Unaligned DNA sequences viewed on LCD screen
Credit: alanphillips / Getty Images / E+

Researchers from Baylor College Medicine mined data from the Clinical Proteomic Tumor Analysis Consortium (CPTAC) and other publicly available data sites to identify possible new therapeutic targets represented across 10 different cancer types. Their study, published in Cell, delves into the intricacies of proteomics, genomics, and epigenomics data to identify several different types of targets, many of which were validated as promising candidates for novel therapeutic strategies.

CPTAC’s history of pioneering proteogenomics research began in 2011. In 2014, the group published one of its seminal papers in Nature demonstrating the application of proteomics in combination with genomics and transcriptomics to study colon cancer. “This was the first demonstration of the approach,” says Bing Zhang, PhD, senior author of the present study but an early career scientist and first author of that 2014 paper. Since then, CPTAC has conducted numerous studies on ten different cancer types, laying a robust foundation for the current pan-cancer analysis.

“After we generated the data for all these kinds of cancers, we thought it would be interesting to put them together to get more cancer insights and look at the data from all the angles instead of just one cancer,” explains Zhang, who became part of the CPTAC pan-cancer working group which published the collective omics data from the 10 cancer types for pan-cancer proteogenomics research in a 2023 Cancer Cell paper.

However, in this study, the team went a step further—integrating the CPTAC dataset with other public datasets to investigate the similarities and differences among gene and protein alterations found in diverse tumor types and identify potential new protein targets for cancer treatment.

To achieve that goal, the team integrated all of CPTAC’s proteomic data with other omics data to identify potential therapeutic targets. Zhang explains, “The difference between our analysis and previously published analyses is our leverage of CPTAC data, which is primarily protein and phosphorylation-centric. This is the first time we have global protein and post-translational modification (PTM) measurements from more than 1,000 samples from 10 cancer types, driving our analysis.”

Tumor and data types of the CPTAC pan-cancer database. [Li, Yize, et al., Proteogenomic data and resources for pan-cancer analysis, Cancer Cell, 2023]
In this study, Zhang and colleagues integrate proteogenomic data comprising genome-wide information on DNA, RNA, and proteins that was generated by the CPTAC. The data included tissue samples from 1,043 treatment-naïve primary tumors and 524 normal adjacent tissues for comparison across 10 cancer types. The team applied the data integration approach to systematically identify proteins and genes that are important for cancer growth and development. They focused on identifying three different types of protein-related targets.

The first group includes proteins that are either overexpressed or hyperactivated in tumors. Says Zhang, “We identified a few hundred of these targets across all cancer types. The most interesting ones are those shared by multiple cancer types.” The researchers required that at least five cancer types have the same overexpressed or hyperactivated protein to consider it a strong candidate. This approach highlights potential targets that could have broader therapeutic applications across different cancers.

Another group included protein dependencies associated with the loss of tumor suppressor genes. “Our hypothesis is that some tumor suppressors, when lost, may induce dependency on other proteins,” Zhang explains. “We found relationships where the loss of a tumor suppressor gene creates a dependency on another protein, potentially leading to new therapeutic strategies.” This discovery opens new avenues for targeting cancers that have traditionally been challenging to treat due to the loss of tumor suppressor genes.

They also searched for new tumor antigens, including neoantigens —cancer-specific peptides derived from gene mutations in tumors. “We developed a computational pipeline to identify new antigens, which are mutation-derived peptides presented by MHC and can be targeted by T-cell receptor (TCR) therapies,” says Zhang. The team identified new antigens in 30–80% of patients across different cancer types, with some recurrent antigens observed in multiple patients, a finding the researchers believe has important implications for personalized cancer immunotherapy.

The study identifies new targets and suggests practical applications, including biomarker-guided treatments. As an example, the team found higher TOP2A abundance and activity in TP53-mutant uterine cancer. This suggests that doxorubicin, an FDA-approved drug for uterine cancer, could be more effective in this subgroup of patients, whereas, in the general population of uterine cancer, the drug shows activity in only about 16% of patients. “Insights like this could lead to biomarker-guided treatments, improving efficacy for specific patient groups,” says Zhang.

Moreover, the study opens possibilities for drug repurposing. The team found that an FDA-approved antifungal drug could inhibit cancer cell growth potentially expanding its use for cancer treatment.

“Through the integration of 6 omics data types from 10 cancer types with external cell line and human tissue data, we have created a comprehensive resource of protein and peptide targets that covers various therapeutic modalities,” the authors write. Zhang adds, “This study shows that the integration of proteomic data with other omics data provides a more comprehensive understanding of cancer biology, paving the way for innovative therapeutic strategies.”

The team has developed a web portal to facilitate further research that makes all the data easily accessible. “Researchers can use this portal to check whether the data supports their hypotheses or explore drug repurposing opportunities,” says Zhang. The team has identified promising targets and is seeking collaborations for testing in model systems or clinical trials.

Also of Interest