Tet2: Breaking Down Barriers to T Cell Cytokine Expression
Chao Zhong,Jinfang Zhu
DOI: https://doi.org/10.1016/j.immuni.2015.04.003
IF: 32.4
2015-01-01
Immunity
Abstract:It has been unclear whether alteration in DNA methylation at cytokine genes during T helper (Th) cell differentiation is a cause or consequence of gene expression. In this issue of Immunity, Ichiyama et al., 2015Ichiyama K. Chen T. Wang X. Yan X. Kim B.S. Tanaka S. Ndiaye-Lobry D. Deng Y. Zou Y. Zheng P. et al.Immunity. 2015; 42 (this issue): 613-626Abstract Full Text Full Text PDF PubMed Scopus (222) Google Scholar show that oxidation of 5-methylcytosine by the methylcytosine dioxygenase Tet2 regulates cytokine production in Th cells. It has been unclear whether alteration in DNA methylation at cytokine genes during T helper (Th) cell differentiation is a cause or consequence of gene expression. In this issue of Immunity, Ichiyama et al., 2015Ichiyama K. Chen T. Wang X. Yan X. Kim B.S. Tanaka S. Ndiaye-Lobry D. Deng Y. Zou Y. Zheng P. et al.Immunity. 2015; 42 (this issue): 613-626Abstract Full Text Full Text PDF PubMed Scopus (222) Google Scholar show that oxidation of 5-methylcytosine by the methylcytosine dioxygenase Tet2 regulates cytokine production in Th cells. Cell-fate determination, including T helper (Th) cell differentiation from naive CD4+ T cells, is often accompanied by the induction of lineage-specifying transcription factors and epigenetic changes at the lineage-specific gene loci (Kanno et al., 2012Kanno Y. Vahedi G. Hirahara K. Singleton K. O’Shea J.J. Annu. Rev. Immunol. 2012; 30: 707-731Crossref PubMed Scopus (262) Google Scholar, Zhu et al., 2010Zhu J. Yamane H. Paul W.E. Annu. Rev. Immunol. 2010; 28: 445-489Crossref PubMed Scopus (2272) Google Scholar). T-bet, GATA-3, RORγt, and Foxp3 are the key transcription factors in determining the fate of Th1, Th2, Th17, and inducible regulatory T (iTreg) cells, respectively. Besides a direct role in transcriptional activation and/or repression, these key transcription factors might also induce specific epigenetic changes at cytokine gene loci, which include modifications to the DNA and associated histones (Wilson et al., 2009Wilson C.B. Rowell E. Sekimata M. Nature Reviews. 2009; 9: 91-105Google Scholar). Genome-wide mapping of the histone modifications in CD4+ T lineages indicates that permissive (H3K4me3) and suppressive (H3K27me3) modifications at the signature cytokine loci are associated with the activation or silencing of these genes, respectively, in specific lineages (Wei et al., 2009Wei G. Wei L. Zhu J. Zang C. Hu-Li J. Yao Z. Cui K. Kanno Y. Roh T.Y. Watford W.T. et al.Immunity. 2009; 30: 155-167Abstract Full Text Full Text PDF PubMed Scopus (893) Google Scholar). In addition, changes in DNA methylation at cytokine gene loci are observed when cells gain the capacity to produce these cytokines during Th cell differentiation (Ansel et al., 2006Ansel K.M. Djuretic I. Tanasa B. Rao A. Annu. Rev. Immunol. 2006; 24: 607-656Crossref PubMed Scopus (532) Google Scholar). However, to date there is little direct evidence to suggest that DNA demethylation is necessary for induction of cytokine gene expression under physiologic conditions. DNA methylation predominately occurs at the 5-position of cytosine (5-methylcytosine, 5mC) within CpG dinucleotides. 5mC pattern is inheritable during DNA duplication and actively maintained by DNA methyltransferases (Dnmts) (Koh and Rao, 2013Koh K.P. Rao A. Curr. Opin. Cell Biol. 2013; 25: 152-161Crossref PubMed Scopus (68) Google Scholar). Hypermethylation in promoter and/or enhancer regions is associated with heterochromatin and/or silenced genes. Oxidation of 5mC by ten-eleven translocation (Tet) family proteins results in production of 5-hydroxymethylcytosine (5hmC), 5-formylcytosine (5fC), and then 5-carboxylcytosine (5caC). The methyl group is finally removed by thymine DNA glycosylase (TDG) through excision repair. 5hmC, the first intermediate product during cytosine demethylation, is relatively abundant in the genome and much more stable than 5fC or 5caC, indicating that 5hmC might serve as an important epigenetic mark. The function of 5hmC, particularly in T cells, is poorly understood because the conventional bisulfite method to determine DNA methylation status used in early studies does not distinguish 5mC from 5hmC. In this issue of Immunity, Ichiyama et al. use DNA immunoprecipitation (DIP) with specific antibodies to 5mC and 5hmC followed by high-throughput sequencing (DIP-Seq) to map these two modifications genome-wide in distinct T helper cells that were cultured in vitro (Ichiyama et al., 2015Ichiyama K. Chen T. Wang X. Yan X. Kim B.S. Tanaka S. Ndiaye-Lobry D. Deng Y. Zou Y. Zheng P. et al.Immunity. 