CD101 is expressed on memory CD4 T cells and highly upregulated on regulatory CD4 T cells

While CD101 has been investigated in chronic viral infection on CD8 T cell populations using the LCMV mouse model, its role on CD4 T cells has not been investigated in the rhesus macaque model and there have been no studies of CD101⁺ CD4 T cells in acute and chronic viral infection [34]. To understand the role of CD101 across CD4 T cell subsets in healthy animals, we performed flow cytometric analyses of cryopreserved peripheral blood mononuclear cells (PBMC), lymph node mononuclear cells (LN) and rectal mucosa biopsies (RB). While CD101 was not expressed on naïve CD4 T cells (CD28+ CD95-), it was moderately expressed on memory CD4 T cells (CD95+) and significantly enriched on memory CD4 Tregs (CD95+ CD25+ CD127- FoxP3+), with more than 60% of Tregs expressing CD101 in both blood and LN (Fig 1A, 1B and 1C) (S1 Fig). Despite the significant levels of CD101 expression on CD4 Tregs, they make up only a small fraction of the total CD101-expressing CD4 pool, as the overall contribution of CD101⁺ CD4 T cells largely comes from central and effector memory cells, due to their higher frequency of the total pool of CD4 T cells (Fig 1C). To further investigate the phenotype of CD101-expressing cells, we utilized Uniform Manifold Approximation and Projection (UMAP) of our flow cytometry analysis and focused on memory CD4 T cells. In combination with Phenograph analysis, this revealed distinct clusters of CD101⁺ memory CD4 T cells (Figs 1D and 1E and S2). In particular, CD101 was found highly expressed in clusters (2, 6, 17) also expressing markers of T regulatory cells, including CD25, FoxP3, CD39 and CTLA-4 (Fig 1D and 1E). An additional group of clusters (8, 10, 15, 23) confirmed the expression of CD101 in the central memory compartment, with co-expression of CCR7 and CD127 (Fig 1D and 1E).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Phenotype of CD101-expressing CD4 T cells in healthy rhesus macaques.

A) Representative staining of CD101 on CD4 subsets in healthy RM PBMC. B) Representative histograms of CD101 on detailed CD4 subsets in healthy RM PBMC. C) Expression of CD101 within detailed CD4 subsets in PBMC and LN from healthy RM (top); total contribution of CD101⁺ cells to the overall CD4 pool from each CD4 subset in PBMC and LN. D) UMAP plot of flow cytometry data showing overall Phenograph cluster location and expression intensity of markers of interest. E) Hierarchical clustering of expression (z-score) for selected markers from flow cytometry data in each Phenograph cluster. F) Heatmap of the top 50 differentially expressed genes between CD101⁺ and CD101^- Tregs (CD95+ CD25+ CD127-) by scRNA-seq analysis of healthy RM PBMC. G) Violin plots showing RNA expression levels of 4 genes of interest. H) Cytokine expression levels in CD101- and CD101+ memory CD4 after 3hr PMA/ionomycin stimulation of total PBMC from healthy individuals.

https://doi.org/10.1371/journal.ppat.1010723.g001

Given the high expression of CD101 on CD4 Tregs, we performed single-cell RNA-seq (scRNA-seq) analysis on CD4 Tregs (CD95+ CD25+ CD127-) that were sorted into CD101^- and CD101⁺ populations. scRNA-seq analysis revealed clearly distinct transcriptional profiles of CD101⁺ Tregs as compared to CD101^- Tregs (Fig 1F and S1 File). Specifically, we found an expression profile suggesting CD101⁺ Tregs were more differentiated, with lower levels of TCF7, and showed higher potential for suppressive activity with lower levels of SATB1 (a chromatin modulator that represses FoxP3 expression) and higher levels of TNFRSF18 (which encodes for GITR), both of which have been strongly linked to the ability of Tregs to exert suppressive activity in vivo [35–37]. Additionally, CD101⁺ Tregs were highly enriched in expression of LGALS3 (which encodes for Galectin-3), which has been shown to increase N-glycan branching on CD8 T cells, increasing the antigenic threshold for activation of these cells and contributing to long-term viral persistence in chronic LCMV model by limiting CD8 T cell antiviral functions [38].

