Phenotypic analysis of productively SIV-infected cells in blood and lymphoid tissues of rhesus macaques

SUMMARY

Simian immunodeficiency virus (SIV) primarily infects CD4⁺ T cells, yet the relative contribution of different subsets to viral production in blood and tissues remains unclear. We develop SIV-Flow, a flow cytometry-based assay, to quantify and characterize the phenotype of productively infected cells in blood, spleen, and multiple lymph nodes from eight chronically SIV-infected rhesus macaques. We measured a consistent average of ~1,000 infected cells per million CD4⁺ T cells, with limited variation by the anatomical site. Most productively infected cells displayed a memory phenotype, dominated by the effector memory subset. Despite modest tissue-specific differences, infected cells were uniformly enriched for the activation (PD-1 and CD69), integrin (VLA-4), and proliferation (Ki-67) markers. A unique subset of memory cells expressing high levels of Ki67 and PD-1 represented 44% of all infected cells. These data identify proliferating effector memory CD4⁺ T cells as key contributors to viral production and suggest that their phenotypic profile is conserved across diverse tissues.

In brief

The anatomical distribution and phenotype of cells producing SIV remain unclear. Using SIV-Flow, a sensitive flow cytometry-based assay, Trifone et al. show that proliferating CD4 T cells expressing PD-1, CD69, and VLA-4 constitute the largest fraction of productively infected cells and exhibit conserved phenotypic features across lymphoid tissues.

Graphical Abstract

INTRODUCTION

Following mucosal transmission, human immunodeficiency virus (HIV) infects local CD4⁺ T cells and disseminates in multiple lymphoid organs.^1–4 Although autopsy studies have provided critical information on viral persistence in people with HIV on antiretroviral therapy (ART),^4–8 little is known on the relative contributions of different tissues to HIV replication during untreated infection. Non-human primate models, such as simian immunodeficiency virus (SIV)-infected rhesus macaques, offer a unique opportunity to address this question in a comprehensive manner.^9–11 Like HIV, SIV is known to rapidly disseminate across various compartments,^12–14 including the spleen, lymph nodes, gut, brain, blood, liver, and lungs.^15–17 Lymphoid tissues are major sites of HIV and SIV production, as they contain abundant CD4⁺ T cells, particularly T follicular helper (Tfh) cells that are highly susceptible to viral infection.^4,18–27 Moreover, cytotoxic T cells have limited access to B follicles, where productively infected cells predominantly reside.^28–30 Consequently, in both humans and non-human primates, the highest frequencies of cells harboring viral DNA and/or RNA are typically found in lymph nodes, the spleen, and the gut.^19,27,31 However, because these measurements are based on nuclear acid quantification, they do not directly reflect the frequency of productively infected cells. Additional experimental approaches are, therefore, required to more accurately assess the relative contribution of each tissue to overall viral production.

Several studies have investigated the phenotype of the early cellular targets of SIV infection. Within days of exposure, a rapid and profound depletion of activated, memory CD4⁺ T cells occurs in the intestinal mucosa of SIV-infected macaques.^12,32,33 Through all stages of infection, both HIV and SIV preferentially infect activated and proliferating CD4⁺ T cells.^34–37 In blood and lymphoid tissues, memory CD4⁺ T cells (including central [T_CM], transitional [T_TM], and effector memory [T_EM] subsets) constitute the main viral targets during both untreated and ART-treated infection.^{27,31,35,38,39} Moreover, the extent of SIV replication in non-human primates is influenced by the size of the pre-existing Th17 cell compartment,⁴⁰ highlighting the prominent role of this subset in the early stages of infection. CD4⁺ Tfh cells are also highly permissive to HIV/SIV infection and contribute disproportionally to viral production relative to other CD4⁺ T cell subsets.^20,41–43 Remarkably, even in SIV-infected elite controller rhesus macaques, Tfh cells sustain productive SIV replication in B cell follicles in lymph nodes and the spleen.²⁸ Additional phenotypic markers have also been associated with SIV-infected cells. For instance, several immune checkpoint molecules are preferentially expressed by infected cells: CTLA-4 marks cells harboring SIV DNA in both treated and untreated animals, while PD-1 expression is enriched in reservoir cells during ART.⁴⁴ Beyond the intrinsic properties of target cells, SIV infection induces several phenotypic changes in host CD4⁺ T cells, including rapid internalization of CD4 and CD3, which result in decreased surface expression.^45–47

Despite their role in viral dissemination and reservoir establishment prior to ART initiation, the anatomical distribution and phenotype of productively infected cells remain poorly defined. This is largely due to the reliance of most previous studies on cell sorting of predefined and limited numbers of cellular subsets. In this study, we investigated the phenotype of productively infected CD4⁺ T cells in blood and lymphoid tissues during untreated SIV infection. Using a novel flow cytometry-based assay, SIV-Flow, we characterized both the anatomical distribution and the phenotypic features of productively SIV-infected cells.

RESULTS

SIV-Flow approach can specifically detect SIV-infected cells with high sensitivity

To identify productively infected cells (e.g., translationally active), we tested a panel of monoclonal antibodies that specifically recognize the intracellular SIV p27 protein and screened a panel of antibodies, using in vitro Simian Human Immunodeficiency Virus (SHIV)-infected human CD4⁺ T cells (Figure S1). Two antibodies, clones 4B7 and 55–2F12, demonstrated the highest sensitivity and specificity and were subsequently conjugated to bright, spectrally distinct fluorochromes (CF568 and CF647, respectively). Dual staining with these antibodies revealed a clear double-positive population of p27⁺ cells in SHIV-infected primary human CD4⁺ T cells (Figure 1A). To assess the specificity of our assay, we sorted the double-positive, double-negative, and single-positive populations and quantified the frequency of SHIV-infected cells in each subset by qPCR for SHIV DNA. p27 double-positive cells were highly enriched in SHIV DNA, averaging approximately 1 copy of SHIV DNA per cell. In contrast, SHIV DNA frequencies in the single-positive populations were comparable to those measured in the double-negative subset (Figure 1B), validating the specificity of our dual-antibody staining strategy to detect infected cells.

Figure 1. Development of a flow cytometry-based assay to identify SIV-infected cells.

(A) Dot plots of the double p27 staining in human-derived uninfected CD4 T cells (left) and SHIV-in-vitro-infected CD4 T cells (right).

(B) Left: representative dot plot of the SIV-Flow staining, identifying the four cell populations sorted by flow cytometry: p27 double-negative cells (gray), p27 single-positive cells (4B7 clone, blue), p27 single-positive cells (55–2F12 clone, red), and p27 double-positive cells (violet). Right: SHIV DNA quantifications in each sorted subset. Data are represented as mean ± SEM.

(C) HUT78 productively SIV-infected cells were spiked at different ratios into non-infected PBMCs. The graph shows the theoretical frequencies of HUT78 SIV-infected cells vs. the measured frequencies of p27⁺ cells detected when the detection of p27 protein was performed with the 4B7 antibody alone (blue), the 55–2F12 antibody alone (red), and both antibodies (violet). The black dotted line represents the predicted curve.

