Highlights
After a grueling, all-out Sumo style wrestling match broke out at EpiGenie HQ over who should go to Costa Rica to for this conference, we decided the safest thing to do for all involved, was to send someone else. So, for this event we teamed with Abcam to send a special guest reporter to cover all of the action.
Our EpiGenie Field Correspondent was Marloes de Groote, a grad student at University Medical Center Groningen in the Netherlands.
Abcam’s Fifth ‘Chromatin: Structure & Function’ Conference took place in the Hilton Papagayo Resort in Costa Rica. One could not think of a more magnificent location in terms of surroundings. Not only is the flora in Costa Rica extraordinary, but the wildlife is amazing. As many presenters referred to the wonderful species they had seen during their stay, I guess I was not the only one enjoying it. The resort itself served great food and nice drinks to enjoy during the significant amount of time held free for networking. Also from a scientific point of view, in my opinion, the conference was great.
The meeting started off on November 16th with a talk from John Mattick. Overall, the conference made clear that next to the still evolving world of DNA (de-)methylation and histone modifications an immense diversity of non-coding RNAs appear to play an important role in epigenetics, more essential than a lot of us would probably have guessed. This actually is not so strange since the majority of our genome appears to be transcribed although only a very small part is coding for proteins. All the other transcripts rationally seen must have some function too.
RNA Regulation of the Epigenome
John Mattick, University of Queensland
John Mattick discussed in some detail the importance of RNA in the regulation of gene expression. One of the great sentences he used covered a big part of the vibe of this conference:
“Rather than oases of protein-coding sequences in a desert of junk, the genome is better thought of as islands of protein-coding sequences in a sea of regulation… most of which is transacted by RNA”.
He shared that RNA might be the missing link when talking about the targeting of epigenetic enzymes to their specific place of actions.
Mattick explained how the relative amount of non-protein coding DNA increases with the complexity of the organism. This non-protein coding DNA is very important for gene expression regulation. That is, according to John Mattick, why us humans are so different from C. elegans even though we do have the same amount of protein coding genes, which is even less than the amount in plants and protozoa.
The non coding RNAs (ncRNAs) themselves are again epigenetically regulated and thus differentially expressed. The functions of ncRNAs mentioned by John Mattick are among other things:
- Chromatin modification/epigenetic memory
- Transcriptional regulation
- Control of alternative splicing
Recently his group identified tiny RNAs (18 nt) that are associated with transcription initiation sites (tiRNAs). These tiRNAs might play a role in nucleosome positioning. Other than that he talked about the fascinating phenomenon of RNA editing, which seems to play a big role in implementing environmental factors in the epigenome.
Chromatin Associated Large Intergenic Non-coding RNAs (lincRNAs) in Cancer and Stem Cells
John Rinn, Harvard Medical School
Another person highlighting the role of RNA in epigenetics was John Rinn. His lab uses chromatin features to predict sites where long non-coding RNAs (the “missing lincs” as he named them very appropriately) are transcribed: e.g. H3K4me3 is associated with an active promoter and H3K36me3 with elongated transcripts. Any position where these marks follow each other, a long non-coding RNA might be transcribed. In this way up to approximately 5000 large intergenic non-coding RNAs (lincRNAs) were found by John Rinns lab in the human genome.
Their function is diverse, including roles in cell cycle regulation and in maintaining embryonic stem cell pluripotency. E.g. lincRNA p21 is positively regulated by P53 in order to repress genes. John Rinn told us that lincRNA p21 phenocopies p53, since knocking down lincRNA p21 leads to death of the cell after DNA damage. LincRNA p21 appears to be a global repressor in the p53 pathway via binding of hnRNP-K. LincRNA p21 repression occurs via recruitment of chromatin modifying complexes.
In general lincRNAs are associated with repression via association with chromatin remodeling complexes. In addition, a very exciting lincRNA was found by John Rinns group that shows to be regulated by Oct4, Sox2 and Nanog, and is required for correct reprogramming of fibroblasts into pluripotent cells.
