4
$\begingroup$

I realized that in all cases of "RefSeq Genes" annotations of hg19 I looked at spliced transcripts start (and end) with an exon. From the annotation there is no evidence of any sequence upstream or downstream of these exons that remain in the nascent RNA.

Does this reflect the biology or is it just a consequence of the spliced read mapping? In other words: Do transcribed regions upstream of the first exon and downstream of the last exon exist in nascent RNA?

These regions, if they exist, cannot be called introns when those are defines as 'spliced sequences between exons'. How would one call them?

$\endgroup$
  • $\begingroup$ Welcome to BiologySE... I agree with what you say at the end (i.e. "I guess if you have an answer for 1 and 2..." - I would narrow your question to get an answer (it's easier to answer a simple, straightforward question instead of a broad question with 5 subquestions). You can always ask more questions - the more specific the better $\endgroup$ – Vance L Albaugh Apr 26 '16 at 15:06
  • $\begingroup$ Very Good (and well-researched) question!! I am looking forward to answers here $\endgroup$ – Failed Scientist Apr 26 '16 at 15:06
  • 1
    $\begingroup$ @TalhaIrfan Thank you. I condensed the text to the basic questions. $\endgroup$ – Herr Jemine Apr 26 '16 at 15:57
  • $\begingroup$ Could you explain what you mean by "spliced read mapping" please. $\endgroup$ – David Apr 26 '16 at 22:05
  • $\begingroup$ @David, spliced read-mapping refers to the mapping of spliced sequences (e.g. RNA-derived sequence reads) to an un-spliced reference (DNA of reference genome). The algorithm that performs this kind of task has to be able to accept long insertions on the reference sequence without penalizing those as mismatches in the alignment. These read-to-reference insertions (or reference-to-read deletions) are introns. $\endgroup$ – Herr Jemine Apr 27 '16 at 9:37
6
$\begingroup$

Most (almost all, AFAIK) mRNAs and lncRNAs start with exons for the reasons already mentioned by David. In a typical splicing event, the nucleotide that is 5' to the splice donor site (lets call it pre-donor) and the one that is 3' to the acceptor site (lets call it post acceptor) are joined together and the intronic sequence between them is removed.

If you look carefully at the mechanism, you'll notice that the pre-donor (i.e. the 3' end of the first exon) renders a nucleophilic attack on the phosphorous of the post acceptor (i.e. the 5' end of the second exon) which leads to the release of the intron.

enter image description here Image courtesy: http://mips.helmholtz-muenchen.de/proj/yeast/reviews/intron/spliceo_splicing.html

Therefore there is always an exon preceding an intron, in the transcript.

However, RNA processing can occur via a different mechanism in which the ends are just chopped off. This happens in tRNA processing, where RNAseP chops off a portion 5' to the mature tRNA region, called the leader sequence (see the figure below). I am not aware of such a mechanism happening in the case of mRNAs but a possibility certainly exists. One possible reason for mRNAs not having this kind of mechanism is because their 5' end is capped (co-transcriptionally) and cleavage of the 5' would lead to loss of cap and destabilization of the mRNA (all guesses, though). Similarly, the 3' of mRNAs is polyadenylated which also renders stability to them. Many lncRNAs are also capped and polyadenylated but there are exceptions (mRNAs too).

Since this mechanism is not very common, there is no systematic name for the excised ends. For the case of tRNA, it is simply called the "leader" sequence.