To evaluate the linearity and sensitivity of this approach, we spiked cells from an SIV-infected cell line, HUT 78,⁴⁸ into non-infected non-human primate-derived peripheral blood mononuclear cells (PBMCs) at different ratios. To isolate a pure population of productively infected cells, we pre-sorted HUT 78 cells based on gp120 surface expression. Between 1 and 100,000 SIV-infected cells were then spiked into 1 million uninfected PBMCs, and the frequency of infected cells was measured using the SIV-Flow p27 dual-staining protocol or each individual anti-p27 antibody. Detection of productively infected cells using a single p27 antibody (55–2F12 or 4B7) largely overestimated the frequency of infected cells at low concentrations, which resulted in a lack of linearity when the frequency of p27⁺ cells was lower than 1,000 cells per million cells. In contrast, the combination of the two monoclonal antibodies exhibited strong linearity (R = 1) and high sensitivity, as demonstrated by the detection of a single positive event in a million cells (Figure 1C).

In summary, these findings demonstrate that SIV-Flow using two antibodies to SIV p27 is a specific and sensitive assay to detect individual SHIV- and SIV-infected cells, with a broad linear detection range spanning from 1 to 100,000 p27⁺ cells per million cells.

SIV-Flow approach can be applied to detect SIV-infected cells in blood and different lymphoid tissues

Using enriched CD4⁺ T cells from blood and lymphoid tissues (spleen as well as axillary, inguinal, and mesenteric lymph nodes) from eight chronically SIVmac239-infected rhesus macaques (median time of infection, 15.1 weeks; median plasma viral load, 1.7 × 10⁶ SIV RNA copies/mL; Table 1), we quantified the frequency of productively infected cells by using the SIV-Flow assay. In parallel, we measured SIV DNA and cell-associated SIV RNA by qPCR. SIV-Flow enabled the detection of productively infected cells in all analyzed tissues, with comparable median frequencies of 818, 956, 672, 767, and 1,326 p27⁺ cells/10⁶ CD4⁺ T cells for axillary lymph nodes, inguinal lymph nodes, mesenteric lymph nodes, blood, and spleen, respectively (Figure 2A). Although these differences did not reach statistical significance, in most animals (6/8), highest frequencies of p27⁺ cells were detected in the spleen. Notably, we observed reduced surface expression of CD4 and CD3 on p27⁺ cells compared with the total CD4⁺ T cells (fold changes = 17.2 and 1.9 and p < 0.0001 and < 0.0001, respectively), further confirming the specificity of our p27 dual-staining for productively infected cells (Figure 2B; Figure S2).

Table 1.

Characteristics of the 8 animals used in this study

Animal ID	Sex	Age (years)	Virus strain	Time of infection (weeks)	Viral load (copies/mL)
34675	male	4.9	SIVm239X	16.0	2.6E+07
33447	male	5.7	SIVm239X	16.1	1.6E+06
33122	male	5.9	SIVm239X	16.4	1.7E+06
33921	male	5.8	SIVm239X	13.4	4.0E+06
35026	male	4.8	SIVm239X	23.9	6.8E+04
34254	male	5.1	SIVm239X	25.8	4.5E+05
37170	male	2.5	SIVm239X	8.0	4.7E+06
37172	male	2.7	SIVm239X	8.1	3.1E+06

Figure 2. Frequencies of productively infected cells measured by SIV-Flow in blood and lymphoid tissues.

(A) Frequencies of p27⁺ cells measured by SIV-Flow in blood and lymphoid tissues from 8 chronically SIV-infected rhesus macaques. Data were analyzed using a non-parametric two-way Kruskal-Wallis test and Dun’s multiple-comparison test. Because of poor cell viability, the analysis of axillary lymph node cells from animal 33447 was not performed.

(B) Cell surface expressions of CD4 and CD3 in p27⁺ cells from all compartments and animals, expressed as the median fluorescence intensity. Data were analyzed using a non-parametric Wilcoxon test.

(C) Proportion of genetically intact genomes in p24⁺ cells. A total of 28 sequences from 2 animals were analyzed.

(D) Frequencies of p27⁺ cells were compared with frequencies of cells harboring SIV DNA and the level of SIV cell-associated RNA. Measures from all animals and all compartments are included. Each tissue is represented by a different color: axillary lymph node (red), inguinal lymph node (green), mesenteric lymph node (blue), blood (violet), and spleen (pink). Data are represented by median values and were analyzed using a non-parametric Wilcoxon test.

(E) Correlations between the frequency of infected cells determined by SIV-Flow and SIV DNA (left) and SIV cell-associated RNA (right) measures. Measures from all animals and all compartments are included. Each tissue is represented by a different symbol color as in (C). Data were analyzed using the Spearman’s correlation test.

(F) Correlation matrix showing the associations between the frequencies of p27⁺ cells and other virologic markers. The results were analyzed per compartment, from left to right: axillary, inguinal, and mesenteric lymph nodes; blood; and spleen. R values are represented in the upper right portion of each matrix. p values are shown in the lower left portion of each matrix. Data were analyzed using the Spearman’s correlation test.

*, p < 0.05; **, p < 0.01; ****, p < 0.0001.

To determine the fraction of p27⁺ harboring genetically intact SIV genomes, we sorted individual p27⁺ cells from 2 animals and performed near-full-length genome amplification and sequencing. From the 28 proviruses analyzed, 89.3% were found to be genetically intact, while 10.7% harbored deletions, primarily in the env gene (Figure 2C). We conclude that most cells detected by SIV-Flow harbor intact proviruses and, therefore, likely contribute to viral replication.

To further validate the SIV-Flow assay, we tested associations with conventional virologic assays including plasma viral load, SIV DNA, and cell-associated SIV RNA. In all compartments, frequencies of SIV-infected cells determined by SIV-Flow were lower than frequencies of cells harboring SIV DNA (median fold difference of 25.8, p < 0.0001) (Figure 2D), suggesting that only a small fraction of cells harboring HIV genomes actively produce virions, even during uncontrolled viremia. Despite this difference, frequencies of p27⁺ cells correlated with the frequencies of cells harboring SIV DNA and with the levels of cell-associated RNA (p = 0.0019 and p = 0.004, respectively; Figure 2E). Plasma viral loads positively correlated with the frequency of p27⁺ cells measured in the axillary and inguinal lymph nodes (p = 0.042 and 0.038, respectively) but not in the mesenteric lymph node, neither blood nor the spleen (Figure 2F). When analyzing each compartment individually, we observed strong positive associations between the frequencies of p27⁺ cells and the levels of SIV-DNA and cell-associated SIV RNA in axillary lymph nodes. In inguinal lymph nodes, SIV-Flow values correlated only with cell-associated HIV RNA. In contrast, p27⁺ cell frequencies measured in the mesenteric lymph node and blood correlated exclusively with SIV DNA levels in these compartments. Notably, there was no correlation between the SIV-Flow measures and any virologic markers in the spleen. We conclude that in untreated SIV infections, SIV DNA and cell-associated SIV RNA measures do not necessarily reflect the frequency of productively infected cells as measured by p27 expression. Frequencies of p27⁺ cells in the axillary and inguinal lymph nodes correlated with cell-associated SIV RNA in these sites and with plasma viral load, suggesting that these sites contribute to the production of SIV virions.

