A number of previous studies reported the efficient Vif-dependent degradation of APOBEC3G by cellular proteasomes 36828. However, we and others noted only a moderate reduction of the cellular APOBEC3G levels in response to Vif expression 45. This is exemplified in figure 1 where APOBEC3G was expressed either in the presence or absence of Vif. Specifically, HeLa cells were transfected with pcDNA-APO3G together either with wild type pNL-A1 (Fig. 1A, lane 2) or its vif-defective variant, pNL-A1vif(-) (lane 3). Mock transfected cells were included as a control (lane 1). Cells were harvested 24 hr post-transfection and whole-cell lysates were subjected to immunoblot analysis as described in Methods using an APOBEC3G-specific polyclonal antibody (Fig. 1A, top panel) or a Vif-specific monoclonal antibody (Fig. 1A, middle panel). To control for loading errors, the filters were re-probed with an antibody to α-tubulin (Fig. 1A, bottom panel). Consistent with our previous results, expression of Vif from pNL-A1 only moderately reduced the steady-state levels of APOBEC3G in HeLa cells. Quantitation of the data confirmed that expression of APOBEC3G in the presence of Vif was reduced by only about 20% (Fig. 1B).

An earlier study investigating the coexpression of Vif and APOBEC3G by confocal microscopy concluded that APOBEC3G was virtually excluded from Vif-expressing COS7 cells 6. To verify this observation, we investigated the effects of Vif on APOBEC3G expression on a single cell level by performing a series of immunocytochemical analyses. For that purpose, HeLa cells were transfected with the Vif expression vector pNL-A1 together with pcDNA-APO3G for the expression of human APOBEC3G. Cells were grown on cover slips, fixed 24 hr later with cold methanol, and stained with antibodies to APOBEC3G (Fig. 2, panels A & D) and Vif (panels B & E). The results of this experiment show that APOBEC3G can be expressed in Vif-positive cells (white arrow heads) without a dramatic reduction in its expression level when compared to Vif-negative cells (yellow arrow heads). Furthermore, these data confirm that APOBEC3G is localized to the cytoplasm while Vif was observed in this experiment in some cells both in the cytoplasm and the nuclei of cells (red arrow heads). Finally, we also observed cells expressing Vif that were APOBEC3G-negative (blue arrow heads). Overlay of the Vif and APOBEC3G channels revealed a partial co-localization of Vif and APOBEC3G in the cytoplasm apparent by the yellow staining in panels C & F of figure 2. Interestingly, a significant number of cells in this experiment were single-positive expressing either APOBEC3G or Vif alone. The appearance of Vif-positive, APOBEC3G-negative cells could be explained by a Vif-dependent restriction of APOBEC3G expression as proposed by Marin et al 6. However, cells expressing Vif only were rare when compared to cells expressing APOBEC3G alone (data not shown). The preponderance of APOBEC3G single positive cells cannot be explained by a Vif-dependent restriction but more likely represents a technical, albeit reproducible, artifact.

In the experiment shown in figure 2, APOBEC3G was expressed under the control of a CMV promoter while Vif was expressed from the HIV-LTR promoter under the regulatory control of Tat and Rev. Because of the independent expression of Vif and APOBEC3G it cannot be ruled out that the large number of single-positive cells in that experiment – while statistically improbable – was caused by the selective transfection of cells with either the Vif or the APOBEC3G expression vector. To check this possibility, we expressed APOBEC3G from the HIV-1 long terminal repeat (LTR) promoter driven vector pHIV-APO3G 4. Because of its dependence on Tat, APOBEC3G expression from pHIV-APO3G is restricted to cells also expressing Vif. Indeed, transfection of pHIV-APO3G into cells in the absence of pNL-A1 or any other Tat expression vector did not reveal any APOBEC3G expression as judged by immunofluorescence analysis or immunoblotting attesting to the strict Tat-dependence of this APOBEC3G expression vector (data not shown).

In addition of measuring the context-dependent expression of APOBEC3G, we also wanted to determine the influence of the fixation procedure on the efficiency of Vif:APOBEC3G co-staining. It is well known that the choice of fixative can affect the ability of a given antibody to recognize a specific epitope on its target protein. Frequently, epitopes are masked because of the folding properties of a protein in vivo or because of pre-existing protein-protein interactions that may compete for antibody binding. To test this possibility we compared a formaldehyde fixation procedure employed previously 6 with the methanol fixation procedure employed in our own studies 22.

