HIV-1 Vpr drives a tissue residency-like phenotype during selective infection of resting memory T cells

Abstract

HIV-1 replicates in CD4⁺ T cells, leading to AIDS. Determining how HIV-1 shapes its niche to create a permissive environment is central to informing efforts to limit pathogenesis, disturb reservoirs, and achieve a cure. A key roadblock in understanding HIV-T cell interactions is the requirement to activate T cells in vitro to make them permissive to infection. This dramatically alters T cell biology and virus-host interactions. Here we show that HIV-1 cell-to-cell spread permits efficient, productive infection of resting memory T cells without prior activation. Strikingly, we find that HIV-1 infection primes resting T cells to gain characteristics of tissue-resident memory T cells (T_RM), including upregulating key surface markers and the transcription factor Blimp-1 and inducing a transcriptional program overlapping the core T_RM transcriptional signature. This reprogramming is driven by Vpr and requires Vpr packaging into virions and manipulation of STAT5. Thus, HIV-1 reprograms resting T cells, with implications for viral replication and persistence.

Keywords: CP: Immunology; CP: Microbiology; HIV-1; Vpr; cell-cell; permissivity; resting memory T cell; tissue residency; transcriptional reprogramming.

Graphical abstract

Figure 1

HIV-1 exploits cell-to-cell spread to preferentially infect resting memory CD4⁺ T cells (A) HIV-1 NL4.3 infected mitogenically activated primary CD4⁺ donor T cells co-cultured with resting autologous primary CD4⁺ target T cells separated by a 0.4 μm transwell (cell-free) or in direct co-culture (cell-cell). Target cell infection was measured by intracellular staining for HIV-1 Gag protein. Representative flow cytometry plots are shown. Bar graphs show mean of independent experiments (n = 4). (B) Cell-to-cell spread into resting or αCD3/αCD28-activated CD4⁺ target T cells measured by intracellular Gag expression (n = 5). (C and D) Cell-to-cell spread from activated primary donor CD4⁺ T cells to resting primary target CD4⁺ T cells preferentially infects CD45RA^- memory CD4⁺ T cells. A representative flow cytometry plot and quantification are shown (n = 4). (E) Quantification of infection performed as in (C) (n = 11). (F) HIV-1 infection of target CD4⁺ T cells as part of the total resting CD4⁺ T cell population (total) compared with pre-isolated naive and memory CD4⁺ target T cells (isolated) (n = 9). (G and H) Quantification of infection of CXCR4 (X4)- and CCR5 (R5)-tropic viruses (n = 4) (G) and transmitter/founder viruses HIV-1 CH040 and CH077 (n = 7) (H). (I) Representative flow cytometry plots of cell-to-cell infection of resting CD4⁺ T cells with CCR5-tropic HIV-1 NL4.3 BaL and transmitter founder viruses HIV-1 CH040 and CH077 as performed in (C). (J and K) Cell-to-cell infection of resting CD4⁺ T cells is reduced by the HIV-1 fusion inhibitor T20 (n = 6) (left) and the reverse transcriptase inhibitor efavirenz (n = 6) (right) measured by intracellular Gag staining (median fluorescence intensity, MFI) (J) or HIV-1 LTR-driven GFP-reporter gene expression (n = 4) (K). (L) HIV-1 infection downregulates CD4 expression. Shown is the percentage of CD4⁺ cells in the total CD3⁺ target cell population (n = 6). (M–O) Resting CD4⁺ memory T cells were isolated after 72 h of cell-to-cell spread by fluorescence-activated cell sorting (FACS) and cultured for 4 days. HIV-1 infection was measured by intracellular Gag staining (M) and virus release was measured by culture supernatant reverse transcriptase (RT) activity (N) (n = 5–7). T cells recovered at day 1 or 4 post-isolation were then cultured with uninfected eFluor450⁺ target Jurkat T cells, and infection of Jurkat T cells was measured after 72 h (O) (n = 3). All measurements were made after 72 h or at the indicated time post co-culture. Data are the mean ± SEM. Paired two-tailed t test or one-way ANOVA with Bonferroni post test was used. For (L), the median + IQR is shown and Friedman test with Dunn’s post test was used. For (O), unpaired one-tailed t test was used. ^∗p < 0.05; ^∗∗p < 0.01; ^∗∗∗p < 0.001; n.s., not significant.

