By feeding male worms on bromodeoxyuridine (BrdU) labelled E. coli and using a monoclonal antibody to BrdU, we have been able to identify paternal DNA strands within the next generation of developing embryos. The following is our current protocol, although it is still being improved. A thymidine requiring E. coli strain was grown to saturation in M9 salts supplemented with 0.4 % glucose, 1mM MgSO4, 1.25 g/ml vitamin B1, 9 M BrdU and 1 M thymidine. The culture was concentrated and ampicillin was added so as to prevent contamination of plates with unlabelled bacteria that the worms seem to find tastier. A mixed population of males and hermaphrodites (N2) at dauer larva stage were transferred onto M9-agarose plate with this BrdU-labelled E.coli. After two days at 20 C, male worms were transferred onto a new M9- plate, and were continued being fed on BrdU-bacteria for an additional two days. The incorporation of the label can be seen marching down the gonad over a period of a few days. The labelled males were mated to spermless hermaphrodites,
fem-2, (raised at 25 C) on a NGM-plate with non-labelled E. coli (OP50). The hermaphrodites also carried a spontaneous dumpy mutation that turned out to be convenient for separating unlabelled hermaphrodites from any contaminating labelled hermaphrodites inadvertantly brought in with the males. After about five hours of mating, dumpy hermaphrodites were picked up and cross- progeny were obtained by cutting the worms in half followed by gentle hypochlorite treatment (0.4-0.6 % NaOCl in egg-salts, 3 min.). The embryos, from which egg-shells were removed by chitinase-chymotrypsin digestion, were transferred onto a subbed slide and gentle pressure was applied on the cover-slip to spread the cells to avoid cell-cell overlap as well as to permeabilize the embryos. To detect the labelled DNAs, we used monoclonal anti-BrdU antibody conjugated with fluorescein (Becton-Dickinson). Prior to antibody binding, DNA was denatured in 6N HCl for 3 minutes, followed by neutralization in a large excess of 0.4M borate buffer, pH 8.5 for ten minutes. After antibody binding (30 l of a 1:10 dilution per slide), the DNA was stained with DAPI. Distinct green spots appear within the embryos under epifluorescence. Straightforward controls seem to work; green spots are not seen in unlabelled worms; antibody is competed off with BrdU; green spots go away with DNase I. In embryos with total cell number of 100 or more, the number of green spots seems to converge on 10 +/- 2. Our initial conclusions are that: (1) sister chromatid exchange does not appear to occur at high frequency, and (2) sperm DNA strands do not co-segregate, since this would predict only two labelled cells in embryos of all ages. A few preliminary experiments with larval worms suggest that strand segregation could be random, i.e. green spots appeared at different positions in different larvae. We should be able to test far more rigorously whether DNA strand segregation is truly random. Further experiments should be possible by variation on the above technique; for example, following maternal DNA strands, following both paternal and maternal strands in the same embryo, following strand segregation in post-embryonic cell divisions and so on.