- Email: [email protected]
An Alternative Model for Cell Sheet Migration on Fibronectin during Heart Formation
J. theor. Biol. (1996) 179, 33–39
An Alternative Model for Cell Sheet Migration on Fibronectin during Heart Formation D J. W† Department of Biology, University of Northern Iowa, Cedar Falls, IA 50614, U.S.A. (Received on 17 February 1995; Accepted in revised form on 9 October 1995)
The emergence of animal form and function depends on cell migrations in the embryo. Some migrations are accomplished by cells individually, and the mechanism of movement is predictable by contemporary models of cell adhesion and cytoskeletal function. However, other migrations occur that involve layers or sheets of cells connected by junctions, and the mechanism of migration is obscure. An example is the precardiac mesoderm, an epithelium that migrates anteriorly and ventrally in the early amniote embryo to the position of heart formation. It moves upon and is influenced by the adjacent endoderm, which has produced an extracellular matrix. The matrix contains the cell adhesion and cytoskeleton-activating glycoprotein fibronectin. Some immunolocalization studies have reported that fibronectin is arrayed in an anterior-to-posterior gradient, and it has been suggested that directional migration results from a haptotactic response of each cell to the gradient, a model derived from and supported by experiments with individual cells in culture. However, we have produced evidence from immunostaining that suggests fibronectin is arrayed as a localized anterior patch rather than a gradient. We propose an alternative model for precardiac epithelial migration in which only the anterior cells attach effectively to fibronectin. Thus adhered, their cytoskeletal contractile activity generates force which propagates throughout the layer of connected cells, and efficiently pulls them in the proper direction, following the bending and extending movements of the foregut, notochord and other structures of the head. Theoretical implications of the two models are discussed. 7 1996 Academic Press Limited
The advance of a layer of interconnected cells would not necessarily be the sum of the advances of the individual cells that make it up. A layer might more efficiently operate as a collective unit with different behaviors. The purpose of this paper is to examine and compare the theoretical implications of epithelial cell migration as derived from individual cell or collective cell action, using the precardiac mesoderm as an illustration.
Introduction The mechanism by which cell layers advance in embryos, toward a position of organ formation or consequent differentiation, is unknown. The cytoskeleton within embryonic cells is instrumental in providing motive force for cell shape changes and cell migrations, and extracellular adhesion molecules provide a substratum for motility. Indeed, it was demonstrated almost 30 years ago by Carter (1967), that an adhesive substratum arrayed in a gradient can provide not only adhesiveness and stimulus to cytoskeletal activity for crawling cells, but also an orientation for direction of migration. However, this paradigm is well-developed for migration of individual cells, not for joined cells within epithelial layers.
Cell Migration and Morphogenesis The initiation of morphogenesis in vertebrate embryos occurs at gastrulation, the early process of germ layer formation that yields the ectoderm, mesoderm, and endoderm. From these layers organ rudiments will form; for example, the heart forms very early in development from paired portions of the
† E-mail: wiens.cobra.uni.edu 0022–5193/96/050033 + 07 $18.00/0
33
7 1996 Academic Press Limited
34
. .
middle germ layer, the mesoderm. Both gastrulation and ensuing organ formation involve the migration and profound rearrangement of embryonic cells, bringing them and their secreted products into new relationships with other cells. Successful cell differentiation often depends on this migration and rearrangement. Thus, the generation of the form and function of the animal body ultimately relies on embryonic cell migrations. In the migration of germ cells, trunk neural crest cells, myoblasts, pre-neural and pre-glial cells, movement occurs individually or in small clusters. The cells have broken away from an epithelial layer or other coherent tissue. The mechanism by which they move, though not comprehensively understood, is at least predictable by cogent models of cytoskeleton-based cell motility. The direction of their migration is similarly predictable by the models. Electrical fields, mechanical tension, the character of physical terrain, diffusible chemoattractants, hormones, growth factors, and extracellular matrix components can all act through the surface of cells to activate signalling mechanisms that typically involve protein kinases and second messengers (Barritt, 1992). These, in turn, alter localized cytoplasmic calcium concentrations, or affect cytoskeletal proteins and their associates directly. Both actions regulate the function of the cytoskeleton, and can bring about directional cell migration. Convergent extension, a cell behavior involved in the formation of the notochord, is an example of a morphogenetic process resulting from the activities of individually migrating cells that achieve tissue elongation by a coordinated action of crawling over one another (Shih & Keller, 1992). But in other cases, the cells migrate or rearrange collectively. This may take the form of the bending of a layer as in invaginations, neurulation, or cleft formation during branching morphogenesis. Wellsupported models exist to explain how such bending might be brought about through coordinated cell shape changes from cuboid to wedged (Burnside, 1971; Spooner, 1974; Jacobson et al., 1986). Collective cell migration may also take the form of spreading of a cell layer. The tendency of isolated, cultured amphibian ectoderm to spread by epiboly was described in the pioneering work of Spemann (1931) and Holtfreter (1933, 1934) as a flattening of cube-shaped surface cells into a thin layer accompanied by migration of deeper ectoderm cells into the spreading surface layer (Holtfreter, 1943) or into the layer just under the surface layer (Keller, 1980) by an intercalation process (radial intercalation) in the deep layer. The direction of the spreading of the
ectoderm is simply outward with no intrinsic polarity, and is determined by the geometric position in the occupied embryo (Holtfreter, 1943). Theoretically, the direction of spreading of any epibolizing layer may also be an outcome of the withdrawal of migrating cells from contact with other cells (contact inhibition), as first described by Abercrombie (1970). There is also substantial evidence to support the hypothesis that cell layers move during gastrulation as liquid-like masses. The subunits (cells) move over one another to relax shear stresses, and the movements are governed by surface tension forces (Holtfreter, 1943; Phillips & Davis, 1978; Davis, 1984) and by differential sorting and adhesion of the cell types involved based on thermodynamic principles (Townes & Holtfreter, 1955; Steinberg, 1963, 1975; Steinberg & Takeichi, 1994). The role (if any) of active or passive cell shape changes and the cytoskeleton in these mechanisms is unclear. However, it is noteworthy that some collective cell migrations have also been described as involving pulling actions of a subgroup of cells at a leading edge. The amphibian marginal zone layer converges towards the blastoporal groove, a movement noted by Holtfreter (1944) as involving a ‘‘pull exerted through the layer’’ by the blastoporal cells as they undergo apical surface constriction and convert to flaskshaped cells, decreasing their outer surface area. The amphibian involuting chordamesoderm then spreads anteriorly along the roof of the blastocoel during early gastrulation. In anuran species this is believed to occur through convergent extension of cells in the deep involuting marginal zone (Keller & Schoenwolf, 1977; Keller, 1986; Wilson & Keller, 1991) and is substrate-independent. However cell migration also plays a role in the process, particularly in urodeles (Shi et al., 1987) but also in Xenopus in the anterior-most prospective axial mesoderm where it acts synergistically with convergent extension (Winklbauer & Nagel, 1991). This migration is given directional guidance by a uniform array of fibronectin fibrils (Winklbauer & Nagel, 1991). Precardiac Cell Migration Heart muscle development clearly involves the differentiation of myocytes (muscle cells) from an epithelial sheet of cells, one that is migratory as a collective unit, the precardiac mesoderm. This is in contrast to the development of skeletal muscle, in which myocytes differentiate from free embryonic connective tissue cells (mesenchyme) that have migrated individually from the myotomes of the somites into position. From the time chick embryo
precardiac cells first condense within the primary mesenchyme (about 18 hr of incubation; Hay, 1968, 1990), they are interconnected by junctional complexes (Rosenquist & De Haan, 1966; Manasek, 1968; Stalsberg & De Haan, 1969; Spira, 1971). As gastrulation proceeds at this time, the precardiac mesoderm migrates in an anterior and then medial direction from its paired axial locations flanking the primitive knot (the anterior-most site of cell involution where cells move into the interior cavity to form the mesoderm layer). During the next few hours, as the head with notochord and foregut begin to form, a cleavage between these mesodermal cells occurs in this same region. The cleavage divides the mesoderm into somatic and visceral layers. The cells that will form the heart enter the ventral, visceral side of the mesoderm before the separation occurs (De Haan, 1965), and change from a loosely connected sheet of spindle-shaped cells into a columnar-like epithelium, oriented with cell apices toward the forming coelom and bases directed toward the adjacent endoderm (Drake et al., 1990). Migratory behavior changes dramatically at 21–22 hr of incubation as a strong directional influence of the closely adjacent endoderm comes into effect, causing the precardiac cells to move in smoothly curved, parallel pathways toward the anterior intestinal portal (AIP) (De Haan, 1963a,b, 1964). The Mechanism of Precardiac Cell Migration and the Role of Fibronectin The mechanism by which this cell layer advances anteriorly in the embryo, toward the position of heart tube formation near the AIP (occurring at about 33 hr of incubation in the chick embryo) is unknown. It seems probable that the activities of individual cells provide the motive force, and that an adhesive substratum and directional guidance for the oriented movement of the cells are provided by the adjacent endoderm (De Haan, 1963a,b) with its array of extracellular matrix (ECM) molecules, including fibronectin (FN) and collagen. All precardiac mesodermal cells, though joined, exhibit irregular shapes suggestive of motile activity, with numerous filopodia and processes (Trelstad et al., 1967; England & Wakely, 1977). However, there is evidence from the observation of chick embryo mesodermal explants placed in culture that suggests motile activity is confined to a limited subgroup of these cells. When the explants are cultured on a plastic substratum coated with parallel lines of FN, the cells at the periphery of the mesoderm were actively motile, pulling other cells along passively (Toyoizumi et al.,
35
1991), in a manner similar to the epithelial sheet migration that occurs during wound healing (Lash, 1955; Trinkaus, 1976). Isolated early mesoderm from gastrulating frog embryos also exhibits a spreading, leading anterior margin with a compact, retracting posterior end when cultured on a continuous FN substratum (Winklbauer, 1990). If migration of the precardiac mesoderm is in some way stimulated and guided by the endodermal ECM, then insight concerning the mechanism of directed migration may be gained by analysis of the spatial and temporal organization of the ECM. Among the known components of this ECM, the cell adhesion glycoprotein FN is prominent and interesting because of its known involvement in many cell motility processes, and its binding sites for other ECM molecules such as collagens and proteoglycans, and for cell surface receptors (see review by Hynes, 1985). Several studies that employed antibodies to FN in immunofluorescence staining have indicated its presence between the precardiac mesoderm and adjacent endoderm (Icardo & Manasek, 1983; Duband & Thiery, 1982; Drake et al., 1990), including that of Linask & Lash (1986) which described an anterior-to-posterior gradient of FN staining that appears by the time of migration. On the basis of this pattern of immunofluorescence observed, a haptotactic mechanism (migration directed by binding with a specific peptide within FN acting as the binding site) of precardiac cell migration was suggested in which the cells continually move toward areas of greater adhesion (Linask & Lash, 1986). An illustration of this arrangement is presented in Fig. 1A. These authors also reported an inhibition of heart development when the embryos were incubated with antibodies to FN or to its receptor, or with arginine-glycine-aspartic acid synthetic peptide, which mimics FN’s binding site and blocks its binding to cell surface receptors (Linask & Lash, 1988a). In another experimental study by these authors (Linask & Lash, 1988b), microsurgical rotation of the precardiac mesoderm together with adjacent endoderm 180° resulted in abnormal or ectopic heart development. Control experiments in which the two-layered explants were removed and then replaced in original orientation developed normally. These studies confirm an important role for FN, though they do not necessarily indicate a mechanism of haptotactic migration in response to a FN concentration gradient. Abberrant directional migration and ectopic heart tube formation would also be predicted after rotation if an endodermally produced, anterior, confined patch of fibronectin, residing in and near the head fold, and included in the explant, is responsible
. .
36
for creating a pulling tension through the precardiac mesoderm that drives migration. In sham-operated controls with no rotation, the tension would be regenerated after a short healing time.
Evidence for an Anterior, Confined Patch of Fibronectin We have examined the distribution of FN using sagittal and transverse sections, and whole mounts of embryos, during the time of migration. The complete sagittal profile of FN immunostaining in the precardiac mesoderm offers the most definitive assessment of its anterior-posterior deployment. We employed an immunostaining method that allows visualization of FN in eosin-hematoxylin stained tissue sections, facilitating the interpretation of exact localization. We used a highly specific monoclonal antibody to FN, and a specific antiserum to chicken type IV collagen and another to fibronectin for comparison purposes. Our findings led me to propose a different mechanism of precardiac mesoderm migration. We found (Plate 1) that in the pre-heart embryo, FN is localized with striking intensity in two areas: the prospective proamnion (outside the embryo body) and the forming head fold (Plate 1, A–C). It appears as a concentrated, crescent-shaped region associated with the basement membrane of the anterior germ layers, particularly the endoderm, and as the layers fold to yield a head fold with AIP, the FN-stained region also folds (Plate 1, C). This action results in a ventrally located, foregut endodermal crescent that appears to maintain a ‘‘grip’’ on the adjacent
precardiac mesoderm. The folding movements of the foregut endoderm have long been implicated in cardiac morphogenesis (Bellairs, 1953). As development of the heart proceeds, this area increases to an apparent maximum of staining intensity shortly before the tubular heart is formed (Plate 1, F). Pre-endocardial cells then move into the extracellular space between the two layers and become coated with FN. These cells adhere to each other and quickly form paired endocardial tubes that are covered with and connected by FN (Plate 1, F). Meanwhile, the remaining precardiac mesoderm surrounds the endocardial tubes as they fuse, forming the myocardium. The primitive heart tube takes on definitive form and function (contractions) at about 33 hr of incubation. Type IV collagen was found to be mainly co-localized with FN during precardiac cell migration, but it was not observed to be accentuated in the head fold as fibronectin was (Plate 1, D). This result is in agreement with that of Drake et al. (1990). Our localization of type IV collagen demonstrates an ECM molecule distribution that parallels that of FN in many respects, but also stands in contrast to the striking restricted patch of intense FN staining, which was displayed with both the monoclonal and polyclonal antibodies. The existence of a highly concentrated anterior region of FN within the pre-heart chick embryo has not been previously reported, although the staining that we also observed in the anterior extraembryonic region has been (Critchley et al., 1979). Linask & Lash (1986) presented some evidence for a gradient distribution shown by immunofluorescence photographs in an anterior-to-posterior sequence of
F. 1. Sketch illustrating two alternative models of precardiac mesodermal cell migration along FN-containing basement membrane of adjacent endoderm. A. The precardiac cells all respond to a gradient of FN, utilizing the gradient to determine direction for migration. Each cell moves individually even though the cells are interconnected. B. The precardiac cells depend on adhesion of the anterior-most cells of the layer to a concentrated, crescentic patch of fibronectin. As a result of adhesion they develop tension which is transmitted posteriorly throughout the layer. The cells then advance anteriorly as a layer.
