Cadherin-Mediated Cell Adhesion Is Critical for the Closing of the Mouse Optic Fissure

Coloboma is a congenital disease that contributes significantly to childhood blindness. It results from the failure in closing the optic fissure, a transient opening on the ventral side of the developing eye. Although human and mouse genetic studies have identified a number of genes associated with coloboma, the detailed cellular mechanisms underlying the optic fissure closure and coloboma formation remain largely undefined. N-cadherin-mediated cell adhesion has been shown to be important for the optic fissure closure in zebrafish, but it remains to be determined experimentally how cell-cell adhesions are involved in the mammalian optic fissure closing process. α-catenin is required for cell adhesion mediated by all of the classic cadherin molecules, including N-cadherin. In this study, we used the Cre-mediated conditional knockout technique to specifically delete α-catenin from the developing mouse eye to show that it is required for the successful closing of the optic fissure. In α-catenin conditional mutant optic cups, the major cell fates, including the optic fissure margin, neural retina and retinal pigmented epithelium, are specified normally, and the retinal progenitor cells proliferate normally. However, adherens junctions components, including N-cadherin, β-catenin and filamentous actin, fail to accumulate on the apical side of α-catenin mutant retinal progenitor cells, where adherens junctions are normally abundant, and the organization of the neural retina and the optic fissure margin is disrupted. Finally, the α-catenin mutant retina gradually degenerates in the adult mouse eye. Therefore, our results show that α-catenin-mediated cell adhesion and cell organization are important for the fissure closure in mice, and further suggest that genes that regulate cell adhesion may underlie certain coloboma cases in humans.     

Curvature of the caudal region is responsible for failure of neural tube closure in the curly tail (ct) mouse embryo.

Delayed closure of the posterior neuropore (PNP) occurs to a variable extent in homozygous mutant curly tail (ct) mouse embryos, and results in the development of spinal neural tube defects (NTD) in 60% of embryos. Previous studies have suggested that curvature of the body axis may delay neural tube closure in the cranial region of the mouse embryo. In order to investigate the relationship between curvature and delayed PNP closure, we measured the extent of ventral curvature of the neuropore region in ct/ct embryos with normal or delayed PNP closure. The results show significantly greater curvature in ct/ct embryos with delayed PNP closure in vivo than in their normal littermates. Reopening of the posterior neuropore in non-mutant mouse embryos, to delay neuropore closure experimentally, did not increase ventral curvature, suggesting that increased curvature in ct/ct embryos is not likely to be a secondary effect of delayed PNP closure. Experimental prevention of ventral curvature in ct/ct embryos, brought about by implantation of an eyelash tip longitudinally into the hindgut lumen, ameliorated the delay in PNP closure. We propose, therefore, that increased ventral curvature of the neuropore region of ct/ct embryos imposes a mechanical stress, which opposes neurulation and thus delays closure of the PNP. Increased ventral curvature may arise as a result of a cell proliferation imbalance, which we demonstrated previously in affected ct/ct embryos.     

Vax2 inactivation in mouse determines alteration of the eye dorsal-ventral axis, misrouting of the optic fibres and eye coloboma

Vax2 is a homeobox gene whose expression is confined to the ventral region of the prospective neural retina. Overexpression of this gene at early stages of development in Xenopus and in chicken embryos determines a ventralisation of the retina, thus suggesting its role in the molecular pathway that underlies eye development. We describe the generation and characterisation of a mouse with a targeted null mutation of the Vax2 gene. Vax2 homozygous mutant mice display incomplete closure of the optic fissure that leads to eye coloboma. This phenotype is not fully penetrant, suggesting that additional factors contribute to its generation. Vax2 inactivation determines dorsalisation of the expression of mid-late (Ephb2 and Efnb2) but not early (Pax2 and Tbx5) markers of dorsal-ventral polarity in the developing retina. Finally, Vax2 mutant mice exhibit abnormal projections of ventral retinal ganglion cells. In particular, we observed the almost complete absence of ipsilaterally projecting retinal ganglion cells axons in the optic chiasm and alteration of the retinocollicular projections. All these findings indicate that Vax2 is required for the proper closure of the optic fissure, for the establishment of a physiological asymmetry on the dorsal-ventral axis of the eye and for the formation of appropriate retinocollicular connections.     

