PROJECT SUMMARY
Diffuse gliomas are incurable brain tumors despite aggressive therapies 1–4. Understanding the mechanisms that
initiate gliomas is a critical step towards improving therapies. The majority of diffuse gliomas are driven by
missense mutations in Isocitrate Dehydrogenase 1 (IDH1) which cause genome-wide DNA hypermethylation, a
characteristic of IDH1mut gliomas 4–7. DNA hypermethylation may promote gliomagenesis by silencing tumor
suppressor genes 8. Alternatively, DNA hypermethylation may also promote gliomagenesis by activating proto-
oncogenes through disruption of CTCF insulator sites 9–13. CTCF insulator sites define the three-dimensional
shape of the genome by dictating the boundaries of topologically associated domains (TADs) 14,15. Enhancers
and promoters can interact when located in the same TAD but not across different TADs 14. IDH1mut gliomas are
characterized with CpG hypermethylation around CTCF sites, effectively leading to insulator disruption and TAD
reorganization, allowing for the activation of the platelet-derived growth factor receptor α (PDGFRA) proto-
oncogene 16. Importantly, the effects of TAD disruption critically depend on the enhancers in the affected loci,
which are strongly cell-type specific 12,14,17–19. I present preliminary evidence that an insulator downstream
PDGFRA is disrupted in IDH1mut gliomas, allowing for aberrant interactions with a strong enhancer in a
neighboring TAD, driving PDGFRA expression. This glioma enhancer is also active in oligodendrocyte-progenitor
cells (OPCs), candidate cells-of-origin for glioma 20–23. I also present evidence that robust hypermethylation of
the CDKN2A promoter silenced this tumor suppressor in IDH1mut gliomas. These strongly methylated elements
can be maintained throughout cell division by DNA methyltransferase 1 (DNMT1), which maintains methylation
patterns through cell replication 24–26. The objectives of this proposal are to (1) Model glioma-relevant epigenetic
lesions at the PDGFRA and CDKN2A loci in OPCs, (2) Test whether disruption of the PDGFRA insulator and
CDKN2A is sufficient to initiate gliomagenesis from OPCs in vivo, and (3) Test whether the epigenetic lesions at
the PDGFRA insulator and CDKN2A promoter are mitotically propagated as tumor drivers in glioma. These
studies will uncover whether mitotically heritable epigenetic lesions are sufficient to initiate diffuse gliomas.
Dr. Bernstein is an internationally respected leader in epigenetics and an outstanding mentor with a history of
successful trainees. The proposed research will be carried out at MGH, a research and medical institute, part of
a vibrant collaborative community that includes the Broad Institute and other local institutes. A research advisory
committee of world-class scientists will provide guidance: Drs. Suzanne Baker, Mario Suvà, and Miguel Rivera.
Critical aspects of the research will be completed through collaborations with Drs. Chao Cheng and Mario Suvà.
The K99/R00 award will provide me with the best opportunity to succeed in my career and will be invaluable for
a successful transition to independence, allowing me to start an ambitious research program.