sign in sign up
<<
Home , Chromatin remodeling, Histone methylation, Thomas Jenuwein

Downloaded from genesdev.cshlp.org on September 23, 2021 - Published by Cold Spring Harbor Laboratory Press

MEETING REVIEW

Wild : regulation of eukaryotic genes in their natural chromatin context

Valerio Orlando1 and Katherine A. Jones2,3 1Dulbecco Telethon Institute, Institute of Genetics and Biophysics National Research Council, 80131 Naples, Italy; 2Regulatory Biology Laboratory, The Salk Institute for Biological Studies, La Jolla, California 92037, USA

Nearly 40 years ago, François Jacob and Jacques Monod ings highlight a key role for modification of individual provided the first genetic evidence that genes are con- in gene regulation. Ultimately, the stabil- trolled through the opposing actions of activators and ity of open or closed chromatin is dictated by reversible on specific DNA targets. Although minimal- tail modifications, which occur in patterns that istic, the classic paradigm provided by the regulation of correlate with gene activation or repression. A major the Lac in Escherichia coli quite accurately re- task is to unravel how gene- and locus-specific patterns flects the rationale of eukaryotic gene regulation, even in of modifications are initiated and maintained. the context of highly organized structure. At the meeting, Dave Allis (University of Virginia, Transcriptional activation of most eukaryotic genes is Charlottesville, VA, USA) showed that the Saccharomy- now known to require several 10s of factors, many of ces cerevisiae Rad6 gene product regulates which are responsive to developmental or environmen- at Lys 4 (K4) on , often considered a hallmark tal signaling pathways. Some regulators are enzymes of gene activation. Remarkably, RAD6 is not a histone that modify or reconfigure chromatin, whereas others but, rather, functions as a - influence the subnuclear organization of conjugating enzyme specific for -K120 (K123 factors or, in some cases, change the relative chromo- in yeast). Thus, ubiquitination of histone H2B is epi- somal environments of specific gene targets. Posttrans- static to the histone H3-K4 methylation. The converse is lational modifications of or factors de- not true, namely, methylation at histone H3-K4 is not fine localized chromatin domains as well as the activity required for ubiquitination of histone H2B. This example and turnover of enhancer complexes. A crucial problem illustrates the importance of nucleosomal modification in the field now is to decipher how proper regulation cross-talk, in which a modification at one site can pro- springs from combinations of positive and negative fac- foundly influence modification at nearby sites or, re- tors, generally organized in multiprotein complexes or markably, even on distant histone tails. Recent studies coupled to molecular engines, which recognize and have suggested sequence similarities between histone modify regulatory loci embedded in chromatin. deacetylase and deubiquitinating enzymes, which raises Remarkably, after an early period in which nucleo- the intriguing possibility that inhibi- somal DNA was seen as a negative and somehow passive tors such as trichostain A (TSA) might impact ubiquitin- template for transcription, chromatin and chromosome dependent pathways, which, in organization in the nucleus have now emerged as central turn, affect histone . The Allis group have parameters with direct and essential effects on the con- also found that at Ser 14 on histone H2B trol of . The workshop “Regulation of occurs selectively on apoptotic chromatin in vertebrates Eukaryotic Genes in Their Natural Chromatin Con- and may be mediated by Mst1 (mammalian sterile-like text,” hosted by Fundaciòn Juan March in Madrid, Spain kinase), an apoptotic-inducing kinase that acts immedi- (April 22–24, 2002), was organized by Miguel Beato and ately downstream of caspase-3s. Ken Zaret to address some of the most pressing issues Louis Mahadevan (University of Oxford, UK) ad- concerning how complex regulatory networks work in dressed whether phosphorylation of histone H3 at S10 their natural molecular environment: the chromosome. and acetylation at residues K4 and K9, which are both early events in the induction of the c-fos and c-jun genes, Histone modifications in the control of gene occur as concerted or independent events. Disruption of expression and plasticity the MSK1 and MSK2kinases resulted in the loss of S10 phosphorylation without affecting the overall levels of Contrary to the view that chromatin functions as a natu- histone H3 acetylation, indicating that histone phos- ral barrier to most DNA-dependent processes, new find- phorylation and acetylation may be regulated indepen- 3Corresponding author. dently. These findings contrast with previous reports E-MAIL [email protected]; FAX (858) 552-8285. Article and publication are at http://www.genesdev.org/cgi/doi/10.1101/ that phosphorylation at S10 precedes acetylation and gad.1017402. gene activation. This group also reported that whereas

GENES & DEVELOPMENT 16:2039–2044 © 2002 by Cold Spring Harbor Laboratory Press ISSN 0890-9369/02 $5.00; www.genesdev.org 2039 Downloaded from genesdev.cshlp.org on September 23, 2021 - Published by Cold Spring Harbor Laboratory Press

Orlando and Jones

H3 phosphorylation is an inducible event, acetylation is some maintained H3-K9 methylation. This implies the continuous and dynamic, even in quiescent cells and in existence in of more than one type of the absence of transcription. The latter finding may be histone H3 methylation, and also indicates that more particularly relevant for maintenance programs, in than one HMT controls the methylation state of hetero- which a locus specifies an appropriate level of acetyla- chromatin. Whether histone methyl groups are tion through the regulated actions of opposing enzymat- removed enzymatically is an open question at the pres- ic activities. To define the function of different chroma- ent. tin modifications in the various stages of mammalian Much discussion focused on the fact that not all his- X-chromosome inactivation, Bryan Turner (University tone modifications are created equal. In particular, K4 of Birmingham, UK) presented immunohistochemical and K9 of histone H3 can be mono-, di-, or trimethylated, evidence that global changes in the acetylation of all four and existing antisera do not appropriately distinguish core histones occur in parallel and follow down-regula- these modifications. Should these distinct methylation tion (silencing) of X-linked genes in female ES cells. Re- states differentially impact gene regulation or chromatin markably, analyses using novel antisera capable of dis- compaction, the combinatorial power of signaling tinguishing di- and trimethylated lysine residues (H3-K4, through histone modification would be enhanced signifi- H3-K9) revealed that X inactivation is accompanied by cantly. Moreover, because most in vivo studies of his- changes in the relative abundance of specific methyl- tone modification rely on chromatin immunoprecipita- ation modifications. Thus, although the inactive X-chro- tion or immunolocalization, the quality and specificity mosome (Xi) lacks H3 dimethyl K4, H3 trimethyl K4, of existing antibodies restricts our understanding of the and H3 trimethyl K9, it retains H3 dimethyl K9 in state of native chromatin. Further, Louis Mahadevan amounts equivalent to those on the active X-chromo- (University of Oxford, UK) reported that second-site some (Xa). Evidence was presented that the relatively modifications in histone tails can strongly interfere with enhanced immunostaining of Xi with certain antisera to epitope recognition. In particular, he reported the case of H3 methyl K9 is fixation-dependent, and likely reflects a commercial K9/K18 acetyl H3 antiserum that could differences in chromatin conformation rather than in- not detect the epitope if Ser10 is phosphorylated. In this creased levels of H3 methylation. case, therefore, phosphorylation at S10 would mask de- The mammalian genome undergoes extensive repro- tection of histone H3 acetylation. Bryan Turner (Univer- gramming after fertilization. The paternal genome is sity of Birmingham Medical School, Birmingham, UK) widely demethylated in the zygote, whereas the mater- has shown that antibodies raised against dimethylated nal genome is demethylated during cleavage division in H3 tail peptides share little or no cross-reactivity with the preimplantation embryo. After these epigenetic antibodies raised against the equivalent trimethylated marks are removed, a wave of de novo methylation oc- peptides, and vice versa. This issue will deserve particu- curs postimplantation. Wolf Reik (Babraham Institute, lar attention in the future, and as knowledge about com- Cambridge, UK) examined the possible relationship be- plex histone modification profiles emerges, it will be- tween DNA and histone methylation in preimplantation come crucial to develop reagents that more precisely de- embryos. K9-histone H3 methylation is asymmetric at fine the state of regulated chromatin in vivo. fertilization, and is immediately followed by DNA de- methylation of the paternal genome. Later, de novo at activated genes in vivo DNA methylation occurs, accompanied by global his- tone H3-K9 remethylation, specifically in inner-cell- An early step in gene activation involves -spe- mass cells of the blastocysts. Reik also addressed why cific recruitment of energy-dependent -re- nuclear transplantation is such an inefficient process in modeling and histone-modification complexes. Several animal cloning experiments. One explanation was sug- distinct complexes, defined by the identity of the gested by looking at fibroblast nuclei implanted into ATPase subunit, have been shown to disrupt nucleo- enucleated oocytes, where immunolocalization studies somal arrays and open chromatin structure. Thus, SWI2/ revealed that implanted nuclei do not recapitulate the SNF2or ISWI-containing complexes, as well as histone DNA and histone methylation patterns that are nor- acetyltransferases and are differen- mally observed during early mouse embryogenesis. tially recruited to specific subsets of genes through A key goal in the field is the identification of evolu- DNA-bound transcription factors. In this way, remodel- tionarily conserved histone-specific methyltransferases ing complexes recruited by one activator can facilitate (HMTs) that give rise to the epigenetic marks that im- binding and activation of other transcription factors that pact genome stability and plasticity. may be unable to recruit these enzymes. Similarly, en- (IMP, Vienna, Austria), examined knockout mice lacking hancer factors may directly recruit specific histone acet- both mammalian Su(var)3-9 HMTases (Suv39h1 and yltransferases, methyltransferases, and ubiquitin ligases Suv39h2) and showed that these mice showed reduced to modify local nucleosomes as well as the assembled viability and that the survivors acquire lymphomas at enhancer complex. Protein modifications of enhancer ∼30% penetrance. Remarkably, the constitutive hetero- complexes may control other steps in transcription, in- chromatin in cells lacking both Suv39h genes was no cluding elongation, reinitiation, and exchange of coacti- longer reactive to antisera recognizing histone H3 Lys 9 vators and . (H3-K9) methylation, whereas the inactive X-chromo- Several reports presented at the workshop examined

