- Email: [email protected]
T cell differentiation: a mechanistic view
654
T cell differentiation: a mechanistic view Orly Avni* and Anjana Rao† It has recently become clear that cytokine expression by T cells involves epigenetic changes in chromatin structure, locus accessibility and DNA methylation that occur during differentiation of naive T cells into mature effector T cells. These changes require the coordinate actions of antigen- and cytokine-induced transcription factors, chromatin remodeling complexes, histone-modifying enzymes and subset-specific transcription factors. Addresses Department of Pathology, Harvard Medical School and The Center for Blood Research, 200 Longwood Avenue, Boston, MA 02115, USA *e-mail: [email protected] † e-mail: [email protected] Current Opinion in Immunology 2000, 12:654–659 0952-7915/00/$ — see front matter © 2000 Elsevier Science Ltd. All rights reserved. Abbreviations BAF BRG1-associated factors CNS conserved noncoding sequence DH site DNase I hypersensitive site HDAC histone deacetylase kb kilobases
Introduction When naive T helper cells encounter antigen for the first time in the periphery, they begin a process of peripheral differentiation that includes commitment to a specific pattern of cytokine production. Th1 cell populations produce INF-γ but not IL-4 whereas Th2 cell populations transcribe the IL-4, IL-5 and IL-13 genes but not the INF-γ gene (individual T cells may show varied cytokine expression). These patterns of gene expression are strongly influenced by the local cytokine milieu: IL-12 and IL-4 strongly potentiate Th1 and Th2 differentiation, respectively, via the transcription factors STAT4 and STAT6. Clinically, INF-γ production is associated with inflammatory responses to bacterial antigens and with autoimmune disease whereas IL-4, IL-5 and IL-13 mediate allergic reactions and immune responses against helminthic parasites. These basic aspects of T cell differentiation have been thoroughly covered in many excellent reviews [1–7]. When first stimulated with antigen, naive T cells produce IL-2 but are incapable of producing large quantities of certain other cytokines, including IL-4 and IFN-γ. Stimulation initiates a differentiative phase in which those genetic loci that are destined to be transcriptionally active in the differentiated cells become ‘competent’ for future transcription, a process associated with changes in chromatin structure of cytokine genes ([8–10]; reviewed in [4,11]). Three stages of cytokine expression may be distinguished: an initiation phase that is highly dependent on antigen, cytokines and cytokine-induced STAT factors [5];
a commitment phase — mediated by subset-specific transcription factors such as GATA3 and cMaf (Maf) in Th2 cells and by T-bet and ERM in Th1 cells [6,7] — in which the differentiated phenotype is stabilised and maintained in the absence of further stimulation; and a phase of acute gene transcription, elicited by secondary contact of differentiated T cells with antigen, that requires the antigen-induced transcription factor NFAT [12,13••]. Here we review recent data on these phases of T cell differentiation, focusing on the IL-4/IL-5/IL-13 genes, which are particularly important for allergy.
Structural changes in chromatin during lymphocyte differentiation and activation Early chromatin decondensation
As pointed out by Zhao et al. [14•], antigen-induced activation leads to rapid and visible decondensation of the highly heterochromatic nucleus of naive T cells. These workers have suggested that chromatin decondensation might be mediated by the BAF (BRG1-associated factors) family of chromatin-remodeling complexes (for a recent review of chromatin-remodelling complexes, see [15,16]). TCR stimulation caused tight, rapid (i.e. after 10–15 minutes) and selective association of BAF complexes with the nuclear matrix of primary murine T cells, an effect apparently mediated by phosphatidylinositol (4,5) bisphosphate (PIP2) via actin and an actin-like protein (BAF53) that is present in the complex. The effect may not be limited to naive T cells but may constitute a general T cell activation response. BAF complexes may exert global effects in the nucleus or may be selective for certain target genes, for instance when gene expression is polarised by combined antigen/cytokine stimulation. Changes in DNase I hypersensitivity
In the IL-4/IL-13 locus, naive T cells exhibit only two DNase I hypersensitive sites (DH sites) in an interval of ~43 kilobases (kb) spanning the IL-4 and IL-13 genes; within 48 hours of primary stimulation in the presence of IL-4, 10 additional clusters of Th2-specific DH sites are observed ([8,9]; S Agarwal, D Lee, A Rao, unpublished data; Figure 1a). At this time, IL-4 mRNA is barely detectable, indicating that chromatin remodelling is an early differentiative event. Two intergenic DH sites and the 3′ DH site V are contained within conserved noncoding sequences (CNS-1 and -2) that were identified by sequence comparison of syntenic regions of mouse chromosome 11 and human 5q31 regions [17••] (Figure 1a). All 10 DH sites are stable markers of differentiated Th2 cells. Stimulation of differentiated cells leads to appearance of a novel hypersensitive site (VA) 3′ of the IL-4 gene (Figure 1a); this site corresponds to an inducible IL-4 enhancer, which binds the antigen-induced transcription factor NFAT and the Th2-specific nuclear factor GATA3 [13••].
T cell differentiation: a mechanistic view Avni and Rao
655
Figure 1 (a)
(b)
IL-13 /IL-4
Naive cells
IL-13
HSS3
IV
IL-4 12
Exon 1 234
I II III
Exon 1 234
IV
Activated cells
Initiation
23 Antigen, IL-12
Commitment
T-bet? Others
CNS-2 V
KIF3
Exon
1 234
5 kb
3 21
I Exon
Others
0
NFAT
IV
V
KIF3
3 4 Maf
4 Initiation Commitment
III 4 Acute transcription
NFAT Others
I II III 12
II
23 Antigen
1
Acute transcription VA (enhancer)
I II III
1
Differentiated cells
34
12 Antigen NFAT
II IFN- γ Exon
GATA3 Others CNS-1 321 0
Naive cells
34
Antigen, IL-4/STAT6
Differentiated cells I II III
KIF3
IFN-γ
NFAT
Activated cells
I 1
Exon GATA3
1 kb
III 23
4
NFAT Current Opinion in Immunology
Diagram showing changes in DNase I hypersensitivity over (a) ~43 kb of the IL-4/IL-13 locus and (b) ~11 kb of the IFN-γ locus during differentiation of murine CD4+ T cells and following activation of differentiated CD4+ T cells. Together with antigen, IL-4/STAT6 and IL-12/STAT4 play important roles in the initiation of Th2 and Th1 differentiation. Transcription factors that are preferentially expressed in the Th2 and Th1 lineages — GATA3, T-bet and others — are especially
important in the commitment phase; NFAT is especially important in acute gene transcription. At the bottom of the figure, NFAT, GATA3 and Maf are shown binding to the IL-4 promoter; for a discussion of other transcription factors implicated in IL-4 and IFN-γ gene transcription, see [7]. Numbered arrows indicate DH sites. In the middle and lower panels of (a), 3,2 and 1 above the lines refer to DH sites HSS3, HSS2 and HSS1, respectively. KIF3 is a brain-specific protein whose gene is near the IL-13/IL-4 locus.
In the IFN-γ locus, naive murine CD4+ T cells show only one weak DH site over an ~11 kb region spanning the IFN-γ gene; their differentiation into Th1 cells is accompanied by the appearance of at least two additional constitutive DH sites (Figure 1b). Stimulation of differentiated T cells attenuates the intensity of one of the DH sites in the murine IFN-γ gene [13••]. Inducible DH sites have been observed 5′ of the human IFN-γ promoter [18].
Maf-deficient T cells do not show detectable changes in the pattern of either the constitutive or inducible DH sites in the IL-4 gene [13••] but in vivo footprinting indicates that the Maf-binding site in the IL-4 promoter is constitutively occupied in resting Th2 cells [24].
The proteins that form the constitutive DH sites in the IL-4/IL-13 and IFN-γ loci have not yet been identified. DH sites mark nucleosome-free regions or other perturbations of canonical nucleosome structure, generally as a result of protein binding to adjacent DNA elements [19,20]. Appearance of DH sites in the IL-4 gene is far less efficient in T cells lacking STAT6 [8]. STAT6 synergises with antigen to induce the Th2-specific nuclear factors GATA3 and Maf [21••]. Ectopic expression of GATA3 suffices for appearance of all the Th2-specific DH sites in the IL-4 gene [22••,23], consistent with the possibility that GATA3 binds directly to at least some of these DH sites. GATA3 also binds directly (with NFAT) to the region of the inducible DH site VA, which marks an IL-4 enhancer [13••].