The epithelial-mesenchymal transition of the Drosophila mesoderm requires the Rho GTP exchange factor Pebble

Drosophila pebble (pbl) encodes a Rho-family GTP exchange factor (GEF) required for cytokinesis. The accumulation of high levels of PBL protein during interphase and the developmentally regulated expression of pbl in mesodermal tissues suggested that the primary cytokinetic mutant phenotype might be masking other roles. Using various muscle differentiation markers, we found that Even skipped (EVE) expression in the dorsal mesoderm is greatly reduced in pbl mutant embryos. EVE expression in the dorsalmost mesodermal cells is induced in response to DPP secreted by the dorsal epidermal cells. Further analysis revealed that this phenotype is likely to be a consequence of an earlier defect. pbl mutant mesodermal cells fail to undergo the normal epithelial-mesenchymal transition (EMT) and dorsal migration that follows ventral furrow formation. This phenotype is not a secondary consequence of failed cytokinesis, as it is rescued by a mutant form of pbl that does not rescue the cytokinetic defect. In wild-type embryos, newly invaginated cells at the lateral edges of the mesoderm extend numerous protrusions. In pbl mutant embryos, however, cells appear more tightly adhered to their neighbours and extend very few protrusions. Consistent with the dependence of the mesoderm EMT and cytokinesis on actin organisation, the GTP exchange function of the PBL RhoGEF is required for both processes. By contrast, the N-terminal BRCT domains of PBL are required only for the cytokinetic function of PBL. These studies reveal that a novel PBL-mediated intracellular signalling pathway operates in mesodermal cells during the transition from an epithelial to migratory mesenchymal morphology during gastrulation.     

Abnormalities of neural tube formation in pre-spina bifida splotch-delayed mouse embryos

The splotch-delayed homozygous mutant (Spd/Spd) develops spina bifida with or without exencephaly, has spinal ganglia abnormalities, and delays in posterior neuropore closure and neural crest cell emigration. The heterozygote (Spd/+) has a pigmentation defect, and occasionally neural tube defects. To investigate the underlying mechanisms, we compared the neuroepithelium in the posterior neuropore region of cytogenetically identified 15-18 somite pair Spd/Spd, Spd/+, and +/+ mouse embryos by transmission electron and light microscopy. The notochordal area and cell number in the non-fused neuroepithelium region of Spd/Spd and Spd/+ embryos were significantly reduced compared to those of normal (+/+) embryos, which suggests an abnormality in notochord elongation. In the mesoderm, the mean cell number and mean ratio of cell number to area in the non-fused region were significantly lower in the Spd/Spd compared with +/+ embryos. The distance of exposed neuroepithelium above the mesoderm in the just-fused region was significantly lower in the Spd/Spd versus +/+ embryos, which may indicate an insufficient force exerted by the mesoderm during neural tube closure. Within the neuroepithelium, significantly more intercellular space was found in Spd/Spd than in +/+ embryos indicating disorganization. The basal lamina was poorly organized and the formation delayed around the neural tube in Spd/Spd and Spd/+ embryos. All together, these results suggest an early abnormality in interactions among the neuroepithelium, mesoderm, and notochord, which may lead to the delay or inhibition of neural tube closure observed in Spd/Spd mutants.     

Retinal morphogenesis in Drosophila: Hints from an eye-specific decapentaplegic allele

decapentaplegic (dpp) regulates many aspects of imaginal disc growth and patterning in Drosophila. We have analyzed the phenotype of an eye-specific dpp allele, dpp^blk, which causes a reduction in the size of the retina due to a loss of ventral ommatidia. Prior to the onset of differentiation, dpp^blk eye discs are normal regarding size, shape, and ability to express dorsal and ventral markers. However, expression of a dpp-lacZ reporter is reduced at the ventral margin. Additional dorsoventral asymmetry appears during retinal differentiation: the morphogenetic furrow (MF) initiates normally at the posterior tip of the disc, but fails to propagate into the ventral epithelium. This defect can be rescued by increasing dpp expression along the ventral margin by local removal of patched function. We propose that the primary defect in dpp^blk is an inability to activate dpp expression properly at the ventral margin. This has two consequences: it prevents initiation from the ventral margin, and it renders the ventral epithelium unresponsive to differentiation signals emanating from the MF. Dev. Genet. 20:197–207, 1997. © 1997 Wiley-Liss, Inc.     

Drosophila ventral furrow morphogenesis: a proteomic analysis

Ventral furrow formation is a key morphogenetic event during Drosophila gastrulation that leads to the internalization of mesodermal precursors. While genetic analysis has revealed the genes involved in the specification of ventral furrow cells, few of the structural proteins that act as mediators of ventral cell behavior have been identified. A comparative proteomics approach employing difference gel electrophoresis was used to identify more than fifty proteins with altered abundance levels or isoform changes in ventralized versus lateralized embryos. Curiously, the majority of protein differences between these embryos appeared well before gastrulation, only a few protein changes coincided with gastrulation, suggesting that the ventral cells are primed for cell shape change. Three proteasome subunits were found to differ between ventralized and lateralized embryos. RNAi knockdown of these proteasome subunits and time-dependent difference-proteins caused ventral furrow defects, validating the role of these proteins in ventral furrow morphogenesis.     

EphB receptor tyrosine kinases control morphological development of the ventral midbrain

EphB receptor tyrosine kinases and ephrin-B ligands regulate several types of cell-cell interactions during brain development, generally by modulating the cytoskeleton. EphB/ephrinB genes are expressed in the developing neural tube of early mouse embryos with distinct overlapping expression in the ventral midbrain. To test EphB function in midbrain development, mouse embryos compound homozygous for mutations in the EphB2 and EphB3 receptor genes were examined for early brain phenotypes. These mutants displayed a morphological defect in the ventral midbrain, specifically an expanded ventral midline evident by embryonic day E9.5-10.5, which formed an abnormal protrusion into the cephalic flexure. The affected area was comprised of cells that normally express EphB2 and ephrin-B3. A truncated EphB2 receptor caused a more severe phenotype than a null mutation, implying a dominant negative effect through interference with EphB forward (intracellular) signaling. In mutant embryos, the overall number, size, and identity of the ventral midbrain cells were unaltered. Therefore, the defect in ventral midline morphology in the EphB2;EphB3 compound mutant embryos appears to be caused by cellular changes that thin the tissue, forcing a protrusion of the ventral midline into the cephalic space. Our data suggests a role for EphB signaling in morphological organization of specific regions of the developing neural tube.     

The temporal requirement for vitamin A in the developing eye: mechanism of action in optic fissure closure and new roles for the vitamin in regulating cell proliferation and adhesion in the embryonic retina

Mammalian eye development requires vitamin A (retinol, ROL). The role of vitamin A at specific times during eye development was studied in rat fetuses made vitamin A deficient (VAD) after embryonic day (E) 10.5 (late VAD). The optic fissure does not close in late VAD embryos, and severe folding and collapse of the retina is observed at E18.5. Pitx2, a gene required for normal optic fissure closure, is dramatically downregulated in the periocular mesenchyme in late VAD embryos, and dissolution of the basal lamina does not occur at the optic fissure margin. The addition of ROL to late VAD embryos by E12.5 restores Pitx2 expression, supports dissolution of the basal lamina, and prevents coloboma, whereas supplementation at E13.5 does not. Surprisingly, ROL given as late as E13.5 completely prevents folding of the retina despite the presence of an open fetal fissure, showing that coloboma and retinal folding represent distinct VAD-dependent defects. Retinal folding due to VAD is preceded by an overall reduction in the percentage of cyclin D1 positive cells in the developing retina, (initially resulting in retinal thinning), as well as a dramatic reduction in the cell adhesion-related molecules, N-cadherin and β-catenin. Reduction of retinal cell number combined with a loss of the normal cell-cell adhesion proteins may contribute to the collapse and folding of the retina that occurs in late VAD fetuses.     

