Molecular control of pathogenic Th17 cells in autoimmune diseases.

Emerging evidence suggests that non-pTh17 and pTh17 cells are distinct cell subsets that controlled by a distinct molecular program. One would expect unique transcriptional factors to direct the generation and function of pTh17 cells. Nonetheless, the identified molecules including RORγt, BATF and IRF4 are indistinguishably required for the generation of both pathogenic- and non-pathogenic-Th17 cells [ 35 , 101 – 103 ]. Therefore, increasing efforts have been made to shed light on the specific molecular control of pTh17 cells. Studies of the transcriptional signatures of non-pathogenic and pathogenic Th17 cells enable us to better understand the molecular network of these two subsets. Lee et. al analyzed the molecular programs associated with pTh17 cells compared to the non-pTh17 cell by genome wide-microarray analysis [ 28 ]. To further address the signaling network of pTh17 cell development in vivo , Gaublomme et. al used single-cell RNA sequence technology to investigate the molecular network governing the heterogeneity and the pathogenicity of Th17 cells isolated from EAE diseased mice [ 49 , 104 ]. Results showed that TGFβ3- and IL-1β/IL-6/IL-23-induced pTh17 cells reveal similar transcriptional gene profiles distinct from that of TGFβ1-induced non-pTh17 cells. The establishment of pathogenicity not only increased a pro-inflammatory genes module ( Il23r, Csf2, Tbx21, Il17a , and Il17f ) but also reduced the expression of immune-suppressive genes ( Il10, Il4, Cd5l, Ahr and cmaf ). Single-cell transcriptomics analysis from in vivo isolated IL-17A + Th17 cells shows that there is a zone of overlapping cell states between non-pTh17 and pTh17 cells [ 104 ]. Through co-variation module analysis, a proinflammatory module co-variance with IL-17A and a regulatory module correlation with IL-10 are further established. Intrigued by the sequence data, some putative regulators are selected showing a potential role in the regulation of pathogenicity of Th17 cells: Gpr65, Toso, Plzp , and Cd5l . GPR65, Toso, and Plzp were found to promote pTh17 cells and are essential for the progress of EAE, with up-regulated pro-inflammatory genes [ 104 ]. For instance, Gpr65-deficient CD4 + T cells show comparable IL-17A + percentage with WT counterpart under TGF-β1/IL-6 condition but exhibit defective differentiation of Th17 cells under IL-1β/IL-6/IL-23 condition, indicating a differential requirement of Gpr65 in non-pTh17 and pTh17 subsets. However, little is known about whether and how IL-23 regulates those molecular factors. Additional studies focus on other top-ranked candidates such as Hif1a , Fosl2 , Stat4 , Med12 , Etv6 , Gem , Foxp1, Rbpj and Procr . Hif1α is a key metabolic sensor and induced in CD4 + T cells in a STAT3/mTOR-dependent manner which is also critical for IL-17A production, through the binding with RORγt and recruiting p300 in non-pTh17 cells [ 105 , 106 ]. In addition, Notch signaling mediator RBPJ drives the expression of IL-23R and reciprocally suppresses IL-10 expression during Th17 differentiation [ 107 ]. Overexpression of IL-23R could rescue the defective pathogenicity of RBPJ-deficient Th17 cells. The role of another candidate molecule, protein C receptor (PROCR), in correlation with IL-10 production in Th17 cells, was also identified in the pathogenicity of Th17 cells [ 108 ]. Overexpression of PROCR can reduce the pro-inflammatory gene profile including Il1r and Il23r , indicating a negative role of PROCR in the pathogenicity of Th17 cells. Additionally, microRNA-183 cluster C, which is upregulated by IL-6 and downregulated by TGF-β1 [ 109 ], promotes the pathogenic module of pTh17 cells by repressing Foxo1, a suppresser of Th17 cells [ 109 , 110 ]. Recently, RIP2 has been reported to be involved in the balancing of homeostatic and pathogenic features of Th17 cells [ 111 ]. Deletion of RIP2 reduced IL-17A production in non-pTh17 cells. In contrast, RIP2 deficiency enhanced IL-17A expression under pTh17 cell condition via the increase of RORα that promotes Th17 cell differentiation [ 111 ]. Are there any molecules that can function in both development and pathogenicity adaption for pTh17 cells but not in non-pTh17 cells? Recently, our group revealed that RASA3 (RAS p21 protein activator 3), a GTPase activating protein of GAP1 sub-family, is specifically required for both the generation and pathogenicity of pTh17 cells and dispensable for non-pTh17 cells ( Figure. 1 ). RASA3 does so by balancing the reciprocal molecular programs of pTh17-Th2 cells via a RASA3-IRF4-Cbl-b axis under pTh17 condition [ 112 ]. Although IRF4 is essential for the development of both non-pTh17 and pTh17 cells, our finding revealed that the levels of IRF4 expression appeared to be critical: high levels of IRF4 expression mediated Th2 program suppressed and medium levels of IRF4 expression enhanced the pathogenic-Th17 cell generation of RASA3-deficient T cells. These findings suggest that IRF4 controls the pathogenic-Th17 program in a biphasic, dose-dependent manner. The above studies suggest that unique molecular programs exist to dictate the generation of pTh17. Both lineage development and pathogenic function adaption contribute to the generation of pTh17 cells in vitro and in vivo . Other molecules and pathways are likely to be discovered to specifically control pTh17 cell generation and immune pathologies. The studies to unveil critical factors in controlling pTh17 cell function will provide valuable therapeutic targets to interfere with the lineage commitment and/or the pathogenicity of Th17 cells and to treat autoimmune diseases.

