the of biological chemistry vol. no. march pp. in u.s.a ...€¦ · from the departments of...

THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1986 by The American Society of Biological Chemists, Inc.

Vol. 261, No. 9, Issue of March 25, pp. 4328-4336,1986 Printed in U.S.A.

Macromolecular Organization of Basement Membranes CHARACTERIZATION AND COMPARISON OF GLOMERULAR BASEMENT MEMBRANE AND LENS CAPSULE COMPONENTS BY IMMUNOCHEMICAL AND LECTIN AFFINITY PROCEDURES*

(Received for publication, August 16, 1985)

Panaiyur S. Mohan and Robert G. SpiroS From the Departments of Biological Chemistry and Medicine, Harvard Medical School and the Elliot P. Joslin Research Laboratory, Boston, Massachusetts 02215

The macromolecular components of bovine glomerular basement membrane (GBM) and lens capsules (anterior and posterior) solubilized by sequential extractions with denaturing agents were quantitated and characterized by polyacrylamide gel electrophoresis, CL-GB filtration, and DEAE-cellulose chromatography with the help of immunochemical techniques. Laminin, entactin, fibronectin, and heparan sulfate proteoglycan were primarily recovered (over 80%) from both basement membranes in a guanidine HCl extract which contained only a limited amount of the total protein (10-14%); most of the remainder of these noncollagenous components could be solubilized by the guanidine in the presence of reducing agent. Although a portion of the Type IV collagen could be obtained by these treatments, effective extraction of this protein de- pended on exposure to sodium dodecyl sulfate under reducing conditions. Immunoblot analysis revealed a remarkably similar pattern for GBM and lens capsule Type IV collagens with prominent bands of M, = 390,000,210,000, and 190,000 being evident. Fibro- nectin was present in much greater amounts in GBM than lens capsule while the reverse was true for entactin. In both GBM and lens capsules, the entactin (M, = 150,000) exceeded laminin; the latter protein on immunoblotting was found to contain primarily the a- subunit (M, = 200,000). The size of the heparan sulfate proteoglycan from anterior (M, = 400,000) and posterior lens capsule (M,. > 500,000) was substantially larger than that from GBM (M, = 200,000). During DEAE-cellulose chromatography under nonreducing conditions in a denaturing solvent, a portion of the Type IV collagen coeluted with the proteoglycan from these membranes.

Considerable Bandeiraea simplicifolia I binding activity (a-D-galactose specific) was observed in GBM and lens capsule extracts and column fractions which could not be accounted for by laminin alone. Several components which reacted with this lectin were seen on transblots and among these Type IV collagen was identified. In contrast to the basement membranes from bovine tissues, the constituents from human GBM did not react with the B. simplicifolia I lectin.

It is now generally recognized that basement membranes

* This work was supported by Grant AM 17325 from the National Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

3 To whom correspondence should be addressed Elliot P. Joslin Research Laboratory, One Joslin Place, Boston, MA 02215.

are multicomponent expressions of the extracellular matrix in which collagenous and noncollagenous glycoconjugates are assembled to provide filtration barriers and substrata for cell attachment and orientation. Although a number of these macromolecules including laminin, entactin, fibronectin, heparan sulfate proteoglycan, and Type IV collagen have been characterized (l), many of the most revealing studies have been carried out on basement membrane-like matrices from the Engelbreth-Holm-Swarm tumor (2, 3) or cells in culture (4-10) rather than on the more well-defined tissue-derived basement membranes. Consequently information in regard to the macromolecular organization of these mature morpholog- ical entities is still quite limited and is to a large measure based on immunomicroscopic approaches (11-17).

In the present investigation we have focused on two well- known basement membranes with quite different morpholog- ical and functional properties, namely bovine GBM’ and lens capsule, which have been the subject of previous studies in our laboratory (18-21). Sequential extraction procedures permitted us to characterize and quantitate the major cunstitu- ents of these basement membranes and explore their inter- relationships with the help of immunochemical techniques. While our findings indicate a fundamentally similar architec- ture on both basement membranes, substantial differences were noted in the relative proportion in which the macromolecular components occur as well as in nature of the proteoglycan component present. Of particular interest was our obervation that the Bandeiraea simplicifolia I’ reactivity reported for the basement membranes of some species (22) appears to be due to the occurrence of a-D-galactosyl residues not only in laminin (23) but also in other glycoconjugates including Type IV collagen.

EXPERIMENTAL PROCEDURES

Preparation of Basement Membranes-Previously described procedures were employed for the preparation of bovine GBM (24, 25) and calf lens capsules (19). Both anterior and posterior capsules were isolated with average dry weights, respectively, of 0.55 and 0.25 mg/ capsule.

Sequential Extraction of Basement Membranes-All extractions of the basement membranes and subsequent fractionations were carried out in the presence of a mixture of protease inhibitors consisting of 1 mM phenylmethylsulfonyl fluoride, 1 mM benzamidine HC1,0.4 mM N-ethylmaleimide, and 0.01 M p-chloromercuriphenyl sulfonic acid.

In the first step of the extraction procedure, GBM (10 mg/ml) as well as anterior and posterior lens capsules (respectively, 12 and 20 capsules/ml) were suspended in 5 M guanidine HCl, 0.05 M sodium

The abbreviations used are: GBM, glomerular basement membrane; DTT, dithiothreitol; SDS, sodium dodecyl sulfate; ME, 2- mercaptoethanol; PBS, 0.05 M sodium phosphate buffer, pH 7.4, containing 0.15 M NaCl; BSA, bovine serum albumin.

* This lectin is also known as Griffonia simplicifolia I.