HeLa cells were transfected with pNL-A1 and pHIV-APO3G at a 1:1 molar ratio. Cells were fixed 24 hr later either with methanol (MeOH) as in figure 2 or with formaldehyde (FA) as described in Methods. Cells were stained with Vif- and APOBEC3G-specific antibodies as described in figure 2. The results of this experiment show that expression of APOBEC3G under the control of the HIV-1 LTR indeed increased the proportion of double-positive cells both in methanol-fixed samples (Fig. 3, panels A-C) and formaldehyde-fixed specimens (Fig. 3, panels D-F). This suggests that the high proportion of single-positive cells observed in figure 2 was not the result of a Vif-dependent restriction of APOBEC3G but was caused by the independent expression of APOBEC3G from a Tat-independent promoter. Again, in methanol-fixed samples APOBEC3G expression levels in Vif-positive cells (Fig. 3A, white arrow heads) were indistinguishable from those observed for neighboring Vif-negative cells (Fig. 3A, yellow arrow head). Interestingly, the APOBEC3G fluorescent intensity appeared to be reduced in Vif-positive formaldehyde-fixed specimens when compared to Vif-negative cells or cells expressing low levels of Vif (Fig. 3D; compare white and yellow arrow heads). Because the methanol-fixed samples did not show a Vif-dependent reduction in APOBEC3G signals in these experiments, we conclude that the reduction in APOBEC3G signals observed in formaldehyde-fixed samples is not the result of Vif-induced degradation of APOBEC3G but is a technical artifact.

For a more quantitative analysis and to determine possible effects that arise from the use of different antibodies, we extended the experiment shown in figure 3 to include three different antibodies for the identification of APOBEC3G. As before, HeLa cells were co-transfected with a 1:1 ratio of pNL-A1 and pHIV-APO3G plasmid DNAs. Cells were grown on cover slips and fixed 24 hr later either with formaldehyde (Fig. 4, panels A-C) or methanol (panels D-F) as described in figure 3. Cells were then stained with either a monoclonal antibody to the C-terminal Myc-epitope in APOBEC3G together with a polyclonal Vif antibody (Fig. 4, panels A & D), or a polyclonal Myc antibody together with a monoclonal Vif antiserum (Fig. 4, panels B & E). A third set of cells was stained with a polyclonal APOBEC3G-specific antiserum together with the monoclonal Vif antibody (Fig. 4, panels C & F). Representative fields are shown for each combination.

To quantify the results, multiple optical fields were analyzed (n = 5–10) with a total of at least 100 transfected cells for each parameter. As can be seen in figure 5, methanol-fixed samples showed a relatively modest variation among the three antibodies used. All three antibodies identified between 45% to 60% of the cells as double-positive for Vif and APOBEC3G. In contrast, formaldehyde-fixed samples exhibited a larger antibody-dependent variation. Staining with the 9E10 monoclonal antibody to the Myc-epitope in APOBEC3G yielded the lowest efficiency of staining and identified little more than 40% of the transfected cells as double-positive for Vif and APOBEC3G. In contrast, staining of APOBEC3G was more efficient with the polyclonal Myc antibody, which identified approximately 80% of the transfected cells as double-positive in formaldehyde-fixed samples. Finally, the polyclonal APOBEC3G-specific antibody was slightly less efficient for the staining of FA-fixed samples than methanol-fixed samples and identified about 40% of the formaldehyde-fixed samples as double-positive. Since all samples were derived from the same transfected culture, variations in the co-expression of Vif and APOBEC3G in the individual samples can only be explained by the differential sensitivity of the antibodies to the fixation procedure.

Under optimal conditions, wild type Vif and APOBEC3G were coexpressed in about 80% of transfected cells (see figure 5). Thus, 20% of the transfected cell population either was expressing Vif but not APOBEC3G or was single-positive for APOBEC3G. To investigate whether the presence of such single-positive cells is due to an activity of Vif or is a general characteristic of transiently transfected cells, we studied the effects of a biologically inactive Vif variant. For this purpose, we employed a Vif mutant carrying a deletion of residues 23–45 in Vif. We previously showed that this mutant is unable to rescue viral infectivity in APOBEC3G-expressing cells 4. Like wild type Vif, VifΔ23–43 was expressed in the context of the subviral expression vector pNL-A1. HeLa cells were cotransfected with pNL-A1/VifΔ23–43 and pHIV-APO3G, fixed with methanol and processed for confocal microscopy as described for figure 2. As shown in figure 6, coexpression of VifΔ23–43 and APOBEC3G yielded a significant number of double-positive cells (white arrow heads). However, as observed before with wild type Vif, we also identified cells that were Vif-positive but had significantly reduced APOBEC3G levels (Fig. 6, blue arrow heads) or cells that were APOBEC3G positive but did not express Vif (yellow arrow heads). As with wild type Vif, overlay of the Vif and APOBEC3G image channels suggested a partial colocalization of the two proteins. In cells, in which Vif had spontaneously collapsed into a perinuclear aggregate (green arrow head), APOBEC3G did not exhibit a similar change in subcellular distribution. This is in contrast to the Vif-induced reorganization of vimentin reported previously 22. Thus, VifΔ23–43 is either unable to interact with APOBEC3G or forms complexes that are unstable. These results also imply that the partial colocalization of APOBEC3G and Vif noted in this study may not reflect a true physical interaction of the two proteins.