Histone modifications
Typically, acquisition of transcriptional competence is marked not only by an altered pattern of DH sites but also by overall increases in histone acetylation and general sensitivity to DNase I [25]. Other histone modifications, including phosphorylation and methylation, have been described [26]. Together, these modifications may induce an ‘open’ DNA conformation that permits greater access of transcription factors to their nucleosomal binding sites; moreover, the modified histone tails may serve as docking sites for nuclear proteins that positively influence transcription [26]. Like other transcription factors, NFAT and STAT6 bind the histone acetyltransferases p300 and CBP [27–29], and may recruit these coactivators to the IL-4/IL-5/IL-13 locus during Th2 differentiation. Conversely BCL-6, a zinc finger transcriptional repressor — which binds sites recognized by STAT factors and interacts with
656
Atopic allergy and other hypersensitivities
a histone deacetylase (HDAC)-containing corepressor complex [30] — represses Th2 differentiation as judged by the pronounced Th2 bias of BCL-6-deficient mice [31]. A similar Th2 bias is observed in T cells lacking both STAT6 and BCL-6 [32], suggesting either that BCL-6 inhibits a pathway of Th2 differentiation that does not involve STAT6 or that, in the absence of BCL-6, STAT6 is not as crucial for promoting chromatin accessibility. DNA demethylation
It has been known for many years that silent genes are present in regions of hypermethylated DNA whereas active genes are marked by hypomethylation. DNA methylation recruits methyl-CpG-binding proteins such as MeCP2, which in turn recruits corepressors such as the SIN3–HDAC complex [33]. Consistent with this paradigm, Th1 and Th2 differentiation are associated with selective demethylation of the IFN-γ and IL-4/IL-5/IL-13 genes, respectively [10,34]; however demethylation may be more relevant to commitment than to the initiation phase of T cell differentiation ([34]; S Agarwal, D Lee, A Rao, unpublished data). The bulk of DNA methylation occurs symmetrically on CpG dinucleotides. Newly replicated DNA strands contain cytosine in place of methylcytosine; following replication, DNA methyltransferases recognise the hemi-methylated CpG and recreate the fully methylated state. As a result, the methylation status of DNA is stably maintained even in dividing cells. The mechanism by which cytokine genes become demethylated during T cell differentiation is not understood: demethylation could occur through action of a demethylase or through local inhibition of the methyltransferase [35]. Probabilistic aspects
Cytokine expression during T cell differentiation is variegated, showing evidence of probabilistic, ‘on/off’ regulation in individual cells (reviewed in [3,4]). Even after optimal stimulation with antigen and cytokine, the fraction of naive T cells that actually commits to cytokine expression is surprisingly low. This behaviour is also apparent as a propensity to monoallelic expression of cytokine genes ([36,37] and references therein). The situation is reminiscent of position effect variegation, a phenomenon in which gene expression is determined in a stochastic fashion by chromatin accessibility [38]. Indeed, stimulation of naive T cells in the presence of butyrate (a general inhibitor of histone deacetylases) or azacytidine (an inhibitor of cytosine methylation) markedly increases cytokine expression; together the drugs strongly synergise to enhance IL-4 expression even in STAT6-deficient T cells [10]. The conserved CNS-1 region, containing the Th2-specific DH sites HSS1 and HSS2, also influences the probability rather than the level of Th2 cytokine expression [17••]. Transgenic mice carrying yeast artificial chromosomes containing the human IL-4/IL-13/IL-5 locus showed appropriate expression of the human cytokines in Th1,
Th2 and NK T cells; however, when CNS-1 was deleted from the transgene by Cre-mediated recombination, there was a 50–75% decrease in the number of Th2 cells expressing human IL-4 and IL-13, without a significant effect on the level of cytokine expression per cell. CNS-1 also affected expression of the human IL-5 gene, located 120 kb away; but it did not affect expression of the nearby RAD50 and KIF3 genes (these show ubiquitous and brain-specific expression, respectively). Thus CNS-1 acts selectively over long distances to increase the probability of IL-4/IL-5/IL-13 gene expression in a chromatin context, in a manner reminiscent of locus control regions (LCRs) and certain enhancers that can induce uniform expression and suppress silencing of transgenes that have been integrated into repressive locations such as centromeres [25,38,39].
Sequential actions of transcription factors during lymphocyte differentiation and activation Initiation Cell cycle dependence
In response to a single antigenic stimulation of naive T cells, cytokine expression showed a striking hierarchy of dependence on cell division: IL-2 was expressed prior to cell division, IFN-γ production required S phase entry, and IL-4 expression required a threshold number of three to four cell divisions [10,40]. In contrast, if naive T cells were first stimulated with antigen and then their ability to produce IL-4 was assessed in response to a second, short stimulation with PMA (phorbol myristate acetate) and ionomycin, cells that had not divided at all were capable of making IL-4 [41]; however, optimal IL-4 production still required multiple cell divisions and the use of cell-cycle inhibitors indicated that passage through S phase was necessary. Nucleosome disassembly occurring during the S phase might be needed for binding of nuclear factors to cytokine genes (or genes encoding other inducible proteins, such as T-bet or GATA3, that are needed for cytokine expression). DNA replication is also required for passive demethylation of DNA, necessary to displace the repressive chromatin structure induced by MeCP2–SIN3–HDAC complexes [35]. The role of antigen- and cytokine-induced transcription factors
TCR stimulation is essential to initiate Th1/Th2 differentiation. Riviere et al. [37] generated mice in which one IL-4 allele was replaced with DNA encoding a cell surface marker, human CD2. This allowed them to monitor independently the expression of each allele, by combining cell surface staining for the ‘knocked-in’ human CD2 marker and intracellular staining for endogenous IL-4. As the strength of initial antigenic stimulation was increased, a greater fraction of T cells expressed either or both alleles [37], implying that antigen-induced processes or transcription factors are limiting at early stages of differentiation. IL-4/STAT6 and IL-12/STAT4 increase the probability of T helper cell differentiation and impose efficiency and selectivity [22••,42,43]. STAT6 synergises with antigen to
T cell differentiation: a mechanistic view Avni and Rao
upregulate GATA3 [21••], a powerful inducer of Th2 differentiation [44,45]. Forced retroviral expression of STAT6 or GATA3 skews differentiation towards the Th2 phenotype, suppressing IFN-γ expression and increasing the frequency of cells expressing Th2 cytokines [21••,22••,23,46]. Forced expression of T-bet has the opposite effect [47••]. A small number of STAT6-deficient T cells are capable of IL-4 expression; strikingly, these are exactly those that detectably express GATA3 [22••], again indicating that STAT6 acts through GATA3 to increase the probability of Th2 cytokine expression. The transient activation of NFAT and STAT proteins leads us to postulate that they use a ‘hit-and-run’ mechanism to initiate cytokine gene expression, such as that utilised during mating type switching in yeast [48]. In this system, a very transient binding of the SWI5 transcription factor to the HO endonuclease gene suffices to recruit the SWI–SNF chromatin remodelling complex and the SAGA histone acetyltransferase complex, and the resulting chromatin structural changes facilitate the stable binding of a cell-type-specific transcription factor, SBF. Consistent with this hypothesis, NFAT is known to bind directly to promoter/enhancer regions of cytokine genes [12,13••] and may cooperate at some of these regions with cytokine-induced STAT factors. Binding sites for NFAT and STAT proteins have been identified in the IL-4 promoter, the IL-4 3′ enhancer and an intronic region of the IFN-γ gene [13••,49,50]. As mentioned above, both NFAT and STAT6 are capable of recruiting the coactivators p300 and CBP, which possess histone acetyltransferase activity [27–29]. Commitment
Commitment is mediated by subset-specific transcription factors such as GATA3, Maf and T-bet (reviewed in [6,7]). GATA3 and T-bet act globally on Th1 and Th2 differentiation, exerting both positive and negative effects on multiple cytokine genes, whereas Maf has a strong negative effect on Th1 differentiation but its positive effects are restricted to IL-4 [7,51,52]. The positive effects of these transcription factors must reflect their ability to maintain an accessible chromatin configuration of cytokine genes, most probably by direct recruitment of chromatinremodelling complexes and histone acetyltransferases to the regulatory regions of cytokine genes. The repressive effects could be mediated either directly (if the proteins bound and recruited repressive complexes to opposing cytokine genes) or indirectly (if they induced secondary factors that actually mediated repression). The basis for selective activation and repression remain to be understood: potentially, a single transcription factor might interact with distinct coactivator and corepressor proteins at different genetic loci in the same cell. T cell commitment requires a mechanism for maintaining constitutive expression of subset-specific nuclear factors in resting cells. So far, this requirement has been met by
657
GATA3, which autoregulates its own expression [22••]. Potentially, subset-specific factors could regulate — positively or negatively — not only their own expression but also that of other subset-specific factors, thus forming dynamic regulatory networks that maintain phenotypic homeostasis. Notably, cytokines and subset-specific transcription factors exert their skewing effects most efficiently when expressed at early stages of Th1/Th2 differentiation [21••,22••,23,46,53] although overexpressed GATA3 and T-bet do show a limited ability to reverse the phenotype of committed Th1 and Th2 clones, respectively [23,47••]. Cytokine loci may be particularly prone to irreversible silencing, perhaps by dense DNA methylation or repositioning to heterochromatin. The Ikaros family of haemopoietic-cell-specific transcription factors may be involved in maintaining epigenetic gene repression in replicating lymphocytes. Ikaros proteins are found in Mi-2–HDAC complexes with chromatin remodeling and histone deacetylase activity [54]; they colocalize with transcriptionally silent genes in heterochromatin [55,56]. Acute gene transcription
Gene transcription by differentiated Th1 and Th2 cells in response to TCR stimulation is mediated by inducible transcription factors such as NFAT [12]. Arrays of NFATbinding sites are found not only in proximal promoters but also in distal enhancers such as those identified in the IL-3, IL-4 and GM-CSF genes [13••,57,58]. Although NFAT is expressed at equivalent levels in T cell subsets, it binds cytokine regulatory regions in a cell-type-specific manner in vivo, associating with the IFN-γ promoter only in stimulated Th1 cells and with the IL-4 promoter and enhancer only in Th2 cells [13••]. The mechanism of selective association is unknown: it may reflect selective alterations in chromatin structure that occur during T helper cell differentiation or a need for cooperation with subset-specific transcription factors such as GATA3.
