A monoclonal antibody (mAb 8C8) that recognizes the Xenopus β1-integrin chain was used to study the appearance, synthesis and distribution of this integrin subunit during the early development of Xenopus. Both the precursor and the mature form of β1-integrin are provided maternally. They do not increase significantly in amount until early gastrula when the level of both forms begins to rise gradually.

Synthesis of β1-integrin from maternal mRNA is observed throughout the pregastrula phase, though it seems to add only little to the total β1-integrin of the embryo. Until late blastula only small amounts of precursor are processed into the mature form.

Starting with the formation of the first cleavage membrane, mature β1-integrin is inserted into the newly formed plasma membranes of all cells. The membrane domains forming the outer surface of the embryo remain devoid of the antigen. The data suggest an as yet unknown function of β1-integrin during the cleavage phase.

During vertebrate development, the individual embryonic cells become integrated to form an increasingly complex system of cell-cell and cell-matrix connections. Important features of development like the maintenance of the integrity and shape of the embryo, the ordered performance of the morphogenetic movements, and the proper positioning of the embryonic regions during inductive interactions depend on the spatial structure and the molecular composition of this system. It may therefore be anticipated that membrane receptors involved in cell-cell and cell-matrix adhesion play an important role in embryonic development. Integrins are a family of transmembrane receptors, which exhibit the required properties for this role.

A functional integrin consists of one α and one β glycoprotein subunit. A multitude of members within the α and β families of polypeptides and the way the variant chains are combined allow for a large number of different integrins to become expressed on the plasma membrane (for recent reviews see Hynes, 1987; Ruos-lahti and Pierschbacher, 1987; Akiyama et al., 1990; Albeda and Buck, 1990; Ruoslahti, 1991).

The structural variability of the integrins correlates with a great variety of integrin functions. Thus, integrins enable cells to adhere to a number of compounds of the extracellular matrix; prominent ligands of integrins are fibronectin, vitronectin, collagens and laminin. Some, though not all integrins bind their ligands via the RGD-sequence that is found in many of the extracellular matrix proteins. This binding can be characteristically inhibited by synthetic peptides that contain the RGD-sequence (Pierschbacher and Ruoslahti, 1984). Integrins have also been found to be engaged in direct cell-cell adhesion (Kishimoto et al., 1989; Campanero et al., 1990). However, the role of integrins may go beyond these mechanical aspects of tissue organisation. The binding of a ligand to an integrin may be a signal that fundamentally affects the physiological condition of a cell (Kornberg et al., 1991).

The pathway by which functional integrins are formed has been studied most extensively for β1-chain-containing integrins of mammalian cultured cells. The β1-chain polypeptide with an apparent Mr of 89 × 103becomes N-glycosylated to form a precursor molecule with an apparent Mr of 105 × 103 (Akiyama et al., 1989). The precursor is often synthesized in excess and, depending on the cell line and the physiological condition of the cell, an intracellular store of precursor may accumulate (De Strooper et al., 1991). Further glycosylation yields the mature β1-integrin with an apparent Mr of 125×103 (Akiyama et al., 1989). This maturation is a remarkably slow process. In mouse and human fibroblasts, it occurs with a half time of 3–4 hours (Akiyama and Yamada, 1987; De Strooperet al., 1991). Complete glycosylation is required for β1-integrin to form functional integrin complexes (Akayima et al., 1989). The mature form of β1003E-integrin becomes quickly inserted into the plasma membrane and no intracellular pool of mature β1-chains is observed (De Strooper et al., 1991). The association of β1-chains with α-chains seems to occur early in the pathway of synthesis and chains free of n-chains have not been found on the cell membrane (Cheresh and Spiro, 1987; Akiyama et al., 1989).

The role of integrins in early amphibian development has been discussed recently in a review by Keller and Winklbauer (1991). Using a heterologous antibody raised against the chicken fibronectin receptor, Darri-bere et al. (1988) observed the presence of the antigen in Pleurodeles embryos from early on. Though functional assays indicated that the antigen is a fibronectin receptor, its subunit constitution was not elucidated in detail. Injection of antibodies against this fibronectin receptor (Darribere et al., 1988) or RGD peptide (Boucaut et al., 1984) into the blastocoel cavity interrupts gastrulation, as does the injection of antibodies against the intracellular domain of β1-integrin into early blastomeres (Darribere et al., 1990). Taken together, these observations suggest the participation of integrins in Pleurodeles gastrulation.

In Xenopus development, integrin function becomes first apparent at the late blastula stage. This is the time when explanted cells gain the ability to adhere to and to spread on a fibronectin substratum (Johnson and Silver, 1986; Winklbauer, 1988). Since these cellular activities are inhibited by the addition of RGD peptide, the participation of an integrin-type fibronectin receptor was inferred (Winklbauer, 1988; Winklbauer, 1990). At the onset of gastrulation, a network of fibronectin fibrils forms on the blastocoel roof. The directed migration of the prospective head mesoderm during gastrulation seems to be supported and guided by this matrix (Winklbauer et al., 1992; Winklbauer and Nagel, 1991). The function of an integrin-type fibronectin receptor for the establishment and maintenance of this network and also for the adhesion of the mesodermal cells to the network is implied (Winklbauer and Nagel, 1991). In addition DeSimone and Hynes (1988) showed that during Xenopus development β1-integrin mRNA does not appear until the early gastrula. Thus, the observations on the appearance of β1-integrin mRNA and on the timing of a putative β1-integrin function seemed to complement each other in a satisfying way.

However, conclusions from the temporal coincidence of β1-integrin mRNA expression and the commencing putative β1-integrin functions in the embryo must remain preliminary as long as no detailed data exist concerning the β1-integrin glycoprotein itself. Indeed, Smith et al. (1990) reported that induction with XTC-MIF enhances the ability of late animal blastomeres to spread on fibronectin though no significant stimulation of β1-integrin synthesis occurs. This finding, that commencing synthesis is not the cause of the onset of β1integrin function in the late blastula embryo, raises a number of questions: When does β1integrin first appear in the embryo? When is it synthesized? When is the precursor transformed into the mature form and when and where is this mature form inserted into the blastomere membranes? This paper is intended to answer these questions and thereby to provide a precondition for the understanding of the regulation of β1-integrin function in the early Xenopus embryo.

Screening of a cDNA clone library and DNA sequencing

105 pfu of a λgt10 cDNA library from Xenopus neurulae stage 17 (Kintner and Melton, 1987) were screened by hybridisation at high stringency (5×SSPE, 1,5% N-lauryl-sarcosine, 0.05% salmon sperm DNA, 5 g/30 ml dextran sulphate, at 68°C, washing in 2×SCP, 0.1% SDS, at 58°C and finally in 0.2×SCP, 0.1% SDS, at 58°C) with an 1.5 kb EcoRl/NcoI fragment of the human β1-integrin cDNA clone pGEMI P-32 (Argraves et al., 1987). The 2.9 kb insert of a single purified clone was subcloned in pBluescript II SK+ and characterized by restriction mapping. Three isolated restriction fragments were sequenced by the dideoxy chain termination method (Sanger et al., 1977) using the Sequenase Kit Version 2.0 (United States Biochemistry Corporation).

