Micro RNA homeostasis: Its stability and degradation
By- Subodh Kumar Sinha
National Research Centre on Plant Biotechnology, New Delhi

miRNAs are important for gene regulation in numerous cellular and developmental processes, therefore they themselves are supposed to be subjected to extensive regulation. On the contrary much attention has not been paid towards regulation of miRNA levels through degradation of the mature and functional miRNA. This is partly because of the perception of miRNAs being inherently stable molecules, consistent with the finding that mature miRNAs persist for many hours or even days after their production is stopped. Nonetheless, many miRNAs show a dynamic expression pattern during development, including rapid downregulation in some instances. Moreover, specific mature miRNAs have been found to be expressed in a tissue- or stage-specific manner without variation in the expression pattern of the precursor forms (pri- and pre-miRNAs), supporting the notion of regulatory mechanisms acting on the mature miRNA. These findings suggest that steady-state levels of miRNAs can be regulated through both biosynthetic and degradation processes.

Cellular conditions affecting miRNA stability

Recent studies have shown that individual miRNAs, or miRNAs in specific environments, are subject to accelerated decay thereby altering miRNA levels and hence activity.

Several miRNA families target, for example, components of cyclin/CDK complexes and hence function in cell cycle regulation. Interestingly, the reverse is also true; that is, cell cycle stage affects accumulation of certain miRNAs, for instance miR-29b. In HeLa cells, miR-29b is polycistronically transcribed together with its 'sister' miR-29a, from which it differs by a nucleotide at position 10 as well as its six 3'-terminal nucleotides. However, whereas miR-29a levels change little during progression through the cell cycle, miR-29b is enriched in mitotic cells. 'Pulse-chase' experiments using transfection revealed a half-life of miR-29b of 4 hr in cycling cells, compared to more than 12 hr in mitotically arrested cells, whereas miR-29a has a half-life of more than 12 hr in either case. Mutational analysis suggested that the uracils at nucleotide positions 9-11 are necessary, although not sufficient, for the fast degradation of miR-29b. Expression of variant pre-miRNA from transgenes revealed that residues in the seed region and at the 3'- end coordinately destabilize miR-503.

In human MCF10A immortalized breast epithelial cells, the levels of several miRNAs rapidly decreased upon epidermal growth factor (EGF) stimulation. After MCF10A cells had been starved of serum to arrest their proliferation, stimulation by addition of EGF caused a reduction by ≥ 50% of 23 miRNAs within 1 h. This suggests that rapid miRNA downregulation contributes to the physiological responses (i.e., proliferation or migration) of a cell to EGF. However, it remains unclear whether EGF acts by inducing miRNA degradation or, alters transcription or processing of inherently unstable miRNAs.

miRNA-degrading enzymes

Several miRNA-degrading enzymes such as 3' to 5' and 5'to 3' exoribonucleases have been reported, however similar results have not been achieved so far for endoribonuclease. Distinct ribonucleases were found to function in turnover of different sets of miRNAs and/or different organisms, but their substrate specificity and phylogenetic conservation remains largely unknown.

Small RNA degrading nucleases (SDNs)

In Arabidopsis thaliana active degradation of miRNAs is mediated by the small RNA degrading nucleases (SDNs). Experiments with recombinant SDN1 and synthetic miRNAs have revealed that SDN1 uses a 3'-to-5' exonucleolytic mechanism, yielding a final degradation product of 8-9 nt. SDN1 can degrade single-stranded RNA in the range of 17-27 nt with comparable efficiency, but not pre-miRNAs, longer RNAs, double-stranded RNA or single stranded DNA. In vivo, plant miRNAs are 2'-O-methylated at their 3'-ends; which slowed down but did not prevent miRNA degradation by SDN1 in vitro. 2'-O-methylation by the methyltransferase HEN1 (HUA ENHANCER1) also stabilizes miRNAs in vivo by preventing 3'-end oligouridylation by HESO1 (HEN1 SUPPRESSOR1), a terminal nucleotidyl transferase. However, because uridylation, at least in vitro, failed to promote and in fact attenuated SDN1-mediated degradation, it appears that uridylation influences miRNA degradation through distinct enzymes that remain to be identified. Nevertheless, it remains to be shown whether HEN1 is used as a physiological regulator of miRNA degradation. Uridylation of miRNAs and siRNAs also contributes to their decay in the green alga Chlamydomonas reinhardtii. The terminal nucleotidyl transferase MUT68 was found to uridylate the 3'-ends of these small RNAs in vivo and to stimulate their degradation by RRP6 (ribosomal RNA-processing protein 6), a component of the 3'-to-5' exosome RNase complex, in vitro. Furthermore, depletion of RRP6 elevated miRNA and siRNA levels in vivo. The 2'-O-methyl group present on endogenous C. reinhardtii miRNAs prevented both uridylation and degradation in vitro.

