaction between CypA and SIV CA has been indicated [17,27,28]. A recent study has shown a suppressive effect of CypA on replication of vif-deleted SIV in human Jurkat cells, which was counteracted by SIV Vif inhibiting CypA incorporation during virus assembly [27]. This Vif function can be distinguished from its anti-human APO-BEC3G (apolipoprotein B mRNA-editing enzyme-catalytic subunit 3G) function.
In the present study, we have investigated the effect of CsA on wild-type SIV replication in human or macaque T cells. SIV replication in human T cells was not inhibited but rather enhanced by treatment with CsA, which inhibits HIV-1 replication. In contrast, CsA treatment abrogated SIV replication in macaque T cells, indicating a species-specific effect of CsA on SIV replication. CypA knock-down or overexpression suggested that CypA affects SIV replication differently in human and macaque T cells and suggested possible contribution of CypA to the determination of SIV tropism.

Effect of CsA treatment on SIV replication in human T cells
We investigated the effect of CsA treatment on wild-type SIV replication in human CEM-SS T cells. Replication of wild-type SIVagm, SIVmac, and HIV-1 in the presence of CsA was compared with that in the absence of CsA (Figure 1A). Consistent with previous reports, CsA treatment inhibited the packaging of CypA into HIV-1 particles ( Figure 1B) and impaired HIV-1 replication in CEM-SS cells ( Figure 1A). CypA was incorporated into SIVagm and SIVmac progeny virions, although not efficiently, and CsA treatment further abrogated this low level of CypA incorporation ( Figure 1B) without the reduction of endogenous CypA (data not shown). Interestingly, however, CsA treatment did not inhibit but rather enhanced SIV replication in CEM-SS cells ( Figure 1A). This CsAmediated enhancement of SIV replication was also observed in human A3.01 T cells ( Figure 1C).

Effect of CsA treatment of target human T cells on SIV infection
Recent studies have indicated that CypA in target cells is crucial for an early HIV-1 replication step in human cells [15,19,25,26,29]. We therefore studied the effect of CsA treatment of target cells on SIV infection. Viruses were produced from CsA-untreated or CsA-treated CEM-SS cells and used to infect CsA-untreated or CsA-treated target human LuSIV cells. Cell lysates were prepared from the target LuSIV cells 24 h post-infection, and luciferase activity was measured to assess the efficiency of SIV infection (Figure 2A). Similar to the results in Jurkat cells [27], CsA treatment of either the producer cells or the target cells resulted in suppression of HIV-1 infection confirming the importance of CypA both in producer and target human cells for efficient HIV-1 replication [16,30]. In contrast, SIVagm infection was not inhibited but rather enhanced by CsA treatment of target cells although it was decreased by CsA treatment of producer cells. Similarly, SIVmac infection was not inhibited by CsA treatment of target cells but decreased by CsA treatment of producer cells.
Infection efficiency was also determined by measuring the amounts of viral cDNA synthesized in target CEM-SS cells after viral infection by quantitative PCR ( Figure 2B). As a negative control, cells were also infected with heatinactivated viruses. Consistent with the above results (Figure 2A), the amounts of viral cDNA synthesized after HIV-1 infection were reduced by CsA treatment of target cells ( Figure 2B). In contrast, CsA-mediated effect on viral cDNA synthesis after SIV infection into CEM-SS cells ( Figure 2B) was consistent with the results shown in Figure 2A. CsA treatment did not inhibit SIV infection in another human T cells, A3.01, either (data not shown). Thus, CsA treatment of human producer cells reduced infectivity of progeny SIVs, whereas CsA treatment of human target cells did not inhibit but rather enhanced SIV infection.