Preparation of the fusion protein

The lacZ-pUR expression system (Rüther and Müller-Hill, 1983) was used to producé a β-gal-ft-integrin fusion protein. The lacZ-β-integrin fusion gene was obtained by cloning the 1.8 kb ApaI/SphI-fragment into the SalI sites of pUR 288. The fusion protein was expressed and harvested as described by Fey and Hausen (1990).

Polyclonal anti--integrin antibodies

Polyclonal antibodies against the fusion protein were produced as described (Fey and Hausen, 1990). For affinity purification of β1-integrin antibodies, rabbit antiserum was preabsorbed exhaustively with an affinity matrix carrying protein from the E. coli strain XII and E. coli β-galactosidase. Integrin antibodies were then isolated by affinity chromatography on a column with coupled lysate of pUR 288/β1-integrin-transformed, induced E. coli XII cells.

Polyclonal rabbit antiserum against a synthetic peptide of 39 residues from the C-terminal sequence of human β1 integrin was a gift from Dr R. O. Hynes, Cambridge, USA (Marcantonio and Hynes, 1988).

Establishment of hybridoma cell lines, production and purification of monoclonal antibodies

Mice were immunized with a protein fraction that was enriched in β1-integrin from XTC-cells by affinity chromatography on matrix-bound polyclonal rabbit anti-β1-integrin. Mouse hybridoma cell lines were established as described by Galfre and Milstein (1981) and Kearney et al. (1979). Screening was performed on western blots of total XTC-cell protein and on fixed and permeabilized XTC cells by immunofluorescence selecting for typical focal contact staining. Antibodies of the established line 8C8 were purified from culture supernatants as described by Harlow and Lane (1988).

Isolation of RNA and northern blotting

Isolation of total RNA from staged embryos, northern blotting, and hybridisation to 32P-labelled antisense RNA probes were carried out as described by Ellinger-Ziegelbauer and Dreyer (1991). Poly(A)+ RNA was selected by a mRNA isolation kit (Stratagene) according to the manufacturer’s instructions.

Embryos

Xenopus embryos were obtained as described by Fey and Hausen (1990) and staged according to Nieuwkoop and Faber (1967).

Preparation of extracts from embryos and from tissue culture cells

Embryos of different stages were sonicated in lysis buffer A (2% NP40,120 mM NaCl, 1 mM CaCl2,1 mM MgCl2, 50 mM Tris/HCl pH 7.4, 2μ M Pepstatin/Leupeptin, 2 mM PMSF, 1 mM iodoacetamide) for direct electrophoresis or in RIPA (0.1% SDS, 0.5% sodium desoxycholate, 1% NP40,150 mM NaCl, 1 mM CaCl2, 50 mM Tris/HCl pH 7.4, protease inhibitors as before) for immunoprecipitation. 10 μl of lysis buffer were used per embryo. Yolk was sedimented by centrifugation (1 minute at 14000 revs/minute). The supernatant was then extracted with an equal volume of 1,1,2 trichlor-trifluorethane and centrifuged as before. Extracts were immediately heated in sample buffer to 95°C (Laemmli, 1970) or stored in liquid nitrogen until further use.

Isolation of 8C8 antigens

2 mg purified mAb 8C8 were coupled to 200 μl CNBr-activated Sepharose (Pharmacia) according to the manufacturer’s instructions. The beads were packed into a 1 ml column, equilibrated with RIPA and washed with three cycles of elution buffer (100 mM glycine, 0.1% NP40, pH 2.5). A RIPA lysate of about 10 ml stage-30 embryos was loaded on the column. After extensive washing, the column was eluted in 150 μl fractions with elution buffer. The eluates were neutralized and immediately mixed with sample buffer and heated to 95°C before electrophoresis (Laemmli, 1970) or stored in liquid nitrogen until further use.

Digestion of proteins with endoglycosidase H and N-glycosidase F

Digestion of isolated ft-integrin protein with endoglycosidase H and N-glycosidase F (Boehringer) was carried out according to the manufacturer’s instructions. Briefly, 100 ng of purified antigen each were reduced with β-mercaptoethanol and treated overnight with 5 mU endoglycosidase H or 1 U N- glycosidase F in digestion buffer. The reaction was stopped by treament with sample buffer and heating to 95°C (Laemmli, 1970).

Immunohistological procedures

Whole-mount staining of different embryonic stages was performed with fluorescent antibodies as described by Angres et al. (1991). The specimens were embedded in glycolmethacrylate (Technovit 7100 from Kulzer, FRG). 5 pm sections were prepared for microscopic observation. Immunofluorescence was detected with a Zeiss Axioplan microscope equipped with epifluorescence optics.

SDS-PAGE and immunoblotting

SDS-PAGE was carried out according to Laemmli (1970) using 7% gels. Non-reduced samples were treated with sample buffer containing 2 mM iodoacetamide before heating to 95°C. Reduction of electrophoresed proteins in the gel was carried out by incubating the gel in 125 mM Tris/HCl pH 6.8, 5% β-mercaptoethanol for 30 minutes.

Silver staining of gels was performed as described by Morrissey (1981). For immunoblotting, proteins were electro-phoretically transferred to nitrocellulose membranes. Blots were blocked for 30 minutes in PBS, 0.05% Tween 20, 10% low fat milk and immunostained as described by Fey and Hausen (1990). mAb 8C8 and rabbit anti-ft-integrin antibody were diluted to approximately 1 μg/ml in blocking buffer. Polyclonal antiβ1-integrin-peptide antiserum was diluted 1:1500 in blocking buffer. Secondary peroxidase-coupled antibodies (Dianova) were diluted 1:3000 in blocking buffer. Staining was detected using the ECL western blot detection system (Amersham). Under these conditions, polyclonal anti-β1integrin-peptide antiserum showed about 1/10 of the reactivity of the two other antibodies.

Labelling of cleavage blastomeres with biotin and isolation of biotinylated proteins by affinity to streptavidin

Surface labelling of cleavage-stage embryos was carried out according to the method of von Boxberg et al. (1990). Fertilized eggs were devitellinized and cultured in an agarose-coated Petri dish in MBS-H without Ca2+/Mg2+ (88 mM NaCl, 1 mM KC1, 2.4 mM NaHCO3,10 mM Hepes, 10 mg/ml each of Penicillin and Streptomycin, pH 7.4) until they reached stage 6. Care was taken that all damaged cells were removed from the disaggregated embryos. A stock solution of 4 mg Biotin-X-NHS (Calbiochem) in 20 μl DMSO was added along the periphery of the dish. After 10 minutes the reaction was terminated by the addition of 500 μl 500 mM glycine pH 7.4. Intact cells were purified from most of the non-bound biotin by sedimenting them at 1 g through a cushion of 500 mM glycine pH 7.4. Lysates of labelled cells (buffer A) were dialysed extensively against S-buffer (0.3% NP40, 120 mM NaCl, 1 mM CaCl2, 1 mM MgCl2, 10 mM Tris/HCl pH 8.0) and centrifuged for 30 minutes in an Eppendorf centrifuge.

50 μl of streptavidine-agarose (Sigma) were washed two times in S-buffer and incubated overnight at 4°C with the lysates of labelled cells. The beads were washed three times with S-buffer and eluted by heating them to 95°C in sample buffer (Laemmli, 1970). Eluted proteins were separated by PAGE and analysed by immunoblotting. For controls un-labelled extracts were processed likewise.

