4

Click here to load reader

Epigenetic Flexibility Underlying Lineage Choices in the Adaptive Immune System

  • Upload
    k

  • View
    213

  • Download
    1

Embed Size (px)

Citation preview

Page 1: Epigenetic Flexibility Underlying Lineage Choices in the Adaptive Immune System

PERSPECTIVE

Epigenetic Flexibility UnderlyingLineage Choices in theAdaptive Immune SystemDimitris Kioussis1* and Katia Georgopoulos2

Although fundamental models have emerged in recent years describing how chromatin and transcriptionregulation interface with one another in the developing immune system, the order of events and theirbiological impact are still being resolved. Recent advances have provided a flexible, rather than static, viewof chromatin regulation to reveal how both positive and negative forces work concomitantly to establishspecific chromatin structures and regulate gene expression. The challenge will now be to explore newepigenetic models and validate them during lymphocyte development, with the ultimate goal of unravelingthe long-sought mechanisms that support the emerging complexity of the adaptive immune response.

Lymphoid development is ultimately deter-mined by a succession of gene expressionprograms and by stage-specific networks

of classical transcriptional factors, which act asdrivers in the progression to specific immunecell types (1, 2). The activity of such cell fate–determining transcription factors is intimately linkedto dedicated chromatin modifiers that alter accessi-bility of lineage-specific gene loci and provide ulti-mate control over this process (3).A consideration ofthe regulated development of the adaptive immunesystem from a nuclear perspective must take intoaccount the extended potential to choose betweenalternative fates that characterizes lymphoid cells,from their earliest stages of development to their laterspecialization as immune cell types (Fig. 1). Suchcell fate choices, acting as they do at numerousbranching points, result in a diverse yet balancedimmune system and depend on the coordinatedacquisition of a gene expression program that favorsone cell fate over another. These programs are theresult of a combination of gene-activating and gene-silencing events and provide the molecular “sig-nature” for a particular cell fate. At cell stagespreceding fate choices, a subset of genes within aprogrammaybe transcriptionally primed (expressed)or poised (not expressed but readily activatable) fortranscription, providing cells with the potential fordifferentiation into alternative lineages (Fig. 1). Thisstate of affairs has been increasingly recognized aslineage priming that is regulated at both the tran-scriptional (4–6) and epigenetic (7, 8) level. At thesepre-decision stages, apparently opposing chroma-tin structures may coexist on a gene locus butresolve in subsequent stages into exclusively ac-tivating or silencing structures (Fig. 1). This type of

resolution may be decided at the point of cell divi-sion, when asymmetric distribution of regulatoryproteins in daughter cells can lead to differentialgene expression patterns (9). The retention of op-posing chromatin structures on lineage-specificgene loci may provide a potential at a later pointfor further differentiation by allowing epigeneticflexibility on key differentiation factors. Activationor repression of genes may be stably maintainedthrough the rest of the differentiation process, orthey may be “flexible,” reverting to earlier states atlater steps in the pathway.

During lymphocyte development, a certainset of “fixed” transcriptional decisions appearsto coexist with flexible changes in gene expres-sion. For example, T cell receptor expression isactivated at the double negative (DN) stage andis maintained at subsequent stages of T cell dif-ferentiation, whereas expression of the CD8 andCD4 co-receptors fluctuates during the develop-ment of cytotoxic and helper T cell lineages.

Epigenetic flexibility may endow developingimmune cells with their extended potential foralternative effector fate choices during terminaldifferentiation and may allow early progenitorsand their late progeny to share key molecularproperties. For example, both primitive hema-topoietic stem cells (HSCs) and memory immunecells survive for extended periods, possibly byusing similar genetic programs that contribute totheir capacity to self-renew (10, 11).

Modes of Epigenetic RegulationThe outcome of genetic programs set up duringlymphocyte development is influenced in partby the developmentally regulated gain or loss ofexpression of nuclear factors that modulate basaltranscription. It is also controlled by specificchanges in chromatin structure in the vicinity oflineage-specific genes. Chromatin structures havebeen classified as closed or open-permissive, de-pending on whether the genes included are silencedor expressed (12). And a number of histone mod-

ifications (known as histone code) have been as-sociated with such states (13).

