Peter- I don't know what significance to ascribe to this observation, but I don't think it is terribly unusual. There are several thousand distinct genes (around 6K in UCSC genes) that have 5' UTR that are 25bp or smaller. Perhaps you could run a Gene Ontology analysis and see if there is any underlying biological function linking these groups of genes. I suppose that one can safely infer that whatever functions one would normally ascribed to 5' UTRs are achieved through other means (other types of post-transcriptional regulation) for this class of genes/transcripts.

Thanks Daniel for your answer. It is actually quite reassuring that there are quite a few other genes which also have such short 5' UTR's as if the mRNA that i was looking at was the only one then i would be a bit more dubious about it. The shortness of the 5'UTR is quite interesting does not seem to leave much space for the translation machinery to "hop" onto the mRNA transcript and produce the subsequent protein.

Peter- maybe translational efficiency of these mRNAs is reduced compared to those with bigger 5' UTRs. Alternatively, it may be the case that ribosomes don't need to be recruited to the 5' UTR per se, but can initially bind 3' to the initiation codon, and subsequently find the correct location for translation initiation by ribosome scanning. Beyond my expertise, but perhaps someone else in the ResearchGate community can offer insights as to which of these scenarios is most likely.

Peter, the best way to be sure that you've got a full length 5' UTR is by RACEing it (is that your oligo capping?).

In the old good days of cDNA cloning, people used to work until a full ORF was cloned , usually not being interested in the 5'UTR (trying to obtain longer 5'end sequences was difficult and time consuming). This is why most of the sequences at the GenBank are incomplete by their 5'UTR.

Peter,

Generally, 5' UTR is highly correlated with translational regulation, especially for the regulation on initiation. Thus, I think the variation of the length of 5' UTR can significantly affect the regulation of gene expression. The following papers maybe be useful.

1 Zhenguo Lin, Wen-Hsiung Li (2012), Evolution of 5′ untranslated region length and gene expression reprogramming in yeasts.

2 Honglei Liu, et al. (2012), Characterization and evolution of 5′ and 3′ untranslated regions in eukaryotes

Thanks for your answers Estanis and Y.H.

In relation to your question estanis the mRNA sequence which i found was obtained from the NEDO sequencing project which as far as i am aware was some sort of attempt by japan to sequence full length cDNA's. Here is what is says on the pubmed page

"NEDO human cDNA sequencing project supported by Ministry of International Trade and Industry of Japan; cDNA full insert sequencing: Research Association for Biotechnology; cDNA library construction, 5'- & 3'-end one pass sequencing and clone selection: Helix Research Institute (supported by Japan Key Technology Center

etc.) and Department of Virology, Institute of Medical Science, University of Tokyo."

There is also more information on the internet where they say that they optimised the 5' capping method to obtain full length cDNA's with a success rate of 90%. Because i can detect a protein migrating at very close to the predicted size of the translated protein im quite optimistic now that this mRNA transcript is legitimate. in the future i would like to do some 5' and 3' race experiments to confirm though and also get some information as to possible regulatory elements in the untranslated regions.

thanks for your suggested papers y.h will have a read through them as they sound quite interesting and should provide me with some good background.

Peter, it sounds good but that means that the authors optimized the "global" cDNA reaction to obtain full-lenght clones, not individual cDNAs (as the one you are interested in). If you are interested in regulatory features of the 5' end I think that it is important that you actually have the entire 5' end of your cDNA. You can start by BLASTing your 5'end against the "transcripts"/"EST" sections of the GenBank to see if there is any sequence longer than yours. If you find longer 5'end transcripts then I would 5'RACE the cDNA to be sure that you are working with the entire sequence.