miRNA binding detection from RNA-seq reads

Question

[bharata1803]

Hello,

I have a question regarding miRNA binding to mRNA. If a mRNA binds with miRNA, it will form miRNA-mRNA binding complex. What is the effect of this complex from the point of view of RNA-seq especially RNA-seq experiment to prepare the sample (if using Illumina technology)?

I don't really understand about biology but I think about the possible effect.

First possibility is the mRNA which binds to miRNA will not be able to bind with the sequencing kit because miRNA will block the binding between mRNA and the sequencing kit (I don't know the name but I think during amplification). The consequence is the expression result from RNA-seq comes from mRNA which doesn't bind to miRNA and we can not detect miRNA at all.

Second possibility is mRNA is fragmented. So, the region of mRNA which is not bound to miRNA can bind to the sequencing kit. The consequences of this is the region which miRNA binds to mRNA will have small or no reads aligned to it. I imagine the reads coverage after the alignment would be like a valley between 2 mountains. Area not bound to miRNA will have a good coverage and area bound to miRNA will have small coverage.

What is your opinion about this? Is it possible to extract information about miRNA binding from RNA-seq?

[Michael]

I think it would really help to get the biology right and understand the technology of sequencing. Therefore it might be a better fit for seqanswers, I don't see much bioinformatics relevance here. A few comments though:

• miRNA does not bind mRNA, a RISC protein complex loaded with miRNA does
• the RISC complex either cleaves or binds, if it cleaves the mRNA the effect should be visible as a reduction of transcripts depending on their degradation, if it only inhibits translation, it might be invisible to RNA-seq
• RNA-seq of mRNA is mostly used with polyA enrichment for mature transcripts, therefore, miRNA and other small RNA content will be reduced. It is not impossible to have some leak of these small RNAs as the protocols are not perfect.

Answer 1

[apa@stowers]

Never say never I guess, but miRNAs should not affect normal mRNA-seq protocols, in theory. As Michael pointed out, miRNAs don't act alone. But even if they did,

• mRNAs are enriched either with polyA selection or ribodepletion. Ribodepletion ignores mRNAs, so no effect there. PolyA uses poly-dT beads to bind polyA tails, but miRNAs don't target polyA tails.
• mRNAs go through a denaturation and washing step prior to cDNA conversion, which should eliminate RNA trans-duplexes.
• RT primers are frequently polyT or random-kmer. Again, polyT binds the polyA tail and would not be affected by miRNAs. Random-kmer might be affected if duplexes still existed, I don't think RT mixes include helicases. But by this stage, I think most "normal" protocols will have eliminated any duplexes.

On the other hand, there are protocols designed to reveal where miRNAs bind, like CLASH-Seq, PARIS, or just chipping the RISC complex.