STATus of STAT3 in Psoriatic Arthritis

Joint disease and enthesopathy are seen in about one-third of patients with psoriasis and are usually preceded by the cutaneous manifestations of psoriasis. The relationships between the pathogenesis of psoriatic skin disease and the pathogenesis of psoriatic arthritis have not been defined, however, and the central question as to whether psoriatic skin disease initiates development of joint disease remains unresolved. It has been shown that constitutive, hyperexpression of signal transducers and activators of transcription 3 (STAT3) using a STAT3C construct targeted to keratinocytes is sufficient to induce cutaneous psoriasis-like lesions. When combined with global expression of a mutant form of the glycoprotein 130 (gp130) allele, a cytokine receptor signal transducer, the STAT3C construct leads to arthritis in mice [1, 2]. In this issue, Yang and colleagues introduce a new mouse model that is unique in that it embodies all the major features of cutaneous psoriasis and psoriatic arthritis including skin disease, joint disease and an enthesopathy, as well as osteopenia [3]. This model also is based on constitutive, hyperexpression of STAT3 using the STAT3C construct but in this case the targeting of the construct to CD4⁺ T cells has been found to be sufficient to induce all the manifestations.

STAT3 signaling plays a key role in normal cell functions including cell growth, survival, proliferation, apoptosis and inflammation. It is activated by numerous cytokines and growth factors through JAK-mediated phosphorylation of tyrosine 705 or phosphorylation of serine 727 through a variety of serine/threonine protein kinases (MAP kinase, mTOR and PKCδ), which is required for maximal STAT3 transcriptional activity. The STAT3 proteins occur as inactive monomers in the cytoplasm until their phosphorylation which prompts their dimerization. The dimers are translocated to the nucleus where they bind to specific DNA elements that regulate transcription of an array of genes. Almost two decades ago, Darnell and co-workers engineered the constitutively dimerizable STAT3 molecule, STAT3C, by substituting cysteine residues for tyrosine within the c-terminal loop of the SH2 domain of the STAT3 molecule [4]. STAT3C was shown to drive transcription, stimulating not only cell growth and proliferation but also promoting transformation. Subsequently, naturally occurring constitutive activation of STAT3 has been identified in several human malignancies. STAT3C should not be confused with the naturally occurring isoforms of STAT3 function that result from an alternative splicing event of STAT3 at exon 23. This results in the expression of a truncated form of STAT3 (STAT3β) [5] that lacks serine 727. Although initially thought to act in a dominant-negative manner, it is a biologically active molecule capable of modulating the tumorigenic functions of STAT3. The differential expression of these two isoforms represents just one aspect of the complex network of mechanisms that regulate the magnitude of STAT3 signaling and modify its effects on transcription in a manner that depends on physiological conditions and cellular context. Thus, a limitation of the use of the constitutively active STAT3C construct is that it not only exaggerates STAT3 activity but also functions independently of many of these biologically relevant regulatory constraints. Nevertheless, the use of STAT3C in mouse models of diseases has proven useful in providing novel insights into potential pathogenic pathways.

As described above, the constitutive hyperexpression of STAT3 keratinocytes alters keratinocyte proliferation and development resulting in the hyperkeratosis characteristic of skin psoriasis as well as psoriasis-like inflammation of the skin. In this keratinocyte mouse model, the dysregulated growth of the keratinocytes induces an influx of macrophages that produce IL-6 and IL-1, which in the context of the dysregulated cytokine milieu promotes the development of the characteristic inflammatory skin response, including development of Th17 and Th22 cells. In conjunction with the gp130 mutation, this inflammatory response is sufficient to elicit joint disease. The newly introduced model of STAT3C expression in CD4⁺ T cells, provides an alternative paradigm in which STAT-associated dysregulation of CD4⁺ T cells can act as a primary initiating event that drives inflammatory skin disease and hyperkeratosis and the joint involvement (Figure 1).

Figure 1. STAT3 in CD4 T cells promote psoriasis and psoriatic arthritis.

