HIV-1 DNA and Immune Activation Levels Differ for Long-Lived T-Cells in Lymph Nodes, Compared with Peripheral Blood, during Antiretroviral Therapy

ABSTRACT

Elucidating the mechanisms underlying the persistence and location of the HIV reservoir is critical for developing cure interventions. While it has been shown that levels of T-cell activation and the size of the HIV reservoir are greater in rectal tissue and lymph nodes (LN) than in blood, the relative contributions of T-cell subsets to this anatomic difference are unknown. We measured and compared HIV-1 DNA content, expression of the T-cell activation markers CD38 and HLA-DR, and expression of the exhaustion markers programmed cell death protein 1 (PD-1) and T-cell immunoreceptor with immunoglobulin and immunoreceptor tyrosine-based inhibitory motif domains (TIGIT) in naive, central memory (CM), transitional memory (TM), and effector memory (EM) CD4⁺ and CD8⁺ T-cells in paired blood and LN samples among 14 people with HIV who were receiving antiretroviral therapy. HIV-1 DNA levels, T-cell immune activation, and TIGIT expression were higher in LN than in blood, especially in CM and TM CD4⁺ T-cell subsets. Immune activation was significantly higher in all CD8⁺ T-cell subsets, and memory CD8⁺ T-cell subsets from LN had higher levels of PD-1 expression, compared with blood, while TIGIT expression levels were significantly lower in TM CD8⁺ T-cells. The differences seen in CM and TM CD4⁺ T-cell subsets were more pronounced among participants with CD4⁺ T-cell counts of <500 cells/μL within 2 years after antiretroviral therapy initiation, thus highlighting increased residual dysregulation in LN as a distinguishing feature of and a potential mechanism for individuals with suboptimal CD4⁺ T-cell recovery during antiretroviral therapy.

IMPORTANCE This study provides new insights into the contributions of different CD4⁺ and CD8⁺ T-cell subsets to the anatomic differences between LN and blood in individuals with HIV who have optimal versus suboptimal CD4⁺ T-cell recovery. To our knowledge, this is the first study comparing paired LN and blood CD4⁺ and CD8⁺ T-cell differentiation subsets, as well as those subsets in immunological responders versus immunological suboptimal responders.

INTRODUCTION

More than 2 decades since the discovery of human immunodeficiency virus (HIV) persistence in resting CD4⁺ T-cells, the complexity of the HIV-1 reservoir continues to be a major challenge to curing infection. Characterizing the mechanisms underlying the establishment, persistence, and location of the HIV latent reservoir is a critical step for developing cure interventions.

Reservoir levels in cells from anatomic sites such as lymph nodes (LN) may not be equivalent to those from peripheral blood mononuclear cells (PBMC), which are frequently measured in clinical trials aimed at HIV remission. In fact, nonhuman primate (NHP) studies have shown that lymphoid tissue (including gut, LN, and spleen) contains most of the simian immunodeficiency virus (SIV) reservoir before and during antiretroviral therapy (ART) (1). The persistence of HIV in lymphoid tissue has been attributed to a variety of mechanisms, including infection of T follicular helper cells and long-lived CD4⁺ T-cells such as central memory (CM) CD4⁺ T-cells, decreased penetration of ART, and B cell follicle sanctuary sites not highly accessible to effector CD8⁺ T-cells (1–3). Furthermore, work by Buggert et al. revealed that effector memory (EM) CD8⁺ T-cell subsets that express high levels of cytolytic molecules are retained in the intravascular circulation and are rarely detected in tissue or thoracic duct lymph during homeostasis (4). These differences were also seen in HIV-specific CD8⁺ T-cells during chronic HIV infection (4, 5). Moreover, Khoury et al. showed that total CD4⁺ T-cell activation and HIV reservoir size are greater in rectal tissue and LN than in blood, but the relative contributions of CD4⁺ T-cell subsets to potential anatomic differences are unknown (6).

