Localisation of tight junction protein cingulin is temporally and spatially regulated during early mouse development

The molecular maturation of the tight junction in the mouse early embryo has been investigated by monitoring the distribution of cingulin, a 140 x 10(3) M(r) peripheral (cytoplasmic) membrane constituent of the junction, at different stages of development and in different experimental situations. Although tight junction formation does not begin until compaction at the 8-cell stage, cingulin is detectable in oocytes and all stages of cleavage, a factor consistent with our biochemical analysis of cingulin expression (Javed et al., 1992, Development 117, 1145-1151). Using synchronised egg and embryo stages and isolated cell clusters, we have identified three sites where cingulin is localised, the cytocortex, punctate cytoplasmic foci and tight junctions themselves. Cytocortical cingulin is present at the cumulus-oocyte contact site (both cell types), in unfertilised and fertilised eggs and in cleavage stages up to 16-cell morulae, particularly at microvillous domains on the embryo outer surface (eg. apical poles at compaction). Embryo manipulation experiments indicate that cortical cingulin is labile and dependent upon cell interactions and therefore is not merely an inheritance from the egg. Cingulin cytoplasmic foci are evident only in outer cells (prospective trophectoderm) from the 32-cell stage, just prior to cavitation, and decline from approx. 8 hours after cavitation has initiated. The appearance of these foci is insensitive to cycloheximide treatment and they colocalise with apically derived endocytic vesicles visualised by FITC-dextran, indicating that the foci represent the degradation of cytocortical cingulin by endocytic turnover. Cingulin is detectable at the tight junction site between blastomeres usually from the 16-cell stage, although earlier assembly occurs in a minority (up to 20%) of specimens. Cingulin assembly at the tight junction is sensitive to cycloheximide and is identifiable approx. 10 hours after cell adhesion is initiated and ZO-1 protein assembles. Collectively, our results indicate that (i) cingulin from nonjunctional sites does not contribute to tight junction assembly and (ii) the molecular maturation of the junction appears to occur progressively over at least two cell cycles.