Th17 cells that produce the signature cytokine interleukin 17A (IL-17A) and lineage-specific transcriptional factor retinoid-related orphan receptor gamma t (RORγt) have attracted great attention since their identification in 2005 [ 1 , 2 ]. Th17 responses contribute to the tissue inflammation and pathogenesis of diverse autoimmune diseases, including psoriasis, rheumatoid arthritis, inflammatory bowel disease (IBD) and multiple sclerosis [ 3 – 6 ]. Intrigued by the strong correlation of Th17 cells with immuno-pathology in autoimmune diseases, Th17 cells have been exploited in the intervention of these diseases [ 4 , 5 , 7 – 17 ]. However, clinical trials targeting IL-17A signaling with Secukinumab (anti-IL-17A antibody) or Brodalumab (anti-IL-17RA antibody) show inconsistent effects. Secukinumab provides encouraging improvement in patients with multiple sclerosis, active ankylosing spondylitis (AS) or psoriasis [ 18 , 19 ]. In contrast, Secukinumab/Brodalumab failed to improve Th17 cell-mediated Crohn’s and arthritis diseases [ 20 , 21 ], which is in line with a report that IL-17A might play a protective role in a mouse colitis model [ 22 ]. A series of studies suggest that the blockade of IL-17A signaling may even exacerbate disease in intestinal and bone tissues [ 11 , 22 – 24 ]. Indeed, although Th17 cells are essential for driving inflammatory disorders, Th17 cells also benefit the homeostasis of tissues in certain conditions [ 25 – 27 ]. Thus, in vivo generated Th17 cells can be both pathogenic and non-pathogenic to the host in a context-dependent manner. In addition, how naive T cells become non-pathogenic or pathogenic Th17 cells depends on signals from different cytokine milieus in vitro [ 28 , 29 ]. In light of the dichotomy of Th17 cells, increasing efforts have been devoted to reveal how the generation and function of pathogenic and non-pathogenic Th17 cells are controlled.

While the past knowledge of Th17 cell function is mainly acquired from TGF-β1/IL-6-induced non-pathogenic Th17 cells, a discrete pathogenic Th17 subset induced by IL-1β/IL-6/IL-23 has attracted great attention for its critical role in controlling autoimmune diseases. Although increasing findings have extended the understanding of pTh17 cells, several questions must be addressed in the future: (1) Are there any specific transcriptional factors or surface makers that can be utilized to distinguish pTh17 cells during pathological conditions? Can these potential candidates be manipulated for the treatment of pTh17-related diseases? (2) Non-pTh17 cell subset is functionally plastic and can be converted into Th1 and Treg cells [ 113 ]. If and how the function of pTh17 cells are similarly plastic remains to be investigated. (3) Increasing evidence indicates that Th17 cells can establish transient or long-term residency in non-lymphoid tissue [ 51 , 114 , 115 ] to control immunity, homeostasis and pathology. Nonetheless, the definition and function of Th17 Trm cells in vivo remain poorly defined. In addition, whether Th17 Trm cells share similar molecular controls as CD8 + Trm T cells are interesting questions to be addressed. In summary, while great strides have been made in understanding the function and regulation of Th17 cells in the past, much work is needed to reveal specific molecular mechanisms underlying the generation and function pTh17 vs. non-pTh17 cells. This will shed light on how Th17 cells impact inflammatory diseases. Additionally, it will benefit the development of therapeutic approaches to treat Th17 cell related diseases.

