Polarized cell movements shape the major features of the vertebrate body plan during development. The head-to-tail body axis of vertebrates is elongated in embryonic stages by "convergent extension" tissue movements. During these movements cells intercalate between one another transverse to the elongating body axis to form a narrower, longer array. Recent discoveries show that these polarized cell movements are controlled by homologs of genes that control the polarity of epithelial cells in the developing wing and eye of the fruit fly, Drosophila.

Epithelial-mesenchymal transitions (EMTs) are an important mechanism for reorganizing germ layers and tissues during embryonic development. They have both a morphogenic function in shaping the embryo and a patterning function in bringing about new juxtapositions of tissues, which allow further inductive patterning events to occur [Genesis 28 (2000) 23]. Whereas the mechanics of EMT in cultured cells is relatively well understood [reviewed in Biochem. Pharmacol. 60 (2000) 1091; Cell 105 (2001) 425; Bioessays 23 (2001) 912], surprisingly little is known about EMTs during embryonic development [reviewed in Acta Anat. 154 (1995) 8], and nowhere is the entire process well characterized within a single species. Embryonic (developmental) EMTs have properties that are not seen or are not obvious in culture systems or cancer cells. Developmental EMTs are part of a specific differentiative path and occur at a particular time and place. In some types of embryos, a relatively intact epithelium must be maintained while some of its cells de-epithelialize during EMT. In most cases de-epithelialization (loss of apical junctions) must occur in an orderly, patterned fashion in order that the proper morphogenesis results. Interestingly, we find that de-epithelialization is not always necessarily tightly coupled to the expression of mesenchymal phenotypes.Developmental EMTs are multi-step processes, though the interdependence and obligate order of the steps is not clear. The particulars of the process vary between tissues, species, and specific embryonic context. We will focus on 'primary' developmental EMTs, which are those occurring in the initial epiblast or embryonic epithelium. 'Secondary' developmental EMT events are those occurring in epithelial tissues that have reassembled within the embryo from mesenchymal cells. We will review and compare a number of primary EMT events from across the metazoans, and point out some of the many open questions that remain in this field.

The cells of many embryonic tissues actively narrow in one dimension (convergence) and lengthen in the perpendicular dimension (extension). Convergence and extension are ubiquitous and important tissue movements in metazoan morphogenesis. In vertebrates, the dorsal axial and paraxial mesodermal tissues, the notochordal and somitic mesoderm, converge and extend. In amphibians as well as a number of other organisms where these movements appear, they occur by mediolateral cell intercalation, the rearrangement of cells along the mediolateral axis to produce an array that is narrower in this axis and longer in the anteroposterior axis. In amphibians, mesodermal cell intercalation is driven by bipolar, mediolaterally directed protrusive activity, which appears to exert traction on adjacent cells and pulls the cells between one another. In addition, the notochordal^somitic boundary functions in convergence and extension bỳ capturing' notochordal cells as they contact the boundary, thus elongating the boundary. The prospective neural tissue also actively converges and extends parallel with the mesoderm. In contrast to the mesoderm, cell intercalation in the neural plate normally occurs by monopolar protrusive activity directed medially, towards the midline notoplate^£oor-plate region. In contrast, the notoplate^£oor-plate region appears to converge and extend by adhering to and being towed by or perhaps migrating on the underlying notochord. Converging and extending mesoderm sti¡ens by a factor of three or four and exerts up to 0.6 m N force. Therefore, active, force-producing convergent extension, the mechanism of cell intercalation, requires a mechanism to actively pull cells between one another while maintaining a tissue sti¡ness su¤cient to push with a substantial force. Based on the evidence thus far, a cell^cell traction model of intercalation is described. The essential elements of such a morphogenic machine appear to be (i) bipolar, mediolaterally orientated or monopolar, medially directed protrusive activity; (ii) this protrusive activity results in mediolaterally orientated or medially directed traction of cells on one another; (iii) tractive protrusions are con¢ned to the ends of the cells; (iv) a mechanically stable cell cortex over the bulk of the cell body which serves as a movable substratum for the orientated or directed cell traction. The implications of this model for cell adhesion, regulation of cell motility and cell polarity, and cell and tissue biomechanics are discussed.

