Distribution of laminin collagen type IV collagen type I and fibronectin in chicken cardiac jellybasement membrane.код для вставкиСкачать
THE ANATOMICAL RECORD 224:417-425 (1989) Distribution of Laminin, Collagen Type IV, Collagen Type I, and Fibronectin in Chicken Cardiac Jelly/Basement Membrane CHARLES D. LITTLE, DOMINIQUE M. PIQUET, LYNN A. DAVIS, LUANNE WALTERS, AND CHRISTOPHER J. DRAKE Department of Anatomy and Cell Biology, School of Medicine, University of Virginia, Charlottesville, Virginia 22908 ABSTRACT Light microscopic immunolabeling studies were designed to identify and locate structural components within the cell-free extracellular matrix which lies between the embryonic endocardial and myocardial tubes. Affinitypurified antibodies were used to examine stage 15-22 embryonic chicken hearts. Specimens were immunolabeled by using three different methodologies: 1) postembedding labeling of 10 pm cryostat sections, 2) preembedding labeling (en bloc) of whole hearts, and 3) postembedding labeling of ethanol/acetic acid-fixed paraffin sections. Our results establish the spatial distribution of collagen type I and demonstrate for the first time the presence of collagen type IV and laminin in the myocardial-basement-membranekardiac jelly. The formation of a vertebrate heart involves cell-cell data strongly support the contention of Markwald and adhesion, cell-substrate adhesion, and cellular-motile his colleagues that the ECM, commonly referred to as activity. Our interest is centered on the cell surface- cardiac jelly, is in fact an expanded basement memextracellular matrix (ECM) adhesions that occur dur- brane (Markwald et al., 1984; Kitten et al., 1987). ing the formation of endocardial cushions. Kitten et al. MATERIALS AND METHODS (1987) showed that fibronectin is abundant in the ECM En Bloc lmmunolabeling through which endocardial-mesenchymal cells migrate Stage 15-21 embryos were removed from the egg as they begin formation of the cushion tissue. Previous work showed that hyaluronic acid (Markwald and Ad- and placed in phosphate-buffered saline (PBS). The ams-Smith, 19721, glycosaminoglycans (Manasek et hearts were dissected from the embryo and then transal., 19731, glycoproteins (Manasek, 19771, and collagen ferred to a fresh dish of cold PBS. Prior to incubation in I (Johnson et al., 1974) are also present in this ECM. primary antibody the hearts were carried through each The term myocardial-basement-membrane was re- of the steps outlined below. Following each step the cently proposed by Kitten and his colleagues (1987) to hearts were washed three times for 5 min in PBS: 1) replace the term cardiac jelly as a name for the ex- Fixation in fresh 3% paraformaldehydelPBS (weight/ panded cell-free ECM through which the endocardial- volume), pH 7.5, for 30 min. 2) Incubation in 0.1 M mesenchyme cells migrate. In this communication, the lysinePBS, pH 7.5 for 30 min. 3) Incubation in 0.4% term myocardial-basement-membranekardiac jelly Triton X-lOO/PBS (v/v) for 30 min. 4) Incubation in 1 (MBM/CJ), will be use to refer to the expanded extra- mg/ml bovine serum albumidPBS for 24 hr at 4°C. cellular matrix which lies between the primitive myo- Next, the hearts were immersed in PBS containing primary antibody (10 pg/ml), 1 mg/ml of bovine serum cardial and endocardial tubes. In order t o pursue our studies on the cell-ECM ad- albumin, and 0.02%Triton X-100, and then they were hesions which occur during the migration of mesenchy- incubated for 48 hr at 4°C with gentle rocking. The ma1 cells into the MBM/CJ, it was necessary to deter- specimens were then washed in a large excess of PBS mine the identity and spatial distribution of ECM (4°C) for 24 hr. The last steps were then repeated with components within this tissue. Accordingly, immuno- an appropriate fluorochrome-conjugated secondary anfluorescence localization studies were undertaken us- tibody (Cooper Biomedical, Malvern, PA). In the case of ing affinity-purified antibodies to chicken collagen I, laminin immunolabeling a few specimens (Fig. 3f) chicken collagen IV, chicken fibronectin, and mouse were immunolabeled with a tertiary antibody. A fluolaminin. By using three different immunolabeling pro- rescein-conjugated donkey antigoat IgG (Jackson Imtocols we were able to visualize the distribution of each munoresearch Laboratories, Inc., West Grove, PA) was ECM component listed above in stage 15-22 embryonic applied after the secondary antibody (fluorescein-conchicken MBM/CJ. These results show that in addition jugated goat antirabbit IgG). The specimens were then to previously reported macromolecules, collagen IV and laminin are present in this ECM. Consequently, Received August 3, 1988; accepted January 11, 1989. collagen IV and laminin can be added to the list of Address remint reauests to Charles D. Little. DeDt. of Anatomy and putative substrate adhesion molecules for the migra- Cell Biology,-Box 436, Medical Center, University'of Virginia, Chartory endocardial-mesenchymal cells. Moreover, these lottesville, VA 22908. 