Cytoplasmic APOBEC3G Restricts Incoming Vif-Positive Human Immunodeficiency Virus Type 1 and Increases Two-Long Terminal Repeat Circle Formation in Activated T-Helper-Subtype Cells^▿

Human subjects.

Peripheral blood samples were obtained with informed consent from healthy volunteers under a protocol approved by the Vanderbilt University Institutional Review Board.

Cells.

Peripheral blood mononuclear cells were isolated from blood using Ficoll-Hypaque (Amersham Biosciences) gradients. CD4⁺ cells were isolated by negative selection through magnetic separation using autoMacs (Miltenyi Biotec, Auburn, CA) or Robosep (StemCell Technologies, Vancouver, British Columbia, Canada). Naïve CD4⁺ T cells were stained with fluorescein isothiocyanate-labeled anti-CD45RO antibodies (anti-CD45RO-FITC) and phycoerythrin (PE)-labeled anti-CD25 antibodies (anti-CD25-PE) (BD Pharmingen, San Jose, CA) and then sorted (FACSAria; Becton Dickinson, San Jose, CA). Sorted naïve CD4⁺ T cells were differentiated to Th1 cells using anti-CD3 (OKT3; ATCC)-coated plates in RPMI 1640 medium supplemented with anti-CD28 antibodies (BD Biosciences Pharmingen), 0.5 μg/ml neutralizing anti-IL-4 antibody, and 30 ng/ml recombinant IL-12. For Th2 cell differentiation, sorted naïve cells were cultured on plates coated with anti-CD3 in medium supplemented with anti-CD28 antibodies (BD Biosciences Pharmingen), 2.5 μg/ml neutralizing anti-IFN-γ antibody, and 50 ng/ml recombinant IL-4. Cytokines and neutralizing antibodies were obtained from R&D Systems, Minneapolis, MN. In vitro-differentiated Th1 and Th2 cells were removed from the anti-CD3-coated plates after 48 h and were subsequently expanded for 10 days in the medium described above. Differentiation was confirmed by intracellular cytokine staining for IL-4-PE and IFN-γ-allophycocyanin (IFN-γ-APC) (BD Pharmingen), as well as surface staining for CXCR3-PE and CRTh2-APC (BD Pharmingen) (5, 12, 35). To increase APOBEC3G expression in cytokine-polarized Th2 cells, differentiating cultures were transduced with an APOBEC3G-expressing, HIV-derived lentiviral vector at the time of activation. The vector expressed APOBEC3G, and human serum albumin was used as a marker of transduction (37, 38). To reduce APOBEC3G expression in cytokine-polarized Th1 cells, fully differentiated Th1 cells were activated for 48 h with anti-CD3/CD28-coated beads (Invitrogen) in the presence of 10 μg anti-IFN-γ antibody (38) (R&D Systems).

Viruses and infection.

Vesicular stomatitis virus glycoprotein (VSV-G)-pseudotyped, green fluorescent protein (GFP)-expressing HIV-1 was produced by polyethylenimine transfection (13) of 293T cells with a NL4-3 recombinant with GFP replacing nef as well as phCMV-VSV-G. Prior to infection, viral supernatants were treated with 100 U DNase I (Bio-Rad) for 1 hour. In vitro-differentiated Th1 and Th2 cells (1 × 10⁶ cells of each type) were used for infections. Subsequent to the expansion described above in the absence of anti-CD3/CD28 antibody, each cell type was reactivated for 48 h with anti-CD3/CD28 antibody-coated beads (Invitrogen) and then inoculated with virus (400 ng of p24) for 4 h. After infection, the cells were washed and resuspended in fresh RPMI 1640 medium with 10% fetal bovine serum, 1% penicillin/streptomycin, and 50 U/ml IL-2. Cells were fixed in 2% paraformaldehyde 48 h postinfection and analyzed for the percentage of cells expressing GFP on a FACSAria cell sorter/cytometer (Becton Dickinson).

Virions with or without packaged A3G were produced by polyethylenimine transfection of 293T cells with a vif-deleted NL4-3 provirus construct (15 μg of DNA) in combination with an empty control plasmid (A3G-negative virions) or an A3G expression plasmid (3 μg) (A3G-positive virions). Five nanograms of the resulting viral supernatants was used for infection. Th1- and Th2-produced virions were generated as previously described (38), and 5 ng of collected supernatant was used for infection.

PCRs.

