The differentiation of naive CD4+ helper T cells into the various lineages of effec-

tor populations drives immune responses to pathogens, or to self antigens, in the case of autoimmune disorders. Proinflammatory helper T cell effector lineages are defined by the secretion of characteristic profiles of cytokines: interferon-γ is secreted by T helper type 1 cells (TH1 cells), and interleukin 17 (IL-17) is secreted by TH-17 cells. Both of these populations of helper T cells, along with CD25+Foxp3+ regulatory T cells, control tis-sue destruction in multiple sclerosis, an auto-immune disorder in which the myelin sheaths of central nervous system (CNS) neurons are destroyed by inflammatory and degenera-tive processes. Considerable effort has been devoted to elucidating the mechanisms by which autoimmunity and helper T cell differ-entiation are regulated, and studies providing descriptions of several key cyto kines, signal-ing molecules and transcription factors have paved the way for the design of TH-17 cell–specific diagnostic and therapeutic agents. In this issue of Nature Immunology, Du and colleagues1 establish a role for microRNA (miRNA) particles in the regulation of TH-17 differentiation and autoimmune disease. Their work positively associates the amount of the miRNA species miR-326 with active disease relapses in patients with multiple sclerosis and uses overexpression and inter-

MicroRNA-managing the TH-17 inflammatory responseAaron J Martin, Liang Zhou & Stephen D Miller

The differentiation of interleukin 17–producing helper T cells is controlled by a complex network of cytokines, signaling pathways and transcription factors. Regulation by microRNA particles can now be added to this list.

ference tools to manipulate the amount of miR-326 in the experimental autoimmune encephalomyelitis (EAE) animal model of multiple sclerosis, in which TH-17 cells have a critical pathologic role1,2. They show that forced expression of miR-326 worsens EAE symptoms and enhances TH-17 differen-tiation, whereas neutralization of miR-326 improves clinical outcome and impairs TH-17 differentiation. Finally, they demonstrate that miR-326 binds to and prevents translation of Ets-1 mRNA, a known inhibitor of TH-17 dif-ferentiation (Fig. 1).

The miRNAs are short molecules of RNA (approximately 22 nucleotides in length) that are transcribed from noncoding regions of the genome and have considerable second-ary structure3. The nucleases Drosha, Pasha and Dicer are involved in processing the miRNA from a hairpin configuration into a short RNA duplex and finally into a single-stranded miRNA, which is then loaded into the RNA-induced silencing complex. This complex and its associated miRNA then bind complementary sequences in 3′ untranslated regions (called ‘seed regions’) of mRNA spe-cies and inhibit their translation by two dis-tinct mechanisms: degradation of the message by the RNA-induced silencing complex pro-tein argonaute; and prevention of ribosomal binding and translation initiation3. Notably, Du et al. show that miR-326 interferes with Ets-1 synthesis mainly by inhibiting transla-tion initiation, as miR-326 overexpression decreases Ets-1 protein but not the amount of Ets-1 mRNA1.

The differentiation of TH-17 cells from naive precursors is governed by a complex network of cytokines, signaling pathways and transcription factors. It is well accepted that signaling through the receptors for IL-6 and transforming growth factor-β (TGF-β) by STAT3 and presumably through Smad

proteins, respectively, and transcriptional regulation by RORγt are required for the establishment of this lineage4. IL-21 and IL-23 are also known as pro-TH-17 factors that function to stabilize the lineage against antagonism by opposing networks of cytok-ines and transcription factors. The list of anti-TH-17 factors includes but is not limited to IL-2 receptor (IL-2R)-STAT5 signals, IL-27, IL-25, interferon-γ, IL-4, Foxp3 and Ets-1. Du and colleagues demonstrate here that in addi-tion to cytokines and related receptor signals, miRNAs can promote TH-17 differentiation by inhibiting antagonists of the TH-17 lin-eage1. Although it has been previously shown that miRNAs can affect developmental out-comes in thymic T cell precursors, influence regulatory T cell development and affect the production of antibodies to thymus-depen-dent antigens, this is the first demonstration that a specific miRNA directly influences the TH-17 effector subset and contributes to autoimmune pathology. Work continuing to catalog miRNA species and associate them with various helper T cell lineages, immune functions and autoimmune pathology would be of great interest.

