total RNA (for miRNA) extraction from human serum - (Jan/05/2010 )

Hello everyone,

I have been - highly unsuccessfully - attempting to extract total RNA from human serum samples. This is not something new and has been widely described in the literature so I must be doing something fundamentally wrong.

Here is roughly what I do:

Sample collection
- Venopuncture, take blood in serum gel tube

- 1h at room temperature (clotting) (this should be deadly for RNA, but all publications I encountered do it like this)

- centrifuge for 15 Min.

- remove serum and store at -80°C

tRNA extraction
Have tried kits from Qiagen (miRNEasy), Applied (PARIS miRvana) and also modified phenol chloform extraction. All of these protocols were tried multiple times by different operators.
They all yield similarly catastrophic results with OD260/230 ~ 0.1 and 260/280 ~ 1.0. (Nanodrop and photometer, both in aqua dest. and in TrisHCl pH 7.5). All procedures yield very consistent results, no outliers in either direction.

Does anyone have any ideas what the problem might be? I'm pretty frustrated at this point and would highly appreciate any help!

Thanks and best wishes for a great 2010,
mulain

Dear Mulain

I have been doing very similar experiments to you - extracting RNA from serum samples. I've tried a few different methods and have found the Norgen Biotek Plasma/Serum Circulating Nucleic Acid Purification Kit (Slurry Format) to be the best. I've attached a product information sheet.

One thing I've found is that OD methods, such as the nanodrop, for measuring RNA concentration from these samples is highly inaccurate. It can't cope with the small RNA you isolate from serum samples. A better way to assess your RNA quality is using an Agilent Bioanalyzer or something similar. Agilent have a special small RNA chip designed to give you a detailed profile of the RNA you isolate. Unfortunately ours are out of date so I'm hoping to borrow a couple of chips from another group next week to fully analyse my samples.

I hope that has been useful, if you are based in the UK GeneFlow Ltd sell the Norgen kits and they have a good working relationship with Norgen to answer any technical questions.

Let me know how you get on
BW
Emily

Thanks for your help, Emily!
I contacted Norgen, and I'll probably try them out, but the kit is pretty expensive (shipping costs are considerable as well) so I'd like to hear about your results from the Bioanalyzer before buying it.

I tried the Bioanalyzer as well and it pretty much confirmed what the nanodrop was saying for my samples: no usable RNA. So it would be really interesting to hear about your results.

Cheers!

mulain on Jan 7 2010, 09:03 AM said:

Norgen is a good product but it will give you no better results than using TRIZOL. Forget quantitation with nanodrop as Emily rghtly told u, it's a waste of time. My protocol with trizol is 1ml of trizol for 400ul serum/plasma. after acqueous phase has been recovered add (optional but useful to see a pellet) 1-2 ul of glycogen before adding 1ml of isopropanol then proceed as usual. recovery is not so bad if need RNA for Real time pcr purpose. if u want to know if different samples had similar yields just run one test pcr with 1 miRNA (i use mir 17 or 16). Normalization: try a spike in. what i do is using a synthetic miRNA (Cel-mir 39 is ok) that can be added to the sample AFTER adding trizol and before chlorophorm. this needs some setting up for right concentrations of spike in but helps a lot with normalization even if i don't like synthetic miRNAs going around in my lab, contamination is HIGHLY probable!
hope it helps.
please let me know if u have good results with Bioanalyzer, i heard mixed opinions about that. we have no acces to that but we have a Bio Rad experion, but it's not good for miRNAs up to now
bye
Fiz

mulain on Jan 7 2010, 09:03 AM said:

One more thing i thinh can help.
here's a VERY good work by mitchell et al. give a look to supplementary mat. this should be helpful.
supplementary is too big to upload but everything is free just check:
Proc Natl Acad Sci U S A. 2008 Jul 29;105(30):10513-8. Epub 2008 Jul 28.
Circulating microRNAs as stable blood-based markers for cancer detection.

Mitchell PS, Parkin RK, Kroh EM, Fritz BR, Wyman SK, Pogosova-Agadjanyan EL, Peterson A, Noteboom J, O'Briant KC, Allen A, Lin DW, Urban N, Drescher CW, Knudsen BS, Stirewalt DL, Gentleman R, Vessella RL, Nelson PS, Martin DB, Tewari M.
PMID: 18663219

Fizban

Hi Fizban, thanks for your help. What do you make of these Bioanalyzer images?

They are what I got from total RNA recovery from serum samples...




Any help is greatly appreciated!

Cheers!

Love that paper! Have read it, cited it and hung it over my bed

Fizban on Jan 12 2010, 08:50 AM said:

Noone has any opinions on the gel pics in post #6?
Come on people

By the way: the company that makes the miRNA microarrays that we will be using told me that other groups had similar gel pictures (in regard to picture 2) from serum extraction. And these other groups have already (pretty nicely) published their results. But the support person from the company had no explanation for the missing ribosomal RNA...

mulain on Jan 14 2010, 08:55 AM said:

sorry for the delay. i have no experience with that kind of pictures so i can't commento on that. one thing i'd say about lack of rRNA in serum samples: RNAses are strongly present in serum and plasma so i'm not surprised that "canonical" RNAs coming from cells are not present. my personal idea is that circulating miRNAs must be protected in some way by binding to proteins of by being inside vescicles. the latter has been already demonstrated so i wouldn't worry about that. circulating miRNAs ARE stably present and that is now fully established, we now lack the details....
Fiz

Fizban on Jan 14 2010, 12:26 PM said:

Thank you again, Fizban! I guess I'll just try to run an array with the samples from the nicer pics!