Conclusions The model of T cell differentiation discussed here has parallels in many other systems, including the β-globin locus, the chicken lysozyme locus and the human CD2 locus (reviewed in [25,38,39,59]). In each case, differentiating loci are characterised by changes in DNase I hypersensitivity and transcriptional competence; acquisition of transcriptional competence precedes overt gene transcription; differentiation and transcription are subject to binary, probabilistic regulation by enhancers and locus control regions; and full commitment to a specific lineage is established gradually, is effectively irreversible and correlates with DNA methylation, histone deacetylation and repositioning of genes to inactive regions of chromatin. With the advent of transgenic mice, the T cell system offers several advantages: large numbers of primary, antigen-specific naive T cells are available from TCR transgenic mice; differentiation and acute transcription are induced with the same antigenic stimulus but are cleanly
658
Atopic allergy and other hypersensitivities
separated in time; significant progress has been made in identifying transcription factors that influence differentiation, commitment and acute transcription; and a variety of transgenic and gene-targeted animal models have been develped in which to study the effects of transcription factors on allergy and asthma (e.g. [44]).
Acknowledgements We apologise to all those whose work we could not cite because of space limitations. We thank S Agarwal for the original version of Figure 1 (adapted from his PhD thesis, Harvard University) and D Lee, C Lopez-Rodriguez, S Feske and other members of the laboratory for helpful discussions. This work was supported by National Institutes of Health grants CA42471 and AI44432 (to AR). OA is supported by a postdoctoral fellowship from the Cancer Research Institute.
References and recommended reading Papers of particular interest, published within the annual period of review, have been highlighted as:
• of special interest •• of outstanding interest 1.
Abbas AK, Murphy KM, Sher A: Functional diversity of helper T lymphocytes. Nature 1996, 383:787-793.
2.
O’Garra A: Cytokines induce the development of functionally heterogenous T helper cell subsets. Immunity 1998, 8:275-283.
3.
Fitzpatrick DR, Kelso A: Nature versus nurture in T cell cytokine production. J Leukoc Biol 1999, 66:869-875.
4.
Reiner S, Seder R: Dealing from the evolutionary pawnshop: how lymphocytes make decisions. Immunity 1999, 11:1-10.
5.
Wurster AL, Tanaka T, Grusby MJ: The biology of Stat4 and Stat6. Oncogene 2000, 19:2577-2584.
6.
Murphy KM, Ouyang W, Farrar JD, Yang J, Ranganath S, Asnagli H, Afkarian M, Murphy TL: Signaling and transcription in T helper development. Annu Rev Immunol 2000, 18:451-494.
7.
Glimcher LH, Murphy KM: Lineage commitment in the immune system: the T helper lymphocyte grows up. Genes Dev 2000, 14:1693-1711.
8.
Agarwal S, Rao A: Modulation of chromatin structure regulates cytokine gene expression during T cell differentiation. Immunity 1998, 9:765-775.
9.
Takemoto N, Koyano-Nakagawa N, Yokota T, Arai N, Miyatake S, Arai K-I: Th2-specific DNAse I-hypersensitive sites in the murine IL-13 and IL-4 intergenic region. Int Immunol 1998, 10:1981-1985.
10. Bird JJ, Brown DR, Mullen AC, Moskowitz NH, Mahowald MA, Sider JR, Gajewski TF, Wang C, Reiner SL: Helper T cell differentiation is controlled by the cell cycle. Immunity 1998, 9:229-237. 11. Agarwal S, Viola J, Rao A: Chromatin-based regulatory mechanisms governing cytokine gene transcription. J Allergy Clin Immunol 1999, 103:990-999. 12. Rao A, Luo C, Hogan PG: Transcription factors of the NFAT family: regulation and function. Annu Rev Immunol 1997, 15:707-747. 13. Agarwal S, Avni O, Rao A: Cell-type-restricted binding of the •• transcription factor NFAT to a distal IL-4 enhancer in vivo. Immunity 2000, 12:643-652. By DNase I hypersensitivity analysis, the authors identified an inducible enhancer located 3′ of the IL-4 gene. The enhancer bound the Th2-specific transcription factor GATA3 in vivo. The transcription factor NFAT1, which is present in the nuclei of both Th1 and Th2 cells, bound the IL-4 enhancer and the IL-4 promoter only in stimulated Th2 cells; conversely, NFAT1 bound the IFN-γ promoter only in stimulated Th1 cells. These results support a model whereby transcription factors gain access to regulatory elements in chromatin only in cell types capable of expressing the corresponding genes. 14. Zhao K, Wang W, Rando O, Xue Y, Swiderek K, Kuo A, Crabtree G: • Rapid and phosphoinositol-dependent binding of the SWI/SNF-like BAF complex to chromatin after T lymphocyte receptor signaling. Cell 1999, 95:625-636. This paper provided the first clue as to how T cell activation might lead to generalised chromatin decondensation. The authors found that the chromatin-remodeling BAF complex becomes tightly associated with the nuclear
matrix within 10 minutes of lymphocyte activation. In vitro association of the complex with chromatin was enhanced by phosphatidylinositol (4,5) bisphosphate (PIP2), perhaps via an actin-like protein (BAF53) that is present in the complex. 15. Kingston RE, Narlikar GJ: ATP-dependent remodeling and acetylation as regulators of chromatin fluidity. Genes Dev 1999, 13:2339-2352. 16. Vignali M, Hassan AH, Neely KE, Workman JL: ATP-dependent chromatin-remodeling complexes. Mol Cell Biol 2000, 20:1899-1910. 17. ••
Loots GG, Locksley RM, Blankespoor CM, Wang ZE, Miller W, Rubin EM, Frazer KA: Identification of a coordinate regulator of interleukins 4, 13, and 5 by cross-species sequence comparisons. Science 2000, 288:136-140. The first identification of a functional non-promoter regulatory element solely by sequence comparisons of non-coding regions across species. The authors compared ~1 megabase of orthologous human 5q31 and mouse chromosome 11 sequences spanning the IL-4/IL/5/IL-13 and IL-3/GM-CSF loci, for conserved noncoding sequences (CNSs) with 70% or greater sequence identity over at least 100 base pairs. Characterization of the largest element, CNS-1, in transgenic mice suggested that it functioned in coordinate regulation of the IL-4, IL-5 and IL-13 genes. 18. Hardy KJ, Manger B, Newton M, Stobo JD: Molecular events involved in regulating human interferon-γγ gene expression during T cell activation. J Immunol 1987, 138:2353-2358. 19. Gross DS, Garrard WT: Nuclease hypersensitive sites in chromatin. Annu Rev Biochem 1988, 57:159-197. 20. Hernandez-Munain C, Roberts JL, Krangel MS: Cooperation among multiple transcription factors is required for access to minimal T-cell receptor alpha-enhancer chromatin in vivo. Mol Cell Biol 1998, 18:3223-3233. 21. Kurata H, Lee H, O’Garra A, Arai N: Ectopic expression of activated •• Stat6 induces the expression of Th2-specific cytokines and transcription factors in developing Th1 cells. Immunity 1999, 11:677-688. An unambiguous demonstration that STAT6 is responsible for Th2 cell differentiation. The authors introduced a STAT6–estrogen-receptor fusion protein, which can be activated independently of IL-4, into developing Th1 cells by retrovirus-mediated gene transfer. The transduced cells expressed Th2 cytokines and suppressed IFN-γ production if STAT6 was activated early during the first antigenic stimulation. STAT6 activation induced expression of both GATA3 and Maf, showing that these Th2-specific transcription factors are downstream targets of STAT6. 22. Ouyang W, Lohning M, Gao Z, Assenmacher M, Ranganath S, •• Radbruch A, Murphy K: Stat6-independent GATA-3 autoactivation directs IL-4-independent Th2 development and commitment. Immunity 2000, 12:27-37. This paper makes a very strong case for GATA3 as the agent of Th2 commitment. By using a sorting technique to identify viable IL-4-producing cells, the authors recognized that ~8% of STAT6-deficient cells are capable of producing IL-4. The population expressed GATA3 and displayed a stable Th2 phenotype. Retroviral introduction of GATA3 into STAT6-deficient T cells induced theTh2-specific pattern of DNase I hypersensitivity in the IL-4 locus. The authors also demonstrated the existence of an autoregulatory loop whereby GATA3 induced its own expression. 23. Lee HJ, Takemoto N, Kurata H, Kamogawa Y, Miyatake S, O’Garra A, Arai N: GATA-3 induces T helper cell type 2 (Th2) cytokine expression and chromatin remodeling in committed Th1 cells. J Exp Med 2000, 192:105-116. 24. Cron RQ, Bort SJ, Wang Y, Brunvand MW, Lewis DB: T cell priming enhances IL-4 gene expression by increasing nuclear factor of activated T cells. J Immunol 1999, 162:860-870. 25. Blackwood E, Kadonaga J: Going the distance: a current view of enhancer action. Science 1998, 281:60-63. 26. Strahl BD, Allis CD: The language of covalent histone modifications. Nature 2000, 403:41-45. 27.