Effect of CypA knock-down on SIV replication in human T cells
To examine the effect of CypA on SIV replication in CEM-SS cells, we established CypA knocked-down (CypA-KD) CEM-SS cell lines. CypA expression in CEM-SS CypA-KD cells was stably suppressed by CypA-specific shRNA ( Figure 3). We confirmed that both cell proliferation and cell surface levels of CD4 and CXCR4 that are required for HIV-1 entry showed no difference between parental CEM-SS and CypA-KD CEM-SS cells (data not shown). As previously reported [16], CypA knock-down reduced viral cDNA synthesis after HIV-1 infection ( Figure 4A) and inhibited HIV-1 replication ( Figure 4B). In CypA-KD CEM-SS cells, CsA treatment showed little effect on viral cDNA synthesis after HIV-1 infection, indicating that CypA inhibition was largely involved in CsA-mediated reduction of HIV-1 infection. In contrast, CypA knock-down did not reduce viral cDNA synthesis after SIV infection ( Figure  4A) or SIV replication ( Figure 4B). These results indicate that target cell CypA is essential for HIV-1 infection but not required for SIV infection into human T cells. CsA treatment, however, enhanced SIV replication even in CypA-KD cells ( Figure 4A and 4B, SIVagm), suggesting the possibility that CsA neutralizes the  residual CypA population which remains in the CypA-KD cells as shown by Western blotting in Figure 3, or the possible involvement of another host factor in this CsA-mediated enhancement of SIV replication. Similar to the results obtained by replication in CypA-KD CEM-SS cells, the effect of CypA knock-down on SIV infection or SIV replication was also observed in A3.01 cells (data not shown).
We also established a cyclophilin B (CypB), another PPIase, knocked-down (CypB-KD) CEM-SS cell line ( Figure 5A). Overall SIV replication was enhanced in CypB-KD cells ( Figure 5B, SIVagm). As predicted, HIV-1 replication was not affected by the CypB knockdown ( Figure 5B, HIV-1). These results suggest that the CsA-induced enhancement of SIV replication in human T cells observed in Figure 1A is mediated largely if not exclusively by an inhibition of CypB.

Effect of CsA treatment on SIV replication in macaque T cells
We next examined the effect of CsA treatment on SIV replication in macaque cells using three macaque T-cell lines: cynomolgus macaque-derived HSC-F, rhesus macaque-derived HSR-5.4, and pig-tailed macaque- derived Mn-3942 ( Figure 6). Both SIVmac239 and SIVagm replicated well in all three cell lines, with the most efficient replication in HSC-F cells. However, in contrast to the results above for SIV replication in human T cells, CsA treatment inhibited SIVagm replication in all three macaque T cell lines. This inhibitory effect of CsA on SIV replication was also observed in rhesus macaque PBMCs ( Figure 6). The CsA effect on HSC-F cell proliferation was marginal and not considered as responsible for this inhibition of SIV replication. Indeed, treatment of these macaque cells with lower concentration (0.5 μM) of CsA had no effect on cell proliferation but resulted in inhibition of SIV replication (data not shown). CypA was incorporated into SIV virions in the absence of CsA, but its incorporation was inhibited by CsA treatment of HSC-F cells ( Figure 7A) without the reduction of endogenous CypA (data not shown). HIV-1 replication was undetectable in these macaque T-cell lines even in the presence of CsA (data not shown), although the possibility of enhancement of HIV-1 infection by CsA treatment in OWM cell lines has been indicated previously [28,31,32]. Thus, SIV   levels in target cells infected with SIV from CsAuntreated or CsA-treated producer HSC-F cells showed that SIV from CsA-treated producer HSC-F cells had lower amount than that from CsA-untreated producer HSC-F cells. In contrast to the results in human CEM-SS cells (Figure 2B), CsA treatment of target HSC-F cells did not increase viral cDNA synthesis after SIV infection, but rather resulted in a reduction in synthesis. Thus, CsA treatment of macaque T cells has an inhibitory effect on a post-entry step of SIV replication.

Effect of exogenous CypA on SIV infection in human and macaque T cells
Our attempts to knock down CypA in macaque HSC-F cells were unsuccessful. We therefore examined and compared the effect of CypA overexpression in target CEM-SS and HSC-F cells on SIV infection. We first examined the effect of exogenously expressed CypA in human target cells on SIV replication. CEM-SS or CypA-KD cells were transfected (nucleofected) with plasmids expressing HA-control or CypA-HA, respectively. Transfected cells were enriched by magnetic beads sorting as described in Methods. More than 97% of sorted cells were shown to express the marker protein (H-2K k ) without a reduction in cell-surface CD4 levels (data not shown).
We found that exogenous CypA increased HIV-1 cDNA synthesis in CsA-treated CEM-SS cells ( Figure  8A), confirming the positive effect of target cell CypA on the early phase of HIV-1 replication. However, SIV infection was not affected by exogenous CypA expression even in CsA-treated CEM-SS cells ( Figure 8A). In contrast, CypA overexpression in HSC-F target cells did not reduce but rather enhanced viral cDNA synthesis after SIV infection ( Figure 8B). These results suggest that target cell CypA essential for HIV-1 infection is not largely involved in SIV infection in human T cells but has a positive effect on SIV replication in macaque T cells.

