Critical evaluation of specificity in electron microscopical radioautography in animal tissues

Mice injected with 3H-uridine were killed after various intervals of time, and comparable samples of blocks of liver were processed for electron microscopy with various fixatives. Some blocks were embedded and used for high-resolution autoradiography. Others were homogenized in cold perchloric acid after dehydration, subsequently treated to obtain in a soluble form the radioactivity corresponding to nucleic acids and proteins; electrophoresis of the labeled nucleotides of each fraction was then performed. Both techniques gave similar results. Biochemical data show that at least with labeling times up to 1 h, a significant amount of the radioactivity is due to acid-soluble RNA-precursors. It may be as high, or higher than the radioactivity bound to RNA in the case of osmic acid or double fixation. Similarly the autoradiographic studies suggest that the grains not only represent the sites of RNA biosynthesis, but also indicate the uridine transport and the presence of the RNA precursors pool. This phenomenon is due to the thickness of the tissue blocks which, unlike isolated cells or bacteria, are not washed free from the unincorporated labeled precursors by the usual procedure for electron microscopy. We propose to fix the tissues with aldehydes, to wash them with cold buffer for as long as two days, and to postfix them with osmium if necessary. The soluble radioactivity in the blocks is then reduced to one tenth of that obtained with the original technique, although at all times there remains an unnegligible labeling of proteins due to the metabolic pathway of uridine. The same study applied to 3H-leucine labeled liver indicates that here also caution is needed if osmium fixation is used.