The phenotypic and functional properties of Th17 cells have been extensively investigated in vitro and in vivo . Detailed studies suggest that different environmental factors direct naïve CD4 + T cells to differentiation into pTh17 cells or non-pTh17 cells [ 28 , 30 , 31 ]. Current knowledge on the Th17-specifying molecular program is largely obtained from studying TGF-β1+IL-6 induced Th17 cells [ 32 ]. IL-6 is the key factor that directly regulates the balance of Th17 and Foxp3 + Treg cells by inhibiting TGF-β1-induced Foxp3 expression [ 33 ]. In addition, IL-6-induced activation of STAT-3 is essential for the differentiation of Th17 cells from naïve T cells ( Figure. 1 ) [ 34 – 37 ]. STAT3-deficient T cells did not respond to IL-6 and failed to elicit a downstream signal [ 38 ]. However, in the absence of IL-6, pleiotropic cytokine IL-21 produced by Th17 cells can promote the development of Th17 cells via the activation of STAT3 [ 39 – 41 ]. And yet, the relevance and necessity of IL-21 in the polarization of Th17 cells remains uncertain in vivo . The development and function of Th17 cells are unaltered during experimental autoimmune encephalomyelitis (EAE) in IL-21 −/− or IL-21R −/− mice [ 42 , 43 ], suggesting a dispensable role of IL-21 signaling in driving Th17 differentiation and tissue inflammation in vivo . While TGF-β1+IL-6-induced Th17 cells may promote inflammation to a certain degree, they are largely considered as non-pathogenic Th17 cells and even demonstrate a protective role under certain circumstances [ 44 – 47 ]. Non-pTh17 cells exhibit a regulatory program to produce the immune regulatory cytokines IL-4 and IL-10, as well as critical negative regulators of a range of pathophysiological responses including CD5L (CD5 like molecule), IL-9 and GATA3 [ 22 , 34 , 48 , 49 ]. For instance, in steady-state, segmented filamentous bacteria (SFB)-elicited Th17 cells produce a large amount of IL-10 in lamina propria in the intestines, contributing to the immune homeostasis and tissue integrity of the gut [ 50 , 51 ]. Also, stimulation of MOG peptide-specific T cells with TGF-β1/IL-6 abrogates their pathogenic function during EAE induction and progression [ 34 ]. In agreement with the critical role of IL-10, IL-27 and IL-6 trigger STAT3-dependent IL-10 production in Th17 cells and restrain the progression of endometriosis and EAE [ 48 , 52 ]. While the importance of IL-10 in non-pTh17 cells have been demonstrated, it remains to be addressed as how IL-10 expression is regulated at the transcriptional and epigenetic levels during non-pTh17 generation. Recently, IL-10 + Th17 cells are found to be controlled by c-Maf and can be characterized as a tissue-resident non-pTh17 population [ 53 ]. Mechanistic study elucidates c-Maf modulates the immunoregulatory and tissue-residency program by binding to genes such as Pdcd1, Ctla4 , and Cd69 , suggesting c-Maf maybe an important factor to distinguish between non-pTh17 and pTh17 cells [ 53 ]. Furthermore, while the deletion of CD5L does not affect the differentiation of Th17 cells, CD5L regulates lipid metabolism, especially the balance of PUFA/SFA level, to alter RORγt ligand availability in non-pTh17 cells [ 49 ]. Compared to enhanced lipid metabolism in Treg cells, Th17 cells favor glycolysis and glutaminolysis pathways [ 54 – 57 ]. Inhibition of mitochondrial function and oxidative phosphorylation (OXPHS) is associated with non-pathogenic Th17 gene expression and impaired pTh17 cell function, suggesting that fine-tuning cell fitness and metabolic activity may be used to balance the function of pTh17 and non-pTh17 cells [ 58 – 62 ]. While regulatory molecules such as IL-4 and IL-10 participate in the process of metabolism [ 63 – 65 ], whether the increased regulatory program modulates the metabolic program in Th17 cells to restrain Th17 cell-mediated inflammatory disorders remains largely unknown and is of great interest. Researchers have made a substantial effort to interpret the differentiation, maintenance and function of pTh17 cells owing to their critical role in tissue inflammation. In humans, an IL-23R polymorphism has been genetically linked to many autoimmune diseases such as Crohn’s disease and psoriasis [ 66 , 67 ]. Notably, mice deficient in the p19 subunit of IL-23 fail to produce Th17 cells and were resistant to autoimmune diseases such as EAE, IBD and arthritis [ 68 – 70 ]. In line with this notion, the development of pathogenic Th17 cells relies on the cytokine combination of IL-1β/IL-6/IL-23 in vitro . Such pTh17 cells exhibit a detrimental pro-inflammatory program ( Il23r, Csf2, Tbx21, Il17a, Il17f, etc) and decreased regulatory modules such as Il4, Il10 and Cd5l [ 28 , 29 , 71 – 73 ] ( Figure. 