Reverse Translation and Extension of Human Tyk2 Biology to a Murine Model

Abstract

Autoimmune disease is a broad term that encompasses over 80 different syndromes, and collectively affects greater than 23.5 million Americans (Rosenblum et al., 2012). Since cytokines play a pivotal role in inflammation and autoimmunity, more defined treatment regiments targeting selective cytokines and/or their signaling pathways continue to be explored. Type I and type II cytokine receptors lack ability to produce kinase activity relying on the JAK family proteins for internal signaling in response to receptor binding. Tyk2 is a JAK family member that is downstream of multiple cytokines that modulate inflammation and Tyk2 variants show strong GWAS correlations to protection from multiple autoimmune diseases. P1104A is a homozygous SNP which protects from various autoimmune diseases without the adverse effects associated with pan-JAK inhibition. Characterization of immune cells from patients with this SNP indicates loss of Type 1 IFN, IL-12 and IL-23 STAT activity but no change in the IL- 6, IL-10 and IL-13 signaling pathways (Dendrou et al., 2016). In an effort to develop a translational model of the P1104A SNP, we have generated the genetically modified mice (GEM) carrying the correlative of the P1104A mutation (P1124A GEM). In proximal assays P1124A GEM responded to IFNα, IL-12, IL-6 and IL-10 STAT phosphorylation in a similar manner as P1104A PBMCs. A significant variation to the human variant was seen with IL-23 stimulation where STAT activity was unaffected. Cytokine driven proximal activity was consistent with functional cytokine stimulated mediator production in mice. Published data indicates that when PBMCs from P1104A patients are stimulated with CD3ε/CD28 they shift towards a Th2 GATA3+, IL-4 producing phenotype (Couturier et al., 2011; Shimoda et al., 2000). Our data indicates that CD4 T cells loss of IFNγ, IL-10 and IL-21 cytokine production when cells were polarized to Th1 or Th17 phenotypes, respectively, contributes to the shift in P1124A mice. When Th1 differentiated, CD4 cells trend towards a Th2 profile, expressing low levels of GATA3. To further delineate the effect of P1124A, they were evaluated in a delayed-type hypersensitivity (DTH) model, in which they showed an attenuated response. Animal modeling suggest that GEM P1124A is protected from Th1 CD4 dependent DTH inflammation. We also measured the activity of the P1124A variant and compared it to other Tyk2 GEMs including: Tyk2 KO, the K950A kinase-dead knock-in (KDKI) and the Y1074F/Y1075F trans-phosphorylation mutant and have determined that the P1124A, and by extension possibly the P1104A SNP does not function by deficiency of Tyk2 activity, loss of auto-phosphorylative ability nor by loss of TYK2 trans-phosphorylative activity.