Thursday 22 January 2009

Meaning out of nonsense

Imagine you are a eukaryotic cell. You have many genes and make many mRNAs. One problems is mutated genes, pseudogenes and mistakes by RNAPII all make mRNAs which contain Premature Stop Codons (PTC). These then make truncated proteins which could be toxic to the cell. Eukaryotes have evolved a mechanism to recognise these and degrade them. I guess prokaryotes do not have an analogous system (that I know of) because the half-life of these mRNAs are minutes long not hours long like most eukaryotic mRNAs. Eg the average mRNA half-life in Arabidopsis is between 4-6 hours and is very variable between mRNAs. Personally I am not convinced this is an important enough reason to have a quality control check mechanism. The system I am referring to is called Nonsense-Mediated mRNA Decay (NMD). Mediators of it have been found in the early branching Giardia and is conserved between fungi/animals to plants. A PTC is recognised as wrong rather than a correct stop codon at the end of a ORF during the first round of translation by the ribosome. When the ribosome reaches a stop codon it has two choices, terminate correctly or stall and cell effectors of NMD to take it away for degradation by whatever nucleases that cell uses. It was first found in yeast and the mutants were called upf1-3 (for up-frame shift 1-3) and these yeast grow fairly normally but have increased levels of transcripts with stop codons. Raises the question why did it evolve if not very important when you KO it. In C. elegance SMG1-7 were identified having genital defects. Strangely, these were involved NMD with SMG2-4 corresponding to UPF1-3. Homologues of most of these proteins are found in many other eukaryotes. UPF1 is generally phosphorylated when ribosome stalling takes place and condemns it to degradation. Interestingly, it is a DEAD box RNA helicase (SF2 I believe).

The two questions I find interesting is 1) how do organisms tell the difference between a PTC and normal stop codon and 2) how this can be used to regulated gene expression in cell signalling. You may ask ‘James, you charismatic stallion, how can something degrading useless mRNAs be useful or regulated for controlling gene expression’ and I would reply ‘wait and read the rest of this post you $#£!&%’. The first question is rather interesting as different organisms do different things. Yeast and invertebrates such as C. elegance and the fly pushers Drosphilia use something called the 3’ faux model. Something in the 3’ end allows normal termination when the stop codon is here but NMD starts when the stop codon (PTC) is distant from the 3’ end. It has been found the distance from the poly-A tail and therefore the Poly-A Binding Protein (PABP) determines this. When a long 3’ UTR is present the mRNA is degraded. When PABP is tethered close to the stop codon termination occurs normally. This also works in plants and mammals. However, another mechanism also works in both of them. When an introns is spliced out an Exon-Junction Complex (EJC) is left behind and if a EJC is found near a stop codon by the ribosome NMD starts. Why this operates in both plants and mammals but not yeast and invertebrates is curious. Did both mechanisms operate in the last common ancestor of plants and animals and has been lost a few times in the fungi and animal kingdoms or is it convergent evolution? I hope to better understand that.


Another and perhaps more important question is what is NMD role in plants and animals. Does it provide another function other than keeping toxic protein levels low? Short answers, yes. It affect development in C. elegance and embryo lethality in mice. Clearly it plays a role in animal development. In yeast the cells senesce sooner in upf1-3 mutants because telomeres shorten. In mammals, amino acid deletion causes problems for the ribosome and NMD is down regulated and genes for amino acids biosynthesis are up-regulated so more amino acids will be made. In Arabidopsis, when upf-1 is completely KO we see embryo lethality but Knock-down alleles see some developmental phenotypes (such as in the flowers) and altered stress responses.

I will finished up now but feel free to ask questions. I have just finished writing a 4-page grant proposal on this. Just to piss Mel off, I will say this. I didn’t do any reading for this. This was all from memory when I got home. No PDF files were opened – not one!

3 comments:

Argent23 said...

Oh, so the connection between EJCs and NMD could be the reason why the very last exon is generally much longer than internal exons in most of the genes. At least in Arabidopsis, where I notice this all of the time.

Btw, do you know any papers on mRNA half-life time? I mean papers where there are specific numbers given, like the shortest and longest lived mRNAs in human cells, for example. This point just came up recently during a discussion with students, and I couldn't give them more than the standard 'several hours in eukaryotes' answer.

James Lloyd said...

The EJC has been shown to be deposited 20–24 nucleotides
upstream of splice junctions sites and causes stop codons recognised >50-55nt upstream of an exon-exon boundary to be classed as a PTC. I had not noticed long final introns but it does suggest this could be involved.

this is the reference from a good review on mRNA decay in plants which uses microarrays to look at half-lives of 13000 genes but i have not read it.

Narsai R, Howell KA, Millar AH, ’Toole N, Small I, Whelan J: Genome-wide analysis of mRNA decay rates and their determinants in Arabidopsis thaliana. Plant Cell 2008, 19:3418-3436.
i think it cites some animal work in the intro.

Argent23 said...

Thanks alot for the paper! Arabidopsis transcripts are actually more useful for me, so this should help clear up the questions of the students.