New developments implicating IL-21 in autoimmune disease

Abstract

Elevated interleukin (IL)-21 is a common finding in the tissues and/or sera of patients with autoimmune disease. CD4 T cells are the primary producers of IL-21; often the IL-21 producing CD4 T cells will express molecules associated with follicular helper cells (T_FH). Recent work has shown that the CD4 T cell-derived IL-21 is able to promote effector functions and memory differentiation of CD8 T cells in chronic infections and cancer. Autoimmunity has similarities to chronic infections and cancer. However, CD4 T cell-derived IL-21:IL21R signaling in CD8 T cells has not been fully appreciated in the context of autoimmunity. In this review, we assess the current knowledge regarding CD4 T cell-derived IL-21 and IL21R signaling within CD8 T cells and evaluate what implications it has within several autoimmune diseases including systemic lupus erythematous, rheumatoid arthritis, juvenile idiopathic arthritis, type 1 diabetes mellitus, psoriasis, Sjögren’s syndrome, vitiligo, antiphospholipid syndrome, pemphigus, and giant cell arteritis.

1. Introduction

IL-21 produced by follicular helper CD4 T cells (T_FH) acts directly on B cells via their IL-21 receptors (IL21R) to generate high-affinity, class-switched antibodies [1,2]. In addition to this well-documented function of IL-21, CD8 T cells also express IL21Rs [2]; the implications of CD8 T cell intrinsic IL21R signaling are a source of active investigation [3,4]. Much of the work exploring the impact of IL21R signaling on CD8 T cell differentiation and function has been in the context of persistent infection models including LCMV clone 13 and docile, Toxoplasma gondii, and mouse polyomavirus [5–9], but IL21R⁺ CD8 T cells also combat cancer [10–12], and drive autoimmune disease [13]. Persistent infections, cancer, and autoimmunity have the commonality of being conditions of chronic antigen presentation. For immune protection from persistent viral infections and cancer, CD8 T cells need to preserve effector function while balancing off-target tissue damage with control of the persistent infectious pathogen or tumor. Animal models of chronic antigen presentation show that CD8 T cells maintain homeostasis within the tissues by acquiring specific metabolic profiles to maintain longevity and dampen effector responses [14–17]. IL21R signaling in CD8 T cells has been shown to help direct CD8 T cell differentiation and affect cellular metabolism, particularly in conditions of chronic antigen presentation [5–9]. While this pathway is beneficial for eliminating an infectious pathogen or tumor, in the context of CD8 T cell reactivity to self-antigen, IL-21/IL21R signaling may be further driving autoimmune disease. Increased IL-21 production, elevated IL21R expression, and polymorphisms in either gene have been documented in many autoimmune diseases [18,19]. While increased IL-21/IL21R signaling has been described before as a driver of autoantibody generation [20], little attention has been given to the IL-21-IL21R pathway on driving autoimmune pathogenesis through autoreactive CD8 T cells. This review will focus on IL21R signaling in CD8 T cells within the context of certain autoimmune diseases.

2. The CD4 T cell-IL-21:IL21R-CD8 T cell axis

Several immune cell types express Il21 transcripts but the specific CD4 T helper subtypes T_FH, peripheral helper (T_PH), and helper 17 (T_H17) CD4 T cells are often attributed as the primary producers of IL-21 [2,21,22]. CD4 T cells can provide helper functions to CD8 T cells. Work using systemic chronic LCMV infection revealed CD4 T cell-derived IL-21 as a mechanism of CD4 T cell help that circumvented exhaustion of CD8 T cells (a terminally differentiated dysfunctional state) [5–7]. IL-21 from CD4 T cells was also found to promote CD8 T cell effector function during persistent T. gondii and mouse polyomavirus brain infection, particularly by promoting a metabolic profile associated long-lived residence in the tissue, i.e. tissue resident memory (T_RM) [8,9]. Figure 1 illustrates IL-21 production by T_FH or T_PH polarized CD4 T cells in the tissues and how CD8 T cells respond to the CD4 T cell-derived IL-21.

Figure 1.

The CD4 T cell-IL-21:IL21R-CD8 T cell axis. High TCR signaling leads to increased expression of PD-1 as well as of other molecules associated with T_FH and T_PH polarization such as ICOS and the ability to produce IL-21. When these CD4 T cells enter tissues, e.g. the joint space, and are activated they produce IL-21. This CD4 T cell-derived IL-21 signals to CD8 T cells activated through their TCRs. The insert shows the specific IL21R signaling pathway that occurs within the CD8 T cell. JAK1/3 signals primarily through STAT3 which leads to transcription of many different genes associated with T_RM. Signaling also occurs through phosphatidylinositol 3-kinase (PI3K) and mitogen-activated protein kinase (MAPK). The combined results of these signals ascribe the cell to a metabolic profile that supports chronic inflammation including through the release of IFNγ. The IFNγ from the CD8 T cell furthers the inflammation in the joint space leading to tissue destruction and pain. Figure image created with BioRender.com.

2.1. Production of IL-21 by CD4 T cells

T_FH cells are so named for their abundant presence in the follicles of lymphoid organs. It is in these follicles where T_FH interact with B cells to promote somatic hypermutation, affinity maturation, and class-switching of the B cell receptors [23]. The defining phenotype of T_FH is high expression of PD-1 (programed death receptor, which is associated with elevated TCR signaling) and CXCR5 (a chemokine receptor for CXCL13; CXCL13 is highly expressed in germinal center follicles to direct B and T cell migration). T_PH have similar transcriptomes to T_FH. The primary distinguishing features of T_PH is that they are CXCR5⁻ and extra-follicular [24]. T_PH, like T_FH, have been shown to provide help to B cells, although this help primarily occurs within inflamed tissues. T_H17 cells are often distinguished by high expression of the transcription factor RORγT and production of IL-17A. T_H17 cells are important to anti-bacterial and anti-fungal responses against extracellular bacteria and fungi, but have also been associated with autoimmune diseases, particularly multiple sclerosis [25].

