IL-9 and T_h9 cells: progress and challenges

Abstract

T_h9 cells are a new subset of helper T cells, and the signature cytokine for T_h9 cells is IL-9. Both T_h9 cells and T_h9 products are implicated in multiple disease settings. Thus, a clear understanding of how T_h9 cells are induced and controlled is an important and clinically relevant issue. There are different molecular pathways identified thus far in the induction of T_h9 cells, and activation of such diverse pathways requires integration of signals from TGF-β and IL-4 cytokine receptors as well as costimulatory molecules. These signals converge on the induction of multiple transcription factors that collectively drive the development of T_h9 cells.

Introduction

IL-9 was originally cloned in 1989 from murine helper T-cell clones (1), so it is by no means a newly discovered cytokine. IL-9 per se has been extensively studied; IL-9 belongs to a family of cytokines that use the common IL-2Rγc for signal transduction, and similar to other family members (i.e. IL-2, IL-4, IL-7, IL-15 and IL-21), IL-9 was believed to be a T-cell growth factor and its chief function was to drive T-cell proliferation. But later studies showed that IL-9 has a weak effect in proliferation of primary T cells (2), despite the fact that proliferation of certain T-cell clones can be strongly stimulated by IL-9. Instead, IL-9 exhibits other functions, most noticeably in proliferation of mast cells, goblet cells and airway mucin-producing cells. Thus, in many ways, IL-9 is different from other γc cytokines as a T-cell growth factor. IL-9 signals through the JAK/STAT system. Specifically, upon binding to its cell surface receptor, which consists of a private IL-9Rα chain and the common γc, IL-9 induces recruitment of JAK1 and JAK3 to the IL-9Rα chain and the common γc, respectively, followed by cross-phosphorylation and activation of JAK1 and JAK3. This leads to the activation of STAT1, STAT3 and STAT5. Consequently, STAT1 and STAT5 form homodimers, while STAT1 and STAT3 form heterodimers, and such dimeric complexes translocate to the nucleus to drive transcription of IL-9-inducible genes (3). These gene products are involved in cell survival, proliferation and secretion of inflammatory mediators.

IL-9 is often seen in the context of T_h2 cells in vitro or T_h2-associated inflammatory conditions in vivo, especially in allergic inflammation (4, 5). Thus, for a long time, IL-9 was considered just another T_h2 cytokine and thought to be redundant among other T_h2 cytokines (i.e. IL-4, IL-5 and IL-13) (6, 7). Furthermore, IL-9 is not confined to T_h2 cells, and other cell types including mast cells, NKT cells, T_h17 cells or even T_reg cells can become IL-9 producers (1, 8–15). Moreover, a recent study demonstrated by using IL-9-Cre reporter mice that even innate lymphoid cells are significant producers of IL-9 (16). So, IL-9 seems to be one cytokine of many sources, and therefore, interest in IL-9 biology and in its significance is diluted, and study of IL-9 has lagged behind that of others. The recent discovery that IL-9-producing cells are in fact a unique subset of CD4⁺ helper T cells that is different from other subsets, with distinct features and transcriptional controls, generates renewed interest in the field.

In this review, we summarize the latest advances in the study of T_h9 cells, discuss the evolving conditions that promote their differentiation as well as the in vivo relevance of T_h9 cells and finally we highlight some outstanding issues that remain to be resolved.

Defining T_h9 cells

Naive CD4⁺ T cells can be further specialized into functionally different subsets upon activation (e.g. T_h1, T_h2, T_h17, T_h22 and T_reg cells), which are often measured by the distinct cytokine profiles they express (17–20). Subset specialization is driven primarily by the texture of cytokines in the local environment where naive the T cells are activated, with the induction of lineage-specific transcription factors as a critical event in further development of specific subsets (21). T_h9 cells are a recently described new helper T-cell subset; the signature cytokine for T_h9 cells is IL-9 (without IL-4). T_h9 cells, together with other helper T-cell subsets, form a complex array of effector mechanisms in the immune system.