Effect of CsA treatment of target cells on human cell-or macaque cell-derived SIV infection
We then investigated how the infectivity of SIV from producer macaque cells is affected by CsA treatment of target human cells and how the infectivity of SIV from producer human cells is affected by CsA treatment of target macaque cells. We first measured viral cDNA levels in target human CEM-SS cells after infection with SIV from producer macaque HSC-F cells ( Figure 9A). Similar to the results obtained by infection of CEM-SS cells with CEM-SS-derived SIV (Figure 2), CsA treatment of target CEM-SS cells enhanced infection by HSC-F-derived SIV. These data indicate that even macaque cell-derived SIV infection is enhanced by CsA treatment of target human cells. Finally, we examined the effect of CsA on the infectivity of CEM-SS-derived SIV in macaque HSC-F cells ( Figure 9B). The results were similar to those for infection of HSC-F cells by HSC-F-derived SIV ( Figure 7B), i.e. CsA treatment of either producer or target cells diminished infectivity of CEM-SS-derived SIV in HSC-F cells, although previous reports showed that the role of virion-associated CypA is minor compared to that of target cell CypA [15,19,25,26,29].

Discussion
The present study showed that CsA treatment, which inhibits HIV-1 replication, does not inhibit but rather enhances SIV (SIVmac239 and SIVagm) replication in human T cells. In SIV infection of both CEM-SS and A3.01 cells, CsA treatment resulted in production of viruses with lower infectivity but enhanced an early step of replication. The present study indicates that CsA treatment exerts different effects on the early phase of HIV-1 and SIV replication in human T cells. Indeed, previous study also showed the positive effect of CsA on HIV-1 and SIV vectors infection in primary mouse cells, suggesting different effects of CsA on the early phase of HIV-1 between human and mouse cells [33]. Additionally, we found that expression level of CCR5 on cell surface was lower than that of CXCR4 in CEM-SS cells (data not shown). This may be due to different kinetics between CCR5-tropic SIV and CXCR4-tropic HIV-1 in Figure 1A.
Knock down of CypA, an essential host factor for HIV-1 replication, did not reduce SIV replication in human T cells. CypA knock-down in target human cells, which inhibited HIV-1 infection, did not reduce viral cDNA synthesis after SIV infection, indicating that target human cell CypA is essential for HIV-1 but not for SIV infection. These results imply different effects of CypA on HIV-1 and SIV replication in human T cells.
Our results revealed that CypA can be incorporated into SIV virions although not very efficiently. This is consistent with previous results [27,34]. A previous study implied inhibition of viral replication by overincorporation of CypA in vif-deleted SIV [27], whereas the present study suggests that low levels of CypA proteins incorporated into wild-type SIV do not inhibit but may rather contribute to SIV replication in human T cells. Two amino acids in HIV-1 Gag, Gly221 and Pro222, were found to be important for the binding of CypA [21]. Although two amino acids are in fact conserved in our SIVagm isolate, the same region is not conserved in SIVmac239, which also was found to encapsidate CypA ( Figure 1B). Therefore, it seems that CypA is incorporated into SIV virions through a mechanism that is distinct from HIV-1. Target human cell CypA was shown to be essential for HIV-1 but not for SIV infection. This may partially explain different sensitivities of HIV-1 and SIV replication to CsA treatment, although other unknown host factors could also be involved in the CsA-mediated enhancement of SIV replication in human T cells. Indeed, CsA treatment enhanced SIV replication even in CypA-KD cells ( Figure 4B). We also attempted to examine the effect of cyclophilin B (CypB), another CsA-sensitive PPIase, on SIV replication by knock-down of CypB ( Figure 5). CypB knock-down showed no significant effect on HIV-1 replication but enhanced SIV replication in human T cells, suggesting a possible involvement of CypB in SIV replication. Overall SIV replication was enhanced in CypB-KD cells ( Figure 5B, SIVagm) and CsA treatment resulted in only slight enhancement of SIV replication (data not shown). As predicted, HIV-1 replication was not affected by the CypB knock-down ( Figure 5B, HIV-1) but was inhibited by CsA treatment (data not shown). Thus, enhanced replication of SIV in CsA-treated human cells as shown in Figure 1A might be best explained by the neutralization of CypB's inhibitory effect. Previous reports showed that CypB can bind to both HIV-1 and SIV CA [14,35], but CypB incorporation into HIV-1/SIV virions was undetectable in the present study (data not shown). Regarding the cellular localization of CypB, association of CypB with the endoplasmic reticulum (ER) via an ER signal sequence has been reported [36]. In other reports [37][38][39], CypB has been shown to be associated with heparan sulfate proteoglycans (HSPG) on the cell surface. Overall, it remains unclear how CypB affects SIV infection directly or indirectly, because most CypB is believed to express in the endoplasmic reticulum but not in cytoplasm [40].