2040 GENES & DEVELOPMENT Downloaded from genesdev.cshlp.org on September 23, 2021 - Published by Cold Spring Harbor Laboratory Press

Chromatin ‘au naturale’ changes in nucleosome positioning and remodeling at amined how steroid hormone-inducible genes become highly induced genes in vivo, or effects of chromatin on refractory to further stimulation following induction. He the regulation of specific genes during development. showed that the SWI/SNF chromatin-remodeling com- A fundamental question concerns how different chro- plex is recruited by the glucocorticoid receptor (GR) in a matin remodelers participate in different biological pro- hormone-dependent manner via interactions with a cesses. Hua-Ying Fan from Bob Kingston’s lab (Harvard SWI/SNF subunit, termed Brahma-associated factor 60a Medical School, Boston, MA, USA), presented evidence (BAF60a). Further, expression of a mutant form of that two different ATP-dependent remodeling subunits, BAF60a binds the glucocorticoid receptor (GR) in a hor- BRG1 and SNF2h (members of the SWI/SNF and ISWI mone-dependent manner and inhibits the activity of families, respectively) produce distinctly remodeled in- BRG1-containing remodeling complexes. Thus, the termediates at different rates. The continuous intercon- hormone-activated GR may engage SWI/SNF through version of remodeled states indicates that a regulated BAF60a to promote transient remodeling. Archer’s lab equilibrium between change of nucleosome position and also investigated the role of phosphorylation rate of remodelling may contribute to the different roles in the refractory state, and showed that inhibition of of chromatin remodeling enzymes in vivo. Marco Bian- histone H1 kinase CD-K2reduces MMTV transcription chi (Department of Biological and Technological Re- in vivo. Using a mass spectrometric approach, they search, Milan, Italy) demonstrated that the chromatin- showed that prolonged treatment with dexamethasone associated HMGB1 protein facilitates nucleosome slid- led to dephosphorylation of a specific H1 isoform (H1.3) ing and remodeling by ISWI. The HMGB1-induced concomitant with repression. Through the use of photo- loosening, or bulging, of DNA around the nucleosome bleaching experiments, Gordon Hager (National Insti- could provide a linchpin between transcription factors tutes of Health, Bethesda, MD, USA) examined the be- and DNA. havior of tagged enhancer factors bound at the MMTV Wolfram Horz (University of Munich, ) re- , and discovered that most enhancer factors are ported that acetylation of histone H3 by the yeast SAGA rapidly displaced from MMTV DNA. Whereas DNA- complex accelerates the rate of activation of the PHO5 bound factors appear to dissociate within seconds, the promoter, but Gcn5 activity is not essential for its final stably recruited RNAPII remains associated far longer activation. The Pho4 activator accumulated normally in with the template. In vitro, Hager found that GR recruits nuclei in the absence of Gcn5 acetyltransferase activity. a SWI/SNF complex, but is actively displaced from the A similar delay in PHO5 activation and chromatin re- MMTV chromatin template during remodeling, simulta- modeling at the promoter is found in the absence of Snf2. neous with a transient burst of chromatin modifications, Pho4 