T Follicular Regulatory (TFR) Cell Deficiency during Chronic Graft-Versus-Host Disease Is Improved By Low-Dose IL-2 Therapy

Background: Chronic Graft versus Host Disease (cGVHD) reflects the inability to achieve immune tolerance after allogeneic stem cell transplantation (allo-SCT) resulting in the persistence of allo and auto-reactive T cells and B cells. We previously showed that regulatory T cells (Treg) are decreased in patients with cGVHD and may be rescued by low dose IL-2 therapy resulting in improved Treg number and survival and clinical improvement of cGVHD symptoms. More recently, we showed that circulating T follicular helper cells (cTFH) have increased function during cGVHD, thus stimulating B cells to produce auto and allo-antibodies involved in cGVHD. T follicular regulatory cells (TFR) regulate the germinal center (GC) reaction and suppress B cell differentiation and immunoglobulin secretion. TFR reside primarily in lymphoid tissues but also circulate in small numbers. In the present study we characterized the numbers and function of blood CD3+CD4+CD45RA-CD25hiFoxP3+CXCR5+ TFR in healthy donors (HD; N=18) and allo-SCT patients (N=64).

Results: Patients undergoing allo-SCT showed a significant reduction of circulating TFR (cTFR) compared to HD (median: 0.08% vs 0.34% of CD4+T respectively, p<10^-4). Moreover, patients with active cGVHD (N=37) had significantly lower cTFR frequency compared to patients with no cGVHD (N=15) (median: 0.034 vs 0.18% of CD4+, p=0.0012) and resolved cGVHD (N=12) (median: 0.034 vs 0.14, p=0.02). Interestingly, cTFR values tended to be negatively associated with cGVHD grade (median cTFR %: 0.11 vs 0.029, p=0.004 for mild vs moderate; 0.11 vs 0.012, p=0.005 for mild vs severe; 0.029 vs 0.012, p=0.2 for moderate vs severe). The numeric reduction of cTFR in cGVHD patients appears to be a consequence of decreased survival as attested by lower BCL-2 expression (median: 1831 vs 1939, p=0.037 in active cGVHD and without cGVHD, respectively). This favors the survival of TFH effector cells (CD4+CD45RA-CXCR5+ excluding CD25hiFoxP3+) as measured by the ratio of expression of BCL-2 in effectors/regulators in active compared to no cGVHD (BCL-2 E/R ratio: 1.26 vs 1.14, p=0.0013 respectively). Similar to tonsil TFR, blood TFR exhibit suppressive functions in in vitro assays. Purified TFR were co-cultured with TFH and memory B cells at different ratios. The addition of TFR was associated with a significant reduction of plasmablast (PB) generation (CD27hiCD38hiCD19+ cells) (Figure 1A, t-test: p=0.0087) and IgG production (Figure 1B, t-test: p=0.0087). Human cTFR in vivo display similar suppressive function toward B cells, as analysis of active cGVHD patients revealed an inverse correlation between TFR and PB frequency in the peripheral blood (Spearman test: r = -0.49, p=0.003) (Figure 1C). As IL-2 has been shown to enhance regulatory T cells, we assessed its impact on the GC reaction in vitro . Purified TFH and memory B cells from HD were co-cultured at different doses of IL-2 from 0 to 100 IU/mL, simulating a germinal center reaction in vitro . Low dose IL-2 starting at 10 IU/mL was associated with a suppressive effect on PB generation and IgG production that peaked at 50 IU/mL reflecting the high expression level of IL-2 high affinity receptor (CD25). We also evaluated the capacity of IL-2 to restore in vivo cTFR:cTFH ratio using mass cytometry (CyTOF) in a cohort of patients receiving IL-2 (1x10⁶ IU/M²/day) for 12 weeks as previously described (Koreth et al, Blood 2016). We observed an improvement of blood cTFR:cTFH ratio over the course of IL-2 therapy, with a maximum ratio observed at 2 weeks, and subsequently decreasing to reach baseline values after stopping IL-2 (Figure 2).

Conclusion: Our results confirm that blood TFR exert a suppressive function as evidence by an inverse correlation with PB and in vitro suppression of memory B cell Ig production. This suppressive subset is deficient in patients with cGVHD due to altered TFR survival, thus favoring imbalance toward TFH effectors which may promote a GC reaction in vivo resulting in PB and memory B cell pathogenic Ig production. By enhancing TFR, low-dose IL-2 therapy promotes effective control of the GC reaction.

View largeDownload slide

View largeDownload slide

Close modal