Despite the strong association of IL-21 production with CD4 T cell T_FH, T_PH, or T_H17 polarization, yet to be established is why certain CD4 T cells produce IL-21 and other CD4 T cells do not. As production of IL-21 is often included as a functional identifying factor to ascribe a CD4 T cell as being T_FH, T_PH, or T_H17, it is difficult to separate the signaling pathways that result in IL-21 production from the signaling pathways that increase expression of other T_FH-, T_PH- or T_H17-associated molecules. Using 2D micropipette adhesion assays to measure the binding affinity between peptide-MHC class II and its TCR, IL-21-producing CD4 T cells were shown as having higher TCR affinities and more TCRs on their surface than non-IL-21-producing CD4 T cells [9,26], suggesting that high TCR signal strength may be driving IL-21 production. It is worth noting however that in both these experiments measuring the TCR affinities, the CD4 T cells were CXCR5^hiPD-1^hi and expressed other molecules associated with T_FH polarization [9,26]. Both these works also showed that the non-IL-21-producers expressed the IL-2 receptor while the CD4 T cells producing IL-21 did not [9,26]. Also relevant was the finding that IL-21 producers had initially been IL-2 producers and Il2 expression tracked with Bcl6 expression [26]. These findings are consistent with Bcl6 (a T_FH-associated transcription factor) and Il2rα as divergent differentiation pathways in CD4 T cells [27].

Other potential determinants of CD4 T cell fate and IL-21 production are the type of antigen presenting cell and/or the amount and duration of the MHC class II: peptide ligand interaction. Mouse polyomavirus has two primary CD4 T cell epitopes in C57BL/6 mice: VP1–221 (from the viral capsid) and LT678 (from the nonstructural Large T antigen). Our work showed a higher percentage of the IL-21-producing CD4 T cells were VP1–221-specific [9]. As VP1–221 is derived from the capsid, which is an ideal antibody target, it is interesting to speculate whether B cells with VP1 in their BCR are more likely to present peptides derived from VP1 to CD4 T cells and whether antigen presenting B cells are providing specific signals to the CD4 T cells that induce T_FH-polarization. Another intriguing finding regarding the two different CD4 T cell mouse polyomavirus specificities was that the CD4 T cells with the highest TCR affinity were all VP1–221-specific with a population of both VP1–221- and LT678-specific IL-21-producing CD4 T cells with moderate-to-high TCR affinity [9]. VP1–221- and LT678-specific CD4 T cells both produce IFNγ [9]. These data add other questions regarding TCR signaling and specificity with regards to function. One hypothesis is that there is a “sweet spot” (high, but not too high) of TCR-signal strength that could drive CD4 T cells to become dual producers of IL-21 and IFNγ.

2.2. CD8 T cells respond to IL-21

NK cells, B cells, CD4 T cells, and CD8 T cells express IL21R [2], which, like the IL-2, IL-4, IL-7, IL-9, and IL-15 receptors contains the common γ chain. In the case of the IL21R, the common γ chain utilizes JAK3 and the α chain (the region unique to IL21R) uses JAK1 to activate STAT3 and STAT1 (Figure 1, insert) [28]. The STAT5, phosphatidylinositol 3-kinase (PI3K), and mitogen-activated protein kinase (MAPK) pathways are also activated through IL21R [28].

Similar to the production of IL-21, TCR stimulation induces the upregulation of its receptor (IL21R) on T cells [29]. In CD8 T cells, IL-21/IL21R signaling occurring at the same time as TCR stimulation promotes proliferation and the production of effector molecules such as IFNγ [2]. IL21R signaling in a CD8 T cell without TCR stimulation can result in anergy [2]. Batf, Bcl6, Bim, Eomes, Gzma, Gzmb, Il10, Maf, Prdm1, Rorγt, Socs1, and Socs3 have all been identified as target genes of IL21R-signaling [13,30]. Bcl6 and Rorγt are the transcription factors associated with the CD4 T cell helper subtypes T_FH and T_H17, respectively. As previously discussed, IL-21 is known to further induce CD4 T cells into T_FH- or T_H17-polarization. In CD8 T cells, intrinsic expression of Bcl6 was important for the generation and recall expansion of memory [31]. As in CD4 T cells, Rorγt expression is seen in CD8 T cells that produce IL-17 and associated with multiple sclerosis [32]. Gzma, Gzmb, and Prdm1 are associated with the CD8 T cell memory subset tissue resident memory (T_RM) [33,34].

CD8 T_RM are so named due to their sessile lifestyle in tissues. They have been shown to require IL-21/IL21R signaling for their formation, particularly in nonlymphoid organs such as the brain [8,9,35,36]. Several of the molecules associated with CD8 T_RM are also associated as being inducible by IL21R-sigaling within the CD8 T cell. Prdm1, a target gene downstream of IL21R-signaling, encodes Blimp1 which binds within the S1pr1, Ccr7, and Tcf7 loci [33]. The binding of Blimp1 and its homolog Hobit to these loci is thought to directly repress these genes with the end result of preventing cell egress from the tissues [33], a defining characteristic of T_RM. The expression patterns of the molecules expressed on the surface of CD8 T_RM are also associated with prevention of cell egress from the tissues. CD8 T_RM are CD69^hi and often CD103⁺ [34,37]. CD69 is a C-type lectin that causes internalization/destruction of sphingosine-1-phosphate-1 by forming a complex with sphigosine-1-phosphate-1 so that the cell no longer responds to sphingosine-1-phosphate gradients which is required for egress of the cell into the lymphatics. The CD103(αE) subunit pairs with β7 to form an integrin heterodimer which binds E-cadherin and likely tethers the cells to the epithelial cells. Another distinguishing feature of CD8 T_RM is their metabolic profile. To maintain longevity in tissues that may not be amiable to T cells, CD8 T_RM will often rely on exogenous fatty acids to support fatty acid oxidation [38,39]. In autoimmune diseases, CD8 T_RM can be drivers of the disease pathogenesis given their location, effector functions, and ability to act as sentinels in barrier tissues.

Another important feature of IL-21/IL21R signaling in CD8 T cells is that IL-21/IL21R signaling has been shown to prevent terminal exhaustion of CD8 T cells in the setting of repetitive TCR stimulation [5–7]. T_EX express multiple inhibitory receptors (e.g., PD-1, Lag-3, 2B4) and often have compromised functionality. The severity and duration of persistent antigen (in the context of infection, cancer, or autoimmune disease) has been associated with a decrease in CD8 T cell functionality and an increase in the expression of inhibitory receptors [40]. During LCMV Clone 13 infection (a chronic infection model), CD4 T cell-derived IL-21 directs CD8 T cells differentiation into a CX₃CR1⁺ subset that is more cytotoxic and provides better viral control than the CX₃CR1⁻ CD8 T cells [41]. Another study, also in the LCMV chronic infection model, showed that CD4 T cell derived IL-21 was responsible for maintaining CD8 T cell function through BATF [30], and further work from this group showed that this pathway is critical for the anti-tumor effects of CD8 T cells [12], demonstrating the redundancy between chronic viral infections and cancer. IL-21 as a rescue method of exhaustion has been more recently explored using a fusion protein of an anti-PD-1 antibody to IL-21 [42]. Combined stimulation of the IL21R on CD8 T cells with PD-1 blockade resulted in CD8 T cells that had characteristics of memory stem cells and enhanced proliferation potential [42].

