IDENTIFICATION OF TRIM24 DOMAIN ESSENTIALITY IN PRIMARY TRIM24COE CARCINOSARCOMA CELL LINES

Regulation of transcriptional control is a critical feature for organismal development and survival. Because of its effects on chromatin-controlled expression, disruptions of this delicately tuned mechanism are an important factor in development of tumorigenesis. Therefore, there is a growing interest in targeting these control mechanisms for their potential therapeutic values. Our lab previously discovered the epigenetic regulator function of Tripartite motif protein 24 (TRIM24) through its H3K4me0 and H3K23ac dual histone signature reader function, its negative regulatory effect on the p53 tumor suppressor, in addition to its shown oncogenic driving capacity on immortalized mammary epithelial cells when overexpressed. Although TRIM24 is a highly promising target for its therapeutic potential with its previously discovered domains facilitating several different functions, it is still elusive which of its domains are essential for its oncogenic effects. During our lab’s previous studies on identification of the mechanisms behind the oncogenic effects of TRIM24, we generated a mouse model that conditionally overexpresses TRIM24. We successfully showed that a level of 2-3-fold above normal expression of TRIM24 is sufficient for tumor initiation and progression in mouse mammary epithelial cells. Moreover, the pathological examination of the tumors extracted from these mice revealed over 70% concordance with carcinosarcoma and also were reported to recapitulate human Metaplastic Breast Carcinomas (MpBC). Furthermore, in our earlier research on TRIM24 overexpressing mammary cancer epithelial cell lines, our lab also showed that knocking down TRIM24 using siRNA results in significantly impaired cell growth and induction of apoptosis compared to controls in the mammary cancer cell population. These findings were also consistent with gene dependency data from Project Achilles that has genome-wide loss of function screens from hundreds of cell lines. Considering our observations regarding the dependency phenotype of our cell lines on TRIM24, and that TRIM24 functions are facilitated by its multiple domains from as shown by our lab’s previous studies, we hypothesized that the function of TRIM24 in breast cancer is domain-specific. To test this hypothesis, we performed a CRISPR/Cas9 tiling array screen that targets multiple TRIM24 domains to identify the essential domain or domains in two TRIM24 overexpressing cancer cell lines, with the long-term goal of identifying which domain can be therapeutically targeted. In this study, we performed a comprehensive computational analysis of this CRISPR/Cas9 tiling array screen. By evaluating multiple approaches, we established a computational pipeline to analyze our small scale CRISPR tiling screen to identify the essential domain(s) of TRIM24. We observed that B Box and Coiled-Coil domains of TRIM24 showed essentiality in both of our cell lines in our screen. To computationally validate our findings, we performed a systematic analysis of other publicly available datasets including Project Achilles and DepMap to find similar TRIM24 dependencies and integrated orthogonal datasets from multiple sources including Genomics of Drug Sensitivity in Cancer (GDSC) and STRING Protein interaction databases. We found a similar TRIM24 dependency to our cell lines in AML cell lines and showed that these TRIM24 dependent cell lines exhibited a sensitivity to a JAK-3 inhibitor. We finally proposed a possible explanation of the dependency of our cell lines on the B Box domain which recruits STAT3. While we concluded that targeting B Box domain might be most effective in inhibiting TRIM24 oncogenic functions through JAK-STAT pathway inhibition, further studies are needed to evaluate the outcomes of dual targeting of domain pairs simultaneously to observe additive effects