An intriguing observation made in this paper is that miR-326 interferes with the protein synthesis of one TH-17 antagonist (Ets-1) but not another (Foxp3; Fig. 1). It will therefore be a challenge to determine if miR-326 contributes to the commitment of naive precursors to the TH-17 lineage, as does IL-6–TGF-β, or if it functions to stabi-lize helper T cells that are already producing IL-17 in the manner of IL-23–IL-21. Signals from TGF-β, IL-2R and the T cell anti-gen receptor converge to upregulate Foxp3 expression and initiate the differentiation of regulatory T cells, and STAT5 and Foxp3 dis-courage TH-17 differentiation by interfering with RORγ t expression or function and thus

nature immunology volume 10 number 12 december 2009 1229

Aaron J. Martin and Stephen D. Miller are in

the Department of Microbiology-Immunology,

Feinberg School of Medicine, Northwestern

University, Chicago, Illinois, USA. Liang Zhou is

in the Department of Microbiology-Immunology

and Department of Pathology and The

Interdepartmental Immunobiology Center, Feinberg

School of Medicine, Northwestern University,

Chicago, Illinois, USA.

e-mail: s-d-miller@northwestern.edu

N e w S A N D v I e w S

©20

09 N

atu

re A

mer

ica,

Inc.

All

rig

hts

res

erve

d.

preventing Il17 transactivation4. Because of the involvement of Ets-1 in IL-2 synthesis and IL-2R signaling, we speculate that miR-326 functions to regulate IL-2R signals through Ets-1 while allowing TGF-β signals to con-tinue unperturbed. This may allow TGF-β signaling to contribute to the expression of both Foxp3 and IL-17, which is consistent with reports indicating that both Foxp3 and IL-17 are expressed at early times in TH-17 differentiation, when the naive precursors show considerable plasticity5,6. If this is true, then miR-326 serves to uncouple TGF-β sig-naling from IL-2R signals and promote TH-17 differentiation over regulatory T cell differ-entiation at an early developmental stage. Regardless of whether it is true, because of the presence of the TGF-β-induced Foxp3, IL-6 receptor signals transduced by STAT3 will still be needed to antagonize the differ-entiation of regulatory T cells and commit naive precursors to the TH-17 lineage. In fact, data presented by Du and colleagues support this idea, as interference with miR-326 with a ‘sponge’ vector decreases but does not com-pletely ablate IL-17 synthesis1. It is therefore probable that miR-326 contributes to TH-17 differentiation but is not required for lineage commitment. It has been shown that Ets-1 regulates IL-2 mRNA expression but not the amount of RORγt transcript. Upregulation of RORγt transcripts by forced expression of miR-326 suggests that in addition to Ets-1, miR-326 may also perturb a potential inhibi-tor of Rorc, the gene encoding RORγt. Further work describing the precise role of this miRNA relative to that of the other molecules known to be involved in TH-17 differentia-tion is therefore of considerable interest.

There are two additional exciting sugges-tions that result from the paper from Du et al.: the interpretation of data previously generated by knockout and overexpres-sion models and biomarker discovery. This miRNA species is encoded by an untranslated region of the gene encoding β-arrestin-1, and both β-arrestin-1 and miR-326 are positively associated with the development of CNS autoimmunity1,7. Although the mechanis-tic contributions of β-arrestin-1 have been clearly demonstrated, it is possible that β-arrestin-deficient mice also lack miR-326. Knocking out both a gene and an embedded miRNA-coding sequence that both mediate autoimmune disease could potentially affect the interpretation of disease outcome and lead to the improper assignment of an role to the gene being studied. Going forward, it will be of importance to carefully examine candidate genes for miRNA sequences before performing knockout studies.

The second exciting possibility this work opens is the authors’ proposed use of miR-326 as a biomarker for the development or relapse of multiple sclerosis. Such a diagnostic tool would be of tremendous clinical impor-tance in multiple sclerosis, as tissue samples of the target organ (the CNS or cerebrospinal fluid) are difficult or impossible to obtain. In contrast, miR-326 can be detected in easily accessible circulating T cells. Studies aiming to confirm miR-326 as a biomarker in mul-tiple sclerosis should be prioritized.

The finding that miRNAs can directly regulate TH-17 differentiation and contrib-ute to autoimmune pathology is a substantial advance; however, several outstanding ques-tions remain. Although the authors’ claim that miR-326 can serve as a possible diagnos-tic marker is well supported, the therapeutic potential of this target is less clear. It would be much more clinically relevant to see the out-come of treating animals with the miR-326 sponge vector after they present with clinical symptoms of EAE. As previous studies have shown that epitope spreading in the SJL/J mouse model of relapsing EAE involves the

activation of TH-17 cells specific for endog-enously released myelin epitopes mainly in the inflamed CNS8,9, determining the effects of treatment with the miR-326 sponge vector during disease remission on incidence of dis-ease relapses would be extremely interesting. Furthermore, if miRNAs are to be seriously considered as therapeutic targets in autoim-mune disorders, the development of an effi-cient nonviral delivery method is of paramount importance, particularly if activation of T cells to spread epitopes before disease relapse occurs mainly in the CNS. Several studies involving the delivery of small interfering RNA mol-ecules by biopolymer microparticles or nano-particles have been published10, and phase I clinical trial are active or in the recruitment stage (NCT00689065 and NCT00882180). If these obstacles are overcome, a third concern involves the proper timing of any therapeutic designed to target the TH-17 lineage. The TH1 lineage and chemokine receptor–mediated events can initiate EAE by facilitating the entry of TH-17 cells into the CNS11,12. Furthermore, the ratio of TH1 cells to TH-17 cells can affect the location of CNS lesions and the severity

