Inflammatory and immune markers in HIV-infected older adults on long-term antiretroviral therapy: Persistent elevation of sCD14 and of proinflammatory effector memory T cells

Abstract

HIV-positive patients whose viral loads are successfully controlled by active antiretroviral therapy (ART) show no clinical signs of AIDS. However, their lifespan is shorter compared with individuals with no HIV infection and they prematurely exhibit a multitude of chronic diseases typically associated with advanced age. It was hypothesized that immune system aging may correlate with, and provide useful biomarkers for, this premature loss of healthspan in HIV-positive subjects. Here, we tested whether the immune correlates of aging, including cell numbers and phenotypes, inflammatory status, and control of human cytomegalovirus (hCMV) in HIV-positive subjects on long-term successful ART (HIV+) may reveal increased "immunological age" compared with HIV-negative, age-matched cohort (HIV-) in participants between 50 and 69 years of age. Specifically, we expected that younger HIV+ subjects may immunologically resemble older individuals without HIV. We found no evidence to support this hypothesis. While T cells from HIV+ participants displayed differential expression in several differentiation and/or inhibitory/exhaustion markers in different T cell subpopulations, aging by a decade did not pronounce these changes. Similarly, while the HIV+ participants exhibited higher T cell responses and elevated inflammatory marker levels in plasma, indicative of chronic inflammation, this trait was not age-sensitive. We did find differences in immune control of hCMV, and, more importantly, a sustained elevation of sCD14 and of proinflammatory CD4 and CD8 T cell responses across age groups, pointing towards uncontrolled inflammation as a factor in reduced healthspan in successfully treated older HIV+ patients.

Keywords: HIV; antiretroviral therapy; immune aging; sCD14.

FIGURE 1

Inflammatory markers in plasma from successfully treated HIV‐positive participants (HIV+) and HIV‐negative counterparts (HIV−). (a) Comparison of inflammatory marker levels in plasma between HIV+ and HIV−. Means and standard errors in each age/HIV group are shown. Significant differences between HIV+ (black) and HC (gray) were determined by unpaired t‐test. *p < 0.05, **p < 0.01, ***p < 0.001. (b) Analysis of covariance analysis of inflammatory markers with age for HIV+ (black dots and lines) and HIV− (gray dots and lines) using an F‐test for the difference in slopes.

FIGURE 2

T cell subset analysis in PBMCs from successfully treated HIV‐positive participants (HIV+) and HIV‐negative counterparts (HIV−). (a) Gating strategy for CD4 and CD8 T cell subset analysis by flow cytometry shown in (b). Lymphocyte population without cell debris was selected followed by removal of doublets and dead cells. Single positive CD4 and single positive CD8 cells are selected in CD3+ population and naïve, central memory (CM) and total effector memory (EM) cell populations are gated using CD28 and CD95, respectively. Total EM population was then divided into effector memory (EM) and CD45RA+ EM (TEMRA) using CCR7 and CD45RA. (b–d) Absolute numbers of total, naïve, central memory (CM), effector memory (EM) CD4 (b) and CD8 (c) T cell subsets and δγ T cells (d). Means and standard errors in each age/HIV group are shown. Significant differences between HIV+ (black) and HIV− (gray) were determined by unpaired t‐test. *p < 0.05, **p < 0.01, ***p < 0.001. (e–g) analysis of covariance f absolute number of CD4 (e) and CD8 (f) T cell subset and δγ T cells in blood with age for HIV+ (black dots and lines) and HIV− (gray dots and lines), using an F‐test for the difference in slopes. PBMCs, peripheral blood mononuclear cells.

FIGURE 3

Expression of functional markers on CD4 and CD8 T cells in PBMCs from HIV+ and HIV−. (a) Frequency of each functional marker positive cells in total CD4 or CD8 T cells. (b) Mean fluorescence intensity (MFI) of functional marker signals in each respective positive populations of CD4 or CD8 T cells. Means and standard errors in each HIV group are shown. Significant differences between HIV+ (black) and HIV− (gray) in each age were determined by unpaired t‐test. *p < 0.05, **p < 0.01, ***p < 0.001. (c) Regression analysis of frequencies of TIM3 positive cells and MFI of TIM3 signals in TIM3 positive population of CD4 and CD8 T cell in blood with age for HIV+ (black dots and lines) and HIV− (gray dots and lines). p Values are from an F‐test statistic for the difference in slopes. PBMCs, peripheral blood mononuclear cells.

FIGURE 4

T cell proliferation and marker expression profiles in non‐dividing and dividing CD4. And CD8 T cells after CD28/CD3/CD49d stimulation in HIV+ and HIV−. (a) Gating strategy for non‐dividing and dividing CD4 and CD8 T cells using proliferation modeling with FlowJo. The low‐right panel shows gating example of dividing cells and positive populations on TIGIT. (b) Number of cells in each division. (c–g) frequencies of PD‐1+ (c), TIM3+ (d), TIGIT+ (e), CD57+ (f) and TCF‐1+ (g) cells in each divisions. Means (columns) and standard errors (bars) in each HIV group are shown. Significant differences between HIV+ (black) and HIV− (gray) were determined by unpaired t‐test. *p < 0.05, **p < 0.01. (h,i) analysis of covariance of frequencies of PD‐1+ cells in CD4 and CD8 (h) cells or frequencies of CD57+ cells in CD4 and CD8 T cells with age. p Values are the significance from an F‐test statistic for the difference in slopes.

FIGURE 5

Host response to CMV‐specific and non‐specific stimulation. (a,b) Total antibody titers (a) and 80% neutralization titers (b) against CMV in plasma. Significant differences between HIV+ (black) and HIV− (gray) antibody titers were determined by unpaired t‐test. ***p < 0.001. (c) Correlation between total anti‐CMV antibody (tAb) titer and anti‐CMV neutralization antibody (nAb) titer in HIV+ and HIV−. (d) Frequency of IFN‐γ + and/or TNF‐α + cells after 3 h of CMV‐peptide stimulation of PBMCs from HIV+ and HIV−. (e) Frequency of IFN‐γ + and/or TNF‐α + cells after 3 h of PMA/ionomycin stimulation of PBMCs from HIV+ and HIV−. Statistical significance between frequencies of cytokine‐positive cells in HIV+ and HIV− were determined by unpaired t‐test. ***p < 0.001. (f,g) Analysis of covariance showing statistical significance from an F‐test statistic for the difference in slopes in CMV‐peptide stimulated cells (f) and PMA/ionomycin stimulated cells (g). PBMCs, peripheral blood mononuclear cells.

FIGURE 6

Heatmap of the bivariate Pearson correlation matrix. Correlations between inflammatory markers and other significant parameters in HIV+ and in HIV− are shown. Red and green colors represent positive and negative correlation, respectively, and darker colors representing larger values. Pearson's r are shown in the boxes with statistically significant p‐values (represented by stars) based on a z‐test statistic for each correlation under the hypothesis of difference from zero. *p < 0.05, **p < 0.01. D, Dividing cells; ND, Non‐dividing cells.