Molecular cues of gamma delta T cells

An important part of the adaptive immune system is the population of lymphocytes called T cells. Normally these cells require an interaction between a foreign antigen presented by an antigen presenting cell to their T cell receptor, as this will induce a specific effector function and in turn induce a memory response against the specific object. However, a small proportion of T cells called γδ T cells are not depended on this peptide – T cell receptor interaction. Instead, they are functionally programmed during neonatal development into one of two effector functions that can produce either interleukin (IL)-17 or interferon (IFN)-γ. The IL-17 producers are characterized based on their expression of the transcription factor RORγt and the surface markers CCR6 and high expression of CD44 (γδT17), whereas the IFN-γ producers express T-bet and CD27 (γδT1). After development, the IL-17 producing γδ T cells migrate to the tissue and occupy a large proportion of the T cells in barrier sites such as the dermis and intestine. Situated in the barriers site, they can utilize their unique functionality to quickly respond to cytokines like IL-7 and IL-23. These cytokines are known to signal through the three STAT molecules: 3, 4 or 5 in T cells that regulate cytokine signaling. The importance of these cytokines and STATs have been thoroughly investigated for IL-17 producing CD4s; however, the full extend as to how γδT17 cells utilize these STAT proteins in maintaining their functionality is not fully understood yet. Due to their important role in driving inflammation overexpression of STATs or continuous activation have been associated with several types of dysfunctions in mammals, so several regulatory mechanisms are in place to prevent this. Such mechanisms consists of the the tyrosine phosphatases that can prevent activation of STAT molecules by competing with upstream components of the pathway. The two tyrosine phosphatases SHP-1 and SHP-2 have been associated with the immune checkpoint inhibitor BTLA that is known to regulate γδT17 cells in a negative feedback loop with STAT5 and IL-7. But their direct role in γδT17 activation and development is still to be investigated. STAT expression levels have also been shown to be affected by metabolites such as retinoic acid (RA). RA is a vitamin A metabolite that regulates gut homing properties on T cells; however it is alost important for maintaining the balance between different T-helper subsets. Extrinsic RA has been shown to affect IL-17 producing γδ T cells both in-vivo and in-vitro, however how intrinsic RA affect the development and inflammatory response for γδT17 cells is not known.
To approach these issues, we generated a number of transgenic mouse strains targeting all RORγt⁺ positive cells using a cre recombinase system (RORγt^cre). We crossed this strain to mice with floxed versions of STAT3^flox/flox, STAT5^flox/flox, SHP-1^flox/flox, SHP-2^flox/flox or RARdn^flox/flox, generating strains with either deleted- or inactive versions of the genes. This was done to investigate their function in maintaining development and driving γδT17 cell inflammation in various in-vivo models. As the STAT5^flox/flox deletion affects both STAT5A and STAT5B we sought to investigate the contribution of each of the paralogs and used mice expressing a hyperactive mutant version of either STAT5A or STAT5B. Finally, for the role of STAT4 we used a full STAT4^KO mice strain in which all cells lack STAT4 signaling.
We found that absence of STAT5 completely diminished the IL-17 producing subset of γδ T cells in all tissue. Interestingly, although the thymic development of RORγt⁺ γδ T cells were unaffected by the deletion, instead STAT5 is required for the early expansion during the neonatal window in the lymph nodes, intestine and other peripheral tissue. In the intestine, a previously undescribed population of RORγt⁺ T-bet⁺ γδ T cells were identified and found to be depended on the expression of both STAT3 and STAT5. Using the hyperactive mutants, we found that hyperactive STAT5A preferentially expanded γδT17 cells whereas the hyperactive STAT5B expanded the γδT1 population. When we exposed RORγt^creSTAT5^flox/flox mice to EAE (a model for multiple sclerosis) we found that these mice were completely protected against disease.
A similar observation was made using the RORγt^creRARdn^flox/flox mouse strain as these were also protected against EAE induced inflammation. Interestingly, we found that the early expansion of the Vγ4⁺ population was reduced in both lymph nodes and spleen after disease induction. Investigation of the brain 21 days post induction revealed that the number of infiltrating γδT17 cells and pathogenic Th17 were greatly reduced compared to wild type mice. Furthermore, surface expression of the integrin α4, a molecule associated with migration to the CNS, was reduced on both splenic CD4s and γδT17 cells. During homeostasis we found that the absence of functional retinoic acid signaling led to a defective proliferative profile only in the Vγ4⁺ γδT17 cell population compared to their wild type counterparts. Furthermore, we observed a failed negative regulation of CD4 T cells, as the total number of activated CD4s (expressing high levels of CD44) were increased along with their IL-17 and IL-22 production.
The role of STAT3, STAT4 and SHP-1/SHP-2 were all investigated in the context of skin inflammation using a psoriasis like model. We found that STAT3 expression is crucial in γδT17 cells as these mice were completely protected from inflammation, likely due to the inability to respond to IL-23 that is required for γδT17 activation in this model. This led to reduced expansion and production of IL-17 and IL-22 from γδT17 cells compared to wild type controls. Interestingly, although IL-23 is also able to signal through STAT4, absence of this transcription factor had no effect on γδT17 development or disease progression. It did however reduce IL-17F expression by γδT17 during inflammation.
Surprisingly, when we removed either SHP-1 or SHP-2 alone or crossed the two strains for a double KO strain (RORγt^creSHP-1^flox/floxSHP-2^flox/flox) we found no defect during homeostasis or in their inflammatory response. This outcome was a surprise but is likely due to the amount of tyrosine phosphatases present in the cytoplasm that can compensate for the absence of SHP-1 and SHP-2.
Collectively, these findings highlight the important role of certain STATs in γδT17 cells as STAT5 is crucial for the neonatal expansion of γδT17 cells in peripheral tissue while absence of STAT3 diminishes γδT17 cell driven skin inflammation. While the STAT5A and STAT5B paralogs each preferentially differentiate either the γδT17 cells or γδT1 cells. We also show that the metabolite retinoic acid has a specific role in regulating both Vγ4⁺ γδT17 and CD4 T cells, and the absence of RA signaling leads to protection against EAE.