Garcia-Rodriguez C, Rao A: Nuclear factor of activated T cells (NFAT)-dependent transactivation regulated by the coactivators p300/CREB-binding protein (CBP). J Exp Med 1998, 187:2031-2036.
28. Avots A, Buttmann M, Chuvpilo S, Escher C, Smola U, Bannister AJ, Rapp UR, Kouzarides T, Serfling E: CBP/p300 integrates Raf/Racsignaling pathways in the transcriptional induction of NF-ATc during T cell activation. Immunity 1999, 10:515-524.
T cell differentiation: a mechanistic view Avni and Rao
659
29. Gingras S, Simard J, Groner B, Pfitzner E: p300/CBP is required for transcriptional induction by interleukin-4 and interacts with Stat6. Nucleic Acids Res 1999, 27:2722-2729.
45. Zheng W, Flavell RA: The transcription factor GATA-3 is necessary and sufficient for Th2 cytokine gene expression in CD4 T cells. Cell 1997, 89:587-596.
30. Dhordain P, Lin RJ, Quief S, Lantoine D, Kerckaert JP, Evans RM, Albagli O: The LAZ3(BCL-6) oncoprotein recruits a SMRT/mSIN3A/histone deacetylase containing complex to mediate transcriptional repression. Nucleic Acids Res 1998, 26:4645-4651.
46. Ouyang W, Ranganath S, Weindel K, Bhattacharya D, Murphy T, Sha W, Murphy K: Inhibition of Th1 development mediated by GATA-3 through an IL-4-independent mechanism. Immunity 1998, 9:745-755.
31. Dent A, Shaffer AL, Yu X, Allman D, Staudt LM: Control of inflammation, cytokine expression, and germinal center formation by BCL-6. Science 1997, 276:589-592. 32. Dent AL, Hu-Li J, Paul WE, Staudt LM: T helper type 2 inflammatory disease in the absence of interleukin 4 and transcription factor STAT6. Proc Natl Acad Sci USA 1998, 95:13823-13828. 33. Ng H-N, Bird A: DNA methylation and chromatin modification. Curr Opin Genet Dev 1999, 9:158-163.
47. ••
Szabo SJ, Kim ST, Costa GL, Zhang X, Fathman CG, Glimcher LH: A novel transcription factor, T-bet, directs Th1 lineage commitment. Cell 2000, 100:655-669. The authors report the isolation of T-bet, the first transcription factor with a demonstrated major role in Th1 lineage specification. When retrovirally transduced into differentiated Th2 cells, T-bet induced IFN-γ and repressed IL-4 and IL-5 expression. 48. Cosma M, Tanaka T, Nasmyth K: Ordered recruitment of transcription and chromatin remodeling factors to a cell cycleand developmentally regulated promoter. Cell 1999, 97:299-311.
34. Fitzpatrick DR, Shirley KM, McDonald LE, Bielefeldt-Ohmann H, Kay GF, Kelso A: Distinct methylation of the interferon gamma (IFN-gamma) and interleukin 3 (IL-3) genes in newly activated primary CD8+ T lymphocytes: regional IFN-gamma promoter demethylation and mRNA expression are heritable in CD44(high)CD8+ T cells. J Exp Med 1998, 188:103-117.
49. Lederer JA, Perez VL, DesRoches L, Kim SM, Abbas AK, Lichtman AH: Cytokine transcriptional events during helper T cell subset differentiation. J Exp Med 1996, 184:397-406.
35. Wolffe AP, Jones PL, Wade PA: DNA demethylation. Proc Natl Acad Sci USA 1999, 96:5894-5896.
51. Ho I-C, Lo D, Glimcher L: C-maf promotes Th2 and attenuates Th1 differentiation by both IL-4 dependent and independent mechanisms. J Exp Med 1998, 188:1859-1866.
36. Bix M, Locksley RM: Independent and epigenetic regulation of the interleukin-4 alleles in CD4+ T cells. Science 1998, 281:1352-1354. 37.
Riviere I, Sunshine MJ, Littman DR: Regulation of IL-4 expression by activation of individual alleles. Immunity 1998, 9:217-228.
38. Festenstein R, Kioussis D: Locus control regions and epigenetic chromatin modifiers. Curr Opin Genet Dev 2000, 10:199-203. 39. Grosveld F: Activation by locus control regions? Curr Opin Genet Dev 1999, 9:152-157. 40. Gett A, Hodgkin P: Cell division regulates the T cell cytokine repertoire, revealing a mechanism underlying immune class regulation. Proc Natl Acad Sci USA 1998, 95:9488-9493. 41. Richter A, Lohning M, Radbruch A: Instruction for cytokine expression in T helper lymphocytes in relation to proliferation and cell cycle progression. J Exp Med 1999, 190:1439-1450. 42. Croft M, Swain SL: Recently activated naive CD4 T cells can help resting B cells, and can produce sufficient autocrine IL-4 to drive differentiation to secretion of T helper 2-type cytokines. J Immunol 1995, 154:4269-4282.
50. Xu X, Sun Y, Hoey T: Cooperative DNA binding and sequenceselective recognition conferred by the STAT amino-terminal domain. Science 1996, 273:794-797.
52. Kim J, Ho I-C, Grusby M, Glimcher L: The transcription factor c-Maf controls the production of interleukin-4 but not other Th2 cytokines. Immunity 1999, 10:745-751. 53. Murphy E, Shibuya K, Hosken N, Openshaw P, Maino V, Davis K, Murphy K, O’Garra A: Reversibility of T helper 1 and 2 populations is lost after long-term stimulation. J Exp Med 1996, 183:901-913. 54. Kim J, Sif S, Jones B, Jackson A, Koipally J, Heller E, Winandy S, Viel A, Sawyer A, Ikeda T et al.: Ikaros DNA-binding proteins direct formation of chromatin remodeling complexes in lymphocytes. Immunity 1999, 10:345-355. 55. Brown KE, Guest SS, Smale ST, Hahm K, Merkenschlager M, Fisher AG: Association of transcriptionally silent genes with Ikaros complexes at centromeric heterochromatin. Cell 1997, 91:845-854. 56. Brown KE, Baxter J, Graf D, Merkenschlager M, Fisher AG: Dynamic repositioning of genes in the nucleus of lymphocytes preparing for cell division. Mol Cell 1999, 3:207-217. 57.
Cockerill P, Bert A, Jenkins F, Ryan G, Shannon M, Vadas M: Human granulocyte-macrophage colony-stimulating factor enhancer function is associated with cooperative interactions between AP-1 and NFATp/c. Mol Cell Biol 1995, 15:2071-2079.
43. Jankovic D, Kullberg MC, Noben-Trauth N, Caspar P, Paul WE, Sher A: Single cell analysis reveals that IL-4 receptor/Stat6 signaling is not required for the in vivo or in vitro development of CD4+ lymphocytes with a Th2 cytokine profile. J Immunol 2000, 164:3047-3055.
58. Duncliffe K, Bert A, Vadas M, Cockerill P: A T cell-specific enhancer in the interleukin-3 locus is activated cooperatively by Oct and NFAT elements within a DNAse I-hypersensitive site. Immunity 1997, 6:175-185.
44. Zhang D, Yang L, Cohn L, Parkyn L, Homer R, Ray P, Ray A: Inhibition of allergic inflammation in a murine model of asthma by expression of a dominant-negative mutant of GATA-3. Immunity 1999, 11:473-482.
59. Bonifer C, Vidal M, Grosveld F, Sippel A: Tissue-specific and position independent expression of the complete gene domain for chicken lysozyme in transgenic mice. EMBO J 1990, 9:2843-2848.