Silenced chromatin, largely heterochromatic,contains a number of restrictive histone mod-ifications, such as meH3K9, meH3K27, meH4K20,and histone deacetylation, which allow for a higher-order packing and inaccessibility to transcriptionfactors. In addition, silenced chromatin frequentlycontains hypermethylated DNA. Conversely, open-permissive chromatin with histone modifications,such as meH3K4, meH3K36, acH3, and acH4,contains genes that are actively transcribed and isperceived to be accessible to regulators of tran-scription. Recent studies in embryonic stem (ES)cells have provided evidence for a third type ofchromatin, referred to as “bivalent,” as a way ofgenerating developmental plasticity through epi-genetic flexibility (Fig. 1). This type combinesthe characteristics of both closed (meH3K27) aswell as permissive (meH3K4) chromatin struc-tures and marks lineage-specific genes poised forlater lineage-specific activation (7, 8). In a modeldeduced from these studies, the repressive chro-matin modifications keep lineage-specific genesin a transcriptionally inactive state, while thepermissive chromatin modifications keep thempoised for activation once the former influenceis removed. Conversely, removal of the permis-sive chromatin modifications may also allowthe repressive chromatin modifications to pre-vail, thus establishing gene silencing. Increasedexpression of relevant transcription regulatorsduring development may also aid the resolutionof bivalent epigenetic structures into their respec-tive activating or repressing states in a perma-nent fashion.

Epigenetic Flexibility in DifferentiatingImmune Cells?The existence of bivalent chromatin structures andtheir role in providing cell fate flexibility duringsomatic cell differentiation and during developmentof the immune system need further exploration.Lineage-specific genes with flexible expression areobserved throughout the lymphoid pathway, fromthe earliest pre-commitment steps to the later pre-effector cell stages, and these genes represent ex-cellent candidates for testing the presence and roleof bivalent epigenetic states during lymphocyte de-velopment as well as maturation to effector states(Fig. 1). For example, components of the antibody-producing machinery (e.g., immunoglobulin J andsterile transcripts from the immunoglobulin heavychain constant regions implicated in immunoglo-bulin class switching) required in terminally dif-ferentiated plasma cells are expressed in earlyhematopoietic multipotent progenitors and aretemporarily repressed during early B cell differenti-ation (to be reactivated in the periphery at later stagesof B cell development) (14). Similarly, in earlythymocyte precursors that lack both CD4 and CD8co-receptors (DN), the genes for each co-receptormay acquire a bivalent chromatin state poised for

1Molecular Immunology, Medical Research Council (MRC)National Institute for Medical Research, The Ridgeway,London NW7 1AA, UK. 2Cutaneous Biology Research Center,Massachusetts General Hospital, Harvard Medical School, 14913th Street, Charlestown, MA 02129, USA.

*To whom correspondence should be addressed. E-mail:[email protected]

3 AUGUST 2007 VOL 317 SCIENCE www.sciencemag.org620

Challenges in Immunology

on

Dec

embe

r 21

, 201

4w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from

o

n D

ecem

ber

21, 2

014

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

on

Dec

embe

r 21

, 201

4w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from

Page 2: Epigenetic Flexibility Underlying Lineage Choices in the Adaptive Immune System

expression. This state would then resolve into ac-tivating structures at the double positive (DP) stageof T cell differentiation. After thymocytes havereached the DP (CD4+ CD8+) stage of develop-ment, these genes may again become temporarilypoised, before their fate is determined by perma-nently resolving the bivalent state in a com-plementary fashion in mature cytotoxic (CD8) andhelper (CD4) T cells (15). It is possible that

expression of regulatorygenes such asGATA-3,Tbet,Fox-P3, andRORgamma, which influence decisionsboth in early Tcell development and at later steps ofT helper lineage maturation (1), may also be mod-ulated through such flexible epigenetic mechanisms.

This handful of known examples lends evidencefor a broader flexibility in the genetic programs thatbestow lymphocytes with an extended differentia-tion potential. Genome-wide investigations on de-velopmentally relevant cells [for example, withthe use of chromatin immunoprecipitation–on-chiptechnology (16, 17)] in the concurrence of restric-tive and permissive chromatin modifications atvarious branch points of the lymphoid pathways

may offer yet further examples of genetic programspoised for activation and help us define their modeof establishment and resolution. For example, itwill be important to determine whether the resolu-tion to a permissive or silenced state is permanent,or whether these can revert back to a bivalentstate, which may indicate restoration of previouspotential. This type of epigenetic regulation mayalso explain the relative ease with which lym-

phocytes can be reprogrammed through much oftheir ontogeny to their closer relatives within thehematopoietic system: the myeloid lineage. For ex-ample, the inappropriate introduction or removal ofantagonistic lineage-specific transcriptional regulators(e.g., loss of PAX-5 in precursor B cells and ectopicexpression of C/EBPa in mature B cells) may al-low activation of a myeloid genetic program thatmay exist in a bivalent state in these cells (18–20).

