Responsiveness to IL-7 in circulating naïve and memory Treg subsets
As for conventional T cells, also circulating Treg can be categorized according to surface phenotypes corresponding to naïve or memory subsets. According to the expression of CD45RA, CD62L and CD95 we determined the relative proportion (%: median, IQR) of Treg naïve (10.1, 7-16.3 Treg-n; CD45RA+CD62L+CD95-), Treg stem cells memory (1.2, 0.6-2.6 Treg-scm, CD45RA+CD62L+CD95+), Treg central-memory (44.8: 32.5-59.6 Treg-cm, CD45RA-CD62L+CD95+), Treg effector-memory (42.5, 35.8-52.1 Treg-em, CD45RA-CD62L-CD95+) and Treg effector-memory re-expressing CD45RA (0.4, 0.2-0.6 Treg-emra, CD45RA+CD62L-CD95+) in peripheral blood samples from 16 adult healthy subjects (Figure 1A and B). In parallel we determined the relative proportion of the same subsets in circulating conventional CD4+ T cells (Supplementary Figure 1A and 1B). Using this phenotypic classification we found all the subsets described for conventional T cells, including Treg with a memory stem cell phenotype that have never been described before. Since Treg are defined as CD127low cells (14), we measured (MFI; median, IQR) the expression of CD127 in all the Treg subsets identified and compared it to the one of conventional bulk CD4+ T cells (3205, 2064-4692) and CD19+ B cells (57, 49-69) that do not express CD127 (15) (Figure 1C). Treg-n showed the highest expression of CD127 (480, 316-611) followed by Treg-scm (432; 304-515), Treg-cm (328; 246-390), Treg-em (173, 122-236), and Temra (132, 101-159). We next incubated Treg with IL-7 (10ng/ml for 1 minute) and measured the phosphorylation of STAT5 to determine whether CD127 was functional and sufficiently expressed to elicit intracellular signaling (Figure 1D). We compared STAT-5 phosphorylation (MFI; median, IQR) to the one of conventional CD4+ T cells (8056, 6730-10526) and CD127- B cells (167, 126-168). Treg-n showed the strongest STAT-5 phosphorylation (3386, 2794-4629) followed by Treg-scm (2540, 1972-2817), Treg-cm (1500, 950-1827), Treg-em (779, 364-909), Treg-emra (576, 332-776). These data show that Treg were responsive to IL-7, especially those with a naïve phenotype or a stem cell memory phenotype, even though their response in terms of STAT-5 phosphorylation was far less intense than the one of conventional CD4+ T cells. Based on these observations, we tested the capacity of IL-7 to trigger Treg expansion in comparison or in combination with IL-2 (Figure 1E). Treg showed a limited fold expansion (median, IQR) in response to IL-7 (10, 8-11) while expanded vigorously in the presence of IL-2 (85, 68-95). The combination of IL-7 and IL-2 induced an intermediate level of Treg expansion (54, 34-61) that was unexpectedly less than the one with IL-2 alone. We interpreted these data as the result of competition between IL-2 and IL-7 for the common gamma chain providing an intermediate proliferation rate instead of an additive or synergistic effect. Supporting this hypothesis, we have previously shown that blocking the IL-2 receptor alpha with a monoclonal antibody increase the availability of the common gamma chain and render conventional T cells more responsive to IL-7 (16). To have a better insight into this competition mechanism, we performed experiments to determine the binding of IL-2 FITC when IL-7 was present in the culture or the binding of IL-7 FITC when IL-2 was present in the culture. Results showed that in the presence of increasing concentration of IL-7 the binding of IL-2 to Treg progressively decreased with a reduction of 70% in the presence of 104 pg/ml of IL-7 (Figure 2A). On the other hand, the effect of IL-2 on the binding of IL-7 FITC to Treg was less pronounced with a reduction of 45% in the presence of 100 UI of IL-2 (Figure 2B). To formally prove that IL-7 interferes with the association of CD25 and CD132 (and possibly CD122) to form the high affinity IL-2 receptor, we developed a co-precipitation assay to measure receptor association by flow cytometry (Figure 2C) similar to the one we previously shown (16). Treg pre-treated with increasing concentrations of IL-7 showed a progressive reduction of the association of the CD25/CD132 complex, suggesting that IL-7 decreased the availability of CD132 to form the high affinity IL-2 receptor complex. Taken together these data indicate that it is possible to elicit a significant expansion of Treg using a combination of IL-7 and IL-2 and that Treg with a naïve and stem cell memory phenotype are the subset with the strongest responsiveness to IL-7. Based on these data we decided to expand sorted CD45RA+CD62L+ Treg in the presence of IL-7 and IL-2.