Memory CD4⁺ T-cells, including CM, transitional memory (TM), and EM T-cells, are key cellular reservoirs for HIV-1 infection (2, 7). CM CD4⁺ T-cells contribute substantially to HIV persistence, given that this subset can survive for decades and represents a long-lasting reservoir in immunological responders (IR) receiving ART (2, 8). Our group recently showed that maintenance of HIV-1 infection was increased despite several years of ART for immunological suboptimal responders (ISR), compared with IR, with the largest difference being observed in the TM and long-lived CM CD4⁺ T-cell subsets due to a greater abundance of immune checkpoint molecules and expression of homeostatic cytokines (9).

Here, we sought to identify the relationship between blood and LN HIV-1 DNA content in key CD4⁺ T-cell subsets, i.e., naive, CM, TM, and EM cells, among a population of IR and ISR receiving ART. Additionally, we determined whether activation markers and immune checkpoint molecules on CD4⁺ and CD8⁺ T-cell subsets were associated with these differences.

RESULTS

Greater HIV-1 DNA content and immune activation in long-lived CD4⁺ T-cells from LN, compared with blood, with the difference in immune activation observed mainly in ISR.

We measured total HIV-1 DNA in CD4⁺ T-cell subsets isolated from LN and blood samples from 14 people with HIV (PWH) who were receiving ART. Demographic characteristics of the PWH included in this study are presented in Table 1. Samples for which <3,000 cells were assayed were excluded from the analysis. Total HIV-1 DNA was 2.0-fold, 2.2-fold, and 3.5-fold higher in LN CM (P < 0.01), TM (P = 0.02), and EM (P < 0.01) CD4⁺ T-cells, respectively, compared to blood (Fig. 1A). The difference in EM CD4⁺ T-cells remained significant in the IR subgroup (n = 5) (Fig. 1B), while ISR (n = 7) had significant differences in CM and TM CD4⁺ T-cells (Fig. 1C). There was no significant difference in the total HIV-1 DNA content for naive CD4⁺ T-cells for any of the groups (Fig. 1A to C). Next, we compared the expression of CD38 and HLA-DR to assess differences in CD4⁺ T-cell subset immune activation between LN and blood. The percentage of CD4⁺ CD38⁺ HLA-DR⁺ T-cells was significantly higher in naive (mean of 3.1% versus 1.4% [P = 0.01]), CM (mean of 14.4% versus 6.2% [P < 0.01]), and TM (mean of 15.0% versus 6.6% [P < 0.01]) CD4⁺ T-cells from LN versus blood (Fig. 1D). Notably, this difference was observed exclusively in ISR (n = 7) (Fig. 1E and F), and the statistical significance was maintained in CM CD4⁺ T-cells from this group (mean of 23.2% versus 8.6% [P = 0.04]) (Fig. 1F). Of note, the percentage of CD4⁺ CD38⁺ HLA-DR⁺ T-cells was significantly higher in all memory CD4⁺ T-cell subsets from ISR, compared with IR, in both blood and LN (see Fig. S1A and B in the supplemental material). Specifically, the frequencies of CM, TM, and EM CD4⁺ T-cells expressing CD38 and HLA-DR in ISR were 2.2-, 2.2-, and 4.0-fold higher, respectively, than those in IR in peripheral blood and 4.2-, 4.5-, and 6.4-fold higher, respectively, than those in IR in LN.

TABLE 1.

Demographic characteristics

Characteristic	Data for:			P a
	All participants (n = 14)	IR (n = 7)	ISR (n = 7)
Sex (no. [%])
Female	0 (0)	0 (0)	0 (0)	1
Male	14 (100)	7 (100)	7 (100)
Mean age (mean ± SD) (yr)	42 ± 9	45 ± 9	39 ± 10	0.22
Ethnicity (no. [%])
African American	11 (79)	6 (86)	5 (71)	1
White	3 (21)	1 (14)	2 (29)
Other	0 (0)	0 (0)	0 (0)
CD4+ cell count nadir (mean ± SD) (cells/mm3)b	191 ± 146	301 ± 101	80 ± 87	0.0041
Baseline CD4+ cell count (mean ± SD) (cells/mm3)	237 ± 206	392 ± 171	83 ± 84	0.0023
CD4+/CD8+ ratio (mean ± SD)	0.88 ± 0.61	1.35 ± 0.5	0.42 ± 0.18	0.0007
Baseline plasma HIV load (mean ± SD) (log10 copies/mL)c	5.06 ± 0.64	5.07 ± 0.69	5.06 ± 0.63	0.90
Duration of ART (mean ± SD) (yr)	4 ± 3	3 ± 3	5 ± 4	0.56
ART (no. [%])d
2 NRTIs and 1 boosted PI	11 (79)	5 (71)	6 (86)	1
2 NRTIs, 1 INSTI, and 1 PK enhancer	0	0	0
2 NRTIs and 1 INSTI	0	0	0
2 NRTIs and 1 NNRTI	3 (21)	2 (29)	1 (14)