4328

Macromolecular Organization of Basement Membranes 4329

phosphate, pH 7.0, and stirred with a small magnet for 3 h a t room temperature. The residue obtained after centrifugation (700 X g for 20 min at 4 "C) was re-extracted with the same volume of the guanidine HCl reagent for 18 h at 4 "C, and after centrifugation the two supernatants were combined and designated as Extract 1. Sub- sequently the insoluble basement membrane material was stirred at the same concentration with the guanidine HC1 reagent containing 20 mM DTT for 3 h a t room temperature, and the supernatant obtained from this treatment was referred to as Extract 2. After washing the residue with PBS, it was suspended (5 mg GBM/ml; 12 anterior capsules/ml; 20 posterior capsules/ml) in 0.05 M sodium phosphate buffer, pH 7.0, containing 2% (w/v) SDS, 5% (v/v) ME, and heated with stirring at 100 "C for 60 min; this yielded a solubilized fraction (Extract 3) which was centrifugally separated from a small amount of residual material. In some experiments an extraction step with buffered 2% SDS (18 h at room temperature) was employed prior to the SDS/ME treatment.

Gel Filtration of Basement Membrane Components-The guanidine HC1-solubilized material (Extract 1) from GBM and lens capsule were fractionated at 4 "C on a Sepharose CL-GB column (1.5 X 80 cm) equilibrated and eluted with 4 M guanidine HCI in 0.05 sodium phosphate, pH 7.0. A flow rate of 12 ml/h was maintained, and 4-ml fractions were collected. Aliquots of these fractions were analyzed for various components by solid phase radioimmunoassay.

DEAE-cellulose Chromatography of Basement Membrane Compo- nents-After extensive dialysis a t 4 "c against 8 M urea in 0.02 M sodium phosphate, pH 7.0, the guanidine HCl extracts from GBM and lens capsules were chromatographed on a column (1 X 6 em) of DEAE-cellulose (DE-52 microgranular). After application of the sample, 30 ml of the equilibrating buffer (8 M urea, 0.02 M sodium phosphate, pH 7.0) was passed through the column followed by the same volume of an 8 M urea, 0.02 M sodium phosphate, pH 7.0-0.05 M NaCl solution. Further elution was then achieved with 120 ml of a linear concentration gradient in which the NaCl concentration in the 8 M urea, 0.02 M sodium phosphate, pH 7.0 buffer, was raised from 0.05 to 0.5 M. A flow rate of 12 ml/h was maintained, and 3-ml fractions were collected from which aliquots were taken for solid phase radioimmunoassays. Designated tubes were pooled for further study and brought to a small volume in PBS by repeated centrifuga- tions in a Centricon-10 microconcentrator (Amicon).

Digestion with a-Galactosidase-Guanidine HCl- (Extract 1) and guanidine/DTT (Extract 2)-solubilized basement membrane components, after dialysis, were incubated in 125 pl of 0.1 M sodium phosphate buffer, pH 6.5, with coffee bean a-galactosidase (1 unit) for 6 h at 37 "C; the reaction was terminated by heating at 100 "C for 3 min.

Collagenase Digestion-Guanidine HCl/DTT-extracted proteins, after dialysis, were incubated with 20 pg collagenase (Clostridium histolyticum, Sigma, Type VII) in 400 p1 of 0.1 M Tris/acetate, pH 7.4, buffer containing 5 mM calcium acetate for 80 min at 37 "C; the digestion was stopped by the addition of EDTA. Control incubations with BSA (25 pg) and carp swim bladder collagen (25 pg) were carried out to monitor the specificity of the collagenase action.

Source of Antisera-Rabbit antiserum against bovine plasma fibronectin was a gift of Drs. E. Ruoslahti and E. Engvall (La Jolla Cancer Research Foundation) while rabbit anti-mouse entactin was kindly provided by Dr. A. Chung (University of Pittsburgh), and rabbit anti- mouse laminin was purchased from Bethesda Research Laboratories. The following antigens (20-100 pg) were used to prepare rabbit antiserum in our laboratory by a schedule of multiple intradermal injections (26): Type IV collagen from bovine GBM (20), heparan sulfate proteoglycan purified from bovine GBM (21), and calf proteoglycan prepared by DEAE-cellulose chromatography of guanidine HCI extract from calf lens capsules. The specificity of the antisera was checked by solid phase radioimmunoassays as well as by immunoblotting of electrophoretically separated antigens.

titated by a procedure in which flat-bottom microtiter wells (Immulon Solid-phase Radioimmunoassays-The various antigens were quan-

2, Removawells, Dynatech R/D Co.) were coated with 100 pl of sample or standard antigen and kept at 4 "C overnight. After the wells were rinsed 5 times with 250 pl of PBS containing 0.05% Tween 20 and 0.1% BSA, antiserum (100 pl usually diluted 1:lOOO in the PBS/ Tween/BSA solution) was added and permitted to interact for 2 h a t room temperature. This was followed by 5 washes of the wells with PBS/Tween/BSA and the addition of '251-laheled protein A (1 X lo5 cpm, 3-5 ng), radioiodinated by the chloramine-T procedure (27) in the presence of 1 mM methyl-a-D-galactoside (28), in 100 p1 of the

PBS/Tween/BSA solution. After a 2-h incubation at room temperature, the unbound protein A was removed by washes, and the wells were assayed for radioactivity in a gamma counter (model 1290 Gamma Trac). Concentration curves were constructed for the following antigens: bovine GBM Type IV collagen and heparan sulfate proteoglycan (prepared in our laboratory), murine entactin (a gift of Dr. A. Chung, University of Pittsburgh), mouse laminin (a gift of Drs. H. Kleinman and G. Martin, NIH), and bovine plasma fibronectin (purchased from Bethesda Research Laboratories). All samples and standards were diluted with PBS prior to addition to the wells; at the dilutions used no interference with protein binding was caused by the guanidine HCl, guanidine HCI/DTT, SDS/ME, and urea reagents used in this study (see "Res~lts"). The protein content of all assayed samples was kept within the binding capacity of the wells (-300 ng), and multiple assays of each sample assured that analysis was performed in the linear range for the antigen being determined. Control incubations which were routinely carried out with appropri- ately diluted preimmune rabbit serum indicated that nonspecific binding of the lZ5I-labeled protein A was negligible.

