Proteogenomics enhances the identification of therapeutic vulnerabilities in breast cancer
Combined approach defines new subtypes of breast cancer, yields insights into targetable pathways and that could help match tumors to therapy.
By Molly Chiu, Baylor College of Medicine
Breast cancer cells. (Credit: <a href="https://visualsonline.cancer.gov/details.cfm?imageid=10574" target="_blank">NCI</a>)
Researchers at Baylor College of Medicine, the Broad Institute of MIT and Harvard, and other institutions have applied powerful proteogenomics approaches to better understand the biological complexity of breast cancer. With this approach, the researchers were able to propose more precise diagnostics for known treatment targets, identify new tumor susceptibilities for translation into treatments for aggressive tumors and implicate new mechanisms whereby breast cancers resist treatment. The study appears in the journal Cell.
Proteogenomics combines laboratory techniques for next-generation DNA and RNA sequencing with mass spectrometry-based analysis for deep, unbiased quantification of proteins and protein modifications in cancer cells, along with computational methods for integrated analysis of this data. Such proteogenomic approaches have been extensively applied to interrogate cancers by investigators at the National Cancer Institute’s Clinical Proteomic Tumor Analysis Consortium (NCI-CPTAC).
“Importantly, our analysis included identification of phosphorylation and acetylation, protein modifications that reveal information about the activity of individual proteins. Protein acetylation had not been profiled in breast cancer before,” said co-corresponding author Matthew Ellis, a breast cancer oncologist and professor of medicine and cellular and molecular biology at Baylor College of Medicine.
Simultaneously analyzing changes in the genetic code and the resulting alterations in protein function provides a much more complete picture of what is going on inside breast cancer tumors than analyzing each component in isolation.
it included tissue samples that were collected using protocols that specifically preserve protein modifications,
it represents a larger cohort of patients,
the researchers generated genomic and proteomic profiles on exactly the same tissue fragments, and
the scientists carried out protein acetylation profiling in addition to protein phosphorylation, DNA, and RNA measurements.
Proteogenomic analytical techniques have matured substantially in recent years, and those cutting-edge approaches were applied to this dataset.
The researchers completed proteogenomic analyses of 122 treatment-naïve primary breast cancer samples. Their measurements generated a tremendous amount of data ‒ about 38,000 protein phosphorylation sites and almost 10,000 protein acetylation sites per tumor, as well as whole exome and RNA sequencing ‒ necessitating advanced computational methods for analyzing and integrating the information. “Complex analyses like these are now routinely being performed on large-scale proteogenomic data sets and we are developing tools to automate the process,” said DR Mani, a co-corresponding author and principal computational scientist at Broad.
L-R: Steven Carr, Michael Gillette, Karsten Krug, Shankha Satpathy
“We describe here proteogenomic characterization of the largest set to date of breast cancer samples that were purposefully collected for these types of analyses, maximizing the fidelity and accuracy of the results,” Ellis said. “Each tumor cell has literally hundreds of genomic changes. Mostly we don’t understand their significance either clinically or biologically. The approach we illustrate enables a deeper and more complete understanding of each individual’s breast cancer.”
For example, the analyses revealed that some subtypes of breast cancer have certain targetable enzymes called kinases that are more heavily phosphorylated than in other cancers, suggesting greater activity and therefore targetability. These analyses included recently identified drug targets such as CDK4/6 and its regulatory context, as well as programmed cell death receptors and ligands that are the targets of new immunotherapy drugs. The integrated analyses also identified new sets of estrogen receptor-positive breast cancers that could be treated with these agents. This is significant because currently these agents are restricted to estrogen receptor-negative disease.
Additional analyses raised entirely new insights into the metabolic vulnerabilities of ER+ and ER- breast cancer. “Our global analysis of the acetylproteome, the first in breast tumors, exposed new details of breast cancer subtype-specific metabolism,” said co-corresponding author Steven Carr, director of proteomics at Broad.
The researchers – including co-first authors Karsten Krug and Shankha Satpathy of Broad, Eric Jaehnig and Meenakshi Anurag of Baylor College of Medicine, Lili Blumenberg of New York University, and Alla Karpova of Washington University – hope that their findings will motivate breast cancer scientists to explore the therapeutic or diagnostic potential of the new biological alterations they have identified in this study. They also are optimistic that their findings will encourage an effort to translate proteogenomics into a cancer-profiling approach that can be used routinely in the clinic to improve diagnosis and treatment.
“We believe that proteogenomics approaches will continue to help us to identify new candidate therapeutic targets, better understand the immune landscape of breast and other cancers, gain insights into response and resistance, and ultimate progress towards our goal of personalized cancer care,” noted co-corresponding author Michael Gillette, a pulmonary and critical care physician at Massachusetts General Hospital and senior group leader in proteomics at Broad. “The science is powerful and exciting, but in the end it is what we can deliver to the patient that makes it important.”
Support for this study was provided by the National Cancer Institute.