Regulators of Epigenetic FlexibilityTo date, the molecular players and their interactionsthat generate and regulate bivalent flexible chroma-tin structures and their resolution are not well de-

fined. It is possible that epigenetic flexibility may bebrought about by the concomitant action of func-tionally opposed chromatin regulators occupyingthe same chromosomal site. Precedence for such amode of regulation is again provided in ES cells,where positive chromatin regulators and membersof the trithorax complex coexist with negative reg-ulators of the Polycomb group on lineage differen-tiation genes, several of which have been shown toexist in a bivalent chromatin configuration (21).The CD4 and CD8 genes lend themselves as po-tential paradigms for identifying such dueling epige-netic regulators during T cell development. Thecoexistence of opposing chromatin regulators onthe CD4 locus and its regulatory elements providesa ground for identifying such a “bimodal” reg-ulatory network composed of competing activitiesand responsible for setting bivalent chromatin statesand flexible transcriptional outcomes during devel-opment (22–24).

Another example for generating poised chro-matin is suggested by the structure of the nucleo-some remodeling and deacetylase (NuRD) complex,in which the adenosine triphosphate–dependentchromatin remodeler Mi-2b that provides chromatinfluidity coexists with histone deacetylases that areusually associated with repressive chromatin struc-tures (25, 26). The coexistence of antagonistic en-zymatic activities within a protein complex mayensure both proper chromatin regulation and epi-genetic flexibility. The chromatin remodeler Mi-2bin the NuRD complex is one such potential directbimodal regulator of CD4 gene expression duringT cell development (24, 27). A direct partner ofMi-2b in this NuRD-based chromatin-remodelingcomplex is Ikaros: a sequence-specific DNA bind-ing factor implicated in early lymphocyte develop-ment (26, 28, 29). Ikaros, through its association andgene-specific targeting of such a bivalent complexin the HSC and its immediate progeny, may conferlineage plasticity and the potential for differentiationto these cells. It would be important to determinewhether such an Ikaros bimodal complex effects thepriming or poising of lineage-specific gene expres-sion programs in the HSC and its early progeny.

One further challenge will be to determinewhether and how opposing activities within a pro-tein complex on a given genetic locus are regulated.For example, DNA bound chromatin-modifyingcomplexes and their components may be amenableto modifications, such as those to histones, thatcould influence their overall activity and, therefore,chromatin dynamics. Another challenge is to obtaina more global view on the recruitment of suchchromatin regulators to gene loci associated withlineage-specific expression signatures and determinethe DNA binding activities or chromatin (histonecode) platform that determines such targeting.

Nuclear Compartmentalization ofGenetic ProgramsIn addition to the mechanisms described above,classical competition between opposing regulators

Uncommitted

Precursor/pre-effector Fate 1 Fate 3

Fate 2

Fate 4

Fig. 1. Developmental progression from an uncommitted state and a precursor–pre-effector state to distinctcell fates from a gene expression–chromatin perspective. Four programs in gene expression (shown incompartments within the nucleus), whose combinatorial acquisition allows for distinct cell fate choices(depicted by the differently colored cell perimeters) to be made, are shown from the early to the later stepsof the pathway. The poised (not expressed but activatable) or primed (low expression) state of the genes inthese programs is indicated by a half-black, half-white circle representing bivalent chromatin and a dashedbent arrow for low or no transcription. Activated genes are indicated by a solid white circle and a solid bentarrow for transcription. Genes repressed in a permanent fashion are indicated by a solid black circle and ablock (indicated by “×”) on the transcriptional arrow. Different scenarios in the resolution of poised lineage-specific genetic program are entertained, leading to distinct cell fate choices. Fate choices with poisedgenetic programs may be reversible, whereas others without such programs may be permanent.

www.sciencemag.org SCIENCE VOL 317 3 AUGUST 2007 621

SPECIALSECTION

Page 3: Epigenetic Flexibility Underlying Lineage Choices in the Adaptive Immune System

for the same DNA binding sequence at a chromo-somal site or for an established chromatin (histonecode) domain may also determine the transcrip-tional activity of a regulated gene. In these cases,fluctuating concentrations of nuclear factors withdisparate activities during development can allowfor a transcriptional flexibility that would obeymass action rules by establishing working equilib-ria between activating and silencing components(12). Such equilibria may be modified not only byvarying the production of any particular factor butalso by regulating the rate of their synthesis, stability,and degradation, as well as by sequestering them indifferent nuclear compartments. In the latter case,gene activity could be determined by moving genesinto nuclear compartments where different types ofregulators predominate. Indeed, the positioning ofgene loci within nucleus subdomains has emergedas a potentially important determinant of gene ac-tivity (27, 30, 31). Genes associated with hetero-chromatic regions of the nucleus (perinuclear,centromeric clusters) seem to be silent. So far, theassociation is correlative, and it is unclear whetherthe silencing precedes or is the result of this lo-calization. Better characterization (composition anddynamics) of such active or silencing regions willrequire the identification of molecules responsible(i) for setting up the environment in these domainsand (ii) for the movement of genes from one region

of the nucleus to another. Improvements inresolution and specificity of the tools needed forthe identification and visualization of these compo-nents will be one of the most formidable techno-logical challenges in the forthcoming years.

Concluding RemarksThe hematopoietic system, in which cell lineagechoices are well characterized and a substantialnumber of transcription regulators of cell fate andtheir targets have been identified, provides an ex-cellent paradigm to study the mechanisms thatunderlie lineage progression and plasticity. Initialsteps in such studies are already identifying epige-netic states by which lineage priming and plasticityare achieved and are suggesting that the three dis-crete states of chromatin may be achieved by dif-ferent mechanisms at different stages in thehematopoietic lineage. The ability to use alternativemechanisms at multiple steps during differentiationmakes the hematopoietic system an important con-tributor to future research on epigenetic models ofgene regulation in normal development and disease.

References and Notes1. E. V. Rothenberg, Nat. Immunol. 8, 441 (2007).2. K. L. Medina, H. Singh, Curr. Opin. Hematol. 12, 203

(2005).3. K. Georgopoulos, Nat. Rev. Immunol. 2, 162 (2002).4. K. Akashi, Ann. N.Y. Acad. Sci. 1044, 125 (2005).

5. R. Mansson et al., Immunity 26, 407 (2007).6. P. Laslo et al., Cell 126, 755 (2006).7. B. E. Bernstein et al., Cell 125, 315 (2006).8. V. Azuara et al., Nat. Cell Biol. 8, 532 (2006).9. S. L. Reiner, F. Sallusto, A. Lanzavecchia, Science 317,

622 (2007).10. D. T. Fearon, P. Manders, S. D. Wagner, Science 293,

248 (2001).11. J. T. Chang et al., Science 315, 1687 (2007).12. R. Festenstein, D. Kioussis, Curr. Opin. Genet. Dev. 10,

199 (2000).13. T. Kouzarides, Cell 128, 693 (2007).14. A. Delogu et al., Immunity 24, 269 (2006).15. D. Kioussis, W. Ellmeier, Nat. Rev. Immunol. 2, 909 (2002).16. Y. Zheng et al., Nature 445, 936 (2007).17. A. Marson et al., Nature 445, 931 (2007).18. C. Cobaleda et al., Nat. Immunol. 8, 463 (2007).19. H. Xie et al., Cell 117, 663 (2004).20. C. V. Laiosa, M. Stadtfeld, T. Graf, Annu. Rev. Immunol.

24, 705 (2006).21. L. A. Boyer et al., Nature 441, 349 (2006).22. T. H. Chi et al., Nature 418, 195 (2002).23. T. Sato et al., Immunity 22, 317 (2005).24. C. J. Williams et al., Immunity 20, 719 (2004).25. Y. Zhang et al., Cell 95, 279 (1998).26. J. Kim et al., Immunity 10, 345 (1999).27. K. E. Brown et al., Mol. Cell 3, 207 (1999).28. T. Yoshida et al., Nat. Immunol. 7, 382 (2006).29. S. Y. Ng, T. Yoshida, K. Georgopoulos, Curr. Opin. Immunol.

19, 116 (2007).30. T. Misteli, Cell 128, 787 (2007).31. S. T. Kosak et al., Science 296, 158 (2002).32. K.G. is supported by the National Institute of Allergy and

Infectious Diseases, and D.K. is supported by the MRC.

10.1126/science.1143777

PERSPECTIVE

Division of Labor with a Workforceof One: Challenges in SpecifyingEffector and Memory T Cell FateSteven L. Reiner,1* Federica Sallusto,2* Antonio Lanzavecchia2*In the course of the immune response against microbes, naïve T cells proliferate and generate variedclasses of effector cells, as well as memory cells with distinct properties and functions. Owing to recenttechnological advances, some of the most imposing questions regarding effector and memory T celldifferentiation are now becoming experimentally soluble: How many classes of antigen-specific T cellsexist, and how malleable are they in their fate and in their functional state? How might a spectrum of cellfates be imparted to the clonal descendants of a single lymphocyte? Where, when, and how doespathogen-associated information refine the instruction, selection, and direction of newly activated T cellsas they perform their tasks in different locations and times? Some surprising new glimpses ahead on thesesubjects and other yet-unanswered questions are discussed.

Specific immunity adapts to the threat ofpathogen attack with vigorous clonal ex-pansion of a selected lymphocyte whose

antigen receptor binds microbial peptide in the

context of self major histocompatibility molecules.The culmination of specific immunity is the gen-eration of effector cells that are responsible foracute elimination of the pathogen and memorycells that patrol their various tissue domains insearch of evidence of re-attack.

Heterogeneity is a hallmark of antigen-specificT cells. CD4+ T cells make effector choices tobecome T helper cell 1 (TH1), TH2, or TH17 cellsand might likewise choose to become antigen-specific regulatory cells (1–3). In addition to

choice of cytokine repertoire, effector CD4+ Tcells exhibit diversity in homing, such as migrationto peripheral nonlymphoid tissue versus transit tolymph node follicles to promote B cell help (4).Heterogeneity of CD8+ T cell effector gene ex-pression has been described (5), although it is notclear whether this represents physiologically dis-tinct cell fates or simply fluctuation in activationstate. Memory T cells are heterogeneous, with cen-tral memory cells that patrol secondary lymphoidtissues, recapitulating the surveillance of their naïveprogenitor, and effector memory cells that act assentinels standing guard at frontline barriers (6).

Although the role and function of effector andmemory subsets in protection or pathology and thenature of polarizing signals required for their dif-ferentiation are becoming increasingly clear, thereare still outstanding questions that need to beaddressed that relate to the mechanism of T cellfate specification. Many of these questions dealwith fundamental uncertainties that are common tomany areas of blood differentiation, such as theextent of fate diversity, the ontogeny and lineagerelationship between opposing and kindred fates,and the degree of natural and therapeutic plasticityat different stages of differentiation.

“One Cell, One Fate” Versus“One Cell, Multiple Fates”Signaling and transcription during T cell activa-tion have traditionally been viewed as a uniformprocess. Any given naïve precursor cell could be

1Abramson Family Cancer Research Institute of the Universityof Pennsylvania, Philadelphia, PA 19104, USA. 2Institute forResearch in Biomedicine, Via Vincenzo Vela 6, CH-6500Bellinzona, Switzerland.

*To whom correspondence should be addressed. E-mail:[email protected] (S.L.R); [email protected] (F.S.); [email protected] (A.L.)

3 AUGUST 2007 VOL 317 SCIENCE www.sciencemag.org622

Challenges in Immunology

Page 4: Epigenetic Flexibility Underlying Lineage Choices in the Adaptive Immune System

DOI: 10.1126/science.1143777, 620 (2007);317 Science

Dimitris Kioussis and Katia GeorgopoulosImmune SystemEpigenetic Flexibility Underlying Lineage Choices in the Adaptive

This copy is for your personal, non-commercial use only.

clicking here.colleagues, clients, or customers by , you can order high-quality copies for yourIf you wish to distribute this article to others

  here.following the guidelines

can be obtained byPermission to republish or repurpose articles or portions of articles

  ): December 21, 2014 www.sciencemag.org (this information is current as of

The following resources related to this article are available online at

http://www.sciencemag.org/content/317/5838/620.full.htmlversion of this article at:

including high-resolution figures, can be found in the onlineUpdated information and services,

http://www.sciencemag.org/content/317/5838/620.full.html#relatedfound at:

can berelated to this article A list of selected additional articles on the Science Web sites

http://www.sciencemag.org/content/317/5838/620.full.html#ref-list-1, 4 of which can be accessed free:cites 31 articlesThis article

11 article(s) on the ISI Web of Sciencecited by This article has been

http://www.sciencemag.org/content/317/5838/620.full.html#related-urls6 articles hosted by HighWire Press; see:cited by This article has been

http://www.sciencemag.org/cgi/collection/immunologyImmunology

subject collections:This article appears in the following

registered trademark of AAAS. is aScience2007 by the American Association for the Advancement of Science; all rights reserved. The title

CopyrightAmerican Association for the Advancement of Science, 1200 New York Avenue NW, Washington, DC 20005. (print ISSN 0036-8075; online ISSN 1095-9203) is published weekly, except the last week in December, by theScience

on

Dec

embe

r 21

, 201

4w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from