Solid-phase Lectin Binding Assay-Interaction of basement membrane proteins with '9-labeled B. simplicijolia I (1.5 X lo4 cpm/ng) was determined in the microtiter wells by a procedure previously described (28).

Polyacrylamide Gel Electrophoresis-Electrophoresis was performed in SDS on vertical polyacrylamide slab gels (3-mm thick) according to the procedure of Laemmli (29). The separating gel which consisted of a linear 4-15% acrylamide gradient was overlaid by a 4% stacking gel. Prior to being applied to the gel, the samples were incubated at 37 "C for 60 min in 80 mM Tris/chloride pH 6.8 buffer containing 4% SDS, 5% ME, and 1.7 mM phenylmethylsulfonyl fluoride followed by heating for 3 min at 100 "C. The proteins were visualized by staining of the gels with Coomassie Blue or transferred electrophoretically to nitrocellulose sheets.

Electrophoretic Transfer Blotting and Reaction with Antibodies and Lectin-Proteins resolved by electrophoresis were transferred to two nitrocellulose sheets in an apparatus supplied by Bio-Rad Laborato- ries (0.5 A for 8 h at 4 "C) according to the procedure of Towbin et al. (30). Immunological identification of the proteins was carried out at room temperature on the first nitrocellulose sheet with various antisera and radiolabeled protein A (31). After exposure for 30 min to PBS, 0.1% BSA, the nitrocellulose sheet was incubated with antiserum (usually diluted 1500) in the same buffer for 2 h. Subse- quently the sheet was washed 5 times with a PBS, 0.1% BSA, 0.05% Tween 20 solution and then exposed to '251-labeled protein A (1 X lo6 cpm/ml) in the PBS/BSA/Tween for 90 min. After several washes bound radioactivity was detected by autoradiography using X-Omat AR film (Eastman Kodak). Control incubations which were performed with preimmune rabbit serum on the various electrophoretically separated basement membrane fractions demonstrated an absence of radiolabeled components after exposure to the protein A.

For the detection of B. simplicijolia I binding proteins, the nitrocellulose sheets, after exposure to the PBS/O.l% BSA solution, were incubated with the lectin (obtained from E. Y. Laboratories and radioiodinated by the chloramine-T procedure) a t a concentration of 150 ng/ml (2.5 X lo6 cpm/ml) for 18 h at 4 "C; radioautography was then carried after repeated washing of the sheets.

The location of standard proteins was determined by India ink staining (32) of the second nitrocellulose sheet from the transblotted electrophoresis gels.

Chemical Analysis-Total protein was estimated by modification (33) of the Lowry procedure. Amino acid analyses were carried out on the Technicon NC-2 amino acid analyzer after hydrolysis of the samples in sealed tubes under nitrogen with constant boiling HCI for 24 h at 105 "C (21).

RESULTS

Sequential Extraction and Quantitatwn of Basement Mem- brane Macromolecular Components-Guanidine HC1 extraction (Extract 1) of GBM and anterior lens capsule led to a limited solubilization to protein (Fig. 1) which accounted for 14.0 and 12.5 mg/100 mg of dry weight, respectively. Subse- quent treatment with guanidine HC1 in the presence of DTT extracted an additional similar amount of protein (Extract 2), and the residue could then be solubilized to a large extent with SDS/ME (Extract 3). In contrast even prolonged treat-

4330 Macromolecular Organization of Basement Membranes

ment of the residue from the guanidine HCl/DTT extraction with SDS in the absence of reducing agent brought only an insignificant amount of protein into solution (GBM, 0.1 mg/ 100 mg; anterior lens capsule, 0.7 mg/100 me).

The major macromolecular components present in the basement membrane extracts could be quantitated by solid phase radioimmunoassay under conditions in which the extracting reagents did not interfere (Fig. 2). Although our assays were carried out at substantial dilutions (Fig. 2), we observed with standard antigens that reagents in concentrations up to the following limits were compatible with the assay: 2 M guanidine

ALC GEM A L C GEM

fl, 0 T 7 100

100 r

1 2 3

T 7 100 .. .

I 2 3 I 2 3 1 2 3

E X T R A C T

FIG. 1. Distribution of macromolecular components, protein, and B. simplicifolia I binding activity in sequential extracts of basement membranes. Calf anterior lens capsules (ALC) and bovine glomerular basement membrane (GBM) were treated in sequence with 5 M guanidine HCl (Extract l ) , 4 M guanidine HCl, 20 mM DTT (Extract 2), and 2% SDS, 5% ME (Extract 3) as described under “Experimental Procedures.” The bar graphs indicate the percentage distribution of laminin, heparan sulfate proteoglycan, fibronectin, collagen, and E. simplicifolia I binding activity as measured by solid phase radioassays (see “Experimental Procedures”). The protein contents of extracts 1, 2, and 3 expressed as mg/100 mg of basement membrane were 12.8, 8.6. and 67.2 and 14.0, 15.5, and 50.8 for ALC and GBM, respectively.

FIG. 2. Effect of increasing amount of guanidine HCl and SDS/ ME extracts of GBM on solid phase radioimmuno and lectin binding assays. The solid phase assays were carried out on variously diluted Extract 1 (left panel) and Extract 3 (right panel) as described under “Experimental Pro- cedures,” and the amount of protein as well as the concentrations of guanidine or SDS in the 100 gl applied to the wells is indicated. The lZ5I-labeled protein A bound to samples which had bgen exposed to antisera against proteoglycan (PG), fibronectin ( F N ) , laminin ( L M ) , and collagen (CL) as well as preimmune serum (PI) is shown. The lectin binding assay was carried out with ‘251-labeled B. simplicifolia I (ES-I) .

