Transcriptional Reprogramming during Effector-to-Memory Transition Renders CD4+ T Cells Permissive for Latent HIV-1 Infection

Abstract

The latent reservoir for HIV-1 in resting memory CD4⁺ T cells is the major barrier to curing HIV-1 infection. Studies of HIV-1 latency have focused on regulation of viral gene expression in cells in which latent infection is established. However, it remains unclear how infection initially becomes latent. Here we described a unique set of properties of CD4⁺ T cells undergoing effector-to-memory transition including temporary upregulation of CCR5 expression and rapid downregulation of cellular gene transcription. These cells allowed completion of steps in the HIV-1 life cycle through integration but suppressed HIV-1 gene transcription, thus allowing the establishment of latency. CD4⁺ T cells in this stage were substantially more permissive for HIV-1 latent infection than other CD4⁺ T cells. Establishment of latent HIV-1 infection in CD4⁺ T could be inhibited by viral-specific CD8⁺ T cells, a result with implications for elimination of latent HIV-1 infection by T cell-based vaccines.

Figure 1. Latent Infection by CCR5-tropic HIV-1 Is Not Efficient in Naive, Resting Memory, or Activated CD4⁺ T Cells

(A and B) Replication competent HIV-1 (A) was isolated from resting CD4⁺ T cells of eight patients using a limiting dilution virus outgrowth assay. The patient from whom only CXCR4-tropic virus was recovered was highlighted in blue. Genomic DNA (B) was isolated from resting CD4⁺ T cells of 11 patients. Viral env sequences were analyzed by the PSSM system. (C and D) Bcl-2-transduced resting or activated CD4⁺ T cells were infected using a pseudotyped HIV-1 (NL4-3-Δenv-drEGFP) with a R5 (Yu-2) or X4 (NL4-3) envelope. HIV-1 gene expression was assessed by flow cytometry. Productive infection (C) was measured 3 days after infection. To generate latent infection, infected cells were cultured in basal medium without supplement of antibodies or cytokine for another 25 days before removal of GFP-positive cells. Latent infection (D) of Bcl-2-transduced CD4⁺ T cells was measured 2 days after reactivation of latent virus by stimulation with anti-CD3 and anti-CD28 antibodies. The limit of detection of latent HIV-1 infection by flow cytometry is 0.01%. Blood samples from 12 healthy donors were used for the analysis.

Figure 2. Latent Infection of CCR5-tropic HIV-1 Is Not Efficient in CCR5⁺ or CCR5⁻Resting Memory CD4⁺ T Cells

(A) CD4⁺ T cells isolated from a healthy donor were stimulated with anti-CD3 and anti-CD28 antibodies or left untreated for 3 days. Activated and resting (CD25⁻, CD69⁻, and HLA-DR⁻) CD4⁺ T cells were analyzed for CD45RO, CCR5, or CCR7 expression. (B) Resting memory CCR5⁺ or CCR5⁻ CD4⁺ T cells were purified by sorting from freshly isolated healthy donor CD4⁺ T cells and cultured for 24 hr. Culture medium was then collected for the measurement of cytokine production. Activated CD4⁺ T cells were used as control. Error bars represent SEM, n = 3. (C) Productive and latent infection of resting memory CCR5⁺ CD4⁺ T cells. Three populations of Bcl-2-transduced CD4⁺ T cells were purified by FACS for infection experiments: (1) CCR5⁻ activated cells, (2) CCR5⁺ memory cells, and (3) CCR5⁻ memory cells. Purified cells were infected with NL4-3-Δenv-drEGFP pseudotyped with HIV-1-X4 or -R5 envelope. Error bars represent SEM, n = 3. Flow data were presented in standard, 4 log biexponential plots. In (B) and (C), data were pooled from three healthy donors.

Figure 3. EMT CD4⁺ T Cells Are More Susceptible to CCR5-tropic HIV-1 Infection due to Upregulation of CCR5 Expression

(A) HIV-1 co-receptor expression and markers of T cell activation and proliferation were measured and compared in naive (CD25⁻, CD69⁻, HLA-DR⁻, and CD45RA⁺), fully activated, EMT, and resting memory (CD25⁻, CD69⁻, HLA-DR⁻, and CD45RO⁺) CD4⁺ T cells generated from freshly isolated healthy donor CD4⁺ T cells. (B and C) CCR5 gene transcription and surface expression in EMT CD4⁺ T cells. In (B), blood samples from ten healthy donors were used for CCR5 staining. In (C), data were pooled from two donors. (D and E) R5-tropic HIV-1 enters EMT CD4⁺ T cells expressing high level of CCR5. Data were pooled from two donors. (D) Total CD4⁺ T cells; (E) CCR5⁺ CD4⁺ T cells. Error bars represent SEM, n ≥ 3. t test values: ***p < 0.001; **p < 0.01. See also Figures S1–S4.