To confirm that CD101⁺ cells represent a functionally distinct population, we performed PMA/ionomycin stimulation of PBMC from healthy individuals. After a 3hr stimulation, CD101⁺ CD4 T cells produced similar levels of traditional effector cytokines IL-2, IFNy, IL-17a and TNFa as compared to CD101- cells (Figs 1H and S3). However, CD101+ CD4 T cells produced significantly more LAP (Latency Associated Peptide; mean 11.7 fold-difference) than CD101- cells in response to stimulation. LAP is part of the latent TGF-β complex and is produced by Tregs that release TGF-β, suggesting that cells producing LAP are likely to be immunosuppressive [39–41]. Collectively, these observations suggest CD101-expressing cells represent a functionally distinct population of CD4 T cells that can play a key role in regulating inflammatory environments, such as those after a viral infection.

CD101⁺ CD4 T cells are selectively depleted during SIV infection

To determine the role of CD101-expressing CD4 T cells in SIV infection, we assessed the dynamics of CD101⁺ CD4 T cell populations in SIV_mac239-infected RMs. CD101⁺ CD4 T cells are severely depleted during acute SIV infection (14 days post-infection), with a mean 2.6-fold reduction in the frequency of blood memory CD4 T cells expressing CD101 (Fig 2A and 2B). Absolute counts of CD101⁺ and CD101^- CD4 subsets in the periphery were evaluated to determine and compare the absolute loss of these populations during acute infection. CD101⁺ populations were significantly more depleted across all CD4 subsets as compared to CD101^- cells (Fig 2C). This preferential depletion was most severe for Tregs, as 80% of CD101⁺ Tregs were lost compared to only 44% of CD101^- Tregs, and for CM, with 84% of CD101⁺ cells lost compared to 64% of CD101^- (Fig 2C). Notably, CD4 Tregs shift from being a primarily CD101⁺ population (accounting for 66% of total Treg) to being majority CD101^- after SIV infection (only 32% of total Treg being CD101⁺), representing a shift during acute infection toward a population of Tregs with a reduced suppressive potential (Fig 2D and 2E). While depletion of CD101⁺ memory CD4+ T cells, or CD101⁺ Tregs, was not associated with viral load (copies HIV RNA/mL of plasma) or T cell activation at D14 p.i., there was a significant association between the loss of CD101⁺ cells and higher plasma viral load at D42 p.i. (Table 1). Additionally, lower levels of CD101⁺ cells at D42 p.i. was associated with increased plasma levels of the inflammatory cytokines IP-10 and MIP-1a (Table 1). These findings suggest a direct relationship between the depletion of CD101+ cells, an increased inflammatory environment and higher viral burden. Interestingly, increased CD8 T cell activation as assessed by levels of HLA-DR/CD38 was associated with better preservation of CD101+ CD4 T cell, suggestive of a potential for CD8 antiviral responses to limit the depletion of this population (Table 1).To expand this observation from SIV-infected RMs to humans, we assessed these populations in PBMCs from PWH and confirmed that CD101 expression on CD4 Tregs is significantly lower in individuals with active viremia compared to ART-suppressed individuals, despite equivalent overall frequencies of CD4 Tregs (Figs 2F and S4). To examine whether this depletion represented preferential susceptibility of CD101⁺ cells to HIV infection, we sorted CD101^- and CD101⁺ memory CD4 T cells from healthy human PBMC and performed an in vitro HIV infection. After 7 days in culture with ART, we harvested the cells and measured integrated HIV DNA, and found comparable levels of integrated HIV DNA between CD101^- and CD101⁺ cell cultures (S5A Fig).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Dynamics of CD101-expressing CD4 T cells after SIV infection.

