STAT3 gain of function: Too much of a good thing in the skin!

JAK/STAT pathways are activated following engagement of surface receptors by cytokines, growth factors or hormones, and relay signals from the extracellular environment to the nucleus, resulting in transcriptional regulation of myriad target genes that influence cell behavior (Philips et al., 2022). Within the immune system, JAK/STAT signaling has fundamental non-redundant roles in leukocyte development, activation, differentiation, and function. There is no greater evidence of this than the discovery of germline mutations in genes encoding individual JAKs, STATs, and related cytokines and receptors that manifest clinically as immune dysregulatory conditions (Casanova et al., 2012; Philips et al., 2022; Tangye et al., 2017). Currently, inborn errors of immunity (IEI) have been found to result from mutations in JAK1, JAK3, and TYK2; STAT1, STAT2, STAT3, STAT4, STAT5B, and STAT6; IL2RA, IL2RG, IL2RB, IL6R, IL6ST, IL7R, IL10R1/R2, IL12RB1/RB2, IL17RA/RC, IL21R, IL23R, IFNGR1/R2, and IFNAR1/R2 (Casanova et al., 2012; Tangye et al., 2017, 2022). Such mutations cause disease by partial or complete loss of function (LOF), negative dominance, haploinsufficiency or gain of function (GOF) of the encoded protein.

In IEI resulting from loss of expression and/or LOF of a component of JAK/STAT signaling, mechanism(s) of disease pathogenesis are intuitive or can be efficiently determined by appropriate cellular and molecular studies. For instance, mutations in IL2RG, IL7RA, or JAK3 dramatically impede T cell development, thus causing severe combined immunodeficiency, rendering affected individuals susceptible to a broad range of infectious diseases (Casanova et al., 2012; Tangye et al., 2017). Similarly, dominant negative STAT3 mutations disrupt generation of Th17, T follicular helper (Tfh), memory B and CD8⁺ T cells, thereby explaining susceptibility to recurrent fungal infections (Th17 cells), impaired humoral immunity (Tfh, memory B cells), and frequent re-activation of herpes infections (Mackie et al., 2023; Tangye et al., 2017). However, mechanisms by which mutations that result in enhanced protein activity cause disease are often less intuitive and even elusive. An example of this is the clinical syndrome resulting from STAT3 GOF variants.

STAT3^GOF syndrome is a multi-organ immune dysregulatory condition characterized by lymphoproliferation, organomegaly, lymphadenopathy, cytopenias, polyendocrinopathies, enteropathy, short stature, and recurrent infections (Flanagan et al., 2014; Leiding et al., 2023). Based on the requirement for STAT3 in generating Th17 cells (Tangye et al., 2017) and antagonism between STAT3 and STAT5 in generating regulatory T cells (Tregs) (Mackie et al., 2023), it was proposed that immune dysregulation in STAT3^GOF syndrome resulted from excessive production of Th17 cells and reduced Tregs (Milner et al., 2015). While data from some patients supported these mechanisms, a large study revealed Th17 cells were increased and Tregs reduced in 25–40% of STAT3^GOF patients (Leiding et al., 2023). Thus, it is unlikely these cellular perturbations sufficiently explain disease manifestations nor provide mechanistic insights that can be leveraged to implement potential therapies for this condition. For these reasons, it is often necessary to utilize additional models to more fully understand how disease arises in human IEI.

Gene knockout mouse models can clarify disease mechanisms resulting from biallelic null mutations. However, such models are not instructive when pathogenic mutations are heterozygous and cause disease by partial LOF, negative dominance, haploinsufficiency, or GOF. Fortunately, advances in gene editing technologies have enabled rapid generation of mice harboring mutations validated to cause human diseases, thus establishing models that more closely resemble the genetic etiology of human pathologies. Recently, four groups described mice carrying germline Stat3^GOF mutations (T716M, K658N, G421R, K392R) and used these to investigate pathology, cellular phenotypes, and putative mechanisms of STAT3^GOF syndrome (Masle-Farquhar et al., 2022; Schmitt et al., 2022; Warshauer et al., 2021; Woods et al., 2022).