0 1989 ALAN R. LISS, INC. C.D. LITTLE ET AL. 418 washed as described above. After immunolabeling, the hearts were infused with 5% sucrose/PBS (w/v) overnight at 4°C and then incubated with 15% sucrose/PBS (w/v) for 1 h r a t 4°C. The specimens were mounted in Tissue Tek OCT mounting medium (Miles Sci., Inc., Naperville, IL), brought to -2o"C, and sectioned in the frozen state on a DamodIEC Harris-Cryostat (Inter. Equip. Co., Needham Heights, MA). Sections (10 pm) were placed on gelatin-coated microscope slides and then allowed to thaw and dry at 37°C for 30 min. The sections were washed free of OCT mounting medium and sucrose and then mounted in 90% glycerol/lO% PBS (v/v) under a #1 glass coverslip. All solutions contained 0.01% sodium azide (w/v) to prevent microbial growth. Postembedding lmmunolabeling of Frozen Sections This method is essentially the same as described above except that the hearts were not treated with Triton X-100 or antibodies (en bloc) after fixation. Instead, the fixed hearts were sectioned in the frozen state (above), air dried, and then incubated with 10 pg/ml of primary antibody for 1 h r at room temperature. The sections were washed extensively with PBS, incubated with fluorochrome-conjugated secondary antibodies, washed as before, and mounted (above) for light microscopy. Acid/Alcohol-Fixed Paraffin Sections Embryonic hearts were prepared essentially according to Saint-Marie (1962). Briefly, specimens were fixed overnight in 99 parts of 95% ethanol and 1 part glacial acetic acid at 4°C. The hearts were washed 2 times with 4°C absolute ethanol for 30 min and then 2 times with 4°C xylene for 30 min. Next, they were incubated in a 5 0 5 0 mixture of xylene:paraffin, followed by a 25:75 mixture and finally 100% paraffin overnight. Sections were cut at 10 pm thickness and placed on gelatin-coated microscope slides. The sections were cleared of paraffin by 3 x 5 min washes in ice-cold xylene and then rehydrated in a graded series of cold ethanollwater washes, followed by a 2 h r incubation in PBS at 4°C. After this step the sections were immunolabeled as described above. Immunological Reagents against immobilized native chicken collagen I and then affinity purified on a n agarose-chicken collagen IV column. Western immunoblot analyses showed that the cross-adsorbed rabbit antichicken type IV collagen antibodies recognized the immunogen but not chicken type I collagen (data not shown; the complete characterization of this antibody is the subject of a manuscript in preparation). Immunoprecipitation of 35 S - methionine - labeled stage 22 chicken embryonic heart explants was accomplished a s follows: Sixty hearts were dissected free, cut into several small portions, and placed in a 100 mm tissue culture dish with 2 ml of Dulbecco's modified Eagle's Medium plus 10% chicken serum. The explants were allowed to attach overnight; then the medium was replaced with 2 ml of methionine-free DMEM, to which was added 100 pCi/ml of 35S-methionine. The explants were returned to the tissue culture incubator for 24 hr. At the end of the labeling period, the explants were scraped from the dish into 500 pl of radioimmunoprecipitation assay buffer (RIPA) composed of 1% Triton-X 100, 1% sodium deoxycholate, 0.5% sodium dodecylsulfate, 0.1 M Tris/HCl, pH 8.0, and 0.1 M NaC1. The mixture was homogenized with a loose pestle and centrifuged at 10,OOOg for 5 min. A 50 p1 aliquot of cleared lysate was mixed with 10 pg of nonimmune rabbit IgG for 15 min at room temperature, followed by tumbling with Sepharose protein A for 15 min. The beads were pelleted by centrifugation and discarded; the supernatant was transferred to a new tube. Affinity-purified rabbit antimouse laminin (10 pg) or 10 p1 of mouse hybridoma 31-2 ascites fluid was added to the sample and tumbled gently 1h r at 4°C. In the case of the mouse 31-2 antibody a second antibody (rabbit antimouse IgG) was added to the sample and tumbled for 1hr. This was followed by incubation with Sepharose protein A beads for 1 hr. The beads were pelleted and washed repeatedly in RIPA buffer, resuspended in SDS sample buffer, heated to 100°C for 5 min, then pelleted. The resulting supernatant was subjected to SDS-polyacrylamide gel (5%)electrophoresis according to Laemmli (1970). Immunoreactive polypeptides were analyzed by fluorography. To serve