Formation of certain cell-cell adhesive contacts and avoidance of others are essential to morphogenesis. Differences in adhesive specificity often correlate with differences in cell fate, suggesting these two processes are linked. We are using the cells that give rise to the C. elegans male sensory rays as a model to study this linkage. The sensory rays derive from cells that display adhesive specificity: adherens junctions form between cells of the same ray, and not between cells of different rays. The Hox gene
egl-5 and the Pax6 homolog
mab-18 are required for adhesive specificity in the cells in which they are expressed, but how they perform this function is unknown. Both genes encode transcription factors, suggesting that they transcriptionally activate mechanisms that promote adhesive specificity. Signaling pathways involving the Semaphorin MAB-20 and the ephrin MAB-26 are candidates for such mechanisms. Absence of either MAB-20 or MAB-26 causes a high frequency of ray fusion, indicating that both proteins are necessary for adhesive specificity. Ephrins and Semaphorins mediate repulsion to axons and migrating cells in other systems. Hence, as a predicted secreted protein expressed in cells of all rays, MAB-20 was hypothesized to prevent the initial formation of inappropriate cell-cell contacts, via inhibiting extension of processes from ray cells (Roy et al, Development 127, 755-67). We found that a rescuing
mab-26 reporter is expressed in ray cells. As a membrane-bound factor, we hypothesize that MAB-26 acts as a contact repellent between cells of different rays. Interestingly, the
mab-26 reporter is expressed at levels that differ between rays, with an alternating OFF-ON pattern in cells of rays 1, 2, 3 and 4, strong expression in ray 6 cells, and intermediate expression in cells of rays 5 and 7. We tested the hypothesis that
egl-5 and
mab-18 promote adhesive specificity through activating expression of
mab-20 or
mab-26 . In
egl-5 mutants, rays 2, 3, 4 and 5 form a single fusion; in
mab-18 mutants, rays 4 and 6 fuse. Expression of a
mab-20 reporter is wild-type in
egl-5 and
mab-18 null mutants, indicating that neither gene operates through activating
mab-20 expression. In contrast,
mab-26 reporter expression is altered in both
mab-18 and
egl-5 .
mab-26 reporter expression is reduced in ray 6 cells in
mab-18 , suggesting that
mab-18 ray fusion might result from insufficient
mab-26 -mediated repulsive signaling from ray 6 cells to ray 4 cells. However, a transgene that drives
mab-26 expression in ray 6 cells and rescues ray 6 fusion in
mab-26 does not rescue ray 6 fusion in
mab-18 . We conclude that
mab-18 promotes adhesive specificity of ray 6 cells through a mechanism in addition to activating
mab-26 expression. In contrast to
mab-18 ,
egl-5 appears to repress
mab-26 expression. In
egl-5 , the
mab-26 reporter is ectopically expressed in ray 3 cells. Because ephrin-Eph interactions can mediate adhesion, we considered the possibility that fusion of ray 3 with either ray 2 or ray 4 in
egl-5 results from an adhesive interaction between ectopic MAB-26 in ray 3 cells and a receptor in cells of rays 2 or 4. Consistent with such an interaction between cells of rays 3 and 4, driving ectopic expression of
mab-26 in ray 3 cells in a wild-type background causes a low but significant frequency of ray 3-4 fusion. However, ray 3 cells still make ectopic contacts with ray 4 cells in the
egl-5 ;
mab-26 double null mutant, indicating that ectopic
mab-26 expression alone cannot account for ray fusion in
egl-5 . We conclude that
egl-5 promotes adhesive specificity between cells of rays 3 and 4 through a mechanism in addition to repressing
mab-26 expression. Our results indicate that
mab-18 and
egl-5 control expression of
mab-26 in distinct manners that promote adhesive specificity. However, this control is not sufficient for adhesive specificity, indicating that
mab-18 and
egl-5 perform additional functions. These functions could regulate activity of
mab-20 or
mab-26 -dependent pathways, or other pathways yet to be identified.