We set out to develop non-radioactive in situ hybridization methods with the goal of eventually obtaining single-cell revolution. Fluorescent detection (FD) was chosen because this method is sensitive enough to detect single-copy DNA sequences and has allowed detection of mRNA in cultured cells with subcellular resolution. We adapted fixation protocols that were developed for the detection of the
lin-14 protein for detection of mRNA using modifications of procedures for Drosophila. In order to detect hybridized probe, we used a monoclonal antibody specific for digoxigenin-labeled DNA (described by L. Avery WBG 11(1) 68) and fluoresceinated secondary and tertiary antibodies. At present, our protocol is best used for detecting abundant mRNAs in embryos and L1 and L2 larvae. We will send the protocol on request. With this protocol, we are able to detect pharyngeal
myo-3 in mRNA in embryos, L1 larvae, and L2 larvae. Identical results are obtained using either
myo-1 or
myo-2 probes. Our results indicate that pharyngeal myosin mRNA is exclusively synthesized in pharyngeal muscle cells, not a surprising finding. However, we were surprised that the level of resolution was equal to that obtained by using antibodies to pharyngeal myosin. The results indicate that pharyngeal myosin mRNA is first expressed late in embryogenesis at about the three-fold stage. Partially staining pharynges are never seen, indicating that the pharyngeal muscle cells must all synthesize pharyngeal myosin mRNA at about the same time. Pharyngeal myosin mRNA could also be detected using immunocytochemistry (Genius Kit). Comparison of FD with alkaline phosphatase based immunochemistry indicated that FD had greater resolution and lower background, yet had comparable sensitivity to the enzymatic method. An advantage of FD was that staining of interior regions could be visualized in the presence of hypodermal staining. Moreover, the results could also be viewed with the confocal microscope. The current protocol, though still not as sensitive as we would like, has allowed the detection of mRNA for the heterochronic gene
lin-14, the nuclear hormone receptor gene
crf-2, and the POU-homeobox gene ceh- 18. A major problem is that the signals for each of these three lower abundance mRNA's are speckled. Possibly, the speckling is an artifact, due to a failure to detect all of the mRNA or because the antibodies promote the formation of aggregates. Another difficulty is that the conditions for detection of lower abundance mRNA distort the morphology (e.g., swelling of the nuclei), making it hard to assign hybridization to specific cells. Despite these caveats, we have been able to obtain some useful information. Our results indicate that most cells of three-fold stage embryos contain
lin-14 mRNA, whereas no staining is seen prior to the two-fold stage, in agreement with the antibody studies. In contrast,
crf-2 mRNA is detected only during the first half of embryogenesis.
ceh-18 mRNA is first detected at the three-fold stage in the pharynx, ventral cord neurons, and numerous but unidentifiable cells in the head. Our method does not allow detection of mRNA'a for genes that are expressed at an even lower level (e.g.,
unc-86,
lin-11,
ced-4, or
mab-5).