as a positive immunological control, a n embryonic mouse epithelial cell culture was also labeled with 35-methionine, lysed in RIPA, and treated a s above. Antibodies to chicken collagen I and fibronectin were RESULTS characterized previously (Little and Chen, 1982). Antibody Specificity Laminin was purified from the Englebreth-HolmSwarm (EHS) mouse tumor (Kleinman et al., 1982, Laminin purified from the mouse EHS-tumor ap1983) and injected into a rabbit, and the resulting an- pears as two diffuse Coomassie blue-stained bands a t tibodies were affinity purified on a laminin-Sepharose approximately 440 kDa (A chain) and 220 kDa (B column as described (Little and Chen, 1982). A mouse chains) when analyzed on 5% SDS-polyacrylamide antichicken laminin monoclonal antibody (31-2) was slab gels (Fig. l a ; lane 1). Rabbit antibodies raised generously provided by Dr. E. Godfrey, Medical College against a similar preparation of laminin were affinity of Wisconsin. purified on immobilized antigen columns and then Collagen IV was purified from pepsinized adult used to probe 35S-methionine-labeled cultures of an chicken breast muscle by using the method of Mayne embryonic mouse epithelial cell line. Electrophoresis of and Zettergren (1980). A method developed for prepa- immunoprecipitates shows two major polypeptide ration of antibodies to mouse type IV collagen (Chen bands. The radiolabeled polypeptides migrated to and Little, 1985, 1987) was adapted for isolation of an- slightly lower molecular weight positions relative to tibodies to chicken antigen. Briefly, the purified colla- the EHS-tumor laminin chains (Fig. la: lane 2). A congen IV fragments were injected subcutaneously in a trol reaction (Fig. l a : lane 3) using nonimmune rabbit rabbit. The elicited antibodies were cross-adsorbed IgG shows that the bands present in lane 2 are the BASEMENT MEMBRANE ANTIGENS IN EARLY HEART 1 2 3 1 2 3 4 -B Fig. 1. Results from two 5% SDS-polyacrylamide gels (a and b) are shown. Some lanes were stained with Coomassie blue and some show fluorographically detected 35S-methionine-labeledpolypeptides. Bars indicate molecular weight markers, M, = 205,000, 116,000, 97,000. a: Purified mouse-EHS laminin A chains (M, = 400,000) and B chains (M, = 210,000) chains are shown after Coomassie blue staining (lane 1). Embryonic mouse epithelial cell cultures were radiolabeled as described in Materials and Methods and then subjected to immunoprecipitation analysis by using the rabbit antimouse laminin affinity-purified IgG (lane 2).A control reaction in which nonimmune rabbit IgG was used (lane 3).b Purified mouse-EHS laminin A and B chains stained with Coomassie blue (lane 1).Radiolabeled stage 22 embryonic chicken heart lysates were analyzed by immunoprecipitation by using rabbit anti-mouse laminin IgG (lane 2). The same sample was probed with a monoclonal mouse antichicken laminin antibody (lane 3). A control in which nonimmune rabbit IgG was substituted for immune IgG (lane 4). The migration positions of chicken laminin chains are designated by the arrowheads labeled A and B. Equal aliquots of radiolabeled antigen were loaded in each lane and then analyzed identically. 419 180 kDa, designated “A” and “B,” displayed lower electrophoretic mobility than the EHS-laminin. The 180 kDa band immunoprecipitated by the rabbit antimouse laminin antibodies (lane b2) comigrated with the laminin B chain recognized by the mouse antichicken laminin antibodies (lane b3). A control immunoreaction in which 10 bg/ml of nonimmune rabbit IgG was substituted for immune rabbit IgG is shown in Figure l b , lane 4.The control lane shows a sharp band a t approximately 200 kDa as well as faint high molecular weight bands. These nonspecifically precipitated chicken polypeptides are present in all the experimental lanes, including duplicate lanes not shown (compare lane 4 with lanes 2 and 3). Based on the results shown in Figure l b , it appears that the rabbit antibodies preferentially recognize chicken laminin B chains. I n contrast, the mouse antibody (31-2) precipitates laminin A chains (-280 kDa) and B chains (-180 kDa) in addition to other unknown polypeptides. The affinity-purified rabbit IgG was more efficient at immunoprecipitating chicken laminin B chains than the mouse monoclonal antibody. (All experimental conditions used in lanes b2, b3, and b4 were identical, including photographic, fluorographic, biochemical, and immunochemical.) Indirect immunofluorescence microscopy conducted by using the rabbit antimouse laminin (polyclonal) antibody resulted in easily detectable levels of immunostaining (see below). However, under similar conditions, immunostaining with the mouse antichicken