The Betaretrovirus Mason-Pfizer Monkey Virus Selectively Excludes Simian APOBEC3G from Virion Particles

Construction of molecular clones.The HIV-1 proviral expression plasmids pNL-Luc-HXBΔVif and pNL4-3ΔVifΔEnv have been described previously (4, 41), as have the MPMV expression plasmids pMTΔE, pTMO, and pMPMV (7, 17, 36); the HIV-1 Gag-green fluorescent protein (GFP) fusion protein expression plasmid pCR/HIVGag-GFP (29); and the MLV-based expression plasmids pNCS and pFB-Luc (13). pCRV1, a plasmid containing cis and trans regulatory elements that allow the efficient expression of Rev-dependent mRNAs, has been previously described. A DNA fragment carrying the full-length MPMV gag gene was amplified from the proviral clone pSARM4 (36) by using the primers 5′-TTCCCTGAATTCATGGGGCAAGAATTAAGCCAG-3′ and 5′-TTCCCTGCGGCCGCATACTGTGTGGGAGGTGGAAC-3′ and introduced as an EcoRI-NotI fragment into pCRV1. Thereafter, GFP was inserted into the 3′ NotI site of pCRV1/MPMVGag, generating pCRV1/MPMVGag-GFP.

pcDNA3-based plasmids expressing carboxy-terminally influenza hemagglutinin (HA) epitope-tagged forms of hA3G, hA3F, and mA3 have been described previously (4, 41). The mA3 expression plasmid used produces exclusively the shorter, eight-exon form of this protein. A plasmid containing a full-length rA3G cDNA (26) was used as a template to PCR amplify a complete rA3G cDNA, with Asp718 and EcoRI restriction sites engineered at the 5′ and 3′ ends of the fragment, respectively. These sites were then digested, and the resultant DNA fragment was inserted into the equivalent sites in pcDNA3-HA to generate prA3G-HA.

Cell culture and analysis.293T and HeLa cells were cultured in Dulbecco's modified Eagle's medium containing 10% fetal calf serum and were transfected using calcium phosphate or Fugene (Roche). The TZM-bl indicator cell line has been described previously (30) and was maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum and antibiotics.

HIV-1 and MLV infectivity assays were performed as described previously (4, 13, 41). MPMV infectivity assays were performed as follows. 293T cells were transfected with 1 μg of pMTΔE, 1 μg pTMO, and a total of 1 μg of APOBEC3 expression plasmid and/or empty vector filler by using a 35-mm culture dish. Forty-eight hours posttransfection, virus-containing supernatant media were filtered and used to infect TZM-bl cells. A further 48 h was allowed to pass before the TZM-bl cells were lysed and induced luciferase levels determined.

Packaging of APOBEC3 proteins into HIV-1 virions was analyzed as previously described (4, 41). Briefly, 293T cells were transfected with pNL4-3ΔVifΔEnv and an APOBEC3 expression plasmid. Forty-four hours later, the virus-containing supernatant media were collected, filtered, and layered onto a 20% sucrose cushion. Virions were collected by centrifugation at 35,000 rpm for 1.5 h at 4°C in a Beckman SW41 rotor. Pellets were lysed and analyzed by Western blotting. MPMV packaging assays were performed by transfecting 1.5 μg of the MPMV proviral plasmid pMPMV, along with either a mA3-HA or rA3G-HA expression plasmid and filler DNA to a total of 0.5 μg, into each well of a six-well plate. Forty-eight hours posttransfection, the supernatant media were harvested, filtered, pooled, and layered onto a 20% sucrose cushion and subjected to ultracentrifugation as described above for HIV-1. The resultant viral pellets were lysed and analyzed by Western blotting.

Western blot analysis.Cell lysates, virion lysates, and immunoprecipitates were subjected to gel electrophoresis and then transferred to a nitrocellulose membrane. Membranes were probed with a mouse monoclonal antibody specific for the HIV-1 capsid protein (9), a mouse monoclonal antibody specific for the HA epitope tag (Covance), a mouse monoclonal antibody specific for GFP (Clontech), or a rabbit polyclonal antibody specific for the MPMV Gag protein (36). Reactive proteins were detected with Lumi-Light Western blotting substrate (Roche) as previously described (4, 41).

