I need to make RNA:DNA duplexes. I can make 100 to 200 ug of mRNA through in vitro transcription, and I know how to use reverse transcription to make a cDNA library, but I have questions with this.

The mRNA I make with IVT is all poly A RNA, and the reverse transcription kit I have says to use 5ug of total RNA or 500ng of poly A RNA. I need to make at least 10ug of RNA:DNA, which is 20 times the suggested limit for poly A RNA, so a direct scale up probably isn't practical.

I can't find a literature precedent for making such large amounts of DNA through reverse transcription, everything I have found that uses RNA:DNA duplexes used short strands that were chemically synthesized and combined, my mRNA is 1800 bases long, so I can't just order an oligo, I have to make it.

EDIT : I can't comment on my own question.

The duplexes will be about 1800 bases long, it's for firefly luciferase.

Trying to make the DNA for the duplex from the original DNA template presents other issues. While it would be easy to cut it out from the plasmid, it would be double stranded, and I would have to mix with the RNA and heat it up to dissociate everything and hope that the RNA bound before the DNA strands reassociated. I also think it would complicate my controls, because it would be harder to say that any expression I see after delivering these duplexes came from the RNA and not the DNA. Of course synthesizing the DNA from the RNA could also introduce that problem.

Some crude calculations suggests that the dNTPs are the limiting reagent for a transcript this size, and that even with 5ug of mRNA I should have enough oligo dT primer for the reaction. I right now I am running reactions from 0.5 to 5ug of mRNA, 1uL of primer, and 4uL of dNTPs, to make sure I have enough. Hopefully the gel will turn out ok.

EDIT: I got the results. So far it looks like 3.0ug of mRNA works the best. It has the darkest band, and looks like the right length. Less mRNA gives fainter bands that appear too short, and more mRNA gives slightly fainter bands that appear slightly too short. Also, I ran the reaction for 2 hours instead of 1.

  • $\begingroup$ Can't you use the template you used to synthetize the RNA? You'd need two restriction sites to remove the parts you don't want, but using DNA to synthesize RNA and then using that RNA to synthesize DNA again seems a bit cumbersome. $\endgroup$ Oct 24, 2013 at 18:04
  • $\begingroup$ how big should your RNA-DNA duplexes be ? $\endgroup$
    Oct 25, 2013 at 5:01

1 Answer 1


The best way for scaling up these kinds of reactions is to set up many reactions. Prepare a mastermix for lets say 20 reactions. You can pool them together, precipitate the RNA and dissolve to appropriate concentration in nuclease free water. Also, use gene specific primer instead of oligo-dT and don't add poly-A tails (poly-A does not affect the in-vitro stability of RNA).

As Mad Scientist already indicated, you would not need to make DNA again from RNA. Simply take 1:2 molar ratio of your IVT template dsDNA (restriction digested) and the RNA (post IVT). Heat and anneal slowly. DNA-RNA hybrids are stronger then DNA-DNA hybrids and the former is more likely to form after annealing. You can digest the unpaired strands using enzymes like mung bean nuclease.

  • $\begingroup$ I had almost forgotten about this question. I think we've all but abandoned this idea, RNase H seems to be a problem with RNA DNA hybrids, we paired a PolyA tailed mRNA with a 25 base oligo(dT) and dosed it to mice, completely stopped protein expression. If we used an oligo(rU) the signal was not affected, so we suspect it's RNase H cutting the polyA tail off and starting the mRNA degrdation process. $\endgroup$
    – user137
    Nov 24, 2014 at 16:28

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .