β1- and β3-Class Integrins Mediate Fibronectin Binding Activity at the Surface of Developing Mouse Peri-implantation Blastocysts
1995; Elsevier BV; Volume: 270; Issue: 19 Linguagem: Inglês
10.1074/jbc.270.19.11522
ISSN1083-351X
AutoresJeffery F. Schultz, D. Randall Armantt,
Tópico(s)Reproductive System and Pregnancy
ResumoImplanting mouse blastocysts adhere through their abembryonic surfaces to the endometrial extraceUular matrix. Because blastocysts cultured on fibronectin in vitro dissociate to form trophoblast outgrowths, it is unclear whether this adhesion is initially mediated by fibronectin receptors on the apical or basolateral surface of the trophectoderm. Intact blastocysts were examined in a ligand binding assay utilizing the fibronectin cell binding domain attached to fluorescent microspheres. Fibronectin binding activity on the apical surface of the trophectoderm was confined to the abembryonic pole of the blastocyst, where trophoblast differentiation initiates, and was regulated temporally in accordance with blastocyst outgrowth. Soluble fibronectin (IC50, = 0.2 µm) or Gly-Arg-Gly-Asp-Ser-Pro, but not laminin, competitively inhibited fibronectin binding activity. Addition of antibodies against the αv, α5, β1, or β3 integrin subunits also inhibited binding activity. Blastocysts cultured in the absence of an adhesive substratum exhibited fibronectin binding activity only after exposure to immobilized or soluble ligand. Potentiation of binding activity by ligand was unaffected by cycloheximide but was sensitive to brefeldin A inhibition of protein trafficking. These findings suggest that the interaction of fibronectin with the trophectoderm induces a translocation event that up-regulates fibronectin binding β1- and β3-class integrins on the apical surface. Implanting mouse blastocysts adhere through their abembryonic surfaces to the endometrial extraceUular matrix. Because blastocysts cultured on fibronectin in vitro dissociate to form trophoblast outgrowths, it is unclear whether this adhesion is initially mediated by fibronectin receptors on the apical or basolateral surface of the trophectoderm. Intact blastocysts were examined in a ligand binding assay utilizing the fibronectin cell binding domain attached to fluorescent microspheres. Fibronectin binding activity on the apical surface of the trophectoderm was confined to the abembryonic pole of the blastocyst, where trophoblast differentiation initiates, and was regulated temporally in accordance with blastocyst outgrowth. Soluble fibronectin (IC50, = 0.2 µm) or Gly-Arg-Gly-Asp-Ser-Pro, but not laminin, competitively inhibited fibronectin binding activity. Addition of antibodies against the αv, α5, β1, or β3 integrin subunits also inhibited binding activity. Blastocysts cultured in the absence of an adhesive substratum exhibited fibronectin binding activity only after exposure to immobilized or soluble ligand. Potentiation of binding activity by ligand was unaffected by cycloheximide but was sensitive to brefeldin A inhibition of protein trafficking. These findings suggest that the interaction of fibronectin with the trophectoderm induces a translocation event that up-regulates fibronectin binding β1- and β3-class integrins on the apical surface. Interstitial implantation in humans and mice comprises a complex interaction between the primary trophoblast cells of the differentiating blastocyst and the endometrial ECM 1The abbreviations used are: ECM, extracellular matrix; FN-120, 120-kDa fibronectin central cell binding fragment; FN-50, 50-kDa NH3-terminal portion of FN-120 lacking the RGD site; FN-45, 45-kDa gelatin binding fragment; FN-40, 40-kDa high affinity heparin binding fragment; α1-AG, α1-acid glycoprotein; hPBS, high phosphate saline solution; BSA, bovine serum albumin; PBS, phosphate-buffered saline.. At the onset of implantation, the mural trophectoderm of the blastocyst comes into intimate contact with the uterine epithelial cells that eventually slough away, allowing the trophoblast cells to contact the basement membrane, penetrate it, and invade the underlying stroma (Schlafke and Enders, 1975Schlafke S. Enders A.E. Biol. Reprod. 