It is well known that single stranded CpA and UpA steps are typically sites of strong autohydrolysis because the backbone at these steps can most easily adopt the "in-line" conformation that is required for attack by the 2'-OH on the adjacent phosphodiester linkage. For NMR studies, we would often engineer these sites away from YpA to RpA to minimize this problem, if that region was unimportant. So, your solution of changing the hotspot to a more "friendly" sequence is a very reasonable solution.

Because this problem is intrinsic to the RNA, adding RNase inhibitors is likely not going to suppress it. Rather, keeping your RNA in a 1 mM EDTA solution and around pH 6.0 will inhibit the chemistry (which is facilitated by magnesium).

When do the products show up? If they are during transcription, they may be sites where the pol stops. You may always get some degredation. Are you keeping metals out of the way with EDTA? How are you storing? Keep pH buffered around 6.5-7.5 - not 8.0.

Do you denature your RNA properly before in vitro Tx? It could, as was said by Amanda, really be highly structured regions that cause your RT to stop. Try adding some RNAse inhibitors to your reaction, if you not already have.

You could also include some standard RNA samples, where you know the RT should be alright, to check your workflow/reaction mix/degradation, etc.

@Amanda true that EDTA inhibits some nucleases, though each Pol, or more precisely to keep the NTPs in the "good" conformation, needs divalent cations, usually Mg2+. So don't put too much EDTA, better use RNAse inhibitors. Absolutly right to keep the buffer below pH7.

EDTA (usually 1 mM) is not to inhibit nucleases, it is to prevent metal-dependent cleavage of the RNA backbone by activating the 2'OH for nucleophilic attack. Only add during storage and elution, not during reactions. 1 mM is usually a small enough amount that it gets diluted away without incident during reactions.

Rnase inhibitors are a good suggestion too - I was surprised to learn acouple years ago that they're proteins!

It is well known that single stranded CpA and UpA steps are typically sites of strong autohydrolysis because the backbone at these steps can most easily adopt the "in-line" conformation that is required for attack by the 2'-OH on the adjacent phosphodiester linkage. For NMR studies, we would often engineer these sites away from YpA to RpA to minimize this problem, if that region was unimportant. So, your solution of changing the hotspot to a more "friendly" sequence is a very reasonable solution.

Because this problem is intrinsic to the RNA, adding RNase inhibitors is likely not going to suppress it. Rather, keeping your RNA in a 1 mM EDTA solution and around pH 6.0 will inhibit the chemistry (which is facilitated by magnesium).

P.S. Liked your PNAS paper on B12 riboswitch folding!

I agree with Amanda - try lowering the pH of your storage buffers . Also, you might consider using more kinase(?), but incubating less time during your labeling. Also, don't heat denature your kinase when the reaction is done - try P/C extraction (where you can keep the samples on ice). Finally, I don't know how you're purifying your RNA (ion exchange spin column? PAGE?), but you might consider using a native gel, which can be run at low temperatures and is thus easier on your RNA.

I'd only consider changing your sequence as a last resort. Seems to me you might end up wasting a lot of time "chasing your tail", so to speak...