Protein coimmunoprecipitation assays.To detect the binding of APOBEC3 proteins to HIV-1 or MPMV Gag, 293T cells were cotransfected with 1.5 μg of a plasmid expressing a Gag-GFP fusion protein, together with 1.5 μg of a plasmid expressing an HA-tagged APOBEC3 protein or pcDNA3 as a control. Forty-eight hours posttransfection, cells were harvested and lysed in 250 μl of binding buffer (50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 5 mM EDTA, 5% glycerol, 1% Triton X-100). Fifty microliters of the lysate was retained for analysis of input protein expression levels. The remaining lysate was incubated with rabbit polyclonal anti-HIV or anti-MPMV Gag antisera and protein A agarose beads (Invitrogen) for 2 h at 4°C. The beads were then collected and washed extensively with binding buffer and bound proteins eluted by heating in the presence of sodium dodecyl sulfate. Western analyses were then performed as described above.

MPMV is resistant to rA3G but sensitive to heterologous APOBEC3 proteins.As noted above, one would predict that a given mammalian retrovirus would be resistant to APOBEC3 proteins expressed in the relevant target tissues in its normal host species but potentially highly sensitive to APOBEC3 proteins expressed by nontarget species. To address whether this is true for MPMV, we asked whether MPMV infectivity would be inhibited by rA3G, the most widely expressed APOBEC3 protein in its normal host species, the rhesus macaque (26); by hA3G or hA3F, two human APOBEC3 proteins that can effectively inhibit the replication of Vif-deficient but not wild-type HIV-1 variants (34, 41); and finally, by mA3, an APOBEC3 protein expressed in a wide range of mouse tissues (13, 26). The assay used involves transfection of 293T cells with pMTΔE, which encodes a previously described derivative of the MPMV provirus in which the viral env gene has been replaced with the HIV-1 tat gene (36). The pMTΔE plasmid therefore expresses the MPMV Gag and Pol proteins as well as HIV-1 Tat. 293T cells were also cotransfected with the expression plasmid pTMO, which encodes the MPMV env gene (7), and finally a pcDNA3-based plasmid encoding carboxy-terminally HA-tagged versions of full-length rA3G, hA3G, hA3F, or mA3. This tag has been previously shown to have no detectable effect on the biological activity of APOBEC3 proteins (26, 41). Of note, the mA3 expression plasmid used produces exclusively the shorter, eight-exon form of this protein.

At 48 h after transfection, the culture media were harvested and used to infect TZM-bl indicator cells (30). TZM-bl cells are a derivative of HeLa cells that contain an integrated indicator cassette consisting of the HIV-1 long terminal repeat promoter linked to the firefly luciferase (F-luc) gene. Under normal culture conditions, this long terminal repeat is essentially quiescent. However, expression of the HIV-1 Tat protein results in a dramatic activation of F-luc expression. Because the MPMV-derived MTΔE indicator virus expresses HIV-1 Tat, the effect of APOBEC3 proteins on MPMV infectivity can be readily determined by measuring the induced level of F-luc expression after infection of TZM-bl cells with supernatants containing MTΔE virus produced in their absence or presence. By dilution of an MTΔE virus stock, we observed that the level of F-luc activity induced is linearly related to the level of infectious virus added to the TZM-bl cell culture (data not shown).

As shown in Fig. 1, we observed little or no inhibition of MPMV infectivity when rA3G-HA was coexpressed in the virion-producing cells but a profound inhibition when mA3 was present. The hA3G-HA and hA3F-HA proteins displayed intermediate phenotypes. The result did not reflect a substantially higher level of expression for mA3-HA than for rA3G-HA, as Western analysis of the levels of intracellular expression for mA3-HA and rA3G-HA in virus producer cells showed that rA3G-HA was actually expressed at a higher level than mA3-HA, even though it was a less effective inhibitor of MPMV infectivity (Fig. 1).

• Open in new tab
• Download powerpoint

FIG. 1.

