Sample prep for RNA-seq from low abundance RNA

[chadn737]

I've been working with RNA isolated from cells obtained by laser
capture microdissection. Isolating these cells and subsequently its
RNA is very laborious and as the cells are in low abundance, I'm
limited in the RNA that I can obtain. We are talking picogram
quantities of total RNA per cell. Realistically, by pooling samples I
could bring the amount of total RNA up to the nanogram range.

Given the limited amount of RNA that we will have available the
standard RNA-seq protocols which require 1-10 micrograms of total RNA
is simply not reasonable.

Does anyone have any experience in preparing samples for RNA-seq from
low abundance RNA? I would appreciate any recommendations that you
have.

[doxologist]

are you willing to amplify?

we also work with samples with low RNA but go through one or two rounds of amplifications.

[neethav]

Which kit do you use for mRNA amplification?

[james hadfield]

Have you looked at Nugens Ovation kit (http://www.nugeninc.com/nugen/index....-seq-system/)?
Looks good but we have not tried it yet.

[krobison]

A few papers have done SOLiD sequencing from a single cell: Cell Stem Cell. 2010 May 7;6(5):468-78, J Biomol Tech. 2009 Dec;20(5):266-71, Nat Protoc. 2010;5(3):516-35. Epub 2010 Feb 25, Nat Methods. 2009 May;6(5):377-82. Epub 2009 Apr 6.

[kainsteven]

Users of Ovation RNA-Seq

We have had several users generate good results using low input amounts of total RNA with the Ovation RNA-Seq System. If you are interested, we have several customer posters and webinars from recent conferences on the NuGEN website: http://www.nugeninc.com/nugen/index....en-sequencing/

Also note that today we announced our Encore NGS Library Systems which integrate directly with the RNA-Seq protocol, or can be used to construct libraries for any NGS application.

Best,
Steve

[paul z]

Illumina is said to have started with as low a 1 ng total RNA.
Internally, I have started with as low as 10 ng total RNA.

I don't think the problem is obtaining sequencing material (that's pretty easy) -- I think the problem is, how accurate of a representation what it is that ends up being sequenced is of its origin, whether starting with 1 ng, 10 ug, or using the Ovation System. Unfortunately, the received determination of this is typically through correlation coefficients, which are inappropriate as measures of agreement, but agreement (and not association) is what I take as intended to assess when we compare the efficaciousness of starting with 1 ng vs 10 ug, e.g.

[wenhuang]

I have used messageAmp from Ambion for amplification and was able to get sequencing to work, sort of...

One problem I have is reads are severely biased towards the 3'end. So recovery of full length transcript is very difficult.