Highlights
EpiGenie uses more Canadian talent than the NHL, so naturally we sent one of our staff writers, Ben Laufer, to hit up this nice little meeting. He got to enjoy the home court advantage at Canada’s second national conference on Epigenetics at Western University in London, Ontario. Check out Ben’s report below:
Meeting Summary
Tucked away in one of the University’s lecture halls, the 2nd Canadian Epigenetics Conference had a strong focus on biochemistry and clinical application. The term epigenetics was taken somewhat loosely and rather than a pure focus on inheritance and genetics, there was a bit of a slant towards biochemical machinery and protein names. However, through all the cutting edge data presented, plenty of interesting themes arose, including a noticeable impact of Bing Ren’s work, a new appreciation for the complex transcriptional control of nuclear pore complexes, and an engulfing interest into the emerging massive regulatory roles of lncRNAs.
Histone modification and exchange during sense and antisense transcription
Jerry Workman, Stowers Institute for Medical Research
Dr. Jerry Workman talked about his research on chromatin mechanisms. In a yeast model his group looks at how chromatin signatures are reset during transcription. In their research they have found that these chromatin signatures can control strand specific expression and they are established by the Set2/Isw1b/Rpd3S pathway.
Workman’s group has taken an interest in how cryptic transcripts occur due perturbations in this pathway, ultimately resulting in improper chromatin spacing. The chromatin spacing by this pathway influences what DNA sequences are amenable for RNA Polymerase II to bind at regions that resemble promoters. By looking for cryptic transcripts, in Set2 loss of function mutants, they found hundreds of Set2 Repressed Antisense Transcripts (SRAT). Interestingly, 82% of the yeast genome is transcribed as bonafide genes in wild-types and 30% show up in their Set2 mutant as SRATs.
Next, they found that H3K36 methylation mediated by Set2 suppress cryptic transcripts by preventing their initiation and that the antisense transcripts produced upon loss is Set2 mediated methylation are usually intergenic and contained within the gene body. However, he notes that the levels of antisense transcripts are significantly lower than those of protein coding transcripts in his mutants. In another set of experiments they found that Set2-mediated suppression of antisense transcripts does not require the Rpd3s deacetylation pathway, but the Isw1b and Chd1 chromatin remodellers are involved in suppressing the antisense transcripts. Finally, they found that transcription of the antisense RNA overlapping the sense promoter is required to repress transcription in cis.
This leads him to the conclusion that “real genes with real promoters” use this pathway to silence cryptic transcription, with a take home message that the process of antisense transcription over a promoter represses transcription through Set2 and H3K36 methylation. Ultimately, his research suggests that “transcription can regulate transcription.”
Dynamic DNA methylation and gene expression in spermatogonial stem cell differentiation
Hiroyuki Sasaki, Kyushu University
Dr. Sasaki (Fukuoka, Japan) started off his plenary talk with his early research years and the difficulties he had in mouse models of disease due to genomic imprinting leaving a big mark on his transgene of interest. Ultimately, this started him on a life long journey where ‘imprinting left its imprint on him’. These findings left him with the burning question, how does the same machinery (Dnmt3a+Dnmt3L) react differently to establish imprinting in the sexes? His group then went on to show how piwi RNAs (piRNAs), which are male germline specific, can guide the genomic imprinting machinery in the male germline and influence imprinting control regions (using the example of Rasgrf1) and are distinct from siRNA/miRNA. They also showed how endo-siRNAs in the ooctye arise from naturally occurring dsRNA and can go onto regulate transposons and genes.
Next, Sasaki focused on his current research into stem cell epigenomics, specifically how sperm cells are reprogrammed throughout development. Utilizing post-bisuflite adaptor tagging (PBAT), they examined the methylomics of differentiating postnatal sperm stem cells. They found large partially methylated domains (PMDs) that also appear in cancerous and placental cells, but not somatic cells. They also found that a certain stage of differentiation (neonatal prospermatogonia) shows very high levels of non-CpG methylation and that there were a number of stage specific DMRs that were important to the regulation of stem cell properties and differentiation, with there appearing to be distinct classes of DMRs that are enriched for Sox transcription factor binding sites.
XIST: Long non-coding RNA as CIS-acting silencer
Carolyn Brown, University of British Columbia
Dr. Carolyn Brown (Vancouver), a pioneer in X chromosome inactivation, discussed her groups work on imprinting and dosage compensation. Starting off with the classical example of calico cats, her initial interest lies is in the classic example of how dosage compensation results in the inactive X taking on the state of facultative heterochromatin that results in the mosaic expression patterns that produce the recognizable fur patterns. Getting into the specifics, the long non-coding RNA (lncRNA) Xist is her molecule of interest. She views lncRNAs as modular adaptors, bringing proteins to sites of action, in the case of X it recruits epigenetic modifications that silence the 160 MB of DNA on the X chromosome.
In Brown’s earlier research, using a Dox system and transgenes, her team found that nuclear contact and H3K27me3 are key features of X inactivation. Overall, there is a recruitment of repressive marks by Xist, with the exception of H3K9me3, but the cascade of events leading to this remains unknown. So rather than focus on the chain of events, they shifted to understanding what regions of DNA are responsible for certain events, by carrying out systematic deletions.
Brown’s group found that silencing correlates with the number of repeats since it requires stems and loops to form. They are now focused on DNA methylation marks, as their earlier model wasn’t able to recruit them properly. But overall, she thinks that initial expression of Xist induces silencing in cis, which has a different control mechanism in humans and mice, and regardless of the mechanism it recruits the chromosome to a unique cellular region where the repeats are then silenced so their RNA products cannot compete with Xist and thus this allows it to bind and silence the chromosome.