MPMV infectivity is inhibited by heterologous APOBEC3 proteins but not by rA3G. 293T cells were transfected with the MPMV proviral indicator plasmid pMTΔE (1 μg), the MPMV env expression plasmid pTMO (1 μg), 100 ng of the indicated HA-tagged APOBEC3 expression plasmid, or the pcDNA3 parental vector and 900 ng of pcDNA3. At 48 h posttransfection, supernatant media were harvested, filtered, and used to infect the TZM-bl indicator cell line. The virus producer cells were also harvested at this time and used for determination of APOBEC3 protein expression by Western analysis, using an HA-specific antiserum. Forty-eight hours after infection, the TZM-bl cells were lysed and induced F-luc expression levels quantitated. In parallel, we determined the effect of these APOBEC3 proteins on the infectivity of the MLV-based F-luc expression vector pFB-Luc, as previously described (13). The levels of MPMV and MLV infectivity seen in the absence of any APOBEC3 protein, as determined by the levels of F-luc activity observed, were set at 100. Averages for three independent experiments with standard deviations are indicated.

In parallel, we examined the abilities of these APOBEC3 proteins to inhibit the infectivity of an MLV-based retroviral vector, as previously described (13). As previously reported by ourselves and others (3, 13, 21), MLV infectivity is potently inhibited by hA3G, weakly inhibited by hA3F, and entirely resistant to mA3 (Fig. 1). MLV infectivity was also potently inhibited by coexpression of the rA3G protein (Fig. 1). These data therefore show that, while the sensitivities of different retroviral species to APOBEC3 proteins derived from different animal hosts differ widely, retroviruses are frequently resistant to the APOBEC3 proteins expressed by their normal host.

To extend and confirm the data shown in Fig. 1, we performed a dose response analysis of the sensitivities of MPMV and HIV-1ΔVif to a range of levels of expression for mA3-HA and rA3G-HA (Fig. 2). The HIV-1ΔVif virus is predicted (13, 26) to be highly sensitive to inhibition by both mA3-HA and rA3G-HA, and this was in fact observed (Fig. 2B). In contrast, MPMV infectivity was found to be far more sensitive to the heterologous mA3-HA protein than the cognate rA3G-HA protein (Fig. 2A). Nevertheless, MPMV was significantly more resistant than HIV-1ΔVif to low levels of mA3-HA, while high-level expression of rA3G-HA did result in some inhibition of MPMV infectivity (Fig. 2A). These results confirm the initial result shown in Fig. 1, i.e., that MPMV is more sensitive to inhibition by mA3-HA than to that by rA3G-HA, but suggest that MPMV, unlike HIV-1ΔVif, may nevertheless also be partially resistant to inhibition by mA3-HA. Of note, Western analysis of the virus producer cells again showed an approximately fourfold-higher level of expression for rA3G-HA than for mA3-HA in 293T cells transfected with the same level of the mA3-HA expression plasmid (Fig. 2C). Moreover, neither mA3-HA nor rA3G-HA was found to inhibit the expression of the MPMV p27 capsid protein in cotransfected cells (Fig. 2C).

• Open in new tab
• Download powerpoint

FIG. 2.

Differential sensitivities of MPMV to mA3 and rA3G. Results for a dose response experiment analyzing the sensitivities of MPMV or HIV-1ΔVif to a range of concentrations of rA3G and mA3 are shown. (A) This experiment was performed as described in the legend to Fig. 1, except that the levels of the indicated APOBEC3 expression vectors were modulated as indicated. The amount of transfected DNA was kept constant by varying the level of the pcDNA3 parental vector included in the transfection. (B) Similar to panel A, except that 293T cells were transfected with the pNL-Luc-HXBΔVif HIV-1 proviral indicator plasmid in place of pMTΔE and pTMO. (C) Western analysis of the levels of mA3-HA, rA3G-HA, and the MPMV Gag p27 capsid protein expressed in the virus producer cells used for panel A at the time that the supernatant virus was harvested. This analysis used a mouse monoclonal anti-HA antibody (upper panel) or a rabbit polyclonal antiserum specific for MPMV Gag (lower panel). Panels A and B show the averages for three independent experiments, with standard deviations indicated, while panel C shows the results for one representative experiment.