Infected Th cells were collected at 2, 4, 18, 24, and 48 h after infection and washed in phosphate-buffered saline. Cellular DNA was isolated using the DNeasy kit (Qiagen). DNA was quantified by spectrophotometry on a GeneQuant Pro (Amersham Biosciences) and normalized prior to use in quantitative PCR (qPCR) assays on an ABI Prism 7000 (Applied Biosystems). To detect early reverse transcription (RT) products, the primers used were 5′-GTGCCCGTCTGTTGTGTGAC-3′ and 5′-GGCGCCACTGCTAGAGATTT-3′, in conjunction with a probe 5′-FAM-CTAGAGATCCCTCAGACCCTTTTAGTCAGTGTGG-TAMRA-3′, where FAM is 6-carboxyfluorescein and TAMRA is 6-carboxytetramethylrhodamine. Late RT products were detected using primers 5′-TGTGTGCCCGTCTGTTGTGT-3′ and 5′-GAGTCCTGCGTCGAGAGAGC-3′ with probe 5′-FAM-CAGTGGCGCCCGAACAGGGA-TAMRA-3′ (7, 24). For two-long terminal repeat (2-LTR) circle quantification, DNA isolated at 48 h after infection was normalized and used in qPCR. The forward primer was MH535 (5′-AACTAGGGAACCCACTGCTTAAG-3′), the reverse primer was MH536 (5′-TCCACAGATCAAGGATATCTTGTC-3′), and the probe was MH603 (5′-FAM-ACACTACTTGAAGCACTCAAGGCAAGCTTT-TAMRA-3′) (7). For quantification of GFP DNA following reverse transcription, the forward primer used was 5′-AAGCTGACCCTGAAGTTCATCTG-3′, the reverse primer was 5′-TTGAAGAAGTCGTGCTGCTTCAT-3′, and the probe was 5′-FAM-ACCGGCAAGCTGC-MGB NFQ-3′ (19), where MGB NFQ is minor groove binder-nonfluorescent quencher.

Cellular fractionation.

In vitro-differentiated Th1 and Th2 cells (expanded as described above in the absence of an activating stimulus) were either restimulated by anti-CD3/CD28 antibody for 48 h (activated) or not stimulated in IL-2-containing medium for 72 h (nonactivated). Activated and nonactivated cells were stained with CD4-APC-Cy7 and CD25-FITC antibodies (BD Pharmingen, San Jose, CA) to determine differences in this activation marker. Cells were then lysed in 50 mM HEPES (pH 7.4), 125 mM NaCl, 0.2% NP-40, 0.1 mM phenylmethylsulfonyl fluoride, and EDTA-free protease inhibitor cocktail (CalBiochem, San Diego, CA) and subjected to ultracentrifugation as previously described (125,000 × g, TLA 55 rotor [Beckman Coulter]) (28). Equal volumes of supernatant and pellet were subjected to Western blotting using polyclonal anti-APOBEC3G antibody (33), with a goat anti-rabbit secondary antibody conjugated to Alexa Fluor 680 (Invitrogen Molecular Probes, Carlsbad, CA). Band intensity was quantified using the Odyssey infrared imaging system (Li-Cor Biosciences, Lincoln, NE). Values are expressed as the fraction of relative light units (RLUs) of the supernatant band relative to the total RLUs for both supernatant and pellet.

Cellular lysates of activated Th1 and Th2 cells were also subjected to centrifugation through a 4% to 40% sucrose gradient as previously described (39). Gradients were centrifuged overnight at 32,000 rpm in an SW-41 rotor (Beckman), and then 12 1-milliliter fractions were collected, precipitated by trichloroacetic acid, resuspended in sodium dodecyl sulfate sample buffer, and subjected to Western blotting. Values are presented as the percent RLUs in each fraction relative to the total RLUs for each sample.

HIV-1 infection of, and integration into, CD4⁺ Th1 cells relative to Th2 cells.

Following 48 h of T-cell receptor activation by anti-CD3/CD28 antibody-coated beads, cytokine-polarized Th1 and Th2 cell cultures were infected with VSV-G-pseudotyped, vif-competent HIV-1 expressing GFP (2, 6, 9). Forty-eight hours after infection, the percentage of cells with integrated HIV genomes was determined by GFP expression using fluorescence-activated cell sorting analysis. In cells from each of three donors, infection of Th2 cells yielded greater GFP signal than did infection of Th1 cells (Fig. 1A).

Since the block to reverse transcription in some primary cells has been attributed to the “low-molecular-mass” form of A3G (10), we separated A3G forms in Th subtype cells by two methods. Sucrose gradient density centrifugation was validated previously to separate A3G forms (39). Density gradients of activated Th1 cell lysates confirmed that 39% of the total A3G was in the two lower-density fractions (fractions 5 and 6) and that 23% was in the two highest-density fractions (fractions 11 and 12). This compared to 32% of the total A3G in the two lower-density fractions and 28% in the two highest-density fractions, respectively, in activated Th2 cells (Fig. 1B and C).