Figure 4. CCR5-tropic HIV-1 Preferentially Infects EMT CD4⁺ T Cells

CD4⁺ T cells isolated from healthy donor were used to generate resting, activated, and EMT CD4⁺ T cells. (A) 24 hr or 72 hr after infection of different CD4⁺ T cell populations with X4-tropic virus, late reverse transcription products were measured by qPCR and normalized to levels in activated cells. (B and C) Productive expression by R5-tropic HIV-1 was assessed by GFP expression 72 hr after infection. (D) CD4⁺ T cells at different activation states were co-transfected by electroporation (Lonza) with a proviral construct (pNL4-3-Δenv-drEGFP) with eGFP in the env ORF, and a control expression vector in which BFP expression was driven by SV40 promoter. GFP and BFP expression was measured 24 hr after transfection. Error bars represent SEM, n = 3. t test values: ***p < 0.001; **p < 0.01; *p < 0.05. Data were pooled from three independent infections or transfections. See also Figure S5.

Figure 5. Establishment of HIV-1 Latent Infection Occurs Preferentially in EMT CD4⁺ T Cells

Freshly isolated CD4⁺ T cells from healthy donors were used in (A)–(C). Bcl2-transduced CD4⁺ T cells from healthy donors were used in (D) and (E). (A) Microarray analysis of gene transcription profile was performed in CD4⁺ T cells at different activation states. The top 100 most upregulated (comparing to resting cells) and top 50 most up-regulated NF-κB responsive genes in activated CD4⁺ T cells were represented. Three healthy donor CD4⁺ T cells were used for microarray analysis. (B) Cytokine production was measured from CD4⁺ T cells at different activation states. Data were pooled from three independent experiments. (C) CD4⁺ T cells at different activation states were infected with NL4-3-Δnef-EGFP pseudotyped with X4 envelope. Individual GFP-positive cells were purified by single cell sorting 24 hr after infection for mRNA isolation and quantification. Quantitative measurement of pol2a and HIV-1 gag transcripts was performed in individual HIV-1-infected cells. Data were pooled from two healthy donor CD4⁺ T cells. (D and E) NL4-3-Δenv-drEGFP pseudotyped with HIV-1-X4 or -R5 envelope was used to infect CD4⁺ T cells at different activation states. Productive (D) and latent (E) infection was measured as describe in Figure 1. Data were pooled from six healthy donor CD4⁺ T cells. Error bars represent SEM, n ≥ 3. t test values: **p < 0.01; *p < 0.05. See also Figures S6 and S7.

Figure 6. Latent HIV-1 Is Enriched in CCR5⁺ Resting Memory CD4⁺ T Cells from HIV-1 Patients

(A) Frequency of HIV-1 proviral DNA in naive, CCR5⁻CD45RO⁺, and CCR5⁺ CD45RO⁺ resting CD4⁺ T cells was measured by qPCR. t test values: ***p < 0.001; *p < 0.05. (B) Frequency of latent HIV-1 in CCR5⁻ and CCR5⁺ resting memory CD4⁺ T cells was measured by limiting dilution virus outgrowth assay. Ratio-paired t test was performed for statistical analysis. (C) Relative contribution of CCR5⁺ and CCR5⁻ subsets to the reservoir for latent HIV-1 in memory CD4⁺ T cells in each patient was calculated according to the frequency of latently infected cells in each subset and the frequency of each subset in the whole memory CD4⁺ T cell population.

Figure 7. HIV-1-Specific CTLs Inhibits Latent HIV-1 Infection in EMT CD4⁺ T Cells

(A) Patient-derived Bcl-2-transduced EMT CD4⁺ T cells were infected with NL4-3-Δenv-drEGFP pseudotyped with HIV-1 R5 envelope. Autologous CD8⁺ T cells pre-stimulated with Gag peptides mixture were added to the culture 24 hr after infection at the ratio of 1:1. Productive infection with (dotted line) or without (solid line) the presence of autologous CD8⁺ T cells was monitored over time by FACS. (B) GFP-positive cells as well as CD8⁺ T cells described in (A) were removed from the co-culture 15 days after infection. The remaining cells were stimulated with CD3&CD28 antibodies for the measurement of latent infection. Data were pooled from three HIV-1 patient samples. Error bars represent SEM, n = 3. t test values: *p < 0.05.