Chromatin Clubs on the East Coast, USA

1. New York Chairman: Dr. Gregory David (NYU Dept of Pharmacology)

"I was my great pleasure to host the first New Chromatin Club in New York on October 20th, 2004 at the New York Academy of Medicine. For this first meeting, we decided to focus on new roles for chromatin modifiers in development and oncogenesis. Indeed, thanks to the six great presentations delivered at this meeting, it became even more obvious to all of us attendees that tight regulation of chromatin modification is an essential prerequisite for numerous biological processes, ranging from very early development to host defense against pathogens.

"The first speaker was Kristen Johnson [Kathryn Calame’s lab at Columbia University] presenting on the influence of chromatin modifications on V(D)J recombination, and using chromatin immunopreicipitation assays (ChIP). She very elegantly demonstrated that histone methylation on Lysine 9 on Histone H3 inhibits this recombination in vitro. Kristen’s studies established that histone replacement by the variant H3.3, driven by the B cell factor Pax5, allows the removal of the inhibitory histone mark and allows the V to DJ recombination process to happen.

"Yuhong Fan [Art Skoultchi’s lab at Albert Einstein College of Medicine] presented surprising results obtained after generating mouse strains deleted for several histone H1 genes, know as the linker histones. Even though single or even double knock-out mice for those H1 genes did not exhibit obvious phenotypes, Yuhong elegantly demonstrated the necessity of a threshold level of histone H1 proteins to achieve early embryonic development. Using ES cells, she investigated the transcriptional consequences of histone H1 depletion and suggested a surprising and fascinating connection between the level of histone H1 and the transcription repression driven by DNA methylation in vivo.

"The role of the gonad specific protein DNMT3L in gametogenesis was elucidated by Deborah Bourc¹his [Tim Bestor¹s lab at Columbia University]. Deborah generated a mouse strain genetically deleted for DNMT3L. DNMT3L deficiency results in azoospermia, correlated with striking defects in homologous chromosome synapsis at meiosis along with lack of methylation and reactivation of retrotransposable elements. These data suggest a mechanism for DNMT3L which, despite the lack of catalytic methylation activity, would be necessary for the establishment of long-term silencing of specific regions of the genome. This function is not only required for protecting the germ line against parasitic infiltration but also indirectly for proper homology search at meiosis.

"Naoko Tanese [NYU School of Medicine] presented recent data suggesting a role for human Osa2, a component of the Swi/Snf chromatin remodeling complex, in the response to DNA damage in vivo. Upon Osa2 overexpression, the p53 tumor suppressor becomes activated, probably through the ATr kinase. Thus, Naoko elucidated an unsuspected function for a component of a chromatin-modifying complex in the signaling pathway, from sensing DNA damage to cell-cycle arrest.

"Scott Coonrod [Cornell University] studies how the global chromatin modifications, and more specifically histone marks, evolve during fertilization and early embryonic fertilization. State-of-the-art imaging techniques clearly show that histone modifications are highly dynamic during early development. Moreover, proteomics analysis of the mouse egg revealed the presence of a very abundant new protein, ePAD, which was later shown to support a previously unknown enzymatic activity: the removal of methyaltion marks on histones. Identification of ePAD and the related PAD proteins revealed the molecular basis for a long-sought chromatin modification, the removal of lysine marks from histones.

"Angus Wilson [NYU School of Medicine] concluded the meeting with a presentation of his recent work on HCF1, a cellular protein that is targeted by the Herpes Simplex activator VP16 for expression of immediate early genes during early stages of viral infection. A new player in this host-virus reprogramming of transcription, the bromo-domain protein 7 (Brd7), was described. Indeed, Brd7 interacts with HCF1 and is able to modulate nuclear substructures in a HCF1 dependent manner, suggesting a role for histone acetylation in HCF1 nuclear functions in response to DNA strand break.

"In conclusion, it was a great meeting, and we all enjoyed the presentations and discussions that followed, as evidenced by positive feedback from numerous participants. The challenges and surprises that await anyone interested in chromatin as it relates to biological processes justify all the more such forums, and we are confident that many meetings as successful and fruitful as this first one will follow in the near future."