Figure 2

HIV-1 infection induces a T_RM-like phenotype in resting memory CD4⁺ T cells (A) CD69 expression on resting memory CD4⁺ target T cells following co-culture with HIV-1-infected primary donor T cells or uninfected donor T cells (mock) (n = 17). (B) Representative flow cytometry plots from (A). (C) CD69 expression on infected resting memory CD4⁺ T cells ± IL-7 and T20 (n = 7). (D) CD69 expression on infected resting memory CD4⁺ T cells ± IL-7 and ruxolitinib (n = 8). (E) CD69 expression on infected resting memory CD4⁺ T cells in response to IL-7 and IL-15 (n = 11). (F) CD69 expression on infected Gag⁺ resting memory CD4⁺ T cells and uninfected Gag⁻ bystander cells in response to IL-7 and IL-15 (n = 11). (G) CXCR6 surface expression from (F) (n = 11). (H) Representative flow cytometry plots of CD69 and CXCR6 co-expression in the presence of IL-7. (I) Co-expression of CD69 with CXCR6, CD49A, or PD-1 on infected resting memory CD4⁺ T cells (n = 5–7). (J) As for (I) in the presence of IL-7 (n = 4–7). (K) As for (I) comparing infected Gag⁺ memory CD4⁺ T cells and uninfected Gag⁻ bystander cells. (L and M) Blimp-1 expression in CD69⁺ HIV-infected resting memory CD4⁺ T cells and infected CD69⁻ cells in the presence of IL-7 (n = 8). (N) Total lymphocytes from cellularized tonsils co-cultured with HIV-1-infected Jurkat T cells. Infection of resting CD4⁺ T cells shown as CD45RO versus Gag. (O) Representative flow cytometry plots of CD69 and CXCR6 co-expression on infected Gag⁺ and uninfected Gag⁻ tonsil resting memory CD4⁺ T cells ± IL-7. (P) Recall cytokine response by HIV-1-infected Gag⁺ resting memory T cells. At 72 h of co-culture, expression of IFN-γ, IL-2, or TNF was measured after stimulation with PMA/ionomycin and brefeldin A for the indicated duration. Gag⁺ cells were categorized by CD69 expression (n = 6). (Q) Mean proportion of Gag⁺ resting memory T cells expressing one, two, or three of the cytokines IFN-γ, IL-2, or TNF after 6 h of PMA/ionomycin stimulation in the presence of brefeldin A, categorized by CD69 expression (n = 8). All measurements were made after 72 h or at the indicated time post co-culture. Data are the mean ± SEM. Paired two-tailed t test or one-way ANOVA with Bonferroni or Dunnett’s post test were used. ^∗p < 0.05; ^∗∗p < 0.01; ^∗∗∗p < 0.001; n.s., not significant. MFI, median fluorescence intensity.

Figure 3

Vpr drives HIV-1-induced T_RM induction in resting memory CD4⁺ T cells Resting memory CD4⁺ T cells were co-cultured with primary CD4⁺ T cells infected with HIV-1 wild-type (WT) or mutant viruses or with uninfected donor cells (mock). (A) CD69 upregulation in response to IL-7 compared with mock (n = 9). (B) CD69 expression on HIV-1-infected Gag⁺ resting memory CD4⁺ T cells compared with uninfected Gag⁻ bystander cells (n = 9). (C) Quantification of cell-to-cell spread of HIV-1 WT and ΔVpr to resting naive and memory CD4⁺ T cells (n = 9). (D) CD69/CXCR6/CD49a co-expression on resting memory CD4⁺ T cells infected with HIV-1 WT or ΔVpr (n = 9). (E) CD69, CXCR6, and PRDM1 (Blimp1) mRNA levels from FACS-sorted infected resting memory CD4⁺ T cells. Fold change relative to uninfected (mock) is shown (n = 5). (F) IFN-γ expression by HIV-1-infected resting memory CD4⁺ T cells at 72 h in response to IL-7 (n = 3). (G) Total lymphocytes from cellularized tonsils were co-cultured with HIV-1 WT- or ΔVpr-infected Jurkat T cells in the presence of IL-7. Expression of CD69 (left) or CD69/CXCR6 (right) was measured in Gag⁺ infected and Gag⁻ uninfected bystander cells (n = 4). (H) Western blot showing Vpr packaging into HIV-1 WT and Vpr-mutant virions. Values indicate Vpr levels normalized to p24, relative to WT. (I) CD69 upregulation in response to IL-7 on resting memory CD4⁺ T cells infected with HIV-1 WT, ΔVpr, or Vpr mutants (n = 9). (J) Co-expression of CD69 and CXCR6 from (I) (n = 9). All measurements were made after 72 h or at the indicated time post co-culture. Data are the mean ± SEM. Paired two-tailed t test or one-way ANOVA with Bonferroni or Dunnett’s post test was used. 2LTR circles (I) were compared by unpaired one-tailed t test. ^∗p < 0.05; ^∗∗p < 0.01; ^∗∗∗p < 0.001; n.s., not significant.