In many aspects, T_h9 cells are a unique helper T-cell subset. For example, in most studies, the frequency of T_h9 cells is very low (~5%), even under optimal polarizing conditions in vitro (22). This often casts considerable concerns over whether T_h9 cells are truly a distinct helper T-cell subset. Also, T_h9 cells are closely associated with T_h2 cells, as T_h2 cells co-express both IL-4 and IL-9 in the early phase of T_h2 differentiation, and the T_h2 cytokine IL-4 provides one of the key signals for T_h9 induction (23).

Furthermore, some of the transcription factors in T_h2 development are also involved in T_h9 induction. A clear example is that STAT6 knockout CD4⁺ T cells fail to develop to T_h2 cells; they also fail to become T_h9 cells (24). However, T_h9 cells are not T_h2 cells. As discussed below, the culture conditions and cytokine milieu that lead to T_h2 and T_h9 cells are very different. In some T_h2 cultures, CD4⁺ T cells that express IL-4 (T_h2 cells) and IL-9 (T_h9 cells) are completely segregated in that only those that lose the ability to express IL-4 will become IL-9 producers (23). Interestingly, only a small fraction of T_h2 cells acquire the ability to continually express IL-9. Importantly, the transcriptional regulation mechanisms of T_h9 and T_h2 cells are strikingly different from each other, thus clearly setting T_h9 and T_h2 cells apart (23).

Are T_h9 cells progeny of T_h2 cells? In most reports showing low levels of T_h9 cells under TGF-β and IL-4 culture conditions, T_h9 cells often co-express IL-10, which is another T_h2 cytokine (25). It is likely that such T_h9 cells are derivatives of T_h2 cells as a consequence of induction of additional transcription factors such as PU.1 (purine-rich box 1) and IRF4 (interferon regulatory factor 4) (26–29), which shut off IL-4 and turn on IL-9 (see below). In this setting, T_h2 cells are likely intermediaries that may further differentiate to T_h9 cells. However, our own studies suggest another pathway of T_h9 induction in which naive CD4⁺ T cells can be directly converted to T_h9 cells at high levels (up to 80% of the CD4⁺ T cells) by TGF-β and IL-4 when OX40 costimulation is engaged (30). Furthermore, we demonstrated that the non-canonical NF-κB (RelB–p52) pathway (the canonical pathway involves RelA–p50) rather than PU.1 and IRF4 is essential to T_h9 induction (Fig. 1).

Fig. 1.

Pathways of T_h9 induction. Naive CD4⁺ T cells can be converted to IL-9-producing T_h9 cells via different molecular pathways. Depending on the presence or absence of OX40 costimulation, T_h9 cells can develop from a subset of T_h2 cells (shown in purple) or directly from CD4⁺ T-cell precursors under TGF-β/IL-4-polarizing conditions.

T_h9 induction

Cytokines

A complex cytokine milieu is required for T_h9 induction, and integration of multiple cytokine signals is critical to optimal T_h9 development. The best cytokine mixture for T_h9 induction is a combination of TGF-β and IL-4, which contrasts sharply to the role of individual cytokines. TGF-β alone without IL-4 promotes T_reg cells by inducing Foxp3, whereas IL-4 alone without TGF-β supports T_h2 induction. This highlights the complexity of T_h9 induction and also places T_h9 cells as a unique subset that is different from T_h2 and T_reg cells. IL-4 activates STAT6 and IRF4, whereas TGF-β activates PU.1 and represses T-bet (T-box expressed in T cells) and GATA3 (GATA-binding protein 3), and the integration of those events eventually drives IL-9 expression (18, 31–35).