HC1; 2 M guanidine HCl, 10 mM DTT, 4 M urea; and 0.01% SDS, 0.025% ME. The linearity of the assay for all antigens measured was maintained up to at least 20 ng, and at that level addition of up to 300 ng of BSA did not interfere with their binding to the immunowells (data not shown).

The immunochemical measurements indicated an uneven distribution of the basement membrane components in the sequential extracts (Fig. 1) with most (over 80%) of the noncollagenous material being present in the guanidine HC1 fraction (Extract l), while the collagen was recovered primarily in the SDS/ME-solubilized material (Extract 3). The inclusion of DTT in the guanidine HC1 reagent resulted in the additional solubilization (2-15%) of the laminin, heparan sulfate proteoglycan, and fibronectin so that less than 10% of these proteins remained in the residue from this extraction; a similar pattern of solubilization was observed for entactin (data not shown). While only small amounts of collagen (less than 5%) were extracted with guanidine HC1 alone, the guanidine/DTT reagent brought a more substantial amount of this component (GBM, 30%; lens capsule, 15%) into solution (Fig. 1).

The extraction profile of posterior lens capsule was in essence indistinguishable from that of the anterior (data not shown).

While B. simplicifolia I reactive glycoconjugates were to a large extent solubilized by the guanidine HC1 treatment (Ex- tract I), components which bound to this lectin were also recovered in the subsequent extracts, and, indeed in the case of the lens capsule, Extracts 2 and 3 accounted for, respectively, 26% and 19% of this activity (Fig. 1).

The macromolecular composition of the guanidine HC1 extracts from GBM and anterior as well as posterior lens capsules (Table I) provided a useful basis for comparing these basement membranes. The fibronectin content of GBM was strikingly greater than that of the lens capsules in contrast to the entactin, and to a lesser extent laminin, for which values were higher in the lens capsule. Proteoglycan which consti- tuted the major component measured in the guanidine HC1 extract was present in comparable amounts, while more GBM than lens capsule collagen was solubilized by this reagent.

Identification of Collagen and Glycoprotein Components by Immunoblotting and Gel Filtration-Polyacrylamide gel electrophoresis in SDS of unextracted GBM and anterior as well as posterior lens capsules revealed, on the basis of their reaction with Coomassie Blue, a large number of protein components ranging in apparent M, from >400,000 to C29,OOO

20 -

- 16-

.-I

m m

12 -

8 - - 0

c3 4 - a -

0 -

- 8 - 5 - X -

- 6 4 -

- - 4 t

2 3 -

- 2 - a

1 I - - 2 z

- 0 o 1 I I I I I I 0 40 80 120 160

PROTEIN tng)

I I I t I I

0 2 4 6 8 i O GUANIDINE (mM)

0 30 60 90 120 PROTEIN ( n q )

L I I I I ,

0 1 2 3 4 SDS(pg/ml)


TABLE I Content of several macromolecular components in guanidine HC1

extracts of bovine GBM and calf anterior and posterior lens capsules Analysis was performed on 5 M guanidine HCl solubilized fraction

(Extract 1) by solid phase radioimmunoassays as described under “ExDerimental Procedures.”

Basement Type IV Lami- Entac- membrane collagen proteoglycan nin

tin Fibronectin

&I100 m a

Glomerular 910 3040 155 228 1130 Anterior lens 380 4370 365 760 14 Posterior lens 270 2380 218 1020 49 ”Values were determined with the appropriate antiserum from

standard curves of the antigens specified under “Experimental Pro- cedures” which were of bovine origin except for the murine laminin and entactin. Each component in the extracts is reported per 100 mg of dry basement membrane; the protein content on the guanidine HCI extracts of GBM, anterior lens capsule, and posterior lens capsule was, respectively, 14.0 mg, 12.8 mg, and 10.4 mg/100 mg basement membrane.

TOTAL EXTRACT RESIDUE ’ A C P C G I ‘ A C P C G ‘ ‘AC PC G ’

400 K -

200 K -

l l 6 K -

6 6 K -

4 5 K - 2 9 K -

FIG. 3. Comparison of the electrophoretic patterns of bovine glomerular basement membrane (G) and calf anterior (AC) as well as posterior (PC) lens capsules. The total membranes (-120 pg of protein), as well as their guanidine/HCl extracts (-70 pg of protein) and residues (-120 pg of protein), were submitted to polyacrylamide gel electrophoresis as described under “Experimen- tal Procedures,” and the components were visualized with Coomassie Blue. The designated molecular weight markers were: mouse laminin, /+subunit (400,000) and 0-subunit (200,000); Escherichia coli 0-galactosidase (116,000); bovine serum albumin (66,000); hen ovalbumin (45,000); and bovine erythrocyte carbonic anhydrase (29,000).

(Fig. 3). Examination of the material remaining insoluble after guanidine HC1 treatment presented a simpler pattern with three major components (Mr 390,000, 210,000, and 190,000) being evident (Fig. 3). The guanidine HC1 extracts on the other hand contained a large array of proteins most prominent among which was a sharp band with an M , of 150,000 (Fig. 3). Furthermore, the guanidine extract of each membrane contained several components of M, greater than 200,000 as well as doublet which migrated to the M, = 90,000 position; the multiple smaller weight polypeptide components observed in the lens capsules, representing crystallins (34), were recovered in the guanidine HCI-soluble fraction while prominent components with M , of 190,000 and 53,000 were seen in the GBM extract.

