Morphogenesis in C. elegans is initiated when the hypodermal cells migrate to enclose the embryo and contract to force the ball of cells into a worm like shape. Morphogenesis is also dependent on body wall muscle which becomes structurally connected to the hypodermis during morphogenesis. Most
mup-4 mutants arrest late in morphogenesis at three-fold with body wall muscle cells mispositioned to the dorsal side of the worm: mutant embryos have a characteristically kinked shape. This muscle position defect is similar to others in the Mup class. Time lapse video tape analysis reveals that the
mup-4 muscle position defect is a result of body wall muscle detaching halfway through three-fold. These data suggest that muscle initially positions and attaches properly but muscle detaches during contraction. Mosaic analysis demonstrates that
mup-4 is not essential in embryonic muscle since it is required in the AB lineage, which gives rise primarily to hypodermal and neuronal cells. Consistent with a function in the hypodermis, we have observed three classes of embryonic hypodermal phenotypes by immunofluoresence: bean, two-fold and three-folds. For the rare embryos that have either the bean or 2-fold morphology, the hypodermis appears not to enclose these embryos. This arrest morphology is consistent with our light microscope quantitation of variable expressivity of
mup-4 alleles. These two arrest phenotypes suggest an early hypodermal function essential for enclosure or cell-cell interactions during morphogenesis. In contrast, the hypodermis encloses the three-fold Mup class, but hypodermal cells are abnormally shaped. Based on extensive analysis of developing populations we think that the hypodermal defects of three-fold Mups arise coincident with muscle detachment during three-fold. Therefore, we suggest two hypotheses to explain these data: 1-
mup-4 is required directly for muscle/hypodermal and hypodermal/hypodermal interactions or 2-
mup-4 is required in hypodermal cells (either for cell-cell interactions or for structural integrity of the cells) and that muscles detach because the hypodermis cannot withstand the force of muscle contraction. We have tested for interaction of
mup-4 with
mup-1 and
mua-3, two genes that also cause a muscle detachment defect and are hypodermally required. The terminal phenotype of
mup-1 mutants is similar to
mup-4 and
mua-3 has larval muscle attachment defects. These studies did not reveal any interactions or redundant functions. Molecular analysis of
mup-4 provides the opportunity to investigate embryonic cell functions that are essential for morphogenesis. We genetically mapped
mup-4 into the 0.02 cM genetic interval between
mec-14 and
lin-39 III. We have further refined this interval by deficiency mapping with two genetic deficiencies that complement
mup-4 and have breakpoints in this interval (David Baillie). This corresponds to a physical interval which is approximately 100 KB of unsequenced YAC and 50 KB of sequenced cosmids. In the 50 KB we have identified a strong candidate gene
mrp-1 (
mua-3 related protein-thanks to J. Plenefisch).
mrp-1 is a good candidate because of its homology to
mua-3 which, like
mup-4, is hypodermally required and causes muscle detachment. Furthermore, conceptual translation of
mrp-1 predicts a protein of at least 200 KD with 24 EGF-like repeats, cartilage matrix binding protein motif, type XII Collagen and a transmembrane domain. This structure would be consistent for cell-cell interactions and attachment. We are currently testing if
mup-4 is indeed
mrp-1.