J. theor. Biol.
P 1. Sections through embryos of approximately 22 and 30 hr incubation, showing the monoclonal antibody localization of fibronectin and type IV collagen. A–D are sagittal sections of a stage 6 embryo and anterior is to the left; E and F are transverse sections. (A) Low magnification view of entire embryo’s sagittal profile stained anti-FN. Two patches of intense staining are located at the anterior end in this section taken lateral to the head fold. (B) Higher magnification of the section shown in A. (C) Another sagittal section taken closer to the midline showing the early head fold and the disposition of the FN patches: the more anterior patch is localized in the prospective proamnion whereas the more posterior patch is within the head fold. (D) Parasagittal section equivalent to that shown in B, but stained with antibodies to type IV collagen. The localization is generalized in the anterior-posterior axis. (E) Control cross-section through the AIP of a stage 9 chick embryo, with normal mouse serum replacing antibody. Blue color of the tissues is from counterstaining with hematoxylin. (F) Anti-FN stained transverse section through a stage 9 embryo showing red staining associated with all basement membranes, head mesenchyme, and the fusing endocardial tubes of the primitive heart. Staining is intense between precardiac mesoderm and foregut endoderm, and surrounds a pre-endocardial cell in this area dorsal to the forming heart (arrow). It lines and joins the two tubes prominently. Also note the striking staining of neutral crest cells. hf, head fold; fg, foregut; nc, neural crest; nt, neural tube. (A) Scale bar = 200 mm; (B–F) Scale bar = 100 mm.
D. J. WIENS
facing p. 36
separate regions of a section from a stage 5 and stage 7 embryo. The anti-FN used was a rabbit anti-chick antiserum, and staining procedures were carried out in fixed, methanol-permeabilized whole embryos before sectioning. Drake et al. (1990), however, employing a different antiserum (affinity-purified rabbit antibodies to chicken FN, Little & Chen, 1982) also incubated prior to sectioning, were unable to duplicate this observation, and reported uniform immunostaining across the heart-forming region. However, the fluorescence photomicrographs shown in their report were en face views of the ECM stained for FN, and it was not indicated whether they included the anterior-most regions within the head fold. The anti-FN antibodies used have been reported to display some cross-reactivity with collagen (Little & Chen, 1982). It is possible that some non-specific staining with the antisera used in both these reports masked the higher intensity regions that we observed using monoclonal antibodies, or that greater availability of the antigenic sites and more direct washing in sectioned tissue afforded more resolution in immunostaining, having less dependence on antibody penetration of the tissues. Another possibility is that the monoclonal anti-FN recognizes an antigenic site found in only a subset of the whole population of FN molecules. This raises the interesting possibility that a specific FN subtype(s) serves to guide the precardiac cells. However, our results with the polyclonal antibodies were virtually identical to those with the monoclonal antibody.
An Alternative Model The formation and maintenance of this crescentic localized region of FN at the anterior end suggests that precardiac cell migration could depend mainly on the anterior-most mesodermal cells. Their adhesion would be high and their motility activated, allowing them to establish a contractile force leading to a pulling tension throughout the epithelial layer of more passive, interconnected cells. Although a gradient of FN (and perhaps additional fixed ligands) supporting haptotactic migration toward a maximum is logical for cells that move individually, it would be inefficient for joined cells, each of which should have no requirement to extend protrusions as well as form and break adhesions. Joined cells’ motile activities might be restrained by their attachments to other cells. Joined cells, each operating in response to a gradient to find direction and then migrate, would also require the prior graded production, secretion, and deposition of the adhesion
37
molecule by many other cells (in this case cells of the endoderm), a process that would require careful coordination. The precardiac mesoderm could operate more efficiently as a collective unit with anterior cells only required to form a strong adhesive region (Fig. 1B). The contractile force would be generated by cytoskeletal contractile activity within the anterior cells or throughout all the cells. Resistance to the tension would be provided by weak adhesion of all cells posterior to this area, where there is a continuous FN substratum of low concentration, affording little adhesion. The resistance would be gradually overcome as the anterior adhesion and the contraction increase while posterior adhesions release. The release might even be augmented by cytoskeletal changes induced by mechanoreceptive signalling mechanisms. The overall effect would eventually pull the interconnected precardiac cells anteriorly, then ventrally around the anterior edge of the foregut, and finally posteriorly again to the position where cells join the forming heart. The existence of tension in the pre-heart chick embryo has not been demonstrated, however, we have observed that a cut through the precardiac mesoderm and endoderm results in rapid retraction of the edges (Wiens et al., 1984). The process of migration of the precardiac mesoderm and ensuing tubular heart formation can be visualized as a consequence of the transitional sequence of events that occur at the FN-rich locale. When gastrulation produces an anterior group of mesodermal cells bounded above by ectoderm and below by endoderm, FN is secreted locally. The mesodermal cells attach to it. The binding acts as a cell surface signal which is transduced to begin contractile processes within the cells that create contractile force throughout the interconnected mesodermal layer. The region then bends, becomes part of the head fold, and becomes displaced ventrally and posteriorly as the foregut extends anteriorly. It then fosters pre-endocardial cells emergence and subsequent self-adhesion of endocardial cells (evident in Plate 1, F), and finally, it remains as a dorsal mesocardium through which cells continue to migrate and join the growing heart. Further studies on the adhesive, contractile, and migratory activities of the precardiac cells are needed to test the model, and should help elucidate the mechanism of their forward migration. Tests of migratory ability of intact, intact but subdivided, dissociated, and reassociated embryonic epithelia, cultured upon various arrangements of FN would serve as a starting point, and would offer the opportunity to investigate the effects of agents that can perturb the migrating cells in specific ways.
. .
38
The monoclonal antibody to avian fibronectin (B3D6, Gardner & Fambrough, 1983) was obtained from the Developmental Studies Hybridoma Bank, maintained by the Department of Pharmacology and Molecular Sciences at Johns Hopkins University School of Medicine, Baltimore, MD, and the Department of Biology at the University of Iowa, Iowa City, IA, under contract number NO1-HD-2-3144 from the NICHD. Monoclonal antibody IV-IIB12/C2 to avian type IV collagen (Mayne et al., 1983) was a generous gift from Dr. Thomas F. Linsenmayer, Department of Anatomy and Cell Biology, Tufts University, Boston, MA. I thank Jeffrey Rathmell, Tamara Mann Nowling, Clifton Hall, Scott Briggs, John Reppas, Jennifer Becker, and Melissa Appleget, students at the University of Northern Iowa who assisted with immunostaining; and Drs. Jean Gerrath and Carl Thurman of the Biology Department for critical comments on the early drafts of the manuscript. This work was supported by the American Heart Association, Iowa affiliate grant IA-90-G-40, by Summer Research Fellowships from the graduate college of the University of Northern Iowa, and by JOVE (Joint Venture Program) grant No. NAG9-265 to the University of Northern Iowa, D. Wiens and T. Hockey, Project Directors. Contributions to the work were carried out as undergraduate student research projects in partial fulfillment of the B.S. degree in biology by Jeffrey Rathmell, Clifton Hall, and Scott Briggs.