laminin monoclonal antibody did not result in detectable fluorescent images (data not shown). Western blotting of adult chicken muscle pepsin extracts with rabbit anticollagen IV antibodies yielded a set of immunoreactive bands with the molecular weights reported for pepsinized chicken collagen IV (Mayne and Zettergren, 1980). The immunoreactive bands were removed by prior treatment of samples with bacterial collagenase. Affinity-purified rabbit antichicken collagen IV antibodies, which had been crossadsorbed against collagen type I, did not recognize chicken collagen I on immunoblots (data not shown, manuscript in preparation). In addition, the immunostaining patterns of type I collagen and type IV collagen were reproducibly different (see below), strongly suggesting immunological monospecificity. result of a specific immunoprecipitation reaction. These immunochemical data (Fig. la) demonstrate that the rabbit antimouse laminin antibodies will reclmmunostaining ognize mouse laminin A and B chains in a crude cell A differential interference contrast (DIC) microlysate. Next we tested the ability of the rabbit antimouse graph of a section through the outflow tract region of a laminin antibodies to recognize chicken laminin. For stage 15 embryonic chicken heart reveals a n outer mymolecular weight comparison, a sample of purified ocardial tube (m) and inner endocardia1 tube (el. Fine mouse EHS-tumor laminin is also included in this SDS filamentous material, observable by DIC optics, occugel (Fig. l b : lane 1).Stage 22 chicken heart explants pies the space between the two cellular tubes, (Fig. 2a). were labeled with 35S-methionineand lysed with RIPA The corresponding epif luorescence micrograph shows buffer, and the resulting solution was probed with the the location of immunoreactive laminin (Fig. 2b). Imrabbit antilaminin antibodies. A diffuse band a t ap- munostaining within the MBM/CJ is distributed as proximately 180 kDa and a sharp band at 100 kDa irregular granular spots which coincide with the mawere immunoprecipitated (Fig. lb: lane 2). A similar terial seen under DIC microscopy. Considerable imradiolabeled sample was probed with a mouse mono- muno-fluorescence is associated with the heart tubes. clonal antibody (31-2) to chicken laminin. This resulted In order to demonstrate immunostaining within the in the detection of diffuse bands at approximately 280 MBMiCJ it was necessary to overexpose the immunoand 180 kDa and faint bands at 150 and 100 kDa (Fig. staining associated with the cell layers. A similarly prepared stage 15 heart section is shown l b : lane 3). The chicken laminin chains at 280 kDa and 420 C.D. LITTLE ET AL. Fig. 2. This set of micrographs show cryostat sections (10 pm) of stage 15 embryonic hearts in the region of the outflow tract. The hearts were immunolabeled prior to embedding and cryosectioning. a: The differential interference contrast (DIC) image shows sparse ECM material lying between the myocardium (m) and the endocardium (el. b The corresponding immunofluorescent image shows very bright staining for laminin in the heart walls and punctate arrays of staining which codistribute with the material lying between the heart tubes. DIC image (c) and the epiflourescent image (d) of a heart immunostained with collagen type I antibodies show bright strands of immunoreactive material between the primitive heart tubes. x 180. in Figure 2c and 2d. In this case the embryonic heart was incubated with antibodies to collagen I. The immunofluorescence shows bright fibrils of collagen throughout the MBM/CJ. In some cases the fibrils extend across the entire acellular space. In order to demonstate the fine fibrils within the MBM/CJ it was necessary to overexpose the fluorescence associated with the heart tubes. The embryonic tissue shown in Figure 2 was immunolabeled en bloc and then sectioned. Figure 3 also depicts the results of immunolabeling of embryonic hearts en bloc. The DIC image (Fig. 3a) shows the atrioventricular region of a stage 22 heart in which cells (arrowhead) had moved from a n earlier position in the endocardium (e) and migrated into the MBM/CJ. The cells seem to be closely associated with conspicuous fibers, which appear to bridge the region between the endocardium and the myocardium (m). The epifluorescence image (Fig. 3b) shows th at fibro- nectin antibodies strongly immunolabel the fibers which are visible under DIC optics. The cellular endocardial and myocardial tubes are heavily labeled with the fibronectin antibodies. Figure 3c shows a stage 22 heart similar to the previous figure. Migratory cells (arrowhead)are in contact with strands of material visible under DIC optics. The distribution of