3' RNA Seq libraries (creating tags)

[moritzhess]

Hi,
we are planning to profile gene expression in a number of samples. We are not investigating an organism without genome sequence (but reference transcriptome) available. In order to save costs we are looking for a method to sequence only one part (read) per transcript (e.g. the 3' prime end). There is a Kit from NUGEN available but the price is quite high and we are looking forward to use Truseq technology in order to make the handling of a large number of samples more convenient. Has anybody of you have made experiences with "custom" 3' enrichment protocols that fit well into the standard illumina sample processing pipeline ?

Thanks in advance !

Moritz

[steven]

That makes me think of Direct RNA Sequencing -no amplification bias + targets the actual 3' end- but i have no idea about the price (that's helicos, not illumina). It has been used to profile the "polyadenylatome" -just made it up

http://www.ncbi.nlm.nih.gov/pubmed/21145465

[moritzhess]

Hi Steven,

thank you very much ! This Helicos stuff is really interesting, especially in terms of sample prep. Nevertheless we are a little bit "tied" to Illumina Seq as "our" Seq facility "only" has Illumina machines, but this direct RNA sequencing is really something to keep track of.

Moritz

[steven]

Thanks Moritz for your comment. Yes, their library prep protocol is very simple. Especially when comparing with this one -a similar study using illumina also on polyA:
http://rnajournal.cshlp.org/content/early/2011/02/20/rna.2581711.long

[moritzhess]

I do not have access to that paper, so I could not checkout. But in the meantime I just remembered that I heard somewhere that to only sequence the most 3' prime part of a transcript it would be sufficient to do: a) 1 round of poly A enrichment, b) fragmentation, c) another round of poly A enrichment. d) following normal Illumina RNASEQ protocol. That should result in libraries consisting of only the most 3' prime part. OK the problem would be that the insert could be sequenced from the "wrong" (polyA) part, but by using PE sequencing and discarding low quality regions, this issue could be circumvented.
Is this oversimplified or has anybody already tried tag sequencing this way ?

In keen anticipation,

Moritz

[moritzhess]

I Uploaded a picture to outline the method (it's horrible I know!) I wan't to apply. Basically the only modifications are to switch fragmentation and enrichment steps and to use polyT primers for cDNA synthesis instead of random primers. Has somebody done library prep for RNA Seq that way with good results?

[amango]

These authors published a protocol for exactly what you seek to do:
Yoon and Brem. Noncanonical transcript forms in yeast and their regulation during environmental stress. RNA (2010) vol. 16 (6) pp. 1256-1267

Here is a link to an improved protocol for 3'-end Illumina mRNA-seq from the Brem lab: http://blogs.ls.berkeley.edu/bremlab/protocols/

[moritzhess]

Hi Amango,

Thank you very much! Actually I've always wondered if this kind of modification of the protocol works out and it seems to work out pretty well.
Moritz