The RhoGEF protein Plekhg5 regulates medioapical and junctional actomyosin dynamics of apical constriction during Xenopus gastrulation

Canonical Wnt pathway modulation is required to correctly execute multiple independent cellular dynamic programs during cranial neural tube closure

Defects in cranial neural tube closure are among the most common and deleterious human structural birth defects. Correct cranial closure requires the coordination of multiple cell dynamic programs including cell proliferation and cell shape change. Mutations that impact Wnt signaling, including loss of the pathway co-receptor LRP6, lead to defects in cranial neural tube closure, but the cellular dynamics under control of the Wnt pathway during this critical morphogenetic process remain unclear. Here, we use mice mutant for LRP6 to examine the consequences of conditional and global reduction in Wnt signaling and mutants with conditional inactivation of APC to examine the consequences of pathway hyperactivation. Strikingly, we find that regulated Wnt signaling is required for two independent events during cranial neural tube closure. First, global reduction of Wnt leads to a surprising hyperplasia of the cranial neural folds driven by excessive cell proliferation at early pre-elevation stages, with the increased tissue volume creating a mechanical blockade to efficient closure despite normal apical constriction and cell polarization at later stages. Conversely, conditional hyperactivation of the pathway at later elevation stages prevents correct actin organization, blocking apical constriction and neural fold elevation without impacting tissue scaling. Together these data reveal that Wnt signaling levels must be modulated to restrict proliferation at early stages and promote apical constriction at later elevation stages to drive efficient closure of the cranial neural tube.

Mechanical control of neural plate folding by apical domain alteration

Vertebrate neural tube closure is associated with complex changes in cell shape and behavior, however, the relative contribution of these processes to tissue folding is not well understood. At the onset of Xenopus neural tube folding, we observed alternation of apically constricted and apically expanded cells. This apical domain heterogeneity was accompanied by biased cell orientation along the anteroposterior axis, especially at neural plate hinges, and required planar cell polarity signaling. Vertex models suggested that dispersed isotropically constricting cells can cause the elongation of adjacent cells. Consistently, in ectoderm, cell-autonomous apical constriction was accompanied by neighbor expansion. Thus, a subset of isotropically constricting cells may initiate neural plate bending, whereas a 'tug-of-war' contest between the force-generating and responding cells reduces its shrinking along the body axis. This mechanism is an alternative to anisotropic shrinking of cell junctions that are perpendicular to the body axis. We propose that apical domain changes reflect planar polarity-dependent mechanical forces operating during neural folding.

Architecture of the cortical actomyosin network driving apical constriction in C. elegans

Apical constriction is a cell shape change that drives key morphogenetic events during development, including gastrulation and neural tube formation. The forces driving apical constriction are primarily generated through the contraction of apicolateral and/or medioapical actomyosin networks. In the Drosophila ventral furrow, the medioapical actomyosin network has a sarcomere-like architecture, with radially polarized actin filaments and centrally enriched non-muscle myosin II and myosin activating kinase. To determine if this is a broadly conserved actin architecture driving apical constriction, we examined actomyosin architecture during C. elegans gastrulation, in which two endodermal precursor cells internalize from the surface of the embryo. Quantification of protein localization showed that neither the non-muscle myosin II NMY-2 nor the myosin-activating kinase MRCK-1 is enriched at the center of the apex. Further, visualization of barbed- and pointed-end capping proteins revealed that actin filaments do not exhibit radial polarization at the apex. Our results demonstrate that C. elegans endodermal precursor cells apically constrict using a mixed-polarity actin filament network and with myosin and a myosin activator distributed throughout the network. Taken together with observations made in other organisms, our results demonstrate that diverse actomyosin architectures are used in animal cells to accomplish apical constriction.