Expansion, phenotype and suppressive capacity of naïve Treg expanded with IL-2 or a combination of IL-2 and IL-7.
We then compared two methods for expansion of highly purified (Supplementary figure 2 A, B and C) sorted CD45RA+CD62L+ Treg (Treg-n/Treg-scm). A standard method (StM: aanti CD3/CD28 microbeads at 1:1 ratio in the presence of 100UI of rhIL-2) was compared to the IL-7 method (IL-7M) similar to the StM but with the addition of 10ng/ml of rhIL-7. After 14 days Treg expanded more with the StM (fold expansion: 68, 61-93) than with the IL-7M (fold expansion: 49, 47-59,P =0.021) resulting in a 28% decrease in the final cell yield using the IL-7M (Figure 3A). Surface phenotype analysis showed that Treg expanded with the IL-7M were highly enriched in CD45RA+CD62L+CD95+cells (Figure 2B). Dividing Treg into subsets according to the expression of CD45RA, CD62L and CD95 we observed that Treg expanded with the two methods showed a similar percentage (median, IQR) of Treg-n (StM 5.4, 3.4-8.1 vs IL-7M 3.5, 2.4-5.5, p=0.234), Treg-cm (StM 39, 29.7-45.7 vs IL-7M 46, 40.1-57, p=0.458) and Treg-emra (StM 1.1, 0.8-1.2 vs IL-7M 1, 0.3-1.2, p=0.88) after 14 days of expansion (Figure 2C). Treg expanded with the IL-7M were highly enriched in Treg-scm (StM 4.3, 2.6-6 vs IL-7M 23.2, 17.9-31, p=0.0078) and have less Treg-em (StM 51, 40.5-63.5 vs IL-7M 23.5, 11.7-39, p=0.0375) (Figure 2C). After 14 days Treg cultures were magnetically depleted of anti CD3/CD28 microbeads and cultured for additional 3 days in the absence of cytokines (resting period). After resting expanded Treg showed a preservation of the relative percentages of Treg subsets after resting. Treg expanded with the StM or the IL-7M methods were tested in a suppression assay against CD8+ T cells, showing a significant reduction in the suppressive activity of Treg expanded with the IL-7M compared to Treg expanded with the StM (Figure 2D). Since we have shown that IL-7 can reduce the Treg suppressive activity but the suppressive function is re-established once IL-7 is removed, we tested Treg after a 3 day resting period, which resulted in a full recovery of the suppressive function after removal of IL-7. To determine whether the kinetic of proliferation can impact on the differences in the phenotype of expanded Treg we analyzed the phenotype of proliferating Treg according the dilution of CFSE after 5 days of expansion. Treg expanded with the StM showed a homogeneous proliferation pattern in which the large majority of Treg performed several cell cycles and lost the expression of CD45RA (Figure 3E and 3F). Treg expanded with the IL-7M can be divided in two groups, of which one performed multiple cell cycles and lost CD45RA, and a second one which performed a limited number of cell cycles and preserves the expression of CD45RA. Sorted Treg that performed more than two cell cycles from both the StM and the IL-7M were committed to differentiate predominantly into Treg-cm and Treg-em (Figure 3G). On the other hand, sorted Treg that performed less that two cell cycles from both the StM and the IL-7M were committed to differentiate into Treg-n and Treg-scm. These results suggest that a population of slowly proliferating Treg that is predominant in the IL-7M (but very scarce in the StM) is responsible for the abundance of Treg with a stem cell memory phenotype found in the final Treg product after 14 days of culture. Taken together these data suggest that the presence of IL-7 reduces the proliferation rate (and the final cell yield) but preserves a poorly differentiated phenotype. Moreover, the reduced suppressive function can be recovered after a brief culture in the absence of IL-7.