One of the ways IL-21/IL21R signaling likely prevents terminal exhaustion is through the effects it confers to the cell’s metabolic profile. Our work using mouse polyomavirus brain infection to study brain T_RM differentiation showed that IL-21 treatment and CD8 T cell intrinsic IL21R expression was associated with increased expression of genes within the electron transport chain [9]. Others have shown that IL-21 treatment of CD8 T cells in vitro resulted in a decrease in glycolysis and increase in fatty acid oxidation [43]. A fatty acid oxidative metabolic profile is similar to the metabolic profile seen in CD8 T_RM, and, as described above, IL-21 has been shown to drive CD8 T cells to become T_RM [8,9,35,36]. The molecular mechanisms by which IL-21 signaling drives specific metabolic profiles within the CD8 T cell are not clear, but it has been speculated that this may occur through the PI3K pathway as PI3K has been shown to activate the mammalian target of rapamycin (mTOR) [3], which is a known regulator of T cell metabolism.

3. IL-21 in autoimmune disease

Reports differ on the exact mechanism(s) through which elevated IL-21 exerts pathogenic effects on autoimmunity. Most of the focus within autoimmune disease pathogenesis has been on the effects increased IL-21/IL21R signaling has on B cells [20]. However, there is accumulating evidence that IL-21:IL21R signaling in CD8 T cells is also driving the symptoms and pathology seen in autoimmunity as detailed below.

An emerging common theme with autoimmune disease is that CD4 T cells are the primary producers of IL-21 and generally these IL-21-producing CD4 T cells have characteristics consistent with T_FH or T_PH (Table 1). IL21R expression is less studied in the context of autoimmunity but has most often been associated with B cells (Table 1). Current understanding of IL-21/IL21R signaling within the context of specific autoimmune diseases is explored in subsequent paragraphs with an emphasis on CD8 T cell effects.

Table 1.

IL-21 in autoimmune diseases

Disease	Cellular source(s) of IL-21	IL21R expression
Systemic lupus erythematosus	ICOS+CD4 splenocytes[44] PD-1hiCXCR5− circulating CD4 T cells[45]	B cells[21*]
Rheumatoid arthritis	PD-1hiCXCR5− synovial CD4 T cells[22]	B cells[16*,22*,23] IL-17 producing CD4 T cells[22*] Synovial macrophages[48] Fibroblast-like synoviocytes[48,49] CD4 T cells, CD8 T cells, NK cells[47]
Juvenile arthritis	IFNγ and TNFα-producing synovial CD4 T cells[50]	No data
Type 1 diabetes	CD4 T cells with TFH signature[51–53] CD4 T cells with TPH signature[54]	Pancreatic CD4 and CD8 T cells[55]
Psoriasis	Skin CD4 T cells and CD161+ T cells[56]	Skin T cells, skin B cells, skin NK cells, and primary keratinocytes [56]
Sjögren’s syndrome	Circulating CXCR5+CD4 T cells, correlated with high ICOS, Bcl-6, and PD-1[57] + CXCR5 CD4 T cells in the salivary gland [58] TFH and TPH CD4 T cells[59]	Infiltrating lymphocytes and ductal cells[58] Expressed within the ectopic lymphoid structures within the salivary glands which primarily contains Bcl6+CD20+ GC B cells and TFH CD4 T cells[59]
Vitiligo	Limited data, but cytokine profile consistent with TH1/TFH CD4 T cells[60]	No data
Antiphospholipid syndrome	IL-17-producing β2GP1-reactive CD4 T cells[61]	No data
Pemphigus	CXCR5+, IL-17a-producing CD4 T cells[62] Dsg3-specific CD4 T cells[63]	No data
Giant cell arteritis	CD4 T cells[64]	TH1 and TH17 CD4 cells[42*] CD103+CD4 T cells[43*]

^*presumed based on indicated cell type’s response to blockade or stimulation

3.1. IL-21 in Systemic lupus erythematosus

Systemic lupus erythematosus (SLE) is often described as the loss of immune tolerance to nuclear components resulting in damage to various organs including the skin, joints, kidneys, and central nervous system. An elevated antinuclear antibody titer is almost always present in patients with SLE, indicating a strong correlation between immune dysfunction/dysregulation within pathways for the production of antibodies and B cell maturation. Polymorphisms in IL-21/IL21R are correlated with the development of SLE [66], but most reported data supports, that for the development of SLE disease, changes to the IL-21/IL21R pathway must primarily affect the CD4 T cell-to-B cell interactions [1,44,45,67]. In fact, IL-21 signaling to CD8 T cells in the setting of SLE might actually confer some disease protection. In a graft-vs.-host disease model of murine lupus, mice with transferred IL21R^−/− CD8 T cells had worse renal disease than those given IL21R⁺ CD8 T cells likely due to decreased cytotoxic CD8 T cell killing of autoreactive B cells [68]. Similarly, in the BXSB.Yaa mouse model of SLE, mice whose B cells were selectively deficient in IL21R had ameliorated SLE disease and mice with IL21R^−/− CD8 T cells had increased immunoglobulin, elevated antinuclear antibodies, and accelerated morbidity [69]. Another group found CD8 T cells are able to target the CD4 T_FH to limit disease [70]. These findings have prompted several researchers to describe IL-21 as a “double-edged sword” in the context of SLE as it both drives the disease and provides protection depending on through which cell type it signals [13,69].