immunolabeled collagen type I coincides with these filaments. A comparison of Figure 3b and 3d suggests that the ECM fibers visible under DIC optics contain both immunoreactive collagen type I and fibronectin. The myocardium (m) and endocardium (el immunolabel strongly with the collagen type I antibodies. The distribution of laminin and collagen type IV were examined by using en bloc staining in approximately stage 19 embryonic hearts (Fig. 3e-h). The DIC optical image (Fig. 3e) shows material intervening between the atrioventricular myocardium (m) and endocardium (e). Laminin was distributed as a n amor- BASEMENT MEMBRANE ANTIGENS IN EARLY HEART phous array of very fine filaments throughout the MBM/CJ. Codistributed within this fine material were discrete, variable-sized, immunoreactive particles (Fig. 3f). Thus the laminin immunolabeling seemed to distinguish two staining patterns, one granular and the other fine filamentous. The specimen shown in Figure 3f was immunostained by using a three-step procedure which employed a fluorochrome-conjugated secondary and tertiary antibody (see Materials and Methods). In the case of collagen IV, immunolabeling showed fine wisplike filaments throughout the acellular expanse of ECM (Fig. 3h), which were associated with material visible by DIC optics (Fig. 3g). The fluorescence intensity of collagen IV staining was often greater than that of laminin, based on subjective judgement. For comparison with the preembedding labeling technique, Figure 4 shows a stage 19 atrioventricular region that was embedded in paraffin and then immunolabeled after sectioning. Affinity-purified rabbit antilaminin IgG and nonimmune IgG were used a t the same concentrations. The DIC images show conspicuous ECM material lying between the endocardium and the myocardium (Fig. 4a). This material contains immunoreactive laminin, which appears as small dots or granules of variable size associated with fine wisplike staining (Fig. 4b). The staining is localized to areas where the filaments of ECM are apparent under DIC optics. In contrast, a control specimen (Fig. 4c) shows no specific immunostaining within the myocardial basement membrane when the same concentrations (10 p.g/ml) of nonimmune IgG and fluorochrome-conjugated secondary antibody are applied to the specimen (Fig. 4d). The initial photomicrographic exposure times and the photographic printing conditions were identical for the sections shown in Figure 4b and 4d. Hearts prepared by the en bloc staining method, using control antibodies, showed less nonspecific fluorescence than paraffin-labeled hearts. The immunostaining patterns described above are representative of a series of examinations of various stage hearts (15-22) using each of the antibodies (laminin, collagen I, collagen IV,and fibronectin) in the three different labeling protocols described in Materials and Methods. Space limitations preclude presentation of all the data; however, the immunolabeling results shown were consistent with all three labeling procedures for each affinity-purified antibody. The en bloc staining protocol resulted in the best preservation of the delicate ECM filaments between the myocardium and endocardium. Indeed, labeling specimens after cryosectioning resulted in the considerable MBM/CJ being removed during the various washes. In each staining experiment three controls were performed: 1) Some specimens were incubated with only fluorochrome-conjugated secondary antibodies (no primary antibody). 2) Some hearts were incubated in nonimmune primary antibodies followed by fluorochromeconjugated secondary antibodies. 3) Some specimens were not incubated in with antibodies but were carried through the various solutions and inspected for intrinsic fluorescence. In all cases the control specimens showed trace amounts of diffuse fluorescence distributed evenly throughout the microscopic field, most of which was attributable to intrinsic fluorescence. 42 1 DISCUSSION ECM Constituents of Cardiac Jelly The list of specifically identified components contained within the MBM/CJ has increased as more refined techniques and immunological reagents have become available. The current list of constituents includes hyaluronic acid, heparin sulfate proteoglycan (HSPG), collagen I, collagen IV, fibronectin, and laminin (Markwald et al., 1984;Johnson et al., 1974; Kitten et al., 1987; and this report). The immunof luorescent images of MBM/CJ fibronectin in this study differed from those of Kitten and his co-workers (1987). This may be due to the different methods of fixation. The freeze-substitution method used by these workers resulted in a more homogenous distribution of ECM components whereas the