1230 volume 10 number 12 december 2009 nature immunology

Cytoplasm

Plasma membrane

miR-326

RISCAAAAA AAAAA

AAAAA

STAT3 Foxp3

Foxp3 mRNA Ets-1 mRNA

Ets-1

RORγt

Rorc

RORγtinhibitor mRNA

InhibitorIl17

IL-2 synthesisand signaling

IL-6R+IL-6

No significantinteractions

Figure 1 The microRNA miR-326 encourages TH-17 differentiation. First, miR-326 is transcribed from noncoding regions of the gene encoding β-arrestin. In the cytoplasm, miR-326 binds to a complementary untranslated region in the 3′ region of the Ets-1 message and prevents translation of the message. This prevents Ets-1 protein from accumulating and interfering with TH-17 differentiation by way of IL-2 synthesis and signaling pathways. Interestingly, miR-326 has not been found to target Foxp3 mRNA to perturb the Foxp3 protein expression required for the inhibition of RORγt activity. Signaling events downstream of the IL-6 receptor (IL-6R) are still needed to prevent TH-17 lineage antagonism by Foxp3. RORγt expression is upregulated by miR-326, which suggests that miR-326 may somehow enhance the expression of the TH-17 lineage transcription factor through interference with an unknown inhibitor of Rorc. RISC, RNA-induced silencing complex.

NE wS AN d v IE wS

©20

09 N

atu

re A

mer

ica,

Inc.

All

rig

hts

res

erve

d.

of disease13. It is therefore possible that thera-peutics that target only the TH-17 lineage may be effective only at certain times or in certain disease conditions. The most successful treat-ment strategies will probably target both the TH1 and TH-17 cell effector subsets in an autoantigen-specific manner.

Collectively, the data in Du’s publication establish a role for miR-326 as a regulator of TH-17 differentiation and the development of autoimmune disease while identifying a prom-

ising target for use as a diagnostic or therapeu-tic biomarker. Furthermore, this paper lays a fertile groundwork for future research into the precise roles of miRNAs in helper T cell dif-ferentiation, cytokine receptor signaling and transcription factor function.

1. du, C. et al. Nat. Immunol. 10, 1252–1259 (2009). 2. Langrish, C.L. et al. J. Exp. Med. 201, 233–240

(2005).3. Shivdasani, R.A. Blood 108, 3646–3653 (2006).4. dong, C. Nat. Rev. Immunol. 8, 337–348 (2008).

5. Yang, X.O. et al. Immunity 29, 44–56 (2008).6. Zhou, L. et al. Nature 453, 236–240 (2008).7. Shi, Y. et al. Nat. Immunol. 8, 817–824 (2007).8. McMahon, E.J., Bailey, S.L., Castenada, C.v.,

waldner, H. & Miller, S.d. Nat. Med. 11, 335–339 (2005).

9. Bailey, S.L., Schreiner, B., McMahon, E.J. & Miller, S.d. Nat. Immunol. 8, 172–180 (2007).

10. Aouadi, M. et al. Nature 458, 1180–1184 (2009).11. O’Connor, R.A. et al. J. Immunol. 181, 3750–3754

(2008).12. Reboldi, A et al. Nat. Immunol. 10, 514–523

(2009).13. Stromnes, I.M., Cerretti, L.M., Liggitt, d., Harris, R.A.

& Goverman, J.M. Nat. Med. 14, 337–342 (2008).

nature immunology volume 10 number 12 december 2009 1231

T cells need Nod too?Shahram Salek-Ardakani & Michael Croft

Nucleotide-binding oligomerization domain 2 (Nod2) is required for sensing of intracellular bacteria and subsequent inflammatory responses. Unexpectedly, new evidence suggests that Nod2 influences T helper cell signaling, proliferation and differentiation and effector responses against Toxoplasma gondii.