A) Representative staining of CD101 on PBMC memory CD4 pre- and post-SIV infection. B) Expression levels of CD101 on PBMC memory CD4 pre- and post-SIV infection. C) Total loss of CD101⁺ and CD101^- cells within CD4 subsets calculated based on absolute counts in blood pre- and post-SIV infection (mean+SEM). D) Representative staining of CD101 on PBMC CD4 Tregs pre- and post-SIV infection. E) Expression levels of CD101 on PBMC CD4 Tregs pre- and post-SIV infection. F) Expression levels of CD101 on PBMC CD4 Tregs from HIV-infected individuals during active viremia and during suppressive antiretroviral therapy (ART). G) Representative staining of CD101 on rectal biopsy (RB) memory CD4 pre- and post-SIV infection. H) Expression levels of CD101 on RB memory CD4 pre- and post-SIV infection. I) Expression levels of α4β7 and CCR5 on RB memory CD4 CD101^- and CD101⁺ cells from healthy RM. J) Association between depletion of CD101+ CD4 T cells in rectal biopsies at day 14 p.i. (calculated from baseline to acute infection as % of live lymphocytes) and SIV DNA levels in rectal biopsies at day 14 p.i. (spearman correlation, n = 9). K) CD101 expression within PBMC CD4 subsets during longitudinal SIV infection (mean±SEM). L) Overall contribution of CD101⁺ cells to the overall CD4 pool from each PBMC CD4 subset during longitudinal SIV infection (mean±SEM). Lines designate mean values.

https://doi.org/10.1371/journal.ppat.1010723.g002

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Correlations of CD101 dynamics with plasma viral load, immune activation and plasma cytokine levels.

Analyses were done as Spearman correlations with r and p value displayed. All values are from the D14 p.i. or D42 p.i for 28 SIV-infected RMs. CD4 loss was calculated based on absolute counts of CD101+ CD4 pre- and post-infection. Statistically significant findings are highlighted in yellow (n = 28).

https://doi.org/10.1371/journal.ppat.1010723.t001

To understand if the preferential depletion of CD101⁺ cells was limited to peripheral blood or represented a migration of CD101⁺ CD4 to tissues, we evaluated these populations in rectal biopsy samples (RB) before and following SIV infection. Similar to PBMC, there was a marked, rapid (already present at d14 p.i.), depletion of CD101-expressing CD4 T cells in the gut after SIV infection (Fig 2G and 2H). This depletion was preferential, with 97% loss of CD101⁺ CD4 T cells compared to only 83% loss of CD101^- CD4 T cells (calculated from baseline to acute infection as % of live lymphocytes) (S6A Fig). Gut depletion of CD4 T cells is well described to be one of the main drivers of systemic inflammation and disease progression in viremic individuals during SIV/HIV infection, and the loss of highly suppressive CD101⁺ CD4 T cells could strongly contribute to this inflammation [42–44]. To investigate potential mechanisms for preferential depletion of CD101⁺ cells in the gut, we assessed levels of α4β7 and CCR5 on CD101⁺ and CD101^- memory CD4 T cells. These well-described surface receptors are known to drive T cell migration to the gut and to favor SIV/HIV cell entry. In uninfected macaques, CD101⁺ CD4 T cells in the gut express significantly more α4β7 and CCR5 than CD101^- cells (Fig 2I). This finding suggests CD101⁺ CD4 T cells have a high potential to migrate and reside in the gut tissues and is consistent with their propensity to be more depleted in the gut during acute SIV infection, especially given the role of α4β7-high cells as early targets for HIV/SIV infection and of CCR5 as a co-receptor for direct HIV/SIV infection [45]. To understand potential links between this depletion and viral burden in gut mucosa, we assessed SIV DNA and RNA levels in rectal biopsy samples from day 14 post-infection. We found a significant association between loss of CD101⁺ CD4 T cells (calculated from baseline to acute infection as % of live lymphocytes) and higher SIV DNA levels during acute infection (Fig 2J), while there was no association of SIV DNA/RNA levels with depletion of CD101^- CD4 T cells (S1 Table). This finding establishes a specific link between depletion of CD101⁺ cells in the mucosa and disease progression. Finally, we evaluated levels of circulating biomarkers of microbial translocation and mucosal barrier breakdown (sCD14, IFABp, zonulin) pre- and post-SIV infection. Notably, we found that lower levels of CD101⁺ cells in RB 14 days post-infection was significantly associated with a higher fold-change of zonulin from pre-infection to day 42 post-infection (S6B Fig) (S2 Table). Zonulin is a key regulator of intestinal permeability, with higher circulating levels of zonulin being indicative of intestinal epithelial damage, and has been previously shown to be predictive of HIV-related mortality [46].