1975; 12: 41-65Crossref PubMed Scopus (362) Google Scholar; Chavez, 1984Chavez D.J. Van Blerkom J. Motta P.M. Ultrastructure of Reproduction. Martinus Nijkoff, Boston1984: 247-259Crossref Google Scholar). Studies investigating cell adhesion and migration using the blastocyst outgrowth assay (Gwatkin, 1966Gwatkin R.B. Ann. N. Y. Acad. Sci. 1966; 139: 70-90Crossref Scopus (48) Google Scholar) have demonstrated that many ECM glycoproteins found in the uterine endometrium, including fibronectin, laminin, vitronectin, collagen, and entactin (Wewer et al., 1986Wewer U.M. Damjanov A. Weiss J. Liotta L.A. Damjanov I. Differentiation. 1986; 32: 49-58Crossref PubMed Scopus (104) Google Scholar, support trophoblast adhesion (Armant et al., 1986aArmant D.R. Kaplan H.A. Lennarz W.J. Deo. Biol. 1986; 116: 519-523Crossref PubMed Scopus (162) Google Scholar, Armant et al., 1986bArmant D.R. Kaplan H.A. Mover H. Lennarz W.J. Proc. Natl. Acad. Sci. U. S. A. 1986; 83: 6751-6755Crossref PubMed Scopus (92) Google Scholar; Sutherland et al., 1988Sutherland A.E. Calarco P.G. Damsky C.H. J. Cell Biol. 1988; 106: 1331-1348Crossref PubMed Scopus (125) Google Scholar; Carson et al., 1988Carson D.D. Tang J.P. Gay S. Dev. Biol. 1988; 127: 368-375Crossref PubMed Scopus (55) Google Scholar; Yelian et al., 1993Yelian F.D. Edgeworth N.E. Dong L.J. Chung A.E. Armant D.R. J. Cell Biol. 1993; 121: 923-929Crossref PubMed Scopus (81) Google Scholar. The outgrowth of primary trophoblast cells from cultured blastocysts is believed to reflect differentiation of the embryo, leading to trophoblast invasion of the endometrial stroma during implantation in utero (Sherman, 1975Sherman M.I. Cell. 1975; 5: 343-349Abstract Full Text PDF PubMed Scopus (52) Google Scholar; Glass et al., 1979Glass R.H. Spindle A.I. Pedersen R.A. J. Exp. Zool. 1979; 208: 327-336Crossref PubMed Scopus (41) Google Scholar; Enders et al., 1981Enders A.E. Chavez D.J. Schlafke S. Glasser S.R. Bullock D.W. Cellular and Molecular Aspects of Implantation. Plenium Press, New York1981: 365-382Crossref Google Scholar. Blastocyst adhesiveness to fibronectin and other ECM proteins is coordinately regulated through the embryonic developmental program (Sherman and Ateinza-Somols, 1978Sherman M.I. Ateinza-Somols S. Ludwig H. Tauber P.F. Human Fertilization. Thieme, Stuttgart, Germany1978: 179-183Google Scholar; Armant et al., 1986aArmant D.R. Kaplan H.A. Lennarz W.J. Deo. Biol. 1986; 116: 519-523Crossref PubMed Scopus (162) Google Scholar by a mechanism that remains to be elucidated. Fibronectin is a multi-functional ECM glycoprotein that plays a central role in cell adhesion and is thought to regulate cellular polarity, differentiation, and growth (Ruoslahti, 1988Ruoslahti E. Annu. Rev. Biochem. 1988; 57: 375-413Crossref PubMed Scopus (1059) Google Scholar). Increased fibronectin synthesis accompanies ECM remodeling within the uterine endometrium in preparation for implantation (Rider et al., 1992Rider V. Carlone D.L. Witrock D. Cai C. Oliver N. Dev. Dyn. 1992; 195: 1-14Crossref PubMed Scopus (30) Google Scholar. Fibronectin can be cleaved by controlled proteolysis into several functional domains that bind fibrin, heparin, collagen, DNA, and various cell surface receptors (Pierschbacher et al., 1981Pierschbacher M.D. Hayman E.G. Ruoslahti E. Cell. 1981; 26: 259-267Abstract Full Text PDF PubMed Scopus (282) Google Scholar; Ruoslahti, 1988Ruoslahti E. Annu. Rev. Biochem. 