A computational analysis of transcriptional profiles from CD8(+) T lymphocytes reveals potential mechanisms of HIV/AIDS control and progression

Abstract

Cytotoxic and noncytotoxic CD8⁺ T lymphocyte responses are essential for the control of HIV infection. Understanding the mechanisms underlying HIV control in elite controllers (ECs), which maintain undetectable viral load in the absence of antiretroviral therapy, may facilitate the development of new effective therapeutic strategies. We developed an original pipeline for an analysis of the transcriptional profiles of CD8⁺ cells from ECs, treated and untreated progressors. Hierarchical cluster analysis of CD8⁺ cells' transcription profiles allowed us to identify five distinct groups (EC groups 1-5) of ECs. The transcriptional profiles of EC group 1 were opposite to those of groups 2-4 and similar to those of the treated progressors, which can be associated with residual activation and dysfunction of CD8⁺ T-lymphocytes. The profiles of groups 2-4 were associated with different numbers of differentially expressed genes compared to healthy controls, but the corresponding genes shared the same cellular processes. These three groups were associated with increased metabolism, survival, proliferation, and the absence of an "exhausted" phenotype, compared to both untreated progressors and healthy controls. The CD8⁺ lymphocytes from these groups of ECs may contribute to the control under HIV replication and slower disease progression. The EC group 5 was indistinguishable from normal. Application of master regulator analysis allowed us to identify 22 receptors, including interferon-gamma, interleukin-2, and androgen receptors, which may be responsible for the observed expression changes and the functional states of CD8⁺ cells from ECs. These receptors can be considered potential targets of therapeutic intervention, which may decelerate disease progression.

Keywords: CD8+ T lymphocytes; Cluster analysis; Elite controllers; Gene expression; Human immunodeficiency virus; Master regulators.

Graphical abstract

Fig. 1

Heatmap demonstrating clustering results of CD8⁺transcription profiles from ECs. The rows in the heatmap are genes; the columns are samples. The blue, cyan, green, red, and yellow colors of columns represent EC groups 1–5 (Table 1). Row Z-Score is the number of standard deviations by which the value of gene expression in particular sample is above or below the mean value of all samples. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fig. 2

Comparison of five EC groups with cART-treated progressors and healthy controls. The rows in the heatmap are genes; the columns are samples. The blue, cyan, green, red, and yellow colors of columns represent EC groups 1–5 (Table 1); grey and black colors represent cART-treated progressors and healthy controls, respectively. Row Z-Score is the number of standard deviations by which the value of gene expression in particular sample is above or below the mean value of all samples. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fig. 3

Comparison of log fold changes from five EC groups, cART-treated and untreated progressors. The rows in the heatmap are genes; the columns are groups of HIV-infected individuals. The EC groups 1, 2, 3, 4, 5 (EC1-5) are marked by blue, cyan, green, red, and yellow colors. cART-treated progressors (cART) are marked by grey color. Untreated progressors in acute (AI1 and AI2) (GSE6740 and GSE25669 GEO datasets, correspondingly) and chronic (CI) (GSE6740 GEO dataset) phases are marked by black color. Only genes, which were differentially expressed in at least one of the investigated groups (Table 2), were used to create the heatmap. Row Z-Score is the number of standard deviations by which the value of log fold change in particular column is above or below the mean value of all groups. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fig. 4

KEGG pathways are differentially regulated in ECs, cART-treated and untreated progressors. The values (columns 3–5) are t-test scores calculated in gene set enrichment analysis (see Materials and Methods). The positive value and red color mean that pathway is up-regulated, whereas the negative value and blue color mean that pathway is down-regulated compared to healthy control. EC 1–5 are groups of ECs; cART is cART-treated progressors; AI1 and AI2 are untreated progressors in the acute phase from GSE6740 and GSE25669 GEO datasets, correspondingly; CI is untreated progressors in the chronic phase from GSE6740 GEO dataset. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fig. 5

Receptors, identified as MRs, and their transcription changes in EC groups, cART-treated and untreated progressors. EC 1–5 are groups of ECs; cART is cART-treated progressors; AI1 and AI2 are untreated progressors in the acute phase from GSE6740 and GSE25669 GEO datasets, correspondingly; CI is untreated progressors in the chronic phase from GSE6740 GEO dataset.