Shahram Salek-Ardakani and Michael Croft are

in the Division of Molecular Immunology, La Jolla

Institute for Allergy and Immunology, San Diego,

California, USA.

e-mail: ssalek@liai.org or mick@liai.org

The early immune response to most patho-gens is triggered by pattern-recognition

receptors (PRRs), including Toll-like receptors (TLR) and nucleotide binding and oligomer-ization domain (Nod)-like receptors (NLRs). TLRs can stimulate antigen-presenting cells, thus indirectly allowing the effective induc-tion of T cell immunity. Many TLRs induce activation of the canonical (IKKα-IKKβ-IKKγ–dependent) NF-κB pathway through the adaptor MyD88 and the serine/threo-nine kinase RIP2 (Fig. 1). Most early studies assumed that any requirement for TLRs and MyD88 in immunity reflected their roles in innate cells, but this view has been modified by recent data demonstrating that MyD88 can directly control T cell function, either through TLRs now known to be expressed on T cells or through IL-1R family members that can also engage MyD88 (refs. 1,2). In this issue of Nature Immunology, Shaw et al.3 now suggest that this type of innate cell–T cell dual activity is also shown by Nod2, an intracellular PRR. Nod2 can stimulate epithelial cells, macrophages and dendritic cells by sensing muramyl dipeptide (MDP), a component of bacterial peptidogly-can, and can also trigger activation of RIP2 and the canonical NF-κB pathway in these cells4. Shaw et al.3 challenge the current dogma that Nod2 is dedicated for innate immune

recognition of bacterial motifs by showing that Nod2 can provide a T cell–intrinsic signal that does not rely on MDP recognition.

The initial focus that led to this conclusion was the finding that Nod2-deficient animals succumb to T. gondii infection. Through a series of experiments with this pathogen, Shaw et al.3 revealed a specific defect in the ability of T cells from Nod2-deficient mice to produce interferon (IFN)-γ, raising the possibility that Nod2 might exert T cell–intrinsic functions. Consistent with this idea, upon T. gondii chal-lenge, wild-type CD4+ T cells differentiated into T helper type-1 (TH1) effector cells in a Nod2–/– host, whereas Nod2–/– CD4+ T cells failed to differentiate into TH1 cells in a Nod2-sufficient environment.

Does Nod2 then have a more general role in T cell function? Shaw et al.3 found that Nod2 is required for lymphopenia-induced homeo-static T cell proliferation, a response driven by T cell antigen receptor (TCR)–peptide-MHC interactions and cytokines that signal through the common γ-chain. Nod2–/– donor CD4+ T cell populations, if transferred alone, expand normally in a lymphopenic host. However, when transferred with wild-type T cells, Nod2–/– CD4+ T cells show impaired popula-tion expansion. The authors also used a mouse model of colitis whereby the transfer of naive CD4+ T cells into Rag1–/– mice leads to gut inflammation. Recipients reconstituted with Nod2–/– T cells remained healthy and did not succumb to wasting disease.

These results support the idea that, in some situations, Nod2 expressed in T cells can affect T cell activity independently of sensing a

bacterial product. In support of this argument, Shaw et al.3 also found that in the apparent absence of any Nod2 ligand, Nod2-deficient CD4+ TCR-transgenic T cells exposed to anti-gen in vitro produced less interleukin (IL)-2 and showed impaired differentiation into TH1 or TH2 cells. Addition of exogenous IL-2 restored the defective TH cell differentiation, leading the authors to speculate that the pri-mary requirement for Nod2 is to regulate IL-2 secretion. In light of the T. gondii studies, this might be a reasonable conclusion as Il2–/– mice succumb to this pathogen in a time frame simi-lar to that reported for Nod2–/– mice5. However, it is not clear that defective production of IL-2 by Nod2–/– T cells can explain the defects in homeostatic proliferation, as competition for cytokines and a requirement for Nod2 in com-mon γ-chain cytokine receptor signaling is just as feasible. In the colitis experiments, Shaw et al.3 found reduced IFN-γ production by Nod2–/– CD4+ T cells, but this might also be unrelated to a lack of IL-2 production. IL-2 produced by CD4+ T cells can induce peripheral con-version of CD25– T cells into CD25+Foxp3+ inducible regulatory (iTreg) cells, leading one to predict that if the primary defect of Nod2–/– T cells in this situation was in producing IL-2, a more severe disease would ensue as a result of impaired iTreg cell generation.

The present model of Nod-mediated sig-naling in non-T cells holds that after detec-tion of MDP, Nod2 recruits RIP2, leading to a common pathway shared by TLRs, the IL-1R family and members of the tumor necrosis factor receptor (TNFR) superfam-ily, wherein Tab-TAK1-mediated activation

NE wS AN d v IE wS

©20

09 N

atu

re A

mer

ica,

Inc.

All

rig

hts

res

erve

d.