Speakers' abstracts from New York Chromatin Club meeting

2. Washington DC Chairman: Dr. David Clark (LMGR, NICHD, NIH)

"We held the inaugural meeting of the Washington Area Chromatin Meeting on 18th October, 2004, at the National Institutes of Health in Bethesda, Maryland. We were treated to an excellent series of talks from speakers working at NIH and at other local institutions, culminating in a seminar by our invited speaker, Professor Mitch Smith of the University of Virginia.

"Keiko Ozato [National Institute of Child Health and Human Development] began the Meeting by describing her fascinating studies of the mammalian bromodomain proteins, Brd2 and Brd4. Brd4 was shown to persist on chromosomes even during mitosis, when most proteins dissociate from chromatin. Consequently, Keiko suggested that Brd4 might play a role in epigenetic memory.

"The next speaker was Cathy Smith [National Cancer Institute], who described her lab's latest exciting results concerning the cyclic AMP-dependent regulation of histone dephosphorylation. She discussed the phosphatases responsible for this dephosphorylation and the implications for signalling-dependent gene regulation.

"Abou Elkharroubi [Johns Hopkins University] told us about the work in his lab aimed at addressing the role of epigenetic modifications in silencing imprinted genes. He demonstrated very clearly that DNA methylation and histone deacetylation are associated with some, but not all, imprinted genes. His work leads to the very important and interesting proposal that other (as yet unidentified) epigenetic modifications might also be involved in the silencing of imprinted genes.

"Keji Zhao [National Heart, Lung and Blood Institute] gave us a thought-provoking glimpse into a future in which we will be able to map histone modifications on a genome-wide scale. Using a novel technique developed in his lab, "GMAT", the distribution of acetylated histones with respect to the human genome was mapped in T-cells and the changes in acetylation that occur in response to T-cell activation were determined. He identified 47,000 "acetylation islands", about10% of which alter their status in response to T-cell activation.

"Michael Bustin [National Cancer Institute] discussed his latest insights into the roles of the nucleosome-specific binding proteins, HMG-N1/N2, in chromosome structure and gene regulation. The patterns of histone modifications are altered in HMG-N1 knock-out mice. In particular, phosphorylation of HMG-N1 is correlated with changes in H3 phosphorylation and acetylation. Michael is building a model that integrates the roles of histone H1, HMG-N1 and core histone modifications in gene regulation.

"Finally, our invited speaker, Mitch Smith [University of Virginia], presented a riveting seminar focused on yeast histone H2AZ, a variant of histone H2A found in all eukaryotes. He referred to H2AZ as a "deviant histone" because its genetic and structural properties are quite different from those of "normal" H2A, although the epithet did seem a little unkind and generated some sympathy for H2AZ in the audience! Mitch described an elegant screen for isolating mutated genes that are synthetically lethal with a knock-out of HTZ1 (the gene encoding H2AZ). Surprisingly, his lab isolated mutants in subunits of RNA polymerase II, indicating that there is a genetic interaction between this polymerase and H2AZ. He presented additional data suggesting that H2AZ affects elongation as well as induction of transcription. It is now clear that the role of H2AZ in chromatin is multi-faceted.

"We all had the opportunity to discuss the talks and chromatin in general at the buffet and a pleasant time was had by all. I would like to take the opportunity to thank Rhian Hayward [Abcam] for her invaluable help in organising this Meeting."

Speakers' abstracts from the Washington DC Chromatin Interest Group meeting.

3. Boston Area Chairman: Dr Kami Ahmad (Department of Biological Chemistry & Molecular Pharmacology, Harvard Medical School)

"Given the large number of chromosome biology researchers concentrated in the Boston area, there has been great enthusiasm to hear from the community in a local setting. The first meeting was a smorgasbord of research organisms and topics in the community. The roles of unusual RNAs were a thread running through many of the talks.

"Fred Winston [Genetics, Harvard Medical School] described the genetics of a novel mode of transcriptional repression in budding yeast in which intergenic transcription across the promoter of the SER3 gene blocks expression.

"Bob Kingston [Molecular Biology, Massachusetts General Hospital] discussed the biochemistry of protein complexes that repress or activate transcription from chromatin templates. Studies of PRC complexes show that they bind directly to nucleosome particles and can link chromatin arrays together.