Most productively SIV-infected cells display an effector memory phenotype in blood and lymphoid tissues

To identify the memory phenotype of productively SIV-infected cells across various compartments, we used the SIV-Flow assay and performed a multiparametric flow cytometry analysis in 5 tissues from 8 chronically SIV-infected rhesus macaques. The gating strategy employed to distinguish memory subsets is illustrated in Figure 3A. Naive (T_N) cells were characterized as CD28⁺ CD95⁻ within the bulk CD4⁺ T lymphocyte population. After the exclusion of T_N cells, T memory stem cells (T_SCM) were defined as CD45RA⁺ CCR7⁺ T cells. The remaining cells were subdivided into T_CM (CD28⁺ CCR7⁺), T_TM (CD28⁺ CCR7⁻), and T_EM (CD28⁻) CD4⁺ T cells.

Figure 3. Analysis of the memory phenotype of productively SIV-infected cells.

(A) Representative dot plots showing the gating strategy used to analyze subsets of CD4⁺ T cells, with infected cells (p27⁺, in red) overlaid on uninfected cells (p27⁻, in gray). T_N cells were defined as CD28⁺ CD95⁻ within the bulk CD4⁺ T lymphocyte population. Following the exclusion of T_N cells, stem cell memory (T_SCM) cells were defined as CD45RA⁺CCR7⁺. The remaining CD4⁺ T cells were classified into central (T_CM) (CD28⁺CCR7⁺), transitional (T_TM) (CD28⁺ CCR7⁻ ), and effector memory (T_EM) cells (CD28⁻).

(B) Pie charts illustrating the relative contribution of each memory subset to the bulk CD4 T cell population (top row) and to the pool of SIV-infected cells (bottom row) in each compartment. Data were analyzed using the permutation test. *p < 0.05 for comparison between tissues; ^#### p < 0.0001 for comparison between CD4⁺ T cells and p27⁺ cells.

(C) Matrix showing the fold enrichment of p27⁺ cells in each cellular subset and in each compartment. Data are represented as median fold-enrichment values. Fold enrichments were calculated as described in STAR Methods.

(D) Comparison of the contribution of each memory subset to the pool of CD4⁺ T cells and to the pool of p27⁺ cells. Contributions are plotted for each compartment, and individual values per animal are shown. Data were analyzed by 2-way ANOVA followed by a Sidak’s multi-comparison post hoc test. *, p < 0.05; ***, p < 0.001; ****, p < 0.0001.

Analysis of the global pool of CD4⁺ revealed that T_N cells were the predominant CD4⁺ T cell subset across all compartments and constituted 83% in axillary lymph nodes, 84% in mesenteric lymph nodes, 83% in inguinal lymph nodes, 77% in blood, and 64% in the spleen (Figure 3B). In sharp contrast, in all compartments, the majority of productively infected cells exhibited a memory phenotype (T_SCM, T_CM, T_TM, and T_EM), with frequencies of 81% in axillary lymph nodes, 77% in mesenteric lymph nodes, 82% in inguinal lymph nodes, 76% in the blood, and 86% in the spleen (Figure 3B); T_EM cells had the highest proportion of productively infected cells, with frequencies of 54% in axillary lymph nodes, 53% in mesenteric lymph nodes, 58% in inguinal lymph nodes, 48% in blood, and 65% in the spleen (Figure 3B). This resulted in a highly significant enrichment of infected cells within the T_EM subset across all analyzed compartments (10.5- to 36.8-fold enrichment, Figure 3C). Conversely, the contribution of T_N cells to the pool of productively infected cells was substantially lower than their contribution to the pool of CD4⁺ T cell. Of note, a modest enrichment of infected cells was also observed within the T_SCM population in all compartments, but this difference reached statistical significance only in the mesenteric lymph node (Figure 3D). Altogether, these results indicated that productively infected cells are highly enriched in the T_EM subset in blood and multiple lymphoid tissues.

Productively infected cells are enriched in Ki67⁺, PD-1⁺, CD69⁺, and VLA-4⁺ subsets in blood and lymphoid tissues

Having demonstrated an enrichment of productively infected cells in the T_EM subset, we sought to determine if specific markers were preferentially expressed by these cells. We focused on markers associated with T cell activation and exhaustion (CD69, PD-1, and CTLA-4), proliferation and survival (Ki67 and CD127), and migratory capacities (α4β1/VLA-4 and α4β7; Figure 4A). In all compartments, PD-1 was more frequently expressed by p27⁺ cells than their uninfected counterparts (Figure 4B). Similarly, productively infected cells were highly enriched in cells expressing the proliferation marker Ki67 across all compartments, suggesting that rapidly dividing cells are preferential targets of SIV in all tissues. For both PD-1 and Ki67, the highest fold enrichments were observed in blood and the axillary lymph nodes, although the enrichment was obvious in all compartments (Figure 4C). Interestingly, CD69 marked productively infected cells in all lymphoid tissues but not in blood, which suggested that tissue-resident T cells are preferential targets for SIV in lymphoid compartments. VLA-4, an integrin that promotes the migration of T cells to inflamed lymphoid tissues, was more frequently expressed by p27⁺ cells than their uninfected counterparts in blood, the spleen, and mesenteric lymph nodes (Figures 4B and 4D). This was mostly attributed to the heightened expression of the β1 chain because neither the individual expressions of the β7 and α4 chains (Figure 4D) nor their co-expression (Figure 4D) significantly differed between the infected and uninfected cells. Although trends were noted for the enrichments of p27⁺ cells in CD4⁺ T cells expressing CTLA-4 and negative for CD127 expression, these did not reach statistical significance in any of the compartment we studied. Notably, the lower expression of CD127 could be attributed to the enrichment of infected cells in T_EM cells, which typically do not express this marker. Altogether, these results indicated that proliferating cells expressing PD-1 are preferentially targeted by SIV in all compartments, whereas other markers such as CD69 and VLA-4 mark productively infected cells in specific tissues.

Figure 4. Phenotypic analysis of the activation status of productively SIV-infected cells.

(A) Representative dot plots showing the gating strategy used to analyze the expression of activation and proliferation markers, with infected cells (p27⁺, in red) overlaid on uninfected cells (p27⁻, in gray).

(B) Comparisons of the frequencies of CD4⁺ T cells and p27⁺ cells expressing each marker. Contributions are plotted for each compartment, and individual values per animal are shown.

(C) Matrix showing fold enrichments in SIV-infected cells in cellular subsets expressing the activation and proliferation markers (columns) across the different compartments (rows).