recruitment to the promoter is delayed under these including H3 acetylation and methylation, during acti- conditions as shown by ChIP analysis. Thus, nucleo- vation. The dynamics of DNA repair in chromatin was some acetylation and remodeling function to reinforce a illustrated in a talk from Fritz Thoma (Swiss Federal In- rapid response to activation, but are not essential to stitute of Technology-Honggerberg, Zurich, Switzer- reach normal PHO5 transcription levels at much later land), who used photolyase as a molecular tool to com- stages in induction. Ernesto Di Mauro (University La pare the kinetics of repair (NER) of DNA lesions with the Sapienza, Rome, Italy) examined nucleosome destabili- slower nucleotide excision repair mechanism in yeast zation upon activation of the ADH2 gene in S. cerevi- and heterochromatin. The Thoma group siae. High-resolution analysis of nucleosome positioning find that active genes are more rapidly repaired than si- revealed that nucleosomes −1 and +1 surrounding the lenced genes, which is consistent with a more compact site of initiation change positioning by a few nucleotides chromatin structure of silenced genes. Interestingly both in the direction of transcription upon induction. These photolyase and NER appear to have open access to changes require the ADR1 activator and are also ob- nucleolar chromatin. served with mutants with impaired RNAP II catalytic The problem of how to establish and maintain pro- activity, indicating that nucleosome repositioning is not moter competence during development was addressed by a result of ongoing transcription. Ken Zaret (Fox Chase Center, Philadelphia, PA, Dramatic effects of nucleosome positioning are also USA). Two enhancer factors, GATA-4 and HNF3 (FoxA), observed upon induction of the MMTV promoter by ste- bind the albumin enhancer at 8.5–11.5 d of mouse devel- roid hormones. Miguel Beato (CRG/UPF, Barcelona, opment in the endoderm, prior to binding of other factors Spain) showed that the progesterone receptor recruits responsible for liver-specific albumin gene expression. ATP-dependent remodeling activities, which enable These two factors are sufficient to trigger an opening of binding of the enhancer factor NF1. NF1 binding stabi- the chromatin structure that supports the binding of lizes an open chromatin conformation that facilitates other regulatory factors, and are lost in cells that are not binding of additional receptors. In vitro minichromo- committed to differentiate to liver. HNF3 can interact some assembly experiments suggest that linker histone directly with the H3 and H4 core histones, and binding is H1 improves nucleosome positioning and enhances tran- correlated with opening of chromatin structure, al- scriptional synergism between hormone receptors and though not mediated directly through the histone tails. NF1, which is preceded by phosphorylation of histone Taken together, these data suggest that specialized tran- H1 and its depletion from the MMTV promoter. Trevor scription factors can create preset conditions that allow Archer (NIEHS, Research Triangle Park, NC, USA) ex- a gene to undergo subsequent rapid activation. The roles

GENES & DEVELOPMENT 2041 Downloaded from genesdev.cshlp.org on September 23, 2021 - Published by Cold Spring Harbor Laboratory Press