"Jeannie Lee [Molecular Biology, Massachusetts General Hospital] described experiments designed to tease apart the relationships between non-coding RNAs involved in mammalian X chromosome dosage compensation. Gene-replacement studies have tested whether the anti-sense Tsix transcript must base-pair with Xist or if the act of transcription itself is required.

"Mitzi Kuroda [Genetics, Harvard Medical School] introduced a different solution to the problem of dosage compensation. Dosage compensation in fruitfly males doubles expression from their one X chromosome. Again, non-coding RNAs are involved. The two roX RNAs typically originate from X-linked loci, but experiments show that roX RNAs made on autosomes can seek out 'entry sites' on the X chromosome and then ooze across it. The mechanism of spreading remains mysterious.

"James Sherley [Biological Engineering, MIT] revealed to us that there remain surprises in even the most basic chromosome processes. DNA labeling experiments show that stem cells distinguish between template and newly-replicated strands, and that they retain the old strand preferentially in the stem mother cell. This contributes to the low rate of mutation in stem cells, but implies that the division machinery of the cell is asymmetric and connected in some way to the old strand in during replication.

"The new Chromatin Club will continue in two formats, a monthly short meeting for data discussion and, twice a year, a longer session (organised with Abcam as reviewed here) with research talks from postdoctoral fellows and faculty speakers. We're off to a great start!"

Speakers' abstracts from the Boston Area Chromatin Club meeting

4. Abstracts from the New York Chromatin Club meeting October 2004

Abstract 1: Breaking the Silence of the Immunoglobulin Heavy Chain Locus

Kristen Johnson, David Pflugh, Duonan Yu, David Hesslein, Kuo-I Lin, Alfred Bothwell, Andrei Thoms-Tikhonenko, David Schatz and Kathryn Calame

Gene activation is a vital process that results in cell specification and function. Spurious gene expression is prevented by the fact that DNA is packaged within chromatin, a generally inhibitory structure. This inhibition creates a requirement for lineage specific regulatory factors that are able to act on silent chromatin and in turn make it accessible for enzymes involved in processes such as transcription, repair and recombination which use chromatin as a substrate. The molecular mechanisms by which the initial activation occurs are still being unravelled.

B cell development is defined by the expression of several B cell specific surface markers, the expression of B cell specific genes, as well as successive steps in the rearrangement and expression of the antigen receptor. V(D)J recombination, the process responsible for antigen receptor gene rearrangement, allows individual gene segments which are separated in the germline to be brought together in random combinations resulting in the creation of a diverse antigen receptor repertoire. Activation of this process is unique in several aspects including the strict lineage and developmental regulation to which it is subject, as well as the size of DNA loci on which it acts.

Using Chromatin Immunoprecipitation assays (ChIP), we find methylation of lysine 9 on histone H3, a mark of inactive chromatin which can inhibit V(D)J recombination on an engineered substrate, is uniquely associated with the 2 MB V_H locus in cells outside of the B cell lineage. The B cell commitment factor, Pax5, is both necessary and sufficient for the removal of V_HH3K9 methylation. Concomitant accumulation of the histone variant H3.3, suggests that replication-independent histone exchange may be the mechanism by which this inhibitory modification is removed. These data provide evidence for a mechanism by which a Pax5 promotes V_H gene accessibility, promoting lineage specific V_H-to-DJ_H recombination and allowing further B cell development.

Abstract 2: Transposon reactivation and meiotic catastrophe in Dnmt3L-deficient male germ cells

Déborah Bourc’his, Department of Genetics and Development, Columbia University, New York.

Dnmt3L is a catalytically inactive member of the DNA methyltransferase family that is specifically expressed in germ cells at the time genomic methylation patterns are established. Dnmt3L is required in growing oocytes for setting up maternal methylation imprints. Unlike the situation in females where Dnmt3L-mutant oocytes are normally produced, Dnmt3L deficiency results in azoospermia in males. Dnmt3L is expressed during a short period of spermatogenesis, in perinatal prospermatogonia that are the precursors of spermatogonial stem cells and therefore of all the subsequent differentiating germ cells. Dnmt3L deprivation in prospermatogonia interferes with homologous chromosome pairing in spermatocytes at meiosis and triggers a developmental interruption at pachytene. This meiotic failure is associated with a demethylation and derepression of LTR and non-LTR retrotransposable elements. Pericentromeric tandem repeats and paternal imprinted genes are respectively non- and limitedly affected. This specific effect suggests that Dnmt3L might be involved in tagging dispersed repeated sequences in a pre-meiotic scan that occurs around the time of birth and establishes life-long transcriptional repression. We will discuss the sexual dimorphism of Dnmt3L expression and function in imprinting, transposon control and meiosis.

