CD8 + HLA-G + Regulatory T Cells Are Expanded in HIV-1-Infected Patients

Introduction

I t has been recently identified and characterized that a subpopulation of CD8 T cells in human peripheral blood expresses the immune-tolerizing molecule HLA-G (4). These CD8⁺ HLA-G-expressing T cells are FOXP3-negative, are mostly naïve, and exhibit potent suppressive properties that are partially mediated by HLA-G. These HLA-G-positive T cells are present at sites of inflammation and they have been identified in inflammatory-related diseases, suggesting an important function in modulating inflammatory responses in vivo (4).

Materials and Methods

In HIV-1 infection, a vigorous CD8⁺ T-cell response has been associated with better control of the disease in certain circumstances (6,10,11). Also, a loss of HLA-A and HLA-B molecules, and an increase in HLA-G (12) have been observed in HIV-1-infected patients. This HLA-G overexpression has been described in monocytes and also in lymphocytes (7), and it has been associated with poor outcome in HIV-1-infected patients (11). With these observations in mind, we have recently been studying the distribution of the HLA-G molecule on different CD8⁺ T-cell subsets in HIV-1-seropositive individuals with undetectable viral load, nine of whom were receiving highly active antiretroviral therapy (HAART), while nine had a CD4⁺ lymphocyte level higher than 500 cells/μL, and were asymptomatic and antiretroviral naïve.

Eighteen HIV-1-seropositive individuals with undetectable viral load were analyzed, nine of whom had been receiving HAART for at least 2 y, with HIV-RNA levels lower than 50 copies/mL in at least three determinations in the course of 6 mo, with more than 500 cells/μL CD4⁺ lymphocytes during the previous 12 months, were clinically asymptomatic, and were classified in group A of the CDC classification. The other nine had CD4⁺ lymphocyte levels higher than 500 cells/μL, and were asymptomatic and antiretroviral naïve. Eight healthy volunteers served as controls (HC). PBMCs were subjected to inmunostaining using the following conjugated mAbs: anti-CD3-PE, anti-CD8-FITC (Becton Dickinson, Biosciences, San José, CA), and anti-HLA-G MEM-G/09 (Exbio, Prague, Czech Republic), plus Cy5-conjugated goat anti-mouse. CD8⁺HLA-G⁺ T-cell frequencies were analyzed in healthy controls and untreated and treated HIV-1-infected patients in CD3⁺ gated cells. Student's t-test (Kolmogorov-Smirnov) was used for the statistical analysis.

Results

The results we obtained showed an expansion of the CD8⁺HLA-G⁺ subset in HIV-1-infected treated (8.65 ± 0.86%; p < 0.001) and untreated patients (4.09 ± 0.28%; p < 0.001), compared with healthy individuals (1.06 ± 0.54%) (Fig. 1). In this expanded CD8⁺HLA-G⁺ population, we have also analyzed the maturation stages using the CD45RA surface marker and the CCR7 chemokine receptor marker. The results show that cells with a naïve phenotype (CD45RA⁺CCR7⁺) are the major subset within the CD8⁺HLA-G⁺ T cells in HIV-1-positive patients in both treated and untreated individuals. But importantly, only in HIV-1-infected individuals were there CD8⁺HLA-G⁺ cells with an effector memory phenotype.

OPEN IN VIEWER

FIG. 1.

HLA-G-expressing CD8⁺ T cells in HIV-1 infected patients. (A) A representative example of HC and untreated patients are shown in the left panel. In the right panel, the average of CD8 T cells expressing HLA-G is shown (*p < 0.001). (B) Maturation phenotype according to the CD45RA and CCR7 (anti-CD45RA and anti-CCR7 mAbs provided by Becton Dickinson) surface markers in both, CD8⁺HLA-G⁺ (right panel) and CD8⁺HLA-G⁻ (left panel) cells (CM, central memory [CD45RA-CCR7⁺]; N, naïve [CD45RA⁺CCR7⁺]; EM, effector memory [CD45RA⁻CCR7⁻]; E, effector [CD45RA⁺CCR7⁻]). Results are expressed as means ± SD.

Discussion

These results showing an expansion of the CD8⁺HLA-G⁺ subset point toward their involvement in the regulation of the immune response, and appear to be of the same regulatory T-cell subset as that described by Feger et al. (4). The reasons for this expansion of the CD8⁺HLA-G⁺ naïve T-cell subset in HIV-1 infection are unknown. Nevertheless, one important issue may be that IL-10 levels in HIV-1-positive patients are increased, and this cytokine has been shown to upregulate HLA-G expression (1,8,9). The elevated frequency of this suppressive subset observed in untreated, and even more obviously in treated HIV-1⁺ patients, might partially suppress the HIV-specific CD8 T-cell response. This regulatory function could also be mediated by other regulatory T cells induced by HLA-G (4). It may be that ILT2, which is a ligand of HLA-G and is upregulated in HIV-1 infection (2,3,5), could boost the inhibitory effects of this regulatory subset in these patients. Thus these results indicate an expansion of an HLA-G-expressing CD8⁺ T-cell subset in HIV-1-infected patients that may be associated with the impaired CD8⁺ response seen in HIV-1-infected patients.