The perfect RT??? which control? - (Dec/15/2009 )

Hi Guys,

I'm pretty sure there are some topics quite similar but I need some precisions and no topic gave me all the answers. So sorry if this topic is redundant, and thank you for your help.

Here is my situation:
I have a lot of RNA extracted from Blood. Half were extracted using the Paxgene kit (Qiagen) and the second half are being extracted using trizol (Invitrogen). We use 2 different kits because the first bloods were collected by ourselves, not the 2nd one...

When I arrived in the lab all the samples were converted into cDNA (fortunately not the whole RNA). But the guy who was here before me never assessed the quality and quantity of the RNA and convert 10ul of RNA each time...


My questions are the following:
1- To assess the quality what do you recommend? I use the bioanalyzer 2100 from agilent. But what is the minimum RIN we should use? My blood from paxgene are around 8~9, but the trizol RNA are between 6.5~7.5.

2- For the quantification? Spectrophotometer, nanodrop, bioanalyzer? Did someone make a comparison of these different techniques?

3-Which concentration should I normalize my samples?

4-For the no-RT control, a lot of people still use agarose gel... Do you think the sensitivity of agarose gel is good enough? I did a no-RT control on our real-time PCR using the SYBR kit from ABI. And unfortunately, I got amplifications around 33 cycles for the paxgene RNA and a bit earlier for the trizol RNA. Paxgene kit contains a DNAse step, so my former colleague expected that no DNA remained before cDNA conversion. But finally that's not the case. Is there any tolerance for a gDNA contamination? Is a real time PCR too sensitive?

5-With my trizol samples, I tried to treat some with DNAse I free DNA kit from Ambion. Unfortunately, I still have an amplification around 36~37 cycles. Moreover, on agilent (Nano) it seems that it completely killed my RNA... I still need to launch it with the pico kit to see if I still have something


Thanks a lot for the ones who will answer me, and can you give me some references so support your argument, cause I will need to explain it to my PI...

All the best

Grake

Hi there,

1. I would say don't base only on RIN only. Look at the gel picture also and the nanoDrop A260/A280 ratios also. After all, a higher RNA RIN would just give you a better probability of getting good cDNA.

2. Yes all three would give rather different concentrations. Just choose one and go with it.

3. Depends on what you do with your cDNA and what you want to study. I would say either adjust your RNA mass (eg. 1ng each RT reaction) if you have or not have a housekeeping gene to normalize to OR RNA volume (eg. 10uL each reaction) only if you normalize to a housekeeping gene.

4. Hmm 33 cycles. I would say run that reaction on agarose gel/Bioanalyzer DNA and check for primer dimers. If you have DNA contamination, it is likely that the Ct would be much earlier. Check the DNase-treated RNA on FA gel to see if you have a DNA band/smear above the RNA bands. Is real time more sensitive? I would say with great technology comes great sensitivity.

5. What is your average Ct for the samples? If they are about 28 or less, your 35-37 Ct is probably primer dimers. How're your NTC controls?
Also, what do you mean by it killed your RNA? Degraded?

chrisbelle on Dec 16 2009, 01:40 AM said:

Thanks a lot for your answer.

Concerning :
1- don't you think the 28s/18s ratio using an agarose gel is not quantitative enough?
2- I usually use the nanodrop that give (in my opinion) the best results in term of results quality/labor intensity... But since I really have a lot of sample and not so much time I want to use a 96 well plate spectrophotometer, but not really sure about the reliability.
4-I didn't run a gel but according to my melting curve it seems to be my amplicon and not primers...
5-My average Ct is around 27~28 cycles. my NTC is most of the time clean, sometimes with a late amplification. I said the treatment killed my RNA because I did not see anything left on agilent... really small peaks giving a RIN at 4 max.

Thanks,

Grake on Dec 16 2009, 09:38 AM said:

Problem is, if you don't run a gel, you really cannot know what size product QPCR is seeing. Some primers do funky things if they are amplified enough. One control is to do a no template well with the suspect primers. They will still amplify at 33 cycles if primer dimer is the issue. In general 33 cycles is really late to amplify for a "real" product. You can also try adding more template for the suspect reaction - like 4x as much. If it is genomic contamination, and not primer dimers, you should see the ct value move up to 31 (make sure to run side by side with the lesser amount of template).

I always run a gel if I have issues with my QPCR. The gel may not be quantititive, but if your product has amplified, you will see it (and you'll know if it's the product and not primer dimer).

Gel is also the only way you can know if your RNA is really degraded. I don't really know why your RNA would be completely lost after DNAseI digest. My guess is you probably lost your pellet in a later precipitation step? If your RNA was degraded by RNAse, you should still see something on nano, though it will have become degraded when seen on a gel (rRNA will be lost and you will have a smaller mRNA "smear").

Also of course nanodrop and other specs can't distinguish between RNA and DNA. So if your RNA was already degraded before DNAse digest, you actually might have lost your "RNA" signal because what you were really seeing was only DNA contamination. Again the only way you can know is by running your RNA out on a gel.

If you have tons of money you can use an agilent bioanalyzer instead of a gel. This has the added benefit of being able to distinguish RNA and DNA. It's really finicky though, from what others who use it have told me.

Grake on Dec 16 2009, 07:15 AM said:

Regarding 1st and 2nd question. Somene told me this before:
Certain sample has no problem to get RIN of 10 (my friend who work with C.elegans always got 10). My cell lines has no problem to give me 8 and above. However, my another labmate who uses animal model has problem getting that high. Her samples always around 6-8. If your samples behave like that, always choose samples with similar range of RIN. If majority of your samples ranged around 6-7, don't compare to those RIN 8-10.

I used both Bioanalyzer and Nanodrop. They often give different reading (esp. concentration of RNA). So if you use Nanodrop to quantify RNA concentration, continue to use the same machine throughout your experiment/ project. Same thing goes to Bioanalyzer. Hard to tell which is more accurate. The trick is, if you start your experiment with that particular machine, stick to it. Then it should be fine. I used to quantify my RNA concentration using Nanodrop but run Bioanalyzer just to check the RIN no.