Abstract 3: A role for the largest human SWI/SNF complex subunit in the activation of the DNA-damage signaling pathway

Stavros Giannakopoulos, Hiroko Inoue, Takako Furukawa, Naoko Tanese, New York University School of Medicine, New York, NY 10016

Eukaryotic cells are constantly challenged by genotoxic stress, and efficient detection of damaged DNA is essential to maintaining the integrity of the genome and survival of the organism. Recent studies suggest that the transcriptional machinery designed to function in the context of chromatin might also play a role in the recognition of damaged DNA and activation of the response pathway. hOsa2/BAF250b is an integral component of the human SWI/SNF chromatin remodeling complex. We established inducible HeLa cell lines over-expressing FLAG-tagged hOsa2 to facilitate purification of the hOsa2 complex. Mass spectrometric analysis of the purified hOsa2-associated proteins identified peptides representing BRG1, but not hBRM or hOsa1, suggesting that hOsa1 and hOsa2 form distinct complexes. Furthermore, we found induced over-expression of hOsa2 to cause growth arrest and accumulation of cells in G1. The p53 protein level, but not the mRNA, was increased and the accumulated p53 was found phosphorylated at serine 15. The finding led us to hypothesize that hOsa2 expression might mimic the events that accompany cellular response to DNA damage. Because hOsa2 is a chromatin-associated protein, we think it might function to trigger a pathway leading to the activation of ATM/ATR. Treatment of these cells with caffeine, an inhibitor of the DNA damage checkpoints, but not wortmanin, an inhibitor of ATM, blocked serine 15 phosphorylation of p53 upon induction of hOsa2, suggesting a role for ATR in this process. Knockdown of ATR by siRNA reduced p53 phosphorylation upon induced expression of hOsa2. Immunoflourescence studies confirmed accumulation of p53 phosphorylated at serine15 in cells co-expressing hOsa2 and p53. Additionally, siRNA-mediated knockdown of endogenous hOsa2 reduced accumulation of p53 in cells treated with ionizing radiation, suggesting that hOsa2 might be involved in the activation of the DNA damage response pathway.

Abstract 4: Human PAD4 regulates histone arginine methylation levels via demethylimination

Scott Coonrod, Department of Genetic Medicine, Weill Medical College of Cornell University, Room 405 Whitney, 1300 York Ave. New York, NY 10021.

Methylation of arginine (Arg) and lysine (Lys) residues in histones has been correlated with epigenetic forms of gene regulation. Although histone methyltransferases are known, enzymes that demethylate histones have not been identified. Here, we demonstrate that human peptidylarginine deiminase 4 (PAD4) regulates histone Arg methylation levels by converting methyl-Arg to citrulline (Cit) and releasing methylamine. PAD4 targets multiple sites in histones H3 and H4, including those sites methylated by co-activators CARM1 (H3 Arg17) and PRMT1 (H4 Arg3). A decrease of histone Arg methylation, with a concomitant increase of citrullination, requires PAD4 activity in human HL-60 granulocytes. Moreover, PAD4 activity is linked with the transcriptional regulation of estrogen responsive genes in MCF-7 cells. These data suggest that PAD4 mediates gene expression by regulating Arg methylation and citrullination levels in histones.

Abstract 5: Nucleolar disruption by bromodomain protein Brd7 is dependent on HCF-1 and p53.

