Are RNA cleavage sites always polyadenylated?

Question

During RNA processing (including both mRNA and non-coding RNAs), is the end of 3'UTR always polyadenylated after cleavage?

I could not find the answer to my question in the Wikipedia entry for polyadenylation.

If a recent reference could be provided, that will be great. I've done quite a bit search and hasn't found a good source yet.

Some background to this question: we are developing bioinformatics tools to predict potential cleavage sites, we need to understand this problem better in order to design a suitable algorithm

I'm not sure what you mean by "mRNA processing, including both coding and non-coding" or why you need to qualify this. Surely all you are interesting is that part of the mRNA processing that recognizes the cleavage signal AAUAAA (or variants). I have edited your question, but left this for the moment. I suggest you delete "including both coding and non-coding" unless you can clarify it. — David, Aug 16 '18 at 12:53
What sort of cleavage are you talking about? Do you transcription termination? — Nicolai, Aug 16 '18 at 13:02
@David, sorry for being confusing. I meant both coding and non-coding RNAs. — zyxue, Aug 16 '18 at 13:03
@Nicolai, not really, after pre-mRNA is transcribed, it will be cleaved somewhere towards the 3' end. — zyxue, Aug 16 '18 at 13:05

David · Accepted Answer · 2018-08-16T13:17:12.013

2

This answer attempts to cover only the question of mRNA processing. It does not address the polyadenylation of non-coding RNAs.

As far as I can see from the recent review by Neve et al. (2017) cleavage and polyadenylation are tightly coupled. It is interesting, in this regard, that histones mRNAs which lack polyA tails, have a different structure, lacking the canonical AAUAAA, at their 3ʹ ends (see review by Marzluff et al.).

As distinct enzyme activities are involved in cleavage and polyadenylation, it would, I imagine, be possible (either genetically or chemically) to disrupt the protein complex responsible for polyadenylation and hence prevent it occurring (this has probably been done). However this would not be physiologically relevant.

There is an earlier review by Tian and Graber (2007) of the sequence requirements of the cleavage site which makes no distinction between cleavage and polyadenylation.

However this is not my field, so others may have additional information I have missed.

edited Aug 16 '18 at 13:17

answered Aug 16 '18 at 13:08

David

25,583
8
53
95

1

@zyxue — I think you are having trouble or typos with your negatives again. Surely you mean "It DOES seem… most are polyadenylated". Agreed. – David Aug 16 '18 at 21:53
Quote from Neve et al. (2017): "With the exception of replication-dependent histone genes, metazoan protein encoding mRNAs contain a uniform 30 end consisting of a stretch of adenosines." It does seem, at least for protein-coding mRNAs, except for histone genes, most are polyadenylated. – zyxue Aug 16 '18 at 21:56
Thanks for the catch, @David. Do you have any lead on non-coding mRNAs, please? – zyxue Aug 16 '18 at 21:57
Based on Table 1 from this Review paper, Strategies to Annotate and Characterize Long Noncoding RNAs: Advantages and Pitfalls Author links open overlay panel, 2018, for non-coding RNAs, they are more often non-polyadenylated than polyadenylated. – zyxue Aug 16 '18 at 22:33

Are RNA cleavage sites always polyadenylated?

1 Answers1