2015; 42 (this issue): 613-626Abstract Full Text Full Text PDF PubMed Scopus (222) Google Scholar). Dynamic changes of these two modifications are noted in many lineage-specific gene loci. The global distribution patterns of either modification are similar among various Th subsets, with a slight enrichment of both modifications at gene bodies. Interestingly, 5hmC is more abundant in naive CD4+ T cells than in differentiated T helper cells, which is consistent with another genome-wide profiling of 5hmC distribution in T cells using a different method (Tsagaratou et al., 2014Tsagaratou A. Äijö T. Lio C.W. Yue X. Huang Y. Jacobsen S.E. Lähdesmäki H. Rao A. Proc. Natl. Acad. Sci. USA. 2014; 111: E3306-E3315Crossref PubMed Scopus (116) Google Scholar). Binary marks of both 5mC and 5hmC are present at ∼20% of the 5hmC-occupied regions indicating a potential active DNA demethylation process within these regions during T cell differentiation. Binary marks might exist at a higher frequency if the 5mC DIP-seq were to work more efficiently; many known 5mC marks in T cells from the literature including those located in an intron of Foxp3 gene are not identified by 5mC DIP-Seq and there is inconsistency between 5mC DIP-Seq and 5mC DIP-PCR at certain regions suggesting that 5mC DIP-Seq in this study might not be sensitive enough to detect all the 5mC marks. Nevertheless, the 5hmC marks are more likely to co-localize with H3K4me3 modification than 5mC marks. Therefore, 5hmC marks are enriched in active and/or poised genes. Indeed, many lineage-specific 5hmC marks are identified within the evolutionarily conserved non-coding sequence (CNS) regions of the lineage-specific genes, particularly effector cytokine genes including Ifng and Il17a. By contrast, fewer lineage-specific 5hmC marks are present within the gene loci of key transcription factors, such as Gata3 and Rorc. While 5mC marks are present at the CNS (−6 kb) of the Ifng gene in naive CD4+ T cells and Th17 cells judged by DIP-PCR results, 5hmC marks at this region are only detected in Th1 cells (Figure 1). 5hmC, but not 5mC, is preferential found in the regions with p300 binding, which defines active enhancers. In addition, the binding sites for key transcription factors such as T-bet, RORγt, and STAT proteins substantially overlap with 5hmC, but not 5mC in Th1 and Th17 cells, respectively. Therefore, 5hmC modification occurs at the regulatory DNA elements of lineage-specific genes, many of which serve as active enhancers in distinct Th lineages. Conversion from 5mC to 5hmC is catalyzed by Tet family proteins, Tet1, Tet2, and Tet3 (Koh and Rao, 2013Koh K.P. Rao A. Curr. Opin. Cell Biol. 2013; 25: 152-161Crossref PubMed Scopus (68) Google Scholar). Ichiyama et al. show that although all three Tet proteins are expressed in naive CD4+ T cells, their expression is downregulated during T cell activation, leaving only Tet2 expressed in substantial amounts in various Th lineages. Therefore, Tet2 seems to be a reasonable candidate that is responsible for 5hmC modification in distinct Th lineages. Indeed, Tet2 binding is detected at the CNS (−6 kb) of Ifng and the CNS2 of Il17 locus, where 5hmC marks are present in Th1 and Th17 cells, respectively. The recruitment of Tet2 to these regions and thus 5hmC modification depend on the expression of key transcription factors T-bet or RORγt, respectively. Conditional deletion of the Tet2 gene in T cells with Cd2-Cre affects the generation of cytokine-producing cells. Tet2-deficient cells differentiated into Th1 and Th17 lineages have a substantial reduction of 5hmC content at a genome scale. However, only very few genes, such as Ifng, Il17a, and Il10, are significantly affected. Tet2 deficiency not only results in a reduced 5hmC modification in the DNA regulatory elements in these cytokine genes, it also affects histone modification such as H3K4me3, p300 binding, and key transcription factor binding to these elements, indicating that Tet2 and key transcription factors such as T-bet might regulate each other’s function in inducing cytokine production. Ichiyama et al. further show that Tet2 regulates autoimmune responses in