paraformaldehyde fixation used here resulted in filamentous bundles of ECM material within the MBM/CJ. The en bloc fixation and labeling method was chosen to maximize the immunofluorescence signal-to-noise ratio. In this regard, the en bloc labeling method results in a virtually black background. ECM Adhesions During Cushion Mesenchyme Migration Our interests lie in the molecular basis of cellular adhesion to the ECM during morphogenesis. All of the ECM components in the list above, except HSPG, are known to have corresponding cell surface adhesion receptors (Underhill et al., 1985; Rubin et al., 1986; Buck and Horwitz, 1987; Dedhard et al., 1987; Ruoslahti and Pierschbacher, 1987; Hynes, 1987; Terranova et al., 1983; Lesot et al., 1983, Brown et a]., 1983; Ogle et al., 1986; Graf et al., 1987; Ogle and Little, 1989). Thus, each of these macromolecules, individually or in combination, is a candidate for the adhesive scaffold to which endocardial-derived mesenchymal cells attach during migration into and through the MBM/CJ. Recent work in this laboratory showed that embryonic cushion-mesenchyme cells will migrate vigorously through 3-dimensional gels of laminin. Moreover, addition of the laminin B1 chain synthetic pentapeptide tyrosyl-isoleucyl-glycyl-seryl-arginine to the cultures will abolish the migration of cushion-mesenchyme cells. Various Arg-Gly-Asp (RGDI-containingsynthetic peptides, however, have no effect on migration of cushion-mesenchyme cells into 3-dimensional gels of laminin, collagen, or collagen + fibronectin (Davis et al., 1989; and unpublished data). The ability of endocardial cushion-mesenchymal cells to migrate through laminin gels in vitro, coupled with the presence of laminin in the MBM/CJ, suggests that laminin may be one of the ECM components to which the mesenchymal cells adhere during formation of the endocardial cushions. lmmunochemical Characteristics of Avian Laminin The rabbit antimouse laminin antibodies immunoprecipitate a polypeptide with the characteristics of laminin B chains but not of A chains. It is possible that the affinity-purified rabbit IgG cross-reacts only with chicken B chains whereas the mouse antibodies preferentially recognize A chain:B chain complexes. If the foregoing is true, then “free” B chains must be present in the embryonic heart lysates. It then follows that the rabbit antibodies recognize epitopes on chicken lami- 422 C.D. LITTLE ET AL. BASEMENT MEMBRANE ANTIGENS IN EARLY HEART 423 Fig. 4. Paraffin-embedded stage 19 embryonic chicken hearts were sectioned, deparaffinized, rehydrated, and immunolabeled. a: A DIC image showing a section of atrioventricular myocardium (m), endocardium (e), and intervening ECM. b The strands of ECM material lying between the two cardiac tubes immunolabel with laminin antibodies. c: A DIC micrograph of the ECM within the acellular region of a n embryonic heart similar to the specimen described above. d The nonspecific fluorescence that results when nonimmune IgG and second antibody is applied to the section. Except for the primary antibodies, all conditions, including photography, are identical to the specimen shown in b. x 500. Fig. 3. The atrioventricular region of embryonic hearts shown in this figure was immunolabeled prior to embedding and frozen sectioning. In each panel the primitive endocardial tube (e) and myocardial tube (m) are shown. a: The DIC image of a stage 22 endocardial cushion. Cells derived from the endocardium (arrowhead) have seeded into the region between the primitive endocardium and myocardium. x 600. b Immunostaining with fibronectin antibodies shows brightly labelled filaments within the MBM/CJ, some of which are closely associated with the migratory cells. The walls of the heart tubes are heavily labeled. c: A similar cryosection showing migratory cells (arrowhead). x 500. d The pattern for collagen type I immunostaining shows filamentous strands between the heart endocardium (e) and myocardium (m); some of the filaments are apparently associated with the migratory cells (arrowhead). The myocardium and endocardium are brightly labeled. e: DIC optics shows the endocardium and myocardium of a stage 19 embryonic heart. x 375. f: Laminin immunostaining is observed throughout the MBM/CJ, a punctate granular pattern codistributed with fine wisplike arrays. g: A stage 19 heart that was immunolabeled with collagen type IV antibodies. x 330. h: The collagen type IV staining pattern shows wisplike patches of fluorescence within the MBMICJ. nin B chains which are cryptic when complexed with A chains. Otherwise, one would expect chicken A:B complexes to be precipitated by the rabbit antimouse laminin IgG. That this may be the case is