Stat3^GOF mice recapitulated numerous aspects of human disease, including runting, skin pathology, hepatosplenomegaly, lymphadenopathy, lymphoproliferation, and systemic inflammation (Masle-Farquhar et al., 2022; Schmitt et al., 2022; Warshauer et al., 2021; Woods et al., 2022). Under conditions of disease exacerbation, Stat3^GOF mice developed diabetes and colitis more rapidly than wild-type (WT) mice (Schmitt et al., 2022; Warshauer et al., 2021; Woods et al., 2022). Stat3^GOF mice accumulated activated CD4⁺ T cells skewed towards a Th1/IFNγ⁺ fate and CD8⁺ T cells resistant to exhaustion and enriched for effector memory cells over-expressing molecules required for pro-inflammatory and cytotoxic functions (Masle-Farquhar et al., 2022; Schmitt et al., 2022; Warshauer et al., 2021; Woods et al., 2022).

Interestingly, Stat3^GOF mice generated Tregs in vivo that were numerically and functionally comparable to WT Tregs. Furthermore, few Th17-type cells were detected in inflamed tissues (lamina propria, pancreas) in Stat3^GOF mice (Schmitt et al., 2022; Warshauer et al., 2021; Woods et al., 2022). These findings suggest that quantitative or qualitative Treg defects or unbridled production of Th17 cells are not the main driver of STAT3^GOF disease. Consistent with these observations, disease onset was T cell intrinsic, particularly CD8⁺ T cells, revealing a role for highly cytotoxic CD8⁺ T cells in the pathology in Stat3^GOF mice (Masle-Farquhar et al., 2022; Schmitt et al., 2022; Warshauer et al., 2021; Woods et al., 2022). Collectively, these studies provided some insights into possible causes of immune dysregulation in human STAT3^GOF syndrome, which have implications for targeted and cellular therapies. However, as the clinical phenotype of human STAT3^GOF syndrome involves many tissues and organ systems, it is necessary to continue to explore the role of other immune and non-immune cells in disease pathogenesis.

Consistent with frequent skin disease in STAT3^GOF patients, spontaneous skin disease was reported in homozygous Stat3^GOF mice (Masle-Farquhar et al., 2022). In this issue of JEM, Toth et al. (2024) found that aged heterozygous Stat3^G421R GOF mice (>18 wk) develop spontaneous skin inflammation in the form of ear swelling, increased epidermal thickness and immune cell infiltrates (neutrophils, monocytes, CD4⁺ and CD8⁺ T cells). Interestingly, while IFNγ⁺CD4⁺ T cells were increased in spleens from adult Stat3^GOF mice, IL17A⁺IL-22⁺ and RORγt⁺ CD4⁺ T cells, as well as IL-22⁺ and IL17A⁺IL-22⁺ but not IL17A⁺ γδ T cells, were dominant in inflamed skin. Together, these data suggest that increased Th17 differentiation in Stat3^GOF mice is tissue specific i.e., the skin, and that skin-specific disease may be driven by IL-22 rather than IL-17A.

To investigate if inflammatory challenge in younger Stat3^GOF mice could recapitulate spontaneous skin disease observed in aged Stat3^GOF mice, Toth et al. (2024) utilized the imiquimod (IMQ) model of psoriasiform dermatitis, where STAT3-activating cytokines are induced via TLR7 signaling. IMQ-treated in Stat3^GOF mice had more severe disease than WT littermates with increased ear swelling, epidermal thickening and cellular infiltrates. Stat3^GOF mice had increased IL-17A⁺ and RORγt⁺ CD4⁺ T cells and IL-22–producing γδ T cells, but fewer Tregs, in the skin following IMQ treatment compared to Stat3 WT mice, establishing a common role for tissue-specific inflammatory Th17/IL-22⁺ cells in spontaneous and induced models of skin disease in Stat3^GOF mice.