(D) Matrix showing fold enrichments in SIV-infected cells in cellular subsets expressing integrins (columns) across the different compartments (rows). Fold enrichments were calculated as described in STAR Methods. Data were analyzed using 2-way ANOVA followed by a Sidak’s multi-comparison post hoc test.

Specific subsets of CD4⁺ T cells encompass most productively SIV-infected cells in blood and lymphoid tissues

To identify shared phenotypic characteristics of productively infected cells across different compartments, we performed an unsupervised analysis and generated a two-dimensional UMAP (uniform manifold approximation and projection) plot (Figure 5A, left). In parallel, we performed a clustering analysis using the Phenograph algorithm, which allowed us to identify 21 distinct clusters in the CD4⁺ T cell compartment (Figure 5B). By overlaying the p27⁺ population on the UMAP of total CD4⁺ T cells, we observed a distinctive aggregation of these infected cells in specific regions (Figure 5A, right). Notably, p27⁺ cells were frequently observed in clusters 8, 12, and 17 (Figure 5C). Cells in cluster 8 displayed a memory phenotype and were characterized by high expression levels of PD-1, CTLA4, and CD69 and lack of Ki67 expression. This cluster was, therefore, named “exhausted memory.” Cells in cluster 12 expressed high levels of CD69, Ki67, PD-1, CTLA-4, CD127, and the integrin VLA-4. This cluster was, therefore, renamed “proliferative memory.” Cluster 17 was almost exclusively made of T_EM cells expressing high levels of PD-1 and VLA-4. This cluster was renamed “T_EM PD-1⁺.”

Figure 5. Unsupervised phenotypic analysis of productively SIV-infected cells.

(A) Two-dimensional reduction UMAP generated after pooling CD4⁺ T cells from all compartments and from all animals. The plot on the left shows the distribution of all clusters in the UMAP; the plot on the right shows the distribution of p27⁺ (in red) overlaid on uninfected cells (in gray).

(B) Heatmap showing the frequency of expression of all markers in each cluster identified by the Phenograph algorithm.

(C) CD4⁺ T cell clusters matching the spatial localization of the majority of p27⁺ events.

(D) Fold enrichments in p27⁺ cells in the three clusters of interest, shown for each compartment. The dotted line indicates no enrichment. Fold enrichments were calculated as described in STAR Methods.

(E) Contribution of each cluster to the total pool of p27⁺ cells. The three clusters of interest are represented in orange, pink, and blue, and all other clusters are shown in gray.

(F) Contribution of each cluster to the pool of CD4⁺ T cells (top row) and to the pool of p27⁺ cells (bottom row) in each compartment. The three clusters of interest are represented in orange, pink, and blue, and all other clusters are shown in gray.

We then calculated the fold enrichment in p27⁺ cells in each cluster as well as its contribution to the total pool of productively infected cells. In all 4 lymphoid tissues, the exhausted memory subset was only slightly enriched in productively infected cells (Figure 5D). Overall, this subset contributed 15% to all productively infected cells (Figure 5E). The proliferative memory cluster was highly enriched in SIV-infected cells across all 5 compartments (3.4- to 8.1-fold enrichment, Figure 5D). Importantly, 43.7% of all productively infected cells belonged to this cluster (Figure 5E). The T_EM PD-1⁺ cluster was also highly enriched in p27⁺ cells in all 5 compartments, particularly in the inguinal lymph node (fold enrichment of 11.8, Figure 5D). Nonetheless, because cells displaying this phenotype were relatively rare, the contribution of this cluster to the total pool of infected cells was modest (3.5%, Figure 5E).

The three clusters had similar contributions to the total pool of CD4⁺ T cells in the different compartments (Figure 5F, top). In contrast, their relative contributions to the pool of productively infected cells varied in a tissue-specific manner. The contribution of the proliferative memory cluster to the pool of p27⁺ cells was higher in blood and the spleen (38.5% and 57.8%, respectively) than in the other lymphoid tissues (mean of 21.1%). The contribution of the T_EM PD-1⁺ cluster to the pool of infected cells was highest in the inguinal lymph node and lowest in the spleen (16.7% and 4.3%, respectively). The contribution of the exhausted memory subset to the pool of infected cells was similar across all tissues (range: 13.9%–19.4%; Figure 5F).

Altogether, our results indicate that only three subsets of cells, expressing PD-1, CTLA-4, and a markedly memory phenotype, encompass more than 60% of productively SIV-infected cells in chronically infected rhesus macaques, indicating that the phenotypic signatures of infected cells are relatively conserved across multiple anatomic compartments.

DISCUSSION

In this study, we developed the SIV-Flow assay, a novel approach that enables the characterization of individual productively SIV- and SHIV-infected cells, through dual staining of the viral protein p27. Similar to what we previously reported for the detection of HIV Gag/p24,³⁵ the combination of two monoclonal antibodies significantly enhanced the detection accuracy of SIV-infected cells. As expected, frequencies of cells expressing Gag were consistently lower than those harboring SIV DNA (24-fold), suggesting that only a small fraction of cells harboring SIV genomes produce SIV proteins, similar to what was previously described in HIV (39-fold).³⁵ Nonetheless, we observed strong correlations between SIV-Flow frequencies and conventional virologic measures, including SIV DNA and cell-associated RNA. Interestingly, these associations were tissue dependent: stronger associations between p27⁺ cell frequencies and plasma viral loads were observed in the axillary and inguinal lymph nodes, suggesting that these two compartments may be major producers of viral particles during untreated SIV infection.

Near-full-length sequencing of SIV genomes from p27⁺ cells demonstrated that the majority (~90%) of these cells harbor genetically intact proviruses, suggesting that in the setting of untreated infection, most translationally active cells contribute to viral replication. These findings support the use of SIV-Flow to identify cells likely contributing to productive infection in vivo.

Our phenotypic analysis of p27⁺ cells revealed that, in all compartments, most infected cells exhibited a memory phenotype and were dominated by infection of cells with a T_EM phenotype. This finding confirms and extends previous studies reporting the greater susceptibility of memory CD4⁺ T cells to SIV infection compared with T_N cells.^12,49 It also aligns with several studies demonstrating that T_EM cells are the major producers of virions in people with HIV who are not receiving ART.^1,35 Of note, we also observed an enrichment of SIV-infected cells within the T_SCM subset, characterized by heightened self-renewal and proliferation capacities, which may contribute to viral persistence during ART.⁵⁰

Our SIV data largely corroborate previous HIV studies demonstrating the preferential infection of cells expressing specific markers. As in HIV,^1,35 CD4⁺ T cells expressing activation and proliferation markers, such as CD69 and Ki67, were significantly enriched in SIV across all analyzed compartments. We also confirmed PD-1 as a cellular marker of SIV-infected cells in blood and all lymphoid tissues of non-human primates.^44,51,52 As we previously reported for HIV,³⁵ productively SIV-infected cells also expressed higher levels of the integrin VLA-4 than their uninfected counterparts.³³

