Protocol Online logo
Top : New Forum Archives (2009-): : PCR, RT-PCR and Real-Time PCR

Upscaling cDNA synthesis using SuperScript™ III Reverse Transcriptase - (Aug/31/2012 )

Hi all,
I am planning to do RT-PCR on RNA samples (5ug) which are quite precious because they are not mine, they were a gift.
For the cDNA synthesis, I am going to use Superscript III RT with Oligo dTs. For the RT-PCR, we generally dilute the cDNA 1:10 and use 5ul per well, and set up each reaction in triplicates.
Since I am going to test 10 genes (including normalisation), and because the RNA is not mine and I can't go back and scrounge more of it, I calculated all the volumes I will be needing and it turns out the standard reaction volume of 20ul for cDNA synthesis with this RT is going to be enough for about 13 reactions. In case something goes wrong, I don't have a big margin to play around with the RT-PCRs.

So, I was wondering whether it is possible to upscale the reaction volume for the cDNA synthesis from the standard 20ul (in the protocol) to 30ul.
I haven't found any info anywhere- except there is another Invitrogen kit which allows for upscaling cDNA synthesis, but I'd rather not buy that if I can get around it.

Does anybody here have experience with upscaling the Invitrogen Superscript RT? Is it not recommended because e.g. the reaction is most efficient in 20ul? Or is it okay to increase the volumes?
I'm slightly hesitant to go ahead with it because it's not mentioned on any MDS or handbook on the Invitrogen website that it can be upscaled.

Some insight into this would be greatly appreciated.

I'm not familiar with Superscript RTs, what amount do you regulary put to the RT reaction?
You need to consider that there is usually no way to tell the abundance of your genes, so some amount may be OK for one gene, but insufficient for a low abundant.

We have Transcriptor from Roche and standard amount of RNA we put there is 1 ug. We got 20 ul reaction and use 2 ul in 20ul reactions or 1 ul in 10ul (10% max as recommended). So in case of 20ul reaction that means only 3 different genes in triplicates (2 ul in triplicates, i.e. maximum that can be used is 18 ul) can be run single cDNA.
If we need to run more genes, we increase the amount of RNA put into RT (Transcriptor limit is 5 ug) and then dilute the RT mix. So after that you have 40 ul of cDNA with similar concentration as the cDNA mentioned before. This can go further to 3 ug, 4 ug,...
Transcriptor manual says they kind of ensure the linearity within recommended range.
This is our setting.
BUT if I remember Superscript III is considered to have higher efficiencies, so the ratios may be different in your case. However this is due to reduced RNase H activity of SIII, but that is definitelly recomended for real-time quantitative experiments, so the efficiency of PCR is then lower than for sscDNA.

Also, Roche kit can use random hexamer, oligo dT or combination of both (which we use), oligo dT is said (particulary by Invitrogen) to have more more cosisntency, but lower yield. I think by combining both you can have compromise.

You didn't write how much RNA are you goint to put to RT. Because if you put there 2.5 ug and then dilute 10 times and run your genes, if that failes, you have another 2.5 ug of RNA to use.

You upscale not by incresing volume, but increasing RNA in reaction up to recommended volume (though the manual for S III First strand kit I read specifies to use bigger volume when adding more than 5 ug, which you won't be anyway). You didn't said which kit you have, but I think it's the same for all of them, they jus vary in components.


Hi Trof,
thanks for your quick response.
You need to consider that there is usually no way to tell the abundance of your genes, so some amount may be OK for one gene, but insufficient for a low abundant.
That is exactly the reason why I was thinking of using all of the 5ug of RNA I have for the cDNA synthesis is because I am going to look at isoforms of transcription factors in response to different treatments. I suppose being transcription factors they won't be abundantly expressed so I thought by using all of my RNA I should hopefully get enough cDNA to see a difference- if there is any at all.
I don't have any experience at all with RT-PCR and expression levels of transcription factors and whether they can easily be detected by RT-PCR, so I thought the more I put in, the better chances to see something.
However, maybe 2.5ug would be already enough to detect the isoforms , and then I'd have spare in case everything goes wrong. But if the genes have very low expression levels and I have already used 2.5ug with no effect, then the other 2.5ug are kind of useless.