Figure 4

Incoming Vpr is sufficient to drive T_RM induction in resting memory CD4⁺ T cells (A) CD69 (left) and CD69/CXCR6 (right) co-expression in response to IL-7 in the presence of integrase inhibitor raltegravir (n = 6). (B) Quantification of integrated provirus and 2LTR circles in FACS-sorted target CD4⁺ memory T cells after 72 h of cell-to-cell spread in the presence or absence of raltegravir. (C) Schematic depicting the viruses and VLPs used in (D–H). (D) Western blot showing Vpr packaging into virions of HIV-1 WT_PM, Vpr_PM, WT, ΔVpr, and ΔVpr complemented with FLAG-tagged Vpr in trans (ΔVpr⁺Vpr_trans). (E) CD69 (left) and CD69/CXCR6 (right) upregulation in response to IL-7 on Gag⁺ resting memory CD4⁺ T cells at 72 h infected with the indicated HIV-1 viruses (n = 8). (F) Expression of CD69 (left) and CD69/CXCR6 (right) in response to IL-7 on Gag⁺ resting memory CD4⁺ T cells at 72 h post spinoculation of HIV-1 WT, Vpr, and ΔVpr⁺Vpr_trans (n = 10). (G) Western blot showing packaging of FLAG-tagged Vpr into Env-VLPs or full-length HIV-1 WT or ΔVpr. (H) Expression of CD69 (left) and CD69/CXCR6 (right) in response to IL-7 on Gag⁺ resting memory CD4⁺ T cells at 72 h post spinoculation of Env-VLPs with or without Vpr (n = 5). All measurements were made after 72 h or at the indicated time post co-culture or spinoculation. Data are the mean ± SEM. Paired two-tailed t test or one-way ANOVA with Bonferroni or Dunnett’s post test was used. ^∗p < 0.05; ^∗∗p < 0.01; ^∗∗∗p < 0.001; n.s., not significant. EV, empty vector.

Figure 5

Transcriptional profiling of HIV-1-infected resting memory CD4⁺ T cells (A) Heatmap showing hierarchical clustering of 226 differentially expressed genes (DEGs) of infected (HIV-1 WT) over uninfected (mock) resting memory CD4⁺ T cells at 72 h post co-culture (adjusted p < 0.01, fold change ±1.2). Mean log2 TPM of four biological repeats are shown. Cytokine indicates presence or absence of IL-7. Virus indicates infection with HIV-1 WT, HIV-1 ΔVpr, or uninfected (mock) condition. (B) Principal-component analysis (PCA) of (A), with ellipses indicating 95% CI. (C and D) Scatterplots of mean log2 TPMs of DEGs from HIV-1 WT/mock (gray circles) or HIV-1 ΔVpr/mock (orange circles) in the absence (C) or presence (D) of IL-7 (adjusted p < 0.01, fold change ±1.2). Lines indicate line of identity (LOD). Genes above or below the LOD are up- or downregulated, respectively. (E and F) Venn diagrams showing overlap of DEGs comparing expression profiles of HIV-1 WT/mock with HIV-1 ΔVpr/mock (E) or HIV-1 ΔVpr/HIV-1 WT (F). (G) GSEA was performed on expression profiles comparing HIV-1 WT/mock (black) or HIV-1 ΔVpr/HIV-1 WT (gray). Normalized enrichment scores (NES) are shown for significantly enriched hallmark gene sets (false discovery rate [FDR] q < 0.05 and NES > 1.75). (H and I) Top 10 significantly enriched canonical pathways predicted by ingenuity pathway analysis (IPA) of DEGs in HIV-1 WT/mock (H) or HIV-1 ΔVpr/HIV-1 WT (I) (adjusted p < 0.05). (J and K) Cytokines (J) and transcription regulators (K) predicted to be upstream regulators by IPA of gene expression signatures for HIV-1 WT/mock (black) or HIV-1 ΔVpr/mock (gray); line indicates p = 0.05. TPM, transcripts per million.

