How much total RNA for a transcriptome?

[Dangermouse]

We'd like to do a 454 transcriptome experiment on some primary, cultured cells and we were planning to just send total RNA to 454. They're asking for 50ug

On a good day we might get 10ug from a large flask - so how much total RNA is required for a transcriptome experiment? Having trouble getting a straight answer from anywhere and I'm pretty sure 50ug was never mentioned in our initial enquiries...

[shurjo]

50ug is absurd. Depending on the kind of transcriptome experiment (whether you want polyA selection/Ribominus) between 2 and 10ug (even 10 is a LOT) sounds more rational. On Illumina, even 0.5ug is comfortable. Also, as you have (presumably) fresh, intact RNA from the cells you are growing (i.e. not archival samples), you should be able to get by with the lower boundaries of the input amount.

[Dangermouse]

Thank you for that, I thought so too. We just want to look at gene expression so poly A selection would be the way to go. And yes, we've got fresh RNA with RINs up around 9.6. Hoping and presuming somebody has left out a decimal point somewhere along the line...

[jordi]

Hi,
We have done a 454 pyrosequencing with 5ug of RNA and poly A priming with a enzyme restriction site to remove the long run of A's. It could be useful for the following analysis steps.

[Dangermouse]

Well it wasn't a typo and 454 are standing firm on 50ug of sample.

They will accept a 10ug sample and try to construct a library from it, but have quoted US$8,900 to do the work and they think that there is nearly 100% chance of failure ie not getting enough mRNA from this.

Am I missing something obvious here? This sounds nuts.

[shurjo]

Try this another way: ask them if the can start with mRNA, and if so, how much mRNA they need. I suspect that you are getting a stock answer from someone at the company who is not aware of the day-to-day realities of laboratory science.

[Dangermouse]

That's a possibility I suppose; the GS FLX cDNA Library Prep. Manual calls for ≥200ng of mRNA and our original intention had been to take the sample prep. as far as mRNA ourselves. We have access to an Agilent Bioanalyzer and our total RNA preps look very nice but after attempting the polyA enrichment a few times the results didn't look great and still had some rRNA contamination.

We reasoned that we could continue to spend time and $$$s on perfecting our mRNA purifications or alternatively we could hand the task over to those who do this on a daily basis and entrust the mRNA prep. to them. But we hadn't reckoned on them insisting on 50ug. After all, if the experts can't confidently expect to purify sufficient mRNA for a transcriptome experiment from 10ug, what chance have we got? And I'm finding it difficult to reconcile their statement that 'there is nearly 100% chance of failure ie not getting enough mRNA from 10ug'; I guess it comes down to the proportion of mRNA to total RNA in any particular sample. Whereas we had anticipated it to be 5-10%, 454 quoted a value of 1% (although neither they, nor we, are basing that figure on any empirical data, we just don't know yet...).

If we can't get over this hurdle I guess we'll have to keep refining our mRNA purifications in the hope of securing enough to do our experiment or switch technologies. Certainly the Illumina route is evermore tempting for our application...

[shurjo]

Just to give you some ballpark figures, I estimate I have done ~100 mRNA extractions from 10ug amounts of total RNA, mostly from LCLs and blood. I usually get between 100-150ng of polyA RNA after two rounds of DynaBeads extraction. The BioAnalyzer pics for almost all of these will show some faint hint of the 18S and 28S bands; from speaking with people with much more experience than me, what I gather is that it is very very hard to get rid of this AND have good mRNA yields. Bioinformatically, I see ~5% of reads in the final data that map to the ribosomal clusters (this is on Illumina).

[jordi]

What about this??:

http://www.ncbi.nlm.nih.gov/pubmed/16738135

Maybe not only contamination...

