Recent studies have suggested that regulatory T-cells (Tregs) play a critical role in the pathogenesis of follicular lymphoma (FL). We have previously shown that the Tregs infiltrating FL lymph nodes (FLN Tregs) are quantitatively and qualitatively different than those infiltrating normal and reactive nodes (NLN, RLN, respectively). To gain insight into how such Treg populations differ, and how FLN Tregs may impact FL progression and response to therapy, we performed RNA sequence analyses on flow sorted Tregs (CD3+CD4+CD25brightCD127−) from NLN (n=10), FLN (n=12) and RLN (n=7).
Using the RefSeq annotated set of mRNA we first identified 11686 annotated genes that were expressed in each of the Treg groups. Using pairwise unequal variance t-tests at p<0.01, we identified 798 genes that differed between NLN and FLN, 589 that differed between NLN and RLN, and 475 genes that differed between FLN and RLN. The transcriptome profile of the FLN Tregs differed most from that of NLN Tregs, with 281 genes increased in FLN relative to NLN, and 517 genes decreased in FLN Tregs relative to NLN Tregs. In contrast, 404 genes were increased and 70 genes decreased in RLN Tregs relative to FLN Tregs suggesting that differences in the transcriptome profile between NLN and FLN Tregs are due, in part, to inflammatory signals shared between FLN and RLN. Tregs, which express FOXP3, are a heterogeneous population which share a “core suppressive module” of CD25 and CTLA-4, but which adapt to microenvironmental signals by co-opting/or up-regulating the canonical transcription factor (TF) of the T-cell population the Treg needs to suppress. This results in modulation of the Treg chemokine receptors in such a fashion that allows these Tregs to traffic to and suppress the specific T-cell population defined by their TF expression. Relative to NLN, FLN Tregs have similar expression of FOXP3 and CD25 but greater expression of CTLA-4 (confirmed at the protein level), which may account, in part, for the greater suppressive activity of FLN Tregs compared to NLN Tregs. Expression of other genes that differ between NLN and FLN which encode for proteins that modulate Treg suppression include IL-10 (up in FLN relative to NLN), S1PR1 and ENTPD1 (both down in FLN relative to NLN), which all favor greater suppression by FLN Tregs. In contrast LGALS1 and NT5E expression (both down in FLN relative to NLN) would favor less suppression by FLN Tregs. Relative to NLN, FLN Tregs have greater BCL-6 expression, consistent with a population that suppress follicular helper T-cells (Tfh), which support normal germinal center (GC) and FL-B cell survival. Consistent with their increased expression of BCL-6, FLN Tregs have greater expression of CXCR5, CXCR4 and CXCL13 and decreased expression of CCR7 compared to that of NLN Tregs, a repertoire associated with GC trafficking. The significant lower expression of both S1PR1 and SELL (CD62L; p<0.000002, and fold changes of −9.0 and −2.8, respectively) that is found in FLN Tregs compared to NLN Tregs is consistent with the greater nodal retention. The lower expression of RORc found in the FLN Tregs would predict for decreased IL-17 secretion and their elevated levels of STAT3 which would support the ability of the FLN Tregs to inhibit Th17 cells.
In conclusion, the comparison of the transcriptome profiles of FLN and NLN Tregs; a) suggest that the observed differences are shaped, in part, by inflammatory signals also present in RLN; b) lay the foundation of a hypothesis to explain the molecular basis for their greater immunosuppressive capacity; and c) suggest that FL Tregs are ‘poised’ to localize in the GC, be strongly retained in the lymph node and have enhanced potential to suppress Tfh and Th17 cells. Such findings offer new insights into how FL Tregs may contribute to the biology of these B cell lymphomas.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.