An immunoblot with anti-Type IV collagen serum (Fig. 4) revealed a remarkably similar pattern for all the basement

TOTAL RESIDUE EXTRACT ’ A C PC G ” A C PC G I ‘ AC PC G ’

4 0 0 K - - 2 0 0 K - -- J ~h =

l l 6 K - 6 6 K -

4 5 K - 29K -

FIG. 4. Immunological identification of Type IV collagens in electrophoretically resolved bovine glomerular basement membrane ( G ) and calf anterior (AC) as well as posterior (PC) lens capsules. The total basement membranes as well as the guanidine HC1 extracts and residues were electrophoresed as in Fig. 3. The resolved proteins were then transferred to a nitrocellulose sheet and reacted with anti-Type IV collagen serum as described under “Exper- imental Procedures.” Bound antibody was detected with lZ5I-labeled protein A and visualized by autoradiography; the molecular weight markers were the same as Fig. 3.

membranes and indicated that the 390,000, 210,000, and 190,000 molecular weight bands seen in the Coomassie Blue- stained gels (Fig. 3) are the major collagen components. Indeed, as might be anticipated from the sequential extraction data (Fig. l), the pattern revealed by anti-collagen immunoblotting was strikingly similar to that observed in the stained guanidine HCl residue fractions. It is evident however that the antibodies reacted more favorably with the M, = 190,000 collagen chains than did the Coomassie Blue dye (cf. Figs. 3 and 4) resulting in a quite different impression of the relative proportion of the 210,000 and 190,000 collagen components. As suggested by the solid phase radioimmunoassays, the guanidine HCl extracts also contained some components which reacted with the anti-collagen antibodies, and the most prominent of these co-migrated with the 210,000 and 190,000 molecular weight bands seen in the residue fraction. The higher molecular weight ( M , = 390,000) band however was not apparent in the guanidine HCl extract suggesting a dif- ferential solubility of the Type IV collagen forms. Moreover, in the GBM extract the M , = 190,000 component appeared to be more prominent than the 210,000 band, in contrast to the pattern observed in the residue, and a relatively large proportion of the collagenous material migrated as smaller molecular weight polypeptides (Fig. 4).

Reaction of the nitrocellulose sheets containing transferred proteins from electrophoretically resolved guanidine HC1 extract with anti-fibronectin serum revealed one major component (Mr = 240,000) in the anterior and posterior lens capsules; in the GBM a number of prominent bands were seen in addition to the 240,000 molecular weight component, and the pattern was similar to that of the plasma standard (Fig. 5). The electrophoresis confirmed the relatively high content of this protein in GBM observed by radioimmunoassay (Table I).

Immunoblotting with anti-laminin of electrophoretically separated guanidine HCI-solubilized proteins visualized a major band in both lens capsule and GBM extracts which co- migrated with the a-subunit ( M , = 200,000) of mouse laminin (Fig. 6). However, in contrast to the mouse protein (351, the @-subunit ( M , = 400,000) was not evident in GBM and present only in small amounts in lens capsule; furthermore, the lens capsule @-component on repeated examination appeared to

4332 Macromolecular Organization of Basement Membranes AC PC G FN

4 0 0 K - 2 0 G K -

116K -

-0

6 6 K - 4 5 K - 2 9 K -

FIG. 5. Immunological identification of fibronectin after electrophoresis of guanidine HCl extracts of bovine glomerular basement membrane (G) and calf anterior (AC) as well as posterior (PC) lens capsules. Electrophoresis of the extracts as well as standard bovine plasma fibronectin ( F N ) was carried out as in Fig. 3, and this was followed by immunoblotting with anti-fibronectin serum as described under “Experimental Procedures.” The components were visualized on the nitrocellulose sheet by autoradiography after exposure to 1Z51-labeled protein A; the molecular weight markers were the same as in Fig. 3.

LAMININ ENTACTIN ANTI- ANTI-

‘ L M AC G I ‘ A C G ‘ 4 0 0 K - 2 0 0 K -

l l 6 K -

6 6 K - 4 5 K - 2 9 K -

collagen, laminin, fibronectin, and entactin (Fig. 7). The collagen and a portion of the fibronectin appeared in the void volume, followed in order by laminin (K,, = 0.087), a major fibronectin peak (Ka, = 0.22), and entactin (Kav = 0.35). A second anti-laminin reactive peak coincided with the entactin and is presumed to represent the lack of specificity already noted for this antiserum in the immunoblots (Fig. 6).

A similar elution pattern was observed (data not shown) for these components upon filtration of a guanidine extract from anterior lens capsule with the exception that most of the fibronectin emerged in an early (Kav = 0.087) rather than in the more retarded ( ICav = 0.22) peak.

Characterization of Proteoglycan Components by Zmmuno- blotting and Gel Filtraton-The electrophoretic migration of the heparan sulfate proteoglycan from lens capsules and GBM was determined by immunoblotting with anti-bovine GBM proteoglycan serum (Fig. 8). A single band ( M , = 200,000) was apparent in the guanidine HC1 extract from the GBM, similar in appearance to that previously observed by staining procedures (21), which moved to the same position as the purified GBM component (Fig. 8). The lens capsule proteoglycans were visualized as broad bands with substantially higher molecular weights; indeed the anterior and posterior capsules differed in the size of their proteoglycans with re- spective molecular weights of 400,000 and >500,000 (Fig. 8).

