Advantages of in situ hybridisation over direct or indirect in situ reverse transcriptase-polymerase chain reaction for localisation of galanin mRNA expression in rat small intestine and pituitary.

In situ hybridisation (ISH) and direct or indirect in situ reverse transcriptase-polymerase chain reaction (RT-PCR) were used to detect galanin mRNA in paraffin sections of rat intestine and pituitary. With conventional ISH, a subset of intestinal neuronal ganglion cells and anterior pituitary endocrine cells were labelled. Direct in situ RT-PCR also labelled some cells in pituitary but not in intestine. Negative controls were unlabelled, but sections with 3' primer alone for RT-PCR appeared positive. No signal was apparent using the indirect in situ RT-PCR method. Investigation of the specificity of solution phase RT-PCR using RNA extracts from pituitary or intestine revealed that additional PCR products were detected under some conditions. The sequences of these PCR products suggested that one was the result of mispriming and single primer PCR, which could also have occurred in situ. Alternative galanin primers gave only the predicted RT-PCR product in solution phase yet still gave artefacts in tissue sections using direct in situ RT-PCR. ISH with probes transcribed from the correct PCR product gave identical labelling to the original galanin riboprobe. In conclusion, direct in situ RT-PCR is unreliable and requires validation, while indirect in situ RT-PCR may fail even though sufficient target exists for detection with conventional sensitive riboprobe ISH.