Orlando and Jones of different coactivators in the developmentally regu- sitioning of chromosomal domains within the inter- lated Wnt and Notch enhancer complexes was discussed phase nucleus facilitates transcriptional repression in by Kathy Jones (The Salk Institute, La Jolla, CA, USA). yeast. Analysis of GFP-tagged in inter- The Wnt-associated ␤-catenin strongly en- phase yeast cells by time-lapse microscopy revealed that hances binding of the LEF-1 HMG protein to chromatin internal tagged sites along yeast chromosomes are highly and supplies an activation domain sufficient to support mobile in G1 phase. This rapid chromosome movement transcription initiation on chromatin in vitro. In con- is ATP-dependent and linked to the metabolic state of trast, the Notch intracellular domain (ICD) bound to the the cell, and becomes constrained in . In contrast, CSL enhancer-binding protein was unable to support chromatin movement is constrained at and transcription in vitro in the absence of a third coactiva- centromeres in both G1 and S phases, when these ele- tor, Mastermind. Mastermind was also shown to couple ments are anchored near the nuclear envelope. The peri- activation with disassembly of the Notch enhancer com- nuclear attachment of telomeres requires the heterodi- plex and to facilitate phosphorylation in vivo of two as- meric Ku complex and an unidentified nuclear envelope sociated factors, the Notch ICD and CBP/p300, causing component. Importantly, a second, Sir-dependent, path- the latter to redistribute to large foci in cells expressing way seems to tether natural telomeres at the nuclear Mastermind. Emerging studies from other systems have periphery in S phase. Because previous work has shown suggested a key role for protein phosphorylation and that yeast cells must pass through S phase to establish ubiquitination in coupling of activation with disassem- heritable silent chromatin, this S-phase-dependent teth- bly of activator complexes, and can impact the timing ering correlates with subtelomeric repression and may and progression of transcription initiation and subse- provide a mechanistic link between repression and an- quent steps. chorage. Another interesting possibility suggested by these studies is that silent chromatin may play a direct role in the anchoring of telomeres. The organization and function of nuclear domains Peter Becker (Ludwig-Maximilians University, Mu- Several talks highlighted important advances in defining nich, Germany) discussed the targeting of the histone the effects of long-range regulatory elements and sub- acetyltransferase MOF to the male X-chromosome for nuclear positioning on the control of gene expression. dosage compensation in flies. Dosage compensation in Gary Felsenfeld (National Institutes of Health, Bethesda, flies involves the increase in expression of X-chromo- MD, USA) discussed the action of insulators, which can some genes in the male, as both sexes must express shield enhancers from the actions of external promoters equivalent levels of X-chromosome genes. Proper dosage and prevent condensed chromatin from spreading into compensation requires the epigenetic formation of stable active gene loci. Studies of the chick ␤-globin 5Ј insula- chromatin structure and the coating of the X-chromo- tor showed that a regulatory DNA-binding protein, some by an X-encoded, noncoding regulatory RNA. The CTCF, mediates enhancer blocking activity but does not hyperactive male X-chromosome was shown to be acety- convey protection against position effects resulting from lated at -K16 by the MOF histone acetyltrans- different sites of integration. Remarkably, CTCF also ferase, a member of the MYST family that is targeted to functions at the Igf2boundary element, and sites for this the X-chromosome as part of the dosage-compensation factor lie near Xist and may contribute to X-chromosome complex. This complex also includes the noncoding roX inactivation. Analysis of the modification state of RNA and several male-specific lethal proteins including nucleosomes within the insulated transgene suggested a the MLE and a histone H3 kinase. The Becker model in which insulators directly facilitate nucleosome lab is taking a biochemical approach to define how the acetylation, and that the resulting open chromatin struc- dosage-compensation complex is assembled and targeted ture supports the binding of factors that protect the gene to the X-chromosome, and have further elaborated on against DNA methylation. A new zinc finger protein, the roles of the MLE helicase, MOF, and MSL-3 in bind- termed BGP1, was shown to bind to G-string elements in ing to roX RNA. Although MOF contains a chromodo- the and may be involved in protecting against main related to those of other proteins that recognize position effects. Frank Grosveld (Erasmus University, methylated histones, this region of MOF has been impli- The Netherlands) discussed models for stochastic gene cated to play a role in binding to RNA. Becker reported activation, which have been invoked to explain monoal- on the role of a conserved C2HC zinc finger module for lelic gene expression patterns as well as other aspects of interaction of MOF with the nucleosomal substrate. In gene activation. Analysis of allelic transcription patterns collaboration with the Kuroda Laboratory (Houston, TX, and mRNA levels in single cells showed an imbalance of USA), the Becker Laboratory has mapped an X-chromo- ␣-globin versus ␤-globin gene expression, indicating that somal dosage-compensation complex entry site to a once a cell chooses a particular pattern of expression, the short DNA fragment. Using chromatin immunoprecipi- pattern becomes fixed in that cell. The findings empha- tation (X-ChIP) they are presently searching for addi- size the importance of stochastic events in the pheno- tional high-affinity sites for the complex. The goal is to typic differences that are observed in otherwise geneti- define the principles that allow the targeting and coating cally identical cells or organisms. of the X-chromosome by epigenetic regulators. Keiko Susan Gasser (University of Geneva, Switzerland) de- Ozato (National Institutes of Health, Bethesda, MD, scribed a series of elegant studies showing that the po- USA) discussed the regulation of the mammalian double

2042 GENES & DEVELOPMENT Downloaded from genesdev.cshlp.org on September 23, 2021 - Published by Cold Spring Harbor Laboratory Press

Chromatin ‘au naturale’ -containing BET gene family. The bromo- tin, David Gross (Louisiana State University Health Sci- domain is a chromatin-targeting module found in many ences Center, New Orleans, LA, USA) showed that in nuclear factors that mediates binding to acetylated his- the yeast S. cerevisiae silent, SIR-generated heterochro- tone H4. Histone deacetylase inhibitors were found to matin is permissive to constitutive-binding transcrip- stabilize the binding of BRD4 to acetylated chromatin, tion factors, TBP, and both hypo- and hyperphosphory- and fluorescence resonance energy transfer (FRET) ex- lated RNAPII. Moreover, the Sir2and Sir3 proteins re- periments with coexpressed tagged histone H4 revealed main bound to the promoter even under conditions of colocalization to noncentromeric chromatin on mitotic robust transcription. The heat-shock chromosomes. Ectopic expression of BRD4 blocked the HSF can override silencing in the absence of either the

G1/S transition, and BRD4 was shown to interact di- Sas3 or Gcn5 acetyltransferases, suggesting that histone rectly with DNA replication factor C (RFC140) in a man- acetylation is not a prerequisite step in release of hetero- ner that correlates with inhibition of DNA replication chromatic gene silencing. On the other hand, SWI/SNF and entry to S phase. These findings indicate that BRD4 is required, indicating a role for ATP-dependent chroma- may recruit the RFC to acetylated chromatin or affect tin remodeling. the timing of RFC activation during S phase.

Heterochromatin: open or closed? Perspectives Cell identity is largely determined by cell-specific tran- This conference focused on understanding how DNA- scription profiles. After determination, gene-specific si- dependent processing events function in the context of lencing is maintained and transmitted to daughter cells native chromatin structure and , by epigenetic mechanisms involving chromatin modifi- particularly as defined by physiological regulatory path- cation and subnuclear compartmentalization. In Dro- ways. Collectively, the talks highlighted chromatin as a sophila, the Polycomb group proteins (PcG) control cell programmable substrate that regulates the assembly of identity by maintaining homeotic and other selector molecular engines, depending on its local modification genes in a repressed state. PcG proteins act as multipro- profile. A continued theme is the importance of indi- tein complexes that act via specialized DNA elements vidual histones and specific covalent modifications in called Polycomb response elements (PREs). PREs are the formation and stability of these protein–protein in- thought to maintain silencing by driving the building-up teractions. Because nucleosomes are major targets of cell of heterochromatin-like structure. Details of the mecha- signaling, gene-specific histone modification pathways nism by which PcG maintains silencing are largely un- will have to be precisely identified and deciphered. Re- known. cent technical developments, such as chromatin immu- Valerio Orlando (Telethon and Institute of Genetics noprecipitation and high resolution in vivo genomic and Biophysics, National Research Council, Naples, analysis, have opened new possibilities for addressing Italy) examined in vivo the role of PcG at promoters and the role of chromatin structure and its modifications. reported that PcG-repressed chromatin does not exclude Other key parameters are the dynamics of nucleosome general transcription factors, but, rather, contains a re- positioning and the importance of higher-order chroma- activatable RNAPII complex. The bithorax complex re- tin structure in regulating gene loci. In some systems, pressed promoters look similar to active promoters in the packaging of DNA into chromatin determines which that they contain K4 (not K9) methylated H3, hyperphos- regulatory proteins interact with their cognate elements phorylated RNAPII, and elongation factors. Further- and, thus, the outcome of the transcriptional response. It more, the PcG proteins were shown to coimmunopre- now appears that many genes are preset for activation, cipitate with elongation factors and cyclin T, suggesting and that transcriptional activators as well as general that the block to transcription lies downstream of phos- transcription factors may reside on the DNA and become phorylation of the RNAPII CTD. In Drosophila, at least activated only later, for example, in response to specific two major PcG protein complexes have been character- signaling pathways. Tissue-specific promoters appear to ized; an early complex (ESC–EZ complex), which is part be preset by transcription factors, giving genes the of a determination cascade that establishes the homeotic chance to be switched according to developmental pro- gene transcription state, and a later (PRC1) complex, grams. As many of these factors regulate more than one which acts as the system. An unsolved key is- gene, one can envisage that pathway-specific transcrip- sue is how the memory system reads the promoter state tion complexes sit on the chromosome and can be coor- in early embryogenesis, and then sets up the memory dinately switched. Genome-wide identification of target lock. Vincenzo Pirrotta (University of Geneva, Switzer- genes by ChIP will provide global views of these dynam- land) showed that the ESC–EZ complex contains histone ics. This is true also for putatively heterochromatinized methyltransferase activity and that a burst of histone promoters regulated by the SIR and Polycomb proteins, acetylation (provided by transient GCN5 expression) in and indicate that silenced, developmentally regulated early development impairs methylation and mainte- genes may retain the ability to be reactivated at all nance in a PRE-reporter construct. Also in this case, a times, which has intriguing implications for cell repro- stalled RNAPII was reported to sit at repressed promot- gramming. Beyond the widespread role of histone modi- ers. In the context of gene silencing and heterochroma- fication, posttranslational modifications of enhancer fac-

GENES & DEVELOPMENT 2043 Downloaded from genesdev.cshlp.org on September 23, 2021 - Published by Cold Spring Harbor Laboratory Press

Orlando and Jones tor complexes can impact the activity, localization, and timing of transcription for tightly regulated genes. As always, continued progress relies on novel specific technical advances and overcoming existing technical limitations. For example, it is becoming clear that op- posing forces trafficking on chromatin represent the sum of events and interactions that may take only a few sec- onds, whereas conventional ChIP protocols fix cells for minutes. Thus, time is clearly a fundamental parameter in gene regulation, especially in light of stochastic events. Similarly, it is necessary to define where deter- mined events take place within the nucleus. An accurate understanding of all dimensions of native chromatin also directly impacts on efforts to model regulation through biochemical reconstitution studies. The continued ap- plication of novel approaches and concepts such as those described at the meeting, and the further evolution of new techniques such as laser cross-linking and live cell imaging, will, therefore, make chromatin biologists hap- pier in their natural context.

Acknowledgments The Fundación Juan March provided an interactive setting con- ducive for thoughtful talks and discussions. We are grateful to Ken Zaret and Miguel Beato for conceiving and hosting this stimulating meeting, and to Andrés Gonzáles and Lucia Franco for their outstanding management of the workshop as well as to all of our colleagues who participated in Madrid.

2044 GENES & DEVELOPMENT Downloaded from genesdev.cshlp.org on September 23, 2021 - Published by Cold Spring Harbor Laboratory Press

Wild chromatin: regulation of eukaryotic genes in their natural chromatin context

Valerio Orlando and Katherine A. Jones

Genes Dev. 2002, 16: Access the most recent version at doi:10.1101/gad.1017402

License

Email Alerting Receive free email alerts when new articles cite this article - sign up in the box at the top Service right corner of the article or click here.

Cold Spring Harbor Laboratory Press