Adapting TruSeq RNA-Seq Protocol

[JayS]

Hello everyone,

I have been given a little funding for an undergraduate research project and decided on performing differential gene expression through RNA-Seq on the two mouse models I have access to. After designing my experiment, and budgeting for a couple of lanes on the Illumina HiSeq, I was surprised to learn that the university core facility charges a per sample fee to prepare the libraries, and those fees make me go over my budget.

Because I do have access to a few different kits and reagents in the lab I'm wondering whether I can adapt the TruSeq protocol to fit my needs. My objective in developing a modified protocol is to:

1) Decrease per sample costs of library prep in order to afford a higher number of replicates;
2) Limit the introduction of bias (lane, batch, pcr, etc.).

After reading a few papers and many posts on seqanswers.com I came up with the following protocol. Will this work? What kinds of problems will I encounter if I prepare my samples in this manner? Any insights will be greatly appreciated.

1. Isolate mRNA (Invitrogen Dynabeads)
2. Verify purity
3. Fragment mRNA (Ambion Fragmentation kit)
4. Verify size range
5. Repair ends (T4 polynucleotide kinase, ATP)
6. Ligate 6-bp barcode to 5' end (NEB T4 RNA Ligase, Custom made barcodes)
7. Verify ligation, purity
8. Pool samples in equimolar concentrations
9. Submit as one single sample to core facility for remaining steps of TruSeq library prep

[pbluescript]

There will have to be some sort of reverse transcription step in there. The Illumina recommended RT is Superscript II, but I have used Superscript III before and I know others here have done the same.
Also, are you suggesting the barcode and the adapter will be ligated separately? The barcodes should be part of your adapter.

[elaney_k]

Tbh if you're going that far with the protocol I'd suggest buying custom-made Illumina barcoded adapters and PCR enrichment primers and finishing the sample prep yourself, I've pooled RNA-seq samples in this fashion many times. That way you can submit a finished library to the sequencing facility with no need for them to prep the samples, but you definitely need to let them know to expect a barcode in the first 6 bases as it can make sequencing very difficult if too high a concentration of that kind of library (i.e. only 2 indexed samples) are loaded onto the flowcell. Another big thing to remember is that you should make sure that no 2 bases of the 2 seperate indexes are the same at any position (all the better if you multiplex 4 samples or more so that the base bias in the index is lessened).
With the way that you're suggesting, they'll probably still process the samples in the same fashion that they would any other DNA sample i.e. end-repair, a-tail and adapter ligation followed by enrichment PCR.

Elaine

[JayS]

Thanks for your replies. I'll take your advice in consideration before I make a final proposal to my professor.