1988; 57: 375-413Crossref PubMed Scopus (1059) Google Scholar). A fibronectin receptor has been identified in several cell lines by affinity chromatography on a 120-kDa fragment of fibronectin (FN-120) containing the central cell binding domain and specific elution with a synthetic peptide containing an Arg-Gly-Asp (RGD) sequence (Pytela et al., 1985Pytela R. Pierschbacher M.D. Ruoslahti E. Call. 1985; 40: 191-198Scopus (688) Google Scholar; Cardarelli and Pierschbacher, 1987Cardarelli P.M. Pierschbacher M.D. J. Cell Biol. 1987; 105: 499-506Crossref PubMed Scopus (40) Google Scholar). This fibronectin receptor is now known to be a member of the integrin family of integral membrane proteins (Ruoslahti, 1988Ruoslahti E. Annu. Rev. Biochem. 1988; 57: 375-413Crossref PubMed Scopus (1059) Google Scholar; Hynes, 1992Hynes R.O. Cell. 1992; 69: 11-25Abstract Full Text PDF PubMed Scopus (9026) Google Scholar). The integrins are a family of about 20 heterodimeric glycoprotein receptors that consist of a 120–180-kDa α-subunit and a 90–110-kDa β-subunit. In addition to mediating cellular adhesion to ECM components, integrins also play important roles in cell-cell adhesion, the immune response, cell differentiation, thrombosis, and inflammation (Hynes, 1992Hynes R.O. Cell. 1992; 69: 11-25Abstract Full Text PDF PubMed Scopus (9026) Google Scholar). Several integrins bind fibronectin through the RGD site in the central cell binding domain or through non-RGD sites located in an alternatively spliced IIICS domain (Hynes, 1992Hynes R.O. Cell. 1992; 69: 11-25Abstract Full Text PDF PubMed Scopus (9026) Google Scholar). Mouse primary trophoblast cells appear to interact with fibronectin exclusively through the RGD recognition site (Armant et al., 1986bArmant D.R. Kaplan H.A. Mover H. Lennarz W.J. Proc. Natl. Acad. Sci. U. S. A. 1986; 83: 6751-6755Crossref PubMed Scopus (92) Google Scholar; Yelian et al., 1995Yelian F.D. Yang Y. Hirata J.D. Schultz J.F. Armant D.R. Mol. Reprod. Dev. 1995; (in press)PubMed Google Scholar, indicating that the trophoblast fibronectin receptor may, in fact, be an integrin similar to that isolated by Pytela et al., 1985Pytela R. Pierschbacher M.D. Ruoslahti E. Call. 1985; 40: 191-198Scopus (688) Google Scholar. Trophoblast outgrowth is also mediated by RGD sequences in vitronectin (Armant et al., 1986bArmant D.R. Kaplan H.A. Mover H. Lennarz W.J. Proc. Natl. Acad. Sci. U. S. A. 1986; 83: 6751-6755Crossref PubMed Scopus (92) Google Scholar, collagen (Sutherland et al., 1988Sutherland A.E. Calarco P.G. Damsky C.H. J. Cell Biol. 1988; 106: 1331-1348Crossref PubMed Scopus (125) Google Scholar; Carson et al., 1988Carson D.D. Tang J.P. Gay S. Dev. Biol. 1988; 127: 368-375Crossref PubMed Scopus (55) Google Scholar, and entactin (Yelian et al., 1993Yelian F.D. Edgeworth N.E. Dong L.J. Chung A.E. Armant D.R. J. Cell Biol. 1993; 121: 923-929Crossref PubMed Scopus (81) Google Scholar but not in intact mouse laminin (Armant et al., 1986bArmant D.R. Kaplan H.A. Mover H. Lennarz W.J. Proc. Natl. Acad. Sci. U. S. A. 1986; 83: 6751-6755Crossref PubMed Scopus (92) Google Scholar; Armant, 1991Armant D.R. Biol. Reprod. 1991; 45: 664-672Crossref PubMed Scopus (42) Google Scholar; Yelian et al., 1993Yelian F.D. Edgeworth N.E. Dong L.J. Chung A.E. Armant D.R. J. Cell Biol. 1993; 121: 923-929Crossref PubMed Scopus (81) Google Scholar. The trophectoderm of the blastocyst appears to be a typical polarized epithelium (Ducibella et al., 1975Ducibella T. Albertini D.F. Anderson E. Biggers J.D. Dev. Biol. 1975; 45: 231-250Crossref PubMed Scopus (230) Google Scholar; Wiley et al., 1990Wiley L.M. Kidder G.M. Watson A.J. BioEssays. 