T cell differentiation: a mechanistic view Orly Avni* and Anjana Rao† It has recently become clear that cytokine expression by T cells involves epigenetic changes in chromatin structure, locus accessibility and DNA methylation that occur during differentiation of naive T cells into mature effector T cells. These changes require the coordinate actions of antigen- and cytokine-induced transcription factors, chromatin remodeling complexes, histone-modifying enzymes and subset-specific transcription factors. Addresses Department of Pathology, Harvard Medical School and The Center for Blood Research, 200 Longwood Avenue, Boston, MA 02115, USA *e-mail: [email protected] † e-mail: [email protected] Current Opinion in Immunology 2000, 12:654–659 0952-7915/00/$ — see front matter © 2000 Elsevier Science Ltd. All rights reserved. Abbreviations BAF BRG1-associated factors CNS conserved noncoding sequence DH site DNase I hypersensitive site HDAC histone deacetylase kb kilobases
Introduction When naive T helper cells encounter antigen for the first time in the periphery, they begin a process of peripheral differentiation that includes commitment to a specific pattern of cytokine production. Th1 cell populations produce INF-γ but not IL-4 whereas Th2 cell populations transcribe the IL-4, IL-5 and IL-13 genes but not the INF-γ gene (individual T cells may show varied cytokine expression). These patterns of gene expression are strongly influenced by the local cytokine milieu: IL-12 and IL-4 strongly potentiate Th1 and Th2 differentiation, respectively, via the transcription factors STAT4 and STAT6. Clinically, INF-γ production is associated with inflammatory responses to bacterial antigens and with autoimmune disease whereas IL-4, IL-5 and IL-13 mediate allergic reactions and immune responses against helminthic parasites. These basic aspects of T cell differentiation have been thoroughly covered in many excellent reviews [1–7]. When first stimulated with antigen, naive T cells produce IL-2 but are incapable of producing large quantities of certain other cytokines, including IL-4 and IFN-γ. Stimulation initiates a differentiative phase in which those genetic loci that are destined to be transcriptionally active in the differentiated cells become ‘competent’ for future transcription, a process associated with changes in chromatin structure of cytokine genes ([8–10]; reviewed in [4,11]). Three stages of cytokine expression may be distinguished: an initiation phase that is highly dependent on antigen, cytokines and cytokine-induced STAT factors [5];
a commitment phase — mediated by subset-specific transcription factors such as GATA3 and cMaf (Maf) in Th2 cells and by T-bet and ERM in Th1 cells [6,7] — in which the differentiated phenotype is stabilised and maintained in the absence of further stimulation; and a phase of acute gene transcription, elicited by secondary contact of differentiated T cells with antigen, that requires the antigen-induced transcription factor NFAT [12,13••]. Here we review recent data on these phases of T cell differentiation, focusing on the IL-4/IL-5/IL-13 genes, which are particularly important for allergy.
Structural changes in chromatin during lymphocyte differentiation and activation Early chromatin decondensation
As pointed out by Zhao et al. [14•], antigen-induced activation leads to rapid and visible decondensation of the highly heterochromatic nucleus of naive T cells. These workers have suggested that chromatin decondensation might be mediated by the BAF (BRG1-associated factors) family of chromatin-remodeling complexes (for a recent review of chromatin-remodelling complexes, see [15,16]). TCR stimulation caused tight, rapid (i.e. after 10–15 minutes) and selective association of BAF complexes with the nuclear matrix of primary murine T cells, an effect apparently mediated by phosphatidylinositol (4,5) bisphosphate (PIP2) via actin and an actin-like protein (BAF53) that is present in the complex. The effect may not be limited to naive T cells but may constitute a general T cell activation response. BAF complexes may exert global effects in the nucleus or may be selective for certain target genes, for instance when gene expression is polarised by combined antigen/cytokine stimulation. Changes in DNase I hypersensitivity
In the IL-4/IL-13 locus, naive T cells exhibit only two DNase I hypersensitive sites (DH sites) in an interval of ~43 kilobases (kb) spanning the IL-4 and IL-13 genes; within 48 hours of primary stimulation in the presence of IL-4, 10 additional clusters of Th2-specific DH sites are observed ([8,9]; S Agarwal, D Lee, A Rao, unpublished data; Figure 1a). At this time, IL-4 mRNA is barely detectable, indicating that chromatin remodelling is an early differentiative event. Two intergenic DH sites and the 3′ DH site V are contained within conserved noncoding sequences (CNS-1 and -2) that were identified by sequence comparison of syntenic regions of mouse chromosome 11 and human 5q31 regions [17••] (Figure 1a). All 10 DH sites are stable markers of differentiated Th2 cells. Stimulation of differentiated cells leads to appearance of a novel hypersensitive site (VA) 3′ of the IL-4 gene (Figure 1a); this site corresponds to an inducible IL-4 enhancer, which binds the antigen-induced transcription factor NFAT and the Th2-specific nuclear factor GATA3 [13••].
T cell differentiation: a mechanistic view Avni and Rao
655
Figure 1 (a)
(b)
IL-13 /IL-4
Naive cells
IL-13
HSS3
IV
IL-4 12
Exon 1 234
I II III
Exon 1 234
IV
Activated cells
Initiation
23 Antigen, IL-12
Commitment
T-bet? Others
CNS-2 V
KIF3
Exon
1 234
5 kb
3 21
I Exon
Others
0
NFAT
IV
V
KIF3
3 4 Maf
4 Initiation Commitment
III 4 Acute transcription
NFAT Others
I II III 12
II
23 Antigen
1
Acute transcription VA (enhancer)
I II III
1
Differentiated cells
34
12 Antigen NFAT
II IFN- γ Exon
GATA3 Others CNS-1 321 0
Naive cells
34
Antigen, IL-4/STAT6
Differentiated cells I II III
KIF3
IFN-γ
NFAT
Activated cells
I 1
Exon GATA3
1 kb
III 23
4
NFAT Current Opinion in Immunology
Diagram showing changes in DNase I hypersensitivity over (a) ~43 kb of the IL-4/IL-13 locus and (b) ~11 kb of the IFN-γ locus during differentiation of murine CD4+ T cells and following activation of differentiated CD4+ T cells. Together with antigen, IL-4/STAT6 and IL-12/STAT4 play important roles in the initiation of Th2 and Th1 differentiation. Transcription factors that are preferentially expressed in the Th2 and Th1 lineages — GATA3, T-bet and others — are especially
important in the commitment phase; NFAT is especially important in acute gene transcription. At the bottom of the figure, NFAT, GATA3 and Maf are shown binding to the IL-4 promoter; for a discussion of other transcription factors implicated in IL-4 and IFN-γ gene transcription, see [7]. Numbered arrows indicate DH sites. In the middle and lower panels of (a), 3,2 and 1 above the lines refer to DH sites HSS3, HSS2 and HSS1, respectively. KIF3 is a brain-specific protein whose gene is near the IL-13/IL-4 locus.
In the IFN-γ locus, naive murine CD4+ T cells show only one weak DH site over an ~11 kb region spanning the IFN-γ gene; their differentiation into Th1 cells is accompanied by the appearance of at least two additional constitutive DH sites (Figure 1b). Stimulation of differentiated T cells attenuates the intensity of one of the DH sites in the murine IFN-γ gene [13••]. Inducible DH sites have been observed 5′ of the human IFN-γ promoter [18].
Maf-deficient T cells do not show detectable changes in the pattern of either the constitutive or inducible DH sites in the IL-4 gene [13••] but in vivo footprinting indicates that the Maf-binding site in the IL-4 promoter is constitutively occupied in resting Th2 cells [24].
The proteins that form the constitutive DH sites in the IL-4/IL-13 and IFN-γ loci have not yet been identified. DH sites mark nucleosome-free regions or other perturbations of canonical nucleosome structure, generally as a result of protein binding to adjacent DNA elements [19,20]. Appearance of DH sites in the IL-4 gene is far less efficient in T cells lacking STAT6 [8]. STAT6 synergises with antigen to induce the Th2-specific nuclear factors GATA3 and Maf [21••]. Ectopic expression of GATA3 suffices for appearance of all the Th2-specific DH sites in the IL-4 gene [22••,23], consistent with the possibility that GATA3 binds directly to at least some of these DH sites. GATA3 also binds directly (with NFAT) to the region of the inducible DH site VA, which marks an IL-4 enhancer [13••].