The article by Ding and colleagues in the current issue shows that constitutive hyperexpression of STAT3 in mouse CD4⁺ T cells is sufficient to promote psoriasis-like skin disease, psoriasis-like arthritis and enthesopathy. Taken together with other data, the results suggest a model in which constitutive expression of STAT3, with or without IL-6 signaling, promotes an inflammatory environment including IL-1, TGFβ and IL-23 that promotes development and differentiation of Th17 cells (upper pathway). The same conditions in the presence of TNFα and IL-23 lead to the sustained development of Th22 T cells (lower pathway). Th17 cells express characteristic transcription factors, STAT3 and RORᵧt; chemokine receptors, CCR4 and CCR6; and cytokines IL-17a, IL-17f, IL-22 and IL-21. Collectively, these promote development of a psoriasis-like arthritis and individuals with higher STAT3 transcriptional activity are more susceptible to joint disease (upper pathway). Th22 cells express canonical transcription factors, STAT3, NF-kB and aryl hydrocarbon receptor (AhR); chemokine receptors, CCR4, CCR6 and CCR10; and cytokines, IL-22, TNF, IL-13 and IL-26. Collectively, these promote development of skin psoriasis and individuals with high STAT3 transcriptional activity are more likely to develop skin disease (lower pathway).

In this report, Yang and colleagues provide a detailed description of the disease manifestations. In terms of the potential pathogenic mechanism, they focus primarily on the examination of the role of STAT3-driven augmentation of the activity of the Th17 subset of CD4⁺ T cells. The role of IL-22 is examined through use of IL-22-knockout mice. While Th17 cells produce IL-22 as well as IL-17, IL-22 is produced predominately by a distinct lineage of CD4+ T cells (Th22 cells). Increased frequencies of IL-22-producing cells with characteristics of Th22 cells are a feature of psoriasis and psoriatic arthritis in addition to the increased frequencies of Th17 cells [6]. Cytokine-induced STAT3 signaling has been shown to play a role in the development, differentiation and expansion of both Th17 and Th22 cells. It is increasingly clear that although Th17 and Th22 cells share many commonalities, their differences extend well beyond their cytokine-expression profile. Commonalities include their induction by IL-6-induced upregulation of STAT3; their reliance on IL-23 for sustained activity and differentiation, which is of great interest as genetic susceptibilities associated with IL-23 and IL-23-R have been reported for both psoriasis and psoriatic arthritis; and their expression of the chemokine receptors, CCR6 and CCR4. Differences include the induction/promotion of Th17 development by IL-6 in combination with TGFβ and IL-1 signaling [7], whereas Th22 T-cell development is induced/promoted by IL-6 in combination with TNFα signaling; IL-23 leads to production of IL-17a, IL-17f and IL-22 as well as TNFα, IL-6, IL-21 and IL-26 by the Th17 cells, whereas it leads to production of IL-22 and TNFα, as well as IL-13 and IL-26 by Th22 cells; and Th22 cells express CCR10 in addition to CCR6 and CCR4, as well as expressing the aryl hydrocarbon receptor (AhR).

Analysis of samples from patients with psoriasis has indicated increased levels of IL-22 and Th22 cells in the involved skin or peripheral blood [7]. Somewhat surprisingly, Th22 cells are very low in involved joints from patients with psoriatic arthritis [7] although they are high in involved joints of approximately 60% of patients with rheumatoid arthritis (RA) [8]. These differences could potentially reflect the relative ability of the microenvironment at the different sites to promote or sustain the Th22 versus Th17 cells. It also has been suggested that preferential homing of subpopulations of T cells, including Th17 and Th22 cells, to specific sites may affect the manifestation of skin systems in psoriasis and the differences in the arthritis associated with psoriatic arthritis versus RA. Although the Th17 and Th22 cells share some chemokine receptors, Th22 cells also express CCR10 and skin-specific homing receptors that have been described in psoriasis [9]. Further analysis of the new model may provide insights into this critical question.

An unexpected finding was that on crossing the mice with constitutive hyperexpression of STAT3 in CD4⁺ T cells to IL-22 knock-out mice, a significant reduction in the psoriatic arthritis manifestations was observed as well as a reduction in the skin phenotype. While the reduction in the skin phenotype is consistent with the evidence that IL-22 has a more prominent effect on psoriatic skin disease, joint disease has been more closely associated with IL-17. An association between IL-17, RANK ligand expression and osteoclastogenesis is well established. Based on the observed reduction in RANK ligand levels in the IL-22 knockout mice, the authors speculated that this previously undescribed role for IL-22 in psoriatic joint disease is due to the ability of IL-22 to stimulate of RANK-ligand expression. This was based on the report that IL-22 was found to stimulate RANK ligand expression by synoviocytes from patients with RA and IL-22-treated fibroblasts stimulated osteoclast differentiation from monocytes in the absence of RANK ligand [8]. As indicated above the increase in Th22 cells was more prominent in the joints of patients with psoriatic arthritis compared to RA. Thus, the intriguing finding of a potential role for IL-22 in psoriatic joint disease will need to be further evaluated to determine the extent to which the mouse model corresponds to psoriatic joint disease in humans.

