Posted: November 21, 2011
TCGA Data Analysis Identifies a microRNA-Mediated Network of RNA-RNA Interactions Regulating Gene Expression
Pritty Patel Joshi
Glioblastoma multiforme (GBM) is the most common primary brain tumor and one of the most aggressive of all tumor types. Current treatment options for patients with GBM are limited, and despite multiple efforts to overcome the high therapeutic resistance of this disease, fewer than half of patients with GBM survive longer than two years with this cancer.
Using data from The Cancer Genome Atlas (TCGA), researchers at Columbia University recently uncovered a novel layer of gene regulation involving RNA-RNA interactions. Their work, which was published October 2011 in Cell, has the potential to answer long-standing questions about diseases like GBM.
Only a few decades ago, scientists discovered a class of noncoding RNA molecules called microRNAs. These regulatory molecules are small RNAs that silence genes by binding individual messenger RNAs and preventing the production of specific proteins. Scientists know that messenger RNAs of a single gene can affect the messenger RNA levels of another gene by essentially acting as a ‘sponge’ to soak up microRNAs that regulate both genes. However, the extent of this layer of gene regulation had not been thoroughly investigated until now.
A research team, led by Andrea Califano, Ph.D., used a computational approach in their quest to understand gene regulation in cancer. They developed a screening algorithm, Hermes, to identify a network of genes that influence microRNA-mediated gene regulation. To obtain statistically significant results with this algorithm, the researchers needed both a large number of samples and a matched set of microRNA and gene expression profiles. The publicly available TCGA datasets, the authors say, were ideal for this analysis since they are one of few options that meet these requirements. The researchers decided to study the matched GBM microRNA and gene expression profiles from TCGA.
The team’s screening algorithm uncovered a network of approximately 248,000 microRNA-mediated interactions involving over 7000 genes. The scientists were surprised to discover a regulatory network of this magnitude. Moreover, this group of microRNA-mediated regulators, or mPR network, included many established drivers of GBM including PTEN, PDGFRA, RB1, and VEGFA. Thus, the investigators propose that the mPR network represents a new avenue by which multiple oncogenic and tumor suppressor pathways could interact. To better understand the role of microRNA-mediated interactions in cancer, the researchers focused on PTEN since it is often mutated or deleted in GBM.
The scientists used different biochemical techniques to modify the levels of PTEN that could be regulated by microRNAs in GBM cell lines. They found that disrupting the balance of microRNA-mediated regulation of PTEN also altered the messenger RNA levels of 13 genes predicted to be part of the PTEN mPR network. Moreover, alterations in the expression of PTEN and its regulators affected the growth rates of GBM cell lines. These findings provide insight into how disruptions in the mPR network could lead to dysregulation of cellular functions and eventual tumor formation.
Finally, researchers found that deletions in the 13 PTEN mPR network genes often occurred in tumors that lacked PTEN expression but did not have deletions or mutations at the PTEN locus. Consequently, the researchers deduced that working through microRNAs, deletions in these mPR network genes were sufficient to halt PTEN protein activity. Strikingly, many of these PTEN mPR network genes had no prior implications in cancer.
Potential Breakthrough in the Search for Novel Therapeutic Targets
The investigators in this study focused on the microRNA-mediated regulation of the tumor suppressor PTEN in GBM. However, they emphasize that gene regulation through an mPR network does not simply apply to this tumor suppressor or tumor type. Rather this novel level of gene regulation likely contributes to the pathogenesis of many genetic diseases.
TCGA has collected and analyzed tumor samples of many different types, with the ultimate goal of characterizing the genetic and molecular changes in over 20 different cancers. Therefore, similar high throughput analyses of TCGA data from different cancer types could identify numerous molecular targets that might prove useful in the discovery of novel therapeutics.
Sumazin, P., Yang, X., Chiu, H.S., Chung, W.J., Iyer, A., Llobet-Navas, D., Rajbhandari, P., Bansal, M., Guarnieri, P., Silva, J., et al. (2011) An extensive microRNA-mediated network of RNA-RNA interactions regulates established oncogenic pathways in glioblastoma. Cell. 147(2):370-381. View PubMed abstract.