Skip to main content
Premium Trial:

Request an Annual Quote

Top Down Proteomics Identifies Protein-Level Effects of KRAS Mutations in Colorectal Cancer


NEW YORK (GenomeWeb) – Researchers from Northwestern University and the National Cancer Institute have used top-down mass spec to look at the protein-level effects of mutations in the KRAS oncogene.

Detailed in a study published this month in Proceedings of the National Academy of Sciences, the analysis identified several mutation-linked post-translational modifications involved in KRAS activation and membrane attachment, demonstrating the potential of top-down mass spec as a tool for understanding the functional consequences of genetic mutations.

The approach is broadly applicable to other oncogenes and members of cancer signaling pathways, said Neil Kelleher, professor of chemistry at Northwestern and senior author on the study. He added that he now hoped to extend the analysis to other members of the KRAS signaling network.

The PNAS work "is a roadmap for the future to show that you can just go down the KRAS signaling cascade and show how the proteoforms [expressed] change [due to mutations]," he said.

Kelleher is a leading expert in top-down proteomics, which uses mass spectrometry to look at intact proteins. This is in contrast to the more commonly used bottom-up approach, wherein proteins are digested into peptides before mass spec analysis.

Compared to bottom-up proteomics, top-down provides more comprehensive information on the different forms — termed proteoforms — a given gene product can take depending on the particular post-translational modifications or genetic mutations affecting it. However, top-down is more technically challenging and less sensitive than conventional bottom-up approaches, and is, at this point, used by a relatively small group of expert labs.

Kelleher said, though, that his lab's KRAS analysis had drawn more widespread interest than much of his previous top-down work, which has largely focused on analysis of large numbers of intact proteins, as opposed to targeted analysis of the proteoforms generated by a specific oncogene.

"If I measure 10,000 proteoforms, people don't [care]," Kelleher said. "But we measure 11 KRAS4b proteoforms, people go berserk."

In the study, Kelleher and his colleagues used antibody-based enrichment of KRAS followed by top-down analysis on a Thermo Fisher Scientific Q Exactive HF instrument to look at mutated and wildtype KRAS proteoforms in colorectal cancer (CRC) cell lines and CRC patient samples. They looked specifically at proteoforms of the KRAS4b isoform either with or without the Gly13Asp mutation, which is commonly found in CRC.

In the cell line experiments, the researchers observed that upon knockout of the mutated Gly13Asp allele in lines that were heterozygous for mutated KRAS4b, the remaining wildtype proteoforms saw complete nitrosylation of their cysteine 118 residues.

This nitrosylation was previously known, Kelleher said, but, he noted, it was "not known to be at 100 percent stoichiometry in certain mutational backgrounds."

Looking at the top-down data, "it was just staring you in the face," he said. "In one genetic background you had a nitrosylation on 95 percent or more of the cysteine 118 residues, and in another genetic background you had 0 percent."

That finding, Kelleher said, demonstrated the ability of top-down analysis "to serve as a filter for which [protein modifications] are dynamic, for which ones matter."

In the case of the cysteine nitrosylation, the researchers hypothesized its presence might lead to lower levels of KRAS-mediated signaling and lower levels of proliferation than lines with the mutated Gly13Asp allele.

Looking at KRAS4b proteoforms in CRC patient samples, the researchers observed differences in the C-terminal processing of these proteoforms. Specifically, they found a proportion of proteoforms missing the C-terminal COOMe modification, which previous studies have demonstrated is involved in membrane association of KRAS. Looking at samples from two patients, they found that the one with earlier-stage disease (IIA) had a higher percentage of KRAS4b (69 percent) with the COOMe modification than did the patient with later-stage (IIIB) disease (27 percent with COOMe).

Looking at a pair of tumor samples — one with a glycine 12 mutation/wildtype genotype and one that was WT/WT, they determined that 9 percent of the mutant KRAS4b had the C-terminal modification, while 49 percent of the wildtype protein did. The researcher suggested that loss of the COOMe modification might affect subcellular localization of KRAS, altering downstream signaling.

The findings, Kelleher said, demonstrate the potential of top-down mass spec to distinguish protein changes that are relevant to a disease process from those that are not.

"You look at any protein, and it's like, well, there are ten, 20, 200 [post-translational modifications] that are possible," he said. "But which ones matter? And in what context?"

"This study demonstrates a proof-of-concept and real world example of a novel workflow to filter [post-translational modifications]," added Henry Rodriguez, director of the NCI's Office of Cancer Clinical Proteomics Research and a co-author on the paper. "Using a targeted top-down proteomic workflow, the team was able to detect PTM profiles to KRAS4b isolated from both human colorectal cancer cell lines and tumor samples with KRAS gene mutations." 

"This is important because direct experimental evidence at the cellular or biochemical level for different functions of any RAS isoform, including K-Ras4A versus K-Ras4B, is still lacking," he said. "This approach could help overcome that hurdle."

Kelleher noted that the analysis he and his colleagues did in the PNAS study would be essentially impossible using conventional bottom-up protein analysis.

To begin with, he said, there are three RAS genes, and they are 90 percent identical. This means that there is significant overlap among the tryptic peptides that comprise the RAS proteins, making it difficult to determine which peptide came from which protein.

Additionally, the KRAS gene produces two isoforms, KRAS4a and KRAS4b, each of which have different biology.

"With bottom-up, it's very tricky to get isoform specific information," Kelleher said.

Having identified post-translational modifications and modification sites of interest, though, researchers can now develop targeted mass spec assays to these specific portions of the protein, he said. "Now that you know about these KRAS4b proteoforms, you can devise the right targeted assays to measure the pieces [of interest]."

He added that he views the KRAS study as the first step in a larger effort to apply top down to a range of proteins known to be involved in cancer signaling.

"This is a road map for the future, where you can just go down these KRAS signaling cascades, these signaling networks, and show how the proteoforms change," he said.