REFERENCES A, M. (1970). Contact inhibition in tissue culture. In Vitro 6, 128–142. B, G. J. (1992). Communication Within Animal Cells. Oxford: Oxford Science Publications. B, R. (1953). Studies on the development of the foregut in the chick blastoderm. II. The morphogenetic movements. J. Embryol. Exptl. Morphol. 1, 369–385. B, B. (1971). Microtubules and microfilaments in newt neurulation. Dev. Biol. 26, 416–441. C, S. B. (1967). Haptotaxis and the mechanism of cell motility. Nature 213, 256–260. C, D. R., E, M. A. & H, R. O. (1979). Distribution of fibronectin in the ectoderm of gastrulating chick embryos. Nature (London) 280, 498–500. D, G. (1984). Migration-directing liquid properties of embryonic amphibian tissues. Amer. Zool. 24, 649–655. D H, R. L. (1963a). Organization of the cardiogenic plate in the early chick embryo. Acta Embryol. Exptl. 6, 26–38. D H, R. L. (1963b). Migration patterns of the precardiac mesoderm in the early chick embryo. Exptl. Cell Res. 29, 544–560. D H, R. L. (1964). Cell interactions and oriented movements during development. J. Exp. Zool. 157, 127–238. D H, R. L. (1965). Morphogenesis of the vertebrate heart. In: Organogenesis, pp. 377–419. New York: Holt, Rinehart & Winston Co. D, C. J., D, L. A., W, L. & L, C. D. (1990). Avian vasculogenesis and the distribution of collagens, I, IV, laminin, and fibronectin in the heart primordia. J. Exp. Zool. 255, 309–322. D, J. L. & T, J. P. (1982). Appearance and distribution of fibronectin during chick embryo gastrulation and neurulation. Dev. Biol. 94, 337–350. E, M. A. & W, J. (1977). Scanning electron microscopy of the developing mesoderm layer in the chick embryo. Anat. Embryol. 150, 291–300. G, J. M. & F, D. M. (1983). Fibronectin expression during myogenesis. J. Cell Biol. 96, 474–485.
H, E. D. (1968). Organization and fine structure of epithelium and mesenchyme in the developing chick embryo. In: Epithelial-Mesenchymal Interactions (R. Fleischmajer & R. E. Billingham, eds) pp. 31–55. Baltimore: Williams and Wilkins. H, E. D. (1990). Epithelial-mesenchymal transitions. Sem. Dev. Biol. 1, 347–356. H, R. O. (1985). Molecular biology of fibronectin. Annu. Rev. Cell Biol. 1, 67–90. H, J. (1933). Die totale Exogastrulation, eine Selbstablo¨sung des Ekdoderms vom Entomesoderm. Arch. Entw. Mech. 129, 669–793. H, J. (1934). Formative Reize in der Embryonalentwicklung der Amphibien, Dargestellt an Explantationsversuchen. Arch. exp. Zellf. 15, 281–301. H, J. (1943). A study of the mechanisms of gastrulation. Part I. J. Exp. Zool. 94, 261–318. H, J. (1944). A study of the mechanics of gastrulation. Part II. J. Exp. Zool. 95, 171–212. I, J. M. & M, F. J. (1983). Fibronectin distribution during early chick embryo heart development. Dev. Biol. 95, 19–30. J, A. B., O, G. F., O, G. M. & C, L. Y. (1986). Neurulation and the cortical tractor model for epithelial folding. J. Embryol. Morphol. 96, 19–49. K, R. E. & S, G. C. (1977). An SEM study of cellular morphology, contact, and arrangement as related to gastrulation in Xenopus laevis. Wilhelm Roux Arch. Dev. Biol. 182, 165–186. K, R. E. (1980). The cellular basis of epiboly: an SEM study of deep cell rearrangement during gastrulation of Xenopus laevis. J. Embryol. exp. Morphol. 60, 201–243. K, R. E. (1986). The cellular basis of morphogenesis. In: Developmental Biology: A Comprehensive Synthesis (L. Browder, ed.) pp. 241–327, Vol. 2. New York: Plenum. L, J. W. (1955). Studies on wound closure in urodeles. J. Exp. Zool. 128, 13–28. L, K. K. & L, J. W. (1986). Precardiac cell migration: fibronectin localization at mesoderm-endoderm interface during directional movement. Dev. Biol. 114, 87–101. L, K. K. & L, J. W. (1988a). A role for fibronectin in the migration of avian precardiac cells. I. Dose-dependent effects of fibronectin antibody. Dev. Biol. 129, 315–324. L, K. K. & L, J. W. (1988b). A role for fibronectin in the migration of avian precardiac cells. II. Rotation of the heart-forming region during different stages and its effects. Dev. Biol. 129, 324–329. L, C. D. & C, W.-T. (1982). Masking of extracellular collagen and the co-distribution of collagen and fibronectin during matrix formation by cultured embryonic fibroblasts. J. Cell Sci. 55, 35–50. M, F. J. (1968). Embryonic development of the heart. I. A light and electron microscopic study of myocardial development in the early chick embryo. J. Morphol. 125, 329–366. M, R., S, R. D., W, H., F, J. M. & L, T. F. (1983). The use of monoclonal antibodies to fragments of chicken type IV collagen in structural and localization studies. J. Biol. Chem. 258, 5794–5797. P, H. M. & D, G. S. (1978). Liquid-tissue mechanics in amphibian gastrulation: germ-layer assembly in Rana pipiens. Amer. Zool. 18, 81–93. R, G. C. & D H, R. L. (1966). Migration of precardiac cells in the chick embryo: a radioautographic study. Carnegie Inst. Wash. Contributions to Embryology 38, 111–121. S, D.-L., D, M., D, T., R, J.-F. & B, J.-C. (1987). Experimental analysis of the extension of the dorsal marginal zone in Pleurodeles waltl gastrulae. Development 100, 147–161. S, J. & K, R. (1992). Cell motility driving mediolateral intercalation in explants of Xenopus laevis. Development 116, 901–914.
S, H. (1931). Uber den Anteil von Implantat und Wirtskeim an der Orientierung und Beschaffenheit der induzierten Embryonalanlage. Arch. Entw. Mech. 123, 390–517. S, A. W. (1971). Cell junctions and their role in transmural diffusion in the embryonic chick heart. Z. Zellforsch. 120, 463–487. S, B. S. (1974). Morphogenesis of vertebrate organs. In: Concepts of Development (J. Lash & J. R. Whittaker, eds) pp. 213–240, Stamford, CN: Sinauer & Associates, Inc. S, H. & D H, R. L. (1969). The precardiac areas and formation of the tubular heart in the chick embryo. Dev. Biol. 19, 128–159. S, M. S. (1963). Reconstruction of tissues by dissociated cells. Science 141, 401–408. S, M. S. (1975). Adhesion-guided multicellular assembly: a commentary upon the postulates, real and imagined, of the differential adhesion hypothesis, with special attention to computer simulations of cell sorting. J. theor. Biol. 55, 431–443. S, M. S. & T, M. (1994). Experimental specification of cell sorting, tissue spreading, and specific spatial patterning by quantitative differences in cadherin expression. Proc. Nat. Acad. Sci. U.S.A. 91, 206–209. T, P. L. & H, J. (1955). Directed movements and
39
selected adhesion of embryonic amphibian cells. J. Exp. Zool. 128, 53–120. T, R., S, K. & T, S. (1991). The behavior and cytoskeletal system of chick gastrula mesodermal cells on substrata coated with lines of fibronectin. J. Exp. Zool. 260, 345–353. T, R. L., H, E. D. & R, J. P. (1967). Cell contact during early morphogenesis in the chick embryo. Dev. Biol. 16, 78–106. T, J. P. (1976). On the mechanism of Metazoan cell movements. In: Cell Surface Reviews (G. Poste & G. L. Nicholson, eds), Vol. 1, pp. 225–329. Amsterdam: NorthHolland Publishing Co. W, D., S, M. & S, B. S. (1984). Precardiac mesoderm differentiation in vitro: actin-isotype synthetic transitions, myofibrillogenesis, initiation of heartbeat, and the possible involvement of collagen. Differentiation 28, 62–72. W, P. & K, R. (1991). Cell rearrangement during gastrulation of Xenopus: direct observation of cultured explants. Development 112, 289–300. W, R. (1990). Mesoderm cell migration during Xenopus gastrulation. Dev. Biol. 142, 155–168. W, R. & N, M. (1991). Directional mesoderm cell migration in the Xenopus gastrula. Dev. Biol. 148, 573–589.