1990; 12: 67-73Crossref PubMed Scopus (55) Google Scholar that rests on a basement membrane oriented toward the blas-tocoele (Nadijcka and Hillman, 1974Nadijcka M. Hillman N. J. Embryol. Exp. Morphol. 1974; 32: 675-695PubMed Google Scholar; Thorsteinsdottir, 1992Thorsteinsdottir S. Anat. Rec. 1992; 232: 141-149Crossref PubMed Scopus (52) Google Scholar). Less typical of epithelial cells is the role of the apical trophectoderm surface in adhesion to ECM. ECM receptors, such as the integrins α5β1 and α6β1, are restricted to the basolateral surface, although αvβ3 has been observed at the apical surface of mouse blastocysts (Sutherland et al., 1993Sutherland A.E. Calarco P.G. Damsky C.H. Development. 1993; 119: 1175-1186Crossref PubMed Google Scholar. In vivo, after apoptosis of the epithelial cells, the blastocyst contacts the surrounding uterine basement membrane through the apical surfaces of its trophoblast cells and adheres to a matrix that, unlike a culture plate, is neither inflexible nor flat (Schlafke and Enders, 1975Schlafke S. Enders A.E. Biol. Reprod. 1975; 12: 41-65Crossref PubMed Scopus (362) Google Scholar). The dissociation and outgrowth of trophoblast giant cells in vitro is therefore morphologically abnormal. It is unclear whether primary trophoblast adhesion in vitro is initially mediated by receptors on the apical surface of adhesion-competent blastocysts or by receptors at the basolateral surface that gain access to the culture plate as the blastocyst dissociates. To address this question, we have developed a novel ligand binding assay to quantify and localize fibronectin binding activity on the apical surface of differentiating trophoblast cells comprising the intact blastocyst. Fibronectin binding activity was characterized biochemically during blastocyst differentiation, focusing on the central cell binding domain represented by FN-120. We now report that fibronectin binding activity is both temporally and spatially regulated by integrins on the apical surface of cells comprising the trophectoderm of the developing peri-implantation blastocyst. Proteolytic fragments of fibronectin, including the 120-kDa central cell binding fragment (FN-120), the 50-kDa NH3-terminal portion of FN-120 lacking the RGD site (FN-50), the 45-kDa gelatin binding fragment (FN-45), and the 40-kDa high affinity heparin binding fragment (FN-40), were all purchased from Life Technologies, Inc. and have been previously characterized (Pierschbacher et al., 1981Pierschbacher M.D. Hayman E.G. Ruoslahti E. Cell. 1981; 26: 259-267Abstract Full Text PDF PubMed Scopus (282) Google Scholar. Synthetic hexapeptides, GRGDSP and GRADSP, were purchased from Calbiochem. Purified monoclonal antibodies against the extracellular domains of the mouse integrin subunits, α5 (5H10–27, rat IgG), β3 (Hmβ3, hamster IgG), and αv (H9.2B8, hamster IgG), were purchased from Pharmingen. A polyclonal antiserum to the mouse β1 integrin subunit (Ab3675) was a generous gift from Dr. Steven Akiyama (National Institutes of Health, Bethesda, MD). Non-immune rat and hamster IgG was purchased from Jackson ImmunoResearch Laboratories (West Grove, PA). Non-immune rabbit serum was from Sigma. Mouse blastocyst stage embryos were generated as previously described (Yelian et al., 1993Yelian F.D. Edgeworth N.E. Dong L.J. Chung A.E. Armant D.R. J. Cell Biol. 1993; 121: 923-929Crossref PubMed Scopus (81) Google Scholar, using 8–10-week-old NSA females (Colony 202, Harlan Spargue-Dawley, Indianapolis, IN) that were superovulated with pregnant mare serum gonadotropin and human chorionic gonadotropin (Sigma) and mated with B6SJL/J males (Jackson Laboratories, Bar Harbor ME). Blastocysts were collected on gestational day 4 (day 1 = day of vaginal plug) by