Gel filtration under nonreducing conditions of the guanidine extracts further indicated there was a size difference between the lens capsule and GBM proteoglycans (Fig. 9); the anterior lens glycoconjugate eluted earlier (K., = 0.087) from

FIG. 6. Immunological identifications of laminin and entactin after electrophoresis of guanidine HCl extracts of calf anterior lens capsule (AC) and bovine glomerular basement membrane (G). Immunoblotting with anti-laminin or anti-entactin sera was carried on the electrophoretically resolved extracts as well as murine laminin ( L M ) as described under “Experimental Proce- dures.” After treatment with 1251-labeled protein A, the components on the nitrocellulose sheets were detected by autoradiography. The electrophoretic conditions and molecular weight markers were the same as in Fig. 3. The immunoblot of posterior lens capsule was similar to that of anterior capsule.

have a molecular weight ( M , = 360,000) somewhat lower than that of the murine protein (Fig. 6). The anti-laminin serum also revealed in both lens capsule and GBM a faint component which migrated to the position of entactin (Mr = 150,000) suggesting that the laminin preparation used for immuniza- tion contained small amounts of this protein.

Indeed the anti-entactin immunoblot clearly defined the 150,000 molecular weight component (Fig. 6) which was also distinctly seen in the Coomassie Blue-stained gels of the lens capsule and GBM guanidine HC1 extracts (Fig. 3). The anti- entactin also bound to a lower molecular weight protein ( M , = 90,000) which corresponded to a band revealed by Coomas- sie Blue stain (Fig. 3) and might be equivalent to one of the faster anti-entactin reactive components observed in murine membrane sacs (6). The anti-entactin serum however did not react with the laminin (Fig. 6).

Filtration of the guanidine HCl extract of GBM on Sepha- rose CL-GB in 4 M guanidine HCl under nonreducing conditions resulted in the elution of immunoreactive Type IV

9 -

6 -

3 -

0 -

- 3

- 2

- I

- 0

l C 8 P r n t ~ ~ I 0 IO 20 30

TUBE NUMBER (4rnl)

FIG. 7. Filtration on Sepharose CL-GB of guanidine HCl extract of bovine glomerular basement membrane. The sample from 6 mg of GBM was placed on a column (1.5 X 82 cm) equilibrated with 4 M guanidine HCI in 0.05 M sodium phosphate, pH 7.0, and elution of the indicated proteins as followed by solid phase radioimmunoassay in which Iz5I-labeled protein A was used to quantitate the bound antibody as described under “Experimental Procedures”; values are expressed per ml of each fraction. The void volume (V,) and the total volume (V,) of the column as well as the elution position of standard IgG ( M , = 150,000) are indicated; calf thyroglobulin ( M , = 670,000) was excluded from the column.


AC P C G PG

4 0 0 K -

200K -

116K - 6 6 K -

4 5 K - 2 9 K -

FIG. 8. Immunological identification of heparan sulfate proteoglycan in electrophoretically resolved guanidine HCl extracts of bovine glomerular basement membrane (G) and calf anterior (AC) as well as posterior (PC) lens capsules. After electrophoresis was performed on the extracts as well as on the purified GBM proteoglycan (PC), the components were transferred to a nitrocellulose sheet and reacted with anti-bovine GBM heparan sulfate proteoglycan serum as described under “Experimental Proce- dures.” Bound antibody was detected with lZ5I-labeled protein A and visualized by autoradiography. Electrophoresis and molecular weight markers were the same as in Fig. 3.

2 -

l -

0 -

0 IO 20 30

TUBE NUMBER ( 4 m l )

FIG. 9. Comparison of proteoglycan elution from lens capsule and glomerular basement membrane during Sepharose CL-GB filtration. The guanidine extracts from 3 mg of anterior lens capsule and 6 mg of bovine GBM were chromatographed sepa- rately on a column (1.5 X 82 cm) equilibrated with 4 M guanidine HCl in 0.05 M sodium phosphate, pH 7.0. The column fractions were analyzed by solid phase radioimmunoassay employing antiserum against lens capsule and GBM proteoglycan, respectively. Radioiodi- nated protein A was used to quantitate bound antibody, and the values are expressed per ml of each fraction. The designations for V,,, V, and ZgG are as in Fig. 7.

a Sepharose CL-GB column than that from GBM (KaV = 0.30). DEAE-cellulose Fractionation of Basement Membrane Com-

ponents-Chromatography on DEAE-cellulose provided an- other means for resolving the components present in the guanidine HC1 extracts of the basement membranes. The GBM extract yielded three major peaks during this fractionation which represented components which were not retained (peak A), eluted with 0.05 M NaCl (peak B), and emerged with the NaCl gradient (peak C) (Fig. 10). As indicated in Table I1 the proteoglycan, eluted almost exclusively in peak C, while the laminin, entactin, and fibronectin were distributed in various ratios in peaks A and B. While most of the

304 0.5

0 10 20 30 40 50

FIG. 10. Chromatography on DEAE-cellulose of guanidine extract of bovine glomerular basement membrane. The sample originating from 92 mg of GBM was dialyzed against 8 M urea, 0.02 M sodium phosphate, pH 7.0, before being placed on a column (1 X 6 cm) of DE-52 equilibrated with the same buffer. At tube 11 elution was carried with 8 M urea in 0.02 M sodium phosphate, pH 7.0, containing 0.05 M NaC1, and at tube 22 a linear NaCl concentration gradient (0.05-0.5 M) was started as described under “Experimental Procedures.” The elution of indicated components was monitored by solid phase radioimmunoassays in which 1251-labeled protein A was used to quantitate bound antibody, and the values are expressed per ml of each tube. Fibronectin, entactin, and B. simplicifolia I binding were also analyzed (elution not shown). Lettered areas designate tubes which were combined for further study, and the analyses of these pools are reported in Table 11.

TUBE NUMBER ( 3 m l )

TABLE I1 Distribution in DEAE-cellulose peaks of guanidine HCI-solubilized

components and B. simplicifolia I binding activity from bovine GBM and calf lens capsules

peaka gen nectin nin tin glycan binding

G L G L G L G L G L G L

Colla- Fibro- Lami- Entac- Proteo- Lectin

”~”-

Per centC A 51 42 22 14 28 15 49 37 1 4 17 7 B 22 24 78 86 72 85 51 63 3 2 53 42 c 2 1 3 4 - d - - - - - 96 94 30 51

a Peaks designated in Fig. 10 for glomerular basement membrane (C); a similar fractionation was carried out on anterior lens capsule (L) guanidine extract (elution not shown).