Labelling of fertilized eggs with L-[35S]methionine

Fertilized eggs were dejellied in 2% cysteine pH 8.0 about 30 minutes post-fertilization and immediately injected with 2 μCi L-[35S]methionine (cell labelling grade, Amersham, concentrated to 1/10 of the original volume) and incubated in 1/10 MBS-H. At stages 6, 9 and 11, forty embryos each were lysed in RIPA containing 10 mM L-methionine and subjected to immunoprecipitation with mAb 8C8. The total injected radioactivity and the amount of protein-incorporated label was determined as described by Müller et al. (1992).

Immunoprecipitation

Immunoprecipitation was carried out as described by Müller et al. (1992). Briefly, 10 μg of purified antibody were coupled to 15 μl of protein G-sepharose preabsorbed with lysates of non-labelled Xenopus embryos. Lysates for immunoprecipitation were precleared by centrifugation (1 hour, 25000 revs/minute, TL 100 ultracentrifuge, TLA 100.2 rotor) and incubated for two hours with the protein G-sepharose-coupled antibodies. Beads were washed extensively and precipitated proteins were eluted with sample buffer at 95°C.

Preparation of anti-integrin antibodies

Using a cDNA of human β 1 -integrin to probe a Xenopus neurula λgt10 cDNA library a Xenopus-β1- integrin cDNA-clone was isolated. The restriction fragment pattern and the sequence of three fragments of different regions of that clone revealed a 100% identity with the Xenopus β 1 *-integrin cDNA described by DeSimone and Hynes (1988) (Fig. 1C).

Fig. 1.

Cloned c-DNA fragments of the Xenopus β 1*-integrin chain. (A) β 1-integrin with the extracellular domain (ECD), the transmembrane domain (TMD) and the intracellular domain (ICD) is illustrated schematically. Comparing A and B reveals the position of the coding sequence on the cDNA cloned by DeSimone and Hynes (1988). (B) The restriction map of the cloned c-DNA described by DeSimone and Hynes was deduced from the published nucleotide sequence. All restriction sites of the endonucleases used in our analysis are indicated. (C) The positions of the restriction sites on the newly isolated clone as they were determined by restriction analysis are shown to agree with the deduced restriction map. Sequenced fragments are indicated in bold and are framed with the base pair numbers taken from DeSimone and Hynes (1988). (D) The ApaI/SphI-fragment taken for the construction of the lacZ/β 1*-fusion protein and the fragment of the 3’non-coding sequence that has been used in the northern blot analysis to reproduce the results of DeSimone and Hynes (1988) are depicted.

Fig. 1.

Cloned c-DNA fragments of the Xenopus β 1*-integrin chain. (A) β 1-integrin with the extracellular domain (ECD), the transmembrane domain (TMD) and the intracellular domain (ICD) is illustrated schematically. Comparing A and B reveals the position of the coding sequence on the cDNA cloned by DeSimone and Hynes (1988). (B) The restriction map of the cloned c-DNA described by DeSimone and Hynes was deduced from the published nucleotide sequence. All restriction sites of the endonucleases used in our analysis are indicated. (C) The positions of the restriction sites on the newly isolated clone as they were determined by restriction analysis are shown to agree with the deduced restriction map. Sequenced fragments are indicated in bold and are framed with the base pair numbers taken from DeSimone and Hynes (1988). (D) The ApaI/SphI-fragment taken for the construction of the lacZ/β 1*-fusion protein and the fragment of the 3’non-coding sequence that has been used in the northern blot analysis to reproduce the results of DeSimone and Hynes (1988) are depicted.

An ApaI/SphI-fragment of the β 1 -integrin cDNA coding for a polypeptide with an Mr of 68× 103 (Fig. ID) was recloned to generate a β gal-β 1 -integrin fusion protein in E. coli. A rabbit was immunized with the purified fusion protein and antibodies against Xenopus β 1 -integrin were obtained by affinity chromatography using substrate-bound proteins of E. coli XII and E. coli β 1 galactosidase for preclearing and the β 1 -integrin fusion protein as absorbent of the specific antibodies.

Though the antibody obtained gave a strong signal on immunoblots from embryo extracts, results of immuno-stainings either of Xenopus embryo sections or of XTC-cells were unsatisfactory. As assayed by immunoprecipitation, the reaction of the antibody with the native antigen in embryo extracts was also rather weak. Making use of the residual binding capacity of the rabbit antibody to the native antigen a fraction enriched in β 1 -integrin was obtained from extracts of XTC-cells by immunoaffinity chromatography. Mice were immunized with the eluate and 11 hybridoma lines that produce mAbs directed against Xenopus β 1 -integrin were established. The characterization of one of these antibodies is given below.

Specificity of mAb 8C8

On immunoblots from stage 30 embryo extracts, mAb 8C8 recognized two antigenic components with an apparent Mr of 100 ×103 and 115 ×103 when the PAGE was run under non-reducing conditions (Fig. 2 lane A). Under reducing conditions, no signal was obtained (not shown). By way of contrast, the rabbit anti-β 1 integrin antibody reacted only poorly with the two antigenic components when they were left unreduced (not shown). Under reducing conditions, it reacted strongly with two components but these exhibited an apparent Mr of 125 ×103 and 135 ×103 (Fig. 2 lane B).

Fig. 2.

Monoclonal antibody 8C8 recognizes β1integrin in Xenopus embryos. (A,B) Embryo extracts were electrophoresed under non-reducing conditions and immunoblotted with mAb 8C8 (A) or electrophoresed under reducing conditions and immunoblotted with rabbit anti-β1-integrin antibody (B). (C,D) The antigen was purified by affinity chromatography with mAb 8C8 and electrophoresed under non-reducing (C) or reducing (D) conditions and silverstained. (E-G) The purified antigen was electrophoresed under non-reducing conditions and immunoblotted with mAb 8C8 (E). For immunoblotting with anti-β1-integrin antibodies, the purified antigen was electrophoresed under reducing conditions (F) or under non-reducing conditions and reduced after the electrophoresis (G). (H,l) Purified antigen was electrophoresed as in F and G, respectively and immunoblotted with rabbit antibody against a synthetic β1- integrin oligopeptide (Marcantonio and Hynes 1988). (K,L) Purified antigen was digested with endoglycosidase H (K) or N-glycosidase F (L), electrophoresed under reducing conditions and immunoblotted with rabbit anti-β1-integrin antibody. For A and B extracts of two embryos were applied to the gel; for C and D 15 μl and for E-L 0.3 μl of affinity purified antigen were used. The positions and sizes [Mr × 10−3] of marker proteins are indicated.

Fig. 2.

Monoclonal antibody 8C8 recognizes β1integrin in Xenopus embryos. (A,B) Embryo extracts were electrophoresed under non-reducing conditions and immunoblotted with mAb 8C8 (A) or electrophoresed under reducing conditions and immunoblotted with rabbit anti-β1-integrin antibody (B). (C,D) The antigen was purified by affinity chromatography with mAb 8C8 and electrophoresed under non-reducing (C) or reducing (D) conditions and silverstained. (E-G) The purified antigen was electrophoresed under non-reducing conditions and immunoblotted with mAb 8C8 (E). For immunoblotting with anti-β1-integrin antibodies, the purified antigen was electrophoresed under reducing conditions (F) or under non-reducing conditions and reduced after the electrophoresis (G). (H,l) Purified antigen was electrophoresed as in F and G, respectively and immunoblotted with rabbit antibody against a synthetic β1- integrin oligopeptide (Marcantonio and Hynes 1988). (K,L) Purified antigen was digested with endoglycosidase H (K) or N-glycosidase F (L), electrophoresed under reducing conditions and immunoblotted with rabbit anti-β1-integrin antibody. For A and B extracts of two embryos were applied to the gel; for C and D 15 μl and for E-L 0.3 μl of affinity purified antigen were used. The positions and sizes [Mr × 10−3] of marker proteins are indicated.