The integrin α4β7 is a well-established determinant of SIV pathogenesis that critically shapes viral dissemination and reservoir establishment. It defines CD4⁺ T cell subsets that are highly susceptible to infection⁵³ and directs their homing to the gut-associated lymphoid tissue, a primary site for early viral replication and CD4⁺ T cell depletion.^54,55 Consistent with this role, intervention studies in macaque models have demonstrated that anti-α4β7 monoclonal antibodies protect gut CD4⁺ T cells, reduce mucosal viral loads, and alter disease progression.^56,57 This strategy appears to limit the size and distribution of the viral reservoir when administered early,^56,58 though its effects during chronic infection may be more complex, potentially redistributing rather than clearing persistent virus.⁵⁹ Despite this evidence, our data did not reveal a higher expression of α4β7 at the surface of productively SIV-infected cells compared with their uninfected counterparts. A critical consideration is that our sampling strategy did not include the gut, the compartment where α4β7-expressing cells are predominantly localized and likely infected. Consequently, our analysis may be biased, potentially underestimating the association between α4β7 expression and SIV infection in vivo.

Our data revealed that several markers, including PD-1, VLA-4, Ki67, and CD69, were preferentially expressed by SIV-infected cells. However, no individual marker demonstrated sufficient specificity to uniquely identify infected cells. Rather than defining a singular phenotypic signature, these findings support the notion that productive SIV infection is associated with a distinct combination of activation and differentiation features.

We employed unsupervised analysis combining UMAP plotting with Phenograph clustering to investigate the relative contributions of different CD4⁺ T cell subsets to the pool of productively SIV-infected cells. We identified three subpopulations of memory cells of particular interest that we renamed exhausted memory, proliferating memory, and T_EM PD-1⁺ CD4 T cells.

Consistent with an exhaustion phenotype, the exhausted memory subset expressed high levels of PD-1 and CTLA-4, together with low levels of Ki67. This subpopulation was found in similar percentages in the different compartments analyzed. It encompassed approximatively 15% of all p27⁺ cells, and its contribution to the pool of infected cells was similar across all tissues (13.9%–19.4%). This exhausted memory cluster was minimally enriched in SIV-infected cells, suggesting that it was not particularly prone to viral replication. However, because CD4⁺ T cells expressing exhaustion markers have been described as a preferential reservoir during ART in both HIV and SIV infections,^44,52,60,61 viral latency may be established in a fraction of the cells displaying this exhausted memory phenotype.

The T_EM PD-1⁺ CD4 T subset was characterized by elevated expressions of PD-1 and VLA-4. Previous studies have shown that T_EM cells, particularly those expressing PD-1 and VLA-4, are highly susceptible to SIV replication and contribute to viral persistence.^21,27,28 Of note, cells displaying this phenotype were relatively rare in all compartments (1.4%–2.2%). However, these rare cells were highly enriched in SIV-infected cells in all compartments, encompassing up to 16.7% of the infected cells in the inguinal lymph node.

The third cluster (proliferative memory) was composed of activated CD4⁺ T cells, as shown by the elevated expression of CD69 and Ki67. Because CD69 also marks tissue-resident CD4⁺ T cells,⁶² it is possible that this subset encompasses tissue-resident CD4⁺ T cells that are dividing in lymphoid tissues. Of note, this cluster expressed high levels of CD127, a marker associated with T cell survival and memory potential, suggesting that it may contribute to the establishment of a long-lived reservoir. This cluster also shared several markers with the two aforementioned populations, including PD-1, CTLA4, and VLA-4. Although these cells were homogeneously distributed among the different compartments (5.5%–7.2%), their contribution to the pool of SIV-infected cells differed: these cells encompassed 57.8% of the p27⁺ cells in the spleen and 38.5% in blood. Remarkably, this cluster similarly contributed to the pool of infected cells in the 3 different lymph nodes (20.8%–21.7%). Overall, the proliferative memory cluster was a major contributor to the pool of infected cells, encompassing 43.7% of all p27⁺ cells.

In conclusion, our study demonstrates that the SIV-Flow assay can be used to detect and characterize SIV- and SHIV-infected CD4⁺ T cells across multiple tissues, advancing our understanding of CD4⁺ T lymphocyte phenotypes, viral dynamics, and host immune responses. By enabling detailed exploration of cellular interactions within tissue-specific microenvironments, SIV-Flow provides a valuable framework for developing targeted and effective therapeutic interventions to control SIV infection. Our findings underscore the heterogeneous nature of CD4⁺ T cell subsets in the context of SIV infection, revealing distinct populations that critically contribute to viral production. Our results highlight the pivotal role of lymphoid-resident CD4⁺ T cells during chronic SIV infection. Future studies should aim to elucidate the mechanisms governing interactions among these diverse cellular populations. Moreover, longitudinal analyses of these subsets throughout the course of infection, including samples from virally suppressed animals, will deepen our understanding of viral replication and persistence, thereby guiding the development of more effective therapeutic strategies.

Limitations of the study

Our flow cytometry-based phenotyping was constrained by the number of parameters included in the cytometry panel. Because some key markers such as CXCR5 were not included in our panel, we cannot exclude the possibility that a fraction of the proliferative memory or T_EM PD-1⁺ clusters might have overlapped with Tfh or Th17 cells. Future studies incorporating markers of additional CD4⁺ T cell subsets are needed to determine the precise contribution of these subsets to the pool of productively infected cells. Similarly, in future studies, it will be important to include markers expressed by subsets of myeloid cells (monocytes, macrophages, and dendritic cells) to measure their contributions to the pool of SIV-infected cells, particularly in tissues that were not analyzed in this study, including gut and brain samples. An inherent limitation of the SIV-Flow assay is its inability to provide spatial information regarding the tissue localization of SIV-infected cells.

RESOURCE AVAILABILITY

Lead contact

Requests for further information and resources should be directed to and will be fulfilled by the lead contact, Nicolas Chomont (nicolas.chomont@umontreal.ca).

Materials availability

This study did not generate new unique reagents.

Data and code availability

• All data reported in this paper will be shared by the lead contact upon request.

• This paper does not report original code.

• Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request

STAR★METHODS

EXPERIMENTAL MODEL AND STUDY PARTICIPANT DETAILS

This study was undertaken with the approval of the Oregon National Primate Research Center’s Animal Care and Use Committee (project number IP00001452), under the standards of the US National Institutes of Health Guide for the Care and Use of Laboratory Animals. For these studies a total of 8 rhesus macaques (Macaca mulatta) of Indian ancestry and free of simian retrovirus type D (SRV), simian T-lymphotropic virus type 1 (STLV-1), SIV and macacine herpesvirus 1 (herpesvirus B) were included. The animals were socially housed in pairs or groups. Animals 37170 and 37172 were infected intravenously with a single dose of the genetically barcoded SIVmac293M⁶³ at 10,000 IU. At week 8 post infection, animals were sacrificed to allow extensive collection of tissue samples. Animals 34675, 33447, 33122, 33921, 35026, 34254 were infected by low dose intra-rectal inoculation with a sequence-tagged variant of SIVmac239 (“SIVmac239X″).⁶⁵ Animals 35026 and 34254 were treated with 20 mg/kg GSK-360 at weeks 10 and at week 12, which had not impact on plasma viral loads (data not shown). Animals were sacrificed from week 14 to week 26 for extensive tissue collection (see Table 1 for exact times). Tissues were taken from the lymph nodes (axillary, mesenteric, and inguinal), and spleen. Tissues were homogenized, strained through a 70 μM filter to mechanically separate lymphocytes. From blood samples, mononuclear cells were isolated through density gradient centrifugation.