The rA3G protein is excluded from MPMV virions.Previously, we and others have reported that MLV, a gammaretrovirus, is resistant to inhibition by mA3, but not hA3G, because it selectively excludes the former from progeny virion particles (13, 21). To test whether the betaretrovirus MPMV uses an analogous strategy to avoid inhibition by rA3G, we isolated MPMV and HIV-1ΔVif virions produced in the presence or absence of mA3-HA or rA3G-HA and analyzed the levels of mA3 and rA3G present in both the virions themselves and the producer cells. In this experiment, the 293T cell cultures were transfected with a level of the pmA3-HA expression vector fourfold higher than that of the prA3G-HA expression vector, in order to express similar steady-state levels for each protein.

As shown in Fig. 3, HIV-1ΔVif virions packaged both mA3-HA and rA3G-HA with equal efficiencies, consistent with the equivalent sensitivities of HIV-1ΔVif virions to inhibition by both proteins (Fig. 2B). In contrast, MPMV virions clearly discriminated against the rA3G-HA protein compared to what was found for the mA3-HA protein (Fig. 3). As shown in the lower panel of Fig. 3, neither mA3-HA nor rA3G-HA had any effect on the production of MPMV or HIV-1 virions, as determined by Western blot analysis using antisera specific for either the MPMV or the HIV-1 Gag protein.

• Open in new tab
• Download powerpoint

FIG. 3.

MPMV virions package mA3 but exclude rA3G. 293T cells were transfected with pMPMV or pNL4-3ΔVifΔEnv (1.5 μg), together with either 500 ng of pmA3-HA, 125 ng of prA3G-HA and 375 ng of pcDNA3, or 500 ng of pcDNA3 (None). Forty-eight hours posttransfection, the supernatant media were collected and the producer cells lysed. After MPMV and HIV-1 virions were collected from the supernatant by centrifugation, both the producer cell lysate and the virion lysate were subjected to Western analysis using antisera specific for the HA epitope tag, MPMV Gag, or HIV-1 Gag, as indicated. The figure shows only the p27 MPMV and the p24 HIV-1 capsid proteins, as little unprocessed Gag protein was detected in either virion preparation (not shown). α HA, anti-HA antisera; α Gag, anti-Gag antisera.

The MPMV Gag protein binds mA3 but not rA3G.In the case of MLV, the poor incorporation of mA3 into MLV virions, compared to that of hA3G, was found to correlate with the fact that the MLV Gag polyprotein bound to the hA3G protein effectively in vitro while binding of MLV Gag to mA3 was barely detectable (13). If the mechanism used by MPMV to avoid restriction by rA3G is similar to the mechanism used by MLV to avoid restriction by mA3, then one would predict that MPMV Gag would bind rA3G ineffectively while MPMV Gag binding by mA3 might be readily detectable. To test this hypothesis, we transfected 293T cells with prA3G-HA, pmA3-HA, or pcDNA3 as a control. The cells were also cotransfected with pCRV1/MPMVGag-GFP, which encodes a full-length MPMV Gag protein fused to a carboxy-terminal GFP moiety, or pCR/HIVGag-GFP, which encodes full-length HIV-1 Gag fused to GFP. After 48 h, the transfected cells were lysed and subjected to immunoprecipitation using rabbit polyclonal antisera specific for MPMV Gag or HIV-1 Gag and the antibody complexes collected using protein A agarose. Bound proteins were then analyzed by Western analysis using an HA-specific monoclonal antibody while input proteins were detected by Western analysis of the transfected cell lysate using monoclonal antibodies specific for the HA or GFP tags (Fig. 4).

• Open in new tab
• Download powerpoint

FIG. 4.

MPMV Gag binds mA3, but not rA3G, in coexpressing cells. 293T cells were transfected with 1.5 μg of pCRV1/MPMVGag-GFP or pCRV1/HIVGag-GFP, together with either 500 ng of pmA3-HA and 1 μg of pcDNA3, 125 ng of phA3G-HA and 1.375 μg of pcDNA3, or 1.5 μg of pcDNA3 (None). At 48 h after transfection, the cells were lysed and Gag proteins collected by incubation with polyclonal rabbit antisera specific for MPMV Gag or HIV-1 Gag in the presence of protein A agarose beads. Bound proteins, and proteins present in the input cell lysate, were then detected by Western blot analysis using anti-HA (α HA) or anti-GFP (α GFP) antisera. Only full-length MPMV and HIV-1 Gag-GFP fusion proteins were detected.

