3a). At the CD8 T-cell level, a significant amount of AICD in effector memory and effector subsets was observed at baseline, while naïve and central memory subsets were less sensitive to AICD (Fig. 3b). Under ART, the amount of AICD decreased in all CD8 subsets from week 4 to week 24, while the expression of Ki67 in all subsets was low at baseline and slightly decreased under ART (Fig. 3b).

For unknown reasons, the amounts of AICD increased in most subsets at week 48, while immune activation was still suppressed. Altogether, taking into consideration the CHIR-99021 cost balance between priming for AICD and homeostatic proliferation, these observations may account for the differences in CD4 and CD8 T-cell subset kinetics of restoration under enfuvirtide therapy (Fig.

1a). The effect of enfuvirtide-based therapy on parameters affecting HIV entry, i.e. CCR5, chemokine (C-X-C chemokine) receptor 4 (CXCR4) and chemokines, was evaluated. A progressive decrease in the percentage of CCR5-expressing cells was detected in CD4 and CD8 T cells from all RP patients, affecting all four CD4 T-cell subsets and leading to very low CCR5 expression at week 48 in these subsets (Fig. 4a). The proportions of CXCR4-expressing CD4 and CD8 T cells were quite high at 5-Fluoracil datasheet baseline, slightly decreased until week 12 and then returned to baseline values at week 48. Considering the different subsets, CXCR4 expression was high in naïve and central memory CD4 T cells and did not change during the 48-week follow-up period. Regarding effector memory and effector CD4 T-cell

subsets, almost 40% expressed CXCR4 at baseline, and this percentage of CXCR4+ cells decreased until week 24 (Fig. 4b). Similar observations were obtained Astemizole for total CD8 T cells (Fig. 4b) and CD8 T-cell subsets (not shown). Importantly, the decrease in the proportion of CCR5-expressing CD4 T cells under enfuvirtide-based therapy was strongly correlated with, on the one hand, the activation state of CD4 T cells (i.e. CD38 or HLA-DR expression) and, on the other hand, plasma VL. Furthermore, the percentage of CCR5+ CD4 T cells was correlated with disease evolution, as estimated from CD4 cell counts (Fig. 4c). Regarding CXCR4 expression on CD4 T cells, no correlation was found with either the VL or CD4 T-cell numbers (Fig. 4c). To identify the key cytokines and chemokines modulated during enfuvirtide-based therapy, we used MAP technology on patients’ sera. Figure 5 shows that the levels of the CCR5 ligands macrophage inflammatory protein (MIP)-1α and MIP-1β dropped significantly from week 12. In contrast, the high levels of RANTES persisted. Circulating MIP-1α was correlated with the VL (r=0.43; P=0.007), but not with CD4 cell counts. Other chemokines, such as monocyte chemotactic protein (MCP)-1 and MIG, also dropped (Fig. 5), and their levels correlated positively with VL (P=0.02 and 0.