To assess whether CD101 is being downregulated on CD4 T cells after infection rather than these cells being directly depleted, we evaluated CD101 surface levels after in vitro HIV infection of sorted CD101^- and CD101⁺ CD4 T cell cultures. Sorted CD101⁺ cells remained almost entirely CD101⁺ (median 92% CD101⁺) after 7 days in culture and remained clearly distinct from the CD101^- cultures, which remained CD101^- (S5B and S5C Fig). To confirm CD101 is not downregulated by T cell activation, we performed aCD3/CD28 stimulation of PBMC from healthy individuals. After 5 days of stimulation, CD101 levels on the total CD4 population had significantly increased (mean 2.5-fold increase) (S5D Fig).

Considering the potential inflammatory implications of a rapid and permanent depletion of highly suppressive Tregs, it is important to understand whether the depleted populations of CD101⁺ memory CD4 and Tregs are restored during antiretroviral therapy. To this aim, we quantified the frequency of blood CD4 T cell subsets expressing CD101 at an early (6 weeks), mid (24 weeks), and late (33–60 weeks) on ART time points. Within PBMC, the proportion of cells expressing CD101 within each CD4 subset was restored starting from early ART, and fully to pre-infection levels after long-term ART in RMs that initiated ART during the early chronic phase of infection (Fig 2K), thus suggesting direct viral infection or inflammation, both reduced by ART, as mechanisms for the loss of CD101-expressing cells. This restoration occurred very rapidly in the Treg compartment, suggesting the administration of ART prevents further depletion of CD101⁺ cells and allows a restoration of this suppressive population in response to the highly inflammatory environment. In fact, CD101⁺ cells begin to slowly recover prior to the initiation of ART, which is likely reflective of the tendency of Tregs to expand during the chronic phase of HIV/SIV infection and the emergence of SIV-specific CD8 T cells that are able to partially limit viremia during early chronic infection. The contribution of CD101⁺ cells within each subset to the overall pool of CD4 T cells was also fully restored to pre-infection levels (Fig 2L).

During long-term ART, CD101⁺ cells display a phenotype consistent with quiescence and potential for reservoir persistence

We then investigated whether CD101⁺ cells may contribute to viral persistence during long-term ART following reconstitution. Expression of numerous inhibitory receptors on CD4 T cells have previously been described to potentially delineate subsets enriched for HIV/SIV provirus and contribute significantly to the persistent, latent reservoir [6,12,17,18,47]. Given the more differentiated state of CD101⁺ Tregs and associations of CD101 with inhibitory receptor expression on CD8 T cells during chronic LCMV infection, we hypothesized that CD101⁺ cells may be enriched for these markers during long-term ART [34]. We evaluated expression of PD-1 and CTLA-4 within the CD101^- and CD101⁺ memory CD4 populations in LN samples from SIV-infected RMs after >1 year of ART. While CD101^- cells primarily lie in the PD-1^- CTLA-4^- quadrant, CD101⁺ cells are significantly enriched in the PD-1^- CTLA-4⁺ and PD-1⁺ CTLA-4⁺ compartments (Fig 3A and 3B). Of note, we and others have previously reported that these populations of PD-1^- CTLA-4⁺ and PD-1⁺ CTLA-4⁺ cells harbor high levels of replication competent SIV DNA and include a high frequency of Tregs and Tfh, respectively [6,12,17]. Directly probing the expression of CD101 on the PD-1 and CTLA-4 quadrants of LN CD4 T cells revealed significantly elevated expression of CD101 in cells expressing any combination of PD-1 and CTLA-4 (Fig 3C and 3D). We confirmed similar results of CD101 expression on these cell populations in PBMC from PWH on ART (Fig 3E). We further investigated additional immunosuppressive receptors CD39 and TIGIT that are known to be expressed on functional Tregs and have been associated with the HIV reservoir [17,48–50]. CD101⁺ cells express significantly more CD39 and TIGIT as compared to CD101^- cells and Boolean gating on the 4 receptors of interest reveals a clear enrichment of immunosuppressive receptors on CD101⁺ cells, with 84% of CD101⁺ cells expressing at least one of these receptors compared to only 49% of CD101^- cells, and 63% of CD101⁺ cells expressing two or more receptors as compared to only 23% of CD101^- (Fig 3F, 3G and 3H). Additionally, we found that CD101⁺ cells express significantly higher levels of Ki-67 in both ART-suppressed SIV and HIV infection, suggestive of recent cell cycling of CD101⁺ cells (Fig 3I and 3J). This is notable considering the potential for cycling CD4 T cells to maintain the reservoir via clonal expansion [15,51]. Finally, to assess the functional capacity of CD101⁺ CD4 T cells to contribute to reservoir persistence, we performed PMA/ionomycin stimulation of PBMC from ART-suppressed PWH. Similar to the profile observed in cells from healthy individuals, CD101⁺ CD4 T cells expressed significantly more LAP (TGF-β) as compared to CD101^- cells, while producing similar levels of other effector cytokines (Fig 3K). This is particularly interesting, as TGF-β levels are known to be significantly elevated after HIV infection and remain elevated during suppressive ART and TGF-β has been demonstrated to play a role in potentiating T cell latency [52–56].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Phenotypic profile of CD101+ CD4 during long-term antiretroviral therapy.