vivo. In an experimental autoimmune encephalomyelitis (EAE) model, Tet2 expression in CD4+ T cells restrains the disease severity possibly through regulating IL-10 expression. However, when IL-10 signaling is blocked, Tet2-deficient mice are more resistant to EAE induction, presumably due to impaired induction of IL-17-IFN-γ-dual-producing T cells. Therefore, Tet2 is important for regulating IFN-γ, IL-17a, and IL-10 production both in vitro and in vivo. This study by Ichiyama et al. sheds light on the importance of Tet2-mediated 5mC oxidation at the regulatory elements of effector cytokine genes and the critical function of such epigenetic change in regulating cytokine expression. There are several interesting issues need to be resolved in the future. Is 5hmC, as an intermediate product, simply a readout for active demethylation process or 5hmC itself is important for gene regulation? Many 5hmC-binding proteins have been identified (Koh and Rao, 2013Koh K.P. Rao A. Curr. Opin. Cell Biol. 2013; 25: 152-161Crossref PubMed Scopus (68) Google Scholar). So far, it is unknown which protein(s) can bind to 5hmC in T cells. MeCP2, one of the 5mC binding proteins, is the major molecule binds to 5hmC in neuronal cells; a disease-causing mutant of MeCP2 lost its ability to bind to 5hmC, but is still able to interact with 5mC (Mellén et al., 2012Mellén M. Ayata P. Dewell S. Kriaucionis S. Heintz N. Cell. 2012; 151: 1417-1430Abstract Full Text Full Text PDF PubMed Scopus (705) Google Scholar). T cell-specific deletion of MeCP2 results in impaired cytokine production by Th1 and Th17 cells, although the mechanism for such defect is interpreted differently in the report (Jiang et al., 2014Jiang S. Li C. McRae G. Lykken E. Sevilla J. Liu S.Q. Wan Y. Li Q.J. Sci. Signal. 2014; 7: ra25Crossref PubMed Scopus (49) Google Scholar). Whether MeCP2 has a direct role on effector cytokine production through sensing a change from 5mC to 5hmC at the cytokine loci requires further investigation. In the future, it is also important to determine the relationship among 5hmC marks, histone modifications and transcription factor binding at the cytokine gene loci. Furthermore, what is the function of 5hmC marks at the hundreds of lineage-specific genes whose expression is not altered but 5hmC marks are reduced in the absence of Tet2? Are these 5hmC marks important for terminal differentiation of distinct lineages and/or memory generation? What are the 5hmC patterns in T effector cells that are generated in vivo? Tet2 deficiency does not completely eliminate 5hmC marks at the cytokine gene loci in Th1 and Th17 cells. Also, Tet2 deletion does not affect IL-17f expression in Th17 cells, but the mechanism for differential regulation of IL-17a and IL-17f by Tet2 is unknown. In addition, Tet2 is dispensable for lineage-specific 5hmC marks at the key gene loci in Th2 and iTreg cells. Consistent with this, Tet2 does not bind to these 5hmC regions in these cells and the expression of the signature genes including IL-4 and Foxp3 is not affected by Tet2 deficiency. Our RNA-Seq data and previous reports suggest that both Tet2 and Tet3 are expressed in T cells (Tsagaratou et al., 2014Tsagaratou A. Äijö T. Lio C.W. Yue X. Huang Y. Jacobsen S.E. Lähdesmäki H. Rao A. Proc. Natl. Acad. Sci. USA. 2014; 111: E3306-E3315Crossref PubMed Scopus (116) Google Scholar). Therefore, Tet3 might play an important role in 5hmC production in Th2 and iTreg cells and also have some contribution to 5hmC modification in Th1 and Th17 cells. The importance of 5mC oxidation in promoting cytokine production might be underestimated by Ichiyama et al. in their study. Experiments using Tet2 and Tet3 doubly deficient cells and/or mice might give the answers to these unsolved important questions. C.Z. and J.Z. are supported by the Division of Intramural Research (DIR), National Institute of Allergy and Infectious Diseases (NIAID), National Institutes of Health (NIH), USA. The Methylcytosine Dioxygenase Tet2 Promotes DNA Demethylation and Activation of Cytokine Gene Expression in T CellsIchiyama et al.ImmunityApril 7, 2015In BriefTet proteins are methylcytosine dioxygenases that convert methyalated DNA (5mC) to hydroxymethylated DNA (5hmC). Dong and colleagues generate a genome-wide 5hmC map in T helper (Th) cells and demonstrate that Tet2-mediated DNA demethylation plays a crucial role in control of signature cytokine expression in Th cells. Full-Text PDF Open Archive