In contrast, CsA treatment inhibited SIV (SIVmac239 and SIVagm) replication in macaque T cells; CsA treatment of either virus producer or target cells resulted in suppression of SIV replication. This CsA effect on SIV replication in macaque T cells is similar to that on HIV-1 replication in human T cells. Although our attempts to knock down CypA in macaque HSC-F cells were unsuccessful, we have obtained CypA knocked-down rhesus macaque kidney cell line: LLC-MK2 cells (Additional File 1). Analysis using these cells revealed that CypA is inhibitory for SIV replication in macaque cells and indicated that CypA dysfunction is likely to be largely involved in CsA-mediated reduction of SIV replication in macaque cells (Additional File 1). Taken together, the current study reveals that CsA treatment inhibited SIV replication in macaque T cells but enhanced SIV replication in human T cells, indicating a host species-specific effect of CsA on SIV replication. These data suggest that the effect of CsA on SIV infection seems to be general properties of SIV in human and macaque T cells, and cyclophilins may contribute to host-range control of certain lentiviruses.
Sequence analyses of CypA cDNA from human CEM-SS cells and macaque HSC-F, HSR-5.4, and Mn-3942 cells showed no difference in deduced amino acid sequences between human and macaque CypA (data not shown). Therefore, there may be a possible differential posttranslational modification of CypA such as acetylation [41,42] between human and macaque cells or there may be an additional host cell factor involved in the contribution of CypA to the determination of HIV-1 and SIV tropism by possibly affecting expression, localization, or function of CypA or viral capsid proteins.
CypA overexpression increased SIV infection in macaque T cells, suggesting that CypA may have different effects on SIV replication in human and macaque T cells. TRIM5α is known to restrict HIV-1 infection in macaque T cells, but this restriction has been shown to be relieved by CsA-mediated or small interfering RNA-mediated inhibition of CypA function, indicating involvement of CypA in TRIM5α-mediated restriction of HIV-1 infection in macaque cells [28,31,32]. Thus, CypA promotes HIV-1 infection in human cells but shows an inhibitory effect on HIV-1 infection in macaque cells. We attempted to examine the identification of macaque TRIM5α by PCR from macaque genomic DNA as reported previously [43] to investigate the possible effect of macaque TRIM5α on CsA-mediated reduction of SIV infection in macaque T cells, but macaque TRIM5α was identified in HSC-F cells but not in HSR-5.4 cells, suggesting that the effect of CsA on SIV replication in macaque T cells obtained from the present study ( Figure 6) may be independent of macaque TRIM5α restriction (Additional File 2). In owl monkeys, which belong to new world monkeys, previous reports revealed the existence of a TRIM-CypA fusion protein (TRIMCyp) restricting HIV-1 infection, which was relieved by CsA treatment [44,45]. Recently, TRIMCyp has been found also in OWM, although OWM TRIMCyp did not restrict HIV-1 or SIVmac replication [43,46,47]. We also attempted to examine the identification of OWM TRIMCyp by PCR using primers on either side of the CypA insertion as previously reported [43] and the possibility of TRIMCyp expression by immunoblotting with anti-CypA antibody, but TRIMCyp expression was not detected in any of the macaque T-cell lines used in the present study (data not shown) although the CypA insertion was identified in the genome sequence of macaque T cells (Additional File 2).
Taken together, the present study reveals a speciesspecific effect of CsA on SIV replication. Our results suggest a contribution of CypA to efficient SIV replication in macaque cells. In contrast, analysis in human cells indicated that target cell CypA have no positive effect on SIV infection; rather, CypB was considered inhibitory for SIV replication. These results suggest possible involvement of cyclophilins in the determination of SIV tropism.

