Skip to main content
Premium Trial:

Request an Annual Quote

Pan-Cancer Mutation Study Identifies Protein Kinases Key to Tumor Suppression

Premium

NEW YORK (GenomeWeb) – A team led by researchers from the University of Manchester and the National Cancer Institute have used pan-cancer mutation data to identify protein kinases involved in tumor suppression.

In a study published this week in Science Signaling, the researchers used data from the Cancer Cell Line Encyclopedia and the Cancer Genome Atlas to identify mutational hotspots in kinase regions that would contribute to a loss of function. By screening these hotspots against the full complement of 411 human kinases, they were able to identify a number of functionally important mutations that would have fallen below the signal-to-noise threshold of a conventional mutational analysis, said John Brognard, a principal investigator at the NCI and senior author on the paper.

The study, he noted, fits into a larger emerging trend of researchers looking to protein-level information to help prioritize the vast number of genetic mutations being identified through techniques like next-generation sequencing.

Kinases are well suited to such an approach, given that they have been extensively studied and are broadly implicated in cancer signaling.

"Kinases are so well described that we know that certain amino acids are required for their catalytic activity and then also for them to bind to their substrates," Brognard said. "So, we can basically look in a pan-cancer way through the TCGA data for mutations at those exact residues, and that gives us kind of a snapshot of what kinases are inactivated by somatic mutations in cancer."

The researchers began by assessing the frequency of truncation mutations in the TCGA and CCLE datasets that occurred near or within kinase domain motifs, indicating that they would inactivate those proteins. They used this information to generate a list of candidate tumor-suppressing kinases, and then used sequence alignment of these candidates to identify regions conserved across these kinases that contained mutational hotspots, identifying 15 top hotspot residues. They then analyzed the TCGA and CCLE datasets to rank the 411 canonical human kinases by the frequency of mutations found in these 15 hotspots.

By analyzing not just the frequency of mutations throughout the cancer genome, but specifically the mutations that would inactivate protein kinases, the researchers were able to identify potential driver mutations they might have otherwise overlooked.

"Usually you look and you ask, what is the frequency of mutations in a gene," Brognard said. "If there's a high frequency, in like p53, for example, then you say, obviously this is an important gene for cancer."

The kinase mutations investigated in the Science Signaling study "wouldn't rise above that mutational noise and be characterized as something really important to look at," he said. "But, when you ask where they are occurring, they are occurring at these hotspots and conserved residues and within a family of kinases."

"You put that together, and you see that a [signaling] pathway has mutations that inactivate it in 20 percent [of a given] cancer," he added. "So, all of a sudden it becomes very important."

The work is part of a broader move to improve analysis of genomic mutations by adding protein-level data. For instance, last month, researchers from the Swiss Federal Institute of Technology Zurich and the German Cancer Research Center published a study in which they used mutation data from TCGA and the International Cancer Genome Consortium to identify 180 hotspot amino acid residues in 160 proteins. Of those 160 proteins, 66 percent were not coded by known cancer driver genes.

More generally, a number of initiatives like the NCI's Clinical Proteomic Tumor Analysis Consortium (CPTAC) and the Applied Proteogenomics Organizational Learning and Outcomes (APOLLO) Network, part of the US government's Cancer Moonshot initiative, are applying proteogenomic techniques to, in part, help better identify important mutations in large cancer sequencing data sets.

"It's an approach that is growing within the field, and people are starting to apply it more broadly now," Brognard said. "I think it really helps us to decipher and mine this [mutation] data and figure out what's relevant."

Analyzing how mutations contribute to kinase loss-of-function specifically is particularly relevant to cancer, given the widespread use of kinase inhibitors in cancer treatment, Brognard said.

For instance, he noted, there are cancers where patients are given EGFR inhibitors as a standard first-line therapy. However, if a patient already has an EGFR mutation causing a loss of function, that treatment will likely be ineffective.

"It's good to make the cancer community aware that sometimes some of these oncogenes are inactivated, and these are patients you would never stratify for treatment with certain approved kinase inhibitors," he said.

While kinases are one of the most commonly studied classes of protein in cancer research, much of this work has focused on a relatively narrow set of analytes, Brognard said, noting that the recent study identifies some less studied kinases that could be worth investigating.

For instance, he said, the protein MYO3A was one of the top hits identified in their screen, but "no one has published any work on MYO3A [in cancer]."

"So, [the approach] really reveals novel kinases to explore that may have a tumor-suppressive function in cancer," Brognard said.

The study also provided new insights into the function of established oncogenes, he noted. For instance, one of the study's top hits was the oncogene BRAF. Looking at the effect of loss-of-function mutations in this kinase, the researchers observed that inactivation of BRAF "can paradoxically activate the MEK-ERK pathway to drive cancer," Brognard said, suggesting that a similar dynamic could be at work with other cancer-associated proteins like EGFR.

"There's a lot of interesting biology to explore in that sense, as well," he said.

Moving forward, Brognard said his lab plans to follow up on little-studied hits like MYO3A to better understand their role in cancer. They also plan to do a similar screen looking for hotspots leading to gain-of-function mutations.

Similar to the Science Signaling study, such an analysis could identify activating mutations in kinases that would be missed by a less directed approach, Brognard said, adding that these mutations might suggest kinase inhibitors that could be useful for treating a given cancer or particular kinases that could be novel drug targets.