flushing the uterine horns with pre-warmed HEPES-buffered Eagle's medium (Sigma) supplemented with 2 mg/ml calcium lactate, 60 µg/ml sodium pyruvate, 0.3 mg/ml glutamine, 4 mg/ml BSA, 100 units/ml penicillin, and 0.1 mg/ml streptomycin. Blastocysts were transferred by micropipette to pre-warmed Ham's F-10 medium supplemented with 4 mg/ml BSA, 100 units/ml penicillin, and 0.1 mg/ml streptomycin (all from Sigma) on plastic Petri dishes (Falcon 1008, Falcon Labware, Oxnard, CA) flooded with water-extracted (1:10) light mineral oil (Aldrich) to prevent evaporation. Culture medium was changed daily. All embryo culture was carried out in a 5% CO2 incubator at 37 °C. Blastocyst outgrowths were produced in serum-free Ham's medium by culture on plates precoated with 50 µg/ml fibronectin, as previously detailed (Armant et al., 1986aArmant D.R. Kaplan H.A. Lennarz W.J. Deo. Biol. 1986; 116: 519-523Crossref PubMed Scopus (162) Google Scholar, Armant et al., 1986bArmant D.R. Kaplan H.A. Mover H. Lennarz W.J. Proc. Natl. Acad. Sci. U. S. A. 1986; 83: 6751-6755Crossref PubMed Scopus (92) Google Scholar. Blastocyst attachment to the culture plate surface was quantified by noting whether the embryos moved while media was gently blown across the embryo using a micropipette. Trophoblast adhesion and cell migration, associated with blastocyst outgrowth, were distinguished from attachment by microscopically visualizing a spreading monolayer of cells around the embryo and disappearance of the spherical blastocyst morphology, as previously described (Armant et al., 1986bArmant D.R. Kaplan H.A. Mover H. Lennarz W.J. Proc. Natl. Acad. Sci. U. S. A. 1986; 83: 6751-6755Crossref PubMed Scopus (92) Google Scholar, Armant, 1991Armant D.R. Biol. Reprod. 1991; 45: 664-672Crossref PubMed Scopus (42) Google Scholar. Fluoresbrite™ YG polystyrene fluorescent microspheres (1.0 µm, Polysciences, Warrington, PA) were first vigorously washed to completely remove surfactants used in their manufacture that may inhibit protein adsorption (Bangs, 1990Bangs L.B. Am. Clin. Lab. News Ed. 1990; 9: 16-17PubMed Google Scholar). A 2.5% solution of microspheres was washed five times by centrifugation for 5 min at 6000 × g in a high phosphate saline solution (hPBS) that consisted of 0.1 m NaCl, 0.0027 m KCl, 0.005 m Na2PO4, 0.0085 m KH2PO4, and 0.05 m NaH2PO4, pH 7.4, and resuspended to 2.5% in fresh hPBS. The washed microspheres were centrifuged, resuspended in an equal volume of hPBS containing 144 µg/ml FN-120 or α1-AG (Sigma), and incubated at room temperature for 24 h with continuous agitation on a vortex mixer. The microspheres were then washed five times in hPBS by centrifugation, as described above. To saturate all remaining protein binding sites on the microspheres, they were incubated overnight at room temperature with 1 mg/ml α1-AG while being continuously agitated. Finally, the microspheres were washed in hPBS five times by centrifugation and stored at 4 °C as a 2.5% suspension. For use in an assay, the microspheres were diluted to the desired final concentration in PBS containing 10 mg/ml BSA (PBS/BSA). FN-120-coated fluorescent microspheres were used to semi-quantitatively assay fibronectin binding activity of differentiating mouse blastocysts cultured for 48 or 72 h on a non-adhesive (BSA) surface. Blastocyst fibronectin binding activity was potentiated with immobilized ligand by culturing the embryos in Ham's medium at 37 °C on plates precoated with 50 µg/ml FN-120. Embryos were then carefully removed from the plate surface with a micropipette, and any outgrowing or damaged blastocysts were discarded. For potentiation using soluble ligand, the embryos were cultured on BSA-coated plates at 37 °C in medium supplemented with 50 µg/ml