Figure 6

Vpr drives a T_RM-like transcriptomic program in HIV-1-infected resting memory CD4⁺ T cells (A) Heatmap showing hierarchical clustering based on a T_RM core gene expression signature (Kumar et al., 2017) that was performed to compare transcriptional profiles of in vitro HIV-1-infected resting memory CD4⁺ T cells (mock, HIV-1 WT, HIV-1 ΔVpr) with previously described ex vivo gene expression profiles (Kumar et al., 2017). Kumar et al. “Cell subset” indicates ex vivo CD69⁺ T_RM (TRM [tissue]), CD69⁻ non-T_RM (non-TRM [tissue]), tissue-derived T cells (from lung or spleen), and blood-derived CD69⁻ T cells (non-TRM [blood]). Reuschl et al. “HIV-1” indicates infection with HIV-1 WT, HIV-1 ΔVpr, or uninfected (mock) conditions. Presence of IL-7 is indicated by X. (B) The T_RM signature score for the indicated conditions calculated based on (A). Subsets from Kumar et al. (2017) are indicated in red; shown are CD4⁺ or CD8⁺ T cells from lungs or spleens. T_RM⁺, CD69⁺ T cells; T_RM⁻, CD69⁻ T cells. T_RM signature scores for resting CD4⁺ memory T cells infected or uninfected are shown in the presence or absence of IL-7. Means are shown. One-way ANOVA with Dunnett’s post test was used to compare groups in (B).

Figure 7

Vpr enhances STAT5 activation to drive a T_RM-like resting memory CD4⁺ T cell phenotype (A) GSEA enrichment plot of the hallmark IL-2 STAT5 signaling pathway for HIV-1 WT-infected resting memory T cells versus mock. (B) CD69 expression on infected resting memory CD4⁺ T cells ± ruxolitinib at 72 h (n = 4). (C) CD127 MFI on infected resting memory CD4⁺ T cells ± IL-7 (n = 7) at 72 h. (D) Representative histogram of intracellular STAT5 phosphorylation in resting memory T cells infected by cell-to-cell spread at 72 h. (E) Quantification of (D) shown as P-STAT5 MFI (n = 10). (F) P-STAT5 MFI (left) and %P-STAT5⁺ (right) in Gag⁺ resting memory T cells 24 h post spinoculation with the indicated viruses (n = 12). (G) Representative western blot analysis of P-STAT5 and total STAT5 levels in resting T cells at 0 and 24 h post spinoculation with HIV-1 WT and ΔVpr virus (n = 2). Values indicate P-STAT5 or total STAT5 levels normalized to β-actin and mock at 0 h. (H) Quantification of P-STAT5/STAT5 levels normalized to β-actin from western blots of total CD4⁺ T cells at 24 h post spinoculation with HIV-1 WT and ΔVpr virus (n = 4). (I and J) Quantification of (I) P-STAT5 and (J) total STAT5 levels by single-cell immunofluorescence analysis of Gag⁺ resting T cells 24 h post spinoculation with HIV-1 WT and ΔVpr virus ± IL-7 (P-STAT5 n = 1,829–2,000 cells/condition; STAT5 n = 223–900 cells/condition). Normalized mean intensities (quantifications) and representative images of P-STAT5 in HIV-1 WT- and ΔVpr-infected cells without IL-7 (I, right) are shown. P-STAT5, green; Gag, red; Hoechst 33342, blue. Scale bars, 10 μm. (K) P-STAT5 MFI and %P-STAT5⁺ (right) in Gag⁺ resting memory T cells 24 h post spinoculation with VLPs with or without Vpr (n = 7). (L) CD69 (left) and CD69/CXCR6 (right) expression on infected resting memory CD4⁺ T cells in the presence of IL-7 ± STAT5-inhibitor AC-4-130 at 72 h (n = 6). All measurements were made after 72 h or at the indicated time post co-culture or spinoculation. Data are the mean ± SEM. Paired two-tailed t test or one-way ANOVA with Bonferroni or Dunnett’s post test was used. For (I) and (J), median is indicated and groups were compared using Kruskal-Wallis test with Dunn’s post test. ^∗p < 0.05; ^∗∗p < 0.01; ^∗∗∗p < 0.001; n.s., not significant. MFI, median fluorescence intensity.