Reduction of the intramolecular disulfide bonds induces in β 1 -integrin a conformational change that is paralleled by a characteristic shift in the electrophoretic mobility (Hynes,1987). Thus, it seemed likely that the rabbit anti-β 1 -integrin antibody and mAb 8C8 were directed against the same two β 1 -integrin-related components and that the rabbit antibody preferentially recognized the reduced slow migrating form of ft-integrin, whereas mAb 8C8 reacted with β 1 -integrin only in its non-reduced fast-migrating form.

This assumption was supported by the following experiments: when antigenic material was isolated from embryo extracts by applying mAb 8C8 in affinity chromatography, the isolated proteins responded to reduction with the expected shift in electrophoretic mobility (Fig. 2 lanes C and D) and exhibited with both antibodies the same antigenic reactivity as the antigens in the embryo extracts (Fig. 2 lanes E and F).

In addition, immunoaffinity-purified proteins were electrophoresed under non-reducing conditions and brought into the reduced form by treating the gel with βmercaptoethanol. Blots of the gel were stained with the rabbit antibody against the β 1 -integrin fusion protein and in addition with an antibody directed against a synthetic peptide of the intracellular domain of human β 1 -integrin (Marcantonio and Hynes, 1988). Under such conditions both of these antibodies recognized the same double band of antigens that was recognized by mAb 8C8 on non-reduced immunoblots (compare Fig. 2 lanes G and I with Fig. 2 lane E).

The results show that all three antibodies recognize the same two antigenic components in their different states of reduction. Considering the known specificity of the two rabbit antibodies for β1integrin, we conclude that mAb 8C8 is directed against the nonreduced form of Xenopus β1-integrin.

This conclusion is further supported by the finding that the apparent MrS of the antigens are close to those of mammalian β1-integrin (Akiyama et al., 1989).

In addition, when Xenopus tissue culture cells were stained with mAb 8C8, the antigen was found to be strongly enriched in adhesion plaques, a pattern that is indicative of integrins (Burridge et al., 1988) (data not shown).

Characterization of the two antigenic components as precursor and mature form of Xenopus β1-integrin

The electrophoretic mobility of the two antigenic components recognized by mAb 8C8 is strikingly similar to that of the precursor and mature form of mammalian β1-integrin (Akiyama et al., 1989). The mammalian β1-integrin precursor is reported to be sensitive to endoglycosidase H and may thus be distinguished from the mature molecule which is resistant to this enzyme (Jaspers et al., 1988). As shown in Fig. 2, lane K, the two antigenic components isolated from Xenopus embryo extracts behaved in the same way. Upon digestion with the enzyme, the size of the putative precursor form shifted from an apparent Mr of 125 ×103 to 102 ×103 whereas the putative mature form remained unaffected. The two antigenic components were transformed into one common form with an apparent Mr of 102×103 by complete removal of the N- linked polysaccharides with N-glycosidase F (Fig. 2, lane L). Under non-reducing conditions, this denuded polypeptide was found to move with an apparent Mr of 80×10 (data not shown).

These results show that the different electrophoretic mobility of the two antigens results from their difference in glycosylation and that the antigens represent the precursor and mature form of β1-integrin. This conclusion is further corroborated by the delayed labelling of the mature form in metabolic labelling experiments (Fig. 5) and the observation that only this form is found on the plasma membranes of embryonic cells (Fig. 8).

β1-integrin is present in the embryo during all phases of development

Assaying embryos of different stages for β1 -integrin by immunoblotting with mAb 8C8 revealed that both the precursor and the mature form are already present in the egg (Fig. 3). The amount of precursor does not change significantly until stage 10 of development, when it begins to increase gradually. For the mature form, the results of different experiments from different batches of embryos showed some variation. The amount of mature form either increased slightly (Fig. 3A) or it remained constant between the stages 1 and 10 (Fig. 3B). Thereafter, the level of mature β1 -integrin rises parallel to that of the precursor.

Fig. 3.

β1 -integrin is present throughout early embryogenesis. Extracts from embryos of different stages were separated by electrophoresis under non-reducing conditions and immunoblotted with mAb 8C8. Extracts of two embryos each were applied. Developmental stages and the positions. and sizes [Mr× 10− 3] of relative molecular mass markers are indicated. A and B show the results of two experiments from different batches of embryos.

Fig. 3.

β1 -integrin is present throughout early embryogenesis. Extracts from embryos of different stages were separated by electrophoresis under non-reducing conditions and immunoblotted with mAb 8C8. Extracts of two embryos each were applied. Developmental stages and the positions. and sizes [Mr× 10− 3] of relative molecular mass markers are indicated. A and B show the results of two experiments from different batches of embryos.

For the mature form, we reproducibly observed a slight shift to a lower Mr in the course of development. This observation may indicate some further modification in the glycosylation of the mature form as development proceeds.

Synthesis of β1-integrin

The question whether the β1-integrin found before gastrulation derives exclusively from a maternal pool of protein or whether additional β1-integrin is formed by new synthesis during that phase of development was investigated by metabolic labelling experiments.

Fertilized eggs were injected with L-[35S]methionine. At stages 6, 9 and 11, samples of 40 embryos were removed, washed and lysed. The total radioactivity in the three extracts was found to be similar. The radioactivity incorporated into total protein increased up to stage 9 and then levelled off (Fig. 4). The extracts were subjected to immunoprecipitation with mAb 8C8. Precipitates were separated by electrophoresis and labelling was visualized by fluorography.

Fig. 4.

Labelling of Xenopus embryos with L-[35S]methionine. Fertilized eggs were injected with L-|35S]methionine. At stage 6, 9, and 11 forty embryos each were lysed for immunoprecipitation (compare Fig. 5). Total radioactivity (left columns) and the fraction incorporated into protein (right columns) were determined for each lysate from aliquots. Numbers indicate the percentage of total label incorporated into protein.

Fig. 4.

Labelling of Xenopus embryos with L-[35S]methionine. Fertilized eggs were injected with L-|35S]methionine. At stage 6, 9, and 11 forty embryos each were lysed for immunoprecipitation (compare Fig. 5). Total radioactivity (left columns) and the fraction incorporated into protein (right columns) were determined for each lysate from aliquots. Numbers indicate the percentage of total label incorporated into protein.

As shown in Fig. 5, β1-integrin precursor was heavily labelled in all samples without significant change in the intensity of the signal. Labelling of the mature form was delayed. It was readily detectable at stage 9 of development; only after prolonged exposure of the gel could some label be seen as early as stage 6. The labelling of the mature form increased further between stages 9 and 11, in spite of the fact that additional incorporation of label into total protein was low at these stages, due to the exhaustion of the injected tracer. These observations indicate that β1-integrin precursor is synthesized in the embryo from early on and that precursor is processed into mature form.