3.2. IL-21 in Rheumatoid arthritis

Most rheumatoid arthritis (RA) patients have antibodies against cyclic citrullinated peptide (anti-CCP) and/or against the Fc fragment of IgG (rheumatoid factor) at the time of diagnosis. As such, increased or dysregulated IL-21/IL21R signaling, particularly through B cells, likely contributes to its pathogenesis. CD4 T_PH cells are the reported sources of IL-21 in the synovium which was shown through viSNE analysis of mass cytometry on synovial fluid samples of RA patients [22]. The viSNE analysis identified an expanded CD4 T cell population that was PD-1^hiCXCR5^- and expressed other molecules associated with T_PH including IL-21 production [22]. Co-culturing these synovium-derived T_PH cells with PBMC-derived B cells resulted in the B cells differentiating into plasma cells [22]. The addition of IL21R-Fc or anti-SLAMF5 to the cultures to block IL21R signaling or SLAMF5, respectively, resulted in little to no plasma cell differentiation, indicating that the synovium-derived T_PH cells were driving plasma cell differentiation through IL-21 and SLAMF5 [22]. Despite the strong support for IL-21 acting primarily on the B cells to drive RA disease, there are several reports that show that the joint inflammation and destruction in RA is not exclusively antibody driven and that IL-21 may be inducing some of these changes by acting on osteoclasts [71], fibroblast-like synoviocytes [48,49,72], synovial macrophages [48], and CD8 T cells [47] directly.

A study that compared IL21R expression on lymphocytes isolated from the synovial fluid of RA patients, osteoarthritis patients, and healthy controls showed that ~80% of the CD8 T cells in the synovial fluid of RA patients were IL21R⁺ compared to <40% in synovial fluid from osteoarthritis patients and healthy controls [47]. Culturing peripheral blood- and synovial fluid-derived CD8 T cells from RA patients with anti-CD3 and IL-21 increased expression of the T_RM-associated receptor CD69 and induced production of IFNγ and TNFα [47]. Others have reported CD8 T cells in the synovium of RA patients and in mouse models of RA that have characteristics of T_RM and further drive arthritic disease [73,74]. IL21R.Fc treatment in collagen-induced arthritis in mice and adjuvant-induced arthritis in rats reduced both the clinical and histological signs of arthritis [75]. In the collagen-induced arthritis mouse model, IL21R.Fc treatment was correlated with a decrease in IFNγ mRNA in the paws of mice [74]. This work further showed that IL-21 induced IFNγ production from CD3⁺ T cells [74], but did not differentiate whether the IL-21-induced-IFNγ was primarily produced from CD4 T cells or CD8 T cells.

3.3. IL-21 in Juvenile idiopathic arthritis

Similar to many other rheumatologic diseases, SNPs within the IL2/IL21 gene have been associated with Juvenile idiopathic arthritis (JIA) [76]. A study comparing serum cytokine levels amongst active JIA patients, patients in remission, and healthy controls showed IL-21 to be highest in the serum of patients with active JIA, but the increase was only significant when compared to the healthy controls [77]. Another group assessed the cytokine patterns secreted by peripheral blood and synovial fluid CD4 T cells, and found that while IL-21-secreting CD4 T cells were not highly represented in the peripheral blood of JIA patients, approximately 20% of the CD4 T cells in the patients’ synovial fluid were IL-21 producers [50]. Generally, the IL-21 producers, which were also capable of IFNγ and TNFα production, were significantly more likely to be present in the synovium of JIA patients with positive ANAs [50]. This association of IL-21-producing CD4 T cells with positive ANAs suggests that the CD4 T cells are T_FH-like and provide helper functions to the B cells for antibody production. However, the fact that the IL-21-producing CD4 T cells were also IFNγ and TNFα producers, indicates that these cells may also have transcriptional profiles consistent with T_H1 polarization. This is consistent with what we have observed in IL-21-producing CD4 T cells within the brains of polyomavirus infected mice—i.e. that IL-21-producing CD4 T cells within tissues where active inflammation is occurring have both T_H1 and T_FH gene signatures [9]. As T_H17 cells have been identified as integral to JIA pathogenesis [78], T_H17 cells may also be a source of IL-21 in the synovium in JIA, although this has not definitely been shown. There is also limited data on which cells may be expressing IL21R in JIA, but a case series showing success with tofacitinib (JAK1/3 inhibition) or baricitinib (JAK1/2 inhibition) in refractory JIA-associated uveitis [79] aligns with IL21R as one of the receptors driving inflammation in the eyes and joints. There is also data (albeit limited) suggesting CD8 T_RM are involved in JIA pathogenesis and joint inflammation [73].

3.4. IL-21 in Type 1 diabetes mellitus

CD4 T cells have been associated with the pathogenesis of type 1 diabetes mellitus since the late 1980’s when DNA analysis of type 1 diabetes patients revealed that alleles of HLA-DQ_β, a HLA class II gene, were associated with disease susceptibility [80]. More recently, CD4 T cells with a T_FH signature including IL-21 production were found to be elevated in the circulation of patients with type 1 diabetes mellitus [51–53]. A CXCR5⁻ population of CD4 T cells that follow a T_PH signature were also found to be elevated in the circulation of children with type 1 diabetes mellitus and associated with progression of disease [54]. In several of these studies the elevated T_FH or T_PH levels also correlated with the presence of autoantibodies in their serum against islet cells, insulin, glutamic acid decarboxylase, and/or islet antigen 2, suggesting that these CD4 helper cells may primarily interact with B cells [53,54]. However, β-cell destruction in the pancreas has been shown to be primarily CD8 T cell mediated [81,82] with CD4 T cells presumed to perform essential helper functions for these CD8 T cells either indirectly through antigen presenting cells or directly via cytokines [83,84].

How autoreactive CD8 T cells in the context of type 1 diabetes mellitus develop is an ongoing investigation. Work showing abnormal class I assembly and peptide presentation in NOD mice suggests that autoreactive CD8 T cells are produced due to a defect in negative selection early in the thymus as the MHC I molecule is unable to present a full repertoire of peptides [85,86]. However, there is also work to support IL-21 as a driver of type 1 diabetes mellitus as IL21R-deficient NOD mice do not develop type 1 diabetes mellitus [55,87], and IL-21R/Fc or anti-IL-21 treatment inhibits the development of type 1 diabetes in NOD mice [88,89]. Even more intriguing, C57BL/6 mice overexpressing IL-21 in pancreatic β cells developed type 1 diabetes [55]. This work also showed that CD8 T cells isolated from the pancreata of NOD mice had elevated expression of IL21R and that mice with pancreatic β cell expression of IL-21 had elevated chemokines and cytokines associated with activated CD8 T cells including IFNγ, MCP-1, and MCP-2 in their pancreata [55]. In the study showing that IL-21R/Fc treatment can ameliorate type 1 diabetes in NOD mice, there were significantly fewer CD8 T cells present in the pancreata after 5 days of treatment and of the CD8 T cells that were present, fewer of them expressed CD69 [88]. Anti-IL-21 treatment combined with the glucagon-like peptide-1 receptor antagonist Liraglutide was able to reverse type 1 diabetes in NOD mice and histological sections of the pancreata have a visually smaller CD8 T cell infiltrate into the islets [89].