METHOD DETAILS

Cell culture

CD4⁺ T cells were enriched from mononuclear cells derived from blood, lymph nodes and spleen using the EasySep Non-Human Primate CD4⁺ T cell Isolation Kit (StemCell Technology, Cat#19582) following the manufacturer’s instructions. The purity of the isolated cells typically exceeded 90%. Enriched CD4⁺ T cells were resuspended at a concentration of 1–2×10⁶ cells/mL in RPMI supplemented with 10% Fetal Bovine Serum (FBS, PAA, A15–752). Penicillin-Streptomycin (Life Technologies, 15070–063), antiretroviral drugs (200 nM raltegravir, 200 nM 3TC, 5 μM Ritonavir) and a pan-caspase inhibitor (QVD-Oph, 10 nM) (BD, Cat#563828) were added to the culture medium. Cells were rested for 24 h at 37°C until use.

The productively SIV-infected HUT 78 cell line (HIV Program, Cat#ARP-160) and the SIVmac316-infected CEMx174 cells-Clone 3D8 (HIV Program, Cat#ARP-13239) were thawed and maintained in RPMI medium supplemented with 10% FBS and 100 U/mL penicillin-Streptomycin (Life Technologies, 15070–063) until use. 3D8 cells were counted and aliquoted in pellets of 1 × 10⁶ cells and used as PCR standards. Both cell lines were obtained from the NIH HIV reagent program (BEI resources) and were tested negative for mycoplasma, bacterial and fungi.

In vitro infection of human CD4⁺ T cells

To increase infectivity in primary human CD4⁺ T cells, vesicular stomatitis virus G (VSVG)-pseudotyped SHIV.BG505.T332N.375Y virus was produced in 293T cells, titrated and used for infection.^64,66 Briefly, HEK293T cells were pre-seeded in 10cm dishes one day before, 7.5 μg of proviral vectors and 1 μg of VSVG-encoding plasmid were co-transfected by standard calcium phosphate transfection. 48 h after transfection, cell supernatants were harvested, clarified by low-speed centrifugation (5 min at 1500 rpm) and concentrated by ultracentrifugation for 1h at 4°C at 100,605 × g over a 20% sucrose cushion. Pelleted viruses were resuspended in fresh RPMI, and aliquots were stored at −80°C until use. Virus was then used to infect activated primary human CD4 T cells from HIV negative donors by spin infection at 800 × g for 1 h in 96-well plates at 25°C.

SIV-flow cytometry staining

After 24h of resting, CD4⁺ T cells were resuspended in phosphate-buffered saline (PBS), stained with the FVS780 Live/Dead reagent for 10 min at room temperature, and washed with PBS supplemented with 10% FBS. Cells were then stained with antibodies specific for extracellular markers in PBS supplemented with 4% human serum (Atlanta Biologicals, Cat#540110) for 20 min at RT. Cells were washed with PBS supplemented with 4% human serum and fixed and permeabilized step for 45 min using the FoxP3 Transcription Factor Staining Buffer Set (eBioscience, cat#00–5523-00), according to the manufacturer’s instructions. Cells were then washed with the permeabilization buffer (ThermoFisher, cat#00–8333-56) and stained with the two anti-p27 antibodies (55–2F12 and 4B7) for an additional 45 min at RT in the FoxP3 Buffer. Cells were then washed with permeabilization buffer and resuspended in PBS before flow cytometry acquisition (BD FACS Symphony A5 cytometer). In all experiments, CD4⁺ T cells from an SIV-uninfected control animal were included to establish the positivity threshold.

The anti-p27 antibodies 55–2F12 (ARP-1610) and 4B7 (ARP-13443) were sourced from the HIV Reagent Program (NIH) and labeled using the Mix-n-Stain CF 568 Antibody Labeling Kit (Cat#MX568S100) and the Mix-n-Stain CF 647 Antibody Labeling Kit (Cat#MX647S100), respectively. Both antibodies were subsequently titrated. CD95-BV786 (clone: DX2, cat#740991), CD127-BV750 (clone: HIL-7R-M21, cat#747089), CD279-BV711 (PD-1) (clone: EH12.1, cat#564017), CD29-BV605 (Beta1) (clone: MAR4, cat#743784), CD45RA-BV570 (clone: 5H9, cat#624298), Ki67-BV480 (clone:B56, cat#566109), CD197-BV421 (CCR7) (clone: 2-L1-A, cat#566743), CTLA-4-BB700 (clone: BNI3, cat#566901), CD49d-BB515 (Alpha4) (clone: 9F10, cat#564593), CD27-PE-Cy7 (clone: M-T271, cat#560609), CD8-R718 (clone: RPA-T8, cat#567056), CD103-BUV805 (clone: Ber-ACT8, cat#748501), beta7-BUV737 (clone: FIB504, cat#749664), CD69-BUV661 (clone: FN50, cat#750213), CD28-BUV615 (clone: CD28.2, cat#751477), CD3-BUV496 (clone: SP34–2, cat#741183), CD4-BUV395 (clone: L200, cat#564107) were purchased from BD.

Cell sorting of single p27+, double p27+, and p27-cells for SIV DNA quantification

Frequencies of p27 double-positive cells were determined by flow cytometry within the viable cells gate. The four population: p27 double-positive (55–2F12+ and 4B7+), p27 single-positive (55–2F12+ or 4B7+) and p27 double-negative cells 55–2F12-, 4B7-were sorted using a BD FACS ARIA III into a 96-well PCR plates containing 7.6 μL of DirectPCR Lysis Reagent (Viagen Biotech, Cat#301-C) and 0.4 μL of 10 mg/mL proteinase K (Wisent, Cat#25530–015). The PCR plates were incubated overnight at 55°C to facilitate cell lysis, followed by a 10-min incubation at 95°C to inactivate the proteinase K. The resulting cell lysates were then used for quantification of SIV DNA.

Amplification and sequencing of near full length SIV genomes

Near full-length amplification of proviral sequences was performed at an endpoint dilution of the extracted DNA where <30% of the reactions were positive, obtaining a total of 28 SIV proviral sequences. The protocol was adapted from previous studies.^1,4,39 SIV proviruses were pre-amplified using an Invitrogen Platinum SuperFi II MasterMix (ThermoFisher cat.12358050) with two outer primers (SIV U3 Fw1: 5′-CAG AAG AGT TTG GAA GCA AGT C-3’; SIV U3 Rv1: 5′-ACA TAT GCC TCA TAA GTG TAG G-3′). The first-round PCR cycle conditions were as follows: a denaturation step of 30 s at 98°C and 25 cycles of amplification (98°C for 10 s, 60°C for 10 s, 72°C for 5 min), followed by a final elongation step at 72°C for 5 min. Diluted pre-amplification products were used in a second round of amplification, using inner primers (SIV U3 Fw2: 5′-TTA GAA GAA GGC TAA CCG CAA G-3’; SIV U3 Rv2: 5′-TTG GAT CAA ACT TCC ATG CTA G-3′). The second round of PCR was carried out as follows: denaturation step of 30 s at 98°C and 30 cycles of amplification (98°C for 10 s, 60°C for 10 s, 72°C for 5 min), followed by a final elongation step at 72°C for 5 min 3D8 extracted DNA was used as positive controls for PCR amplification. Proviral amplification was confirmed by running a fraction of the PCR products on an 0.8% agarose gel.

Library preparation and nanopore sequencing

The amplified amplicons were quantified with a Nanodrop Spectrophotometer 1000 prior to barcoding. After barcoding, all amplicons were pooled and purified using AMPure XP beads (Beckman Coulter). The concentration and purity of the pooled product were then measured using the a Nanodrop Spectrophotometer 1000. The prepared DNA library was loaded into the flow cell and sequenced on the Oxford Nanopore Technology (ONT) GridION using the Epi2me software (v2.5.2).

Integrity and clonality analysis

To analyze the intactness and clonality of SIV proviruses, we developed an R pipeline called “IntegrityAlgorithm,” publicly available on GitHub (https://github.com/alemi055/IntegrityAlgorithm), that automates the analysis pipeline previously described.^1,4,39 Proviral sequences containing (in this order) inversions, hypermutations, large internal deletions, stop codons and/or frameshifts, defects in the Ψ locus, or small internal deletions in a coding region, were considered defective. Inversions were manually detected at the alignment step, since only the reverse complement of inversions can be aligned with HXB2. Briefly, the algorithm integrates the results of HIVDatabase QCTool (https://www.hiv.lanl.gov/content/sequence/QC/index.html) to detect hypermutations, HIVDatabase Gene Cutter to detect stop codons (https://www.hiv.lanl.gov/content/sequence/GENE_CUTTER/cutter.html), and ProSeq-IT (https://psd.cancer.gov/tools/pvs_annot.php) to detect large internal deletions (sequences shorter than 8,800 bp, excluding the primer’s regions), frameshifts, small internal deletions, and defects in the Ψ locus (MSD point mutation; stem loop 2 deletion; packaging signal deletion). Defects in the Ψ locus were manually confirmed by looking at the alignment with the reference HXB2 sequence. Additionally, we used the IntegrityAlgorithm to identify clonal sequences, defined as proviral amplicons 100% identical to each other. To do so, each proviral sequence was aligned to each other using MAFFT in the R package ips (v0.0.11) (Heibl et al., 2019). Clones were confirmed using the Geneious Prime diversity tool.

DNA and RNA extractions

DNA and RNA were simultaneously extracted using the AllPrep DNA/RNA/miRNA Universal Kit (Qiagen), according to the manufacturer’s instructions.

Quantification of SIV DNA and cell-associated SIV RNA

Genomic DNA and total RNA were simultaneously extracted using the AllPrep DNA/RNA Universal kit (Qiagen, cat# 80224), according to the manufacturer’s instructions. SIV DNA and cell-associated SIV RNA were measured by ultrasensitive nested qPCR with primers annealing the LTR/gag region. SIVLLTR Fw: 5’ – ATG CCA CGT AAG CGA AAC TGG CAG ATT GAG CCC TGG GAG – 3’; SIVgag Rev: 5’ – TGC TGC CCA TAC TAC ATG CTT C – 3’; LambdaT: 5’ – ATG CCA CGT AAG CGA AAC T – 3’; SIVLTR Rev2: 5’ – CTT TAA GCA AGC AAG CGT GGA G – 3’; 5’ –/56-FAM/GCA GGTAGA/ZEN/GCC TGG GTG TTC CCTGC/3IABkFQ/– 3’.^67,68 Primers specific for the CD3 gene were added to the DNA PCR reaction to precisely quantify the cell input: HCD3 out 5’: 5’ – ACT GAC ATG GAA CAG GGG AAG – 3’; HCD3 out 3’: 5’ – CCA GCT CTG AAG TAG GGA ACA TAT – 3’; HCD3 in 5’: 5′-GGC TAT CAT TCT TCT TCA AGG T-3’; Mamu CD3 in 3’: 5’ – TAA GAT GGC GGT AAC AGG GT – 3’; MamuCD3Zen 5’ –/56-FAM/AGC AGA GAA/ZEN/CAG TTA AGA GGC TCC AT/3IABkFQ/– 3’. SIV DNA was amplified using the Taq DNA polymerase (Invitrogen, cat# 10342020) as follows: the first-round PCR cycle conditions were as follows: a denaturation step of 8 min at 95°C, 12 cycles of amplification (95°C for 1 min, 55°C for 40 s, 72°C for 1 min), followed by a final elongation step at 72°C for 15 min. HIV RNA was also reverse-transcribed, 30 min at 50°C using the Superscript IV one-step RT-PCR System (Invitrogen, cat# 12574026), and pre-amplified as follow: 16 cycles of amplification (95°C for 15 s, 55°C for 30 s, 68°C for 1 min), followed by a final elongation step at 68°C for 5 min. For the second PCR, the pre-amplified SIV DNA and cDNA were amplified by real-time PCR on a Rotor-Gene Q instrument, using the Quantinova Probe Master Mix (Qiagen, cat# 208252). The second rounds of PCR were carried out as follows: denaturation step (95°C for 4 min), followed by 40 cycles of amplification (95°C for 3 s, 60°C for 10 s).

The absolute number of SIV DNA copies per PCR reaction was determined by using serial dilutions of lysed 3D8 cells as a standard curve. Copy numbers of cell-associated SHIV RNA were calculated by using serial dilutions of in vitro transcribed SIV RNA as standard. All measures were performed in triplicate wells. Results were expressed as numbers of SHIV copies per million cells.

Linearity of the assay

SIV-infected HUT 78 cells were resuspended in PBS and stained with FVS780 Live/Dead reagent for 10 min at RT. After washing with PBS, cells were stained with an anti-gp120 (Clone KK45) monoclonal antibody coupled to FITC for 30 min at 37°C. Single sorted SIV-infected HUT 78 cells expressing gp120 were spiked at different ratios into PBMCs cultures derived from a SIV uninfected animal. The linearity of the assay was determined by comparing the predicted frequency of infected cells to the measured frequencies of p27+ cells obtained by single p27 staining with 55–2F12 or 4B7 or by dual staining p27 55 and 2F12/4B7.