Histone modifications
Typically, acquisition of transcriptional competence is marked not only by an altered pattern of DH sites but also by overall increases in histone acetylation and general sensitivity to DNase I [25]. Other histone modifications, including phosphorylation and methylation, have been described [26]. Together, these modifications may induce an ‘open’ DNA conformation that permits greater access of transcription factors to their nucleosomal binding sites; moreover, the modified histone tails may serve as docking sites for nuclear proteins that positively influence transcription [26]. Like other transcription factors, NFAT and STAT6 bind the histone acetyltransferases p300 and CBP [27–29], and may recruit these coactivators to the IL-4/IL-5/IL-13 locus during Th2 differentiation. Conversely BCL-6, a zinc finger transcriptional repressor — which binds sites recognized by STAT factors and interacts with
656
Atopic allergy and other hypersensitivities
a histone deacetylase (HDAC)-containing corepressor complex [30] — represses Th2 differentiation as judged by the pronounced Th2 bias of BCL-6-deficient mice [31]. A similar Th2 bias is observed in T cells lacking both STAT6 and BCL-6 [32], suggesting either that BCL-6 inhibits a pathway of Th2 differentiation that does not involve STAT6 or that, in the absence of BCL-6, STAT6 is not as crucial for promoting chromatin accessibility. DNA demethylation
It has been known for many years that silent genes are present in regions of hypermethylated DNA whereas active genes are marked by hypomethylation. DNA methylation recruits methyl-CpG-binding proteins such as MeCP2, which in turn recruits corepressors such as the SIN3–HDAC complex [33]. Consistent with this paradigm, Th1 and Th2 differentiation are associated with selective demethylation of the IFN-γ and IL-4/IL-5/IL-13 genes, respectively [10,34]; however demethylation may be more relevant to commitment than to the initiation phase of T cell differentiation ([34]; S Agarwal, D Lee, A Rao, unpublished data). The bulk of DNA methylation occurs symmetrically on CpG dinucleotides. Newly replicated DNA strands contain cytosine in place of methylcytosine; following replication, DNA methyltransferases recognise the hemi-methylated CpG and recreate the fully methylated state. As a result, the methylation status of DNA is stably maintained even in dividing cells. The mechanism by which cytokine genes become demethylated during T cell differentiation is not understood: demethylation could occur through action of a demethylase or through local inhibition of the methyltransferase [35]. Probabilistic aspects
Cytokine expression during T cell differentiation is variegated, showing evidence of probabilistic, ‘on/off’ regulation in individual cells (reviewed in [3,4]). Even after optimal stimulation with antigen and cytokine, the fraction of naive T cells that actually commits to cytokine expression is surprisingly low. This behaviour is also apparent as a propensity to monoallelic expression of cytokine genes ([36,37] and references therein). The situation is reminiscent of position effect variegation, a phenomenon in which gene expression is determined in a stochastic fashion by chromatin accessibility [38]. Indeed, stimulation of naive T cells in the presence of butyrate (a general inhibitor of histone deacetylases) or azacytidine (an inhibitor of cytosine methylation) markedly increases cytokine expression; together the drugs strongly synergise to enhance IL-4 expression even in STAT6-deficient T cells [10]. The conserved CNS-1 region, containing the Th2-specific DH sites HSS1 and HSS2, also influences the probability rather than the level of Th2 cytokine expression [17••]. Transgenic mice carrying yeast artificial chromosomes containing the human IL-4/IL-13/IL-5 locus showed appropriate expression of the human cytokines in Th1,
Th2 and NK T cells; however, when CNS-1 was deleted from the transgene by Cre-mediated recombination, there was a 50–75% decrease in the number of Th2 cells expressing human IL-4 and IL-13, without a significant effect on the level of cytokine expression per cell. CNS-1 also affected expression of the human IL-5 gene, located 120 kb away; but it did not affect expression of the nearby RAD50 and KIF3 genes (these show ubiquitous and brain-specific expression, respectively). Thus CNS-1 acts selectively over long distances to increase the probability of IL-4/IL-5/IL-13 gene expression in a chromatin context, in a manner reminiscent of locus control regions (LCRs) and certain enhancers that can induce uniform expression and suppress silencing of transgenes that have been integrated into repressive locations such as centromeres [25,38,39].
Sequential actions of transcription factors during lymphocyte differentiation and activation Initiation Cell cycle dependence
In response to a single antigenic stimulation of naive T cells, cytokine expression showed a striking hierarchy of dependence on cell division: IL-2 was expressed prior to cell division, IFN-γ production required S phase entry, and IL-4 expression required a threshold number of three to four cell divisions [10,40]. In contrast, if naive T cells were first stimulated with antigen and then their ability to produce IL-4 was assessed in response to a second, short stimulation with PMA (phorbol myristate acetate) and ionomycin, cells that had not divided at all were capable of making IL-4 [41]; however, optimal IL-4 production still required multiple cell divisions and the use of cell-cycle inhibitors indicated that passage through S phase was necessary. Nucleosome disassembly occurring during the S phase might be needed for binding of nuclear factors to cytokine genes (or genes encoding other inducible proteins, such as T-bet or GATA3, that are needed for cytokine expression). DNA replication is also required for passive demethylation of DNA, necessary to displace the repressive chromatin structure induced by MeCP2–SIN3–HDAC complexes [35]. The role of antigen- and cytokine-induced transcription factors
TCR stimulation is essential to initiate Th1/Th2 differentiation. Riviere et al. [37] generated mice in which one IL-4 allele was replaced with DNA encoding a cell surface marker, human CD2. This allowed them to monitor independently the expression of each allele, by combining cell surface staining for the ‘knocked-in’ human CD2 marker and intracellular staining for endogenous IL-4. As the strength of initial antigenic stimulation was increased, a greater fraction of T cells expressed either or both alleles [37], implying that antigen-induced processes or transcription factors are limiting at early stages of differentiation. IL-4/STAT6 and IL-12/STAT4 increase the probability of T helper cell differentiation and impose efficiency and selectivity [22••,42,43]. STAT6 synergises with antigen to
T cell differentiation: a mechanistic view Avni and Rao
upregulate GATA3 [21••], a powerful inducer of Th2 differentiation [44,45]. Forced retroviral expression of STAT6 or GATA3 skews differentiation towards the Th2 phenotype, suppressing IFN-γ expression and increasing the frequency of cells expressing Th2 cytokines [21••,22••,23,46]. Forced expression of T-bet has the opposite effect [47••]. A small number of STAT6-deficient T cells are capable of IL-4 expression; strikingly, these are exactly those that detectably express GATA3 [22••], again indicating that STAT6 acts through GATA3 to increase the probability of Th2 cytokine expression. The transient activation of NFAT and STAT proteins leads us to postulate that they use a ‘hit-and-run’ mechanism to initiate cytokine gene expression, such as that utilised during mating type switching in yeast [48]. In this system, a very transient binding of the SWI5 transcription factor to the HO endonuclease gene suffices to recruit the SWI–SNF chromatin remodelling complex and the SAGA histone acetyltransferase complex, and the resulting chromatin structural changes facilitate the stable binding of a cell-type-specific transcription factor, SBF. Consistent with this hypothesis, NFAT is known to bind directly to promoter/enhancer regions of cytokine genes [12,13••] and may cooperate at some of these regions with cytokine-induced STAT factors. Binding sites for NFAT and STAT proteins have been identified in the IL-4 promoter, the IL-4 3′ enhancer and an intronic region of the IFN-γ gene [13••,49,50]. As mentioned above, both NFAT and STAT6 are capable of recruiting the coactivators p300 and CBP, which possess histone acetyltransferase activity [27–29]. Commitment
Commitment is mediated by subset-specific transcription factors such as GATA3, Maf and T-bet (reviewed in [6,7]). GATA3 and T-bet act globally on Th1 and Th2 differentiation, exerting both positive and negative effects on multiple cytokine genes, whereas Maf has a strong negative effect on Th1 differentiation but its positive effects are restricted to IL-4 [7,51,52]. The positive effects of these transcription factors must reflect their ability to maintain an accessible chromatin configuration of cytokine genes, most probably by direct recruitment of chromatinremodelling complexes and histone acetyltransferases to the regulatory regions of cytokine genes. The repressive effects could be mediated either directly (if the proteins bound and recruited repressive complexes to opposing cytokine genes) or indirectly (if they induced secondary factors that actually mediated repression). The basis for selective activation and repression remain to be understood: potentially, a single transcription factor might interact with distinct coactivator and corepressor proteins at different genetic loci in the same cell. T cell commitment requires a mechanism for maintaining constitutive expression of subset-specific nuclear factors in resting cells. So far, this requirement has been met by
657
GATA3, which autoregulates its own expression [22••]. Potentially, subset-specific factors could regulate — positively or negatively — not only their own expression but also that of other subset-specific factors, thus forming dynamic regulatory networks that maintain phenotypic homeostasis. Notably, cytokines and subset-specific transcription factors exert their skewing effects most efficiently when expressed at early stages of Th1/Th2 differentiation [21••,22••,23,46,53] although overexpressed GATA3 and T-bet do show a limited ability to reverse the phenotype of committed Th1 and Th2 clones, respectively [23,47••]. Cytokine loci may be particularly prone to irreversible silencing, perhaps by dense DNA methylation or repositioning to heterochromatin. The Ikaros family of haemopoietic-cell-specific transcription factors may be involved in maintaining epigenetic gene repression in replicating lymphocytes. Ikaros proteins are found in Mi-2–HDAC complexes with chromatin remodeling and histone deacetylase activity [54]; they colocalize with transcriptionally silent genes in heterochromatin [55,56]. Acute gene transcription
Gene transcription by differentiated Th1 and Th2 cells in response to TCR stimulation is mediated by inducible transcription factors such as NFAT [12]. Arrays of NFATbinding sites are found not only in proximal promoters but also in distal enhancers such as those identified in the IL-3, IL-4 and GM-CSF genes [13••,57,58]. Although NFAT is expressed at equivalent levels in T cell subsets, it binds cytokine regulatory regions in a cell-type-specific manner in vivo, associating with the IFN-γ promoter only in stimulated Th1 cells and with the IL-4 promoter and enhancer only in Th2 cells [13••]. The mechanism of selective association is unknown: it may reflect selective alterations in chromatin structure that occur during T helper cell differentiation or a need for cooperation with subset-specific transcription factors such as GATA3.