A definitive connection between T_FH/T_PH CD4 T cells producing IL-21 and IL21R⁺ CD8 T cells responding to it in the pancreas still has to be shown. More recent work on CD8 T cells in type 1 diabetes has focused on their memory subset profiles. One such study that obtained pancreatic samples from recent-onset type 1 diabetes patients revealed that CD8 T cells are the most abundant T cell phenotype in the islets of their pancreata [90]. Immunofluorescent staining of these samples suggest the CD8 T cells as being primarily tissue-resident memory as the majority are CD69⁺CD103⁺ [90]. A different study assessed islet-specific CD8 T cells acquired from type 1 diabetic patients’ PBMCs and showed that these CD8 T cells could be designated into 3 different CXCR3⁺ clusters: an activated transitional memory phenotype with high expression of HELIOS and CD27, a transitional memory phenotype that was CD27^hi but HELIOS⁻, and an exhausted phenotype that was EOMES^hiTbet^int and expressed several inhibitory receptors including PD-1 [91]. Patients with more of the exhausted phenotype had slower type 1 diabetes disease progression [91]. While the IL-21 levels in these patients were not assessed, given the data supporting IL-21 as protective from CD8 T cell exhaustion during chronic LCMV [5–7], one possible explanation for differences in the size of the exhausted CD8 T cell compartment between the two groups could be IL-21—i.e., patients with fewer exhausted CD8 T cells and faster disease progression have increased IL-21 while the patients with slower disease progression have lower IL-21 levels. In another study, epigenetic analysis of β-cell specific CD8 T cells from human PBMCs, pancreatic-draining lymph nodes of NOD mice, and the pancreas of NOD mice showed that CD8 T cells within the circulation and lymph nodes retain a epigenetic profile most similar to a less differentiated stem cell memory T cell phenotype [92]. The cells isolated from the pancreas are more terminally differentiated and express markers consistent with tissue resident memory such as low CD62L and high PD-1 [92]. These findings are reminiscent of the results showing that high TCF-1 expression in CD8 T cells during LCMV Clone 13 infection can be ascribed as “stem cell-like” and are more responsive to PD-1/PD-L1 blockade [93,94]. Of note, the TCF-1^hi PD-1⁺ CXCR5⁺ CD8 T cells that were considered “stem cell-like” were found primarily in the lymph nodes and are considered precursors to the terminally differentiated non-PD-1/PD-L1 blockade responders that are in the tissues [94]. This highlights the importance of studying the CD8 T cells within the pancreas as they are likely to be phenotypically, functionally, and metabolically distinct from the CD8 T cells in the draining lymph nodes and circulation and, thus, they likely respond differently and may rely more on IL-21 to exert their pathogenicity.

3.5. IL-21 in Psoriasis

Psoriasis and psoriatic arthritis are generally associated with elevated IL-22 and IL-23. A publication by Caruso et al in 2009 was one of the first to show psoriatic plaques had elevated expression of IL-21 [56]. This group also showed that the IL-21 was produced primarily by CD4 T cells with the remainder (approximately 30% of the IL-21 producing cells) being CD161⁺ NK T cells [56]. A small percentage of the IL-21-producing CD4 T cells were co-producing IFNγ or IL-17 [56], indicating a few might have T_H1- or T_H17-polarization. They did not assess further expression patterns, however, including that of CXCR5. Analysis of IL21R showed that it was present on T, B, and NK cells within the skin as well as keratinocytes [56], which led this group to focus on IL-21/IL21R signaling within keratinocytes. Using a human psoriasis xenograft mouse model, they showed that IL-21 blockade reduced keratinocyte proliferation and epidermal thickening in the mice, independent of IL-22 and IL-23 [56]. Although they showed that IL-21 was able to act on the keratinocytes to promote proliferation and epidermal thickness directly, the skin from anti-IL-21 treated mice also had decreased expression of IFNγ and IL-17A RNA [56]. Their xenograft mouse model included transfer of autologous PBMCs that were pre-activated through CD3 and CD28 in vitro; the decrease in IFNγ and IL-17A RNA expression after anti-IL-21 treatment suggests IL-21 also signals through CD4 and CD8 T cells to promote inflammation in the skin.

A SCID mouse model of psoriasis supports T_RM as sustaining the psoriatic skin lesions. Injections of lymphocytes derived from the peripheral circulation or psoriatic skin lesion of psoriasis patients into SCID mice resulted in only the mice receiving lymphocytes from the psoriatic skin lesions maintaining the psoriatic phenotype [95]. In other experiments with the xenotransplant mouse model of psoriasis, targeting CD8 T cells with anti-CD8 significantly reduced the papillomatous index and psoriasis development [96]. The CD8 T cells found in the skin of patients with psoriasis (even skin without psoriasis lesion) are skewed towards IL-17 production [97,98], and have an association with requiring IL-21/IL21R signaling for differentiation.

Tofacitinib is a JAK1/3 inhibitor approved for use in psoriasis [99–101]. Interestingly, the IL-22 receptor (IL22R) and IL-23 receptor (IL23R) use JAK1/Tyk2 and JAK2/Tyk2, respectively. While Tofacitinib would have some weak blockade on downstream signaling through IL22R, it primarily would inhibit the IL-2, IL-4, IL-15, and IL-21 receptors which all signal through JAK1/3. The success of tofacitinib in psoriasis treatment further indicates a potential role for elevated IL-21/IL21R signaling in the pathogenesis of this disease.

3.6. IL-21 in Sjögren’s syndrome

Primary Sjögren’s syndrome is characterized as the autoimmune destruction of exocrine glands. Histopathologic examination of minor salivary glands from patients with primary Sjögren’s syndrome revealed an infiltrates of immune cells that were approximately 40% CD4 T cells, 33% B cells, and 15% CD8 T cells in mild disease with a higher percentage (>50%) of B cells and fewer CD4 T cells in more moderate disease [102]. Anti-Ro (SSA) and anti-La (SSB) are the autoantibodies acquired in Sjögren’s syndrome. The presence of autoantibodies within this disease and high percentage of B cells in the salivary gland are consistent with aberrant IL-21 signaling primarily driving disease through B cells. Several studies support this hypothesis by showing elevated IL-21 levels in the sera of patients with Sjögren’s syndrome that correlates with elevated anti-Ro/anti-La titers and plasma cells in the salivary glands [58,59,103].

