Introduction

The early development of in situ hybridisation relied heavily upon the use of radioisotopes and autoradiography for the visualisation of specifically hybridised gene probes (Buongiorno-Nardelli & Amaldi, 1969; Gall & Pardue, 1969, 1971). Radioisotopic labelling retains an important role, especially where maximum sensitivity is required, but in recent years a group of techniques employing non-radioisotopic labelling technology have been developed. They have been designed to avoid the perceived drawbacks of radiolabels in terms of safety and inconvenience.

In this review I hope to provide a flavour of the range of non-isotopic techniques available, to discuss their relative merits and to divine future developments. Some general methodologies are outlined but these must be considered as general guidelines from which optimal protocols must be individually determined (some detailed protocols are provided in later chapters).

Aim

In situ hybridisation identifies specific DNA or RNA sequences in tissue sections. What questions can in situ hybridisation answer that other methods cannot?

Standard molecular biological methods using gel electrophoretic separation of restriction fragments is ideal for characterising genomes, for viral detection or general changes in gene expression. However, in certain respects these methods have severe limitations. Extremely small quantities of nucleic acid can be detected by filter hybridisation but the identity of the cell containing the target sequence is unknown. Therefore electrophoretic separation methods provide no distributional information but are rather an accurate measure of the average nucleic acid content of a tissue homogenate. Only in situ hybridisation provides the capacity for the identification of specific target sequences within a mixed cell population. This depends upon the presence of localised areas of high target sequence concentration within a background of non-homologous sequences.