* Lectin bindings refers to B. simplicifolia I. e Values indicate per cent of component or activity present in peaks

from DEAE-cellulose columns; assays were performed by solid phase assays as described under “Experimental Procedures.”

Dash indicates component could not be detected.

Type IV collagen was present in peak A, substantial amounts were also found in the two later peaks; indeed, immunoreactive collagen was the only component which emerged along with the proteoglycan during the salt gradient. Analysis for B. simplicifolia I lectin binding indicated that this activity was to a large measure distributed between peaks B and C (Table 11).

Fractionation of anterior lens capsule guanidine HCI extract on DEAE-cellulose resulted in a similar elution pattern, and the distribution of components is reported in Table 11.

Characterization of B. simplicifolia I Reactive Components- Although laminin is known to react with B. simplicifolia I (23), our findings of substantial binding activity for this lectin in fractions of the basement membranes which contain little or none of this protein (see Fig. 1 and Table 11) motivated us to further characterize the specificity of the interactions and the macromolecular components involved.

The considerable B. simplicifolia I binding activity of the


GBM and lens capsule guanidine HC1 extracts, which was equivalent to the reaction with about 5 mg of murine laminin/ 100 mg basement membrane could be effectively inhibited by methyl-a-D-galactoside (Table 111). Moreover a-galactosidase treatment of the basement membrane components brought about a pronounced reduction (90% or more) in their B. simplicifolia I binding capacity (Table 111).

When the electrophoretically resolved proteins from calf anterior lens capsule and bovine GBM guanidine HC1 extracts were reacted with ‘251-labeled B. simplicifolia I after transfer to a nitrocellulose sheet a number of strongly reactive components were visualized (Fig. 11). The guanidine HCl extract from human GBM (kindly made available by Dr. H. Shimo- mura of this laboratory) in contrast to the bovine GBM, did not show any components which reacted with the lectin although loaded on the gel in a similar amount. Indeed the human extract when assayed by the solid phase assay was

TABLE I11 Specificity of B. simplicifolia I binding to glycoconjugates i n

guanidine HC1 extracts of bovine GBM and calf anterior and posterior lens capsules

Binding was determined with ‘T-labeled B. simplicifolia I by a solid uhase assay as described under “Experimental Procedures.”

Basement membrane

Sample treatment”

None CH.-ol-D-Gal a-Galase ~ ~~~~ ~ ~~”

Lectin bound (cpm X lO-’))llOO mgb Glomerular 10.1 0.02 0.34 Anterior 11.0 0.03 1.14 Posterior lens 9.2 0.03 0.73

,“Assays on the guanidine extracts were carried out as follows: directly; in the presence of 1 mM methyl-a-D-galactoside (CHs-a -~ - Gal); and after digestion of the extract with a-galactosidase (a- Galase) as described under “Experimental Procedures.”

* Values expressed per 100 mg of dry basement membrane; under the conditions of the assay mouse laminin bound 2 X IO5 cpm/pg of protein.

‘8 H GBM 1 LM

4 0 0 K - 200K -

116K -

6 6 K -

4 5 K - 2 9 K -

FIG. 11. Binding of B. simplicifolia I lectin to electrophoretically separated glycoconjugates from guanidine HCl extracts of calf anterior lens capsule (AC) and bovine ( B ) as well as human ( H ) glomerular basement membranes. After electrophoresis the components from the extracts as well as standard mouse laminin ( L M ) were transferred to a nitrocellulose sheet which was then overlaid with Iz5I-labeled B. simplicifolia I as described under “Experimental Procedures.” Bound lectin was then visualized by autoradiography. Electrophoresis and molecular weight markers were the same as Fig. 3.

found to bind negligible amounts of B. simplicifolia I (0.03% as much the bovine materia1/100 mg GBM). It was noted that the /3-component of the murine laminin standard (Fig. 11) reacted more intensely with this lectin than the a-subunit.

Although the identities of all the B. simplicifolia I binding components seen on electrophoresis are not clear, their migration suggested that in addition to laminin other basement membrane glycoconjugates contain terminal a-D-galactosyl residues. Evidence that the collagen of lens capsule reacts with B. simplicifolia I through a-D-galactosyl groups was obtained by examining the guanidine HCI/DTT extract (Ex- tract 2) from this basement membrane which contains substantial lectin binding activity (26% of total) as well as Type IV collagen (8,200 pg/100 mg membrane) but only trace amounts of laminin (22 pg/lOO mg) (Fig. 1). When a nitrocellulose sheet to which the electrophoretically separated components of anterior lens capsule (Extract 2) had been transferred was overlaid with ‘251-labeled B. simplicifolia I three major bands (Fig. 12, lune 1 ) which had the mobility ( M , = 390,000, 210,000, and 190,000) of the Type IV collagen polypeptides were visualized (compare to Fig. 4). These lectin- binding proteins were no longer evident after a-D-galactosidase digestion (Fig. 12, lune 2 ) or treatment with collagenase (Fig. 12, lane 3 ) .