Conclusions
The present study revealed that CsA treatment enhances SIV replication in human T cells but abrogates SIV replication in macaque T cells, indicating a host cell speciesspecific effect of CsA on SIV replication. CypA knockdown or overexpression indicated a positive effect of CypA on SIV infection into macaque but not into human T cells. These results suggest possible contribution of CypA to the determination of SIV tropism.

Immunoblot analysis
Immunoblot analysis of cell lysates and viral pellets was performed as described previously [53]. Briefly, virus supernatants containing equal RT levels were concentrated by centrifugation through 20% sucrose. Pelleted viruses were analyzed by immunoblotting using anti-CypA or anti-CA antibodies. A polyclonal anti-SIVagm CA antibody provided by Vanessa Hirsch [54], plasma from a SIV-mac239-infected rhesus macaque, and a monoclonal anti-HIV-1 p24 antibody (Abcam Inc, Cambridge, MA) were used to detect SIVagm, SIVmac, and HIV-1 CA, respectively. We used polyclonal anti-CypA antibody (BIOMOL Research Laboratories Inc., Plymouth Meeting, PA) and monoclonal anti-α-tubulin antibody (Sigma-Aldrich). A representative result from four independent experiments is shown in each figure.

Infectivity analysis
LuSIV cells which are derived from CEMx174 cells and contain a luciferase indicator gene under the control of the SIVmac239 LTR were maintained in RPMI 1640 medium containing 10% FBS and hygromycin B (300 μg/ml) [55]. LuSIV cells were cultured for 24 h after viral infection and lysed in 1 × reporter lysis buffer (Promega Corp., Madison, WI). To determine the luciferase activity, lysates were mixed with luciferase substrate (Promega Corp.) and light emission was measured in a luminometer (GloMax™ 96 Microplate Luminometer; Promega Corp.). Two sets of viruses were produced, and each was subjected to two sets of infection experiments. Thus, mean values of viral cDNA levels from four independent experiments are shown.

Measurement of viral cDNA levels after viral entry
Viruses were prepared from virus-infected CEM-SS or HSC-F cells. For infection, 5 × 10 5 target cells were incubated with a SIVagm containing 500 ng of p27, SIVmac containing 500 ng of p27, or HIV-1 containing 10 ng of p24 for 24 hr, and then total cellular DNA was extracted using a DNeasy Tissue Kit (QIAGEN Inc., Valencia, CA).

Establishment of CypA and CypB knocked-down human T cell lines
Both human CEM-SS and A3.01 T cells were transduced with HIV-1-based vectors that confer puromycin resistance and express short hairpin RNAs (shRNA) targeting human CypA and CypB (Sigma-Aldrich). CypA and CypB knocked-down cell lines were obtained after selection with 1 μg/ml puromycin. Representative results obtained from experiments using three CypA and CypB knocked-down human CEM-SS cell lines are shown.

Exogenous expression of CypA in human and macaque T cells
CypA cDNAs amplified by PCR were inserted into the pMACS K k .HA(C) vector (Miltenyi Biotec GmbH, Bergisch Gladbach, Germany) to obtain pMACS K k .HA (C)-CypA expressing HA-tagged CypA (CypA-HA) together with a cytoplasmic domain-truncated mouse major histocompatibility complex class I H-2K k protein, respectively. Human CEM-SS or macaque T cells (4 × 10 6 cells) suspended in 100 μl of nucleofection V solution with 2 μg of pMACS K k .HA(c)-CypA (CypA-HA) or pMACS K k .HA(c) (HA-control) plasmid vector DNAs were subjected to transfection (nucleofection) using a Nucleofector device. The nucleofection parameter was D-023 for CEM-SS cells and U-029 for HSC-F cells. After 1 day of culture, transfected cells were labeled with anti-H-2K k microbeads (MACSelect K k MicroBeads) and enriched by separation using the MACS Separator (Miltenyi Biotec GmbH). The enrichment rate of transfected cells was determined by detection of H-2K k using MACSelect Control FITC Antibody (Miltenyi Biotec GmbH). Additional file 2: Identification of a TRIM5a or a TRIMCyp in macaque T cells. Total DNA from macaque HSC-F and HSR-5.4 T cells was harvested. PCR primers on either side of the CypA insertion were used to detect both a TRIM5α and a TRIMCyp in macaque T cells as described [43]. H 2 O denotes water control.