Adaptive immune cells were required to drive skin pathology in Stat3^GOF mice, evidenced by ear swelling and pathological scores being similar in Rag1^−/−Stat3^WT and Rag1^−/−Stat3^GOF mice. However, CD8⁺ T cells and γδ T cells were not necessary as skin pathology persisted in IMQ-treated Stat3^GOF mice depleted of these cells. Furthermore, transfer of activated Stat3^GOF CD4⁺ T cells into Rag1^−/− mice induced disease, associated with increased expression of IL-22, but not IL17A, following IMQ treatment. Taken together, while CD8⁺ T cells drive type 1 diabetes and colitis in Stat3^GOF mice (Warshauer et al., 2021; Schmitt et al., 2022), Th17 cells cause skin inflammation in IMQ-treated Stat3^GOF mice (Toth et al., 2024). Furthermore, disease in Stat3^GOF mice was reduced but not abolished when crossed onto IL-22–deficient mice, revealing a contributory role for IL-22 in STAT3 GOF-mediated skin inflammation. However, IL-22–independent factors contribute to neutrophil recruitment and histological changes in the skin. Treatment of Stat3^GOF mice with the JAK inhibitor tofacitinib also improved IMQ-induced skin inflammation, even though numbers of IL-17⁺ and IL-22⁺ CD4⁺ T cells were unaffected by tofacitinib, suggesting JAK inhibition does not attenuate accumulation of these cells at sites of inflammation (Toth et al., 2024).

The contribution of STAT3^GOF in hematopoietic vs. non-hematopoietic cells to disease onset and severity was addressed using bone marrow (BM) chimeras. Stat3^GOF in hematopoietic cells led to greater recruitment of immune cells to the skin. However, epidermal hyperplasia required Stat3^GOF in non-immune cells, consistent with the ability of IL-22 to promote keratinocyte proliferation (Seth and Dubey, 2023). Thus, Stat3^GOF hematopoietic cells were not sufficient to fully recapitulate skin inflammation observed in intact Stat3^GOF mice. Intriguingly, these mouse experiments also demonstrated that WT BM failed to efficiently engraft in Stat3^GOF hosts. This parallels the variable success rates of hematopoietic stem cell transplantation (HSCT) as a curative therapy in STAT3^GOF patients (Leiding et al., 2023). Indeed, this murine model may provide valuable insights into ways to improve HSCT in patients.

Stat3^GOF CD4⁺ T cells show increased clonal expansion, with increases in central memory and exhausted cells at baseline and effector CD4⁺ T cells following IMQ treatment compared to WT CD4⁺ T cells. Expanded CD4⁺ T cell clones from untreated Stat3^GOF mice showed greater expression of STAT3 target genes (Socs3, Maf, Cxcr3), as well as genes encoding STAT3 binding proteins (Junb), and genes associated with exhaustion (Tox, Tox2). Following IMQ treatment, Stat3^GOF CD4⁺ T cells transcriptionally resembled Th17 cells (Socs3, Junb, Fos, Il22), while those from WT mice were more similar to Tregs (Foxp3, Il2ra, Ikzf2), suggesting exaggerated STAT3 function alters the Th17/Treg balance during skin inflammation.

The ubiquitous nature of STAT3 signaling in different types of immune and non-immune cells makes it difficult to dissect components and cell-specific mechanisms that contribute to disease in humans. This is clearly complex, evidenced by the broad phenotype displayed by individuals with STAT3^GOF syndrome, even those harboring the same pathogenic variant (Milner et al., 2015; Leiding et al., 2023). This underscores the utility of using mouse models engineered to carry validated mutations to elucidate mechanisms underpinning distinct clinical features of human disease. Indeed, Toth et al. (2024) demonstrate that cell intrinsic increases in Th17 cells drive skin disease in Stat3^GOF mice, largely mediated by IL-22, while this and similar models delineated hematopoietic and non-hematopoietic mechanisms for skin disease, as well as the contribution of different immune cell types to different aspects of disease in Stat3^GOF i.e., CD4⁺ Th17 cells in skin inflammation vs. CD8⁺ T cells in type 1 diabetes, colitis, and lymphoproliferation (Warshauer et al., 2021; Masle-Farquhar et al., 2022). Further utilization of Stat3^GOF models will no doubt reveal putative pathogenic roles for other cell types (Th22, B cells) and also inform strategies for improved therapies for immune dysregulatory diseases.