Copying and Reprogramming of Heterochromatin with RNAi
Robert Martienssen, Cold Spring Harbor Laboratory
Since most DNA is transcribed, heterochromatin areas, including transposable elements, show transcription as presented by Robert Martienssen. The RNA transcribed from heterochromatin is however rapidly processed by RNAi, as observed in plants, animals and fission yeast.
RNAi serves as a guide for heterochromatic H3 methylation via the Rik1 complex. This Rik1 complex is responsible for both H3K9 methylation and H3K4 demethylation. RNAi also allows the inheritance of epigenetic marks during cell division. Heterochromatin is temporarily lost during cell division. During this loss of heterochromatin, transcripts of the DNA accumulate. During S-phase these transcripts are rapidly processed into small RNAs, which support the restoration of heterochromatic features. Not only do these interesting findings of Robert Martienssens lab show that heterochromatin is transcribed, they also provide a mechanism of epigenetic inheritance.
After that he demonstrated that LTRs of transposons of the Tf2 group colocalize with CENP-B (a conserved DNA binding factor related to the POGO transposase) and Sap1. Transposons are occupied by Sap1 to stabilize them and making sure the DNA is only replicated in one direction. In the absence of CENP-B and Sap1, retrotransposons indeed appear to be instable and they get spliced out of the DNA.
Steps Towards Understanding the Phenomena of Epigenetics at a Molecular Level
Danny Reinberg, Howard Hughes Medical Institute
Of course, next to the role of RNA, histone modifiers and histone modifications are also still researched. Danny Reinberg provided more insights in the role of PR-Set7, which is a HKMT known to be specific for H4K20 monomethylation.
In contrast to some previous publications, in this presentation evidence was provided for PR-Set7 having M-phase specific expression rather than in the S-phase. The fact that an S-phase phenotype was seen in PR-Set7-depleted cells can be explained by the loss of H4K20 monomethylation at that time instead of the lack of PR-Set7. PR-Set7 is accumulated at DNA damage and recruits among other things, PCNA, which is part of the DNA repair machinery. PR-Set7 inhibits PCNA dependent DNA polδ activity.
PR-Set7 is important in heterochromatin formation since H4K20me1 is required for condensation; it is recognized by L3MBTL1 which is important for chromatin compaction. Danny Reinberg’s group shows that PRSET7 and L3MBTL1 are necessary for transmission of H4K20me1 to daughter chromosomes. L3MBTL1 binds the methylgroup of H4K20me1 and complexes with PRSET7 which facilitates the addition of a methyl-group to the new H4K20. The presenter showed that when L3MBTL1 is downregulated, there is less H4K20me1 whereas the amount of H3K27me2/3 remains unchanged.
Chromatin Modifying Enzymes: Function and Role in Cancer
Tony Kouzarides, University of Cambridge
The organizer of this conference gave a talk about diverse chromatin modifying enzymes. One of the modifications he spoke about was the phosphorylation of H3Y41 by the JAK2 enzyme. This work, recently published in Nature, shows that JAK2 has nuclear activity next to its function in the cytoplasm. It phosphorylates H3Y41 which on its turn prevents binding of HP1α.
Normally the binding of HP1α leads to gene repression, which is thus prevented by H3Y41ph. In the cytoplasm JAK2 phosphorylates STAT, which is thereby transferred to the nucleus to bind its STAT binding site. Occupation of genes by STAT5 overlaps in approximately one third of the cases with H3Y41ph. So it might be that the JAK/STAT pathway in the nucleus is acting similar to that in the cytoplasm.
The other part of Tony Kouzarides’ presentation was about EMSY, a gene that is known to be amplified in 13% of sporadic breast cancers and in 17% of high grade ovarian cancers, giving a poor outcome. EMSY has a repressor function that is HP1 associated.
