How To Best Analyze Polysome Profiling

Question

[Klaus]

Hi everybody,

we are planning on conducting experiments to analyze different fractions of polysomes for ribosome profiling. We are interested in the differences between the high and low ribosome fractions, separated by sucrose gradient. The goal is to see what genes are differentially regulated between these two fractions

I read a paper (D.R. Morris) about the ribosome profiling experiment using RNA-Seq. As it's too expensive for us at the moment, we thought about running a microarray experiment with several replicates instead.

BUT as it's still extensive work, I would like to ask this community, whether it make sense at all to run an analysis of these two fractions with microarrays?

Does anyone know about published experiment of this kind using microarrays?

Are the two fractions comparable at all?

I would appreciate any kind of help or better suggestions.

Thanks Klaus

Answer 1

[Ido Tamir]

Heh, i'm even coauthor on one of the papers. And it feels like a different life [2003],[method]. With the caveat, that I am not up to date anymore regarding polysome bound RNA-profiling, my thoughts:

a) You anyway have to separate the fractions. Discarding the "free" one at the end does not make it less work.

b) I think you need the free fraction, because you want to know, what is only transcribed in the cell vs. what is also translated. Ideally I think would be a comparison: total/free/polysome-bound.

c) By all means go for RNA-seq. With multiplexing on a high-seq it should not be that more expensive. Unless, and here I forgot the experimental setup, its not possible for you to reduce the rRNA component significantly. Then you have to go for microarrays.

[Gareth Morgan]

You can deplete rRNA by subtractive hybridisation of your cDNA, or possibly with something like a Ribominus kit at the RNA stage. Depending on your organism you might want to test which of these works best. You do need total RNA data for normalisation. Cost-wise, a single HiSeq lane has enough capacity to sequence several barcoded samples at the enough depth to measure differences between your samples, for about the same price as a microarray. There's a good description of the Ingolia/Weissman protocol here.

[Gareth Morgan]

You can deplete rRNA by subtractive hybridisation of your cDNA, or possibly with something like a Ribominus kit at the RNA stage. Depending on your organism you might want to test which of these works best. You do need total RNA data for normalisation. Cost-wise, a single HiSeq lane has enough capacity to sequence several barcoded samples at the enough depth to measure differences between your samples, for about the same price as a microarray.

[Gareth Morgan]

Thee's a good description of the Ingolia/Weissman protocol in Methods in Enzymology 470:119

Answer 2

[Neilfws]

I suggest you start by searching the GEO database using the keyword "polysome". My search returned 595 microarray experiments.

You can then link that search to associated publications by going to the "Find related data" box (on the right side of the results page) and choosing PubMed - currently returns 58 items.

Answer 3

[Eric Fournier]

I cannot say anything about splitting the high and low ribosome fraction, but I know someone in my lab did microarray analysis on scarce polyribosomal fractions:

http://www.ncbi.nlm.nih.gov/pubmed/21324132