A Multilevel Pan-cancer Map Links Gene Mutations to Cancer Hallmarks

Sep. 29, 2015

3 Bullets:

  • Differences at the molecular level enable personalized treatment regimes for cancer patients, whereas, at a higher functional level, a set of “cancer hallmarks” define characteristics observed commonly across all cancer types.
  • Scientists at ISB have developed a multi-level map that links molecular alterations associated with different cancer types to the “pan-cancer” hallmarks with the aim to enhance understanding of the underlying biological mechanisms and encourage ideas for newer therapies.
  • In confirmation with previous studies, a higher percentage of mutually exclusive (ME) mutations were observed at the level of pathway. On the other hand, many co-occurring (CO) mutations were observed at the level of hallmarks, implying that deregulation of multiple distinct biological functions by CO mutations may be necessary to acquire certain hallmark characteristics.

molecules-hallmarks2

By Varsha Dhankani

In the scientific community, cancer is not considered a single disease but a multitude of diseases. Different genes and molecular pathways have been associated with different types of cancer. These differences at the molecular level are being leveraged to enable personalized treatment regimes for cancer patients.

However, at a higher functional level, across all the different cancer types, a set of common denominators has been defined and referred to as the “cancer hallmarks.” The cancer hallmarks identify deregulated biological processes that are observed commonly across all the cancer types.

Title: A multilevel pan-cancer map links gene mutations to cancer hallmarks
Journal: Chinese Journal of Cancer
Authors: Theo Knijnenburg, Tycho Bismeijer, Lodewyk Wessels, Ilya Shmulevich
LinkRead paper

Scientists at ISB wanted to decipher how tumors that are so different at the molecular level are similar when observed at a higher functional level. In order to answer this question, they have developed a multi-level map that links genes associated with different cancer types to the pan-cancer hallmarks. By creating a map that links molecular entities to phenotypic cancer hallmarks, the hope is to enhance understanding of the underlying biological mechanisms and encourage ideas for newer therapies.

This initiative was undertaken as part of The Cancer Genome Atlas (TCGA) Pan-Cancer Analysis which aims to study similarities and differences among the molecular alterations found in the first dozen tumor types profiled by TCGA. Scientists used exome-sequencing data and copy number variation data for 2740 tumor samples from 10 different cancer types to indicate mutation status of genes (mutated or not mutated) per sample.

The multi-level mapping from genes to cancer hallmarks allowed scientists to study the role of mutually exclusive (ME) mutations and co-occurring (CO) mutations in enabling cancer hallmark characteristics. Previous studies have shown that gene mutations that affect a pathway tend to be altered in an ME pattern, i.e. they tend not to occur mutated together in the same patient. Interestingly, in this study the authors found many CO mutations at the level of hallmarks, i.e. genes that are frequently found mutated together across cancer samples. The rationale behind these observations is that a single mutated gene in a  pathway might be sufficient to confer a selective advantage to the cancer cell, whereas the deregulation of multiple distinct biological functions by CO mutations may be necessary to acquire certain hallmark characteristics.

Also of interest are the much less frequently observed mutations that are difficult to validate statistically. By combining data across 10 different tumor types and linking the less frequently mutated genes to pathways and cancer hallmarks in the multi-level map, the authors  aimed to elucidate the potential role of these genes in cancer.

Scientists also observed a large variation in the number of genes and pathways connected to each of the hallmarks. Hallmarks like “resisting cell death”, “sustaining proliferative signaling”, and “tissue invasion and metastasis” were potentially enabled by more than 300 genes (50 different pathways). In contrast, the hallmarks “replicative immortality” and “reprogramming energy metabolism” appeared to be enabled by just 1 or 2 pathways. As expected, some well known cancer genes like PIK3CA and TP53 are hubs in the multi-level map, but most others are connected to just a handful of pathways and hallmarks.

The multi-level map linking genes via pathways to cancer hallmarks provides a conceptual and intuitive comprehension of large cancer genomic data sets. Domain experts can utilize such mappings to formulate novel hypotheses and therapies informed by the elucidated functional role of molecular alterations underlying cancer.