^aGroups were compared with a two-sample t test for continuous variables, an exact chi-square test for categorical variables, and a two-sample Wilcoxon rank sum test for non-normally distributed variables.

^bLowest CD4⁺ T-cell count registered prior to antiretroviral initiation.

^cPlasma HIV-1 RNA viral load data are log₁₀ transformed.

^dNRTI, nucleoside reverse transcriptase inhibitor; PI, protease inhibitor; INSTI, integrase inhibitor; NNRTI, nonnucleoside reverse transcriptase inhibitor; PK, pharmacokinetic.

FIG 1.

Paired analysis of total HIV-1 DNA and CD38 and HLA-DR coexpression in CD4⁺ T-cell subsets from LN and blood. (A to C) Comparison of total HIV-1 DNA in blood and LN in naive, CM, TM, and EM CD4⁺ T-cells in all paired participants (A), IR (B), and ISR (C). (D to F) Comparison of proportions of CD4⁺ CD38⁺ HLA-DR⁺ T-cells in blood and LN in naive, CM, TM, and EM CD4⁺ T-cells in all paired participants (D), IR (E), and ISR (F). The Wilcoxon test was performed for statistical significance. *, P ≤ 0.05; **, P ≤ 0.01.

PD-1 expression and TIGIT expression were consistently higher in CM CD4⁺ T-cells from ISR LN versus blood.

The persistence of cells harboring HIV DNA has been linked to expression of several immune checkpoint molecules, including programmed cell death protein 1 (PD-1), T-cell immunoreceptor with immunoglobulin and immunoreceptor tyrosine-based inhibitory motif domains (TIGIT), lymphocyte activation gene 3 (LAG-3), and cytotoxic T-lymphocyte antigen 4 (CTLA-4) (10–13). To assess differences in CD4⁺ T-cell subsets expressing inhibitory receptors, we compared T-cell expression of PD-1 and TIGIT in LN and blood. PD-1 expression was significantly higher in naive (mean of 6.5% versus 4.4% [P = 0.02]) and particularly CM (mean of 64.4% versus 45.7% [P < 0.01]) CD4⁺ T-cells from LN versus blood (Fig. 2A). This significant difference was observed in both IR (n = 7) and ISR (n = 7) for CM CD4⁺ T-cells (Fig. 2B and C), while differences in naive CD4⁺ T-cells remained significant only in ISR, and IR showed differences in TM CD4⁺ T-cells (Fig. 2B and C). TIGIT expression was significantly higher in CM (mean of 19.1% versus 10.9% [P < 0.01]), TM (mean of 28.5% versus 16.9% [P < 0.01]), and EM (mean of 16.5% versus 4.4% [P < 0.01]) CD4⁺ T-cells from LN versus blood (n = 13) (Fig. 2D). This difference was driven mainly by ISR (n = 6) (Fig. 2E and F).

FIG 2.

Paired analysis of PD-1 and TIGIT expression in LN and blood in CD4⁺ T-cell subsets. (A to C) Comparison of PD-1 expression in blood and LN in naive, CM, TM, and EM CD4⁺ T-cells in all paired participants (A), IR (B), and ISR (C). (D to F) Comparison of TIGIT expression in blood and LN in naive, CM, TM, and EM CD4⁺ T-cells in all paired participants (D), IR (E), and ISR (F). The Wilcoxon test was performed for statistical significance. *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001.

CD4⁺ T-cell TIGIT expression and the peripheral blood CD4⁺/CD8⁺ ratio were correlated with total HIV-1 DNA content.