CD4 T cells are the reported sources of IL-21 in Sjögren’s syndrome. Li et al found elevated CXCR5⁺CCR6⁺ CD4 T cells in the blood of patients with primary Sjögren’s syndrome [57]. Most of these cells were also PD-1⁺, CD40L⁺, ICOS⁺, Bcl-6⁺, and IL-21⁺, and although Li et al referred to the CXCR5⁺CCR6⁺ cell as T_H17-like, these expression patterns are most consistent with T_FH [57]. Similarly, Kang et al showed that IL-21 expressing lymphocytes in the labial salivary glands of primary Sjögren’s syndrome patients were CXCR5⁺ CD4 T cells [58], and Pontarini et al showed Sjögren’s syndrome patient-derived circulating ICOS⁺ CD4 T cells that were either CXCR5⁺ (T_FH) or CXCR5⁻ (T_PH) producing IL-21 [59]. Histology of the salivary gland tissue from patients with Sjögren’s syndrome revealed ectopic lymphoid structures; transcriptomic analysis of these ectopic lymphoid structures was noteworthy for elevated expression of genes associated with a T_FH signature, supporting T_FH cells producing IL-21 within the salivary glands [59].

The CD8 T cells within the glands of Sjögren’s syndrome patients have been localized near acinar epithelial cells [104–106]. While there is limited data linking IL-21 signaling in being important to their differentiation and cytotoxic capabilities within primary Sjögren’s syndrome, these cells are primarily CD103⁺ and utilize both the Fas/FasL and perforin/granzyme B pathways to mediate epithelial cell apoptosis [104–106], identifying them as likely T_RM. Interestingly, Goa et al reported that CD8 T cells outnumbered CD4 T cells in the labial salivary glands of patients with primary Sjögren’s syndrome [106], which stands in contrast to previous assessments showing CD4 T cells as the predominant infiltrate within minor salivary glands of Sjögren’s syndrome patients [102]. This difference may be dependent on the specific gland and whether or not it is able to support an ectopic lymphoid structure such as a germinal center which would mostly contain B and CD4 T cells. In glands or tissues less amendable to ectopic lymphoid structures, CD8 T_RM may play a more central role in the pathogenesis of Sjögren’s syndrome. Furthermore, CD4 T cell help through IL-21 could be fueling the CD8 T cell differentiation into T_RM.

3.7. IL-21 in Vitiligo

Serum IL-21 was found to be increased in vitiligo patients compared to healthy controls with higher IL-21 serum levels associating with higher disease activity [107]. Using the Smyth line chicken animal model of vitiligo, one group showed elevated transcripts of IFNγ, IL-10, and IL-21 in the growing feathers of the chickens during active vitiligo [60]. This group also observed a lymphocyte infiltrate to the active vitiligo lesions which were primarily CD8 T cells [60]. CD4 T cells were also present in the lesions as were a small infiltrate of B cells [60]. In humans afflicted with vitiligo, CD8 T cells isolated from active lesions were identified as T_RM due to high expression of CD69 and CD103 and increased production of IFNγ and TNFα [108]. Many of the CD8 T cells in the lesions were CXCR3⁺, suggesting a pathway by which these cells enter into the skin is reliant on IFNγ-induced expression of CXCL9/10 [108].

3.8. IL-21 in Antiphospholipid syndrome

Antiphospholipid syndrome is driven by the production of autoantibodies against β2-Glycoprotein 1. Benagiano et al showed CD4 T cells that infiltrate the atherosclerotic lesions of lupus-antiphospholipid syndrome patients produced IL-17, IL-21 and IFNγ when stimulated with β2-Glycoprotein 1, indicating most of these CD4 T cells had T_H17 polarization [61]. This same group tested the clones of CD4 and CD8 T cells reactive to β2-Glycoprotein 1, and found that only the CD4 T cells were reactive [61]. This result, combined with their data showing that isolated CD4 T cells when cultured with B cells induced IgM, IgG, and IgA synthesis, indicates that the IL-21 produced by CD4 T cells is primarily working on B cells and not CD8 T cells.

3.9. IL-21 in Pemphigus

Autoimmune diseases against adhesion molecules of the skin are often delineated as loss of intraepidermal adhesion (of which the most common is pemphigus vulgaris) or a subepidermal loss of adhesion (of which the most common is bullous pemphigoid) [109]. Pemphigus vulgaris is highly associated with autoantibodies to desmoglein 1 and 3 [110], and bullous pemphigoid is associated with autoantibodies to the hemidesmosomal protein (anti-BP180 and anti-BP230) [111,112]. Pemphigus vulgaris patients have been shown to have elevated IL-21 in their plasma and increased T_FH CD4 T cells in their whole blood when compared to both myasthenia gravis patients and healthy controls [63]. Assessment of both pemphigus vulgaris and bullous pemphigus lesions revealed a large number of B and T lymphocytes and plasma cells in the lesions as compared to healthy skin [62]. Greater than 50% of the CD3⁺ lymphocytes in the pemphigus vulgaris lesions were CD4⁺ and a proportion of those CD4 T cells were producing IL-21 [62]. This group also showed a proportion of the CD19⁺ lymphocytes isolated from pemphigus vulgaris lesions were specific for desmoglein 1 or 3 using recombinant Dsg 1-His-Flag or Dsg 3-His-Flag [62]. More recently a study assessing serum levels of IL-21 in pemphigus vulgaris patients before and after treatment with rituximab showed that although IL-21 levels increased after treatment, the patients’ Pemphigus Disease Area Index significantly decreased, suggesting that IL-21 is primarily exerting its effects on pathogenesis through B cells [113]. However, it is worth noting that rituximab treatment did not decrease the damage scores [113]. The lack of change in damage scores could be due to limitations regarding the duration of the study (only 2 months of therapy). However, as approximately 50% of the T cells in the pemphigus lesions are presumably CD8 T cells [62], it may indicate that IL-21 is also acting on CD8 T cells in the lesions. While CD8 T cells in the context of the pemphigus diseases have not been studied as extensively as the CD4 T cells, B cells, and antibodies, it has been shown that pemphigus vulgaris and bullous pemphigoid patients develop autoreactive cytotoxic T lymphocytes specific for pemphigus and pemphigoid antigens [114]. It also has been shown that CD8-deficient mice have a lower incidence of pemphigus disease using a passive transfer model when compared to wild-type controls [115].