Fig. 5 also reveals the presence of a labelled component that coprecipitates with β1-integrin. A double band with apparent Mrs of 140×103 and 150 ×103 at stage 6 becomes transformed into a single band with an apparent Mr of 155×103 from stage 9 onwards. This behaviour suggests the formation of a mature protein from a precursor as well, possibly of a β1-associated α-chain.

β1-integrin mRNA is present in the embryo from the egg stage onwards

The finding that β1-integrin is synthesized in the embryo before midblastula transition implies the presence of maternally provided β1-integrin mRNA in these stages. Probing samples of total RNA in a northern blot analysis with full-length antisense transcripts from the cloned β1*-integrin cDNA (see Fig. 1C) showed that mRNA sequences of β1-integrin are present in the egg and remain constant in amount throughout early development (Fig. 6A). The transcripts forming the upper band in the autoradiogram agree in length with the mRNA of β1-integrin described by DeSimone and Hynes (1988). The β1- and β1* -mRNA variants described by these authors are not resolved by the method used in our experiment (their putative position is indicated by arrows in Fig. 6A). In addition, an as yet unidentified RNA species of lower relative molecular mass hybridized with the RNA probe. The experiment has been repeated by using poly(A)+ RNA in an otherwise identical experiment with essentially the same result (data not shown).

Fig. 5.

β1-integrin is synthesized in the early cleavage embryo. The lysates of the labelled embryos (compare Fig. 4) were subjected to immunoprecipitation with mAb 8C8. Isolated proteins were electrophoresed under non-reducing conditions and detected by fluorography. Samples A, B and C are from embryos labelled from stage 1 until stages 6, 9 and 11, respectively. Immunoprecipitation with an inert IgG served as a control (D). The positions and sizes [Mr× 10−3] of relative molecular mass markers are indicated.

Fig. 5.

β1-integrin is synthesized in the early cleavage embryo. The lysates of the labelled embryos (compare Fig. 4) were subjected to immunoprecipitation with mAb 8C8. Isolated proteins were electrophoresed under non-reducing conditions and detected by fluorography. Samples A, B and C are from embryos labelled from stage 1 until stages 6, 9 and 11, respectively. Immunoprecipitation with an inert IgG served as a control (D). The positions and sizes [Mr× 10−3] of relative molecular mass markers are indicated.

Fig. 6.

β1-integrin mRNA is present throughout early development. (A) Total RNA was isolated from embryos of different stages. RNA from 1.5 embryo equivalents was applied to each lane for electrophoresis and blotting. Blots were hybridized with 32P-labelled antisense RNA transcribed from the full-length β1 * -integrin cDNA-clone (Fig. 1C). (B) Poly(A)+ RNA was isolated from embryos of different stages. RNA from 6 embryo equivalents was applied to each lane for electrophoresis and blotting. Blots were hybridized with 32P-labelled antisense RNA transcribed from the 3’ end of the β1*-integrin cDNA-clone as indicated in Fig. ID.

Fig. 6.

β1-integrin mRNA is present throughout early development. (A) Total RNA was isolated from embryos of different stages. RNA from 1.5 embryo equivalents was applied to each lane for electrophoresis and blotting. Blots were hybridized with 32P-labelled antisense RNA transcribed from the full-length β1 * -integrin cDNA-clone (Fig. 1C). (B) Poly(A)+ RNA was isolated from embryos of different stages. RNA from 6 embryo equivalents was applied to each lane for electrophoresis and blotting. Blots were hybridized with 32P-labelled antisense RNA transcribed from the 3’ end of the β1*-integrin cDNA-clone as indicated in Fig. ID.

Using poly(A)+ RNA for the blots and DNA probes from the untranslated 3’ends of β1- or β1*-integrin cDNA for hybridization, DeSimone and Hynes (1988) did not observe the presence of β1- or β1*-integrin mRNA until early gastrula. We have obtained a similar result (Fig. 6B) when we probed blots of poly(A)+ RNA with a stretch of antisense RNA transcribed from the 3’end of the β1-integrin clone as it is depicted in Fig 1C.

Distribution of β1-integrin within the embryo

A prerequisite for a putative function of β1-integrin before gastrulation would be the insertion of the mature protein into the cleavage membranes.

Immunohistology revealed that, starting with the first cleavage, β1-integrin is localizd along the newly formed plasma membranes of all cells (Fig. 7). We reproducibly observed that the cell periphery of early blastomeres showed much heavier labeling than that of later stages (compare Fig. 7A and B with Fig. 7D). During development, the antigen persists on the plasma membranes of all tissues and body regions. At around the end of neurulation β1-integrin disappears from the lateral plasma membranes of the outer epidermal layer (Fig. 7E).

Fig. 7.

Distribution of β1 -integrin in the embryo during early development. Embryos were whole-mount stained with mAb 8C8, embedded in glycolmetacrylate and sectioned. A-D are sections from stage 4, 7, 9 and 12 embryos, respectively. E represents a section through the head of a stage 30 embryo. The staining method applied yields sections of high quality but results in some unspecific autofluorescence of the yolk platelets. Bars represent 200 μm.

Fig. 7.

Distribution of β1 -integrin in the embryo during early development. Embryos were whole-mount stained with mAb 8C8, embedded in glycolmetacrylate and sectioned. A-D are sections from stage 4, 7, 9 and 12 embryos, respectively. E represents a section through the head of a stage 30 embryo. The staining method applied yields sections of high quality but results in some unspecific autofluorescence of the yolk platelets. Bars represent 200 μm.

Though a weak staining was observed in the cytoplasm of the uncleaved egg, the plasma membrane is devoid of the antigen. This absence of β1-integrin of the egg membrane is transmitted to all plasma membrane domains on the surface of the embryo that derive from the egg periphery. Therefore the early blastomeres exhibit a conspicuous polar distribution of the antigen. This polarity is maintained by the cells of the outer layer throughout early development (Fig. 7A-C). In the postgastrula stages, it may still be recognized in the archenteron roof, which is derived from the outer cell layer of the pregastrula embryo (Fig. 7D).

β1-integrin is inserted into the plasma membranes of the early blastomeres

The staining of the cell boundaries suggests the presence of β1-integrin on all internal plasma membranes of the early embryo. However, the microscopic resolution is insufficient to distinguish the plasma membranes from the underlying cortices. Thus β1 integrin might accumulate only in the cortices of the blastomeres and might not become inserted into the plasma membranes.

To investigate this question a surface labelling method was applied: the succinimide ester biotin-XNHS, added to intact cells, labels only protein domains outside the plasma membrane. Internal proteins are protected from labelling by the plasma membrane which is impermeable to the reagent (von Boxberg, 1990).

Blastomeres from dissociated stage 6 embryos were treated with biotin-XNHS and labelled proteins were isolated with streptavidin-agarose and subjected to immunoblot analysis. Fig. 8C shows that mature β1-integrin was present in the biotinylated fraction.

Fig. 8.

β1-integrin is present on the plasma membranes of cleavage embryos. Cell surface proteins of blastomeres from stage 6 embryos were labelled with biotin and biotinylated proteins were precipitated with streptavidine-agarose (C) From the supernatant non-biotinylated β1-integrin was isolated by immunoprecipitation with mAb 8C8 (B) and compared with immunoprecipitates from extracts of non-biotinylated embryos (A). Proteins were separated by SDS-PAGE under non-reducing conditions and immunoblotted with mAb 8C8. For A and B extracts of 5 embryos were applied, for C isolated proteins from about 30 embryos were used. The positions and sizes [Mr × 10−3] of relative molecular mass markers are indicated.