Next, we assessed the correlation of HIV-1 DNA content with the expression of TIGIT and PD-1 in CD4⁺ T-cell subsets, as well as with the blood CD4⁺/CD8⁺ ratio. To increase the power of this correlation analysis in blood, we included 18 additional participants who underwent large volume blood draws (for total n = 32) but from whom LN samples were not collected; therefore, they were not included in the previous blood versus LN analyses. Demographic data on the additional participants can be found in Table S1 in the supplemental material. Notably, total HIV-1 DNA content correlated with TIGIT expression in all T-cell subsets from blood except EM CD4⁺ T-cells and in LN CM CD4⁺ T-cells (Fig. 3A); there was a trend toward a positive correlation in TM and EM CD4⁺ T-cells from LN, but this trend did not reach statistical significance. There was a significant inverse correlation between the peripheral blood CD4⁺/CD8⁺ ratio and HIV-1 DNA content in the blood CM and TM CD4⁺ T-cell subsets and the LN TM CD4⁺ T-cell subset (Fig. 3B). There was a trend toward a positive correlation between HIV-1 total DNA and PD-1 expression in CM CD4⁺ T-cells from both LN and blood, but this did not reach statistical significance. There was no correlation between PD-1 expression and HIV-1 total DNA in the remaining CD4⁺ T-cell subtypes from either tissue site (data not shown).

FIG 3.

Correlation of HIV-1 total DNA and TIGIT expression and the peripheral blood CD4⁺/CD8⁺ ratio in CD4⁺ T-cells subsets from blood and LN. (A) Correlation of HIV-1 total DNA and TIGIT expression in CD4⁺ T-cells subsets from blood (magenta squares) and LN (green circles). (B) Correlation of HIV-1 total DNA and the peripheral blood CD4⁺/CD8⁺ ratio in blood (magenta squares) and LN (green circles). The Spearman correlation was performed for r calculations and statistical significance.

Blood and LN CD4⁺ CM T-cell immune activation, TIGIT expression, and PD-1 expression were inversely correlated with the peripheral blood CD4⁺/CD8⁺ ratio.

Next, we assessed the correlations between the CD4⁺/CD8⁺ ratio and the proportion of CD4⁺ T-cells expressing CD38 and HLA-DR, PD-1 and TIGIT, given that low CD4⁺/CD8⁺ ratios during prolonged ART suppression have been associated with ongoing immune dysfunction, immune senescence, and higher risks of non-AIDS-related morbidity and death (14, 15). There was a significant inverse correlation between the peripheral blood CD4⁺/CD8⁺ ratio and TIGIT expression in naive and CM CD4⁺ T-cell subsets in blood and CM, TM, and EM CD4⁺ T-cell subsets in LN (Fig. 4A). The peripheral blood CD4⁺/CD8⁺ ratio was inversely correlated with PD-1 expression in naive and CM CD4⁺ T-cell subsets from both blood and LN (see Fig. S2). There was a significant inverse correlation between the peripheral blood CD4⁺/CD8⁺ ratio and co-expression of CD38 and HLA-DR in blood naive, CM, and TM CD4⁺ T-cells and in LN CM, TM, and EM CD4⁺ T-cells (Fig. 4B).

FIG 4.

Correlation of the peripheral blood CD4⁺/CD8⁺ ratio and TIGIT expression and the proportion of CD4⁺ CD38⁺ HLA-DR⁺ CD4⁺ T-cell subsets from blood and LN. (A) Correlation of the peripheral blood CD4⁺/CD8⁺ ratio and TIGIT expression in CD4⁺ T-cells subsets from blood (magenta squares) and LN (green circles). (B) Correlation of the peripheral blood CD4⁺/CD8⁺ ratio and the proportion of CD4⁺ CD38⁺ HLA-DR⁺ T-cells in CD4⁺ T-cell subsets from blood (magenta squares) and LN (green circles). The Spearman correlation was performed for r calculations and statistical significance.

PD-1 expression and immune activation were significantly greater in all memory CD8⁺ T-cell subsets from LN versus blood.