3.10. IL-21 in Giant cell arteritis

CD4 T cells with expression patterns consistent with T_H1, T_H17, and T_FH cells have been found elevated in the peripheral blood and temporal arteries of patients with giant cell arteritis [64,65]. These CD4 T cell populations correlate with elevated IL-21 [64,65], suggesting a potential source of the IL-21 as well as indicating these CD4 T cell populations may require IL-21 for their pathogenetic effects. One of these groups also utilized a mouse model of giant cell arteritis that consisted of human-artery mouse chimeras which showed the presence of CD103⁺ CD4 T cells in the artery wall that they identified as CD4 T_RM [65]. These mice were also treated with tofacitinib which resulted in fewer total CD4 T cells, decreased IL-21 mRNA, and less intimal hyperplasia within the arteries [65].

CD8 T_RM have also been implicated in giant cell arteritis with the intensity of the CD8 T cell infiltrate into the artery walls being predictive of disease severity [116]. Expression of CXCR3 on the CD8 T cells and immunohistochemistry studies of CXCL9 and CXCL10 expression in patient temporal artery biopsies supports CXCL9/10 as the mechanism of T cell infiltration [116]. CD4 T cell-derived IFNγ has been shown to induce CXCL9 and CXCL10 expression to recruit CD8 T cells into the female reproductive tract in HSV-2 infection and lungs in influenza virus infection and drive the CD8 T cells to differentiate into T_RM [117,118]. Whether or not CD4 T cell-derived IFNγ or IL-21 can drive CD8 T_RM differentiation in giant cell arteritis is not known.

4. IL-21/IL21R blockade as a treatment

Biologics are becoming more widely used for the treatment of autoimmune disease. These include rituximab, obinutuzumab, and ocaratuzumab, which are monoclonal antibodies against CD20 that lead to B cell apoptosis and are used as treatments for SLE and RA. However, as discussed above, focusing only on eliminating or blocking the B cells could miss the potential pathology induced through the CD4 and CD8 T cells. Modulation of the IL-21/IL21R signaling pathway is a current focus within autoimmune disease therapy (Table 2). Figure 2 illustrates current medications in use or trial for the autoimmune diseases discussed in this review that specifically inhibit steps within the IL-21:IL21R signaling pathway. Of note, current outcome measures in the clinical trials of these molecules have not focused on CD8 T cell-specific effects, likely because IL-21/IL21R signaling of CD8 T cells within the context of autoimmunity is not well-studied. However, given what is known about IL21R signaling within CD8 T cells, these molecules would likely lead to CD8 T cell exhaustion, death, or anergy.

Table 2.

clinical trials assessing IL-21/IL21R blockade as treatment as found on clinicaltrials.gov

Disease	Intervention	Clinical trial #	Details	Results
Systemic lupus erythematosus	BOS161721	NCT03371251	Phase I and II clinical trial. Safety, tolerability, and efficacy of repeat doses of BIO161721 in patients with moderate-severe SLE.	Ongoing.
Rheumatoid arthritis	NNC0114–0006	NCT01565408	Phase I clinical trial. Safety and efficacy of NNC0114–0006 in RA patients.	Results not reported.
		NCT01647451 [120]	Phase II clinical trial. Use of NNC0114–0006 with methotrexate in RA patients vs. placebo and methotrexate.	• Appears safe and well-tolerated. • Significant improvement in DAS28-CRP at week 12. • Higher number of reported infections (24% in treated group vs. 10% in placebo).
	Tofacitinib	NCT01262118 NCT01484561 NCT00147498 NCT00550446 NCT00603512 NCT00687193 NCT01164579 NCT00976599 NCT01059864 NCT01359150 NCT00960440 NCT00847613 NCT00814307 NCT00856544 NCT00853385 NCT01039688 NCT00413699 NCT00661661	Phase I, II, and III clinical trials. CP-0690,550 (Tofacitinib) in patients with active RA	• Appears safe and well-tolerated[129]. • Inhibits radiographic progression of RA disease[130].
	HL237	NCT03278470 & NCT03896594	Phase I clinical trial. Trial to find the maximum tolerable dose of HL237 in healthy male	Ongoing.
Juvenile arthritis	Upadacitinib	NCT03725007	Phase I clinical trial. Study to evaluate pharmacokinetics, safety, and tolerability of upadacitinib in pediatric patients with JIA.	Ongoing.
Type 1 diabetes	NNC0114–0006	NCT02443155 [121]	Phase II clinical trial. Conducted at 94 sites in Europe, US, and Canada. NNC0114–0006 12 mg/kg IV q6weeks with/without liraglutide	• Appears safe and well-tolerated. • No major changes noted to WBC populations in blood. • Participants receiving NNC0114–0006 had higher C-peptide levels at the end of the study. • Of the 153 participants receiving NNC0114–0006, ~15% developed anti-NNC0114–0006 antibodies.
Psoriasis	Tofacitinib	NCT01877668 NCT01882439 NCT01976364 [99–101]	Phase III clinical trials. Assessing long-term safety, tolerability, and efficacy of tofacitinib in patients with active psoriasis.	• Appears safe and well-tolerated. • Improvements in Disease Activity in Psoriatic Arthritis were maintained up to 30 months. • Improvements in Composite Psoriasis Disease Activity Index continued over time with use.
Vitiligo	Ruxolitinib	NCT03099304 [131]	Phase II clinical trial. Multicenter, randomized, double-blind study in 26 US hospitals in 18 states. 52 weeks of treatment with ruxolitinib cream	• Appears safe and well-tolerated. • Substantial repigmentation of vitiligo lesions with use of ruxolitinib cream
		NCT04057573 and NCT04052425	Phase III clinical trial. Crossover assignment of >330 participants to assess topical ruxolitinib treatment in patients with non-segmental vitiligo	Ongoing.
Giant cell arteritis	Upadacitinib	NCT03725202	Phase III clinical trial. Multicenter study to evaluate safety and efficacy of upadacitinib in combination with 26-wk corticosteroid taper vs. 52-wk corticosteroid taper + placebo.	Ongoing.
Immune skin conditions in Down Syndrome	Tofacitinib	NCT04246372	Phase II clinical trial. 46 participants. Through the Linda Crnic Institute for Down Syndrome. Tofacitinib 5 mg PO bid for 16 weeks.	Ongoing.