Conclusions The model of T cell differentiation discussed here has parallels in many other systems, including the β-globin locus, the chicken lysozyme locus and the human CD2 locus (reviewed in [25,38,39,59]). In each case, differentiating loci are characterised by changes in DNase I hypersensitivity and transcriptional competence; acquisition of transcriptional competence precedes overt gene transcription; differentiation and transcription are subject to binary, probabilistic regulation by enhancers and locus control regions; and full commitment to a specific lineage is established gradually, is effectively irreversible and correlates with DNA methylation, histone deacetylation and repositioning of genes to inactive regions of chromatin. With the advent of transgenic mice, the T cell system offers several advantages: large numbers of primary, antigen-specific naive T cells are available from TCR transgenic mice; differentiation and acute transcription are induced with the same antigenic stimulus but are cleanly
658
Atopic allergy and other hypersensitivities
separated in time; significant progress has been made in identifying transcription factors that influence differentiation, commitment and acute transcription; and a variety of transgenic and gene-targeted animal models have been develped in which to study the effects of transcription factors on allergy and asthma (e.g. [44]).
Acknowledgements We apologise to all those whose work we could not cite because of space limitations. We thank S Agarwal for the original version of Figure 1 (adapted from his PhD thesis, Harvard University) and D Lee, C Lopez-Rodriguez, S Feske and other members of the laboratory for helpful discussions. This work was supported by National Institutes of Health grants CA42471 and AI44432 (to AR). OA is supported by a postdoctoral fellowship from the Cancer Research Institute.
References and recommended reading Papers of particular interest, published within the annual period of review, have been highlighted as:
• of special interest •• of outstanding interest 1.
Abbas AK, Murphy KM, Sher A: Functional diversity of helper T lymphocytes. Nature 1996, 383:787-793.
2.
O’Garra A: Cytokines induce the development of functionally heterogenous T helper cell subsets. Immunity 1998, 8:275-283.
3.
Fitzpatrick DR, Kelso A: Nature versus nurture in T cell cytokine production. J Leukoc Biol 1999, 66:869-875.
4.
Reiner S, Seder R: Dealing from the evolutionary pawnshop: how lymphocytes make decisions. Immunity 1999, 11:1-10.
5.
Wurster AL, Tanaka T, Grusby MJ: The biology of Stat4 and Stat6. Oncogene 2000, 19:2577-2584.
6.
Murphy KM, Ouyang W, Farrar JD, Yang J, Ranganath S, Asnagli H, Afkarian M, Murphy TL: Signaling and transcription in T helper development. Annu Rev Immunol 2000, 18:451-494.
7.
Glimcher LH, Murphy KM: Lineage commitment in the immune system: the T helper lymphocyte grows up. Genes Dev 2000, 14:1693-1711.
8.
Agarwal S, Rao A: Modulation of chromatin structure regulates cytokine gene expression during T cell differentiation. Immunity 1998, 9:765-775.
9.
Takemoto N, Koyano-Nakagawa N, Yokota T, Arai N, Miyatake S, Arai K-I: Th2-specific DNAse I-hypersensitive sites in the murine IL-13 and IL-4 intergenic region. Int Immunol 1998, 10:1981-1985.
10. Bird JJ, Brown DR, Mullen AC, Moskowitz NH, Mahowald MA, Sider JR, Gajewski TF, Wang C, Reiner SL: Helper T cell differentiation is controlled by the cell cycle. Immunity 1998, 9:229-237. 11. Agarwal S, Viola J, Rao A: Chromatin-based regulatory mechanisms governing cytokine gene transcription. J Allergy Clin Immunol 1999, 103:990-999. 12. Rao A, Luo C, Hogan PG: Transcription factors of the NFAT family: regulation and function. Annu Rev Immunol 1997, 15:707-747. 13. Agarwal S, Avni O, Rao A: Cell-type-restricted binding of the •• transcription factor NFAT to a distal IL-4 enhancer in vivo. Immunity 2000, 12:643-652. By DNase I hypersensitivity analysis, the authors identified an inducible enhancer located 3′ of the IL-4 gene. The enhancer bound the Th2-specific transcription factor GATA3 in vivo. The transcription factor NFAT1, which is present in the nuclei of both Th1 and Th2 cells, bound the IL-4 enhancer and the IL-4 promoter only in stimulated Th2 cells; conversely, NFAT1 bound the IFN-γ promoter only in stimulated Th1 cells. These results support a model whereby transcription factors gain access to regulatory elements in chromatin only in cell types capable of expressing the corresponding genes. 14. Zhao K, Wang W, Rando O, Xue Y, Swiderek K, Kuo A, Crabtree G: • Rapid and phosphoinositol-dependent binding of the SWI/SNF-like BAF complex to chromatin after T lymphocyte receptor signaling. Cell 1999, 95:625-636. This paper provided the first clue as to how T cell activation might lead to generalised chromatin decondensation. The authors found that the chromatin-remodeling BAF complex becomes tightly associated with the nuclear
matrix within 10 minutes of lymphocyte activation. In vitro association of the complex with chromatin was enhanced by phosphatidylinositol (4,5) bisphosphate (PIP2), perhaps via an actin-like protein (BAF53) that is present in the complex. 15. Kingston RE, Narlikar GJ: ATP-dependent remodeling and acetylation as regulators of chromatin fluidity. Genes Dev 1999, 13:2339-2352. 16. Vignali M, Hassan AH, Neely KE, Workman JL: ATP-dependent chromatin-remodeling complexes. Mol Cell Biol 2000, 20:1899-1910. 17. ••
Loots GG, Locksley RM, Blankespoor CM, Wang ZE, Miller W, Rubin EM, Frazer KA: Identification of a coordinate regulator of interleukins 4, 13, and 5 by cross-species sequence comparisons. Science 2000, 288:136-140. The first identification of a functional non-promoter regulatory element solely by sequence comparisons of non-coding regions across species. The authors compared ~1 megabase of orthologous human 5q31 and mouse chromosome 11 sequences spanning the IL-4/IL/5/IL-13 and IL-3/GM-CSF loci, for conserved noncoding sequences (CNSs) with 70% or greater sequence identity over at least 100 base pairs. Characterization of the largest element, CNS-1, in transgenic mice suggested that it functioned in coordinate regulation of the IL-4, IL-5 and IL-13 genes. 18. Hardy KJ, Manger B, Newton M, Stobo JD: Molecular events involved in regulating human interferon-γγ gene expression during T cell activation. J Immunol 1987, 138:2353-2358. 19. Gross DS, Garrard WT: Nuclease hypersensitive sites in chromatin. Annu Rev Biochem 1988, 57:159-197. 20. Hernandez-Munain C, Roberts JL, Krangel MS: Cooperation among multiple transcription factors is required for access to minimal T-cell receptor alpha-enhancer chromatin in vivo. Mol Cell Biol 1998, 18:3223-3233. 21. Kurata H, Lee H, O’Garra A, Arai N: Ectopic expression of activated •• Stat6 induces the expression of Th2-specific cytokines and transcription factors in developing Th1 cells. Immunity 1999, 11:677-688. An unambiguous demonstration that STAT6 is responsible for Th2 cell differentiation. The authors introduced a STAT6–estrogen-receptor fusion protein, which can be activated independently of IL-4, into developing Th1 cells by retrovirus-mediated gene transfer. The transduced cells expressed Th2 cytokines and suppressed IFN-γ production if STAT6 was activated early during the first antigenic stimulation. STAT6 activation induced expression of both GATA3 and Maf, showing that these Th2-specific transcription factors are downstream targets of STAT6. 22. Ouyang W, Lohning M, Gao Z, Assenmacher M, Ranganath S, •• Radbruch A, Murphy K: Stat6-independent GATA-3 autoactivation directs IL-4-independent Th2 development and commitment. Immunity 2000, 12:27-37. This paper makes a very strong case for GATA3 as the agent of Th2 commitment. By using a sorting technique to identify viable IL-4-producing cells, the authors recognized that ~8% of STAT6-deficient cells are capable of producing IL-4. The population expressed GATA3 and displayed a stable Th2 phenotype. Retroviral introduction of GATA3 into STAT6-deficient T cells induced theTh2-specific pattern of DNase I hypersensitivity in the IL-4 locus. The authors also demonstrated the existence of an autoregulatory loop whereby GATA3 induced its own expression. 23. Lee HJ, Takemoto N, Kurata H, Kamogawa Y, Miyatake S, O’Garra A, Arai N: GATA-3 induces T helper cell type 2 (Th2) cytokine expression and chromatin remodeling in committed Th1 cells. J Exp Med 2000, 192:105-116. 24. Cron RQ, Bort SJ, Wang Y, Brunvand MW, Lewis DB: T cell priming enhances IL-4 gene expression by increasing nuclear factor of activated T cells. J Immunol 1999, 162:860-870. 25. Blackwood E, Kadonaga J: Going the distance: a current view of enhancer action. Science 1998, 281:60-63. 26. Strahl BD, Allis CD: The language of covalent histone modifications. Nature 2000, 403:41-45. 27.
