Polycomb and trithorax: the yin and yang of epigenetics

We met with chromatin experts Luciano di Croce and Ali Shilatifard at the Chromatin: Structure and Function conference last November. Watch our video to learn more about the importance of epigenetics in health and disease and the future of epigenetics.


LDC:      Hello everybody, I am Luciano di Croce; I work at the CRG in Barcelona in an ICREA professorship position. The topic of my lab is centered on understanding the role of polycomb in ES-cells, including ES-cell differentiation on one side, and on the other side it's understanding when polycomb becomes misregulated how does this lead to the cancerogenesis process. As model system in the lab, we are using Leukemia.

AS:         Good afternoon, my name is Ali Shilati and I'm professor and chairman of the Department of Biochemistry and Molecular Genetics at Northwestern University School of Medicine. In my laboratory, for the past 20 years, we've been studying the role of trithorax which is the opposite of what Luciano has been studying. He studies polycomb, which is a negative regulator of gene expression. I study trithorax, which is a positive regulator of gene expression, and the complexes associated with trithorax, their role in regulation of development and other mutations with associated pathogenesis of a large number of cancers.

Epigenetics and human disease

AS:         I think what we have in common in here is that recent genome sequencing has demonstrated that mutations in the polycomb subunit and trithorax subunits are found in thousands and upon thousands of different forms of cancer from hematologic malignancies to solid tumors; indicating that nature has done the genetic for us. If you have a mutation within any of these family members, you're bound to have developmental disorders, including cancer.

LDC:      I would say, actually, together with cancer, mutations in trithorax, the MLL complex and polycomb, are also associated with different types of human disease, including psoriasis, diabetes…

AS:         Absolutely.

LDC:      …and others. So I think those are - these two protein complexes are pillars in the functionality of cell development.

AS:         I had a conversation - we are at the meeting right now, and we had a conversation last night with a colleague and we thought that many of these factors involved in polycomb and trithorax are actually involving other immune disorders such as lupus, and inhibitors of the activity could be quite central.

LDC:      Yeah, so, really, understanding the molecular mechanism of polycomb and trithorax, I think can be a key feature for future target therapeutics.

AS:         Absolutely.

LDC:      And for human health.

Epigenetics drugs

AS:         A fantastic story we published with St Jude Children's Research Hospital, where diffuse pontine glioma patients' genome was sequenced, and they found that 75 per cent of the patients, children who have this diffuse pontine glioma have a single point mutation in H3K27 to methionine. This is exactly the site that polycomb methylates, and we actually modelled this in drosophila melanogaster and we found out that when we modelled H3K27 in melanogaster, there was a loss of K27 methylation. There is huge regulation of K27 acetylation, and binding of the two factors Luciano mentioned - BRD1 and BRD4 - to this acetylated chromatin. So we managed to be published in Science last year, and we proposed that the use of drugs such as iBET or JQ1, or BET domain inhibitors could be central for the treatment of these. We're now testing these things in a mouse model system where we take the patient tumors, putting them in the mouse and asking what happens when we treat these animals with these drugs? I think the finding is going to be fantastic, and there is going to be - although there's been first generation and second generation, I think we're going to have much better third, fourth and fifth generation of drugs coming for these epigenetic regulators.LDC:      Back more than six years ago there was an important effort in identifying chemical compounds that actually can really inhibit these multiprotein complexes, the inhibition of which can lead to a beneficial effect on your own health. Some of these have actually been developed by different companies: Constellation, Epizyme and other companies, which, actually, leads to the blocking of the catalytic core, for example, of PRC2 complex through EZH2 activity. Actually, now, there is a second generation of a chemical compounds, which are involved in the inhibition of another epigenetic machinery, which is the BRD family protein from BRD1 to BRD4. This also has a very possible and important impact on future therapeutic aspects.

LDC:      Actually, if you go back ten years ago, everything started with the inhibition of HDAC enzymes.

AS:         Absolutely.

LDC:      Those chemical compounds like PSA, valproic acid and SAHA, are now being used for treating leukemia patients. So you see that how really from inhibiting an epigenetic enzyme, in this case it was HDAC, you can really have an effect on the health of human people. So, what was mentioned before is that for most of the chemical compounds, the aim is to inhibit the core of PRC2 complexes. We have recently demonstrated that actually there is variation of PRC1 and PRC2 complexes, and some of these variations are associated with high proliferation in tumors, while some others are implicated in cell differentiation. So you don't need to kill all the PRC1 complexes or PRC2 complexes. You need to kill the specific variation which supports or induces hyper-proliferation. So a chemical compound, which can target that variation of PRC complex can be much more beneficial, rather than killing all the polycomb complexes, which can have dramatic side-effects.

AS:         Absolutely.

The future of Epigenetics

LDC:      I think - that's my opinion - that what we have done in the last five years with the implementation of ChIP-seq and RNA-seq data, is to have a global view of how the transcriptome changes, how different factor’s binding to the genome changes upon different stimulus or different perturbations. Now, we need to integrate this in chromatin architecture. We know that chromatin in the nucleus is not linear, it actually assembles in what are called Topologically Associated Domains. Within this Topologically Associated Domains there are multiple genes- like between 10 and 100 genes - that are co-regulated. So really understanding that the function of these Topologically Associated Domains, how they evolved, how they are conserved, and how they are perturbed in different diseases, I think that's going to be the next challenge.

AS:         I think that definitely Luciano hit the nail on the head, this is where the research area, the community, the basic research community is moving. I can see it from our medical center, we have a large assembly of hospitals in Chicago and for us it becomes mostly targeted therapeutics. You have two individuals with identical translocation, and they have leukemias: one responds, one doesn't respond, why? I think going and figuring out what's the difference between this patient and this patient, and designing and tailor-making the drug for each patient, and I think targeted therapeutics. What I think you also need to have is quite a bit of understanding of the basic research.

LDC:      Yeah, for proper patient stratification you need to understand the genome, you need to understand the transcriptome and how the chromosomes are organized.

AS:         I think the past 10, 20 years, or I guess maybe the past 50 years since the discovery of the polycomb and trithorax family of proteins, we're just developing the alphabet of science. Now we are at the stage that maybe we have a set of alphabet and we can start putting a few words together; and that's probably in the next five to ten years. Once we've put those words together, I would say that 20 years from now we can write paragraphs, and then there would be the chapter of epigenetics contribution.

LDC:      I think we are lucky; we have just found the Rosetta Stone…

AS:         That's right.

LDC:      …and we're currently understanding the genome.

AS:         Very exciting!

View more Epigenetics resources