Novel chromatin Effects of Oncogenic EZH2 mutations

Epigenetic modifiers are frequently mutated in almost all cancers, including melanoma, and are also associated with poor survival, response to treatment or relapse.  They are thus a relatively new, but very promising class of cancer therapeutic targets.  Unlike the genetic code, which is static, the epigenome is reversible, creating numerous opportunities for manipulation using small molecule inhibitors to restore a normal epigenetic state, or to take advantage of genetic or synthetic lethal interactions.  Despite the availability of numerous small molecules targeting epigenetic/chromatin modifying enzymes, progress in terms of therapeutic efficacy has been slow.  We are only just beginning to understand the interconnectedness and importance of the various epigenetic mechanisms on gene expression and consequently molecular and cellular functions that play critical roles in oncogenic transformation and cancer progression.  Our understanding of the epigenome is thus quite rudimentary.  The fact that progress has been slow and that clinical trial results are mixed, highlights the significant gap in knowledge in not only the crosstalk among various components of the epigenome, but also the downstream mechanisms of epigenetic/ chromatin regulators that are frequently mutated in cancer.  It is essential, therefore, to better understand the mechanistic underpinnings of these chromatin modifications in order to make progress in developing more effective therapies.

The proposed research plan for the CRF proposal was based on the above premise, to understand novel chromatin dynamics of one such epigenetic modifier called EZH2.  EZH2 is a unique gene because it can function both as a tumor suppressor and as an oncogene.  In other words, in some settings its expression can prevent cancer while under a different setting it can promote cancer.  Understanding the mechanisms, therefore, of how EZH2 functions is very important in designing therapeutic strategies.  In this proposal we set out to a) determine how mutations in the Ezh2 gene cause changes in chromatin methylation across the genome, and b) investigate the genetic interactions between Ezh2 and other epigenetic modifiers.  Overall, we made progress in understanding how Ezh2 works, but also discovered some unexpected interactions with non-chromatin proteins that regulate the tumor immune response.

Interactions with other epigenetic modifiers

We tested several molecular genetic interactions between Ezh2 and other epigenetic regulators, both in vitro and in vivo.  For example, we tested whether activating mutations in Ezh2 cooperated with deletion of another epigenetic regulator called Arid1a, a core component of a protein complex, which usually antagonizes canonical Ezh2 activity.  In vitro testing did not show any definitive results, so we acquired genetically engineered mouse models to test a potential genetic interaction in vivo.  In vivo testing of compound engineered mice showed a mild change in B cell differentiation, however, the combination of the two genetic events did not accelerate the onset of lymphoma.  Arid1a has several other related genes/ homologs, so we next hypothesized that perhaps other Arid genes (Arid1b and Arid2) may be compensating for loss of Arid1a, thus masking a potential phenotype.  We acquired both Arid1b and Arid2 conditional knockout mouse models and we are in the process of combining those alleles for further phenotypic and genetic studies.  In the meantime, we acquired several human B cell lymphoma cell lines that express either wild-type or mutant Ezh2^Y641F and using CRISPR/Cas9 technology, we are currently engineering knockout mutations in the Arid genes to test genetic interactions and dependencies in human cell lines.  These studies are ongoing.

Non-canonical interactions in melanoma

EZH2 is a gene that is mutated in a wide range of cancer, from hematopoietic malignancies such as lymphoma, to solid tumors such as lung, prostate, breast and melanoma.  Specifically, the Y641F mutations that our lab is studying are present in diffuse large B cell lymphoma (DLBCL), follicular lymphoma (FL), and also melanoma.   We therefore study the role of these mutations in both lymphoma and melanoma in order to better understand the context under which they drive development of cancer.

During our studies in melanoma, we noted that cells expressing the mutant Ezh2 protein exhibited a gene expression signature that was dependent on interferon signaling.  Interferons are a group of proteins that can modulate the immune system.  We initially investigated whether the overactive Ezh2^Y641F mutant was directly regulating expression of these interferon genes via direct methylation of the chromatin region (H3K27me3) at the promoters of these genes.  Surprisingly, we found no differences in H3K27me3 at the promoters of these interferon-related genes.  Expression of these genes was still dependent on Ezh2 enzymatic activity, so we investigated whether this was an indirect effect via methylation of a non-histone protein.  We hypothesized that perhaps Ezh2 could methylate an immunomodulatory factor that could in turn regulate expression of interferon genes.  We found that Ezh2 directly interacts and methylates Stat3, a transcription factor that regulates expression several interferon-related genes, and which also can have significant effects on the immune system, and consequently how the immune system recognizes and destroys cancer cells.  We next investigated whether this Ezh2-Stat3 interaction had any consequences on the anti-tumor immune response.  We found that expression of Stat3 was required to maintain an anti-tumor immune response and its depletion promoted melanoma progression and recurrence.  Molecularly, Stat3 and Ezh2 bind together at many genomic loci, and can both activate and repress gene expression in association with the rest of the PRC2 complex.  These results suggest that Ezh2 can mediate melanomagenesis through evasion of the anti-tumor immune response, and that the immunomodulatory properties of Stat3 are context dependent.

The above results have been submitted to the biorXiv pre-print server and are currently under review.

Overall, in this proposal we explored the underlying oncogenic mechanisms of activating mutations in Ezh2.  We initially investigated canonical pathways and also explored interactions with other epigenetic regulators.  At the same time, we pursued unexpected findings and uncovered significant non-canonical interactions that have significant consequences on the anti-tumor immune in melanoma.  Preliminary data from this award were used for other grants and as a result our lab was awarded a one-year grant from The Harry Lloyd Charitable Trust in July 2021 to continue to study the interaction between Ezh2 and Stat3 in melanoma.  A significant amount of this data was also included in an R01 application submitted in June of 2021.