ӳ����ý

Over the past two decades, scientists have shed light on the genetic mutations that drive cancer. More recently, they are focusing on proteins to understand the biological events in cancer cells that lead to disease.

In three new studies, researchers including a team from the ӳ����ý of MIT and Harvard have harmonized and analyzed data on proteins, DNA, RNA, as well as clinical data from more than 1,000 patients across nearly a dozen different cancer types. They discovered several ways in which proteins are involved in important cancer processes that are common across many cancer types.

In one study published today in , the scientists homed in on alterations that occur on proteins and change how they behave, called post-translational modifications (PTMs). The team found key modifications, such as the phosphorylation and acetylation of key proteins, that are associated with various biochemical pathways involved in cancer. A second paper showed how certain mutations drive cancer by modifying the activity and regulation of key proteins. And the third study, published in , describes the team’s efforts to harmonize and share the data with the research community.

Together, the three papers could one day help scientists find new cancer drug targets and predict which tumors might respond better to treatment. The studies are part of a larger effort supported by the Clinical Proteomic Tumor Analysis Consortium () at the National Institutes of Health, which aims to integrate analysis of genes and proteins to learn about the molecular basis of cancer.

“Most of biology operates at the level of proteins and their post-translational modifications,” said , a co-senior author on two of the studies and director of the Cancer Genome Computational Analysis Group and institute member at the ӳ����ý.

“Because proteins and PTMs regulate so many signaling pathways that are activated or deactivated in cancer, they’re very important to study,” said Getz, who is also a professor of pathology at Harvard Medical School and the Paul C. Zamecnik Chair in Oncology at the Massachusetts General Hospital Cancer Center.

To immerse themselves in the study of proteins, Getz’s team, along with multi-institutional collaborators, worked with ӳ����ý’s . The platform harmonized the proteomic data for downstream analyses, including for two of the current publications, using a pipeline developed by Karl Clauser, a principal research scientist in the platform and co-author on the papers.

“The critical starting point in the integrated studies being reported now is generation of high-quality, deep-scale and quantitatively reliable proteome and PTMome data across tumor types,” said , a co-author on the studies and senior director of the Proteomics Platform. “With such data, studies like these can provide a valuable resource for obtaining new biological understanding of cancer.”

Making sense of modifications

In the focused on post-translational modifications (PTM), co-first author Yifat Geffen, a postdoctoral scholar in Getz’s lab, collaborated with the Proteomics Platform and other institutions to analyze genomic, transcriptomic, proteomic, and PTM data from 11 types of cancer. They then characterized 33 molecular signatures to group biologically similar tumors. With this approach, the team uncovered patterns in protein modifications linked to cancer-related processes such as DNA repair. The authors said these patterns would have been undetectable in smaller cohorts or by studying only genomic data.

Phosphorylation, for instance, is a type of PTM that can indicate that a protein is active and hence a potential drug target. The researchers found that even cancers that share similar patterns of genetic mutations — such as colon and endometrial tumors that have similar genomic alterations in a DNA repair mechanism — can have different phosphorylation patterns, which could help explain why tumors with similar patterns of genetic mutations can respond differently to specific treatments.

The researchers also found that tumors with reduced levels of acetylation on metabolic proteins  were more likely to respond to immunotherapy. Geffen said that future work could translate these findings into better cancer diagnostics and treatments.

“How post-translational modifications affect the function of proteins is underexplored,” she said. “This is a rich resource, and I’m excited to see everyone dive into the data to find mechanisms that we can potentially target with therapeutics.”

What drivers do

The second study, led by Washington University in St. Louis researchers and Yize Li with Getz as a co-senior author, used the same pan-cancer approach to focus on cancer “driver” mutations. Scientists know these genetic mutations drive the development of cancer, but don’t fully understand the molecular mechanisms behind the process. Figuring out these mechanisms could help researchers develop more effective drugs, including ones that work for more than one type of cancer.

Ding’s team studied a range of features of more than 5,000 driver mutations such as their frequencies across different cancers and their impact on RNA, proteins, protein complexes, and PTMs. They found that certain genetic changes rewire the interactions between proteins. The team also uncovered pairs of genes that, when both are mutated, result in cancer cell death, and could be promising therapeutic targets. (Ding’s team also uncovered drivers of DNA methylation and studied their role in tumor development in a separate .) 

ӳ����ýly, the researchers say that this kind of pan-cancer proteogenomic analysis has only just begun to connect biological features to their genetic roots, and these insights could one day help improve patient care. For instance, clinical trials using proteomics to study tumor samples from patients before and after treatment could reveal molecular mechanisms underlying drug resistance, and point towards more effective combination treatments. 

A new resource

In the third , a team led by Yize Li and of Brigham Young University with Geffen as a co-first author, describes the challenges and solutions CPTAC researchers encountered as they integrated genomic, transcriptomic, proteomic, and clinical data from different cancer cohorts to create a single cohesive resource. The team, which includes the Getz lab, also for scientists to interact with the data, and hope that their findings and guidelines will inspire other cancer researchers to use a similar approach in their own studies. 

“There’s a lot more to be done with this data, but the biggest thing for the community is the resource we’ve created and the availability of the data,” Getz said. “We hope this will be a useful example for how other kinds of studies can integrate genomic and proteomic data, and that this will be a rich dataset for many years to come.”