FN-120. Initial experiments (see "Results") revealed that the accumulation of heparan sulfate proteoglycans on cultured blastocysts caused protein-coated microspheres to nonspecifically bind to the surface of the embryos. To remove heparan sulfate, blastocysts were incubated for 30 min at 37 °C in Hanks' balanced salt solution with 1 unit/ml each of heparinases I, II, and III (Sigma, H2519, H6512, and H8891, respectively) prior to the binding assay. The binding assay was carried out by incubating embryos in 10-µl drops containing protein-coated microspheres for 30 min at 4 °C. In competitive inhibition assays, blastocysts were pre-exposed to the soluble competitor diluted in PBS/BSA for 30 min at 4 °C prior to incubation with microspheres. In EDTA inhibition experiments, blastocysts were incubated with 5 mm EDTA in PBS/BSA for 15 min at 4 °C and washed free of the EDTA before assay. EDTA-treated blastocysts were subsequently incubated for an additional 30 min at 4 °C in PBS/BSA containing 0–10 mm CaCl2, MgCl2, or MnCl2 and then washed in PBS/BSA. Following incubation with the microspheres, the blastocysts were washed free of unbound microspheres by pipetting through five drops of ice-cold PBS/BSA. The washed blastocysts were fixed in PBS containing 2% paraformaldehyde (Polysciences, Inc.) for 2 h at 4 °C and stored refrigerated in PBS/BSA containing 0.1% sodium azide for subsequent image analysis. The microspheres bound to each blastocyst were observed using a Leitz (Wetzlar, Germany) Fluovert FU inverted epifluorescence microscope, exciting the fluorescent microspheres at 450–490 nm. The amount of microsphere binding was quantified over either the embryonic or abembryonic pole of each blastocyst using a computer-based image analysis system. Embryos were oriented with their inner cell masses to one side and viewed using a 25 × objective. The fluorescent images were monitored using a Dage-MTI (Michigan City, Indiana) CCD 72 video camera, adjusted to electronically reverse the image and generate a relative optical density that was equivalent to the fluorescence intensity. The black level of the camera was set to 70 and the polarity was optimized to obtain a background reading of zero. Electronic images were transmitted to a Dell (Austin, TX) 486sx computer with MCID-M4 image analysis software (Imaging Research Inc., St. Catherines, Ontario) and displayed in pseudo-color corresponding to the relative optical density, which was found in preliminary experiments to be directly related to the number of microspheres present (data not shown). Ten equally spaced relative optical density values were averaged along the in-focus edge of each digitized image over 180° of either the embryonic or abembryonic pole using a small square tool projected by the imaging system and positioned with a computer mouse. Each data point, reported as mean ± S.E., represents the average relative fluorescence intensities pooled from at least three experiments that each included at least five embryos. To determine how rapidly cycloheximide inhibited new protein synthesis, blastocysts were prelabeled with 13.6 µm [35S]methionine (160 µCi/mmol) for 1 h at 37 °C and then exposed to 10 µg/ml cycloheximide. Groups of embryos were periodically removed, and radiolabeled proteins were precipitated from cell extracts with trichloroacetic acid. Washed blastocysts were spotted on Gelman (Ann Arbor, MI) glass fiber filters and placed in 10% trichloroacetic acid for 10 min at 4 °C with agitation. The filters were then transferred to boiling 5% trichloroacetic acid for 3 min. The filters were sequentially washed with 95% ethanol, 95% ethanol/acetone (1:1), and finally acetone, each for 5 min at room temperature. The filters were allowed to air dry