In some models, IL-1 favors induction of T_h9 cells; so does IL-25 or IL-33 (16, 36–38), although the exact mechanism remains to be defined. On the other hand, IFN-γ and IL-23 are potent inhibitors of T_h9 induction. Also, cytokines that stimulate IFN-γ production such as IL-12 and IL-18 also inhibit the induction of T_h9 cells. These cytokines most likely act through the induction of T-bet, which promotes T_h1 cells and opposes other helper T-cell lineages including T_h9 cells. Thus, the texture of cytokines fine-tunes the production of different helper T-cell subsets.

Costimulatory signals

T-cell costimulatory signals control not only the status of T-cell activation but also the character of the T-cell response. We recently showed that OX40, a costimulatory molecule in the TNFR superfamily (TNFRSF), is surprisingly potent in promoting T_h9 cells (30), thus emphasizing the importance of costimulation, in addition to cytokines, in T_h9 induction. OX40 is expressed by activated, but not resting, T cells, especially activated CD4⁺ T cells, and plays an important role in cell survival and proliferation (39). Specifically, we found that under T_h9-polarizing conditions, OX40 ligation on naive CD4⁺ T cells resulted in a remarkable increase in T_h9 induction. Such T_h9 cells did not express detectable levels of IL-4 or IL-10 (30). Furthermore, OX40 ligation under T_reg- and T_h17-polarizing conditions potently inhibited the induction of Foxp3⁺ cells and IL-17-producing cells (30). Thus, the effect of OX40 on T_h9 induction seems to be specific.

The role of OX40 in the induction of T_h9 cells was also observed in vivo where OX40L transgenic mice or injection of agonist anti-OX40 antibody induced allergic airway inflammation, as demonstrated by goblet-cell metaplasia and eosinophil infiltration (30). Mechanistically, OX40 signaling activates and sustains induction of the non-canonical NF-κB pathway (RelB–p52), which is critical to IL-9 transcription. This is further confirmed using both gain-of-function assays and loss-of-function assays. Under T_h9-polarizing conditions, overexpression of RelB–p52 in CD4⁺ T cells led to much greater T_h9 induction by TGF-β and IL-4, and knockout of p52 drastically reduced T_h9 induction (30), which places RelB–p52 as a center piece in the induction of T_h9 cells. While interesting, these data also raise several questions on the role of other TNFRSF members in the induction of T_h9 cells, especially those that activate the non-canonical NF-κB pathway. Studies in this area deserve more attention.

Other costimulatory molecules that are known to affect T_h9 induction include the CD28 and Notch pathways. It has been shown that conditional deletion of Notch1 and Notch2 markedly decreased IL-9 production in T_h9 cultures (40). There are multiple ligands for the Notch receptors, and Jagged2 but not Delta-like 1 was shown to induce IL-9 production under TGF-β-based polarizing conditions. In an experimental auto-immune encephalitis (EAE) model, Jagged2-mediated IL-9 production was involved the EAE pathology, and conditional deletion of Notch1 and Notch2 in T cells attenuated the disease (40).

The exact mechanism of how Notch promotes the generation of T_h9 cells remains unclear. Notch signaling is known to favor T_h2 cells (41, 42), which may indirectly promote T_h9 cells. This notion is supported by the finding that exogenous IL-4 could overcome the effects of Notch1 and Notch 2 deficiency. Other studies indicate that Notch may modulate TGF-β signaling by acting on Smad (small/mothers against decapentaplegic) proteins. For example, in humans, Smad3 binds Notch 1, whereas Jagged2 and Delta-like 1 bind Notch 2 (43). It should be noted that T_h9 cells induced upon Notch stimulation are also at low levels, but PU.1 and IRF4 appear not critical to Notch-mediated induction of T_h9 cells.

Transcription machinery

Cytokine signals and costimulatory signals converge to activate the transcriptional apparatus that eventually drives the differentiation of T_h9 cells. Unlike other helper T-cell subsets, there is a hierarchy of transcription factors involved in both induction and differentiation of T_h9 cells. But a single ‘master’ transcription factor, as shown in other subsets, has not been identified thus far for T_h9 cells.