A key feature of the newly introduced model is the description of the development of an enthesopathy. Enthesopathy is one of the initial features that can be used to clearly distinguish psoriatic arthritis from RA [10]. The understanding of the inter-relationship between the enthesis and synovium and its significance is evolving. A close anatomical integration between the enthesis and synovium resulting in a synovio-entheseal complex has been described [11]. It has been proposed that the enthesis fibrocartilage that is located adjacent to synovium, such as in joints, bursa or tendons, is dependent upon the synovium for lubrication and nutrition. The enthesis functions to stiffen the tendon or ligament and helps to create a more gradual change in mechanical properties between soft and hard tissues. The enthesis insert itself is a vascularized fibrocartilaginous structure although it is unclear if the close association with the joint can lead to inflammation in the adjacent vascular synovium. Which of the novel features of this mouse model leads to the enthesopathy is unclear and is a subject for future investigation.

Limitations of the current analysis of pathogenesis mirror the limitations of analysis of human psoriasis and psoriatic arthritis; however, ultimately, the model may provide resolution of some of the outstanding questions. These include analysis of the functional consequences of the localized production of multiple cytokines by the Th17 and Th22 cells, which have not been explored fully to date. Of importance, the model also may enable informative analysis of the heterogeneity and plasticity of CD4⁺ T cells within the specific pathogenic settings. The Th17 cell population is heterogenous in terms of the levels of production of the signature cytokines. In psoriasis, there are at least three subpopulations of Th17 cells characterized by their production of both IL-22 and IL-17, or production of only IL-17 or only IL-22 [7]. This is consistent with the finding that in the BXD2 mouse model of arthritis CD4⁺ spleen T cells produce either IL-17 or IL-21, but not both [12]. The basis for the propensity of individual Th17 and Th22 cells to produce one cytokine is unknown but may reflect prior signaling interactions that modulate the activity of STAT3 or other transcription factors, such as aryl hydrocarbon receptor [13] or RORᵧt. While the mouse model would enable informative single-cell analyses by excluding the levels of “noise” intrinsic in patient samples, it is possible that the STAT3C construct bypasses these modulating effects and favors the development of specific subpopulations. Testing of the roles of SNPs in modifying the downstream effects of STAT transcription should be possible, however.

As in human psoriasis and psoriatic arthritis, the collection of features manifested by the mice raises the question as to which features are primary events and which are secondary. Resolution of this issue is central to the development of effective therapeutic strategies. This extends beyond the potential cascade of events that could be initiated by dysregulated cytokines and circulating immune cells. One example is the potential release of endogenous proinflammatory mediators from a primary site that then induce responses at secondary sites. This possibility has been invoked in terms of the postulated initiation of psoriatic arthritis by the compromised synovio-entheseal complex [11]. A second example, is the question of the role of the altered microbiota that has been observed in psoriasis and patients with psoriatic arthritis [14]. AhR, which is highly expressed by Th22 cells and regulates their expansion, recognizes numerous naturally occurring molecules, including dietary flavonoids and common bacteria, as well as xenobiotics, including dioxin. IL-22 producing have been shown to be central to maintenance of the intestinal barrier and STAT3/Th IL-22 downregulation compromises the gut barrier [15]. It would be important to develop a method to test a primary versus a secondary effect of the altered microbiome in disease, which has recently been highlighted by findings negating previous reports of altered microbiome in mice with inflammasome deregulation [16].

In summary, the new mouse model highlights the importance of STAT3 signaling in psoriasis and psoriatic arthritis. As in cancer, this should provide a framework for analysis of pathogenesis and disease susceptibility although therapeutic targeting of STAT3 has proven difficult to date due to its central importance in so many critical biological processes.