A critical evaluation of Bernhard's EDTA regressive staining technique for RNA.

The EDTA regressive staining procedure to detect RNA (Bernhard's technique) is based on the proposition that after staining ultrathin sections with uranyl the stain is preferentially removed from DNA rather than RNA by the action of the chelating agent EDTA. Whilst attempting to use the EDTA regressive staining procedure to detect the presence of RNA in the large granule complexes of chicken erythrocyte nuclei, certain anomalous staining patterns were observed in the chromatin of these nuclei. Essentially, these were that the edges of condensed chromatin bodies stained positively for RNA even though this molecule is known not to be present there in significant quantities. The staining patterns suggested that chromatin was retaining its stain in a manner expected of RNA but not DNA as a consequence of EDTA-containing species failing to pass freely through the section. This hypothesis was tested by carrying out the EDTA procedure on embedded specimens of a DNA-containing virus, simian virus 40 (SV40), small enough not to be exposed at the surface of the section. In this way it has been shown that virus particles completely surrounded by resin destain so much more slowly than chromatin, which is accessible at the surface of the section, that without any other information it would be concluded that the viruses contained RNA, not DNA. This apparently anomalous result arises because the difficulties encountered by stain or EDTA molecules in passing through a plastic section were not appreciated at the time of the initial publication of the technique. The observations are discussed in the light of recent knowledge that has been gained on the kinetics of staining by measuring the electron-scattering densities of stained sections, similar measurements having been made on sections stained by Bernhard's technique. A model for the mechanism of the EDTA regressive staining technique consistent with the experimental observations is proposed and the conditions under which Bernhard's staining procedure retains its specificity are defined. Briefly, these conditions are that: (a) the sections be stained for only a short period with uranyl before treating them with EDTA even though such brief staining is undesirable for quantitative measurements of stain uptake into biological material; and (b) that sections stained in lead only be compared as controls with sections stained by Bernhard's technique so that any specificity of lead for sub-cellular components is not confused with a positive indication of the presence of RNA. Unless these conditions are fulfilled, results obtained by the use of the regressive staining technique may be highly misleading.