Finally, we investigated the difference in immune activation of CD8⁺ T-cell subsets from LN and blood by measuring the percentage of CD8⁺ CD38⁺ HLA-DR⁺ T-cells by flow cytometry. The percentage of CD8⁺ CD38⁺ HLA-DR⁺ T-cells was significantly higher in LN than in blood for all cell subsets except for the EM subgroup (Fig. 5A), with the difference being driven by IR (n = 7) for both the CM and TM subsets (Fig. 5B and C). Of note, the percentage of CD8⁺ CD38⁺ HLA-DR⁺ T-cells was significantly greater in blood CM, TM and EM and LN naive, CM, and EM CD8⁺ T-cell subsets from ISR, compared with IR (see Fig. S1C and D). PD-1 expression was significantly higher in CM (mean of 67.2% versus 37.4% [P < 0.01]), TM (mean of 81% versus 64% [P < 0.001]), and EM (mean of 64.0% versus 40.8% [P < 0.01]) CD8⁺ T-cells from LN versus blood, with levels remaining significantly higher in LN than in blood in both IR (n = 7) and ISR (n = 7) (Fig. 5D to F). PD-1 expression was significantly lower in naive CD8⁺ T-cells from LN (mean of 2.6%) than in those from blood (mean of 7.8% [P = 0.003]), and this difference remained significant in IR (P = 0.02) (Fig. 5E). In contrast to PD-1 expression, TIGIT expression was significantly lower in TM CD8⁺ T-cells from LN (mean of 12.6%) than in those from blood (mean of 21.0% [P = 0.02]), and there was a similar nonsignificant trend for CM CD8⁺ T-cells (n = 14) (mean of 10.0 versus 14.3 [P = 0.12]) (Fig. 5G).

FIG 5.

Paired analysis of total HIV-1 DNA, CD38 and HLA-DR coexpression, PD-1 expression, and TIGIT expression in CD8⁺ T-cell subsets from LN and blood. (A to C) Comparison of the proportions of CD8⁺ CD38⁺ HLA-DR⁺ T-cells in blood and LN in naive, CM, TM, and EM CD8⁺ T-cells in all paired participants (A), IR (B), and ISR (C). (D to F) Comparison of PD-1 expression in blood and LN in naive, CM, TM, and EM CD8⁺ T-cells in all paired participants (D), IR (E), and ISR (F). (G to I) Comparison of TIGIT expression in blood and LN in naive, CM, TM, and EM CD8⁺ T-cells in all paired participants (G), IR (H), and ISR (I). The Wilcoxon test was performed for statistical significance. *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001.

DISCUSSION

Our group recently showed that PWH who have a suboptimal CD4⁺ T-cell recovery during ART (i.e., ISR) have important immunological and virological differences compared with individuals having robust CD4⁺ T-cell recovery (i.e., IR), in specific CD4⁺ T-cell differentiation subsets (9). These differences included greater expression of immune checkpoint molecules, increased levels of HIV infection in CM and TM CD4⁺ T-cell subsets, and reduced responsiveness to interleukin 7 (IL-7) and IL-15 in ISR. Specifically, we previously found that PD-1 expression was significantly higher in CM and EM CD4⁺ T-cells from blood and LN in the ISR, compared with the IR. Blood TIGIT expression was significantly higher in all CD4⁺ T-cell subsets except EM cells, and TIGIT expression was significantly higher for CM, TM, and EM CD4⁺ T-cells in LN of ISR versus IR (9). In CD8⁺ T-cells, there was no difference in PD-1 or TIGIT expression in blood between IR and ISR, with similar data for LN, i.e., greater PD-1 expression only in the naive CD8⁺ T-cell subset for ISR versus IR (9). Furthermore, we previously reported that increased stability of HIV-infected long-lived CM and TM cells discriminates ISR from IR and contributes to the increased persistence of the viral reservoir in ISR (9). Thus, analyses of CD4⁺ T-cell subsets have the potential to provide mechanistic insights into HIV pathogenesis and persistence that are likely missed by looking only at bulk CD4⁺ T-cells.