An Alternative Model for Cell Sheet Migration on Fibronectin during Heart Formation D J. W† Department of Biology, University of Northern Iowa, Cedar Falls, IA 50614, U.S.A. (Received on 17 February 1995; Accepted in revised form on 9 October 1995)
The emergence of animal form and function depends on cell migrations in the embryo. Some migrations are accomplished by cells individually, and the mechanism of movement is predictable by contemporary models of cell adhesion and cytoskeletal function. However, other migrations occur that involve layers or sheets of cells connected by junctions, and the mechanism of migration is obscure. An example is the precardiac mesoderm, an epithelium that migrates anteriorly and ventrally in the early amniote embryo to the position of heart formation. It moves upon and is influenced by the adjacent endoderm, which has produced an extracellular matrix. The matrix contains the cell adhesion and cytoskeleton-activating glycoprotein fibronectin. Some immunolocalization studies have reported that fibronectin is arrayed in an anterior-to-posterior gradient, and it has been suggested that directional migration results from a haptotactic response of each cell to the gradient, a model derived from and supported by experiments with individual cells in culture. However, we have produced evidence from immunostaining that suggests fibronectin is arrayed as a localized anterior patch rather than a gradient. We propose an alternative model for precardiac epithelial migration in which only the anterior cells attach effectively to fibronectin. Thus adhered, their cytoskeletal contractile activity generates force which propagates throughout the layer of connected cells, and efficiently pulls them in the proper direction, following the bending and extending movements of the foregut, notochord and other structures of the head. Theoretical implications of the two models are discussed. 7 1996 Academic Press Limited
The advance of a layer of interconnected cells would not necessarily be the sum of the advances of the individual cells that make it up. A layer might more efficiently operate as a collective unit with different behaviors. The purpose of this paper is to examine and compare the theoretical implications of epithelial cell migration as derived from individual cell or collective cell action, using the precardiac mesoderm as an illustration.
Introduction The mechanism by which cell layers advance in embryos, toward a position of organ formation or consequent differentiation, is unknown. The cytoskeleton within embryonic cells is instrumental in providing motive force for cell shape changes and cell migrations, and extracellular adhesion molecules provide a substratum for motility. Indeed, it was demonstrated almost 30 years ago by Carter (1967), that an adhesive substratum arrayed in a gradient can provide not only adhesiveness and stimulus to cytoskeletal activity for crawling cells, but also an orientation for direction of migration. However, this paradigm is well-developed for migration of individual cells, not for joined cells within epithelial layers.
Cell Migration and Morphogenesis The initiation of morphogenesis in vertebrate embryos occurs at gastrulation, the early process of germ layer formation that yields the ectoderm, mesoderm, and endoderm. From these layers organ rudiments will form; for example, the heart forms very early in development from paired portions of the
† E-mail: wiens.cobra.uni.edu 0022–5193/96/050033 + 07 $18.00/0
33
7 1996 Academic Press Limited
34
. .
middle germ layer, the mesoderm. Both gastrulation and ensuing organ formation involve the migration and profound rearrangement of embryonic cells, bringing them and their secreted products into new relationships with other cells. Successful cell differentiation often depends on this migration and rearrangement. Thus, the generation of the form and function of the animal body ultimately relies on embryonic cell migrations. In the migration of germ cells, trunk neural crest cells, myoblasts, pre-neural and pre-glial cells, movement occurs individually or in small clusters. The cells have broken away from an epithelial layer or other coherent tissue. The mechanism by which they move, though not comprehensively understood, is at least predictable by cogent models of cytoskeleton-based cell motility. The direction of their migration is similarly predictable by the models. Electrical fields, mechanical tension, the character of physical terrain, diffusible chemoattractants, hormones, growth factors, and extracellular matrix components can all act through the surface of cells to activate signalling mechanisms that typically involve protein kinases and second messengers (Barritt, 1992). These, in turn, alter localized cytoplasmic calcium concentrations, or affect cytoskeletal proteins and their associates directly. Both actions regulate the function of the cytoskeleton, and can bring about directional cell migration. Convergent extension, a cell behavior involved in the formation of the notochord, is an example of a morphogenetic process resulting from the activities of individually migrating cells that achieve tissue elongation by a coordinated action of crawling over one another (Shih & Keller, 1992). But in other cases, the cells migrate or rearrange collectively. This may take the form of the bending of a layer as in invaginations, neurulation, or cleft formation during branching morphogenesis. Wellsupported models exist to explain how such bending might be brought about through coordinated cell shape changes from cuboid to wedged (Burnside, 1971; Spooner, 1974; Jacobson et al., 1986). Collective cell migration may also take the form of spreading of a cell layer. The tendency of isolated, cultured amphibian ectoderm to spread by epiboly was described in the pioneering work of Spemann (1931) and Holtfreter (1933, 1934) as a flattening of cube-shaped surface cells into a thin layer accompanied by migration of deeper ectoderm cells into the spreading surface layer (Holtfreter, 1943) or into the layer just under the surface layer (Keller, 1980) by an intercalation process (radial intercalation) in the deep layer. The direction of the spreading of the
ectoderm is simply outward with no intrinsic polarity, and is determined by the geometric position in the occupied embryo (Holtfreter, 1943). Theoretically, the direction of spreading of any epibolizing layer may also be an outcome of the withdrawal of migrating cells from contact with other cells (contact inhibition), as first described by Abercrombie (1970). There is also substantial evidence to support the hypothesis that cell layers move during gastrulation as liquid-like masses. The subunits (cells) move over one another to relax shear stresses, and the movements are governed by surface tension forces (Holtfreter, 1943; Phillips & Davis, 1978; Davis, 1984) and by differential sorting and adhesion of the cell types involved based on thermodynamic principles (Townes & Holtfreter, 1955; Steinberg, 1963, 1975; Steinberg & Takeichi, 1994). The role (if any) of active or passive cell shape changes and the cytoskeleton in these mechanisms is unclear. However, it is noteworthy that some collective cell migrations have also been described as involving pulling actions of a subgroup of cells at a leading edge. The amphibian marginal zone layer converges towards the blastoporal groove, a movement noted by Holtfreter (1944) as involving a ‘‘pull exerted through the layer’’ by the blastoporal cells as they undergo apical surface constriction and convert to flaskshaped cells, decreasing their outer surface area. The amphibian involuting chordamesoderm then spreads anteriorly along the roof of the blastocoel during early gastrulation. In anuran species this is believed to occur through convergent extension of cells in the deep involuting marginal zone (Keller & Schoenwolf, 1977; Keller, 1986; Wilson & Keller, 1991) and is substrate-independent. However cell migration also plays a role in the process, particularly in urodeles (Shi et al., 1987) but also in Xenopus in the anterior-most prospective axial mesoderm where it acts synergistically with convergent extension (Winklbauer & Nagel, 1991). This migration is given directional guidance by a uniform array of fibronectin fibrils (Winklbauer & Nagel, 1991). Precardiac Cell Migration Heart muscle development clearly involves the differentiation of myocytes (muscle cells) from an epithelial sheet of cells, one that is migratory as a collective unit, the precardiac mesoderm. This is in contrast to the development of skeletal muscle, in which myocytes differentiate from free embryonic connective tissue cells (mesenchyme) that have migrated individually from the myotomes of the somites into position. From the time chick embryo