Unsupervised analysis

Data was down sampled using the down sample plugin of the FlowJo software to homogenize the contribution of each tissue and each animal to the total pool of events included in the analysis. Using the CytoNorm plugin of FlowJo, down sampled data was normalized to reduce differences in the median fluorescence intensity of each marker between animals. Normalized data was pooled and used to build a preliminary UMAP in FlowJo, after excluding the markers CD4, CD8, CD3, live/dead, and both p27 antibodies. When the events derived from a given animal clustered as a single and separated population in the UMAP, this animal was removed from the non-supervised analysis. A definitive UMAP was built as described and the phenograph plugin was used to identify cellular clusters. Each cluster was then analyzed individually in FlowJo to obtain the expression levels of each marker.

QUANTIFICATION AND STATISTICAL ANALYSIS

Data were analyzed and represented using Graphpad Prism version 9.1.0. Results were represented as median or mean values, with interquartile range or minimum and maximum values, as indicated in the figure legends. Fold enrichments were calculated as the ratio of a given subset’s contribution to the p27+ cell pool relative to its contribution to the total CD4⁺ T cell compartment. Correlations were determined using nonparametric Spearman’s test. For group comparisons, non-parametric Wilcoxon matched pairs signed rank or Mann-Whitney or Fisher’s exact tests were used. Multiple comparison was performed using Dunn’s or Sidak’s post-test. p values of less or equal to 0.05 were considered statistically significant.

Supplementary Material

SUPPLEMENTAL INFORMATION

Supplemental information can be found online at https://doi.org/10.1016/j.celrep.2026.117826.

KEY RESOURCES TABLE.

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
CD95-BV786 (clone DX2)	BD Biosciences	cat#740991 AB_2740614
CD127-BV750 (clone HIL-7R-M21)	BD Biosciences	cat#747089 AB_2871843
CD279-BV711 (PD-1) (clone EH12.1)	BD Biosciences	cat#564017 AB_2640100
CD29-BV605 (b1) (clone MAR4)	BD Biosciences	cat#743784 AB_2741752
CD45RA-BV570 (clone 5H9)	BD Biosciences	cat#624298 AB_2563813
Ki67-BV480 (clone B56)	BD Biosciences	cat#566109 AB_2739511
CD197-BV421 (CCR7) (clone 2-L1-A)	BD Biosciences	cat#566743 AB_2869843
CTLA-4-BB700 (clone BNI3)	BD Biosciences	cat#566901 AB_2869943
CD49d-BB515 (a4) (clone 9F10)	BD Biosciences	cat#564593 AB_2738859
CD27-PE-Cy7 (clone M-T271)	BD Biosciences	cat#560609 AB_1727456
CD8-R718 (clone: RPA-T8)	BD Biosciences	cat#567056 AB_2916407
β7-BUV737 (clone FIB504)	BD Biosciences	cat#749664 AB_2873928
CD69-BUV661 (clone FN50)	BD Biosciences	cat#750213 AB_2874414
CD28-BUV615 (clone CD28.2)	BD Biosciences	cat#751477 AB_2875473
CD3-BUV496 (clone SP34-2)	BD Biosciences	cat#741183 AB_2870750
CD4-BUV395 (clone L200)	BD Biosciences	cat#564107 AB_2738596
SIV p27 antibody (clone 55-2F12)	BEI Resources	cat#ARP-1610 AB_2905607
SIV p27 antibody (clone 4B7)	BEI Resources	cat#ARP-13443 N/A
SIV gp120 antibody (clone KK45)	BEI Resources	cat#ARP-2318 N/A
Bacterial and virus strains
SIVmac239M	Fennessey et al.63	NCBITaxon_239221
SHIV-BG505	Dr. George Shaw et al.64	NCBITaxon_57667
Biological samples
Rhesus macaques derived blood, lymph nodes biopsies (inguinal, axillary and mesenteric) and spleen biopsies	Oregon National Primate Research Center	N/A
Chemicals, peptides, and recombinant proteins
Mix-n-Stain™ CF™ 568 Antibody Labeling Kit	Sigma-Aldrich	Cat#MX568S100
Mix-n-Stain™ CF™ 647 Antibody Labeling Kit	Sigma-Aldrich	Cat#MX647S100
Experimental models: Cell lines
SIVmac316-infected CEMx174 cells (Clone 3D8)	BEI Resources	cat#ARP-13239 CVCL_A0CL
SIVmac251-Infected HUT 78 Cells	BEI Resources	cat#ARP-160 CVCL_X611
Experimental models: Organisms/strains
Rhesus macaques (Macaca mulatta) of Indian ancestry	Oregon National Primate Research Center	NCBITaxon_9544
Oligonucleotides
SIVLLTR Fw: 5’ – ATG CCA CGT AAG CGA AAC TGG CAG ATT GAG CCC TGG GAG – 3′	Integrated DNA technologies	N/A
SIVgag Rev: 5’ – TGC TGC CCA TAC TAC ATG CTT C – 3′	Integrated DNA technologies	N/A
LambdaT: 5’ – ATG CCA CGT AAG CGA AAC T – 3′	Integrated DNA technologies	N/A
SIVLTR Rev2: 5’ – CTT TAA GCA AGC AAG CGT GGA G – 3′	Integrated DNA technologies	N/A
5’ –/56-FAM/GCA GGTAGA/ZEN/GCC TGG GTG TTC CCTGC/3IABkFQ/– 3′	Integrated DNA technologies	N/A
HCD3 out 5’: 5’ – ACT GAC ATG GAA CAG GGG AAG – 3′	Integrated DNA technologies	N/A
HCD3 out 3’: 5’ – CCA GCT CTG AAG TAG GGA ACA TAT – 3′	Integrated DNA technologies	N/A
HCD3 in 5’: 5′-GGC TAT CAT TCT TCT TCA AGG T-3′	Integrated DNA technologies	N/A
Mamu CD3 in 3’: 5’ – TAA GAT GGC GGT AAC AGG GT – 3′	Integrated DNA technologies	N/A
MamuCD3Zen 5’ –/56-FAM/AGC AGA GAA/ZEN/CAG TTA AGA GGC TCC AT/3IABkFQ/– 3′	Integrated DNA technologies	N/A
Software and algorithms
GraphPad Prism v9.3.0	https://www.graphpad.com/	SCR_002798
FlowJo	https://www.flowjo.com/	SCR_008520
UMAP	https://www.flowjo.com/exchange/plugin/umap	N/A
Phenograph	https://www.flowjo.com/exchange/plugin/phenograph	N/A
Cytonorm	https://www.flowjo.com/exchange/plugin/cytonorm	N/A
DownSample	https://www.flowjo.com/exchange/plugin/downsample	N/A

Highlights.

• SIV-Flow detects individual productively infected cells, with high sensitivity and specificity

• Productively infected cells are found at comparable frequencies across lymphoid tissues

• Effector memory CD4 T cells constitute the major source of viral production in vivo

• A proliferative PD-1⁺ memory subset accounts for nearly half of infected cells