So how I would be doing the reaction is synthesise cDNA from either 2.5ug or 5ug RNA, using Superscript III RT with oligo dTs exactly as laid out in the protocol by Invitrogen, then dilute the cDNA 1:10 and run 5ul per well, prepare each condition in triplicate.

Do you have experience with RT-PCR for transcription factors? What would you do in my case?

There are some ways how to estimate the expected level, if you search for literature, if you know that gene would be upregulated rather than downregulated, there can be found expression levels in various tissues for certain genes and so on.

But anyway, regardles of what you found out (unless you find out all genes are massively abundant) you can always dilute more, but you can't concentrate after you decide to use certain amount od RNA. So maybe better approach would be to transcribe all 5 ug, dilute and run the genes, and if there are lower Cts than quantifiable limit (usualy 15 Ct), you can dilute the rest of cDNA another 10x (that shifts Cts around 3,3 cycle), that should be sufficient for a whole new run. Just remember that if you found some of the genes too concentrated in the first dilution, and you dilute another 10x, you need to run all housekeepings again with these genes.

So transcribe first some other similar-tissue/similar treatment or whatever RNA you have, that's not precious and if the housekeepings itself on the test RNA are out of limit, use less ug of your sample (but, there maybe some extremely abundant housekeepings, that can make problems, especially 18S is known for that, but that's not your case I suppose, since you use oligo dT for RT which is incompatible, in that case using less abundant houseekeping gene would be prefered). If you don't have usable values from housekeepings, you won'd be able to normalize anything, so that is the only limit or reason why not to use all 5 ug, in that case using all of the amount would be pointless.

I think otherwise using 5 ug at once would be the most safe solution, at certain pre-requisites:

First, be sure you can handle all neccessary steps well, RT, PCR pipetting, primers are OK and tested, reagents do work. Make sure, that RT really works well with 5 ug, test some other RNA first and measure the housekeepings Cts. Make all possible tests to be sure it would work, because you will only have one shot.

Then plan your work on your samples carefully, not in stress, picture everything in your head before you do that, make all calculations and double check them (running some other RNA before your samples may help you realize the things that have to be taken care of especially if you don't have much experience). Then maybe consider not running all the genes on one plate, pippeting to 96-well plate is very tiresome and if you do more plates at once you have higher chance of making mistake. Check everything, multiple times, all the time do to be sure you are not making any mistake, don't hustle. Make sure your plate design is right, remember you must put all samples you are comparing onto same plate, though you don't need to put all genes on the same plate. Be sure you know what your calibrator is gonna be and what you want to compare.

And just a little note, I've noticed you're using the notation "1:10" probably in the meaning of 10x dilution, but it is not. It means "1 to 10" while 10 times dilution means "1 in 10" (in the final volume 10 times bigger). It is a ratio of sample and diluent. Many people use and understand it wrong and it could cause many confusions, since 1:10 ratio actually means 11 times diluted.
It's better then to say 10x diluted, that have no possibility for an error.


Hi Trof,
once again your prompt advice is much appreciated. I think that's exactly what I'll do. Luckily, my primers have been designed and optimised by some company, so in theory they should work. Somebody else in the lab already used primers from them and they were just fine. The housekeeping genes are from the literature and have been successfully used (concentration etc optimised) in our lab by somebody else who did lots of RT-PCR (but he left unfortunately).
So in theory, there are established protocols and procedures in our lab which I'll follow exactly, but because I have never done it all by myself I am somewhat apprehensive.

But you are right, I think it is a good idea to first run the RT-PCR on some cDNA which was left by the other guy, it would give me an idea about what Ct of my target genes I can expect and give me some practice in handling. Also, I do have enough primers for 600 reactions, so they are not the limiting factor really. It didn't even occur to me.

btw- you are of course right, I would be diluting 1:9 to get a 10x dilution. 20ul cDNA in 180ul H2O. Just pure laziness.

i'll drop a line once i have gotten somewhere with all of this.
thanks again.