precardiac cells first condense within the primary mesenchyme (about 18 hr of incubation; Hay, 1968, 1990), they are interconnected by junctional complexes (Rosenquist & De Haan, 1966; Manasek, 1968; Stalsberg & De Haan, 1969; Spira, 1971). As gastrulation proceeds at this time, the precardiac mesoderm migrates in an anterior and then medial direction from its paired axial locations flanking the primitive knot (the anterior-most site of cell involution where cells move into the interior cavity to form the mesoderm layer). During the next few hours, as the head with notochord and foregut begin to form, a cleavage between these mesodermal cells occurs in this same region. The cleavage divides the mesoderm into somatic and visceral layers. The cells that will form the heart enter the ventral, visceral side of the mesoderm before the separation occurs (De Haan, 1965), and change from a loosely connected sheet of spindle-shaped cells into a columnar-like epithelium, oriented with cell apices toward the forming coelom and bases directed toward the adjacent endoderm (Drake et al., 1990). Migratory behavior changes dramatically at 21–22 hr of incubation as a strong directional influence of the closely adjacent endoderm comes into effect, causing the precardiac cells to move in smoothly curved, parallel pathways toward the anterior intestinal portal (AIP) (De Haan, 1963a,b, 1964). The Mechanism of Precardiac Cell Migration and the Role of Fibronectin The mechanism by which this cell layer advances anteriorly in the embryo, toward the position of heart tube formation near the AIP (occurring at about 33 hr of incubation in the chick embryo) is unknown. It seems probable that the activities of individual cells provide the motive force, and that an adhesive substratum and directional guidance for the oriented movement of the cells are provided by the adjacent endoderm (De Haan, 1963a,b) with its array of extracellular matrix (ECM) molecules, including fibronectin (FN) and collagen. All precardiac mesodermal cells, though joined, exhibit irregular shapes suggestive of motile activity, with numerous filopodia and processes (Trelstad et al., 1967; England & Wakely, 1977). However, there is evidence from the observation of chick embryo mesodermal explants placed in culture that suggests motile activity is confined to a limited subgroup of these cells. When the explants are cultured on a plastic substratum coated with parallel lines of FN, the cells at the periphery of the mesoderm were actively motile, pulling other cells along passively (Toyoizumi et al.,
35
1991), in a manner similar to the epithelial sheet migration that occurs during wound healing (Lash, 1955; Trinkaus, 1976). Isolated early mesoderm from gastrulating frog embryos also exhibits a spreading, leading anterior margin with a compact, retracting posterior end when cultured on a continuous FN substratum (Winklbauer, 1990). If migration of the precardiac mesoderm is in some way stimulated and guided by the endodermal ECM, then insight concerning the mechanism of directed migration may be gained by analysis of the spatial and temporal organization of the ECM. Among the known components of this ECM, the cell adhesion glycoprotein FN is prominent and interesting because of its known involvement in many cell motility processes, and its binding sites for other ECM molecules such as collagens and proteoglycans, and for cell surface receptors (see review by Hynes, 1985). Several studies that employed antibodies to FN in immunofluorescence staining have indicated its presence between the precardiac mesoderm and adjacent endoderm (Icardo & Manasek, 1983; Duband & Thiery, 1982; Drake et al., 1990), including that of Linask & Lash (1986) which described an anterior-to-posterior gradient of FN staining that appears by the time of migration. On the basis of this pattern of immunofluorescence observed, a haptotactic mechanism (migration directed by binding with a specific peptide within FN acting as the binding site) of precardiac cell migration was suggested in which the cells continually move toward areas of greater adhesion (Linask & Lash, 1986). An illustration of this arrangement is presented in Fig. 1A. These authors also reported an inhibition of heart development when the embryos were incubated with antibodies to FN or to its receptor, or with arginine-glycine-aspartic acid synthetic peptide, which mimics FN’s binding site and blocks its binding to cell surface receptors (Linask & Lash, 1988a). In another experimental study by these authors (Linask & Lash, 1988b), microsurgical rotation of the precardiac mesoderm together with adjacent endoderm 180° resulted in abnormal or ectopic heart development. Control experiments in which the two-layered explants were removed and then replaced in original orientation developed normally. These studies confirm an important role for FN, though they do not necessarily indicate a mechanism of haptotactic migration in response to a FN concentration gradient. Abberrant directional migration and ectopic heart tube formation would also be predicted after rotation if an endodermally produced, anterior, confined patch of fibronectin, residing in and near the head fold, and included in the explant, is responsible
. .
36
for creating a pulling tension through the precardiac mesoderm that drives migration. In sham-operated controls with no rotation, the tension would be regenerated after a short healing time.
Evidence for an Anterior, Confined Patch of Fibronectin We have examined the distribution of FN using sagittal and transverse sections, and whole mounts of embryos, during the time of migration. The complete sagittal profile of FN immunostaining in the precardiac mesoderm offers the most definitive assessment of its anterior-posterior deployment. We employed an immunostaining method that allows visualization of FN in eosin-hematoxylin stained tissue sections, facilitating the interpretation of exact localization. We used a highly specific monoclonal antibody to FN, and a specific antiserum to chicken type IV collagen and another to fibronectin for comparison purposes. Our findings led me to propose a different mechanism of precardiac mesoderm migration. We found (Plate 1) that in the pre-heart embryo, FN is localized with striking intensity in two areas: the prospective proamnion (outside the embryo body) and the forming head fold (Plate 1, A–C). It appears as a concentrated, crescent-shaped region associated with the basement membrane of the anterior germ layers, particularly the endoderm, and as the layers fold to yield a head fold with AIP, the FN-stained region also folds (Plate 1, C). This action results in a ventrally located, foregut endodermal crescent that appears to maintain a ‘‘grip’’ on the adjacent
precardiac mesoderm. The folding movements of the foregut endoderm have long been implicated in cardiac morphogenesis (Bellairs, 1953). As development of the heart proceeds, this area increases to an apparent maximum of staining intensity shortly before the tubular heart is formed (Plate 1, F). Pre-endocardial cells then move into the extracellular space between the two layers and become coated with FN. These cells adhere to each other and quickly form paired endocardial tubes that are covered with and connected by FN (Plate 1, F). Meanwhile, the remaining precardiac mesoderm surrounds the endocardial tubes as they fuse, forming the myocardium. The primitive heart tube takes on definitive form and function (contractions) at about 33 hr of incubation. Type IV collagen was found to be mainly co-localized with FN during precardiac cell migration, but it was not observed to be accentuated in the head fold as fibronectin was (Plate 1, D). This result is in agreement with that of Drake et al. (1990). Our localization of type IV collagen demonstrates an ECM molecule distribution that parallels that of FN in many respects, but also stands in contrast to the striking restricted patch of intense FN staining, which was displayed with both the monoclonal and polyclonal antibodies. The existence of a highly concentrated anterior region of FN within the pre-heart chick embryo has not been previously reported, although the staining that we also observed in the anterior extraembryonic region has been (Critchley et al., 1979). Linask & Lash (1986) presented some evidence for a gradient distribution shown by immunofluorescence photographs in an anterior-to-posterior sequence of
F. 1. Sketch illustrating two alternative models of precardiac mesodermal cell migration along FN-containing basement membrane of adjacent endoderm. A. The precardiac cells all respond to a gradient of FN, utilizing the gradient to determine direction for migration. Each cell moves individually even though the cells are interconnected. B. The precardiac cells depend on adhesion of the anterior-most cells of the layer to a concentrated, crescentic patch of fibronectin. As a result of adhesion they develop tension which is transmitted posteriorly throughout the layer. The cells then advance anteriorly as a layer.