[wildung]

By the time I'm satisfied with the quality of my mRNA I rarely have >1.0% yield. The Rapid cDNA Library calls for 200ng starting material, so 20ug should get you in the ballpark. I usually ask for a bit extra for QC and enough to repeat without bugging the researcher again if I drop it on the floor. For many 50ug of total RNA is easy to come by, but if it isn't, we work with what we have and the Rapid cDNA Library prep can be used with 50ng but you are pushing the stoichiometry and would need to adjust adapter concentrations. The yield becomes too low to quantify w/ FAM and requires more extensive titrations to get the proper enrichment levels. If you can get 10ug a day, it doesn't appear that availability of RNA is limiting to the project. People doing laser microdissection can spend weeks/ug

[wildung]

By the time I'm satisfied with the quality of my mRNA I rarely have >1.0% yield. The Rapid cDNA Library calls for 200ng starting material, so 20ug should get you in the ballpark. I usually ask for a bit extra for QC and enough to repeat without bugging the researcher again if I drop it on the floor. For many 50ug of total RNA is easy to come by, but if it isn't, we work with what we have and the Rapid cDNA Library prep can be used with 50ng but you are pushing the stoichiometry and would need to adjust adapter concentrations. The yield becomes too low to quantify w/ FAM and requires more extensive titrations to get the proper enrichment levels. If you can get 10ug a day, it doesn't appear that availability of RNA is limiting to the project. People doing laser microdissection can spend weeks/ug

[Dangermouse]

For many 50ug of total RNA is easy to come by, but if it isn't, we work with what we have and the Rapid cDNA Library prep can be used with 50ng but you are pushing the stoichiometry and would need to adjust adapter concentrations

Yes, but it appears 454 aren't willing to push the envelope for us. Must be making too much money...

If you can get 10ug a day, it doesn't appear that availability of RNA is limiting to the project...

I wish. These samples are cultured macrophages obtained from particular stud animals and subjected to a treatment; we get one shot. We can't just go and bleed another mouse...

[pmiguel]

Using magnetic bead-based polyA RNA isolation we were usually getting far less than 1% yields. Less than 0.1% was not unusual. We switched to the (older) oligo-dT cellulose spin columns and our yields (thus far) are now in the 1-2% range. (Recent results on another front make me wonder if the issue is not the use of heating blocks vs. water baths for the denaturation/annealing stages of the protocol with magnetic beads. Not interchangeable for this technique, I think.)

Another factor here is that people generally use UV spectrophotometry to determine total RNA amounts. But UV spec is highly confounded by a number of contaminants commonly found in RNA preps. These would include phenol and various guanidinium salts as well as acetate. (Not to mention DNA, or degraded DNA if you used DNAse).

So if you are a core facility accepting total RNA to construct 454 libraries and you use a nanodrop to check the RNA amount given, then you might end up thinking your poly A yields are much lower than they actually are. Hence you might ask for much more RNA than you might otherwise actually require.

That said, 200 nanograms of polyA RNA to make a cDNA library might seem like a lot. That converts to roughly 0.4 trillion (10^12) 1000 nt RNAs. How many library molecules actually end up being added to that emPCR mix? A few million, right?

I think the conclusion is that the Roche cDNA library creation methodology is not very efficient at conversion of mRNA to cDNA. We always ask for 100 ug of total RNA, if we can get it. If not we usually will give polyA isolation a shot. If it looks good, fine. If not, we tell the investigator we would need to use our back-up method (Smart kit) to generate the cDNA. There even 1 ug of total RNA is plenty.

--
Phillip

[grassgirl]

Philip,

I have just started sequencing cDNA with a 454 Junior system and am having lots of trouble getting enough material at the end steps of the cDNA prep. Can you give me some more details about your cDNA prep? After using the Smart Kit is the ds cDNA cleaned with AmPure Beads or some other method? I have optimized the fragmentation and done several different cDNA preps with 200 ng mRNA and don't get much back. When analyzed on a fluorometer prior to proceeding with the library prep, I'm getting 3-5 ng/ul which translates to about 80 ng or less total cDNA.