Overexpression of EMSY now also appears to repress anti-metastatic miRNAs (and thereby increasing metastasis) by binding to the promoter of these miRNAs. EMSY inhibits among others. miR-31 via ETS1 through the ETS binding site.
EMSY is in complex with Jarid1B – a JMJC domain containing H3K4 demethylase. It interacts with Plu1 and ETS1. Plu1 binds miRNA promoters and inhibition of Plu1 increases the expression of the miRNAs. Repression of anti metastatic miRNAs (via H3K4 demethylation) is breast cancer specific (the genes involved are overexpressed). Overexpression of metastatic genes leads to metastasis and increased cell motility. Higher EMSY means that the target genes of miR-31 are activated and amplification of EMSY leads to higher cell motility.
Chromatin Structure and Epigenetic Alteration in Development and Tumorigenesis
Luciano di Croce, European Institute of Oncology
This presenter talked about the role of macroH2A (mH2A) in silencing and development. mH2A is, for example, involved in X chromosome inactivation, it is conserved and only a few target genes were known so far. Its function is not well understood.
By doing ChIP on chip 760 high confident targets were found by Luciano di Croce’s team. The targets of mH2A have low or no transcription and mainly play roles in development. mH2A is preferentially expressed in the brain. Loss of mH2A leads to severe defects. mH2A displacement is reversible on activated promoters. (Nat. Struct. Mol. Biol. 2009) The question arose how the deposition of mH2A is regulated and how it affects the chromatin structure and function.
Luciano di Croce’s group shows that ZRF1 binds specifically to ubiquitylated H2A by its conserved ubiquitin binding domain. ZRF1 showed ubiquitin specific binding at Hox promoters. PRC1 binds to nucleosomes as well in a ubiquitin dependent fashion. Due to this similarity, ZRF1 and Ring1B (a component of the PRC1 complex) compete for H2Aub binding. In vivo an increase of ZRF1 therefore leads to release of PRC1. This is consistent with the fact that ZRF1 has overlapping target genes with PRC members. ZRF1 regulates PRC1 occupancy and converts repressive marks to active. ZRF1 is in this way important for activation of PcG silenced genes.
Conversion of 5-methylcytosine to 5-hydroxymethylcytosine in Mammalian DNA by MLL Partner TET1
Mamta Tahiliani, Harvard Medical School
DNA methylation was also a big topic covered at this conference. The fascinating debate on whether there is active DNA de-methylation taking place and which enzyme is responsible continued in the talk of Mamta Tahiliani. She shows that what was already known to occur for thymines in trypanosomes, now apparently also holds true for methylated cytosines in mammalian cells: a hydroxy-group can be added to the 5mC.
In this presentation, it was demonstrated that the TET1 protein is responsible for converting methylated cytosines into hydroxymethylcytosines (hmC) in mammals. This might be an in-between step towards de-methylating cytosines. An increase of TET1 did show a decrease of methylated cytosines, whereas a TET1 mutant was not able to show this.
hmC is a normal constituent of ES cell DNA, but not of all TET expressing cells. 4% of the MspI sites consists of hmC, whereas 55-60% consists of ‘regular’ 5-methylcytosine. The remaining question in this research is primarily whether this newly discovered mechanism of DNA de-methylation is passive, active or works via a secondary pathway of (prevention) of recruitment of other proteins.
Takeaways
This conference taught us that still new histone modifications and modifiers continue to be found as well as increasing groups of other important new players in epigenetics like ncRNAs and DNA de-methylases. In addition to the talks I reported on, there were many more interesting talks including those by Kristian Helin, François Fuks, Shelley Berger, Ramin Shiekhattar and Ali Shilatifard that I unfortunately was not able to cover for those of you who missed the conference. We of course already knew that epigenetics is very interesting, but it seems to get more interesting by the minute! I am sorry to be back in the cold and rainy Netherlands again, I can’t wait for the next interesting Abcam epigenetics conference in a beautiful warm place!
-Contributed by Marloes de Groote