Cell Surface 01,4-Galactosyltransferase Functions during Neural Crest Cell Migration and Neurulation In Vivo

Mesenchymal cell migration and neurite outgrowth are mediated in part by binding of cell surface 01,4-galactosyltransferase (GalTase) to N-linked oligosaccharides within the E8 domain of laminin . In this study, we determined whether cell surface GalTase functions during neural crest cell migration and neural development in vivo using antibodies raised against affinity-purified chicken serum GalTase . The antibodies specifically recognized two embryonic proteins of 77 and 67 kD, both of which express GalTase activity. The antibodies also immunoprecipitated and inhibited chick embryo GalTase activity, and inhibited neural crest cell migration on laminin matrices in vitro . AntiGalTase antibodies were microinjected into the head mesenchyme of stage 7-9 chick embryos or cranial to Henson's node of stage 6 embryos . Anti-avian GalTase IgG decreased cranial neural crest cell migration on THE mechanisms by which cells recognize and interact with their environment, either with the extracellular matrix or with neighboring cells, are initiated at the cell surface. Several cell surface receptors that function during cellular interactions have been identified, characterized, and subsequently been shown toplay a critical role in embryonic development by mediating cellular adhesion or migration, or by transducing appropriate receptor-mediated signals that dictate an inductive response (Austin and Kimble, 1987; Takeichi, 1988; Drake and Little, 1991 ; Flanagan et al ., 1991 ; Kramer et al ., 1991 ; Olson et al ., 1991) . One ofthese cell surface receptors is 01,4-galactosyltransferase (GalTase),' which participates in a number of cellular interactions, including cell migration and spreading on laminin matrices (Runyan et al ., 1986, 1988 ; Eckstein and Shur, 1989 ; Romagnano and Babiarz, 1990), neurite initiation and outgrowth (Begovac and Shur, 1990 ; Begovac et al ., 1991), sperm-egg binding (Lopez et al ., 1985 ; Miller et al ., 1992), embryonic cell-cell adhesion (Bayna et al ., 1988 ; Hathaway et al ., 1989), and uterine epithelial cell-cell adhesion (Duff et al ., 1987) . GalTase functions as an adhesion Correspondence may be addressed to Barry D. Shur . 1 . Abbreviation used in this paper: GalTase, ßl,4-galactosyltransferase . © The Rockefeller University Press, 0021-9525/92/04/369/14 $2 .00 The Journal of Cell Biology, Volume 117, Number 2, April 1992 369-382 the injected side but did not cross the embryonic midline and did not affect neural crest cell migration on the uninjected side . Anti-avian GalTase Fab crossed the embryonic midline and perturbed cranial neural crest cell migration throughout the head. Neural fold elevation and neural tube closure were also disrupted by Fab fragments . Cell surface GalTase was localized to migrating neural crest cells and to the basal surfaces of neural epithelia by indirect immunofluorescence, whereas GalTase was undetectable on neural crest cells prior to migration . These results suggest that, during early embryogenesis, cell surface GalTase participates during neural crest cell migration, perhaps by interacting with laminin, a major component of the basal lamina . Cell -surface GalTase also appears to play a role in neural tube formation, possibly by mediating neural epithelial adhesion to the underlying basal lamina . molecule by binding to specific oligosaccharide residues in the extracellular matrix or on adjacent cells . At present it is unclear whether cell surface GalTase functions as a lectin, similar to the selectin class of carbohydrate-binding receptors (Larsen et al ., 1990 ; Lowe et al ., 1990), or catalytically, as occurs in the Golgi during glycoprotein biosynthesis . Nevertheless, with the increasing body ofevidence implicating a role for carbohydrates in normal cellular processes (Kunemund et al ., 1988 ; Reddy et al ., 1989 ; Shur, 1989 ; Griffith and Wiley, 1990; Larsen et al ., 1990 ; Lowe et al ., 1990 ; Woo et al ., 1990 ; Stamenkovic et al ., 1991), it is important to determine the function of cell surface carbohydrate-binding glycoproteins during embryonic development . The function of cell surface GalTase during cell migration has been studied extensively in vitro . GalTase is preferentially localized to the leading edges of migrating cells (Eckstein and Shur, 1989), where it mediates cell spreading and migration by binding to N-linked oligosaccharides within the E8 domain of laminin (Begovac and Shur, 1990) . GalTase does not appear to function during initial cell adhesion to laminin (Runyan et al ., 1988 ; Romagnano and Babiarz, 1990), nor during cell migration on matrices not containing laminin (Runyan et al ., 1986 ; Eckstein and Shur, 1989 ; Romagnano and Babiarz, 1990) . These processes rely upon other matrix receptors, notably the integrins (Hynes, 1987) . 369 on July 9, 2017 jcb.rress.org D ow nladed fom Laminin induces GalTase expression onto the growing lamellipodium (Eckstein and Shur, 1989; Romagnano and Babiarz, 1990), where it becomes associated with the cytoskeleton (Eckstein and Shur, 1992), thus facilitating cell spreading and migration . Ultimately, it is critical to determine if molecular models for cell/matrix interactions developed in vitro actually function during embryogenesis in vivo . The microinjection of mRNA into amphibian embryos (Harvey and Melton, 1988 ; Ruiz i Altaba and Melton, 1989) or of antibodies and competing peptides into amphibian or avian embryos (Boucat et al ., 1984 ; Bronner-Fraser, 1985 ; Bronner-Fraser and Lallier, 1988) has become a useful method for identifying molecules that play a role in embryogenesis . Antibodies against specific cell surface receptors, or peptides that compete for binding to these receptors, block neural crest cell migration when injected into developing chick embryos . These results have demonstrated an in vivo role for such cell surface and extracellular matrix components as integrin (BronnerFraser, 1985), fibronectin (Boucat et al ., 1984), the HNK-1 epitope (Bronner-Fraser, 1987), and a laminin/heparin sulfate proteoglycan complex (Bronner-Fraser and Lallier, 1988) . In this study, we examined the role ofcell surface GalTase during avian embryogenesis . Anti-GalTase IgG and Fab fragments, which specifically recognized chick embryo GalTase and inhibited GalTase activity, perturbed avian neural crest cell migration on laminin matrices in vitro. When these antibodies were injected into the head mesenchyme of early chicken embryos, neural crest cell migration was inhibited . Anti-GalTase Fab fragments also induced neural tube defects due to their ability to cross the embryonic midline . GalTase was localized by indirect immunofluorescence to migrating neural crest cells, as well as to the basolateral regions of neural epithelial cells . These results suggest a role for cell surface GalTase in mediating neural crest cell migration and neurulation in vivo . Materials and Methods