To understand how hypodermal cell shape and adhesion regulate organismal morphology in C. elegans, I am analyzing the gene
vab-9.
vab-9 mutants have tail and body shape defects resulting from defects in either the attachment or distribution of circumferential actin filaments at the adherens junctions. In the normal embryo, circumferential actin filament bundles are evenly spaced along the anterior-posterior axis; their contraction results in the elongation of the embryo. VAB-9 is a putative four-pass transmembrane protein homologous to canine BCMP1 (Brain cell membrane protein 1) and similar to the PMP22/Claudin family of proteins. Claudins are the major protein components of tight junctions and mediate calcium independent cell adhesion. Despite the similarity of VAB-9 to the tight junction claudin proteins, VAB-9 localizes to the adherens junctions of all C. elegans epithelia. VAB-9 localization is dependent on other adherens junction components. In the absence of HMR-1 (cadherin), VAB-9 localization at the adherens junction is completely missing, whereas in the absence of HMP-1 (a -catenin) or HMP-2 (b -catenin), VAB-9 localizes at junctions, but the distribution about the periphery of the junction is disrupted. 3D time-lapse analysis of VAB-9-GFP localization in
hmr-1(RNAi) and
hmp-1(RNAi) animals using multi-photon excitation microscopy was carried out by Mathias Kppen and Jeff Hardin at the University of Wisconsin. In
hmr-1(RNAi) animals, VAB-9-GFP apparently fails to reach the plasma membrane, whereas in
hmp-1(RNAi) animals, initial VAB-9-GFP junctional localization is normal, but the distribution about the periphery of adherens junctions is rapidly lost. Ultimately, in
hmp-1(RNAi) animals VAB-9-GFP becomes concentrated into small, isolated puncta at the junction. Interestingly, HMR-1 localization is almost identical to VAB-9 in
hmp-1 and
hmp-2 mutants. In contrast to these findings, mutations in
vab-9 do not obviously disrupt the localization of HMR-1, HMP-1, or HMP-2. Together, these results suggest that VAB-9 is closely associated with the adherens junction and that VAB-9 affects actin organization by regulating the interaction between actin and the cadherin-catenin complex. Alternatively, VAB-9 may mediate interactions with the actin cytoskeleton distinct from those with cadherin and catenins. VAB-9 similarity to PMP22/Claudin proteins suggests that VAB-9 may have a role in cell adhesion, however,
vab-9 mutants alone have no obvious cell adhesion defects. Nevertheless,
vab-9 mutations enhance the cell adhesion defects of
ajm-1 mutants and
dlg-1(RNAi) animals. Transmitting electron microscopy (carried out by Paul Sims and Jeff Hardin, University of Wisconsin) revealed that large gaps are present between hypodermal cells of
vab-9;
ajm-1 animals. These gaps extend along the entire apical-basal axis of the hypodermal cells and are more extensive than the gaps in the apical junctions of
ajm-1 mutants alone. Since VAB-9 is localized apical to DLG-1 and AJM-1, one possibility is that VAB-9 and DLG-1/AJM-1 independently regulate hypodermal cell adhesion. To determine how VAB-9 is associated with the cytoskeleton, I am currently searching for VAB-9-interacting proteins. To test directly for the ability of VAB-9 to mediate cell-cell adhesion, I am expressingVAB-9 in L fibroblasts. I thank the Hardin Lab (experiments), the Priess Lab (lab space, reagents, advice), John White (multi-photon excitation microscopy) and David Hall (TEM advice).