enter image description here Image courtesy: Leigh & Lang, 2004

$\endgroup$
  • 1
    $\begingroup$ Thanks for your detailed explanation. As David and you point out it is very unlikely that the splicing mechanism would be able to remove sequences upstream and downstream of the first and last exon respectively. Another important hint is that 5' CAP processing occurs co-transcriptionally. This, and the fact that a method called 'CAGE' is used to identify promoter regions by CAP-specific sequencing of mRNA 5' ends gives some confidence that 5' removal of a 'pre-exon sequence' is likely not happening in 5' capped mRNA. Yet we have not discussed the 3' end. $\endgroup$ – Herr Jemine Apr 27 '16 at 9:29
  • $\begingroup$ @HerrJemine I don't really know a reason for why that is not common. Even for the 5' part this was just a wild guess. Do not take it as a fact. BTW, the 3' end has poly-A which is also critical for RNA stability. However an RNA can be stable in the absence of cap and poly-A tail too. Since such cases are also very few, we can guess that this could be a factor for low occurrence of end chopping in RNAs. $\endgroup$ – WYSIWYG Apr 27 '16 at 9:31
  • $\begingroup$ I found some hint on post-transcriptional 3' end truncation in alternative polyadenylation on wikipedia (second paragraph of intro and section 'Alternative polyadenylation'). I'd suggest that you add this aspect to your answer and I'll accept it. $\endgroup$ – Herr Jemine Apr 28 '16 at 16:25
  • $\begingroup$ en.wikipedia.org/wiki/Polyadenylation $\endgroup$ – Herr Jemine Apr 28 '16 at 16:32
  • $\begingroup$ @HerrJemine I already did that. Anything more that you expect? I would not say anything too far fetched. $\endgroup$ – WYSIWYG Apr 28 '16 at 17:20
4
$\begingroup$

As far as I am aware, transcripts always start and end with exons. The reasons I wouldn’t expect otherwise (apart from my observations when examining Drosophila transcripts) are given below.

As you will be aware, the spliceosome (at least for mRNA) is a highly sophisticated multi-component ribonuclear protein complex, and has functions to both splice out the intron and ligate the ends of the exon. The diagram below, from a recent review, suggests that recognition of the splice sites also involves the exon.

Part of Fig. 4 from TIBS (2016) vol 41 pp 33-45

And the authors write (my italics):

“Base pairing between the 50 end of U1 snRNA and the first six nucleotides of the intron, and up to three nucleotides of the exon, provides the main driving force for specific sequence recognition.”

So I think that the contemporary machinery for removing mRNA introns requires exon mRNA to ‘get hold of’.

If such removal of introns could serve a functional purpose you would imagine the machinery would have evolved to accomodate such situations, but if the main function of introns is to allow differential splicing of exons that specify regions of the protein, introns at the end of a mRNA would have no purpose.

Another point to consider is the RNA-polymerase binding site on the DNA (the extended and complex promotor). If introns arose originally as mobile elements (something suggested by the authors of the review) then an insertion at the 5' end would likely have inactivated the promotor, so no transcript would have been produced.

Of course, Nature is full of exceptions, and it would be foolish to exclude that introns of the type you postulate do exist (in some galaxy far away...).

$\endgroup$
  • 1
    $\begingroup$ So why the mark-down? If you vote someone down have the guts and the decency to explain why. What is so reprehensible about the idea that splicing evolved because it allowed variants of proteins by differential splicing and the machinery evolved to deal with that? That although there are introns in the UTRs of a mRNA, which perhaps affect ribosome binding etc, these are likely to have been secondary. If you have a complex that needs to bind two parts of a rope before it can snip and join it, it makes perfect sense for it not to work at the end. Or don't you like my last para? $\endgroup$ – David Apr 26 '16 at 21:58
  • 1
    $\begingroup$ Didn't downvote, but I would guess that the theoretical and unsourced nature of this answer would play a significant role. Also, I find it strange why you would ask for someone to explain their downvote in such a defensive manner. $\endgroup$ – March Ho Apr 27 '16 at 2:58
  • $\begingroup$ @MarchHo I won't shoot the messenger, but when you down-vote you are prompted to make a comment. It is most annoying to provide a structure-based answer to a theoretical (more hypothetical) question and end up with -1 so a casual visitor would assume something basically wrong with your answer. My guess now is that the self-deprecatory and whimsical Star Wars reference was misunderstood as an attack on the question. Perhaps I'll change it to Shakespear when I add a pic of the spliceosome later to clarify my argument. $\endgroup$ – David Apr 27 '16 at 6:48
  • $\begingroup$ Your answer is touching the point but a little explanation and some diagrams would make it better. $\endgroup$ – WYSIWYG Apr 27 '16 at 8:18
  • $\begingroup$ @David Thanks for the answer to my question. I like that you pointed out the structural difference between a splicing between exons and a potential sequence removal upstream and downstream without a 'splicig partner'. By the way I didn't down-vote either, but my upvote did not count yet (missing exp.). $\endgroup$ – Herr Jemine Apr 27 '16 at 9:04
2
$\begingroup$