A) Representative staining of PD-1 and CTLA-4 within naïve, memory CD101^- and CD101⁺ CD4 T cells in lymph node (LN) from ART-suppressed SIV-infected RM. B) Quadrant distribution of PD-1 and CTLA-4 expression within CD101^- and CD101⁺ LN memory CD4 from ART-suppressed SIV-infected RM. C) Representative histogram of CD101 expression across PD-1 and CTLA-4 expression quadrants in LN memory CD4 from ART-suppressed SIV-infected RM. D) CD101 expression across PD-1 and CTLA-4 quadrants in LN memory CD4 from ART-suppressed SIV-infected RM. E) CD101 expression across PD-1 and CTLA-4 quadrants in PBMC memory CD4 from ART-suppressed HIV-infected individuals. F,G) Expression levels of CD39 (F) and TIGIT (G) within CD101^- and CD101⁺ LN memory CD4. H) Pie chart representing the frequency of cells expressing PD-1 and/or CTLA-4 and/or CD39 and/or TIGIT within CD101- and CD101⁺ LN memory CD4. I,J) Expression levels of Ki-67 within CD101^- and CD101⁺ memory CD4 from LN from ART-suppressed SIV-infected RM (I) and PBMC from ART-suppressed HIV-infected individuals (J). K) Cytokine expression levels in CD101- and CD101+ memory CD4 after 3hr PMA/ionomycin stimulation of total PBMC from ART-suppressed PLWH. Lines designate mean values.

https://doi.org/10.1371/journal.ppat.1010723.g003

To assess the potential contribution of CD101⁺ CD4 T cells to the latent HIV reservoir, we measured HIV DNA levels in sorted CD101^- and CD101⁺ memory CD4 T cells derived from PBMC from PWH on suppressive ART (Fig 4A). CD101⁺ CD4 harbored similar levels of total proviruses, defective HIV DNA, and intact HIV DNA as compared to CD101^- cells when assessed by the intact proviral DNA assay, suggesting that CD101⁺ cells are contributing to the overall pool of latently infected cells during long-term ART at levels comparable to the CD101^- counterpart (Fig 4B, 4C and 4D) [57]. Due to their high levels of intact DNA and a phenotype consistent with cell quiescence and survival, CD101⁺ CD4 cells have the potential to critically contribute to the long-term maintenance of the SIV/HIV reservoir.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. HIV DNA levels in sorted memory CD4 T cells.

A) Sorting strategy for memory CD101^- and CD101⁺ memory CD4 T cells. B-D) Total HIV proviruses (B), defective HIV DNA copies (C) and intact HIV DNA copies (D) measured by IPDA in sorted CD101^- and CD101⁺ memory CD4 T cells from PBMC from ART-suppressed individuals.

https://doi.org/10.1371/journal.ppat.1010723.g004

Ethics statement

All animal experimentation was conducted following guidelines set forth by the Animal Welfare Act and by the NIH’s Guide for the Care and Use of Laboratory Animals, 8th edition. All studies were reviewed and approved by Emory’s Institutional Animal Care and Use Committee (IACUC; permit numbers 201700655, 201800047, 201700665) and animal care facilities at Yerkes National Primate Research Center are accredited by the U.S. Department of Agriculture (USDA) and the Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC) International. Proper steps were taken to minimize animal suffering and all procedures were conducted under anesthesia with follow-up pain management as needed. Individuals with HIV infection were consented for a study approved by the Emory University Institutional Review Board and written consent for research was obtained from the individuals.

Animals, SIV infection, antiretroviral therapy and sample collection

53 Indian rhesus macaques (RMs), housed at the Yerkes National Primate Research Center, were included in this study. All RMs were screened to ensure Mamu-B*08- and Mamu-B*17- status. Animals were infected intravenously with 300 TCID50 SIVmac239. Animals initiated a daily subcutaneous antiretroviral therapy regimen of FTC (40 mg/kg), TDF (5.1 mg/kg) and DTG (2.5 mg/kg) during the early chronic phase of infection (week 6–8 post-infection) and were maintained on ART up to 66 weeks post-infection (S3 Table). Peripheral blood, lymph node and rectal biopsy sample collections were conducted at critical timepoints during the study and processed as previously described [12]. Infection timepoints were defined as follows: acute (days 14–18 post-infection), early chronic (days 42–56 post-infection), early ART (6 weeks post-ART initiation), mid ART (24 weeks post-ART initiation), late ART (33–60 weeks post-ART initiation).

Samples from participants with HIV infection

For analysis of depletion of CD101-expressing CD4 T cells during viremia, PBMC were obtained prior to initiation of antiretroviral therapy in individuals with a confirmed HIV diagnosis. For analysis of CD101 expression during long-term antiretroviral therapy, PBMC were obtained from individuals with undetectable viremia for at least 1 year prior to sample collection. Individuals were on a variety of combination ART regimens containing a protease inhibitor, integrase inhibitor and/or nucleoside reverse-transcriptase inhibitors. Cryopreserved cells were used for all analyses.

Plasma viral load

Levels of SIV RNA in plasma were measured by RT-qPCR as previously described (limit of detection 60 copies/mL) [78].

scRNA-seq analysis

Single cells were sorted directly into 96 well plates with lysis buffer. Single cell lysates were then converted to cDNA following capture with Agencourt RNA Clean beads using the SmartSeq2 protocol as previously described [79]. The cDNA was amplified using 20–24 PCR enrichment cycles prior to quantification and dual-index barcoding with the Illumina Nextera XT kit. The libraries were enriched with 12 cycles of PCR, then combined in equal volumes prior to final bead cleanup and sequencing. All libraries were sequenced on an Illumina HiSeq 3500 by either single-end 150 bp reads or short paired-end reads. Alignment was performed using STAR version 2.5.2b and transcripts were annotated using MacaM Rhesus genome assembly and annotation (v7.8.2: https://www.unmc.edu/rhesusgenechip/index.htm#NewRhesusGenome) [80]. Transcript abundance estimates were calculated internal to the STAR aligner using the algorithm of htseq-count [81].

Flow cytometry

Flow cytometric analyses for this study were performed on peripheral blood, LN and gut-derived cells according to previously optimized and standardized procedures with anti-human antibodies with confirmed cross-reactivity with RMs. For T cell phenotyping analyses, the following antibodies were used at pre-optimized staining concentrations: anti-Ki-67-Alexa700 (clone B56), anti-CD3-BUV395 (clone SP34-2), anti-CD8-BUV496 (clone RPA-T8), anti-CD45-BUV563 (clone D058-1283),anti-CD28-BUV737 (clone CD28.2), anti-CD45RA-BUV737 (clone HI100), anti-CTLA-4-BV421 (clone BNI3), anti-Ki-67-BV480 (clone B56), anti-CD27-BV605 (clone L128), anti-CCR5-APC (clone 3A9), anti-CCR7-BB700 (clone 3D12), Fixable Viability Stain 700 (all from BD Biosciences); anti-FoxP3-AF647 (clone 150D), anti-CD4-APC/Cy7 (clone OKT4), anti-CD95-BV605 (clone DX2), anti-HLA-DR-BV650 (clone L243), anti-CD25-BV711 (clone BC96), anti-CD39-BV711 (clone A1), anti-PD-1-BV786 (clone EH12.2H7), anti-CD39-PE/Dazzle594 (clone A1), anti-CD4-PE/Cy5 (clone OKT4), anti-CD101-PE/Cy7 (clone BB27) (all from Biolegend); anti-CXCR5-PE (clone MU5UBEE), anti-CD127-Pe/Cy5 (clone ebioRDR5), anti-TIGIT-PerCP-eF710 (clone MBSA43), LIVE/DEAD fixable aqua (all from ThermoFisher); anti-CD38-FITC (clone AT-1) (StemCell); anti-α4β7-PE (clone A4B7R1) (NHP reagent resource). To detect intracellular expression of FoxP3, cells were fixed and permeabilized with FoxP3 Fix/Perm solution (Tonbo) and subsequently stained for intracellular markers of interest. Acquisition of stained cells was performed on a minimum of 100,000 live CD3+ T cells for LN and PBMC samples on an LSRFortessa or FACSymphony (BD biosciences) driven by FACSDiva software and analyzed using FlowJo software (version 10.8, Treestar). FCS files were imported into FlowJo, compensated electronically, gated on memory CD4 T cells (CD3+ CD4+ CD8- CD95+) and equivalent numbers of cells were input into UMAP analysis (Uniform Manifold Approximation and Projection for Dimension Reduction) was for unbiased evaluation of the distribution of the key markers. Projection of the density of cells expressing markers of interest were visualized/plotted on a 2-dimensional UMAP (https://arxiv.org/abs/1802.03426, https://github.com/lmcinnes/umap). We used the Phenograph clustering approach (https://github.com/jacoblevine/PhenoGraph).