The IL-9 promoter region plus two additional regions of conserved non-coding sequences upstream of the promoter region form the cis- and trans-regulatory elements collectively regulating IL-9 gene expression (26, 35). Sequence analysis identified binding sites for a plethora of transcription factors, which include PU.1, IRF4, STATs, NFAT (nuclear factor of activated T cells), GATA1, GATA3, Smads and Notch as well as NF-κB and AP-1 (activator protein 1) (35), highlighting the complexity of IL-9 gene regulation. In a broad sense, the transcriptional control of T_h9 cells induced by polarizing cytokines (TGF-β and IL-4) and the polarizing cytokines plus costimulatory signals appears to be very different.

Under TGF-β- and IL-4-polarizing conditions, PU.1 and IRF4 have been identified as key transcription factors in the induction of T_h9 cells (26, 27). Overexpression of PU.1 in CD4⁺ T cells facilitates T_h9 induction by TGF-β and IL-4, and deficiency of PU.1 inhibits the induction of T_h9 cells. Furthermore, PU.1 knockout mice exhibit reduced allergic lung inflammation in which T_h9 cells are known to be involved (26). Using a similar experimental strategy, IRF4 was shown to display the same effect on T_h9 development as PU.1 (27). Mechanistically, both transcription factors have been shown to bind to the promoter region of IL-9 and are capable of promoting IL-9 gene expression.

It should be noted that T_h9 development requires signals from both TGF-β and IL-4 cytokine receptors. In the absence of IL-4, TGF-β promotes T_reg cells, and in absence of TGF-β, IL-4 leads to the development of T_h2 cells. As T_h9 cells are closely related to T_h2 cells, IL-4-mediated induction of STAT6 and the STAT6 target gene GATA3 are both required for T_h9 development (24, 25). However, it is not clear how STAT6 and GATA3 function to promote T_h9 cells at the cost of T_h2 cells, nor it is clear how STAT6 and GATA3 collaborate with PU.1 and/or IRF4 in T_h9 induction. In addition, the target molecules downstream of TGF-β signaling pathways that are critical to T_h9 development are incompletely defined. An intriguing point is that TGF-β and IL-4 only convert a small fraction of T_h2 cells to T_h9 cells (26); what renders some T_h2 cells responsive to the switch while other cells are resistant under the same T_h9-polarizing conditions warrants further clarification. Interestingly, a recent study from Chen Dong’s group uncovered the importance of cytokine-induced SH2 protein (CIS) in the control of Th2 and Th9 differentiation (44). CIS is induced by IL-4 and suppresses the activation of STAT3, STAT5 and STAT6 in T cells. They found that STAT5 and STAT6 promote IL-9 expression by directly binding to the IL-9 promoter, and therefore, CIS-deficient T cells exhibit enhanced differentiation into T_h9 cells. Consequently, CIS-deficient mice spontaneously develop airway inflammation in which Th9 cells are required (44).

In our own studies, we identified a new molecular pathway by which T_h9 cells develop, and this pathway is triggered by OX40-mediated costimulation (30). One striking feature is that when OX40 costimulation is delivered to CD4⁺ T cells, up to 80% of the CD4⁺ T cells can be converted to IL-9-producing T_h9 cells by TGF-β and IL-4. Unlike previously reported T_h9 cells, such T_h9 cells have no detectable levels of IL-10 and are highly pathogenic in a mouse model of allergic lung inflammation (30). Thus, we believe that T_h9 cells developed under OX40 costimulation are bona fide T_h9 cells. However, OX40 must act in concert with polarizing cytokines in T_h9 development. We showed that OX40 signaling blocks the induction of inducible T_reg cells and T_h17 cells and selectively diverts the cells to a T_h9 phenotype. However, without the polarizing cytokines, OX40 signaling instead supports T_h1 and T_h2 development (30).