Garcia-Rodriguez C, Rao A: Nuclear factor of activated T cells (NFAT)-dependent transactivation regulated by the coactivators p300/CREB-binding protein (CBP). J Exp Med 1998, 187:2031-2036.
28. Avots A, Buttmann M, Chuvpilo S, Escher C, Smola U, Bannister AJ, Rapp UR, Kouzarides T, Serfling E: CBP/p300 integrates Raf/Racsignaling pathways in the transcriptional induction of NF-ATc during T cell activation. Immunity 1999, 10:515-524.
T cell differentiation: a mechanistic view Avni and Rao
659
29. Gingras S, Simard J, Groner B, Pfitzner E: p300/CBP is required for transcriptional induction by interleukin-4 and interacts with Stat6. Nucleic Acids Res 1999, 27:2722-2729.
45. Zheng W, Flavell RA: The transcription factor GATA-3 is necessary and sufficient for Th2 cytokine gene expression in CD4 T cells. Cell 1997, 89:587-596.
30. Dhordain P, Lin RJ, Quief S, Lantoine D, Kerckaert JP, Evans RM, Albagli O: The LAZ3(BCL-6) oncoprotein recruits a SMRT/mSIN3A/histone deacetylase containing complex to mediate transcriptional repression. Nucleic Acids Res 1998, 26:4645-4651.
46. Ouyang W, Ranganath S, Weindel K, Bhattacharya D, Murphy T, Sha W, Murphy K: Inhibition of Th1 development mediated by GATA-3 through an IL-4-independent mechanism. Immunity 1998, 9:745-755.
31. Dent A, Shaffer AL, Yu X, Allman D, Staudt LM: Control of inflammation, cytokine expression, and germinal center formation by BCL-6. Science 1997, 276:589-592. 32. Dent AL, Hu-Li J, Paul WE, Staudt LM: T helper type 2 inflammatory disease in the absence of interleukin 4 and transcription factor STAT6. Proc Natl Acad Sci USA 1998, 95:13823-13828. 33. Ng H-N, Bird A: DNA methylation and chromatin modification. Curr Opin Genet Dev 1999, 9:158-163.
47. ••
Szabo SJ, Kim ST, Costa GL, Zhang X, Fathman CG, Glimcher LH: A novel transcription factor, T-bet, directs Th1 lineage commitment. Cell 2000, 100:655-669. The authors report the isolation of T-bet, the first transcription factor with a demonstrated major role in Th1 lineage specification. When retrovirally transduced into differentiated Th2 cells, T-bet induced IFN-γ and repressed IL-4 and IL-5 expression. 48. Cosma M, Tanaka T, Nasmyth K: Ordered recruitment of transcription and chromatin remodeling factors to a cell cycleand developmentally regulated promoter. Cell 1999, 97:299-311.
34. Fitzpatrick DR, Shirley KM, McDonald LE, Bielefeldt-Ohmann H, Kay GF, Kelso A: Distinct methylation of the interferon gamma (IFN-gamma) and interleukin 3 (IL-3) genes in newly activated primary CD8+ T lymphocytes: regional IFN-gamma promoter demethylation and mRNA expression are heritable in CD44(high)CD8+ T cells. J Exp Med 1998, 188:103-117.
49. Lederer JA, Perez VL, DesRoches L, Kim SM, Abbas AK, Lichtman AH: Cytokine transcriptional events during helper T cell subset differentiation. J Exp Med 1996, 184:397-406.
35. Wolffe AP, Jones PL, Wade PA: DNA demethylation. Proc Natl Acad Sci USA 1999, 96:5894-5896.
51. Ho I-C, Lo D, Glimcher L: C-maf promotes Th2 and attenuates Th1 differentiation by both IL-4 dependent and independent mechanisms. J Exp Med 1998, 188:1859-1866.
36. Bix M, Locksley RM: Independent and epigenetic regulation of the interleukin-4 alleles in CD4+ T cells. Science 1998, 281:1352-1354. 37.
Riviere I, Sunshine MJ, Littman DR: Regulation of IL-4 expression by activation of individual alleles. Immunity 1998, 9:217-228.
38. Festenstein R, Kioussis D: Locus control regions and epigenetic chromatin modifiers. Curr Opin Genet Dev 2000, 10:199-203. 39. Grosveld F: Activation by locus control regions? Curr Opin Genet Dev 1999, 9:152-157. 40. Gett A, Hodgkin P: Cell division regulates the T cell cytokine repertoire, revealing a mechanism underlying immune class regulation. Proc Natl Acad Sci USA 1998, 95:9488-9493. 41. Richter A, Lohning M, Radbruch A: Instruction for cytokine expression in T helper lymphocytes in relation to proliferation and cell cycle progression. J Exp Med 1999, 190:1439-1450. 42. Croft M, Swain SL: Recently activated naive CD4 T cells can help resting B cells, and can produce sufficient autocrine IL-4 to drive differentiation to secretion of T helper 2-type cytokines. J Immunol 1995, 154:4269-4282.
50. Xu X, Sun Y, Hoey T: Cooperative DNA binding and sequenceselective recognition conferred by the STAT amino-terminal domain. Science 1996, 273:794-797.
52. Kim J, Ho I-C, Grusby M, Glimcher L: The transcription factor c-Maf controls the production of interleukin-4 but not other Th2 cytokines. Immunity 1999, 10:745-751. 53. Murphy E, Shibuya K, Hosken N, Openshaw P, Maino V, Davis K, Murphy K, O’Garra A: Reversibility of T helper 1 and 2 populations is lost after long-term stimulation. J Exp Med 1996, 183:901-913. 54. Kim J, Sif S, Jones B, Jackson A, Koipally J, Heller E, Winandy S, Viel A, Sawyer A, Ikeda T et al.: Ikaros DNA-binding proteins direct formation of chromatin remodeling complexes in lymphocytes. Immunity 1999, 10:345-355. 55. Brown KE, Guest SS, Smale ST, Hahm K, Merkenschlager M, Fisher AG: Association of transcriptionally silent genes with Ikaros complexes at centromeric heterochromatin. Cell 1997, 91:845-854. 56. Brown KE, Baxter J, Graf D, Merkenschlager M, Fisher AG: Dynamic repositioning of genes in the nucleus of lymphocytes preparing for cell division. Mol Cell 1999, 3:207-217. 57.
Cockerill P, Bert A, Jenkins F, Ryan G, Shannon M, Vadas M: Human granulocyte-macrophage colony-stimulating factor enhancer function is associated with cooperative interactions between AP-1 and NFATp/c. Mol Cell Biol 1995, 15:2071-2079.
43. Jankovic D, Kullberg MC, Noben-Trauth N, Caspar P, Paul WE, Sher A: Single cell analysis reveals that IL-4 receptor/Stat6 signaling is not required for the in vivo or in vitro development of CD4+ lymphocytes with a Th2 cytokine profile. J Immunol 2000, 164:3047-3055.
58. Duncliffe K, Bert A, Vadas M, Cockerill P: A T cell-specific enhancer in the interleukin-3 locus is activated cooperatively by Oct and NFAT elements within a DNAse I-hypersensitive site. Immunity 1997, 6:175-185.
44. Zhang D, Yang L, Cohn L, Parkyn L, Homer R, Ray P, Ray A: Inhibition of allergic inflammation in a murine model of asthma by expression of a dominant-negative mutant of GATA-3. Immunity 1999, 11:473-482.
59. Bonifer C, Vidal M, Grosveld F, Sippel A: Tissue-specific and position independent expression of the complete gene domain for chicken lysozyme in transgenic mice. EMBO J 1990, 9:2843-2848.