NEW YORK (GenomeWeb News) – A whole-exome sequencing and copy number variation study of pancreatic cancer published online today in Nature suggests that the disease sometimes involves alterations to genes and pathways best known for their role in axon guidance during embryonic development.
The work was conducted as part of the International Cancer Genome Consortium effort by researchers with the BCM Cancer Genome Project, the Australian Pancreatic Cancer Genome Initiative, and the Ontario Institute for Cancer Research Pancreatic Cancer Genome Study.
As they reported today, the investigators identified thousands of somatic mutations and copy number alterations in pancreatic ductal adenocarcinoma cancer, the most common form of pancreatic cancer. Some of the mutations affected known cancer genes and/or pathways implicated in pancreatic cancer in the past. Other genetic glitches pointed to processes not previously linked to the disease including mutations to axon guidance genes such as SLIT2, ROBO1, and ROBO2.
"This is a category of genes not previously linked to pancreatic cancer," Baylor College of Medicine researcher William Fisher, a co-author on the new paper, said in a statement. "We are poised to jump on this gene list and do some exciting things."
Pancreatic cancer is among the deadliest types of cancer, he and his colleagues explained, with a grim five-year survival rate of less than 5 percent. But despite its clinical importance, direct genomic studies of primary tumors had been stymied in the past due to difficulties obtaining large enough samples for such analyses.
"Genomic characterization of pancreatic ductal adenocarcinoma, which accounts for over 90 [percent] of pancreatic cancer, has so far focused on targeted polymerase chain reaction-based exome sequencing of primary and metastatic lesions propagated as xenografts or cell lines," the study authors noted.
"A deeper understanding of the underlying molecular pathophysiology of the clinical disease is needed to advance the development of effective therapeutic and early detection strategies," they added.
For the current study, researchers started with a set of tumor-normal samples from 142 individuals with stage I or stage II sporadic pancreatic ductal adenocarcinoma. Following a series of experiments to assess tumor cellularity and other features that can impact tumor analyses, they selected 99 patients whose samples were assessed in detail.
For whole-exome sequencing experiments, the investigators nabbed coding sequences from matched tumor and normal samples using either Agilent SureSelectII or Nimblegen capture kits before sequencing the exomes on SOLiD 4 or Illumina sequencing platforms. They also used Ion Torrent and Roche 454 platforms to validate apparent somatic mutations in the samples.
For its copy number analyses, meanwhile, the team tested the pancreatic cancer and normal tissue samples using Illumina HumanOmni1 Quad genotyping arrays.
When they sifted through data for the 99 most completely characterized pancreatic tumors, researchers uncovered 1,628 CNVs and roughly 2,000 non-silent, somatic coding mutations. More than 1,500 of the non-silent mutations were subsequently verified through additional sequencing experiments.
On average, each of the tumors contained 26 coding mutations. And despite the variability in mutations present from one tumor to the next, researchers identified 16 genes that were mutated in multiple tumor samples.
Some were well-known cancer players such as KRAS, which was mutated in more than 90 percent of the 142 pancreatic tumors considered initially. Several other genes belonged to cell cycle checkpoint, apoptosis, blood vessel formation, and cell signaling pathways, researchers reported, or to pathways involved in chromatin remodeling or DNA damage repair.
For example, some 8 percent of tumors contained mutations to ATM, a gene participating in a DNA damage repair pathway that includes the ovarian/breast cancer risk gene BRCA1.
Genes falling within axon guidance pathways turned up as well. That pattern was supported by the researchers analyses of data from published pancreatic cancer studies — including two studies based on mutagenesis screens in mouse models of the disease — and by their own gene expression experiments in mice.
The team also tracked down a few more pancreatic ductal adenocarcinoma cases involving mutations to axon guidance genes such as ROBO1, ROBO2, and SLIT2 through targeted testing on 30 more pancreatic cancer patients.
The findings are consistent with those found in some other cancer types, according to the study's authors, who noted that there is evidence indicating that some axon guidance components feed into signaling pathways related to cancer development, such as the WNT signaling pathway. If so, they explained, it's possible that mutations to axon guidance genes might influence the effectiveness of therapies targeting such downstream pathways or serve as potential treatment targets themselves.
Still, those involved in the study cautioned that more research is needed not only to explore such possibilities but also to distinguish between driver and passenger mutations in pancreatic cancer.
"The potential therapeutic strategies identified will … require testing in appropriate clinical trials that are specifically designed to target subsets of patients stratified according to well-defined molecular markers," the study's authors concluded.