Flow cytometry cell sorting

In order to perform scRNA-seq on sorted CD101+ and CD101- Tregs, cryopreserved PBMC were thawed and stained with the following antibodies: anti-CD95-BV421 (clone DX2), anti-CD28-PE/CF594 (clone 28.2) (from BD Biosciences); anti-CD4-BV650 (clone OKT4), anti-CD101-PE/Cy7 (clone BB27) (from Biolegend); anti-CD8-FITC (clone MHCD08014), anti-CXCR5-PE (clone MU5UBEE), anti-CD127-PE/Cy5 (clone ebioRDR5), LIVE/DEAD fixable aqua (from ThermoFisher). Cells were gated as live, CD3+, CD4+, CD8- and were then sorted as Naïve (CD28+ CD95-), CD101- Tregs (CD95+, CXCR5-, CD127-, CD25+, CD101-) and CD101+ Tregs (CD95+, CXCR5-, CD127-, CD25+, CD101+).

For HIV DNA measurements in CD101+ and CD101- CD4 T cells, cryopreserved PBMC were thawed and stained with the following antibodies: anti-CCR7-BB700 (clone 3D12), anti-CD45RA-APC (clone HI100), anti-CD3-APC-Cy7 (clone SP34-2), anti-CD8-FITC (clone RPA-T8) (from BD biosciences); anti-CD4-APC/Cy7 (clone OKT4), anti-CD101-PE/Cy7 (clone BB27) (from Biolegend). Cells were sorted as memory CD101⁺ or CD101^- CD4 (gating strategy shown in Fig 4A) and DNA was extracted using a QIAmp DNA minikit (Qiagen) and intact proviral DNA was measured as previously published, with sample processing and IPDA analysis performed by Accelevir Diagnostics in a blinded fashion [57].

In vitro infection

In vitro infection of PBMC was performed as previously described [56]. Briefly, PBMC were isolated from buffy coats from HIV-negative healthy donors and enriched for memory CD4 T cells by negative selection using the EasySep human CD4 memory T cell enrichment kit (StemCell). Cells were rested overnight and then sorted as CD101⁺ or CD101^- using a FACS Aria II system (BD Biosciences). After another overnight rest, cells were spinoculated with 89.6, a clinical isolate of HIV, at 100ng/mL p24/million cells. Viral supernatant was removed and infected cells were cultured in RPMI supplemented with 30 U/mL IL-2 (R&D Systems) and 5uM saquinavir (NIH AIDS Reagent Program) to prevent viral spread and preintegration latency. On day 4, cells were directed into latency with addition of 40ng/mL recombinant human IL-7 (R&D Systems), 20 ng/mL recombinant human TGF-B1 (PeproTech) and an ART cocktail of 5uM saquinavir, 100nM efavirenz and 200nM raltegravir (NIH AIDS Reagent Program). Cells were harvested on day 7 for flow cytometric analysis and measurements of integrated HIV DNA levels. The following antibodies were used for analysis: anti-CD3-BUV395 (clone SP34-2), anti-CD8-BUV496 (clone RPA-T8), anti-CD45RA-BUV737 (clone HI100), anti-CD27-BV605 (clone L128), anti-CCR7-BB700 (clone 3D12), anti-CD45RA-APC (clone HI100), anti-CD3-BV421 (clone SP34-2), anti-CD8-PE/CF594 (clone RPA-T8), (from BD biosciences); anti-CD4-APC/Cy7 (clone OKT4), anti-CD101-PE/Cy7 (clone BB27) (from Biolegend). Acquisition of stained cells was performed on an LSRFortessa or FACSymphony (BD biosciences) driven by FACSDiva software and analyzed using FlowJo software (version 10.8, Treestar). Measurements of integrated HIV DNA on cells from day 7 was performed as previously described [82].