A second method of characterizing A3G forms was also used. “Low-molecular-mass” A3G has previously been shown to remain in the supernatant after ultracentrifugation (28). Following expansion, Th1 and Th2 cells were either placed in fresh medium with no activation signal (nonactivated) or restimulated with anti-CD3/CD28 antibody-coated beads (activated) (Fig. 2A) and then subjected to ultracentrifugation. After ultracentrifugation, activated Th1 cell lysates had less supernatant A3G than nonactivated lysates did (Fig. 2B and C). A lower amount of supernatant A3G was also seen in activated Th2 cells relative to nonactivated cells. A substantial proportion of A3G remained in the supernatant of activated lysates from both Th1 and Th2 cells after ultracentrifugation. The amount of total and supernatant A3G was lower in the Th2 cell lysates than in Th1 cell lysates (Fig. 2B and C).

Reverse transcription in CD4⁺ Th1 lymphocytes relative to Th2 lymphocytes.

Given the reduced amount of GFP from integrated virus in Th1 cells, we next determined at which step of infection this restriction was occurring relative to Th2 cells. A qPCR assay for early and late reverse transcription products showed that the two cell types differed in their ability to support reverse transcription (Fig. 3). Figures 3A and B indicate that infected Th1 cells lag behind Th2 cells from one donor in the amount of both early and late RT products formed. There was a significant difference at a single time point (18 h after infection) for multiple donors in both early and late RT products (Fig. 3C and D), as well as a significant difference between Th1 and Th2 early and late RT products from three replicate infections of cells from a single donor (Fig. 3E).

2-LTR circle formation in CD4⁺ Th1 lymphocytes relative to Th2 lymphocytes.

The measurement of 2-LTR circles formed following HIV-1 infection serves as a marker for nuclear entry of reverse transcripts and abortive integration events (11). We studied whether 2-LTR circle formation differed after A3G-free virion infection of activated Th1 versus Th2 cells. Infection of activated Th1 cells with A3G-free virions led to significantly more 2-LTR circles than did infection of activated Th2 cells (Fig. 4A). This is consistent with relatively more reverse transcripts not integrating and forming abortive 2-LTR circle forms in activated Th1 cells than in Th2 cells. Previous studies reported, however, that A3G packaged in the virion reduced 2-LTR circle formation as well as decreasing integration, consistent with impaired host cell-mediated 2-LTR circle formation and integration (3, 21). Therefore, we also evaluated the effects of virion-packaged A3G on 2-LTR circle formation.

To study virion packaged A3G, TZM-bl cells, which do not contain A3G, were infected with vif-deleted virions that contained A3G or did not contain A3G. These virions were produced by transfection of 293T cells with or without a cotransfected A3G expression plasmid. Infection with vif-deleted viruses containing A3G led to fewer 2-LTR circles relative to TZM-bl cells infected with A3G-free virions (Fig. 4B). Fewer 2-LTR circles were also seen in TZM-bl cells infected with Th1 cell-produced virions than in those infected with Th2 cell-produced virions (Fig. 4C), consistent with the previously documented greater packaging of A3G in Th1 cell-produced virions (38). Thus, virion-packaged A3G and cytoplasmic A3G have different effects on 2-LTR circle formation.

Changes in APOBEC3G expression modulate reverse transcription, 2-LTR circle formation, and integration.

To determine whether the difference in cytoplasmic A3G between Th1 and Th2 lymphocytes causes the observed differences in reverse transcription, 2-LTR circle formation, and integration between these cells when infected, the expression level of A3G was altered in each cell type. Retroviral transduction of A3G was used to increase A3G expression in Th2 cells as previously reported (38). Since the A3G-transducing vector was derived from HIV and the virus used for subsequent infection expresses GFP in place of nef, GFP DNA was quantified by PCR rather than using the early and late HIV-1 reverse transcription product primers. The amount of reverse-transcribed GFP DNA was reduced in Th2 cells overexpressing A3G relative to control vector-transduced Th2 cells (Fig. 5A and B). The formation of 2-LTR circles was also increased after infection in Th2 cells overexpressing A3G relative to control vector-transduced cells (Fig. 5C). A3G overexpression in Th2 cells also reduced the amount of integrated genomes as measured by GFP production (Fig. 5D).

As further confirmation of the role of endogenous cytoplasmic A3G in modulating reverse transcription, 2-LTR circle formation, and integration, Th1 cells were incubated with a neutralizing anti-IFN-γ antibody previously documented to reduce A3G expression in Th1 cells (38) or with an isotype control antibody. Infection of Th1 cells treated with the neutralizing anti-IFN-γ antibody with GFP-expressing HIV-1 led to increased early and late RT products (Fig. 6A), reduced 2-LTR circle formation (Fig. 6B), and an increase in integrated genomes based on increased GFP expression (Fig. 6C) compared to the cells treated with the isotype antibody control.