J. theor. Biol.
P 1. Sections through embryos of approximately 22 and 30 hr incubation, showing the monoclonal antibody localization of fibronectin and type IV collagen. A–D are sagittal sections of a stage 6 embryo and anterior is to the left; E and F are transverse sections. (A) Low magnification view of entire embryo’s sagittal profile stained anti-FN. Two patches of intense staining are located at the anterior end in this section taken lateral to the head fold. (B) Higher magnification of the section shown in A. (C) Another sagittal section taken closer to the midline showing the early head fold and the disposition of the FN patches: the more anterior patch is localized in the prospective proamnion whereas the more posterior patch is within the head fold. (D) Parasagittal section equivalent to that shown in B, but stained with antibodies to type IV collagen. The localization is generalized in the anterior-posterior axis. (E) Control cross-section through the AIP of a stage 9 chick embryo, with normal mouse serum replacing antibody. Blue color of the tissues is from counterstaining with hematoxylin. (F) Anti-FN stained transverse section through a stage 9 embryo showing red staining associated with all basement membranes, head mesenchyme, and the fusing endocardial tubes of the primitive heart. Staining is intense between precardiac mesoderm and foregut endoderm, and surrounds a pre-endocardial cell in this area dorsal to the forming heart (arrow). It lines and joins the two tubes prominently. Also note the striking staining of neutral crest cells. hf, head fold; fg, foregut; nc, neural crest; nt, neural tube. (A) Scale bar = 200 mm; (B–F) Scale bar = 100 mm.
D. J. WIENS
facing p. 36
separate regions of a section from a stage 5 and stage 7 embryo. The anti-FN used was a rabbit anti-chick antiserum, and staining procedures were carried out in fixed, methanol-permeabilized whole embryos before sectioning. Drake et al. (1990), however, employing a different antiserum (affinity-purified rabbit antibodies to chicken FN, Little & Chen, 1982) also incubated prior to sectioning, were unable to duplicate this observation, and reported uniform immunostaining across the heart-forming region. However, the fluorescence photomicrographs shown in their report were en face views of the ECM stained for FN, and it was not indicated whether they included the anterior-most regions within the head fold. The anti-FN antibodies used have been reported to display some cross-reactivity with collagen (Little & Chen, 1982). It is possible that some non-specific staining with the antisera used in both these reports masked the higher intensity regions that we observed using monoclonal antibodies, or that greater availability of the antigenic sites and more direct washing in sectioned tissue afforded more resolution in immunostaining, having less dependence on antibody penetration of the tissues. Another possibility is that the monoclonal anti-FN recognizes an antigenic site found in only a subset of the whole population of FN molecules. This raises the interesting possibility that a specific FN subtype(s) serves to guide the precardiac cells. However, our results with the polyclonal antibodies were virtually identical to those with the monoclonal antibody.
An Alternative Model The formation and maintenance of this crescentic localized region of FN at the anterior end suggests that precardiac cell migration could depend mainly on the anterior-most mesodermal cells. Their adhesion would be high and their motility activated, allowing them to establish a contractile force leading to a pulling tension throughout the epithelial layer of more passive, interconnected cells. Although a gradient of FN (and perhaps additional fixed ligands) supporting haptotactic migration toward a maximum is logical for cells that move individually, it would be inefficient for joined cells, each of which should have no requirement to extend protrusions as well as form and break adhesions. Joined cells’ motile activities might be restrained by their attachments to other cells. Joined cells, each operating in response to a gradient to find direction and then migrate, would also require the prior graded production, secretion, and deposition of the adhesion
37
molecule by many other cells (in this case cells of the endoderm), a process that would require careful coordination. The precardiac mesoderm could operate more efficiently as a collective unit with anterior cells only required to form a strong adhesive region (Fig. 1B). The contractile force would be generated by cytoskeletal contractile activity within the anterior cells or throughout all the cells. Resistance to the tension would be provided by weak adhesion of all cells posterior to this area, where there is a continuous FN substratum of low concentration, affording little adhesion. The resistance would be gradually overcome as the anterior adhesion and the contraction increase while posterior adhesions release. The release might even be augmented by cytoskeletal changes induced by mechanoreceptive signalling mechanisms. The overall effect would eventually pull the interconnected precardiac cells anteriorly, then ventrally around the anterior edge of the foregut, and finally posteriorly again to the position where cells join the forming heart. The existence of tension in the pre-heart chick embryo has not been demonstrated, however, we have observed that a cut through the precardiac mesoderm and endoderm results in rapid retraction of the edges (Wiens et al., 1984). The process of migration of the precardiac mesoderm and ensuing tubular heart formation can be visualized as a consequence of the transitional sequence of events that occur at the FN-rich locale. When gastrulation produces an anterior group of mesodermal cells bounded above by ectoderm and below by endoderm, FN is secreted locally. The mesodermal cells attach to it. The binding acts as a cell surface signal which is transduced to begin contractile processes within the cells that create contractile force throughout the interconnected mesodermal layer. The region then bends, becomes part of the head fold, and becomes displaced ventrally and posteriorly as the foregut extends anteriorly. It then fosters pre-endocardial cells emergence and subsequent self-adhesion of endocardial cells (evident in Plate 1, F), and finally, it remains as a dorsal mesocardium through which cells continue to migrate and join the growing heart. Further studies on the adhesive, contractile, and migratory activities of the precardiac cells are needed to test the model, and should help elucidate the mechanism of their forward migration. Tests of migratory ability of intact, intact but subdivided, dissociated, and reassociated embryonic epithelia, cultured upon various arrangements of FN would serve as a starting point, and would offer the opportunity to investigate the effects of agents that can perturb the migrating cells in specific ways.
. .
38
The monoclonal antibody to avian fibronectin (B3D6, Gardner & Fambrough, 1983) was obtained from the Developmental Studies Hybridoma Bank, maintained by the Department of Pharmacology and Molecular Sciences at Johns Hopkins University School of Medicine, Baltimore, MD, and the Department of Biology at the University of Iowa, Iowa City, IA, under contract number NO1-HD-2-3144 from the NICHD. Monoclonal antibody IV-IIB12/C2 to avian type IV collagen (Mayne et al., 1983) was a generous gift from Dr. Thomas F. Linsenmayer, Department of Anatomy and Cell Biology, Tufts University, Boston, MA. I thank Jeffrey Rathmell, Tamara Mann Nowling, Clifton Hall, Scott Briggs, John Reppas, Jennifer Becker, and Melissa Appleget, students at the University of Northern Iowa who assisted with immunostaining; and Drs. Jean Gerrath and Carl Thurman of the Biology Department for critical comments on the early drafts of the manuscript. This work was supported by the American Heart Association, Iowa affiliate grant IA-90-G-40, by Summer Research Fellowships from the graduate college of the University of Northern Iowa, and by JOVE (Joint Venture Program) grant No. NAG9-265 to the University of Northern Iowa, D. Wiens and T. Hockey, Project Directors. Contributions to the work were carried out as undergraduate student research projects in partial fulfillment of the B.S. degree in biology by Jeffrey Rathmell, Clifton Hall, and Scott Briggs.
