Metabolic reprogramming of naïve regulatory T cells by IL-7 and IL-15 promotes their persistence and performance upon adoptive transfer

Abstract

Tregs for adoptive therapy are traditionally expanded ex vivo using high doses of IL-2. However, the final Treg product has limited survival once infused in patients, potentially affecting therapeutic effectiveness. Here, we tested a novel expansion protocol in which highly purified naïve Tregs were expanded with a combination of IL-7 and IL-15, in the absence of IL-2. The final Treg product was enriched with cells displaying an immature CD45RA+CD62L+CD95+ phenotype, reminiscent of conventional memory stem T cells. The combination of IL-7 and IL-15 confers Tregs a glycolytic metabolism and improved metabolic fitness, characterized by an increased capacity to adapt metabolism according to glucose and oxygen availability. Tregs expanded with IL-7 and IL-15 showed longer persistence and an improved capacity to control xeno-GvHD in NSG mice. This work suggests that metabolic reprogramming induced by IL-7 and IL-15 provides better Treg performance for adoptive therapy.

Fig. 1. Treg express a functional IL-15 receptor alpha chain (CD215) and proliferate in response to a combination of IL-7 and IL-15.

A Representative FACS plot showing the expression of CD215 in gated Treg (CD4+CD25^brightCD127^low) and Tconv (CD4+CD25^lowCD127^high) from a PBMC sample. The graph (right panel) shows the mean florescence intensity (MFI) of CD215 in Treg and Tconv classified as naïve and memory subsets (n = 4). B Representative FACS plot of phosphorylated STAT-5 after 1 min of incubation of Treg and Tconv with 10 ng/ml of rhIL-15. The graph (right panel) shows the MFI of phospho-STAT-5 in Treg and Tconv divided into naïve and memory subsets after incubation with IL-15 (n = 4). C Fold expansion of StM-Treg and IL-7/15M-Treg. D Final cell yield of StM-Treg and IL-7/15M-Treg. E Viability measured as trypan blue exclusion of StM-Treg and IL-7/15M-Treg. F Percentage of annexin V+ cells after apoptosis induction with an agonistic anti fas antibody in StM-Treg and IL-7/15M-Treg. G Changes in the number of viable cells after first expansion, contraction due to withdrawal of cytokines and microbeads, and secondary expansion of residual cell. H Percentage of suppression of proliferation of CD8+ (upper panel) and CD4+ (lower panel) allogenic T cells at different Treg/Tcell ratios. The Wilcoxon matched paired t-test with the Dunn–Bonferroni post hoc test was used for comparisons. *P < 0.05, **P < 0.01, ***P < 0.001, ns, P > 0.05.

Fig. 2. IL-7/15M-Treg preserves an immature phenotype and display a distinct gene expression profile.

A Representative FACS plot showing the expression of CD62L and CD45RA in StM-Treg and IL-7/15M-Treg. B Percentage of StM-Treg and IL-7/15M-Treg with a Tn, Tscm, Tcm, Tem and Temra phenotype. C Expression of genes that defines the Treg subset in StM-Treg and IL-7/15M-Treg. D Expression of genes coding for transcription factor FoxP3, T-bet, Gata4 and Ror-gamma-C. E Expression of genes that were differentially expressed between StM-Treg and IL-7/15M-Treg. F ssGSEA enrichment score of the signature of IL-7/15M-Treg (E) in Treg isolated from cord blood and adult peripheral blood. G Expression of genes involved in metabolic pathways. The Wilcoxon matched paired t-test with the Dunn–Bonferroni post hoc test (B) or Wilcoxon Rank Sum test (F) were used for comparisons.

Fig. 3. IL-7/15-Treg display an efficient metabolism based on glycolysis.

A Glucose and mitochondrial dependencies and relative FAO/AAO and glycolytic capacities in StM-Treg and IL-7/15M-Treg after 14 days of in vitro expansion. B Dynamic uptake of the fluorescent glucose analog 2NBDG in StM-Treg or IL-7/15M-Treg. Parameters of the two curves were calculated and compared. C Lactate concentration (mMol/ml) in the supernatant of 106 StM-Treg or IL-7/15M-Treg cultured for 24 h. D Mitochondrial mass measured as fluorescence of the MitoGreen dye in freshly isolated Treg and in StM-Treg or IL-7/15M-Treg. The Wilcoxon matched paired t-test (A–D) was used for comparisons. E Confocal images showing the shape of mitochondria in StM-Treg and IL-7/15M-Treg after 14 days of in vitro expansion. F Oxygen consumption rate (OCR) and G extracellular acidification rate (ECAR) in StM-Treg and IL-7/15M-Treg measured by Seahorse technology at basal and after the injection of oligomycin (1.5 μM) and a mixture of rotenone and antinomycin (0.5 μM each). H Total ATP production and I ATP produced by glycolysis (Glyco ATP) and oxidative phosphorylation (Mito ATP) measured by Seahorse technology with the real-time ATP rate assay. Data are presented as mean ± SEM, n = 8/group; statistical analyses were performed by t-test (H) and two-way Anova (I). **P < 0.01, ns P > 0.05.

Fig. 4. IL-7/15-Treg persist longer in NSG mice and are more protective against xeno-GvHD.

Heatmaps showing A GFP+Treg (number/ml), B CD3+ Tconv (number/ml), C GvHD score (1–6, Supplementary Table 4B) and D weight (% relative to day 0) in NSG mice infused with 10⁷ PBMC and 5 × 10⁶ Treg. E Kaplan–Meier survival curve comparing NSG mice receiving human PBMC alone or in combination with expanded Tregs (16 animals for each group). Statistical analysis was performed using the using the Log-rank (Mantel–Cox) test. F Pictures of spleens collected from untreated, StM-Treg- and IL-7/15-Treg-treated mice NSG mice. G Representative H&E in tissue sections obtained from spleen and liver of treated mice at day 35. H Glucose and mitochondrial dependencies and relative FAO/AAO and glycolytic capacities in StM-Treg and IL-7/15M-Treg retrieved from NSG mice spleen 7 days after injection. I Puromycin fluorescence in StM-Treg and IL-7/15M-Treg both after 14 days of in vitro expansion, and 7 days after injection in NSG mice. The Wilcoxon matched paired t-test (A–D) was used for comparisons.