Mechanistically, OX40 ligation activates, and most strikingly sustains, activities of the non-canonical NF-κB pathway (RelB–p52), which potently mediates T_h9 induction by TGF-β and IL-4. In fact, the promoter region of IL-9 has multiple NF-κB binding sites and RelB–p52 is directly involved in IL-9 transcription (30). These studies uncover additional complexities in IL-9 transcription and further suggest that, for T_h9 cells, additional signals besides those downstream of cytokine receptors are critically important. Our studies also raise other questions. For example, are other costimulatory molecules that also activate the non-canonical NF-κB pathway involved in T_h9 development? RelB–p52 is not acting alone, and PU.1 and IRF4 are not involved in OX40-mediated induction of T_h9 cells (30). Thus, what are the molecular partners downstream of TGF-β and IL-4 receptors that conspire with RelB–p52 in driving development of T_h9 cells? Are there any roles for the classical T_h2 transcription factors STAT6 and GATA3 in this process? Clearly, more studies are required to further clarify these questions.

Clinical relevance

What are T_h9 cells made for? There are several clinical settings in which T_h9 cells are implicated in the disease process. This suggests that intervention of T_h9 development may be therapeutically important. Studies from many laboratories including our own highlight the importance of T_h9 cells in allergic lung inflammation (30, 45, 46). However, the inflammatory response in the lung also involves other cell types besides T_h9 cells, and most prominently T_h2 cells. The interactions among different cell types in development and progression of the disease remain unclear, but T_h9 cells appear to play a particularly important role in airway epithelial alterations, which include goblet-cell hyperplasia, mucus production and infiltration of the airspace by mast cells and eosinophils. Indeed, blocking T_h9 cells markedly reduced the airway pathology while that in the lung parenchyma was not significantly affected (30), suggesting that T_h9 cells are a part, but may not the entirety, of allergic lung inflammation.

The role of T_h9 cells in other inflammatory conditions, especially chronic inflammation, remains to be determined. There are reports supporting a role for T_h9 cells in certain autoimmune diseases including EAE (40), suggesting that targeting T_h9 cells may provide an additional approach in treatment of such autoimmune conditions. There are other conditions where promotion of T_h9 cells might be therapeutically beneficial. It has been shown that IL-9 from mast cells promotes transplant tolerance to skin allografts by recruiting Foxp3⁺ T_reg cells to the grafts (15). Thus, neutralizing IL-9 resulted in failure of tolerance induction (15). By the same token, T_h9 cells might also be tolerogenic in transplant settings. However, considering the inflammatory nature of mast cells, T_h9 cells and cells recruited by IL-9, the exact role of T_h9 cells and T_h9 cell products in immunity and immune tolerance deserves careful clarification.

Another area that T_h9 cells recently attracted considerable attention is cancer therapy. Two laboratories independently reported that T_h9 cells exhibit remarkable therapeutic efficacy in cancer models (47, 48). In a highly aggressive B16 melanoma model, it has been shown that induction of T_h9 cells is associated with potent anti-cancer effects and favorable outcomes of cancer-bearing mice (47, 48). This is a significant area, considering the growing incidence of cancers and the limited choices in treatment of cancer patients.

Conclusions

T_h9 cells are a new and evolving subset of helper T cells. There are different molecular pathways identified thus far supporting the development of T_h9 cells, and integration of multiple signaling pathways downstream of cytokine receptors and costimulatory molecules is essential for specification and induction of T_h9 cells. The greatest effect on T_h9 induction is achieved by engaging OX40 under TGF-β- and IL-4-polarizing conditions. T_h9 cells and T_h9 products are highly pathogenic in allergic lung inflammation as well as in some autoimmune conditions, but they may be therapeutically desirable in other conditions such as cancer therapies. However, compared with other T helper subsets, T_h9 cells are less well studied. Many questions regarding T_h9 induction, the molecular machinery involved, their relationships with other helper T-cell subsets, especially T_h2 cells, and the exact role of T_h9 cells in immunity and immune pathology deserve further attention in future studies.

Funding

National Institutes of Health (R01 AI057409, R01 AI070315).