IDENTIFICATION OF DISEASE-SPECIFIC VULNERABILITY STATES AT THE SINGLE-CELL LEVEL

Area of Science:

• Oncology
• Genomics
• Computational Biology

Background:

• Glioblastoma (GBM) exhibits significant intratumor heterogeneity, complicating treatment strategies.
• Identifying specific, vulnerable cell populations within GBM is crucial for effective therapeutic targeting.

Purpose of the Study:

• To integrate single-cell transcriptomic data with loss-of-function screens to pinpoint critical GBM cell vulnerabilities.
• To develop a computational pipeline for identifying and characterizing distinct cellular vulnerability states within GBM tumors.

Main Methods:

• Utilized CRISPR screen data from the Dependency Map (DepMap) Consortium to define a GBM Dependency Signature (GDS).
• Applied similarity scoring of GDS profiles in single-cell RNA-sequencing (scRNA-seq) data and hierarchical clustering.
• Integrated scRNA-seq and spatial transcriptomic data to characterize three single-cell vulnerability states (VS) across 49 GBM tumors.

Main Results:

• Identified three distinct single-cell vulnerability states (VS) in GBM, reflecting unique gene dependencies and spatial distributions.
• Demonstrated differential drug sensitivity across VS, with VS2 showing specific responsiveness to temozolomide.
• Observed variable proportions of VS within individual GBM tumors, suggesting potential for patient stratification.

Conclusions:

• Developed a novel computational pipeline to identify unique vulnerability states in GBM and potentially other cancers.
• The identified VS offer a framework for stratifying patients in clinical trials and guiding the development of targeted therapies.
• This approach can aid in identifying existing or novel drugs for treating intractable cancers like GBM.