In 2017, Khoury et al. found a significantly greater percentage of CD4⁺ T-cells that coexpressed the activation markers CD38 and HLA-DR in rectal tissue and LN, compared with blood, in a cohort of participants with robust CD4⁺ T-cell recovery (6). Given our prior findings on the importance of long-lived CD4⁺ T-cell subsets in reservoir persistence and CD4⁺ T-cell recovery, as outlined above, we sought to better understand the relative contributions of CD4⁺ T-cell subsets to the reported differences between LN and peripheral blood. In the current study, we found significantly greater CD38 and HLA-DR coexpression on CD4⁺ T-cells from LN versus blood, particularly for naive, CM, and TM subtypes. Importantly, this difference was observed mainly in ISR and remained statistically significant in the CM CD4⁺ T-cell subset from this group. Furthermore, CD4⁺ T-cells high levels of expression of checkpoint proteins such as PD-1 and TIGIT have been implicated in reservoir maintenance, T-cell exhaustion, and failure of CD4 recovery despite adequate viral control with ART (16–19). In the current study, we found that PD-1 expression was consistently and significantly higher in CM CD4⁺ T-cells from LN versus blood for all participants, IR, and ISR. In contrast to the CM CD4⁺ T-cell findings, PD-1 expression was significantly greater in TM CD4⁺ T-cells from LN versus blood exclusively in the IR group. TIGIT expression was significantly greater in CM, TM, and EM CD4⁺ T-cells from LN versus blood, and this difference was driven by ISR. Additionally, TIGIT expression was correlated with total HIV-1 DNA content. Taken together, these data indicate that memory CD4⁺ T-cell subsets from LN, including T follicular helper cells, have T-cell activation and exhaustion markers expressed at higher levels than in blood, which could explain the higher levels of HIV-1 DNA in this compartment. Conversely, this increased T-cell activation and exhaustion could be a result of ongoing stimulation in the setting of higher HIV-1 DNA levels in LN versus blood.

Similar to findings described above for CD4⁺ T-cells, CD8⁺ T-cells displayed a greater percentage of CD38 and HLA-DR coexpression in LN than in blood, particularly in naive, CM, and TM subsets. However, this difference was observed mainly in IR and remained statistically significant only for the naive subset in ISR. Khoury et al. demonstrated a positive correlation between CD8⁺ T-cell activation and virus persistence, proposing that this may be explained by ongoing viral production and HIV antigen presentation in tissue, compared with blood (6). Based on the aforementioned findings, one would expect increased immune activation in CD8⁺ T-cells from LN from ISR compared with blood. It is possible that, given the global increases in CD8⁺ T-cell activation observed in ISR versus IR across both compartments, we were not able to detect a difference between LN and blood in the ISR group (see Fig. S1 in the supplemental material).

Finally, prior studies showed that increased expression of PD-1 and TIGIT on CD8⁺ T-cells is related to HIV disease progression and CD8⁺ T-cell exhaustion (18, 19). Here, we found that PD-1 expression levels were significantly higher in memory CD8⁺ T-cell subsets, including TM, CM, and EM cells, from LN versus blood. This difference remained significant in both IR and ISR except for the EM subset, which remained statistically significant only in the ISR group. Notably, and in contrast to PD-1 expression, TIGIT was expressed at significantly lower levels in TM CD8⁺ T-cells from LN versus blood; there was a similar nonsignificant trend for the CM subset. Interestingly, a recent study showed a strong correlation between the level of TIGIT expression and the intrinsic cytotoxic activity of CD8⁺ T-cells in peripheral blood, suggesting that these cells maintain an intrinsic capacity to kill targets in PWH in the absence of active viral replication (20). The authors proposed that a subset of TIGIT⁺ CD8⁺ T-cells, especially those lacking the expression of certain inhibitory receptors, does not represent a state of immune exhaustion but instead the cells maintain intrinsic cytotoxicity (20). Moreover, recent studies showed that EM CD8⁺ T-cells with cytolytic potential are limited to the intravascular circulation, while those found in LN are largely noncytolytic (4, 5). The significance of the difference in TIGIT expression found for TM and EM CD8⁺ T-cells from LN versus blood needs further investigation. However, if this protein is associated with the adequate cytolytic function of a subset of CD8⁺ T-cells, as described in the study above, a lower level of expression of TIGIT in LN CD8⁺ T-cells may reflect their lower cytolytic activity compared to CD8⁺ T-cells from blood-supporting our overall findings of increased residual dysregulation in LN.