Regulatory functions of target RNAs on miRNAs

The extent of sequence complementarity between miRNA and mRNA determines the mode of mRNA silencing. Extensive complementarity, reminiscent of the siRNA-mRNA interaction, can result in endonucleolytic cleavage of the target mRNA which constitutes a major means by which miRNAs regulate mRNAs in plants. In metazoans, miRNAs base-pair with mRNAs mainly through partial complementarity, resulting in translational repression or exonucleolytic degradation. Recent studies now provide evidence for reciprocal regulation, such that target RNAs can modulate miRNA stability.

In flies, mice, and human HeLa and HEK293T cells, miRNAs are destabilized if they are supplied with an artificial target exhibiting extensive complementarity. The decline of a miRNA in the presence of a highly complementary target is accompanied by the emergence of longer ('tailed'; typically multiple or individual added uridines or adenosines) and shorter ('trimmed' species of the original miRNA). At this point, it is not known whether tailing precedes trimming, or rather defines a separate miRNA fate upon binding highly complementary targets. To solve this puzzle, the identification of the enzymes which mediates tailing and trimming in a target dependent manner is required. Although Drosophila Nibbler, a member of the DEDD family of exonucleases, trims the 3'-ends of some miRNAs by a few nucleotides, it does not appear to function in miRNA turnover. However, miRNA 'seed' binding sites, that is, those with complementarity to nucleotides 2-8 of the miRNA only, do not induce tailing and trimming. Hence, because miRNA complementarity is limited to the seed for most endogenous targets, these targets will not usually induce miRNA degradation.

Target-induced miRNA degradation in plants

Unlike in animals, plant target mRNAs are frequently highly complementary to their cognate miRNAs. Moreover, there is precedence for the idea of tailing and trimming, which occurs in A. thaliana when 2'-O-methylation of small RNA 3'- termini is lost through mutation of hen1. The tails almost exclusively consist of uridines and also occur on trimmed small RNAs. Although it remains to be shown that endogenous targets can indeed induce plant miRNA degradation, artificial, highly complementary target RNAs containing two target sites were found to cause a severe reduction of cognate miRNA levels. Further, no change in primary miRNA levels confirmed that the effect was post-transcriptional, and partial restoration of mature miRNA levels in sdn1 sdn2 double mutant plants provided further evidence that targets acted by inducing miRNA degradation.

Target-mediated miRNA protection

In contrast with target-induced degradation of miRNAs, target mRNAs in C. elegans have been found to stabilize miRNAs in vivo by preventing their release from AGO proteins. It has been found that reduced availability of endogenous targets decreased accumulation of the cognate miRNAs, whereas miRNA levels increased in the presence of artificial target RNAs. This process, termed target-mediated miRNA protection (TMMP) counteracts miRNA decay mediated by XRN-1 and XRN-2. Together, miRNA decay and TMMP could thus serve as a proofreading mechanism that ensures preferential occupation of AGO with functional, that is, target engaged miRNA.

References:

Ruegger and Grobhans (2012) Trends Biochem. Sci. 37: 436-446.
Meng et al. (2011) Pl. Physiol. 157:1583-1595.

About Author / Additional Info: