Dynamics of pulmonary mucosal cytotoxic CD8 T-cells in people living with HIV under suppressive antiretroviral therapy

Abstract

Background: Despite the success of antiretroviral therapy (ART), people living with HIV (PLWH) suffer from a high burden of pulmonary diseases, even after accounting for their smoking status. Cytotoxic CD8 T-cells are likely implicated in this phenomenon and may act as a double-edged sword. While being essential in viral infection control, their hyperactivation can also contribute to lung mucosal tissue damage. The effects of HIV and smoking on pulmonary mucosal CD8 T-cell dynamics has been a neglected area of research, which we address herein.

Methods: Bronchoalveolar lavage (BAL) fluid were obtained from ART-treated PLWH (median duration of supressed viral load: 9 years; smokers: n = 14; non-smokers: n = 21) and HIV-uninfected controls (smokers: n = 11; non-smokers: n = 20) without any respiratory symptoms or active infection. Lymphocytes were isolated and CD8 T-cell subsets and homing markers were characterized by multiparametric flow cytometry.

Results: Both smoking and HIV infection were independently associated with a significant increase in frequencies of total pulmonary mucosal CD8 T-cell. BAL CD8 T-cells were primarily CD69 + expressing CD103 and/or CD49a, at least one of the two granzymes (GzmA/GzmB), and little Perforin. Higher expression levels of CD103, CD69, and GzmB were observed in smokers versus non-smokers. The ex vivo phenotype of GzmA + and GzmB + cells revealed increased expression of CD103 and CXCR6 in smokers, while PLWH displayed elevated levels of CX3CR1 compared to controls.

Conclusion: Smoking and HIV could promote cytotoxic CD8 T-cell retention in small airways through different mechanisms. Smoking likely increases recruitment and retention of GzmB + CD8 Trm via CXCR6 and CD103. Heightened CX3CR1 expression could be associated with CD8 non-Trm recruitment from the periphery in PLWH.

Keywords: CD8 T-cells; HIV; Lung; People living with HIV (PLWH); Pulmonary immunity; Smoking; Tissue resident; Trm.

Fig. 1

Impact of chronic HIV infection and smoking on BAL CD8 T-cell frequencies and the expression of Trm markers. (a) Frequencies of CD8 + CD4- cells in total BAL CD3 + live lymphocytes are shown. (b-d) Expression levels of CD103, CD69, and CD49a were assessed within the total CD8 T-cells pool in BAL (HIV-NS: n = 8; HIV-SM: n = 6; HIV + NS: n = 6; HIV + SM: n = 6). Data points were stratified by smoking status (right), HIV status (middle), or both (left). Comparisons were made using Mann–Whitney rank-sum test (* p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001)

Fig. 2

Distribution of Trm subsets and cytotoxic effector subsets among total BAL CD8 T-cells. (a) Trm subsets stratified by CD103/CD69/CD49a expression in BAL based on median cell population frequencies in each study group. (b) Cytotoxic effector CD8 T-cell subsets stratified by GzmA/GzmB/Perforin expression in BAL and blood based on median cell population frequencies in each study group. (HIV-NS: n = 8; HIV-SM: n = 6; HIV + NS: n = 6; HIV + SM: n = 6)

Fig. 3

Expression of CD8 T-cell cytotoxic effector molecules GzmA/B and perforin among study groups. Expression levels of GzmA (a), GzmB (b), and Perforin (c) were assessed within the total CD8 T-cell pool in BAL (HIV-NS: n = 8; HIV-SM: n = 6; HIV + NS: n = 6; HIV + SM: n = 6). Data points were stratified by smoking status (right), HIV status (middle), or both (left). Comparisons were made using Mann–Whitney rank-sum test (* p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001)

Fig. 4

Increased migration and retention of GzmB + CD8 Trm from the interstitium to alveolar space in smokers. (a-d) Expression levels of CD103, CD49a, CXCR6, and CX3CR1 in GzmB + BAL CD8 T-cells (HIV-NS: n = 8; HIV-SM: n = 6; HIV + NS: n = 6; HIV + SM: n = 5), (e-f) as well as CXCR6 and CX3CR1 in GzmA + BAL CD8 T-cells (HIV-NS: n = 8; HIV-SM: n = 5; HIV + NS: n = 6; HIV + SM: n = 5) are shown. Data points were stratified by smoking status (right), HIV status (middle), or both (left). Comparisons were made using Mann–Whitney rank-sum test (* p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001)

Fig. 5

FlowSOM analysis reveals five distinct BAL CD8 T-cell populations that vary by smoking status. (a) t-distributed stochastic neighbor embedding (tSNE) plot of sub-population in BAL CD3 + lymphocytes pooled from four study groups (n = 5 per group; 3905 CD3 + live BAL lymphocytes per group; 15,620 cells in total). Five different CD8 T-cell populations identified using FlowSOM clustering and visualization technique are shown. (b) tSNE plots of BAL CD8 T-cell populations stratified by either HIV, smoking status, or both with (c) their corresponding cell frequencies shown as bar plots. (d) The heat map displays relative expression of markers used for the FlowSOM analysis in each population identified

Fig. 6

Significantly higher levels of GzmB in effector memory (EM) and terminally differentiated (TD) BAL CD8 T-cells cells in smokers. (a) Expression levels of GzmA, GzmB, and Peforin in BAL CD8 T-cell memory subsets (Naïve/ CCR7 + CD45RA+: HIV-NS n = 6, HIV-SM n = 2, HIV + NS n = 5, HIV + SM n = 1; central memory (CM) / CCR7 + CD45RA-: HIV-NS n = 6, HIV-SM n = 4, HIV + NS n = 4, HIV + SM n = 3; effector memory (EM) CCR7-CD45RA-: HIV-NS n = 8, HIV-SM n = 5, HIV + NS n = 6, HIV + SM n = 4), terminally differentiated (TD) CCR7-CD45RA+: HIV-NS n = 8, HIV-SM n = 3, HIV + NS n = 6, HIV + SM n = 2; (b) Distributions of BAL CD8 T-cell memory subsets between different study groups (HIV-NS: n = 8; HIV-SM: n = 6; HIV + NS: n = 6; HIV + SM: n = 6); (c) Frequencies of GzmB + cells in CM (left) and TD (right) BAL CD8 T-cell subsets stratified by participant smoking status. Wilcoxon matched-pairs signed rank test and Mann-Whitney rank-sum test were used to compare paired and unpaired study variables respectively (* p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001)