Consistent with the observed packaging of both mA3-HA and rA3G-HA into HIV-1 virion particles (Fig. 3), we readily detected the coimmunoprecipitation of both mA3-HA and rA3G-HA with HIV-1 Gag. In contrast, while mA3-HA was efficiently coimmunoprecipitated with MPMV Gag, very little rA3G-HA was bound by MPMV Gag in coexpressing cells (Fig. 4). Together, these data argue that the poor incorporation of rA3G into MPMV virion particles (Fig. 3) is reflective of a weak interaction between rA3G and MPMV Gag (Fig. 4), while the efficient incorporation of mA3 into MPMV virions (Fig. 3) results from the efficient formation of MPMV Gag-mA3 complexes in coexpressing cells (Fig. 4). In contrast, both mA3 and rA3G bind HIV-1 Gag, and are incorporated into HIV-1ΔVif virions, with equal efficiency (Fig. 3 and 4).

MPMV Gag partially colocalizes with mA3, but not with rA3G, in vivo.MPMV, a betaretrovirus, differs from lentiviruses and gammaretroviruses in that virion assembly occurs inside the cytoplasm at a pericentriolar region (33). The subsequent movement of assembled MPMV virions to the cell surface is dependent on the MPMV Env protein, and expression of MPMV Gag in the absence of MPMV Env results in the accumulation of MPMV virus-like particles (VLPs) in the cytoplasm and particularly at sites adjacent to the nucleus. We therefore asked whether we could see recruitment of mA3, but perhaps not rA3G, to these assembling MPMV VLPs. To perform this experiment, we expressed a form of the MPMV Gag protein that contains a carboxy-terminal GFP moiety and that can therefore be readily detected by intrinsic fluorescence. As before, the rA3G and mA3 proteins that were expressed bore a carboxy-terminal HA epitope tag and were detected by indirect immunofluorescence using a mouse monoclonal antibody specific for the HA tag.

Expression of rA3G-HA and mA3-HA in the absence of the MPMV Gag protein revealed a subcellular localization very similar to that described previously for hA3G and hA3F (6, 37, 40) (Fig. 5D and H). That is, both rA3G and mA3 were restricted to the cytoplasm of expressing cells and were found to give an essentially diffuse localization, although some bright speckles that could result from the reported accumulation of APOBEC3 proteins in cytoplasmic processing bodies could be observed (40).

• Open in new tab
• Download powerpoint

FIG. 5.

The mA3 protein partially colocalizes with MPMV Gag in coexpressing cells. HeLa cells were transfected with plasmids expressing mA3-HA, rA3G-HA, and/or MPMVGagGFP, as indicated, or with empty vector (-). At 48 h after transfection, the cells were fixed and incubated with a mouse monoclonal anti-HA antibody followed by a tetramethylrhodamine B isothiocyanate-conjugated goat anti-mouse antiserum and analyzed using an Olympus IX71 fluorescence microscope. The MPMV Gag-GFP fusion protein was detected by intrinsic fluorescence.

Expression of MPMV Gag in the absence of an overexpressed APOBEC3 protein resulted in the detection, using the intrinsic fluorescence of the GFP moiety attached to the carboxy terminus of MPMV Gag, of concentrated regions of fluorescence in the cell cytoplasm. These appear similar to the localization pattern previously reported for MPMV Gag (33) and presumably constitute sites of MPMV VLP assembly (Fig. 5A). MPMV VLP assembly did not appear to be affected by expression of either mA3 or rA3G (Fig. 5E and I). However, expression of MPMV Gag did affect the subcellular localization of the APOBEC3 proteins. Specifically, the mA3-HA protein, which binds to MPMV Gag and is assembled into MPMV virions (Fig. 3 and 4), lost its diffuse cytoplasmic localization and was instead found to partially colocalize with the MPMV Gag protein (Fig. 5I and J). In contrast, the rA3G-HA protein, which binds MPMV Gag poorly and is not effectively incorporated into MPMV virions (Fig. 3 and 4), maintained a largely diffuse staining pattern in the presence of MPMV Gag and indeed appeared to be excluded from sites of MPMV virion assembly (Fig. 5E and F).