Our study has some limitations, including a small number of participants with paired blood and LN samples (n = 14). Notably, we were still able to detect important differences despite this small sample size. Additionally, >90% of the participants included in our correlation analyses and 100% of the participants included in the paired analyses were cisgender men; therefore, our findings have limited generalizability to other genders. Furthermore, our DNA content analysis in CD4⁺ T-cell subsets is limited to total HIV-1 DNA, which limits our ability to comment on the contents of intact and replication-competent reservoirs. Nevertheless, total HIV-1 DNA still provides a measure of the overall viral burden, and it is plausible that cells with HIV-1 DNA still contribute to immune activation even if not intact. Finally, we acknowledge that dichotomizing CD4⁺ T-cell reconstitution into IR and ISR may not reveal subtle differences that may be seen using a continuous measurement.

Despite those limitations, our study provides new insights into the contributions of the specific CD4⁺ and CD8⁺ T-cell subsets to the anatomic differences between LN and blood in individuals with HIV who have optimal versus suboptimal CD4⁺ T-cell recovery. To our knowledge, this is the first study comparing paired LN and blood CD4⁺ and CD8⁺ T-cell differentiation subsets, as well as those subsets in IR versus ISR. Overall, our findings indicate increased residual dysregulation in LN, particularly for long-lived CD4⁺ T-cell subsets. Differences in HIV-1 DNA content, T-cell immune activation, and exhaustion in CD4⁺ T-cells in LN and blood were more pronounced in or specific for ISR, thus highlighting increased residual disease in LN as a critical feature of and a potential mechanism for individuals with poor CD4⁺ T-cell recovery during ART. Further investigation will be needed to determine whether expression of TIGIT on CD8⁺ T-cells plays a pivotal role in reducing LN tissue HIV reservoirs.

MATERIALS AND METHODS

Ethics statement.

This study was approved by the Emory University Institutional Review Board (approval number IRB00068326). Written informed consent was obtained from the participants. Portions of these methods were published in our previous work (9).

Study participants.

Individuals with HIV were enrolled in a longitudinal cohort as part of the Emory University Center for AIDS Research (CFAR) specimen repository between 2010 and 2016. Fourteen samples from participants receiving ART were included in the study (IR, n = 7; ISR, n = 7). An additional group of participants who underwent large blood draws only was included for the correlation analysis with blood (n = 32 [IR, n = 13; ISR, n = 19]). A detailed table with our inclusion and exclusion criteria can be found in the supplemental material of our prior study (9). IR were defined as those who achieved a CD4⁺ T-cell count of >500 cells/μL ≤2 years after ART initiation, and ISR were defined as those who did not achieve a CD4⁺ T-cell count of >500 cells/μL in the first 2 years after ART initiation.

Sample processing.

Blood samples were used within 1 h after phlebotomy. Density gradient centrifugation was performed for PBMC isolation from whole blood. LN tissue was homogenized and passed through a 70-μm cell strainer to isolate lymphocytes. All samples were processed, frozen in 10% dimethyl sulfoxide (DMSO) in heat-inactivated fetal bovine serum (FBS), and stored in liquid nitrogen until needed. Cells were thawed, washed twice in R10 (RPMI 1640 medium [Sigma-Aldrich] containing 10% FBS and 10 U/mL DNase I [Roche Diagnostics]), and resuspended in phosphate-buffered saline (PBS) for staining.

Flow cytometric analysis.

Multiparametric flow cytometric analysis was performed on PBMC and LN mononuclear cells according to standard procedures. Cell suspensions were stained with the viability dye LIVE/DEAD Aqua dye (Molecular Probes) before being incubated with the following antibodies: anti-CD3-allophycocyanin (APC)-Cy7 (clone SP34-2), anti-CCR7-phycoerythrin (PE)-Cy7 (clone 3D12), anti-CD45RA-fluorescein isothiocyanate (FITC) (clone L48), anti-CCR5-APC (clone 3A9), anti-CD38-BV421 (clone HIT2), anti-HLA-DR-texas-red PE (TRPE) (clone G46-2), and anti-CD4-BV650 (clone L200) from BD Biosciences; anti-PD-1-PE (clone EH12.2H7) and anti-TIM-3-BV605 (clone F38-2E2) from BioLegend; anti-LAG3-FITC (clone FAB2319) from R&D Systems; anti-CD8-BV705 (clone 3B5) from Life Technologies; anti-CD27-PE-Cy5 (clone 1A4CD27) from Beckman Coulter; and anti-TIGIT-PerCP-Cy5.5 (clone MBSA43) from eBiosciences. After staining, cells were fixed (1% paraformaldehyde) and analyzed within 24 h after collection. Flow cytometric acquisition was performed with ≥100,000 CD3⁺ T-cells on an LSRII cytometer (BD Biosciences) with FACSDiva software. Cell markers presenting <10,000 events acquired were excluded from the analyses. The data acquired were analyzed using FlowJo v9.8.5 software (TreeStar).

FACS.

PBMC and mononuclear cells isolated from LN tissue were stained with anti-CD3-APC-Cy7 (clone SP34-2), anti-CCR7-PE-Cy7 (clone 3D12), and anti-CD45RA-FITC (clone L48) from BD Biosciences; anti-CD8-BV705 (clone 3B5) from Life Technologies; anti-CD27-PE-Cy5 (clone 1A4CD27) from Beckman Coulter; and anti-CD4-BV421 (clone OKT4) from BioLegend. Different CD4⁺ T-cell subsets were then sorted based on their expression of CD45RA, CCR7, and CD27, using a FACSAria II system (BD Biosciences). Sorted CD4⁺ T-cell subsets were on average >96% pure, as determined by postsorting fluorescence-activated cell sorting (FACS) analysis. Gating strategies for different T-cell subsets are shown in Fig. S3 in the supplemental material.

Measurements of total HIV-1 DNA.

Total HIV-1 DNA was measured based on a previously described protocol (21). Briefly, cell pellets of sorted cells were resuspended in a lysis buffer (10 mM Tris-HCl [pH 8.0], 50 nM KCl, 400 μg/mL proteinase K [Invitrogen]) and digested for 12 to 16 h. Proteinase K was inactivated, and cell lysates were directly used in all preamplification reactions. In all PCRs, primers specific for the human CD3 gene were used to quantify the exact number of cells present in the reaction tube. HIV-1 DNA genomes (long terminal repeat [LTR]-gag) together with the CD3 gene were preamplified for 12 cycles. The second round of PCR was carried out in real time on a Rotor-Gene Q instrument (Qiagen) with the Rotor-Gene Probe Master Mix (Qiagen), following the manufacturer’s instructions. In all experiments, serial dilutions of digested ACH2 cells were used as standards for quantification. All measurements of HIV-1 DNA were performed in triplicate PCRs. Samples for which <3,000 cells were assayed were excluded from the analyses. The following primer sequences were used: HIV total primers, first PCR (preamplification): ULF1: 5′-ATG CCA CGT AAG CGA AAC TCT GGG TCT CTC TDG TTA GAC-3′; UR1: 5′-CCA TCT CTC TCC TTC TAG C-3′; HIV total primers, second PCR (quantitative PCR [qPCR]): LambdaT: 5′-ATG CCA CGT AAG CGA AAC T-3′; UR2: 5′-CTG AGG GAT CTC TAG TTA CC-3′; HIV total probe: UHIV FamZen: 5′-/56-FAM/CA CTC AAG G/ZEN/C AAG CTT TAT TGA GGC/3IABkFQ/-3′; CD3 primers, first PCR (preamplification): 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′; CD3 primers, second PCR (qPCR): HCD3 in 5′: 5′-GGC TAT CAT TCT TCT TCA AGG T-3′; HCD3 in 3′: 5′-CCT CTC TTC AGC CAT TTA AGT A-3′; CD3 probe: CD3 FamZen: 5′-/56-FAM/AG CAG AGA A/ZEN/C AGT TAA GAG CCT CCA T/3IABkFQ/-3′.

Statistical analysis.

Analyses and figures were generated with GraphPad Prism v9.0 software. The Wilcoxon test was performed for statistical significance for paired analyses, and the Mann-Whitney test was performed for unpaired analyses. Spearman correlation was performed for r calculations and statistical significance in correlation analyses. For all tests, statistical significance was defined on the basis of P values of <0.05.