We are interested in understanding how the epidermis is specified as a tissue. We previously showed that
lin-26 null mutations result in degeneration of epidermal cells and support cells, suggesting that
lin-26 is required to specify and/or maintain the fates of non-neuronal ectodermal cells. To determine if
lin-26, which encodes a putative transcription factor, acts by repressing neural-specific genes or by activating epidermal and support cell-specific genes, we ectopically expressed the LIN-26 protein under a heat-shock promoter. Early heat-shock treatment (prior to the 50-cell stage) resulted in embryos that arrested with twice fewer cells than normal (250 to 300). Up to 80% of the cells (150-200 cells) expressed as expected the LIN-26 protein, and concommitantly, the MH27-responsive antigen, a marker of adherens junctions present in all epithelial cells (epidermis, support cells, intestine, pharynx, spermatheca and uterus). By contrast, heat-shock at the time of neurogenesis or expression of LIN-26 under the control of the
mec-3 promoter did not affect neuronal differentiation, suggesting that
lin-26 acts as an activator rather than a repressor. Because MH27 is not specific of a unique tissue, we attempted to determine if other markers normally expressed in the epidermis were similarly ectopically induced upon LIN-26 ectopic expression. The antigen recognized by the mAbs MH4 and MH5 (specific for hemidesmosomes), a
dpy-7::gfp construct (collagen) and a
let-502::gfp construct (seam-cell marker) were found to be expressed in 20 to 40 cells after induction of LIN-26 expression, i.e. in about 5 to 10 times fewer cells than the MH27-reactive epitope. We therefore suggest that ectopic
lin-26 expression endows cells with the potential to become epithelial. There are two genes located immediately upstream of
lin-26, which together with
lin-26 define a new zinc-finger motif, hence their names
lir-1 and
lir-2, for
lin-26 related genes. Characterization of
lir-1 and
lir-2 showed that
lir-2 and one group of
lir-1 isoforms are expressed in neurons, whereas another group of
lir-1 isoforms are expressed in the same cells as
lin-26 (i.e. epidermal and support cells). The presence of three adjacent genes encoding transcription factors with related DNA-binding motifs is reminiscent of the achaete-scute complex in Drosophila. To determine if
lir-1 acts in the same pathway as
lin-26, we performed RNA interference experiments. Injection of double-stranded
lir-1 RNA corresponding to the isoforms expressed in non-neuronal ectodermal cells, but not to those expressed in neurons, yielded 100% dead embryos. These embryos displayed a phenotype very similar to that of moderately strong
lin-26 alleles, i.e. a two-fold to three-fold arrest with defective integrity and/or enclosure. Moreover LIN-26 expression was down-regulated in these embryos suggesting that
lir-1 is involved in the same pathway as
lin-26 and that it might be an activator of
lin-26 expression. We will further report on the phenotype of a 3 kb deletion within
lir-1 that we recently isolated through a PCR approach.