Vanderbilt Researchers Discover Protein Signature Behind Colon & Rectal Cancer

Sep 04, 2014 at 03:33 pm by Staff

A scientific breakthrough at Vanderbilt is changing the future of colon cancer treatment worldwide.

Daniel Liebler, PhD, and his team at the Jim Ayers Institute for Precancer Detection and Diagnosis at the Vanderbilt-Ingram Cancer Center, have identified protein “signatures” of genetic mutations that drive colorectal cancer, currently the nation’s second leading cause of cancer deaths. Published in July’s issue of the journal Nature, the discovery is being hailed as “the first integrated ‘proteogenomic’ characterization of human cancer.”

What is Proteogenomics?

Simply put, proteomics is the study of proteins, and proteogenomics is the combination of proteomic and genomic study. In cancer, genetic abnormalities sometimes result in changes to proteins. By analyzing various genomes of cancers, researchers are generating a wealth of information about abnormalities.

In this case, researchers at Vanderbilt and five other institutions used advanced mass spectrometry to gather proteomic data on 95 human colorectal tumor samples characterized previously by The Cancer Genome Atlas. Bing Zhang, PhD, an associate professor of Biomedical Informatics at Vanderbilt, led the analysis, which integrated the proteomic data with a vast amount of pre-existing genomic data.

“The challenge is trying to understand which abnormalities in genes actually drive the characteristics or behaviors of cancer,” Liebler explained. “That’s very difficult to infer since living systems are complicated, but recent improvements in proteogenomic technology make it possible to understand abnormalities in genes, and which ones are likely to have the biggest driving roles.

Strategic Partnerships

The project first got legs in 2006, shortly after development of the Jim Ayers Institute. The Ayers family’s $10 million commitment allowed VICC to focus on development of new diagnostics for colon and rectal cancers and provided a platform to hire researchers, purchase instrumentation and compete for additional federal funding.

Around that time, the National Cancer Institute also was launching an initiative to use proteogenomic technology to advance cancer diagnostics and soon named VICC one of five large grant recipients in their Clinical Proteomic Tumor Analysis Consortium. Led by Vanderbilt University, the five CPTAC teams represent a network of seven Proteome Characterization Centers located at Washington University and the University of North Carolina; the Pacific Northwest National Laboratory; Johns Hopkins University; The Broad Institute and Fred Hutchinson Cancer Research Center.

The first five years of the initiative focused on removing technical barriers and enabling accurate and reproducible identification and quantification of proteins.

Teams were then tasked with identifying cancer biomarkers in proteins using the newly pioneered technology.While select cancer-related biomarkers like prostate-specific antigen (PSA) are well known, the measurement of individual biomarkers has clinical limitations with respect to both sensitivity and specificity. According to CPTAC researchers, existing studies of the recognized 1,000-plus cancer protein biomarker candidates derive mostly from diverse research groups working independently on available clinical specimens. Consequently, the findings are typically based on an insufficiently low number of samples to provide the adequate statistical power required for rigorous evaluation of protein changes. Relatively few of these candidates have been validated, and even fewer have made it into diagnostic products.

Producing Results

CPTAC’s discovery marks a tremendous milestone for proteogenomics research, which Liebler says experienced a 10-15 year technology lag behind the field of cancer genomics.

“Our network had to spend so much time working out technological challenges, and when those issues were largely resolved, the question was, ‘Now that you can conduct these analyses what good are proteogenomics?’” Liebler said. “It was a ‘put up or shut up’ issue, and our paper ‘put up.’”

It certainly did.

Liebler’s team substantiated proteogenomic research by identifying high impact genomic abnormalities, which help identify the strongest drivers in cancer. They also identified five subtypes of colon cancer including one believed to possess protein molecular features associated with poor outcomes and rapid spread to other tissues. “We know that colon cancers have a variety of abnormalities,” Liebler said. “Some are thought to be pretty well understood, and others not so much. In colon cancer, as well as other cancers, genomics suggests there are multiple subtypes.”

That’s very important, since colon cancer patients are typically diagnosed mid-stage while cancer is still confined to the colon. Following surgery, patients face a 20 percent risk of relapse with no reliable test to identify those most in need of aggressive chemo afterward.

“The general problem in cancer is figuring out which subtype of tumor the patient has, as that will guide the decision of who to treat with chemo and which drugs will be most effective,” Liebler said.

Liebler noted identifying a proteogenomic signature is a concept that needs to be further honed through independent studies using other tumors from other patients. Additional studies are already underway at Vanderbilt and other CPTAC facilities.

Though still early in the game, researchers are optimistic that identification of proteogenomic subtypes could provide new ideas for targets of drugs directed at different subtypes of tumors, including those in the breast and lungs.

“We figured out which proteins are indicators of which subtypes,” Liebler said. “Our next step is to measure proteins in tumors and figure out which subtypes are characteristic of which cancers. Our goal is to translate this to a diagnostic test that can be run on every colon tumor after surgery, which would help physicians make the right decision about what to do after surgery.”

Sections: Archives