Randy Luciano, Søren Ottosen, Shahana S. Mahajan, and Angus Wilson, Department of Microbiology and NYU Cancer Institute, New York University School of Medicine, 550 First Avenue, New York, NY 10016 USA

Besides ribosome production, the nucleolus plays a critical role in the cellular response to environmental stress. Exposure to DNA damaging agents and other insults induces major changes in the nucleolus, termed nucleolar disruption (ND), resulting in reduced rRNA gene transcription and p53 stabilization. Here we show that expression of bromodomain protein 7 (Brd7, BP75, Celtix1) brings about ND in the absence of DNA damage. Brd7 is a chromatin-associated protein that interacts with the C-terminal subunit of host cell factor-1 (HCF-1). ND induced by Brd7 requires the acetyl-lysine binding bromodomain and leads to a striking accumulation of HCF-1 in subnucleolar granules at the centers of the restructured nucleoli. A similar relocalization of HCF-1 occurs when cells are exposed to radiation or drugs that induce double strand breaks. ND induced by Brd7 or DNA damage requires p53 expression and can be blocked by short-interfering RNAs against either Brd7 or HCF-1. These results identify two chromatin-associated factors Brd7 and HCF-1 as components of the cellular machinery that maintains genomic integrity.

5. Abstracts from the Washington DC Area Chromatin Club meeting October 2004

Abstract 1: Double bromodomain protein Brd4: a role in transcriptional memory?

Moon Kyoo Jang, Anup Dey, Tomohiko Kanno, Yuka Kanno, Kazuki Mochizuki and Keiko Ozato

Laboratory of Molecular Growth Regulation, National Institute of Child Health and Human Development, National Institutes of Health, Bethesda MD 20892

Brd4 is a mammalian member of the conserved BET family and carries two bromodomains. The most salient feature of Brd4 is its association with chromosomes during mitosis, a feature shared by another BET family protein Brd2. In contrast to the Brd proteins, other chromatin modifying proteins and transcription factors dissociate from chromatin during mitosis. This dissociation event coincides with chromatin hypoacetylation and the general shut down of transcription during mitosis. Based on their behavior, the Brd proteins are thought to have a potential role in epigenetic memory. The significance of Brd4-chromosome association is further reinforced by our recent observation that Brd4 interacts with the papillomavirus transactivator E2 and tethers viral DNA to cellular chromosomes, thereby ensuring equal segregation of viral genomes into newly divided cells (1).

To study a real time interaction of BET proteins with chromatin, we employed live cell technologies including fluorescence resonance energy transfer and bi-fluorescence complementation (2,3). We found that Brd2 and Brd4 bind to acetylated lysine residues of core histones with exquisite specificity, indicating that the Brd proteins recognize acetylated histone codes in living cells. Photobleaching experiments showed that Brd2 and Brd4 are mobile and transiently interact with chromatin, revealing a highly dynamic nature of Brd-chromatin interactions. Furthermore, Brd-histone interactions persisted during mitosis, indicating that the Brd proteins maintain their histone-code recognition across cell division.

By immunopurification and mass-spectrometry analysis we identified cyclinT/cdk9 (P-TEFb) as a complex that interacts with Brd4 (4). P-TEFb is a kinase that phosphorylates the CTD of RNA polymerase II and is required for transcriptional elongation. Our data show that Brd4 positively regulates cdk 9 kinase activity and enhances the activity of many cellular promoters, thus establishing a role for Brd4 in transcription. Together, Brd4 forms a dynamic molecular network that links acetylated chromatin and transcription. These data may further support a role for Brd4 in the transmission of transcriptional memory from one generation of cells to the next.

(1) You, J et al, Cell. 117: 349, 2004.

(2) Kanno, T et al, Mol. Cell. 13:33, 2004

(3) Dey, A et al., Proc Nat Aced Sci USA. 100: 349, 2003

(4) Jang, M et al, submitted.

Abstract 2: Regulation of Histone H3 phosphorylation by a novel cAMP signaling pathway

Pedro Rodriguez, Sara K. Snyder, Rebecca C. Chiffer, and Catharine L. Smith, Signal Transduction Group, LRBGE, CCR, NCI.