No abstract

During gastrulation, a single epithelial cell layer, the ectoderm, generates two others: the mesoderm and the endoderm. In amniotes (birds and mammals), mesendoderm formation occurs through an axial midline structure, the primitive streak, the formation of which is preceded by massive 'polonaise' movements of ectoderm cells. The mechanisms controlling these processes are unknown. Here, using multi-photon time-lapse microscopy of chick (Gallus gallus) embryos, we reveal a medio-lateral cell intercalation confined to the ectodermal subdomain where the streak will later form. This intercalation event differs from the convergent extension movements of the mesoderm described in fish and amphibians (anamniotes): it occurs before gastrulation and within a tight columnar epithelium. Fibroblast growth factor from the extraembryonic endoderm (hypoblast, a cell layer unique to amniotes) directs the expression of Wnt planar-cell-polarity pathway components to the intercalation domain. Disruption of this Wnt pathway causes the mesendoderm to form peripherally, as in anamniotes. We propose that the amniote primitive streak evolved from the ancestral blastopore by acquisition of an additional medio-lateral intercalation event, preceding gastrulation and acting independently of mesendoderm formation to position the primitive streak at the midline.

We describe mesendoderm morphogenesis during gastrulation in the frog Xenopus laevis and investigate the mechanics of these movements with tissue explants. When a dorsal marginal zone explant is plated onto fibronectin, the mesendoderm moves away from the dorsal axial tissues as an intact sheet. Mesendodermal cells within these explants display monopolar protrusive activity and radially intercalate during explant extension. Live time-lapse confocal sequences of actin dynamics at the margin of these extending explants prompt us to propose that integrin-mediated traction drives these movements. We demonstrate that integrin alpha(5)beta(1) recognition of the synergy site located within the type III(9) repeat of fibronectin is required for mesendoderm extension. Normal mesendoderm morphogenesis occurs with a unique "cup-shaped" geometry of the extending mesendodermal mantle and coincides with a higher rate of tissue extension than that seen in the simpler dorsal marginal zone explant. These higher rates can be reconstituted with "in-the-round" configurations of several explants. We propose several mechanically based hypotheses to explain both the initial fibronectin-dependent extension of the mesendoderm and additional requirement of tissue geometry during the high-velocity closure of the mesendodermal mantle.

Force-producing convergence (narrowing) and extension (lengthening) of tissues by active intercalation of cells along the axis of convergence play a major role in axial morphogenesis during embryo development in both vertebrates and invertebrates, and failure of these processes in human embryos leads to defects including spina bifida and anencephaly. Here we use Xenopus laevis, a system in which the polarized cell motility that drives this active cell intercalation has been related to the development of forces that close the blastopore and elongate the body axis, to examine the role of myosin IIB in convergence and extension. We find that myosin IIB is localized in the cortex of intercalating cells, and show by morpholino knockdown that this myosin isoform is essential for the maintenance of a stereotypical, cortical actin cytoskeleton as visualized with time-lapse fluorescent confocal microscopy. We show that this actin network consists of foci or nodes connected by cables and is polarized relative to the embryonic axis, preferentially cyclically shortening and lengthening parallel to the axis of cell polarization, elongation and intercalation, and also parallel to the axis of convergence forces during gastrulation. Depletion of MHC-B results in disruption of this polarized cytoskeleton, loss of the polarized protrusive activity characteristic of intercalating cells, eventual loss of cell-cell and cell-matrix adhesion, and dose-dependent failure of blastopore closure, arguably because of failure to develop convergence forces parallel to the myosin IIB-dependent dynamics of the actin cytoskeleton. These findings bridge the gap between a molecular-scale motor protein and tissue-scale embryonic morphogenesis.

These findings support a role for integrin signaling in regulating the protrusive activity that drives axial extension. We hypothesize that the planar spatial arrangement of the fibrillar fibronectin matrix, which delineates tissue compartments within the embryo, is critical for promoting productive oriented protrusions in intercalating cells.