The combined results from the experiments shown in figures 2,3,4,5,6, do not support the notion that Vif expression leads to the elimination of APOBEC3G from Vif-positive cells. It can be argued, however, that under the experimental conditions employed in our experiments, the Vif expression levels were insufficient or ineffective. To control for this possibility, we compared the infectivity of viruses produced in the presence of various ratios of Vif and APOBEC3G. To allow a direct comparison with the experiments shown in figures 2 to 6, Vif and APOBEC3G were expressed in trans from pNL-A1 and pHIV-APO3G respectively in the presence of a Vif-defective NL4-3 proviral vector. The ratios of Vif to APOBEC3G expression vector were 1:1, 2:1, and 5:1, respectively. Note that the Vif to APOBEC3G ratio in the experiments shown in figures 2 to 4 was 1:1 throughout. A Vif-negative sample was analyzed as negative control. Virus-containing supernatants were harvested 24 hr after transfection, normalized for equal reverse transcriptase activity and used for the infection of LuSIV indicator cells. Relative virus infectivity was determined by comparing the Tat-dependent expression of luciferase in the target cells (Fig. 7). Interestingly, viruses produced at the lowest Vif:APOBEC3G ratio were virtually as infectious as viruses produced in the presence of higher levels of Vif. In fact, increasing the Vif:APOBEC3G ratio to 2:1 or 5:1 did not significantly increase viral infectivity. Instead, at the 5:1 ratio viral infectivity was slightly reduced, presumably due to the inhibitory effect of Vif at high concentrations as reported previously 39. Taken together, our data suggest that the inability of Vif to prevent co-expression of APOBEC3G in transiently transfected HeLa cells is not caused by sub-optimal levels of Vif or a lack of Vif activity in our system.

The full-length molecular clone pNL4-3 33 was used for the production of wild type infectious virus. For transient expression of Vif, the subgenomic expression vector pNL-A1 34 was employed. This plasmid expresses all HIV-1 proteins except for gag and pol products. A vif-defective variant of pNL-A1, pNL-A1vif(-) was constructed by deletion of an NdeI/PflMI fragment 4. Plasmid pNL-A1/VifΔ23–43 expresses a Vif variant carrying a 21 amino acid deletion (residues 23 to 43) in its vif gene as reported elsewhere 4. This Vif variant is inactive and does not target APOBEC3G 4. Plasmids pcDNA-APO3G and pHIV-APO3G are vectors for the expression of human APOBEC3G under the control of the CMV immediate early promoter or the HIV promoter, respectively, and were constructed as described elsewhere 4.

Serum from an HIV-positive patient (APS) was used to detect HIV-1-specific capsid (CA) proteins. A monoclonal antibody to Vif (MAb #319) was used for all immunoblot analyses and some of the immunohistochemical analyses as indicated in the text and was obtained from Michael Malim through the NIH AIDS Research and Reference Reagent Program 2335363637 For all other immunocytochemical analyses our Vif-specific polyclonal antibody (Vif93) was employed. APOBEC3G, carrying a C-terminal Myc epitope tag was identified either with the Myc-specific 9E10 monoclonal antibody or a polyclonal antibodies to the Myc epitope tag (both antibodies were obtained from Sigma-Aldrich, St. Louis). Alternatively, APOBEC3G was identified using a polyclonal rabbit serum against recombinant APOBEC3G 4. Tubulin was identified using a monoclonal antibody to α-tubulin (Sigma-Aldrich, St. Louis).

HeLa cells were propagated in Dulbecco's modified Eagles medium (DMEM) containing 10% fetal bovine serum (FBS). LuSIV cells are derived from CEMx174 cells and contain a luciferase indicator gene under the control of the SIVmac239 LTR 38. These cells were obtained through the NIH AIDS Research and Reference Reagent Program and were maintained in complete RPMI 1640 medium supplemented with 10% FBS and hygromycin B (300 μg/ml).

For transfection of HeLa cells, cells were grown in 25 cm² flasks to about 80% confluency. Cells were transfected using LipofectAMINE PLUS™ (Invitrogen Corp, Carlsbad CA) following the manufacturer's recommendations. A total of 5–6 μg of plasmid DNA per 25 cm² flask was used. Cells were harvested 24 hr post-transfection. Transfection efficiency in our analyses was generally 30% to 40%.