Prevention of spinal neural tube defects in the mouse embryo by growth retardation during neurulation

Homozygous mutant curly tail mouse embryos developing spinal neural tube defects (NTD) exhibit a cell-type-specific abnormality of cell proliferation that affects the gut endoderm and notochord but not the neuroepithelium. We suggested that spinal NTD in these embryos may result from the imbalance of cell proliferation rates between affected and unaffected cell types. In order to test this hypothesis, curly tail embryos were subjected to influences that retard growth in vivo and in vitro. The expectation was that growth of unaffected rapidly growing cell types would be reduced to a greater extent than affected slowly growing cell types, thus counteracting the genetically determined imbalance of cell proliferation rates and leading to normalization of spinal neurulation. Food deprivation of pregnant females for 48 h prior to the stage of posterior neuropore closure reduced the overall incidence of spinal NTD and almost completely prevented open spina bifida, the most severe form of spinal NTD in curly tail mice. Analysis of embryos earlier in gestation showed that growth retardation acts by reducing the incidence of delayed neuropore closure. Culture of embryos at 40.5 degrees C for 15-23 h from day 10 of gestation, like food deprivation in vivo, also produced growth retardation and led to normalization of posterior neuropore closure. Labelling of embryos in vitro with [3H]thymidine for 1 h at the end of the culture period showed that the labelling index is reduced to a greater extent in the neuroepithelium than in other cell types in growth-retarded embryos compared with controls cultured at 38 degrees C.(ABSTRACT TRUNCATED AT 250 WORDS)     

Phactr4 regulates neural tube and optic fissure closure by controlling PP1-, Rb-, and E2F1-regulated cell-cycle progression

Here we identify the humpty dumpty (humdy) mouse mutant with failure to close the neural tube and optic fissure, causing exencephaly and retinal coloboma, common birth defects. The humdy mutation disrupts Phactr4, an uncharacterized protein phosphatase 1 (PP1) and actin regulator family member, and the missense mutation specifically disrupts binding to PP1. Phactr4 is initially expressed in the ventral cranial neural tube, a region of regulated proliferation, and after neural closure throughout the dorsoventral axis. humdy embryos display elevated proliferation and abnormally phosphorylated, inactive PP1, resulting in Rb hyperphosphorylation, derepression of E2F targets, and abnormal cell-cycle progression. Exencephaly, coloboma, and abnormal proliferation in humdy embryos are rescued by loss of E2f1, demonstrating the cell cycle is the key target controlled by Phactr4. Thus, Phactr4 is critical for the spatially and temporally regulated transition in proliferation through differential regulation of PP1 and the cell cycle during neurulation and eye development.     

Building proteomic pathways using Drosophila ventral furrow formation as a model

Ventral furrow formation is the first morphogenetic movement to occur during Drosophila gastrulation causing the internalization of mesodermal precursors. A previous proteomic screen for ventral-specific proteome changes identified a set of about forty "difference-proteins" that spanned many cellular functions. To understand the connections between these disparate proteins, we initiated a pathway-building scheme using cycles of protein expression manipulation and proteome analysis. This pathway-building exercise started with the proteasomal subunit, Pros35, one of three proteasome subunits found to be ventral-specific difference-proteins. Here we show that Pros35 is a key regulator in ventral furrow formation. Altering the level of Pros35 led to ventral furrow defects. Proteome analysis of the changes induced by Pros35 RNAi showed extensive overlap with the original set of ventral-specific difference-proteins. One of the most prominent changes was in the extracellular iron carrier, Transferrin (Tsf1). Tsf1 is normally less abundant in ventral cells relative to lateral cells; however, RNAi of Pros35 in ventralized embryos negated this ventral-specific difference. Increasing Tsf1 in wild-type embryos blocked ventral furrow formation and caused proteome changes that were similar to the previously seen ventral-specific difference-proteins, including Pros35, which indicates the existence of an unprecedented regulatory loop between the proteasome and iron homeostasis. Additionally, we show that the iron regulatory protein, Irp-1A, also plays an important role in ventral furrow formation. Together these three proteins are part of a regulatory loop that coordinately controls a large number of ventral-specific protein changes.     

Sequence of the twist gene and nuclear localization of its protein in endomesodermal cells of early Drosophila embryos.

The twist gene is involved in the establishment of germ layers in Drosophila embryos: twist homozygous mutant embryos fail to form the ventral furrow at gastrulation and lack mesoderm and all internal organs. We have determined the sequence of the twist gene, that contains 'CAX' repeats in its 5' moiety, and codes for a protein of 490 amino acids. We have raised anti-twist antibodies that were used to study the distribution of the twist protein in whole mounts and tissue sections of wild-type embryos. Twist protein appears to be a nuclear protein at all developmental stages. It is present over both poles and in the midventral region (endoderm and mesoderm anlagen) at cellular blastoderm stage; later in development, it is detected within the mesodermal layer until its differentiation into somatopleura and splanchnopleura in which some cells are still labelled by anti-twist antibodies.     

Spinal ganglia reduction in the splotch-delayed mouse neural tube defect mutant

Splotch and splotch-delayed mutants have anomalies in certain neural crest cell derivatives as well as neural tube defects. A genetic marker was used to identify mutant, heterozygote, and wild-type embryos within a litter, which enabled us to make intergenotypic comparisons. Histological studies of the lumbosacral region of day 15 and day 16 embryos indicated that the splotch-delayed mutant had similar but less severe defects in spinal ganglion development than those reported for splotch (Auerbach: Journal of Experimental Zoology 127:305-329, 1954). The ganglia were extensively reduced in size, residual, or missing in the splotch-delayed mutant, whereas in the splotch mutant, they were virtually nonexistent. Paired comparison analyses showed that all mutant embryos had a significant reduction in their volume of lumbosacral spinal ganglia when compared to their heterozygous and/or wild-type littermates. Also, some heterozygotes were found to have spinal ganglia volumes that were significantly reduced when compared to wild-type embryos. The volume of spinal ganglia was not related to the severity of the neural tube defect. In fact, three mutant embryos, which did not exhibit a neural tube defect, had spinal ganglia volumes comparable to or less than those mutants with open neural tube lesions or curly tails. This shows that the formation of abnormal neural crest cell derivatives is not a result of the neural tube closure defect. We hypothesize that the two anomalies observed in these mutants have a common etiological basis.     

Cytological and biochemical analyses of the maternal-effect mutant embryos with abnormal cleavage furrow formation in Xenopus laevis

We describe an embryonic lethal mutation in Xenopus laevis that provokes regression of cleavage furrow formation. The mutant females (designated as af) were obtained by the back-cross of a female with one of her sons. All the fertilized eggs laid by the mutant females, regardless of the wild-type male used in the mating, failed to cleave although each furrow ran at a proper position superficially. Light and electron microscopic observations of the embryos revealed that the cleavage furrows stayed on the surface and cytoplasmic divisions did not take place at all, while nuclear divisions did. Two-dimensional gel-electrophoretic comparisons of af and wild-type embryos demonstrated that two proteins, having estimated molecular masses of about 38 kDa (pI 6.6) and 78 kDa (pI 7.6), were missing in af embryos. Microinjection of clear cytoplasm from a wild-type egg into fertilized af eggs provoked partial surface contraction and cleavage furrow formation in recipient af eggs. The results showed that the af females carry a lethal maternal-effect mutation which causes cleavage furrow regression by being deficient in a few proteins, and that cytoplasm of wild-type eggs can partially rescue the cleavage furrow formation of af eggs by furnishing the corrective material, presumably a product of the normal allele of af.     

Nuf, a Rab11 effector, maintains cytokinetic furrow integrity by promoting local actin polymerization

Plasma membrane ingression during cytokinesis involves both actin remodeling and vesicle-mediated membrane addition. Vesicle-based membrane delivery from the recycling endosome (RE) has an essential but ill-defined involvement in cytokinesis. In the Drosophila melanogaster early embryo, Nuf (Nuclear fallout), a Rab11 effector which is essential for RE function, is required for F-actin and membrane integrity during furrow ingression. We find that in nuf mutant embryos, an initial loss of F-actin at the furrow is followed by loss of the associated furrow membrane. Wild-type embryos treated with Latrunculin A or Rho inhibitor display similar defects. Drug- or Rho-GTP-induced increase of actin polymerization or genetically mediated decrease of actin depolymerization suppresses the nuf mutant F-actin and membrane defects. We also find that RhoGEF2 does not properly localize at the furrow in nuf mutant embryos and that RhoGEF2-Rho1 pathway components show strong specific genetic interactions with Nuf. We propose a model in which RE-derived vesicles promote furrow integrity by regulating the rate of actin polymerization through the RhoGEF2-Rho1 pathway.     

The Maternal-Effect Gene cellular island Encodes Aurora B Kinase and Is Essential for Furrow Formation in the Early Zebrafish Embryo

Females homozygous for a mutation in cellular island (cei) produce embryos with defects in cytokinesis during early development. Analysis of the cytoskeletal events associated with furrow formation reveal that these defects include a general delay in furrow initiation as well as a complete failure to form furrow-associated structures in distal regions of the blastodisc. A linkage mapping-based candidate gene approach, including transgenic rescue, shows that cei encodes the zebrafish Aurora B kinase homologue. Genetic complementation analysis between the cei mutation and aurB zygotic lethal mutations corroborate gene assignment and reveal a complex nature of the maternal-effect cei allele, which appears to preferentially affect a function important for cytokinesis in the early blastomeres. Surprisingly, in cei mutant embryos a short yet otherwise normal furrow forms in the center of the blastodisc. Furrow formation is absent throughout the width of the blastodisc in cei mutant embryos additionally mutant for futile cycle, which lack a spindle apparatus, showing that the residual furrow signal present in cei mutants is derived from the mitotic spindle. Our analysis suggests that partially redundant signals derived from the spindle and astral apparatus mediate furrow formation in medial and distal regions of the early embryonic blastomeres, respectively, possibly as a spatial specialization to achieve furrow formation in these large cells. In addition, our data also suggest a role for Cei/AurB function in the reorganization of the furrow-associated microtubules in both early cleavage- and somite-stage embryos. In accordance with the requirement for cei/aurB in furrow induction in the early cleavage embryo, germ plasm recruitment to the forming furrow is also affected in embryos lacking normal cei/aurB function.     

Reassessing the role and dynamics of nonmuscle myosin II during furrow formation in early Drosophila embryos

The early Drosophila embryo undergoes two distinct membrane invagination events believed to be mechanistically related to cytokinesis: metaphase furrow formation and cellularization. Both involve actin cytoskeleton rearrangements, and both have myosin II at or near the forming furrow. Actin and myosin are thought to provide the force driving membrane invagination; however, membrane addition is also important. We have examined the role of myosin during these events in living embryos, with a fully functional myosin regulatory light-chain-GFP chimera. We find that furrow invagination during metaphase and cellularization occurs even when myosin activity has been experimentally perturbed. In contrast, the basal closure of the cellularization furrows and the first cytokinesis after cellularization are highly dependent on myosin. Strikingly, when ingression of the cellularization furrow is experimentally inhibited by colchicine treatment, basal closure still occurs at the appropriate time, suggesting that it is regulated independently of earlier cellularization events. We have also identified a previously unrecognized reservoir of particulate myosin that is recruited basally into the invaginating furrow in a microfilament-independent and microtubule-dependent manner. We suggest that cellularization can be divided into two distinct processes: furrow ingression, driven by microtubule mediated vesicle delivery, and basal closure, which is mediated by actin/myosin based constriction.