for 2 h and then placed in scintillation vials containing 2 ml of Scintisol scintillation liquid (Isolab Inc., Akron OH). The radioactivity was then determined using a Taurus Automatic Liquid Scintillation Counter (ICN Micromedic Systems, Huntsville, AL). Preliminary experiments revealed that the abembryonic pole of the mouse blastocyst bound FN-120-coated microspheres in a concentration-dependent manner, whereas microspheres coated with the blocking protein α1-AG did not bind. However, in competition assays, binding was equally inhibited by soluble fibronectin, FN-120, and heparan sulfate (data not shown). We suspected that the observed binding of the microspheres was mediated by heparan sulfate, which accumulates at the blastocyst surface (Farach et al., 1987Farach M.C. Tang J.P. Decker G.L. Carson D.D. Dev. Biol. 1987; 123: 401-410Crossref PubMed Scopus (96) Google Scholar. To determine whether heparan sulfate proteoglycans were responsible for the observed binding, we measured the binding of FN-120-coated microspheres over the course of blastocyst differentiation with and without exposure to heparinase. A gradual increase was observed in microsphere binding on the surface of the embryos between 24 and 48 h of culture, followed by a rapid decrease at 52 h (Fig. 1A). Conversely, the group exposed to heparinase displayed little binding throughout. Therefore, it appeared that microsphere binding was, in fact, due to the expression of heparan sulfate. As shown in Fig. 1B, blastocyst attachment mirrored the pattern of heparan sulfate-mediated binding, whereas blastocyst outgrowth correlated temporally with the loss of heparan sulfate between 48 h and 72 h. Blastocyst attachment occurred transiently on BSA-coated surfaces, perhaps due to the presence of strongly charged heparan sulfate groups. These results suggested that proteoglycans on the surface of the mouse embryo may mediate the initial attachment of the blastocyst to the substratum, while adhesion and cell migration proceed through a more specific ligand-mediated mechanism. Accordingly, in all subsequent experiments, embryos were treated with heparinase prior to incubation with microspheres to specifically assay fibronectin binding activity. Normally, blastocysts collected on gestation day 4 and cultured on fibronectin-coated surfaces begin to adhere and outgrow within 72 h (Fig. 1B) (Armant et al., 1986aArmant D.R. Kaplan H.A. Lennarz W.J. Deo. Biol. 1986; 116: 519-523Crossref PubMed Scopus (162) Google Scholar. However, as suggested by Fig. 1A, when embryos cultured 72 h on BSA were assayed with FN-120-coated fluorescent microspheres, no appreciable fibronectin binding activity was detected (Fig. 2B). To determine whether the adhesiveness of the apical trophoblast surface could be potentiated by exposure to fibronectin, blastocysts cultured for 72 h without an adhesive substrate were incubated on Petri plates precoated with FN-120 for 3 h, carefully removed before they began to adhere, and assayed for their ability to bind FN-120-coated microspheres. Using this protocol, FN-120 bound to the abembryonic surface of blastocysts (Fig. 2D), where mouse trophoblast differentiation initiates (Dickson, 1963Dickson A.D. J. Reprod. Fértil. 1963; 6: 465-466Crossref PubMed Google Scholar; Kirby et al., 1967Kirby D.R. Potts D.M. Wilson I.B. J. Embryol. Exp. Morphol. 1967; 17: 527-532PubMed Google Scholar. Embryos that were cultured for 48 h and then similarly potentiated with FN-120 bound considerably less FN-120 (Fig. 2F), correlating with the relatively low rate of outgrowth found at this earlier developmental stage (Fig. 1B). Quantification of the fibronectin binding activity of blastocysts cultured for both 48 and 72 h by image analysis revealed that binding activity at the embryonic pole of ligand-potentiated blastocysts was almost undetectable (Fig. 3A). However, low levels of FN-120 binding to the abembryonic pole of ligand-potentiated blastocysts were detectable at 48 h and were approximately 6 times greater at 72 h (Fig. 3A).FIG. 3Developmental regulation and specificity of microsphere binding. A, microspheres coated with FN-120 were incubated with blastocysts cultured 48 or 72 h on BSA and then 3 h on plates coated with FN-120, as in Fig. 2, C–F. The fluorescence intensity of the bound microspheres was then quantified along the embryonic (solid bars) or abembryonic (open bars) pole of each embryo by image analysis. B, blastocysts cultured on BSA for 72 h and treated as indicated in Fig. 2, C and D, were incubated at 4 °C in various concentrations of fluorescent microspheres coated with FN-120 (□) or α1-AG (○). The fluorescence intensity of the bound microspheres was quantified along the abembryonic pole of each embryo by image analysis. Microspheres containing FN-120 bound to the embryos in a concentration-dependent manner up to 0.5% and then began to show signs of saturation. The blocking protein, α1-AG, did not mediate significant binding of the microspheres.View Large Image Figure ViewerDownload Hi-res image Download (PPT) The relationship of FN-120 binding activity to microsphere concentration was determined over the mural trophectoderm of blastocysts cultured for 72 h, demonstrating a linear relationship with microsphere concentration up to a point approaching saturation (Fig. 3B). Microspheres containing only the blocking protein α1-AG did not significantly bind to the embryos (Fig. 3B). Based on these binding kinetics, a microsphere concentration of 0.2% was chosen for all subsequent ligand binding assays to work within a range that was linear with microsphere concentration. Biochemical characterization of the fibronectin binding activity along the abembryonic surface of blastocysts cultured for 72 h was consistent with trophoblast adhesive behavior during culture on fibronectin. Binding was inhibited by addition of soluble fibronectin or FN-120 but not by fragments representing other functional domains of fibronectin or by heparan sulfate (Fig. 4A). The concentration of intact fibronectin required to produce 50% inhibition of binding (IC50) was 0.2 µm, while FN-120 had an IC50 of 0.8 µm. Soluble laminin did not compete with binding by the FN-120-coated microspheres (Fig. 4A), supporting our earlier position (Armant et al., 1986bArmant D.R. Kaplan H.A. Mover H. Lennarz W.J. Proc. Natl. Acad. Sci. U. S. A. 1986; 83: 6751-6755Crossref PubMed Scopus (92) Google Scholar, Armant, 1991Armant D.R. Biol. Reprod. 1991; 45: 664-672Crossref PubMed Scopus (42) Google Scholar; Yelian et al., 1993Yelian F.D. Edgeworth N.E. Dong L.J. Chung A.E. Armant D.R. J. Cell Biol. 1993; 121: 923-929Crossref PubMed Scopus (81) Google Scholar that fibronectin and laminin mediate trophoblast adhesion through separate receptors. To determine whether the fibronectin binding activity of blastocysts was mediated by the RGD recognition site, competitive inhibition assays were performed using the synthetic peptides GRGDSP and GRADSP. GRGDSP inhibited FN-120-coated microsphere binding to embryos in a dose-dependent manner, while the control peptide, GRADSP, demonstrated no ability to inhibit binding over the same concentration range (Fig. 4B). The fibronectin binding activity of blastocysts cultured for 72 h therefore appeared to be specifically mediated by the RGD recognition sequence. Inhibition of fibronectin binding activity by GRGDSP suggested a likely role for integrins. Also in support of the idea that the fibronectin receptors were integrins, w
Referência(s)