1 ). In addition, Th17 cells generated via TGF-β1 plus IL-6 do not readily promote EAE disease until further exposure to IL-23 [ 30 , 74 ], suggesting an essential role for IL-23 signaling in pTh17 cell-induced inflammation. However, whether there are cell intrinsic factors regulated by IL-23 and/or other cytokines that can drive the differentiation of pTh17 in vivo is still unclear and warrants further investigation. Interestingly, IL-23 is not the inducer of Th17 cells due to the lack of IL-23 receptor (IL-23R) on naïve CD4 + T cells. Nevertheless, IL-23/IL-23R signal promotes the stability and survival of Th17 cells and is indispensable for Th17 cells to gain pathogenic features [ 5 , 30 , 68 , 74 , 75 ]. IL-23 does so through multiple mechanisms. Transcription factor Blimp-1, downstream of IL-23, drives the pathogenic program of pTh17 cells while suppresses IL2 and Bcl6 that are suppressors of Th17 cells [ 76 ]. In addition, GM-CSF is highly produced by pTh17 cells and is essential for the pathogenicity of Th17 cells but not required for Th17 cell generation [ 77 – 79 ]. Two mechanisms are involved in increased GM-CSF in Th17 cells: 1) IL-23 and RORγt drive the production of GM-CSF in Th cells [ 79 ]; 2) GM-CSF also acts on dendritic cells (DCs) to enhance their production of IL-23, which in turn promotes further activation of Th17 cells and GM-CSF production [ 80 ]. Additionally, IL-23 suppresses CD5L expression and regulates the metabolic status of Th17 cells, which is required for its pathogenicity [ 49 ]. p-STAT3/STAT4 is shown to be required for IL-23-IL-23R mediated CNS autoimmunity and such STAT4 activation is IL-12 independent. IL-12 signal disruption did not ameliorate EAE progression [ 81 ]. However, STAT4-deficient mice are resistant to EAE and colitis [ 82 – 86 ]. Global gene expression analysis indicates that in the absence of STAT4, the levels of pathogenic Th17 genes including Tbx21, Il22 and Cxcl3 are significantly reduced, while the expression of non-pathogenic genes including Il10 and Ahr is increased. Given the importance of IL-23 signal in pTh17 cells, specifically interfering with the IL-23 signal with Isankizumab, Guselkumab or Ustekinumab blockade antibodies led to promising outcomes in several clinical trials in patients with rheumatoid arthritis, psoriasis and Crohn’s disease [ 12 , 87 – 91 ]. However, anti-IL-23 monoclonal antibody treatment did not achieve favorable outcome in patients with advanced relapsing-remitting multiple sclerosis or active ankylosing spondylitis, suggesting a more specific approach to target pTh17 cells is required [ 92 – 94 ]. Current studies indicate TGFβ1 is not always required for IL-17A production. IL-17A producing CD4 + T cells are still detectable in the gut from CD4dnTGF-βRII and TGFbr1 fl/fl ;Cd4Cre mice in which TGFβR signaling is abrogated in T cells [ 30 ]. However, acute deletion with tamoxifen in ERcre;TGFβRII fl/fl mice suggest a requirement of TGFβRII signal in the development of Th17 cells [ 95 ]. These findings indicate the requirement of TGF-βRII but not TGF-βRI in the generation of Th17 cells. This is not happening by chance as CD4 + T cells gradually lose TGFβRI but not TGFβRII expression after activation [ 96 ], which could also explain the generation of pathogenic Th17 cells in the absence of TGFβ1. TGF-β superfamily members (e.g., TGFβs, activins, and BMPs) regulate diverse developmental and physiological processes. More interestingly, compared to TGF-β1, TGF-β3, another TGF-β family cytokine, is produced by activated Th17 cells and further increased by the addition of IL-23. TGF-β3 plus IL-6 can also induce pTh17 cells ( Figure. 1 ) [ 28 ]. Although TGFβ3 shares the same TGFβRII with TGFβ1, TGFβ3 induced the activation of Smad1/5 but not canonical Smad2/3 signal in Th17 cells, further indicating that TGFβ3-induced pTh17 cells are developmentally distinct from non-pTh17 cells. Currently, it is still unclear whether other molecules are involved in the generation of pathogenic Th17 cells beyond TGFβ3 and IL-23. Nonetheless, there are numerous TGF‐β superfamily members that share similar structures and functional overlap and yet distinct receptors [ 97 – 99 ]. Indeed, Activin A, a member of the TGF-β superfamily that regulates tissue homeostasis, cell proliferation, and tissue inflammation, is able to promote Th17 differentiation in concert with IL-6 [ 95 ]. In addition, the activity of phosphatase PP2A, a factor modulates TGFβ/Activin A/Nodal signaling, controls Th17 cell generation and mediated EAE by its differential effect on the phosphorylation of Smad2 vs. Smad3 [ 100 ]. These findings suggest a much broader function of the TGF-β superfamily in the generation and function of Th17 cells through more intricate molecular networks than we previously thought. It therefore would be interesting to know whether and how TGF-β superfamily members besides TGFβ may contribute to the generation and functional specification of distinct Th17 cell subsets.