Abstract
A multiomics approach found a master regulator layer integrating effects of distinct mutational landscapes, producing tumor transcriptional identities.
Major Finding: A multiomics approach found a master regulator layer integrating effects of distinct mutational landscapes, producing tumor transcriptional identities.
Concept: Analysis identified 112 distinct transcriptional identities mechanistically determined by small sets of master regulators.
Impact: Master regulators produce conserved tumor dependencies that are virtually independent of the specific mutations in a tumor subtype.
Master regulator (MR) proteins have been suggested to act in concert as part of MR modules (also known as tumor checkpoints) that integrate the impact of genomic alterations and abnormal paracrine and endocrine signals to influence a tumor's transcriptional identity. To further investigate this, Paull, Aytes, Jones, and colleagues developed a novel computational tool named Multi-omics Master Regulator Analysis (MOMA). MOMA can determine a tumor cell's transcriptional identity by integrating multiple sources of omics information, including gene expression data and genomic alteration profiles. Performing MOMA on data from 9,738 samples in The Cancer Genome Atlas (TCGA) revealed 407 MR proteins operating in concert within 24 distinct MR modules, which integrate the effect of mutations in their upstream pathways to regulate distinct cancer hallmarks. These 24 MR blocks identified 112 distinct transcriptional identities (i.e., tumor subtypes) in the 20 TCGA cohorts investigated and were highly enriched in tumor-specific dependencies. One of identified blocks, dubbed MRB:2, was downstream of genetic alterations that drive an aggressive subtype of prostate adenocarcinoma. Genetic silencing of MRB:2-associated genes harboring these mutations increased tumor invasiveness and growth in in vitro and in vivo models of prostate cancer, validating the prediction that MRB:2 was associated with the transcriptional identity of the prostate cancer subtype. Three of the six identified MRB:2-associated genes (MAP3K7, SORBS3, and BCAR1) had not previously been associated with affecting cancer progression. Furthermore, pharmacologic perturbation of proteins from another MR block, dubbed MRB:14, demonstrated that MRB function can be modified to target treatment of specific cancer subtypes. In summary, this study employed a new computational approach to uncover MR modules representing mutation-independent vulnerabilities of cancer cells that are conserved across each tumor subtype.
Note: Research Watch is written by Cancer Discovery editorial staff. Readers are encouraged to consult the original articles for full details. For more Research Watch, visit Cancer Discovery online at http://cancerdiscovery.aacrjournals.org/CDNews.