suggested by the fact that the same rabbit antibodies which do not precipitate chicken laminin A:B complexes will precipitate mouse laminin A:B complexes. The band a t approximately 150 kDa precipitated by the mouse monoclonal antibody is very likely entactin, a basement membrane glycoprotein known to interact strongly with laminin (Carlin et al., 1981; Martin and Timpl, 1987). The identity of the band a t 100 kDa present in both the rabbit antimouse laminin immunoprecipitate and the mouse antichicken laminin immunoprecipitate is unknown. It is important to note that the mouse antibody (31-2j passively precipitated several components that are probably not laminin-derived C.D. LITTLE ET AL. 424 polypeptides. Although data describing avian laminin Carlin B., R. Jaffe, B. Bender, and A.E. Chung 1981 Entactin, a novel basal lamina associated sulfated glycoprotein. J . Biol. Chem., are sparse, the bands at approximately 280 kDa and 256:5209-5214. 180 kDa are almost certainly chicken laminin A and B Chen, J.-M., and C.D. Little 1985 Cells that emerge from embryonic exDlants Droduce fibers of type J . Cell Biol., chains, respectively. Chicken laminin subunits are re_ - IV collagen. 161:1175-1 181. ported to be of lower apparent molecular weight than J.-M., and C.D. Little 1987 Cellular events associated with lung the corresponding mouse polypeptide (Bayne et al., Chen, branching morphogenesis including the deposition of collagen 1984). Avian laminin may have unique features; howtype IV. Dev. Biol., 120~311-321. ever, the general characteristics of laminin in organ- Davis, L.A., R.C. Ogle, and C.D. Little 1989 Embryonic heart mesenchymal cell migration in laminin. Dev. Biol., May Vol. 132. isms as diverse as Drosophila (Fessler et al., 1987; S.E., E. Ruoslahti, and M.D. Pierschbacher 1987 A cell surMonte11 and Goodman, 1988), and human (Martin and Dedhard, face receptor complex for collagen type I recognizes the arg-glyTimpl, 1987) appear to be similar. asp sequence. J . Cell Biol., 104:585-593. Embryonic “Basement Membranes” Vs. Embryonic Mesenchymes The data reported here combined with earlier work (Chen and Little, 1985,1987) suggest that laminin and type IV collagen are not restricted in their distribution to conventional embryonic basement membranes. Indeed, our unpublished observations show that laminin and collagen IV are widely distributed in embryonic tissues. This more general distribution of laminin and collagen IV has been observed by others. For example, Solursh and Jensen (1988) detected collagen IV in limb bud mesenchyme. Moody and her colleagues have recently shown a correlation between the presence of laminin in mandibular arch mesenchyme and the guidance pathway of the trigeminal axon in chicken embryos (Riggott and Moody, 1987). Others have reported laminin in cranial mesenchyme (Tuckett and MorrissKay, 1986) and in early ganglia and nerve roots (Rogers et al., 1986). Two structural glycoproteins, laminin and collagen IV, may now be added to the list of components known to be present in the ECM commonly called cardiac jelly. These two glycoproteins are generally thought t o be restricted to conventional basement membranes. Therefore, their presence in this region of the embryonic heart adds weight to the suggestion of Kitten et al. (1987) that cardiac jelly is an expanded basement membrane. The identification of laminin and collagen IV within such anatomically diverse mesenchymes as heart, lung, mandible, and limb demonstrates a widespread distribution of these “basement membrane” components throughout the embryo. As investigators continue to describe the molecular anatomy of tissue ECM, it may well be that the current dichotomous classification system (basement membrane ECM vs. interstitial ECM) will have to be reevaluated with regard to embryonic tissues. ACKNOWLEDGMENTS This work was supported by grants from the Developmental Biology Program, NSF (#85-178161, NIHLB (#37709), and a Basic Research Grant-March of Dimes Birth Defects Foundation (#1-10661, to C.D.L. LITERATURE CITED Bayne, E.K., M.J. Anderson, and D.M. Fambrough 1984 Extracellular matrix organization in developing muscle: Correlation with acetylcholine receptor aggragates. J . Cell Biol., 99~1486-1501. Brown, S.S., H.L. Malinoff, and M.S. Wicha 1983 Connectin: Cell surface protein that binds both laminin and actin. Proc. Natl. Acad. Sci. USA, 80~5927-5930. Buck, C.A., and, A.F. Honvitz 1987 Cell surface receptors for extracellular matrix molecules. Annu. Rev. Cell Biol., 3~179-205. Fessler, L.I., A.G. Campbell, K.G. Duncan, and J.H. Fessler 1987 Drosophila laminin: Characterization and localization. J . Cell Biol., 105~2383-3291. Graf, J., R.C. Ogle, F.A. Robey, M. Sasaki, G.R. Martin, Y. Yamada, and