Fig. 8.

β1-integrin is present on the plasma membranes of cleavage embryos. Cell surface proteins of blastomeres from stage 6 embryos were labelled with biotin and biotinylated proteins were precipitated with streptavidine-agarose (C) From the supernatant non-biotinylated β1-integrin was isolated by immunoprecipitation with mAb 8C8 (B) and compared with immunoprecipitates from extracts of non-biotinylated embryos (A). Proteins were separated by SDS-PAGE under non-reducing conditions and immunoblotted with mAb 8C8. For A and B extracts of 5 embryos were applied, for C isolated proteins from about 30 embryos were used. The positions and sizes [Mr × 10−3] of relative molecular mass markers are indicated.

This result indicates the presence of β1 -integrin in its mature form on the cell surface of the early blastomeres. At the same time, it provides a control for the selectivity of the method since the precursor form, which under physiological conditions is known to be intracellular (Akiyama et al., 1989), was not found in the biotin-labelled fraction.

Mature form of β1-integrin was also present in abundant amounts in non-biotinylated protein fraction (Fig. 8B). Assuming that the mature form on the plasma membrane was fully biotinylated this observation provides evidence for the existence of a large intracellular pool of mature β1 -integrin.

mAb 8C8 was applied in studies on β1 -integrin during early development of Xenopus laevis. Several fines of evidence indicate that this antibody specifically recognizes the precursor and the mature form of β1 -integrin. (1) The antigens that immunoprecipitate with mAb 8C8 are recognized by two different polyclonal antibodies directed against β1 -integrin epitopes. (2) Upon reduction the two components recognized by mAb 8C8 respond with changes in their electrophoretic mobility that are typical of β1 integrins (Hynes, 1987). (3) The apparent Mt of the two antigenic components recognized by mAb 8C8 are close to those of the mammam-lian precursor and mature form of β1 -integrin (Akiyama et al., 1989). (4) The two antigenic components respond to treatment with endoglycosidases in a way diagnostic for the β1 -integrin precursor and mature form (Jaspers et al., 1988). (5) The 8C8 antigen is concentrated in adhesion plaques of Xenopus tissue culture cells. This subcellular distribution is characteristic of integrins (Burridge et al., 1988).

The data collected by applying mAb 8C8 in studies on the early embryo lead to the following conclusions. (1) β1 -integrin is already present in the egg and continues to be present in the early embryo with no significant changes in quantity up to the early gastrula stage. From then on, the amount of β1 -integrin in the embryo increases gradually. The proportion of precursor and mature form does not show significant changes throughout the pregastrula phase, although a slight increase in the amount of the mature form can sometimes be observed. (2) β1integrin is continuously synthesized during development including the pregastrula phase. The mRNA for this synthesis is provided maternally. During early development the precursor is slowly but continuously transformed into the mature form. (3) Only the mature form of β-integrin is inserted into the plasma membranes. Up to the neurula stages, β-integrin is present on the plasma membranes of all cells. It is restricted to the plasma membrane domains inside the embryo, whereas membrane domains on the surface of the embryo remain devoid of the antigen.

These findings will be discussed in the light of previous reports on integrins in the early amphibian embryo and of their relevance for more general questions of amphibian embryogenesis.

Insertion of β1-integrin into the cleavage membranes

During cleavage the total area of plasma membranes within the embryo increases more than 100-fold (Servetnick et al., 1990) and all new plasma membranes are endowed with mature β1-integrin. If the β1-integrin on the new plasma membranes is due to new synthesis to any significant degree, an increase in the total amount would be observed. Furthermore, if the precursor formed a pool from which the mature β1integrin was continuously recruited onto the plasma membranes, an accumulation of the mature form should be observed. Both of these models can be dismissed on the basis of our experimental data. In relation to the vast increase of plasma membrane area neither the total amount of β1-integrin nor the amount of the mature form in particular increase to a significant degree until the late blastula stage.

These considerations lead to the conclusion that an intracellular pool of mature β1-integrin sufficient to furnish all plasma membranes until late blastula stages is already present in the egg. This conclusion is further supported by the findings that during the first eight hours of cleavage hardly any of the metabolically labelled precursor was processed to the mature form (Fig. 5) and that an appreciable amount of the mature β1-integrin has to be regarded as intracellular as it did not become labelled when the surface of isolated early blastomeres was treated with biotinsuccinimide ester (Fig. 8).

Based on electron microscope observations, Singal and Sanders (1974) have proposed that plasma membrane formation during cleavage occurs by the fusion of preformed membrane vesicles. We suggest that the pool of mature β1-integrin is included in these vesicles.

Our immunostainings revealed that the cell periphery of early blastomeres was much heavier labelled than that of later stages. Whether this indicates that part of the intracellular pool of β1-integrin becomes associated with the submembrane cortical region of the early blastomeres or whether in the early stages more β1integrin is inserted into a given area of plasma membranes than later on remains to be elucidated.

The blastomeres of the outer cell layer of the embryo display a strong polarity of β1-integrin expression. The protein is only present on the basolateral plasma membrane domains, whereas the apical domains at the surface of the embryo, which are derived from the egg plasma membrane, remain free of β1-integrin. U-cadherin, another membrane protein, exhibits the same distribution pattern (Angres et al., 1990). This protein has been shown to be removed by internalisation from the oocyte membrane during oocyte maturation (Millier et al., 1992). Preliminary results indicate that the observations reported there also hold true for β1integrin. The indication that during oocyte maturation the two membrane proteins behave the same way suggests a general mechanism for the generation of the apical properties of the outer embryonic membranes (A.H.J. Müller, V. Gawantka and P. Hausen, unpublished data).

Synthesis of β1-integrin in the early embryo

Though the maternal pool of β1-integrin seems to be sufficient to supply the plasma membranes formed during cleavage, the metabolic labelling experiment indicates continuous synthesis of this protein (Fig. 5). The long period that it takes until an appreciable amount of label is found in the mature form agrees with the data on mammalian cells (Akiyama and Yamada, 1987; De Strooper et al., 1991).

The finding that β1-integrin is continuously synthesized in the early embryo disagrees with the observation of DeSimone and Hynes (1988). Using the untranslated 3’ends of β1- and β1*-integrin to probe northern blots of Xenopus poly(A)+ RNA, the authors found that β1-integrin mRNA was not present in the embryo until the early gastrula stage. We have been able to reproduce this result (Fig. 6B) by using a small piece of the untranslated 3’sequence depicted in Fig. ID. However, using antisense transcripts from the complete cDNA-clone (Fig. 1C) for hybridisation, we found the continuous presence of β1-integrin mRNA (Fig. 6A). The high stringency applied in the blotting technique and the fact that RNA species of the expected length were detected preclude an unspecific cross-reaction in this experiment.

The discrepancy of the two results could be resolved by assumimg that more than two β1-integrin mRNAs with different 3’-untranslated stretches are present in the embryo and that those supplied by the maternal pool have escaped DeSimone and Hynes’ attention. Our observation of a further mRNA with a relative molecular mass lower than that expected from DeSimone’s and Hynes’ data supports this notion. A more detailed analysis of integrin mRNA in the egg and the early Xenopus embryo is required to substantiate these suggestions.