Measurement of intestinal epithelial damage biomarkers

Levels of circulating biomarkers of mucosal damage in plasma were measured by enzyme-linked immunosorbent assay (ELISA) using a commercially available kit as per the manufacturer’s instructions (sCD14: R&D Systems, Cat #DC140l; IFABP: My Biosource, Cat#MBS740424; zonulin: Alpco, Cat #30-ZONSHU-E01).

In vitro stimulations

In vitro stimulations were performed on total PBMC from both healthy donors and ART-suppressed PLWH. For PMA/ionomycin stimulations, cells were stimulated at 37°C with 80 ng/mL PMA and 500 ng/mL ionomycin in the presence of 10ug/mL of BFA and 0.7uL of GolgiStop (BD biosciences). Cells were harvested and stained at 0hr and 3hr post-stimulation, with the 0hr timepoint serving as a negative control. For aCD3/aCD28 stimulations, cells were cultured in RPMI supplemented with 10% heat-inactivated FBS, 100U/mL penicillin, 100ug/mL streptomycin and 20U/mL recombinant human IL-2 (Gemini Bio-products), 5ug/mL anti-CD28-BUV737 (clone CD28.2, BD biosciences), 5ug/mL anti-CD2-purified (clone RPA-2.10, Biolegend), and 5ug/mL anti-CD3-purified (clone SP34-2, BD Biosciences). Fresh IL-2 at 20U/mL was supplemented every 2 days and cells were harvested and stained at days 0, 3 and 5 post-stimulation. For downstream flow cytometry analysis, the following antibodies were used: anti-CD4-APC/Cy7 (clone OKT4), anti-PD-1-BV786 (clone EH12.2H7), anti-CD101-PE/Cy7 (clone BB27), anti-IFNg-PE/Dazzle594 (clone B27), anti-TNFa-APC (clone MAb11), anti-IL-2-BV650 (clone MQ1-17H12), anti-LAP-BV421 (clone TW4-2F8) (From Biolegend); anti-Ki-67-Alexa700 (clone B56), anti-CD3-BUV395 (clone SP34-2), anti-CD8-BUV496 (clone RPA-T8), anti-CD45RA-BUV737 (clone HI100)), anti-CD27-BV605 (clone L128), anti-CCR7-BB700 (clone 3D12) (from BD biosciences); anti-IL-17a-AF488 (clone eBio64DEC17) from eBioscience; LIVE/DEAD fixable aqua from ThermoFisher.

Plasma cytokine levels

Plasma cytokines were measured using the MSD platform as per manufacturer’s instruction with NHP-specific U-PLEX for IL-6 (K156TXK), IP-10 (K156UFK), MIP-1a (K156UJK), TNF-a (K156UCK) from plasma samples from day 42 post-infection. Values below limit of detection were imputed at half of the limit of detection.

SIV DNA/RNA measurements

Levels of SIV DNA and RNA in rectal mucosa were measured as previously described [83, 84]. Briefly, rectal biopsy tissue was homogenized, RNA and DNA were extracted and SIV DNA and RNA levels were measured by qRT-PCR and normalized to cell counts using concurrent rhesus albumin measurements.

Statistical analyses

Analyses were performed using R studio or GraphPad Prism 9. P values ≤ 0.05 were considered statistically significant with the following definitions: *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Differences in unmatched data were evaluated using Mann-Whitney U test and matched data were evaluated using Wilcoxon matched-pairs signed rank test. For samples with multiple comparisons, values were adjusted for multiple comparisons using Dunn’s test, false-discovery rate or Sidak correction. Longitudinal analyses were evaluated using mixed-effects model with Tukey correction. Correlations were calculated using nonparametric Spearman analysis.