I had dreadful experiences with in-situ PCR for HIV many years ago and ended up writing a paper about the difficulties in getting the technique to work. The whole technique relies upon being able to retain a "fixed" cage in which your target and amplicons are retained and are accessible to PCR reagents. I found 1) that most fixatives caused damage to DNA or put in cross-links that meant the DNA could not be denatured properly as it was bound to histones 2) if you could get PCR reagents in then you could also lose target out 3) The "leaked target" could enter neighbouring cell and be amplified in cells you knew did not harbour the target. For DNA at least the cut off size for amplification was the nucleosomal size (120 bp ) as the protease digestion needed to remove histones was usually detrimental to structure.

I had no experience with RT, but think that a number of the problems will be the same. I find it difficult to envisage how a nucleic acid sequence will not diffuse out of a punctured cell and not be amplified outside the cell unless you can anchor it - after all we routinely electroporate plasmids of several kb into cells through much smaller self-sealing poresl

I know this is not very helpful, but hope you do not waste as much time as I did.

Ian