Author ORCID Identifier(s)
Date of Award
4-24-2026
Document Type
Thesis
Degree Name
Doctor of Philosophy (PhD)
First Advisor
Christopher Vakoc MD, PhD
Abstract
Transcriptional dysregulation is a common feature of many cancer types, yet targeting this dysregulation remains an underutilized therapeutic strategy. A major mechanism of transcriptional dysregulation is the upregulation of lineage-specific transcription factors. Consequently, many cancers become dependent on these transcription factors which support an aberrant epigenomic landscape that favors proliferative growth. Previous literature has established the transcription factor SOX10 as a highly potent and selective dependency in melanoma. By identifying the critical surfaces and interacting partners of SOX10, I hope to reveal molecular vulnerabilities that can be targeted to suppress melanoma proliferation. I established a gene complementation assay by stably expressing SOX10 cDNA to rescue knockout of endogenous SOX10. This approach identified the HMG DNA-binding domain of SOX10 as the most critical domain for cell growth. I hypothesized that the SOX10 HMG domain in melanoma is multifunctional and mediates both DNA and essential cofactor interactions. To decouple these two essential functions, I established a SOX family domain-swap assay. Specifically, I created a panel of 19 synthetic cDNAs by replacing the HMG domain of SOX10 with the HMG domain of all other members of the SOX family. These synthetic cDNAs allowed us to perturb the amino acid sequence of the SOX10 HMG domain to probe for regions binding essential cofactors, without sacrificing the DNA binding function. Of the 19 tested, 12 HMG domains supported cell growth while the other 7 could not. Surprisingly, every amino acid perturbation unique to the seven non-rescuing HMG domains faced away from the DNA binding surface. This supports our claim that the SOX10 HMG domain in melanoma supports essential cofactor binding in addition to DNA binding. These findings have allowed us to identify melanoma-specific requirements of the HMG domain at amino acid resolution. ChIP-seq of two of the non-rescuing domain-swapped cDNAs revealed that these constructs have a diminished ability to bind chromatin compared to wildtype SOX10. Furthermore, crosslink immunoprecipitation-mass spectrometry of these same non-rescuing constructs has identified a differential enrichment of members of the SWI/SNF complex compared to wildtype SOX10. A member of the SWI/SNF complex, SMARCD1, was further validated via western blot to interact less with either domain swap compared to wildtype SOX10. I believe that SOX10 interaction with the SWI/SNF complex is essential to SOX10 engagement with the chromatin and thus melanoma proliferation. My work highlights a novel avenue for therapeutic development in melanoma.
Recommended Citation
Kechejian, Vahag, "Lineage Dependency Mechanisms of the Transcription Factor SOX10 in Melanoma." (2026). Donald and Barbara Zucker School of Medicine Theses and Dissertations. 58.
https://academicworks.medicine.hofstra.edu/zsom_dissertations/58