Figure 2.

Current therapies that affect IL-21/IL21R pathway, either on the market or in clinical trials to treat autoimmune diseases. The insert shows the intracellular signaling of IL21R that would likely be occurring in the CD8 T cells, CD4 T cells, and B cells within the tissue simultaneously. NNC0114–0006 is a monoclonal antibody against IL-21 that is currently in clinical trials for RA and type 1 diabetes mellitus; BOS161721 binds to IL-21 so that it can no longer bind the IL21R and currently is in trials for SLE. ATR-107 is a monoclonal antibody against the IL21R. Tofacitinib (JAK1/3 inhibitor) and Upadacitinib (JAK1 inhibitor) prevent IL21R downstream signaling and are in consideration for RA, psoriasis, JIA, vasculitis, and other skin conditions. HL237 is a STAT3 inhibitor with potential for treating RA. Figure image created with BioRender.com.

4.1. NNC0114–0006 (anti-IL-21 monoclonal)

NNC0114–0006 (also known as NNC0114–0005) is an IL-21 neutralizing monoclonal antibody that has undergone clinical trials in the context of SLE, RA, Crohn’s disease, and type 1 diabetes mellitus [119–121]. As shown in Figure 2, NNC0114–0006 binds/inhibits IL-21 which would prevent IL-21 from activating any of the cells expressing its receptor. Most studies show that NNC0114–0005 and NNC0114–0006 are safe and well tolerated, but have shown only mild improvements in disease activity score 28 (DAS28-CRP), which were calculated using C-reactive protein, for patients with RA [119,120] or C-peptide levels in type 1 diabetes [121]. The only mild improvements in disease activity scores observed with NNC0114–0006 use could be due to CD4 T cells producing more IL-21 or that other cytokines with redundant functions to IL-21 are able to overcome or bypass NNC0114–0006 effects.

4.2. BOS161721

BOS161721 is a humanized IgG1 monoclonal antibody that binds to IL-21 and inhibits its bioactivity by preventing its binding to the IL21R, rendering it similar in mechanism to NNC0114–0006. This molecule has been evaluated in healthy subjects with the initial findings that is safe and well-tolerated, suppresses IL-21-induced pSTAT3 dose-dependently, and produces dose-dependent downregulation of the IL-21 target genes BATF, IL6, LAG3, and SOCS3 [122]. Currently, it is in clinical trials to treat SLE.

4.3. ATR-107 (anti-IL21R monoclonal)

ATR-107 is a human monoclonal IgG1 antibody directed against the IL21R. The IgG1 constant region of ATR-107 contains mutations so that this antibody primarily has antagonist activity with the receptor. Safety and efficacy studies in healthy adults indicate that even a single dose can maximally occupy IL21Rs on B cells through 42 days [123]. However, approximately 75% of the participants receiving ATR-107 developed anti-drug antibodies [123,124]. Due to the high incidence of anti-drug antibodies in healthy controls, there have been no further studies of this monoclonal as a therapy.

4.4. Tofacitinib and Upadacitinib (JAK inhibitors)

The IL21R utilizes JAK1 and JAK3 to phosphorylate STAT1 and STAT3 and exert its effects. Therefore, an inhibitor to JAK1 or JAK3 would prevent downstream IL21R signaling. JAK inhibitors are becoming more widely used as therapy for autoimmune diseases. The JAK1/3 inhibitors tofacitinib and upadacitinib have been shown to effectively decrease phosphorylation of STAT3 in CD4 T cells, NK cells, and monocytes [125], indicating effective inhibition of IL21R signaling. Tofacitinib and has so far shown success in treating RA, refractory JIA-associated uveitis [79], psoriasis [99–101], and alopecia areata in Down syndrome [126]. Ruxolinib, a JAK1/2 inhibitor that appears to also inhibit IL21R signaling [127], has shown promising results in treating vitiligo (Table 2). Of note, JAK inhibitors will not only act on IL21R downstream signaling as several other cytokine receptors including the IL-15 receptor signal through JAK1/3.

4.5. HL237 (STAT3 inhibitor)

STAT3 is one of the primary downstream transducers of IL21R signaling. Studies support STAT3 inhibition as a promising therapy for RA [128]. Therefore, the biguanide HL237, which is a STAT3 transcription inhibitor, is currently being investigated for safety and efficacy in RA. As the IL21R does not only signal through STAT3, HL237 may be less effective than JAK inhibitors for treatment. Furthermore, STAT3 is implicated in many different pathways within immune cells which could result in more off-target effects.

5. Conclusion

Increased IL-21 levels are associated with many autoimmune diseases and likely contribute to autoimmune disease pathogenesis. CD4 T cells with T_FH/T_PH polarization are the most likely sources of IL-21 in the tissue(s). The CD4 T cell-derived IL-21 is acting on several immune subsets particularly B cells and CD8 T cells within the tissues to promote antibody production as well as CD8 differentiation into T_RM. Targeting IL-21/IL21R signaling has shown success in inducing remission of several autoimmune diseases, providing a promising future for many patients.

• Elevated serum and/or tissue interleukin-21 is associated with autoimmune disease.

• Follicular helper-like CD4 T cells are often the sources of IL-21 in the tissues.

• CD8 T cells respond to IL-21 in ways that may further drive autoimmune disease.

• Targeting IL-21:IL21R signaling may offer patients with autoimmune disease relief.

Abbreviations

ANA

Anti-nuclear antibody

CCP

Cyclic citrullinated peptide

JAK

Janus kinase

JIA

Juvenile idiopathic arthritis

IL

Interleukin

IL21R

Interleukin 21 receptor

LCMV

Lymphocytic choriomeningitis virus

MAPK

Mitogen-activated protein kinase

mTOR

Mammalian target of rapamycin

NK

Natural killer

NOD

Non-obese diabetic

PBMC

Peripheral blood mononuclear cell

PD-1

Programed death-1

PI3K

Phosphatidylinositol 3-kinase

SLE

Systemic lupus erythematous

SCID

Severe combined immunodeficient

SNP

Single nucleotide polymorphism

STAT

Signal transducer and activator transcription

TCR

T cell receptor

T_EX

Exhausted T cell

T_FH

Follicular helper T cell

T_H

Helper T cell

T_PH

Peripheral helper T cell

T_RM

Tissue resident memory

WBC

White blood cell