Virus stocks were prepared by transfecting HeLa cells with appropriate plasmid DNAs. Virus-containing supernatants were harvested 24 hr after transfection. Cellular debris was removed by centrifugation (3 min, 3000 × g) and clarified supernatants were filtered (0.45 μm) to remove residual cellular contaminants. For determination of viral infectivity, unconcentrated filtered viral supernatants were used for the infection of indicator cells. For immunoblot analysis of viral proteins, virus particles (7 ml) were concentrated by ultracentrifugation through 4 ml of 20% sucrose in PBS as described before 4.

For immunoblot analysis of intracellular proteins, whole cell lysates were prepared as follows: Cells were washed once with PBS, suspended in PBS (400 μl/10⁷ cells), and mixed with an equal volume of sample buffer (4% sodium dodecyl sulfate, 125 mM Tris-HCl, pH 6.8, 10% 2-mercaptoethanol, 10% glycerol, and 0.002% bromphenol blue). Proteins were solubilized by boiling for 10 to 15 min at 95°C with occasional vortexing of the samples to shear chromosomal DNA. Residual insoluble material was removed by centrifugation (2 min, 15000 rpm in Eppendorf Minifuge). Viral proteins were obtained by boiling concentrated viral pellets in a 1:1 mixture of PBS and sample buffer. Cell lysates and viral extracts were subjected to SDS-PAGE; proteins were transferred to PVDF membranes and reacted with appropriate antibodies as described in the text. Membranes were then incubated with horseradish peroxidase-conjugated secondary antibodies (Amersham Biosciences, Piscataway NJ) and visualized by enhanced chemiluminescence (ECL, Amersham Biosciences).

To determine viral infectivity, virus stocks were normalized for equal reverse transcriptase activity and used to infect LuSIV cells (5 × 10⁵) in a 24-well plate total volume 1.2 to 1.4 ml. Cells were incubated for 24 hours at 37°C. Cells were then harvested and lysed in 150 μl of Promega 1x reporter lysis buffer (Promega Corp., Madison WI). To determine the luciferase activity in the lysates, 50 μl of each lysate were combined with luciferase substrate (Promega Corp., Madison WI) by automatic injection and light emission was measured for 10 seconds at room temperature in a luminometer (Optocomp II, MGM Instruments, Hamden CT).

For the analysis of transfected HeLa cells, cells were scraped off the flasks 3 hr after transfection and reseeded into 12 well plates containing 0.13 mm cover slips. Cells were grown for 15 to 24 hrs at 37°C in DMEM containing 10% FBS. Cells were then fixed at -20°C in precooled methanol (-20°C) for 10 min followed by two washes in PBS or fixed in FA buffer (5% formaldehyde + 2% sucrose in PBS) for 20 min at room temperature followed by two washes in PBS. Coverslips were stored in PBS at 4°C until use. FA-fixed samples were permeabilized for 30 min at room temperature in permeabilization buffer (1% Triton X-100, 10% sucrose in PBS) prior to incubation with antibodies. For antibody staining, cover slips were incubated in a humid chamber at 37°C for 30 min with primary antibodies at appropriate dilutions in 1% BSA in PBS. Cover slips were washed once in PBS (5 min, room temp) and incubated with Texas-Red- or Cy2-conjugated secondary antibodies (diluted in 1% BSA in PBS) for 30 min at 37°C in a humid chamber. Cover slips were then washed twice with PBS and mounted onto microscope slides with glycerol gelatin (Sigma-Aldrich, St. Louis) containing 0.1M N-propyl gallate (Sigma-Aldrich, St. Louis) to prevent photo bleaching and were stored at 4°C in the dark until analyzed by confocal microscopy.

For confocal microscopy, a Zeiss LSM410 inverted laser scanning microscope was employed. The microscope was equipped with a krypton/argon mixed-gas laser and was operated by the Microcosm Renaissance 410 (v2.3.4) software package. Images were acquired with a Plan-Apochromat 63x/1.4 oil immersion objective (Carl Zeiss, Thornwood). Additional optical magnification (up to 5-fold) was achieved using the zoom feature of the image acquisition software. For two-color analysis, objects were excited using 488/568 nm laser lines. Green and red emissions were recorded through appropriate filters (515–540 nm band pass filter for Cy2 and 590 nm long pass filter for Texas-Red) and stored in separate (red and green) image channels. At the same time, bright field images (Nomarski optics) were collected and stored in a third (blue) channel. Image quality was enhanced during data acquisition using the Renaissance 410 line average feature (8 or 16x).