If you don't won't to extract the RNA from the gel (which I know migth be low res), you can also try RNAase A incubation, The RNAse A only cleaves single stranded RNA therefore leaving untouched your hybridized one. We use it to clean brains during the whole mount in situ from the probe that just remains trapped in the tissue.

Is the hydridized oligo RNA or DNA? How long are they? Do you need to remove unhybridized oligo only? or both unhybridized mRNA transcript and free oligo?

I would probably go with gel purification here but would use non-denaturing (e.g. TBE-acrylamide without urea or TAE-agarose depending on the size of the mRNA transcript). Only problem with this approach is you might have trouble resolving free transcript from hybridized transcript.

This is actually sort of a tough one I think, especially if you really want to be sure you have only duplex... The best plan I can come up with; use saturating oligo in the hybridization reaction, say 5-10X molar ratio, and assume this will give you nearly 100% hybridized mRNA and a bunch of free oligo (you could check this assumption by using nuclease on a trial sample). At this point either run the product on a non-denaturing gel, or if the oligo and mRNA are sufficiently different in size you could try to remove the oligo with a size exclusion spin column.

Thanks fr all your suggestions

@ george perdrizet..... ok i do hybridization with RNA(1000bp) and oligo(25bp) and i need to remove unhybridized RNA and oligo from this mix. Removing free oligos is reasonably easy.

Yes i thought about running it on gel and eluting it out.... but as you mentioned resolving unhybridized and hybrid is difficult.

So now i guess we have to get maximum efficency for hybridization... So now almost all the RNAshould be hybridized . I will try as you suggested ... thank you :)