Potential role of CD8+CD28- T lymphocytes in immune activation during HIV-1 infection

As CD8+CD28- T cells have been associated with dendritic and T cell suppression, we analyzed whether an increase in CD8+CD28- T cell numbers during HIV-1 infection could lead to impaired T cell responses. In contrast to the in-vitro generated CD8+CD28- suppressors, peripheral blood CD8+CD28- T cells of both HIV-infected and noninfected individuals promoted dendritic cell activation. The CD8+CD28- T cell accumulation during HIV-1 infection may thus contribute to accelerated inflammatory reactions and immune activation.

The CD28- T cell pool is expanded in the peripheral blood of HIV-1 infected or aged individuals reaching up to 80% among CD8+ T cells ^[1-4]. CD28- T cells, as indicated by several studies ^[1,4-8], represent chronically expanded clones specific for a limited set of antigens and characterized by impaired proliferative ability.

Interestingly, CD8+ T cell subsets characterized by immunosuppressive functions were also associated with the lack of CD28 expression ^[9]. These cells were generated through repeated in-vitro allogenic stimulations (type 1 suppressor T cells) or in cultures with autologous monocytes, granulocyte-macrophage colony stimulating factor (GM-CSF) and interleukin (IL)-2 (type 2 suppressor T cells). The suppressor T cells exert their immunosuppressive effects on T cell activation by modulating the T cell activatory potential of dendritic cells or by directly affecting activated T cells through soluble mediators (type I and type II suppressor T cells, respectively) ^[9,10]. Decreased activity of CD8+CD28- suppressor T cells has been associated with the increased risk of transplant rejections, with the active phase of systemic lupus erythematosus, and with HIV-1 infection providing a potential link between suppressor T cell activity and tolerance regulation in vivo ^[9,11,12]. In these studies, however, the detection of any CD8+CD28- suppressor activity required extensive in-vitro induction procedures, including either sequential antigenic stimulation or treatments with GM-CSF, IL-2 and monocytes ^[9,11-14]. Whether CD8+CD28- T cells occurring in vivo are able to elicit suppressor activity per se and under what circumstances the CD28- suppressor T cells develop from their precursors are questions yet to be clarified.

It is intriguing to speculate that the high number of CD28- T cells existing in HIV-1 infected individuals may result in the altered frequency of potential suppressor T cell types and thus may contribute to immune insufficiencies leading to inappropriate T cell responses against HIV, opportunistic infections and malignant cells. FoxP3 expression, detected in the in-vitro generated CD28- T suppressor cells and the decreased IL-7Rα expression on peripheral blood CD28- T cells indicate similarities between CD8+CD28- T cells and CD4+ regulatory T cells ^[15-17]. In the present study, we analyzed whether the CD28- T cells isolated from peripheral blood of HIV-1 infected or noninfected individuals contain FoxP3+ cells that are able to elicit suppressor activity on dendritic cell functions or T cell activation. Blood samples were obtained from 12 HIV-1 infected patients (nine on combination therapy, three treatment naïve, viral load ranged between <50 and 139 000 copies/ml, mean CD4+ T cell count was 474 cells/μl) and from eight noninfected donors.

Our results showed that the CD8+CD28- T cells were FoxP3-negative in all HIV-1 infected (n = 5) and noninfected individuals (n = 5) tested whereas the expression of FoxP3 was exclusively associated with the CD28+ CD4+ IL-7Rα low T cells in the peripheral blood (Fig. 1a,b). We sought to analyze systematically how CD28- T cells isolated from HIV-1 infected or noninfected individuals may influence dendritic cell and T cell activation. CD28+ and CD28- T cell subsets were purified from the peripheral blood of HIV-1 infected (n = 3) and noninfected (n = 3) individuals and then cultured with monocyte-derived dendritic cells for 24 h. The dendritic cells were then activated with lipopolysaccharide (LPS) or left without further treatment for 24 h and thereafter the expression of HLA-DQ, CD83 and CD86, three markers of dendritic cell activation, was analyzed. Interestingly, the presence of both CD28+ and CD28- T cell subsets induced the upregulation of these molecules (Fig. 1c) and none of the T cell populations interfered with LPS-induced maturation (data not shown). The lack of dendritic cell suppression by CD28- T cells is in striking contrast to the effect of the in-vitro generated CD28- T suppressor cells characterized by an inhibitory potential for phenotypic dendritic cell maturation ^[9]. The effect of purified CD28+ and CD28- T cells was also tested on cytokine production by dendritic cells. Instead of any inhibitory effect of CD28- T cells, this population strongly augmented LPS-induced IL-12 and tumor necrosis factor (TNF) production of dendritic cells, whereas IL-10 secretion was not influenced by the presence of any T cell subsets (Fig. 1d).

Fig. 1

The most characteristic function associated with CD28- suppressor T cells generated in vitro is their profound ability to inhibit proliferation of other T cells either by dendritic cell modulation or through directly interfering with T cell activation ^[9,10]. We analyzed whether CD28- T cells isolated from HIV-1 infected (n = 3) or noninfected (n = 3) individuals were able to influence the proliferation of naïve T cells triggered by dendritic cells and anti-CD3. The CD28-, CD28+ and CD28+CCR7- peripheral blood T cells were cultured with dendritic cells for 24 h and, thereafter, purified naïve T cells were added to the cultures and triggered by 2 μg/ml of anti-CD3 mAbs. CD28+ or CD28- T cells preincubated with dendritic cells minimally influenced the proliferation of third party T cells. The ratio of T cells undergoing cell divisions changed 0.9 ± 0.1-fold and 1.0 ± 0.1-fold (in the presence of CD28+ and CD28- T cells, respectively, for HIV-infected donors) or alternatively, 1.1 ± 0.2-fold and 0.8 ± 0.1-fold (in the presence of CD28+ and CD28- T cells, respectively, for noninfected donors) as compared with untreated dendritic cells. The slight reduction in the number of proliferating cells detected in the presence of CD28- T cells of noninfected individuals was similarly observed when the CD28+CCR7- memory subset was added to the dendritic cell cultures (Fig. 1e), possibly indicating negative feedback mechanisms delivered by previously antigen-encountered T lymphocytes, which might compromise any further T cell activation. This inhibitory effect was, however, not CD28- T cell-specific.

Importantly, CD28- T cell mediated dendritic cell activation and the lack of any CD28- T cell specific suppression of T cell proliferation were equally observed in the case of both HIV-1 infected and noninfected donors. These results suggest that the accumulation of CD28- peripheral T cells during HIV-1 infection may not lead to increased frequency of T suppressor cells or precursors prone to readily differentiate into T suppressor cells in the presence of dendritic cells and activated T cells. We observed that the CD28- T cells elicited dendritic cell stimulatory effects, suggesting that the accumulation of CD28- T cells in HIV-1 infected individuals, instead of leading to dendritic cells or T cell suppression, may contribute to accelerated inflammatory reactions and immune activation.

Acknowledgements

The study was supported by grants received from the Swedish MRC, the Swedish International Development Agency (SIDA-SAREC). Bence Rethi was supported by the Hungarian State Eötvös Fellowship.