Does β1-integrin have a function in the early embryo?

The presence of β1-integrin on the plasma membranes of late blastula and gastrula cells complements the reports of integrin function commencing at these stages as they are summarized in the introduction. The ubiquitous abundance of β1-integrin on the plasma membranes of all blastomeres of the earlier stages suggests that it is neither the synthesis of the β1-integrin chain nor a commencing insertion into the plasma membrane that initiates the observed functions. Indeed Smith et al. (1990) found that upon treatment with XTC-MIF animal cap cells begin to display integrin-dependent cell spreading on low concentrations of fibronectin without significant changes in -integrin synthesis. However, in immunoprecipitates with anti-β1-integrin antibodies, the authors observed the appearance of new protein components after induction. These components may be interpreted as β1-integrin-associated β1-chains, but comparing their data with ours an enhanced formation of mature β1-chains from the precursor seems to be a more likely interpretation.

The presence of β1integrin on the plasma membranes before midblastula transition had not been expected from our present knowledge on integrin-dependent events in the embryo. Information on the function of β1-integrin during these early stages may be obtained by the identification of possible ligand molecules. Type VI collagen has recently been postulated to occur in high amounts in the earliest stages of Xenopus embryogenesis and to be secreted into the extracellular space (Otte et al., 1990). Fibronectin has also been found in the embryo from the egg stage on and it is secreted by all cells of the embryo (Lee et al., 1984; R. Winklbauer, personal communication). No indication of an early function of these ECM components, which might interact with β1-integrin, is as yet available.

The identification of integrin α-chains, which are presumed to be present in the early embryo, might also provide hints at an early β1-integrin function. In the experiments on the early synthesis of β1-integrin, two components with an apparent Mr of 140 ×103 and 150×103 coprecipitated with β1-integrin. At later stages, this material was substituted by a single component with an apparent Mr of 155 × 103. It is likely that these components represent the precursor and the mature form of an integrin α-subunit that associates with β1-integrin, but this supposition needs to be substantiated. We have recently established hybridoma lines that produce monoclonal antibodies against a further component that coprecipitates with Xenopus β1- integrin. The physical characteristics qualify the antigen as an integrin β1-chain. This antigen is present in the egg and it remains constant in amount throughout the pregastrula phase. As the labelling experiment did not reveal this chain, we suppose that it might be provided maternally. A closer characterization of these presumed tr-chains might give some clue on the function of β1-integrins in early Xenopus embryogenesis.

We thank Ursula Müller for excellent technical assistance, Dr D. Melton for the Xenopus neurula cDNA library, Dr R. Pytela for the human β1cDNA clone, Dr J. Smith for XTC-cells and Dr R.O. Hynes for the anti-β1integrin peptide antibody. We would also like to thank Dr Rudi Winklbauer, Amo Müller and Metta Riebesell for helpful discussions and critical reading of the manuscript.

