Integration of Gene Dependency Scores and scRNA-Seq Data Reveals ACTN1 and CD44 as Potential Targets in Glioblastoma

Madison Sharp

Mentor: Dr. Nagi Ayad, Department of Oncology, Georgetown University Medical Center Georgetown University Medical Center

Date/Time: August 23rd, 2022 at 1:00pm.

Abstract: Glioblastoma (GBM), the most common and lethal adult primary brain tumor, has had no new treatments approved in the last decade. With a median survival time of approximately 15 months and a 5-year survival rate of around 5-10% for the current standard of care treatment, the prognosis for patients diagnosed with GBM is poor and new treatment modalities are direly needed. Bioinformatics approaches utilizing publicly available data offer the ability to conduct genome-scale investigations into this issue.  By integrating gene essentiality data from DepMap with single-cell RNA sequencing data of GBM patient tumors, novel GBM vulnerabilities can be revealed. DepMap, the cancer dependency map developed by the Broad Institute, provides researchers with access to genome-wide loss-of-function screening and drug sensitivity data from more than 1000 cancer cell lines. By systematically knocking out individual genes with CRISPR and observing the growth response of the cell line, the DepMap developers calculated a gene effect score to represent the likelihood the gene in question is essential for the cell line’s survival. The gene effect score can then be used to predict genetic vulnerabilities specific to a cancer type that may be potential therapeutic targets. To generate a GBM specific gene dependency signature, gene effect scores from DepMap were used to select an initial gene set of essential genes in GBM cell lines. To verify that these essential genes are specific to GBM tumors, a differential expression analysis with scRNA-Seq data from six patient samples was performed and used to further refine the gene set. To account for intratumor heterogeneity, tumor cells were classified by their transcriptional cell states and, using the marker genes for each state, expression signatures were created. Validating these cell state signatures with an independent external dataset showed consistent expression of the MES1-specific signature across datasets. Kaplan-Meier survival analyses identified two genes that are MES1 markers in both datasets and show a correlation between higher expression and decreased patient survival times – ACTN1 and CD44. The role of CD44 in GBM cell growth, proliferation, and migration has been researched extensively, whereas ACTN1 has yet to be investigated for its role in GBM. However, ACTN1 has been implicated in other cancers, including breast and gastric cancer, in which CD44 is one of its downstream targets. Our findings, along with current literature, suggest that targeting CD44 through ACTN1 in GBM may increase patient survival and should be explored as a potential therapy.