REFERENCES A, M. (1970). Contact inhibition in tissue culture. In Vitro 6, 128–142. B, G. J. (1992). Communication Within Animal Cells. Oxford: Oxford Science Publications. B, R. (1953). Studies on the development of the foregut in the chick blastoderm. II. The morphogenetic movements. J. Embryol. Exptl. Morphol. 1, 369–385. B, B. (1971). Microtubules and microfilaments in newt neurulation. Dev. Biol. 26, 416–441. C, S. B. (1967). Haptotaxis and the mechanism of cell motility. Nature 213, 256–260. C, D. R., E, M. A. & H, R. O. (1979). Distribution of fibronectin in the ectoderm of gastrulating chick embryos. Nature (London) 280, 498–500. D, G. (1984). Migration-directing liquid properties of embryonic amphibian tissues. Amer. Zool. 24, 649–655. D H, R. L. (1963a). Organization of the cardiogenic plate in the early chick embryo. Acta Embryol. Exptl. 6, 26–38. D H, R. L. (1963b). Migration patterns of the precardiac mesoderm in the early chick embryo. Exptl. Cell Res. 29, 544–560. D H, R. L. (1964). Cell interactions and oriented movements during development. J. Exp. Zool. 157, 127–238. D H, R. L. (1965). Morphogenesis of the vertebrate heart. In: Organogenesis, pp. 377–419. New York: Holt, Rinehart & Winston Co. D, C. J., D, L. A., W, L. & L, C. D. (1990). Avian vasculogenesis and the distribution of collagens, I, IV, laminin, and fibronectin in the heart primordia. J. Exp. Zool. 255, 309–322. D, J. L. & T, J. P. (1982). Appearance and distribution of fibronectin during chick embryo gastrulation and neurulation. Dev. Biol. 94, 337–350. E, M. A. & W, J. (1977). Scanning electron microscopy of the developing mesoderm layer in the chick embryo. Anat. Embryol. 150, 291–300. G, J. M. & F, D. M. (1983). Fibronectin expression during myogenesis. J. Cell Biol. 96, 474–485.
H, E. D. (1968). Organization and fine structure of epithelium and mesenchyme in the developing chick embryo. In: Epithelial-Mesenchymal Interactions (R. Fleischmajer & R. E. Billingham, eds) pp. 31–55. Baltimore: Williams and Wilkins. H, E. D. (1990). Epithelial-mesenchymal transitions. Sem. Dev. Biol. 1, 347–356. H, R. O. (1985). Molecular biology of fibronectin. Annu. Rev. Cell Biol. 1, 67–90. H, J. (1933). Die totale Exogastrulation, eine Selbstablo¨sung des Ekdoderms vom Entomesoderm. Arch. Entw. Mech. 129, 669–793. H, J. (1934). Formative Reize in der Embryonalentwicklung der Amphibien, Dargestellt an Explantationsversuchen. Arch. exp. Zellf. 15, 281–301. H, J. (1943). A study of the mechanisms of gastrulation. Part I. J. Exp. Zool. 94, 261–318. H, J. (1944). A study of the mechanics of gastrulation. Part II. J. Exp. Zool. 95, 171–212. I, J. M. & M, F. J. (1983). Fibronectin distribution during early chick embryo heart development. Dev. Biol. 95, 19–30. J, A. B., O, G. F., O, G. M. & C, L. Y. (1986). Neurulation and the cortical tractor model for epithelial folding. J. Embryol. Morphol. 96, 19–49. K, R. E. & S, G. C. (1977). An SEM study of cellular morphology, contact, and arrangement as related to gastrulation in Xenopus laevis. Wilhelm Roux Arch. Dev. Biol. 182, 165–186. K, R. E. (1980). The cellular basis of epiboly: an SEM study of deep cell rearrangement during gastrulation of Xenopus laevis. J. Embryol. exp. Morphol. 60, 201–243. K, R. E. (1986). The cellular basis of morphogenesis. In: Developmental Biology: A Comprehensive Synthesis (L. Browder, ed.) pp. 241–327, Vol. 2. New York: Plenum. L, J. W. (1955). Studies on wound closure in urodeles. J. Exp. Zool. 128, 13–28. L, K. K. & L, J. W. (1986). Precardiac cell migration: fibronectin localization at mesoderm-endoderm interface during directional movement. Dev. Biol. 114, 87–101. L, K. K. & L, J. W. (1988a). A role for fibronectin in the migration of avian precardiac cells. I. Dose-dependent effects of fibronectin antibody. Dev. Biol. 129, 315–324. L, K. K. & L, J. W. (1988b). A role for fibronectin in the migration of avian precardiac cells. II. Rotation of the heart-forming region during different stages and its effects. Dev. Biol. 129, 324–329. L, C. D. & C, W.-T. (1982). Masking of extracellular collagen and the co-distribution of collagen and fibronectin during matrix formation by cultured embryonic fibroblasts. J. Cell Sci. 55, 35–50. M, F. J. (1968). Embryonic development of the heart. I. A light and electron microscopic study of myocardial development in the early chick embryo. J. Morphol. 125, 329–366. M, R., S, R. D., W, H., F, J. M. & L, T. F. (1983). The use of monoclonal antibodies to fragments of chicken type IV collagen in structural and localization studies. J. Biol. Chem. 258, 5794–5797. P, H. M. & D, G. S. (1978). Liquid-tissue mechanics in amphibian gastrulation: germ-layer assembly in Rana pipiens. Amer. Zool. 18, 81–93. R, G. C. & D H, R. L. (1966). Migration of precardiac cells in the chick embryo: a radioautographic study. Carnegie Inst. Wash. Contributions to Embryology 38, 111–121. S, D.-L., D, M., D, T., R, J.-F. & B, J.-C. (1987). Experimental analysis of the extension of the dorsal marginal zone in Pleurodeles waltl gastrulae. Development 100, 147–161. S, J. & K, R. (1992). Cell motility driving mediolateral intercalation in explants of Xenopus laevis. Development 116, 901–914.
S, H. (1931). Uber den Anteil von Implantat und Wirtskeim an der Orientierung und Beschaffenheit der induzierten Embryonalanlage. Arch. Entw. Mech. 123, 390–517. S, A. W. (1971). Cell junctions and their role in transmural diffusion in the embryonic chick heart. Z. Zellforsch. 120, 463–487. S, B. S. (1974). Morphogenesis of vertebrate organs. In: Concepts of Development (J. Lash & J. R. Whittaker, eds) pp. 213–240, Stamford, CN: Sinauer & Associates, Inc. S, H. & D H, R. L. (1969). The precardiac areas and formation of the tubular heart in the chick embryo. Dev. Biol. 19, 128–159. S, M. S. (1963). Reconstruction of tissues by dissociated cells. Science 141, 401–408. S, M. S. (1975). Adhesion-guided multicellular assembly: a commentary upon the postulates, real and imagined, of the differential adhesion hypothesis, with special attention to computer simulations of cell sorting. J. theor. Biol. 55, 431–443. S, M. S. & T, M. (1994). Experimental specification of cell sorting, tissue spreading, and specific spatial patterning by quantitative differences in cadherin expression. Proc. Nat. Acad. Sci. U.S.A. 91, 206–209. T, P. L. & H, J. (1955). Directed movements and
39
selected adhesion of embryonic amphibian cells. J. Exp. Zool. 128, 53–120. T, R., S, K. & T, S. (1991). The behavior and cytoskeletal system of chick gastrula mesodermal cells on substrata coated with lines of fibronectin. J. Exp. Zool. 260, 345–353. T, R. L., H, E. D. & R, J. P. (1967). Cell contact during early morphogenesis in the chick embryo. Dev. Biol. 16, 78–106. T, J. P. (1976). On the mechanism of Metazoan cell movements. In: Cell Surface Reviews (G. Poste & G. L. Nicholson, eds), Vol. 1, pp. 225–329. Amsterdam: NorthHolland Publishing Co. W, D., S, M. & S, B. S. (1984). Precardiac mesoderm differentiation in vitro: actin-isotype synthetic transitions, myofibrillogenesis, initiation of heartbeat, and the possible involvement of collagen. Differentiation 28, 62–72. W, P. & K, R. (1991). Cell rearrangement during gastrulation of Xenopus: direct observation of cultured explants. Development 112, 289–300. W, R. (1990). Mesoderm cell migration during Xenopus gastrulation. Dev. Biol. 142, 155–168. W, R. & N, M. (1991). Directional mesoderm cell migration in the Xenopus gastrula. Dev. Biol. 148, 573–589.