Investigation of Mechanisms and Sequence Requirements in Arabidopsis MiR159-mediated Gene Silencing

📖 The Scoop

MicroRNAs (miRNAs) are critical regulators of gene expression in both animals and plants. Although rapid progress has been achieved in identifying endogenous miRNAs, and artificial miRNAs have been utilized extensively in gene silencing experiments, debate and doubt still linger over the general principles of miRNA target recognition and their mode(s) of action in plants. It is widely accepted that animal miRNAs, which generally have central mismatches to targets, act through mRNA decay and translational repression. In contrast, plant miRNAs have extensive complementarity and perfect matches around the cleavage site, thus are thought to act predominantly through mRNA cleavage, and to a lesser extent translational repression. However, there is emerging evidence that translational repression could also be widespread. Additionally, many known rules for target recognition are constantly challenged as our understanding of miRNA progresses, therefore needs further investigation. In this thesis, the highly conserved Arabidopsis miRI59-GAMYB regulatory module was used as a model to address fundamental questions regarding plant miRNAmediated gene silencing. Firstly, to examine the sequence requirements of miR159mediated silencing, different artificial miR159 variants were used to complement a loss-of- function mir159ab double mutant, which exhibits pleiotropic developmental defects caused by the deregulation of its major target genes, MYB33/MYB65. Contrary to expectation, miR159 variants that contained mismatches to nucleotides around the cleavage site (positions 10 and 11) potently silenced MYB33/MYB65 to phenotypically inconsequential levels. Although these miR159 variants could still cleave MYB33/MYB65 transcripts, cleavage was likely to be compromised, as the steady state transcript levels for MYB33/MYB65 increased with the increasing number of mismatches at the cleavage site. However, these un-cleaved MYB33 and MYB65 transcripts were shown to be fully repressed by a non-degradative mechanism. This mechanism must be extremely efficient, as expression of a MYB33 transgene that resulted in MYB33 transcripts levels 10-15 fold higher than endogenous MYB33 levels, was still fully suppressed. Likewise, for the most divergent GAMYB-like miR159 target MYB120, analysis showed that although it could be cleaved, very high levels of noncleaved MYB120 transcripts generated through a transgenic approach, were again totally silenced through a non-degradative mechanism. Supporting a notion that this mechanism inhibits translation, the proportion of MYB33 and MYB65 transcripts increased dramatically in the polyribosome (actively translated) fractions of the mir159ab mutant, when compared to wild-type. Together these data imply that perfect central complementarity is not required for plant miRNA-mediated gene silencing, and even if miRl59-mediated cleavage is compromised, a non-degradative "translational" repression mechanism confers full silencing. Either way these results suggest that the default state of miR159 is to regulate its targets through degradative (mRNA cleavage) and non-degradative (translational repression) mechanisms. However, at odds with the miR159 variants, was the expression of MYB33-1ml2m transgenes. Although these transgenes had central mismatches to miR159 that mirrored the miR159 variant:endogenous MYB33/MYB65 scenario, they were found not to be efficiently silenced by miR159. In-depth molecular analysis suggests that factors other than sequence complementarity are critical for the in vivo miRNA:target silencing outcome. Firstly, the ratio of the miRNA:target abundance is likely to play a significant role. As the expression of either the MYB33-1m12m transgene resulted in much higher transcript levels than the endogenous MYB33 gene (l0-15 fold higher), a lower miRl59:target ratio is likely to compromise silencing. Secondly, mutation of the miRNA binding site within a target gene will not only alter complementarity, but also target site structure, which may influence recognition, and which has been subsequently shown. Therefore I propose a working model which takes these two factors into consideration along with complementarity, and together these determine the silencing outcome of a plant miRNA:target relationship.

Genre: No Category (fancy, right?)