H.K. Kleinman 1987 A pentapeptide from the laminin B1 chain mediates cell adhesion and binds the 76,000 laminin receptor. Biochemistry, 26:6896-6900. Hynes, R.O. 1987 Integrins a family of cell surface receptors. Cell, 48:549-554. Johnson, R.C., F.J. Manasek, W.C. Vinson, and J.M. Seyer 1974 The biochemical and ultrastructural demonstration of collagen during early heart development. Dev. Biol., 36:252-271. Kitten, G.T., R.R. Markwald, and D.L. Bolender 1987 Distribution of basement membrane antigens in cryopreserved early embryonic hearts. Anat. Rec., 21 7:370-390. Kleinman, H.K., M.L. McGarvey, L.A. Liotta, P.G. Robey, K. Tryggvason, and G.R. Martin 1982 Isolation and characterization of type IV procollagen, laminin and heparan sulfate proteoglycan from the EHS sarcoma. Biochemistry, 21:6188-6193. Kleinman, H.K., M.L. McGregor, J.R. Hassell, and G.R. Martin 1983 Formation of a supramolecular complex is involved in the reconstruction of basement membrane components. Biochemistry, 22:4969-4974. Laemmli, U.K. 1970 Cleavage of structural proteins during the asesmblv of the head of bacterioDhaee T4. Nature (Lond.). 227:860-865. Lesot, H., U. Kuhl, and K. von der Mark 1983 Isolation of a lamininbinding protein from muscle cell membranes. EMBO J., 26:861-865. Little, C.D.-T. Chen 1982 Masxing of extracellular collagen and the codistribution of collagen and file ronectin during matxin formulation by cultural embryonic fiberblasts. J . Cell. Sci., 55:35-20. Manasek, F.J. 1977 Structural glycoproteins of the embryonic cardiac extracellular matrix. J . Mol. Cell. Cardiol., 9:425-439. Manasek, F.J., M. Reid, W. Vinson, J . Seyer, and R. Johnson 1973 Glycosaminoglycan synthesis by the early embryonic chick heart. Dev. Biol., 35:332-348. Markwald, R.R., and W.N. Adams-Smith 1972 Distribution of mucosubstances in the developing rat heart. J . Histochem. Cytochem., 29U 1):896 -907. Markwald, R.R., R.B. Runyon, G.T. Kitten, F.M. Funderberg, D.H. Bernanke, and Philip R. Brauer 1984 Use of collagen gel cultures t o study heart development Proteoglycan and glycoprotein interactions during the formation of endocardia1 cushion tissue. In: The Role of Extracellular Matrix in Development. R. L. Trelstad, ed. Alan R. Liss, Inc., New York, pp. 323-350. Martin, G.R., and R. Timpl 1987 Laminin and other basement membrane components. Annu. Rev. Cell Biol., 3:57-85. Mayne, R.M., and, J.G. Zettergren 1980 Type IV collagen form chicken muscular tissues. Isolation and characterization of the pepsin-resistant fragments. Biochemistry, 19:4065-4072. Montell, D.J., and C.S. Goodman 1988 Drosophila substrate adhesion molecule: Sequence of laminin B1 chain reveals a domain of homology with mouse. Cell, 53~463-473. Ogle, R.C., D.M. Piquet, and C.D. Little 1986 Collagen binding proteins associated with the chicken embryonic cell surface. In: Progress in Developmental Biology, Part B. H.C. Slavkin, ed. Alan R. Liss, Inc., New York, pp. 173-176. Ogle, R.C., and C.D. Little 1989 Collagen binding proteins derived from the embryonic-fibroblast cell surface recognize arginine-glycine-aspartic acid. Biosci. Rep., in press. Riggott, M.J., and S.A. Moody 1987 Distribution of laminin and fibronectin along peripheral trigeminal axon pathways in the developing chick. J. Comp. Neurol., 258:580-596. Rogers, S.L., K.J. Edson, P.C. Letourneau, and S.C. McLoon 1986 Distribution of laminin in the developing peripheral nervous system of the chick. Dev. Biol., 113:429-435. . I BASEMENT MEMBRANE ANTIGENS IN EARLY HEART Rubin, K., D. Gullburg, T.K. Borg, and B. Obrink 1986 Hepatocyte adhesion to collagen: Isolation of membrane glycoproteins involved in adhesion to collagen. Exp. Cell Res., 164r127-138. Ruoslahti, E., and M.D. Pierschbacher 1987 New perspectives in cell adhesion: RGD and integrins. Science, 238r491-498. Saint-Marie, G. 1962 A parafin embedding technique for studies employing immunofluorescence. Methods Enzymol., 82:803-831. Solursh, M., and K.L. Jensen 1988 Accumulation of basement membrane components during the onset of chondrogenesis and myogenesis in the chick wing bud. Development, 104:41-49. 425 Terranova, V.P.,C.N. Rao, T. Kalebic, I.M. Margulies, and L.A. Liotta 1983 Laminin receptor on human breast carcinoma cells. Proc. Natl. Acad. Sci. USA, 80r444-448. Tuckett, F., and G.M. Morriss-Kay 1986 The distribution of fibronectin, laminin and entactin in the neurulating rat embryo studied by indirect immunofluorescence. J. Embryol. Exp. Morphol., 94r95-112. Underhill, C.B., A.L. Thurn, and B.E. Lacy 1985 Characterization and identification of the hyaluronate binding site from membranes of SV-3T3 cells. J . Biol. Chem., 260:8128-8133.