Competing endogenous RNAs in microRNA regulation
As described in the previous section, a single mRNA can contain binding
sites for multiple miRNAs. Similarly, a single miRNA can target multiple
mRNAs. However, a set of non-coding transcripts that possess miRNA
response elements (MREs) similar to target mRNA, bind to the miRNAs and
prevent them from acting on target mRNA.108 These
non-coding transcripts control the amount of miRNA available for target
mRNA, thereby indirectly modulating the target protein expression. RNA
transcripts that share a binding site for the same miRNA and regulates
each other’s expression by competing for that miRNA are called competing
RNAs or ceRNAs. These ceRNAs are present in various forms such as long
coding RNAs, circular RNAs, pseudogenes or they could also be
protein-coding mRNAs.10 CeRNAs provides an additional
layer of complexity to the miRNA mediated gene regulation. Experimental
studies indicate that they have major implications in diseases like
cancer.109–111 For example, PTENP1 is a pseudogene
for PTEN and they share similar MREs in the 3’UTR region. PTENP1
captures miRNAs targeting PTEN and helps in its function as a tumor
suppressor. PTENP1 is found to be deleted in various cancers and
overexpression of PTENP1 resulted in increased PTEN expression and
induced growth inhibition by sequestering the
miRNAs.111–113 Several non-coding transcripts that
regulate PTEN are found to be expressed in a correlated manner in colon
cancer cells, where PTEN expression or copy number is
altered.112
Philip et al114 created artificial miRNA sponges with
multiple binding sites for miR-20a and found that the addition of
miR-20a sponges significantly alters the miR-20a target gene
expression.114 Studies also suggest that miRNA
activity does not depend on its levels, but the relative abundance of
target sites can appreciably influence its
activity.9,115–117 However, the ceRNA hypothesis
remains controversial because miRNA target sites are usually present in
abundance, and alteration in the level of one target mRNA might not
considerably affect other miRNA target’s expression
level.115,116 Boson et al117quantified the concentration of endogenous target and miRNAs using
biochemical techniques such as mRNA-seq, small RNA- ]seq, and iCLIP in
mouse cell lines to calculate miRNA: target ratio. They essentially
investigated how modulating the ratio using ceRNAs of varying affinity
affects miRNA binding.117 They have shown that ceRNAs
with high affinity than the target can derepress the miRNA targets only
for miRNA families with low miRNA: target ratio. If the miRNA: target
ratio is high or the ceRNA affinity is low, then there is no noteworthy
effect of ceRNA regulation on the target. They proposed a simple target
competition model with hierarchical affinity for targets, which predicts
that miRNA: target ratio determines the susceptibility to ceRNA
competition at identical level of high affinity ceRNA overexpression.117 However, this aspect of miRNA regulation has to be
investigated in a much-detailed manner using theoretical or
computational modeling.