It is thought that miRNA's bind to the 3'UTRs (more microRNA binding sites) but complementary sequence can be find within the coding regions as well, the statistical analysis on the imperfect binding of seed sequences in microRNAs show that most significant targets are within the 3'UTRs, of course the targets are identified using mathematical/ statistical formula and needs further investigation whether it actually targets that particular mRNA. There are quite a few papers describing online databases used to identify targets and also about actual interactions of microRNA and mRNA's. The question is if it actually bind to the other regions (which is possible) does it still have the potential to repress the translation/degrade the mRNA. Hope that helps a bit

Article Computational Methods for MicroRNA Target Prediction

Article Imperfect centered miRNA binding sites are common and can me...

This is a confusing topic. The key probably is functionality. Perhaps microRNA/RISCs bind non-canonical and canonical sites in non-3'UTR regions but does that have any impact on this mRNA? This is something I am struggling with since RISC-IP data show a considerable number of non-canonical sites and sites outside of the 3'UTR. And some of these mRNAs appear to be inversely correlated with the microRNA (starbase v2.0: http://starbase.sysu.edu.cn/).

To play it conservatively, the rule of thumb is still: microRNAs mainly bind to the 3'UTR and the most effective sites are 8mers/canonical sites. As always with rules, exceptions may exist and once there are too many exceptions the rules will be rewritten. The prototype algorithm following this rule is Targetscan; and I still prefer to use it since it is clean, logic, based on lots of data, and very dogmatic. Therefore, it is great to see the doubt, confusion on this topic, to see your question. Isn't that what drives progress, our curiosity and desire to clarify things? Even if things after "clarification" get messier and the dogma may fall one day. Sorry for the unclear answer.

MicroRNA as pointed out and clearly demonstrated via multiple parallel sequencing can bind to 3'UTR, 5'UTR and coding regions too. You have to enlarge the classical (old) vision of microRNA as simple regulators of transcription. Indeed, microRNA are also potent regulator of translation and these effects are responsible for the regulation of gene levels, via control of RNA stability. Just to make an example, the binding to the 5` may be responsible for a block of translation and an increase of the gene levels, even if then the same protein will be knocked out. In the coding region microRNA can regulate splicing. An increased binding at the 3` site can result in decreased gene levels and increased protein, again via regulation of mRNA stability. You can have also the case of increased microRNA and increase gene expression. In this sense, you cannot predict which is the effect of microRNA expression on the phenotype. You can have with the same binding opposite effects depending on microRNA localization and the involved RNA-binding proteins. This also explain the many contradictory findings reported for the same microRNA in different tissues. Non coding RNAs are difficult to be explained with dogmatic rules.

@ Naveen

Even if a seed sequence is expressed, this does not mean that the microRNA will bind its target. Binding then is affected by other factors:i) there is possible editing of both microRNA/binding sites. Editing can activate binding of some microRNAs and deactivate others. ii) There are other binders to adjacent regions which can mask and/or compete the binding site. iii) There is the possibility that RNA binding proteins sequestrate the microRNA before binding its target.

Via deep sequencing we know that actually the binding of microRNA occurs at the 3' approximately 60% of the cases. In all the others binding occurs elsewhere. It is very popular to perform luciferase assay to predict effects of microRNA in a given tissue on a given gene,In my vision this approach is outdated. It is impossible to predict the effect of a microRNA by looking exclusively to the binding to the 3'UTR in a model. It is very unlikely that you will be capable to identify the mechanism of action of the same microRNA in all the tissues. By definition in order to identify the function of a microRNA it is essential a multidimensional analysis, which is capable to analyze not only the mocroRNA but its target (RNA sequence) and the related RNA-binding proteins which tune the microRNA function in one sense or in the other. We have still so much to learn!

Dear Athul,

MicroRNAs behave differently in different organisms and thus it is important to consider this when discussing their mechanisms of action. Binding to the 3'UTR is a widely accepted concept in mammals whereas binding to the coding region is often seen in plants. There are also reports of up regulation of gene expression due to binding at the 5'UTR.

Good Luck

Multiple miRNAs bind to and functionally regulate protein levels through 5'-UTR sequences. The idea that miRNAs operate only through the 3'-UTR is outdated and only clung to as blind dogma.

Till date several database have predicted binding sites in all three 5UTRs ,

3UTRs and CDS regions of gene but veracity of these bindings is still not fully proven and also not discarded, still lack of such techniques that can show the effect in (in-vivo) real time aspect is not possible especially in mammal. Other peculiar dye or fluorescence based technique that may be invented in future which could answer this question, but till date 3UTRs region binding is considered more than 70% scientific database.