Signaling via cAMP as a second messenger is used by a wide variety of cellular effectors. Traditionally it was thought that cAMP signaling occurs primarily through the activation of protein kinase A. However, the recent discovery of cAMP-regulated guanine nucleotide exchange factors (cAMP-GEFs), such as Epac, has opened up new mechanisms for cyclic nucleotide signaling. We discovered that in a variety of cycling cells, activation of cAMP signaling induces a rapid dephosphorylation of cellular histone H3 as determined by both metabolic labeling and immunological methods. This effect appears to be specific to phosphorylation of histone H3 since phosphorylation of other histones is not significantly changed and other H3 modifications are not effected. An antagonist of PKA, Rp-cAMPs, does not block H3 dephosphorylation induced by 8-Br-cAMP, indicating that this effect is not mediated through the traditional PKA pathway. However, an agonist specific for Epac does not induce H3 dephosphorylation either. Consistent with these results, H3 dephosphorylation can be induced by much lower concentrations of 8-Br-cAMP than are required for activationof PKA. Others have shown that cAMP signaling can activate MAP kinase pathways by both PKA-dependent and -independent mechanisms. However, inhibitors of both ERK and p38 signaling do not block cAMP-induced dephosphorylation of histone H3. However, the phosphatase inhibitor, okadaic acid, prevents cAMP-induced H3 dephosphorylation at high concentrations, indicating the involvement of a phosphatase. Ceramide, a second messenger known to activate PP1 and PP2A also induces H3 dephosphorylation, suggesting that activation of one or both of these phosphatases may be induced through cAMP signaling. Our results lead us to conclude that the dephosphorylation of histone H3 by cAMP signaling is mediated through a unique cyclic nucleotide binding protein which does not work through MAP kinase signaling pathways and leads to activation of phosphatases.

Abstract 3: Epigenetic modifications associated with silencing of imprinted genes in mammals

Abou Elkharroubi, Advanced Academic Programs in Biotechnology and Bioinformatics, Johns Hopkins University, Maryland

Genomic imprinting results in differential expression of the paternal and maternal alleles of a subset of genes required for wide range of developmental processes. The deregulation of imprinting is associated with abnormal embryogenesis, neonatal growth defects and tumorigenesis. A key question about imprinting is how the paternal and maternal alleles are distinguished in the same nucleus leading to differential allele silencing? Epigenetic modifications such as DNA methylation, histone deacetylation, and binding of protein complexes associated with gene silencing to a dozen imprinted alleles were examined. The results clearly demonstrate that DNA methylation and histone deacetylation are associated with silencing at some but not all imprinted loci, suggesting that other uncharacterized chromatin modifications are involved in the control of imprinted genes during mammalian development. The nature and distribution of some of these chromatin modifications along imprinted loci will be discussed.

Abstract 4: Genome-wide mapping of histone modification by GMAT

Keji Zhao, Laboratory of Molecular Immunology, NHLBI, NIH, Bethesda, MD 20892

The eukaryotic epigenome is mainly defined by modifications of chromatin including posttranslational modification of histones. The current “genome-wide” methods for detecting histone modifications are based on DNA microarrays, which contain a fraction of coding sequences and intergenic regions. Truly genome-wide analysis of histone modifications requires DNA arrays containing all the genomic sequences, which are not available for higher eukaryotic genomes. We developed a non-biased genome-wide mapping technique (GMAT) by combining chromatin immunoprecipitation and SAGE to analyze histone modifications of the entire genome including both the coding and intergenic regions. To validate this method, we determined the genome-wide tetra-acetylation of histone H4 tail and K9/K14 acetylation of histone H3 in Saccharomyces Cerevisiae. The GMAT analysis demonstrated the typical pattern of hyperacetylation in active genes and hypoacetylation in heterochromatin. The data showed that the activity of telomeric repression is inversely correlated with the histone acetylation. Our data revealed that inducible genes are marked by H4 acetylation in their promoter and coding region under non-inducible conditions. In contrast to the current belief that the promoter is the most acetylated region within a gene, our analysis of 6000 genes indicates that acetylation peaks in the region from transcription start site to about 400bp downstream. Significant levels of histone acetylation were detected throughout the coding sequences. Using this technique, we studied the genome-wide changes of histones H3 and H4 acetylation caused by mutations in GCN5. Consistent with the notion that GCN5 is a regulator of global histone H3 acetylation, we found that the CGN5 deletion inhibited the peak acetylation of histone H3 in the promoter and 5’ coding regions, while it did not significantly inhibit the acetylation pattern of histone H4. GMAT should find valuable applications in deciphering the histone code of the eukaryotic genomes and identifying genome-wide target sites for chromatin-modifying enzymes and transcription factors.

Abstract 5: A dynamic interplay between nucleosomal binding proteins modulates the levels of histone modification in chromatin.

Michael Bustin, National Cancer Institute, NIH, Bethesda, MD 20892

Competition between the linker histone H1 and HMGs (see MCB: Network of dynamic interactions between histone H1 and high-mobility-group proteins in chromatin. Mol Cell Biol. 2004 May;24(10):4321-8) changes the levels of posttranslational modifications (see Mol. Cell: Chromosomal Protein HMGN1 Modulates Histone H3 Phosphorylation. Mol Cell. 2004 Aug 27;15(4):573-84.) by affecting the accessibility of histone modifiers to their nucleosomal targets.

Abstract 6: Histone variants and gene regulation

M. Mitchell Smith, Maria S. Santisteban, Anne B. Allison, Kurt Jensen, and Jinmei Li., Department of Microbiology and University of Virginia Cancer Center, University of Virginia, Charlottesville, VA 22908

Nucleosome arrays present a formidable challenge to gene regulation, in vitro and in vivo, and can exert both positive and negative control over a wide variety of genes. Thus, cells have evolved a number of mechanisms for modulating chromatin structure during transcription. Two well known mechanisms for chromatin modulation include nucleosome remodeling and covalent histone modifications. A third mechanism involves the incorporation of histone variants, such as H2A.Z, into chromatin.

In order to better understand the role of H2A.Z in transcription, we carried out a synthetic lethal genetic screen to identify mutations that are deleterious when specifically combined with the htz1 null mutation. Among the mutants identified in this screen was a class of dominant synthetic lethals. We cloned the gene for one of these candidates by complementing its linked temperature sensitive phenotype and identified the mutant allele as rpb2-2, a mutation in the second largest subunit of RNA polymerase II. This result shows that RNA polymerase II itself interacts genetically with chromatin. Consistent with this interpretation, we find that htz1 rpb4 and htz1 rpb9 double mutants are also synthetic lethal.

The dominant synthetic lethality of rpb2-2 suggests that H2A.Z functions in transcription elongation by pol II; pol II containing Rpb2-2 may arrest on chromatin lacking H2A.Z, blocking transcription by wild type polymerase and thus causing dominant lethality. This model makes at least three predictions: (1) htz1 should have elongation phenotypes; (2) htz1 should show genetic interactions with genes known to function in elongation; and (3) there should be distinct molecular defects in the elongation complexes in htz1 mutants. The results of experiments to address these three predictions will be presented.

6. Abstracts from the Boston Area Chromatin Club meeting October 2004

Abstract 1: Control of the S. cerevisiae SER3 gene by intergenic transcription

Joseph A. Martens, Lisa Laprade, Pei-Yun Jenny Wu, and Fred Winston, Department of Genetics, Harvard Medical School

Eukaryotic genomes are highly transcribed, including large amounts of transcription of non-coding regions. Our studies of the regulation of the S. cerevisiae SER3 gene have shown that SER3 transcription is regulated by an intergenic transcript. SER3 is tightly repressed during growth in rich medium. Under these conditions, the SER3 regulatory region is highly transcribed, producing a non-protein-coding RNA (SRG1). Transcription of SRG1 is required for repression of SER3. Repression occurs by a transcription interference mechanism in which SRG1 transcription across the SER3 promoter interferes with the binding of activators. More recent studies have shown that SRG1 transcription is regulated in response to serine levels. Taken together, these results have identified a previously unknown class of regulatory factor and have elucidated one role for intergenic transcription.

Abstract 2: Biochemical analysis of complexes involved in epigenetic regulation

Bob Kingston, Nicole Francis, Ian King, Hua-Ying Fan, Richard Emmons, Ting Wu, Chris Woodcock., Department of Molecular Biology, Massachusetts General Hospital.

The Polycomb-group (PcG) of complexes is required to maintain repression of master regulatory genes such as HOX genes throughout development. PcG action is targeted by PRE elements, which contain binding sites for sequence-specific factors such as GAGA, Zeste, PHO, Hb and Psq. The repression effect spreads away from PRE sequences; it has been proposed that spread might occur either linearly (‘oozing’) or non-linearly (‘bridging’ of two sequences). The PcG complex PRC is conserved from Drosophila to mammals. We have made a recombinant mammalian PRC complex which strongly represses transcription. If this complex is pre-bound to one nucleosomal array, it is capable of recruiting a second nucleosomal array from solution, thus bridging the two arrays. The recruited array becomes repressed for transcription and for chromatin remodeling with a time-course that is identical to the time course of recruitment. This provides evidence that this PcG complex is capable of linking two separate regions of the genome via a non-linear mechanism, and that this might be the way the complex spreads repression. To examine the types of repressive structures that PcG complexes form, we have visualized them using electron microscopy. We observe compacted structures that require nucleosomes, but not histone tails, to form. The PSC protein is central to the ability to form this repressive structure, and mutations in PSC that disrupt function in vivo also disrupt formation of a compacted structure. We propose that this compacted structure causes repression of remodeling and transcription.

Abstract 3: Targeting of dosage compensation in Drosophila

Artyom Alekseyenko, Xiaoying Bai, and Mitzi I. Kuroda, Howard Hughes Medical Institute, Center for Genetics & Genomics, Brigham & Women’s Hospital, Dept. of Genetics, Harvard Medical School

MSL proteins and noncoding roX RNAs form complexes that bind hundreds of sites along the single male X chromosome in Drosophila to make X linked gene expression equal in males and females. MSL complexes are proposed to up-regulate X-linked genes in males through site-specific acetylation of histone H4 at Lys16. The roX1 and roX2 noncoding RNAs are dissimilar in size and sequence, but are functionally redundant. The roX RNAs are required for normal targeting of the X chromosome, and in their absence MSL complexes bind poorly to the X.

We currently have evidence for two distinct mechanisms for X chromosome targeting of MSL complexes: spreading from roX genes in cis, and recognition of unknown X determinants in trans. To analyze how MSL complexes become established on the X, we conditionally expressed MSL2 late in development in females or msl2 mutant males. Although MSL complexes normally are established on the X during embryonic development, we found that X chromosome binding can be initiated in late larval stages. With short periods of induction, MSL complexes are concentrated around the roX2 locus in the middle of the polytene X chromosome, while longer induction times result in a wild type pattern along the length of the X. These results suggest that spreading from roX2 is an early event in establishment of MSL complexes on the X chromosome. Spreading from roX1 was not observed in these experiments, consistent with a failure to induce roX1 RNA expression in most nuclei.

Abstract 4: Mechanisms for non-random chromosome co-segregation associated with adult stem cell asymmetric self-renewal

James L. Sherley, Biological Engineering Department, Massachusetts Institute of Technology

Adult stem cells are responsible for renewal, repair, and regenerative processes in adult tissues. Investigation of the molecular basis for these important adult stem cell functions has been largely precluded because of the lack of methods to identify adult stem cells and expand them in culture without loss of their in vivo properties. We have developed several cultured cell lines that model a defining adult stem cell property, asymmetric cell kinetics.

Asymmetric cell kinetics underlie the ability of adult stem cells to produce differentiating progeny cells while at the same time retaining their own undifferentiated character. These experimental cell models have been used to investigate proposed adult stem cell-specific functions. In particular, we have investigated non-random chromosome segregation, which was first proposed by Cairns in 1975 as a critical protective mechanism against the accumulation of carcinogenic mutations in adult stem cells. Selectively retaining chromosomes bearing "immortal" DNA templates, while segregating away those bearing newer DNA copies, would allow adult stem cells to avoid accumulation of mutations that arise from DNA replication errors.

Employing cytological and physicochemical methods, we have now shown that cultured asymmetrically cycling cells do, in fact, non-randomly segregate chromosomes to preserve immortal DNA strands. Studies, so far, have not revealed a chromosome-marking event; but it is clear that like asymmetric cell kinetics, non-random chromosome segregation requires p53-dependent repression of guanine ribonucleotide biosynthesis via the enzyme inosine-monophosphate dehydrogenase (IMPDH). These findings lead us to new hypotheses regarding the role of p53 in adult stem cell function and carcinogenesis. Our current hypothesis for the molecular mechanism responsible for immortal DNA strand co-segregation will also be discussed.

chromatin@abcam.com

Products by research area: