A question about splicing

Question

[kresimir.krizanovic]

Hello,

I have a question I hope that someone here will be able to answer.

Its concerned with RNA splicing. As far as I can understand it, splicing removes introns from pre-mRNA. Alternate splicing can also remove some exons (or parts of exons) together with the introns.

I've been reading up on it on Wikipedia (I know, I should find a better source of information) and elsewhere on the Internet, and it seems to me that the first and the last exon can not be removed. Is this true, or are they sometimes also removed by splicing?

Answer 1

[WouterDeCoster]

You can have alternative first or last exons...

Answer 2

[Devon Ryan]

While I've never heard of the first/last exon getting spliced out (this doesn't mean it never happens, just that I've never heard of it), you can have a large number of different transcription start and stop sites, which has the same end effect.

Answer 3

[reza.jabal]

This is interesting how fundamental topics will stimulate scientific discussion in here!

Generally speaking as the first exon & last exon contain initiation and termination codons respectively they are barely spliced out. In addition spliceosomes function in the exon-intron boundaries which function in immediate downstream of first exon and upstream of last Exon. Furthermore,to protect cells, an mRNA surveillance mechanism known as nonsense-mediated decay (NMD) degrades most RNA transcripts that have an in-frame PTC. The primary NMD pathway is dependent on RNA splicing. Multisubunit protein complexes, called exon junction complexes, are deposited shortly before the 3' end of each transcribed exon during RNA splicing and remain bound at the positions close to the exon-exon boundaries in mature RNA. The first ribosome to bind and move along the mRNA displaces each of these complexes in turn before disengaging from the mRNA at the natural stop codon. If there is an in-frame PTC, however, the ribosome detaches from the mRNA at an early stage, which usually signals mRNA destruction. However, in-frame PTCs within or just before the last exon often scape NMD and are translated to give truncated proteins.

Single exon genes escape NMD because they do not undergo RNA splicing.

[John]

This is pretty much it. There's a selective pressure due to NMD to have dedicated starting and finishing exons - more so for finishing exons than starting exons, but that's only my impression (I dont have numbers to back that up).

A few years ago I was asked to design several million primer pairs to detect all splice variants of all human/mouse/rat genes for some pharma company, and one of the requirements was that while the first primer pair could go anywhere on a splice-boundry, the second primer-pair must/should go somewhere in the last (3') exon. The theory was the last exon is usually transcribed for most/all splice variants. I don't know how true it is, but that's what they wanted.

Answer 4

[Anima Mundi]

Hello, exons are defined as those sequence stretches that are kept after RNA splicing.

Therefore:

1) the concept of exon will always refer to a defined type of transcript, i.e. each transcript variant will have its own set of introns/exons (often, but not necessarily, many introns/exons will be identical among variants);

2) you will always have, somewhere, a first exon and a last exon (they can be the same).

If you want to know whether an immature transcript always begins with a sequence that will be kept as an exon, then I can say that so far I have never seen a gene the TSS of which was not the start of the first exon.

Answer 5

[kresimir.krizanovic]

Thanks to everyone that responded. While I didn't get the answer I was hoping for, I realized that I didn't know enough about the matter to ask a proper question :). I've got more reading to do.

Not to beat around the bush, I'm writing a script that would compare RNAseq mapping to an annotation file to determine how "good" the mapping is (and maybe do some additional analyses). Thats why I'm studying RNAseq and annotation files to better understan what could occur.

If someone is up to it, I have a follow up question:

I'm looking at BED annotation file for S.Cerevisiae S S288c. There are two annotation enteries for transcripts YOR343W-A and YOR343W-A. This puzzles me a lot. They both start at the same position in the genome and on the same strand. That would mean that they relate to the same gene (namely YOR343W). However, transcript A has only one exon of length 1317, and transcript B has two exons, first of length 1294 and second of length 4019 with only one base in between.

Two things puzzle me here:

• Transcript B seems to have an intron of length 1
• The only exon of transcript A is 23 bases longer then the first exon of transcript B and starts at the same position. It actually contains a whole exon1, one skipped base and a small part (22 bases) of exon2 of transcript B.

There are several other pairs of annotations that look really similar to the one above (e.g. YOR192C)

I would be grateful if someone could briefly explain the mechanism behind these two transcripts. Is this alternate splicing, and which type?

[harold.smith.tarheel]

The transcripts in question are derived from a retrotransposon element (Ty in yeast). The first transcript encodes the GAG gene. The second co-encodes the GAG-POL fusion protein. This transcript is not spliced; rather, a +1 ribosomal frameshift occurs during translation to produce GAG-POL. There are multiple Ty copies in the genome, which explains the presence of similarly annotated genes.

[Devon Ryan]

Those are likely artifacts from how the transcriptome annotation was created. It's unlikely that the 1 base long intron actually exists, more likely either that base is included or it doesn't exist in the genome.