Chromatin folding in post-mitotic neurons
Ana Pombo, Humboldt University
Dr. Ana Pombo (Berlin, Germany) discussed her research on nuclear architecture and offered up a refreshing perspective on genome structure and how it contributes to the regulation of gene expression. Her interest lies in the chromatin organization that is responsible for long-range interactions at the DNA level and at the preference of chromosome positions at the cellular level. Her group’s research utilizes chromatin confirmation capture (3C) based technologies (Hi-C) to identify stem cell differentiation systems that lead to distinct populations of neurons, specifically ESCs, Neural Progenitor Cells, and post-mitotic neurons. They observed that chromatin folding greatly changes through out neural differentiation by analyzing distinct topological domains. By examining the interactions of distinct topological domains they noticed unique and consistent profiles throughout differentiation from ESC into post-mitotic neuron. The observed structure of domains-within-domains that span regions of the genome ranged up to 10 Mb.
Overall, the general organization is conserved in the different cell stages looked at but the functional changes are reflected in the more local details. Intriguingly, these meta structures (known as ultrametric trees) correlated with epigenetic features in their regions. They then went on to model the theory behind it in silico. Ultimately, by viewing chromatin as a dynamic polymer with meta domains that interact in a 3D fashion a lot can be learned about differentiation.
Histone methylation mediated silencing of endogenous retroviruses
Matthew Lorincz, University of British Columbia
Dr. Matthew Lorincz (Vancouver) continued on the theme of transcriptional regulation of transposons and talked about his research into epigenetic processes of development. Endogenous retroviruses (ERVs) are repeated elements that have repeatedly integrated into the mouse genome and represent a genome integrity burden as some ERVs may still be active and represent a large potential for mutation. DNA methylation has long been known to play a role in silencing ERVs and its dynamics through out development are extremely important for proper development, a common theme of this meeting.
However, it appears that DNA methylation wasn’t able to explain all of the ERV repression, and many (Class 1 and 2) are marked with H3K9 by the methyltransferase Setdb1, which was previously thought to be unique to somatic cells. Interestingly, when Setdb1 was knocked out conditionally in the germ cells, it was the only gene able to stop the repression of ERVs at H3K9 sites. These sites showed a difference not only in H3K9 methylation but also DNA methylation in their low input (1000 cell) ChIP-Seq method.
Furthermore, while not too many new transcripts were seen, the ones expressed appeared to be coming from cryptic promoters in the repeats of ERVs and there was some sex specificity in the activation. Deficiency of Setdb1 causes cell viability defects and suggests that it is an essential guardian against the retroviral driven expression and maintains ERV repression in the wake of de novo methylation that occurs in the mammalian genome during development, as marked ERVs show relatively low levels of active and passive demethylation by Tet.
Genetic and environmental factors controlling RNA-DNA hybrids
Karim Mekhail, University of Toronto
Dr. Karim Mekhail (Toronto) talked about his research on how ncRNAs regulate chromosome stability and cellular lifespan in the brain. He started off in budding yeast and has now moved to humans. His research has had a strong focus on ribosomal repeats and their transcriptional complexity, due to their exposed repetitive nature they require special considerations in the cell, such as localization in order to protect it from recombination and ncRNA on a destructive mission. So they carried out some RNA-DNA hybrid ChIP, and while examining the deletion of the gene Pbp1 his group observed that it was preventing ncRNA and DNA hybrids within the intergneic spacers of rDNA repeats.
Interestingly, it didn’t interact with RNases but rather a protein called Stm1. When Stm1 was deleted it recovered the defects associated with Pbp1 deletion. Furthermore, caloric restriction was able suppresses hybrids in cells with Pbp1 deleted via RNAseH and Pif1 and recover repeat integrity. Taken their results to the genome wide level the observed a global role for Pbp1 in RNA-DNA hybrid suppression in repetitive elements in the genome and appears to have a role in neurological disease.
Gene regulation via gene looping & noncoding RNA
Musa Mhlanga, CSIR Synthetic Biology
Dr. Musa Mhlanga (Pretoria, South Africa) talked about his group’s research into epigenetic technology and chromatin architecture by presenting a molecular approach into manipulating chromosome loops. His teams interest lies in that gene loops are not a consequence of transcription, but rather a causative factor behind it. He interrogates whether inter and/or intra chromosomal interactions and domains are needed for transcription to occur. There are 3 main types of gene loops: Enhancer promoter interactions, Gene loop regulatory elements (like CTCF sites), and multi gene complex interactions. The latter represents topological domains that share a focus of RNA polymerases and are co-regulated. His research examines how the dynamic movement of chromosomes makes it possible for such gene loops to form.
Interestingly, the higher order regulation in the nucleus is related to the position of DNA in the nucleus and determines gene expression in “jackpot” cells. By using TALEN genome editing (and later confirmed via CRISPR), they disrupted a single gene loop in a well characterized multi gene complex that is co-transcribed. Upon doing this his group observed that the disruption of a single loop was enough to interfere with the whole complex and causes a transcriptional response of the co-regulated genes. Interestingly, they found that chromosomal breaks uni-directionally affect transcriptional ‘kisses’ ranging from Mb to entire chromosomes. Finally, he demonstrated some of the roles of lncRNAs in influencing looping and transcription.
**EpiGenie would like to thank Ben Laufer, who is a PhD student in the Singh Lab at Western University for contributing coverage of this conference**