Limited HIV infection of central memory and stem cell memory CD4+ T cells is associated with lack of progression in viremic individuals

Abstract

A rare subset of HIV-infected individuals, designated viremic non-progressors (VNP), remain asymptomatic and maintain normal levels of CD4+ T-cells despite persistently high viremia. To identify mechanisms potentially responsible for the VNP phenotype, we compared VNPs (average >9 years of HIV infection) to HIV-infected individuals who have similar CD4+ T-cell counts and viral load, but who are likely to progress if left untreated ("putative progressors", PP), thus avoiding the confounding effect of differences related to substantial CD4+ T cell depletion. We found that VNPs, compared to PPs, had preserved levels of CD4+ stem cell memory cells (TSCM (p<0.0001), which was associated with decreased HIV infection of these cells in VNPs (r = -0.649, p = 0.019). In addition, VNPs had decreased HIV infection in CD4+ central memory (TCM) cells (p = 0.035), and the total number of TCM cells was associated with increased proliferation of memory CD4+ T cells (r = 0.733, p = 0.01). Our results suggest that, in HIV-infected VNPs, decreased infection of CD4+ TCM and TSCM, cells are involved in preservation of CD4+ T cell homeostasis and lack of disease progression despite high viremia.

Figure 1. Clinical and immunological characteristics of viremic non-progressors and putative progressors.

(A) Estimated duration of HIV diagnosis based on first HIV+ test. (B) Plasma HIV RNA in VNPs and PPs as measured by RT-PCR. (C) The fraction of CD4+ T cells in VNPs and PPs measured by flow cytometry. (D) Absolute CD4+ T cell count in blood from VNPs and PPs based on CD4+ T cell fraction and lymphocyte count by CBC. p values from Mann Whitney T test. Line reflects median. Circles, VNPs; Squares, PPs.

Figure 2. CD4+ T cell memory counts and representative flow cytometry staining.

(A–C) The absolute number of CD4+ T cell subsets in VNPs and PPs was measured by flow cytometry and lymphocyte count by complete blood cell (CBC) counts. (A) The absolute number of CD4+ T_EM cells; (B) The absolute number of CD4+ T_CM cells; (C) The absolute number of CD4+ T_SCM cells in VNPs and PPs. (D) Representative flow cytometry staining for each CD4+ T cell memory subset. p values from Mann Whitney T test, line reflects median. Circles, VNPs; Squares, PPs.

Figure 3. T cell proliferation and activation in VNPs and PPs.

(A–B) The frequency of activated CD38+HLA-DR+ T cells in VNPs and PPs by flow cytometry: (A) total CD4+ CD38+HLA-DR+ T cells and (B) CD8+ CD38+HLA-DR+ T cells. (C–D) The frequency of proliferating (Ki67+) VNPs and PPs by flow cytometry: (C) total CD4+Ki67+ T cells and (D) total CD8+Ki67+ T cells. (E–G) The absolute count of Ki67+CD4+ T cell memory subsets measured by flow cytometry and lymphocyte count by CBC for: (E) CD4+Ki67+ T_EM cells; (F) CD4+Ki67+ T_CM cells; (G) and CD4+Ki67+ T_SCM cells. p values from Mann Whitney T test. Line reflects median. Circles, VNPs; Squares, PPs.

Figure 4. VNPs have decreased HIV infection of central memory and stem cell memory T cells.

CD4+ T_EM cells (left), T_CM (center), and T_SCM (right) from VNPs and PPs were sorted by flow cytometry and quantitaive real-time PCR was used to determine the HIV infection frequency in each subset. Infection frequency was determined by copies of gag DNA/100 infected cells. p values from Mann Whitney T test (VNPs vs PPs) or paired T test (T_CM vs T_EM within PP cohort). Line reflects median. Circles, VNPs; Squares, PPs.

Figure 5. CD4+ T cell memory subset numbers are associated with differential preservation mechanisms.

CD4+ T_CM cells (A–B): Correlation between frequency of CD4+Ki67+ memory cells and absolute number of CD4+ T_CM cells in VNPs (A) and PPs (B). CD4+ T_SCM cells (C–D) Correlation between frequency of CD4+Ki67+ memory cells and absolute number of CD4+ T_SCM cells in VNPs (C) and PPs (D). (E) Correlation between HIV DNA infection per 100 CD4+ T_SCM cells and absolute number of CD4+ T_SCM cells. CD4+ T_EM cells (F–H): Correlation between frequency of CD4+Ki67+ memory cells and absolute number of CD4+ T_EM cells in VNPs (F) and PPs (G). (H) Correlation between frequency of CD4+CD38+HLA-DR+ memory T cells and absolute number of CD4+ T_EM cells. p and r values from spearman correlations, with linear regression shown as line. Box around significant (p<0.05) values. Circles, VNPs; Squares, PPs.

Figure 6. Transcriptomic signatures between HIV-infected putative progressors and viremic non-progressors.

Gene expression profiles from whole blood were assayed by microarray analysis. (A) Prinicipal components analysis was used to assess the divergence of the entire transcriptomic profile of the patient classes. (B) Volcano plot showing the distribution of fold-changes and significance level between the PP and VNP phenotypes and the p-value assessed by two-sample t-test. Red data points indicate genes upregulated in PP patient samples, defined as >1.5-fold greater than the VNP samples, and a p-value of <0.05. Blue data points indicate genes >1.5-fold in VNPs with 0.05 p-value. The y-axis is plotted as the log of the inverse of the p-value. (C) Pathway analysis of genes differentially expressed between VNPs and PPs was performed using the Gene Ontology database. Fisher's exact test was used to estimate the significance of enrichment an annotations of probesets on the arrays were used as a background. The forest plot shows genes identified by GO as have immune function and divides them into GO subcategories. (D) Heat map showing the overall concordance amongst patients in each class. The top 50 most highly expressed genes were median-normalized and organized by hierarchical clustering using Pearson's dissimilarity metric and average linkage. (E) Forest plot of 52 genes defined in the GO category “Immune Process” in panel C. Branches indicate subcategories, the number of genes with higher expression in each phenotype are indicated by the scale on the x-axis, and the magnitude of the fold-change is indicated by the color scale. (F) Relative expression of ISGs between phenotypes. (G) Forest plots of individual genes contained with GO categories and unincorporated immune genes. *probeset is predicted to hybridize to multiple genes. (H & I) Gene Set Enrichment plots of ISGs and IL6 signaling genes. GSEA performed using geneset permutation, and array data ranked by signal-to-noise ratio. Vertical dotted lines indicate the margin between gene upregulated in VNPs (left) and PPs (right) of all non-redundant, annotated genes on the array. Data points in red indicate leading edge genes contributing to the majority of the enrichment score.