Competing endogenous RNA networks: tying the essential knots for cancer biology and therapeutics

Abstract

A recently discovered dimension of post-transcriptional gene regulation involves co-regulatory crosstalk between RNA transcripts, which compete for common pools of microRNA (miRNA) molecules. These competing endogenous RNAs (ceRNAs), or natural miRNA sponges, have an active role in regulating miRNA availability within the cell and form intertwined regulatory networks. Recent reports have implicated diverse RNA species including protein-coding messenger RNAs and non-coding RNAs as ceRNAs in human development and diseases including human cancer. In this review, we discuss the most recent discoveries that implicate natural miRNA decoys in human cancer biology, as well as exciting advances in the study of ceRNA networks and dynamics. The structure and topology of intricate genome-scale ceRNA networks can be predicted computationally, and their dynamic response to fluctuations in ceRNA and miRNA levels can be studied via mathematical modeling. Additionally, the development of new methods to quantitatively determine absolute expression levels of miRNA and ceRNA molecules have expanded the capacity to accurately study the efficiency of ceRNA crosstalk in diverse biological models. These major milestones are of critical importance to identify key components of ceRNA regulatory networks that could aid the development of new approaches to cancer diagnostics and oligonucleotide-based therapeutics.

Figure 1

Involvement of ceRNA-mediated regulation in human cancers. Schematic representation of a simplified ceRNA network with two transcripts and one miRNA (blue circle), and the different cancers for which ceRNA activity has been experimentally verified (orange circles). Different ceRNA subnetworks are represented by squares, where ceRNAs (in blue) interact among each other by binding and competing for common miRNA molecule pools (in green). Validated ceRNA interactions have been reported for breast cancer [39,72-74], melanoma [35], endometrial cancer [38], glioblastoma [36], liver cancer [27,74,75], gastric cancer [40,60], colorectal cancer [17,34,76], prostate cancer [17,34,39], lung cancer [39,55], and lymphoma [41].

Figure 2

Molecular dynamics of ceRNA interactions. (A) Optimal molecular conditions for effective ceRNA crosstalk depend on miRNA/ceRNA ratios. Schematic representation of a simplified ceRNA network involving two transcripts (with blue backbones) harboring two different MRE sites (orange and green circles) that bind to miRNA1 (orange) and miRNA2 (green), respectively. Variations in ceRNA expression levels would generate a response in the expression of co-regulated ceRNAs only in a narrow miRNA concentration window, and only for a range of ceRNA concentration. When miRNA concentration is in excess, the titration capacity of competing MRE is diminished and no effect would be observed on ceRNA crosstalk. Conversely, in excess of ceRNA, most miRNA molecules will bind their MRE sites, and the system would be insensitive to changes in the relative miRNA/ceRNA ratios. (B) The presence of indirect interactions can amplify ceRNA crosstalk. Schematic representation of two ceRNA networks involving three ceRNAs and two (top panel) or three (bottom panel) miRNAs. When ceRNA 2 is able to co-regulate ceRNA 3 by sequestering miRNA 3, the effect of ceRNA 1 upregulation on ceRNA 2 and ceRNA 3 levels may be amplified by this secondary crosstalk. This is termed an indirect interaction as it does not involve a direct relationship between ceRNA 1 and either ceRNA 2 or ceRNA 3.