Akiyama
,
S. K.
,
Nagata
,
K.
and
Yamada
,
K. M.
(
1990
).
Cell surface receptors of extracellular matrix components
.
Biochim. Biophys. Acta
.
1031
,
91
110
.
Akiyama
,
S. K.
and
Yamada
,
K. M.
(
1987
).
Biosynthesis and acquisition of biological activity of the fibronectin receptor
.
J. Biol. Chem
.
262
,
17536
17542
.
Akiyama
,
S. K.
,
Yamada
,
S. S.
and
Yamada
,
K. M.
(
1989
).
Analysis of the role of glycosylation of the human fibronection receptor
.
J. Biol. Chem
.
264
,
18011
18018
.
Albelda
,
S. M.
and
Buck
,
C. A.
(
1990
).
Integrins and other cell adhesion molecules
.
FASEB J
.
4
,
2868
2880
.
Angres
,
B.
,
Müller
,
A. H. J.
,
Kellermann
,
J.
and
Hausen
,
P.
(
1991
).
Differential expression of two cadhenns in Xenopus laevis
.
Development
111
,
829
844
.
Argraves
,
W. S.
,
Suzuki
,
S.
,
Aral
,
H.
,
Thompson
,
K.
,
Pierschbacher
,
M. D.
and
Ruoslahti
,
E.
(
1987
).
Aminoacid sequence of the human fibronectin receptor
.
J. Cell Biol
.
105
,
1183
1190
.
Boucaut
,
J. C.
,
Darribere
,
T.
,
Poole
,
T. J.
,
Aoyama
,
H.
,
Yamada
,
K. M.
and
Thiery
,
J. P.
(
1984
).
Biologically active synthetic peptides as probes of embryological development: a competitve peptide inhibitor of fibronectin function inhibits gastrulation in amphibian embryos and neural crest cell migration in avian embryos
.
J. Cell Biol
.
99
,
1822
1830
.
Burridge
,
K.
,
Fath
,
K.
,
Nuckols
,
G.
and
Turner
,
C.
(
1988
).
Focal adhesions: Transmembrane junctions between the extracellular matrix and the cytoskeleton
.
Ann. Rev. Cell Biol
.
4
,
487
525
.
Campanero
,
M. R.
,
Pálido
,
R.
,
Ursa
,
M. R.
,
Rodríguez-Moya
,
M.
,
De Landazuri
,
M. O.
and
Sanchez-Madrid
,
F.
(
1990
).
Alternative leucocyte homotypic adhesion mechanism LFA1/ICAM-1-independent triggered through the human VLA-4 integrin
.
J. Biol. Chem
.
110
,
2157
2165
.
Cheresh
,
D. A.
and
Spiro
,
R. C.
(
1987
).
Biosynthetic and functional properties of an Arg-Gly-Asp-directed receptor involved in human melanoma cell attachment to vitronectin, fibronectin and von Willebrand factor
.
J. Biol. Chem
.
262
,
1703
1711
.
Darribere
,
T.
,
Guida
,
K.
,
Larjava
,
H.
,
Johnson
,
K.
,
Yamada
,
K. M.
and
Thiery
,
J. P.
(
1990
).
In vivo analysis of integrin subunit function in fibronectin matrix assembly
.
J. Cell Biol
.
110
,
1813
1823
.
Darribere
,
T.
,
Yamada
,
K. M.
,
Johnson
,
K. E.
and
Boucant
,
J. C.
(
1988
).
The 140-kDa fibronectin receptor complex is required for mesodermal cell adhesion during gastrulation in the amphibian Pleurodeles waltlii
.
Dev. Biol.
126
,
182
194
.
DeSimone
,
D. W.
and
Hynes
,
R. O.
(
1988
).
Xenopus laevis integrins. Structural conservation and evolutionary divergence of integnn /J subunits
.
J. Biol. Chem
.
263
,
5333
5340
.
De Strooper
,
B.
,
Van Leuven
,
F.
,
Carmeliet
,
G.
,
Van den Berghe
,
H.
and
Cassiman
,
J.
(
1991
).
Cultered human fibroblasts contain a large pool of precursor β1-integrin but lack an intracellular pool of mature subunit
.
Europ. Joum. Biochem
.
199
,
25
33
.
Ellinger-Ziegelbaner
,
H.
and
Dreyer
,
C.
(
1991
).
A retinoic acid receptor expressed in the early development of Xenopus laevis
.
Gen. Dev
.
5
,
94
104
.
Fey
,
J
, and
Hausen
,
P.
(
1990
).
Appearance and distribution of laminin during development of Xenopus laevis
.
Dtff
.
42
,
144
152
.
Galfre
,
G.
and
Milstein
,
C.
(
1981
).
Preparation of monoclonal antibodies: Strategies and procedures
.
Meth. Enzym
.
73
,
3
46
.
Harlow
,
E.
and
Lane
,
D.
(
1988
).
Antibodies: a laboratory manual
.
Cold Spring Harbour Laboratory
.
Hynes
,
R. O.
(
1987
).
Integrins: a family of cell surface receptors
.
Cell
48
,
549
554
.
Jaspers
,
M.
,
De Strooper
,
B.
,
Spaepen
,
M.
,
Van Lenven
,
F.
,
David
,
G.
,
Van den Berghe
,
H.
and
Cassiman
,
J-J.
(
1988
).
Post-translational modification of the -subunit of the human fironectin receptor
.
FEBS Letters
231
,
402
406
.
Johnson
,
K. E.
and
Silver
,
M. H.
(
1986
).
Cells from Xenopus laevis gastrulae adhere to fibronectin-sepharose beads and other lectin coated beads
.
Scanning Electron Microscopy
2
,
671
687
.
Kearney
,
J. F.
,
Radbruch
,
A.
,
Liesegang
,
B.
and
Rajewski
,
K.
(
1979
).
A new mouse myeloma cell line that has lost immunoglobulin expression but permits the construction of antibody-secreting hybrid cell lines
.
J. Immunol
.
123
,
1548
1550
.
Keller
,
R.
and
Winklbauer
,
R.
(
1991
).
The role of the extracellular matrix in amphibian gastrulation
.
Sem. Dev. Biol.
25
,
33
.
Kintner
,
C. R.
and
Melton
,
D. R.
(
1987
).
Expression of Xenopus N-CAM RNA in ectoderm is an early response to neural induction
.
Development
99
,
311
325
.
Kishimoto
,
T. K.
,
Larson
,
R. S.
,
Corbi
,
A. L.
,
Dustin
,
M. L.
,
Staunton
,
D. E
, and
Springer
,
T. A.
(
1989
).
The leucocyte integrins
.
Adv. Immunol
.
46
,
149
182
.
Kornberg
,
L. J.
,
Shelton Eaerp
,
H.
,
Turner
,
C. E.
,
Prockop
,
C.
and
Juliano
,
R. L.
(
1991
).
Signal transduction by integrins: Increased protein tyrosine phosphorylation caused by clustering pl integrins
.
Proc. Natl. Acad. Sci. USA
88
,
8392
8396
.
Laemmli
,
U. K.
(
1970
).
Cleavage of structural proteins during the assembly of the head of bacteriophage T4
.
Nature
227
,
680
689
.
Lee
,
G.
,
Hynes
,
R.
and
Kirschner
,
M.
(
1984
).
Temporal and spatial regulation of fibronectin in early Xenopus development
.
Cell
36
,
729
740
.
Marcantonlo
,
E. E.
and
Hynes
,
R. O.
(
1988
).
Antibodies to the conserved cytoplasmic domain of the integrin pl-subunit react with proteins in vertebrates,invertebrates and fungi
.
J. Cell Biol
.
106
,
1765
1772
.
Morrissey
,
J. H.
(
1981
).
Silver stain for proteins in polyacrylamide gels: a modified procedure with enhanced uniform sensitivity
.
Anal. Biochem
.
117
,
307
310
.
Müller
,
A. H. J.
,
Angres
,
B.
and
Hausen
,
P.
(
1992
).
U-cadherin in Xenopus oogenesis and oocyte maturation
.
Development in press
.
Nieuwkoop
,
P. D.
and
Faber
,
J.
(
1967
).
Normal Table of Xenopus laevis (Daudin)
,
Amsterdam
:
North Holland. Ed. 2nd
.
Otte
,
A. P.
,
Roy
,
D.
,
Siemerink
,
M.
,
Koster
,
C. H.
,
Hochstenbach
,
F.
,
Timmermans
,
A.
and
Durston
,
A. J.
(
1990
).
Characterization of a maternal type VI collagen in Xenopus embryos suggests a role for collagen in gastrulation
.
Cell Biol
.
111
,
271
278
.
Pierschbacher
,
M. D.
and
Ruoslahti
,
E.
(
1984
).
Cell attachment activity of fibronectin can be duplicated by small synthetic fragments of the molecule
.
Nature
309
,
30
33
.
Ruoslahti
,
E.
(
1991
).
Integrins
.
J. Clin. Invest
.
87
,
1
5
.
Ruoslathi
,
E.
and
Pierschbacher
,
M. D.
(
1987
).
New perspectives in cell adhesion: RGD and integrins
.
Science
238
,
491
497
.
Rüther
,
U.
and
Mflller-Hill
,
B.
(
1983
).
Easy identification of cDNA clones
.
EMBO J
.
2
,
1791
1794
.
Sanger
,
F.
,
Nicklen
,
S.
and
Coulson
,
A. R.
(
1977
).
DNA sequencing with chain-terminating inhibitors
.
Proc. Natl. Acad. Sci. USA
74
,
5463
5467
.
Servetnick
,
M.
,
Schnlte-Merker
,
S.
and
Hausen
,
P.
(
1990
).
Cell surface proteins during early Xenopus development: analysis of cell surface proteins and total glycoproteins provides evidence for a maternal pool
.
Roux’s Arch. Dev. Biol
.
198
,
433
442
.
Singal
,
P. K.
and
Sanders
,
E. J.
(
1974
).
An ultrastructural study of the first cleavage of Xenopus embryos
.
Ultrastr. Res
.
47
,
433
451
.
Smith
,
J. C.
,
Symes
,
K.
,
Hynes
,
R. O.
and
DeSimone
,
D.
(
1990
).
Mesoderm induction and the control of gastrulation in Xenopus laevis: the roles of fibronectin and integrin
.
Development
108
,
229
238
.
von Boxberg
,
Y.
,
Wütz
,
R.
and
Schwarz
,
U.
(
1990
).
Use of the biotinavidin system for labelling, isolation and characterization of neural cell surface proteins
.
Europ. J. Biochem
.
190
,
249
256
.
Winklbauer
,
R.
(
1988
).
Differential interaction of Xenopus embryonic cells with fibronectin in vitro
.
Dev. Biol
.
130
,
175
183
.
Winklbauer
,
R.
(
1990
).
Mesodermal cell migration during Xenopus gastrulation
.
Dev. Biol
.
142
,
155
168
.
Winklbauer
,
R.
and
Nagel
,
M.
(
1991
).
Directional cell migration during xenopus gastrulation
.
Dev. Biol
.
148
,
573
589
.
Winklbauer
,
R.
,
Seichow
,
A.
,
Nagel
,
M.
,
Stolz
,
C.
and
Angres
,
B.
(
1992
).
Mesodermal cell migration in the gastrula
.
In: Gastrulation: Movements, Patterns and Molecules
, edited by
Keller
,
R. E.
,
Clark
,
W.
and
Griffin
,
F.
New York
:
Plenum Press
.