Purification of resting and endogenous activated CD4⁺ T cells. Mononuclear cells were stained with a cocktail of monoclonal antibodies against lineage-specific antigens, including FITC-labeled anti-CD8, -CD14, -CD16, -CD20, and -CD56. Simultaneously, the cells were stained with phycoerythrin-labeled monoclonal antibodies against the activation markers CD69 and HLA-DR. The gates used to define the two populations are shown on the presort panel. Postsort analysis showed that endogenous activated CD4⁺ T cells expressed high levels of activation markers and resting CD4⁺ T cells did not express detectable levels of activation markers. To determine the purity of the resting CD4⁺ T cells, the quadrants were set so that 99% of unstained lymphocytes were in the lower left quadrant. The sorting gates were set conservatively to ensure high purity. We consistently found that only 1% of the resting T cells were contaminants and 1% of the activated cells were contaminants. Dendritic cells copurified with the endogenous activated T cells. We chose not to remove dendritic cells from the activated T cells. We reasoned that the activated CD4⁺ T cells were more likely to remain activated in the presence of dendritic cells. The resting CD4⁺ T cells did not contain dendritic cells.

Reverse transcription and integration have delayed kinetics in resting CD4⁺ T cells. The number of copies per cell of short reverse transcripts and long reverse transcripts are shown at several time points after inoculation. Short reverse transcripts were detected using RU5-specific primers. Long reverse transcripts were detected using primers specific for reverse transcripts that form after second strand transfer. Integrated provirus is shown at several time points after inoculation. The cells were cultured in the presence of the protease inhibitor saquinavir to prevent spreading infection. When we inoculated resting CD4⁺ T cells with HIV and cultured in the presence of the integrase inhibitor L870,810 at 100 nM, the integration level was reduced by 96.7% while the level of reverse transcripts was unchanged, demonstrating the specificity of our integration assay. When we inoculated and cultured the cells in the presence of zidovudine at 10 μg/ml, the number of reverse transcripts was reduced by >99%, demonstrating that our viral supernatant contained minimal amounts of contaminating viral DNA. The error bars (standard deviations) are obscured by the symbols. The coefficient of variation ranged from 3 to 17% in this experiment.

Spinoculation and culture did not induce resting CD4⁺ T cells to express HLA-DR or CD69. Sort-purified resting CD4⁺ T cells were stained with anti-CD69 or anti-HLA-DR-FITC immediately after spinoculation (immed. post spin), 3 days after spinoculation (3d post spin), and 6 days after spinoculation (6d post spin). The gray boxes represent the staining seen with anti-HLA-DR and anti-CD69 antibodies. The black line represents the staining obtained with an IgG1 isotype control antibody immediately after sorting. CD69 expression was also not detected at 2, 6, and 20 h after spinoculation (data not shown). The data are representative of three experiments.

HIV-1 integrated into resting CD4⁺ T cells. (A) The relative levels of total DNA and integrated DNA in the two populations of primary T cells at 3 days postinoculation are shown. Total DNA was measured by primers that detect second strand transfer. Integrated DNA was measured by Alu-PCR (51). The relative levels of total DNA and integrated DNA for the activated T-cell line, CEMss, at 18 h postinoculation are also shown. Immediately after spinoculation only very low levels of total DNA (<0.001 copies/cell) and no integrated DNA were detected (not shown), indicating that our viral supernatants contained minimal amounts of contaminating viral DNA from the initial transfection. Comparing the resting and endogenous activated CD4⁺ T cells, there was a significant difference in the amount of total DNA (P < 0.02) and integrated DNA (P ≪ 0.01) by Student's t test. Similar results were obtained with HIV-1_YU2, a CCR5-tropic virus, and pNL4-3, another CXCR4-tropic virus (not shown). The data are representative of four experiments. (B) Spinoculation did not induce integration. Binding and integration were increased proportionally by spinoculation. Resting CD4⁺ T cells were inoculated under routine and spinoculation conditions and washed to remove unbound virions, and the number of bound virions per cell was measured as described elsewhere (52). The number of integrated proviruses per cell was measured 3 days after inoculation as described previously (51). The data are representative of three experiments.

Spinoculation and culture did not induce resting CD4⁺ T cells to enter the cell cycle. Data are plotted as DNA (7AAD) on the y axis versus RNA (pyronin Y) on the x axis. Resting CD4⁺ T cells were fixed at various time points: immediately before spinoculation (before spin), immediately after spinoculation (immed. post spin), and 3 days after spinoculation (3d post spin). Cell cycle stages were defined by activating T cells by treating PBMCs with SEB in the presence of the inhibitor sodium butyrate for 1 day (G_1a), in the presence of aphidicolin for 4 days (G_1b), or in the absence of an inhibitor for 4 days (G₂). To demonstrate that cells did not enter the cell cycle after spinoculation, quadrants were set using CD4⁺ T cells before spinoculation. The data are representative of three experiments.

A fraction of resting CD4⁺ T cells bearing integrated HIV-1 DNA was latently infected. In each plot, the percentage of events in the boxed region is shown. The experiments were performed in the presence of the protease inhibitor saquinavir to prevent spreading infection. (A) IL-7 alone. Uninfected cells treated with IL-7 showed minimal background staining with anti-Gag antibodies. Uninfected resting CD4⁺ T cells were cultured for 3 days and then stimulated with IL-7 for an additional 3 days. This was a control for background staining with the anti-Gag monoclonal antibody. Isotype control antibodies gave less background staining than anti-Gag antibodies. (B) HIV alone. Unstimulated HIV-inoculated cells stained minimally with anti-Gag antibodies. Resting cells were spinoculated with HIV and 6 days later were permeabilized and stained for intracellular Gag. (C) HIV plus IL-7. IL-7 stimulated 3.0% of HIV-inoculated cells to produce Gag. Resting cells were spinoculated with HIV, 3 days later they were stimulated with IL-7, and 3 days after IL-7 addition they were permeabilized and stained for intracellular Gag. Gag was measured 3 days after IL-7 stimulation because Gag production peaked at that time. (D) HIV plus CD3 and CD28 beads. Anti-CD3 and -CD28 beads stimulated 4.5% of HIV-inoculated cells to produce Gag. Resting cells were spinoculated with HIV, 3 days later they were stimulated with anti-CD3 and -CD28 beads in the presence of the integrase inhibitor L870,810, and 24 h later they were permeabilized and stained for intracellular Gag. Gag was measured 24 h after bead stimulation because Gag production peaked at that time. We demonstrated that our integrase inhibitor was effective by inoculating resting CD4⁺ T cells and culturing them in the presence of the integrase inhibitor L870,810 at 100 nM. Under these conditions, we saw no reduction in the frequency of reverse transcripts, but we found that the frequency of proviruses/cell was reduced by 96.7%.