Thank you for a great question.

I would like to start by clarifying some terminology.

First, nascent RNA refers to an RNA molecule that is currently being transcribed and has not been processed. Processing can include the splicing out of introns or polyadenylation at the 3' end, for example. Mature RNA is (typically) spliced and polyadenylated.

Second, the current consensus definition of an exon is a transcribed region of the genome whose sequence can be found in mature RNA species. Importantly, there are two classes of exons: non-coding exons, and protein-coding exons. I think you may have these confused and I have included an image to help clarify the distinction between introns, coding exons, and non-coding exons, using the human GAPDH gene as an example.

UCSC Genome Browser view of human GAPDH gene

In the figure, the solid blue boxes represent exons, which are present in the mature mRNA. These exons can contain both protein-coding and non-coding sequences, as I mentioned. For example, exon 1 of GAPDH contains only non-coding information that belongs to the 5' untranslated region (UTR) while exon 2 contains both coding and non-coding sequence information.

When you view the RefSeq annotation, you are viewing all exons (coding-and non-coding). For some of the genes, such as GAPDH, the first exon will be non-coding. Other genes may begin with a coding exon (I have yet to find one and since the 5'UTR is involved in ribosome binding, I am doubtful that many examples exist).

The regions of an mRNA molecule beyond its coding sequence are referred to as the 5' and 3' untranslated regions (UTRs).

I hope this helps answer your question.

$\endgroup$
  • 1
    $\begingroup$ This does not actually answer the question. I think the OP knows about exons and UTRs. Their question is about why a transcript always has to start with an exon. Perhaps you can add some explanation for this point. $\endgroup$ – WYSIWYG Apr 27 '16 at 8:20
  • $\begingroup$ I have not heard of a classification such as coding and non-coding exons. You can have exons that have a part of the CDS and a part of the UTR. $\endgroup$ – WYSIWYG Apr 27 '16 at 8:22
  • $\begingroup$ Thanks for your reply. But my question did not concern whether exons contain coding or non-coding sequences. I am aware of the fact that exons can be part of the 5' or 3' UTR in the mature mRNA and also that there are non-coding RNAs in general. As @WYSIWYG pointed out I would like to know if the unprocessed, nascent RNA always starts and ends with an exon or if their might me a removal of 'head' and 'tail' RNA sequences that were transcribed from the DNA. $\endgroup$ – Herr Jemine Apr 27 '16 at 9:17
  • $\begingroup$ @WYSIWYG Thank you for your comment. Perhaps coding vs. non-coding is restricted to certain bioinformatics circles, which is where I came across the term. I think the 5'/3' UTR convention is more intuitive. $\endgroup$ – Sciencophilus_rex Apr 27 '16 at 10:44
  • $\begingroup$ @HerrJemine Thank you for clarifying your specific needs. As someone pointed out above, only a very limited subset of RNAs are cleaved at the 5' end post-transcriptionally (I wasn't able to find any examples of mammalian RNAs that are cleaved at the 3' end). You may be interested in exploring a technique called NET-seq link. It demonstrates pervasive transcription outside of the transcriptional unit in yeasts, flies, and humans. $\endgroup$ – Sciencophilus_rex Apr 27 '16 at 10:47

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.