Populations of retinal neurones marked with DAPI in lower vertebrates

Retinae of teleosts, the marine sparid Boops boops and the fresh water cyprinid Carassius auratus, and amphibians, the anuran Rana pipiens and the urodele Ambystoma tigrinum, were stained in vitro with DAPI (4'6-diamidino 2-phenylindole). In all the preparations tested DAPI consistently stained nuclei in the ganglion cell layer, and three levels of nuclei could be observed from IPL (inner plexiform layer) to INL (inner nuclear layer) in the bogue, goldfish and frog retinae. In the goldfish retina a dense mosaic of stained horizontal cell nuclei was also observed. Both single and double cones were stained in fish; no photoreceptor staining could be found in amphibian retinae. Goldfish and frog retinae were incubated with 5'7'-DHT (dihydroxytryptamine) to compare the distribution of DAPI-stained cells with that of putative serotoninergic neurones. Fluorescent cells were found in the ganglion cell layer, and at two levels distally. In fixed retinae only a regular array of cells was found in the proximal INL, and interestingly the cell density equalled the density of serotoninergic amacrine cells. No other fluorescing cells could be detected. In the fixed frog retina two populations of fluorescent cells were found in the INL. For both species tested 5'7'-DHT and DAPI fluorescent populations did not overlap. The images of fluorescent cells were then processed, in order to improve image quality and assess if, in living tissue, cell types could be separated according to differences in intensity of fluorescence. It emerged that the size of fluorescent nuclei is inversely proportional to their optical density.