DISCUSSION

The present study indicates that the molecular organization of GBM and lens capsule, two basement membranes which differ markedly in anatomical location, function, and cell types responsible for their synthesis, is despite some qualita- tive and quantitative variations fundamentally similar. Al- though a substantial amount of information has been obtained about the mature basement membranes of mammalian tissues by immunohistological procedures (11-17), the macromolecular profile of such membranes after isolation has not been explored by biochemical approaches; indeed the difficulty of

I 2 3

4 0 0 K - a 2 0 0 K - 1

116K - 6 6 K -

4 5 K - 29K -

FIG. 12. Occurrence of terminal cr-D-galaCtOSyl residues on a basement membrane collagen. Electrophoresis was carried out on the guanidine HCl/DDT fraction; (60 pg of protein) from Extract 2 of calf anterior lens capsule before (lane I ) and after treatment with either a-galactosidase (lane 2 ) or collagenase (lane 3) as described under “Experimental Procedures.” After transfer of components to a nitrocellulose sheet, they were exposed to 9-labeled B. simplicifolia I, and the bound lectin was visualized by autoradiography. Electrophoresis and molecular weight markers were the same as in Fig. 3.


separating well-defined basement membranes from cells and other extracellular material limits this approach to a few structures among which lens capsule and GBM can be found.

The selective solubilization of noncollagenous components by denaturing reagents under nonreducing conditions, which was initially observed with GBM (36) and has been effectively employed in a study of the Engelbreth-Holm-Swarm sarcoma (3), permitted us to extract a large portion of the laminin, entactin, fibronectin, and heparan sulfate proteoglycan of the basement membranes away from the most of the collagenous matrix with 5 M guanidine HC1. Further treatment with guanidine HC1 in the presence of DTT brought the total extraction of each of these noncollagenous proteins to greater than 90%. While a small portion of the Type IV collagen went into solution with the 5 M guanidine HC1 (2% of GBM 5% of lens capsule) and more was extracted with guanidine HC1/ DTT (30% of GBM; 15% of lens capsule), SDS/ME was required for effective solubilization of this protein. Indeed this is consistent with earlier observations that SDS under strong reducing conditions is the optimal solubilizing agent of GBM proteins (36); even after this treatment, however, some basement membrane material remained insoluble (20% in GBM; 10% in lens capsule) presumably due to the presence of nonreducible cross-links (37).

When visualized by immunoblotting the electrophoretic pattern of the Type IV collagenous components in the guanidine HCl residue from GBM and anterior as well as posterior lens capsules appeared remarkably similar with two closely spaced al- and a2-subunits ( M , = 210,000 and 190,000) and a slower moving component (Mr = 390,000) which may be an aggregate of these chains by way of hydroxylysine- derived cross-links (37); such Type IV collagen components migrating as two distinct a-chains have previously been observed in other undegraded basement membrane material (2, 4, 38, 39). The electrophoretic pattern of the guanidine HC1 insoluble residue as revealed by Coomassie Blue staining matched that observed by immunological detection indicating that this fraction does indeed consist predominantly of collagen; the two procedures for visualization however gave a different impression of the ratio of al- to a2-subunits indicating that the two polypeptide chains have dissimilar affinities for dye and antibody. The limited amount of collagen solubilized with guanidine HCI also contained the 210,000 and 190,000 bands and furthermore, in case of the GBM in particular, a number of lower molecular weight components. The latter which were also detected by the anti-Type IV collagen serum in the whole GBM and the guanidine HC1 residue probably represent the large number of polypeptides of collagen-like composition which were previously characterized in the bovine GBM (40). It has been postulated that they arise by in vivo proteolysis through the action of polymorpho- nuclear leukocyte enzymes (40), and their greater abundance in GBM than in the more sequestered lens capsule argues in favor of this possibility. Since the immunoreactive collagen from the guanidine HC1 extract emerged in the excluded volume during CL-GB filtration under nonreducing conditions, it would appear likely that these collagenous polypeptides are held together as large molecular complexes by disulfide bonds.

Fibronectin, of all the components measured in the present investigation, was present in the most disparate amounts; GBM contained considerably larger quantities of this glycoprotein than either anterior or posterior lens capsules. While the high content of fibronectin in GBM might be derived from plasma during the filtration process at least that portion (17% of total) which required reducing conditions for solubili-

zation would most likely be of local origin. The polydispersity which the GBM fibronectin demonstrated by immunoblotting, which was also observed in the plasma fibronectin standard, has been recognized as a characteristic of this protein and has been attributed to the post-translational modification and to the existence of multiple mRNAs (9,41). Lens capsule which is not exposed to plasma proteins contained only one single major fibronectin subunit ( M , = 240,000).

Both laminin and entactin were predominantly solubilized by the guanidine HC1 extraction of the membranes but were present in quite different ratios in GBM, anterior lens capsule, and posterior lens capsule; indeed the entactin content of the lens capsules was substantially greater than that of GBM, and the posterior capsule had an entactin/laminin ratio of 4.6:l compared to a value of 1:4 reported for the mouse endodermal membrane sacs in which this protein was first observed (6). As anticipated from the studies of Carlin et al. ( 6 ) , the anti-entactin serum did not react with GBM or lens capsule laminin, and the molecular weight which we observed ( M , = 150,000) closely approximates that determined for the mouse protein; the position of elution of entactin on the CL- 6B column under nonreducing conditions suggested that this protein does not occur in covalent association with other polypeptide.

While the a-subunit of laminin was clearly evident in both GBM and lens capsule and co-migrated on electrophoresis with the polypeptide chain from mouse laminin (35), the /3- subunit occurred in only small amounts in lens capsule extracts and could not at all be visualized in the immunoblots of GBM. Although it is possible that the absent or diminished amount of the &component could reflect reduced translation, in view of the well-defined molecular structure of the laminin molecule (35), it appears more likely that proteolysis is responsible for our observations. Indeed the absence of the 400,000 molecular weight component has also been reported in immunopurified human placental laminin (42) and has been attributed to the high protease susceptibility (35) of the P-subunit. Our gel filtration studies under nonreducing conditions do indicate that the GBM and lens capsule laminin occur as entities with M, of about 600,000 which would be consistent with the presence of the three a-subunits observed in the protease degraded molecule (35).

While the high titer antiserum which we were able to raise against the previously purified heparan sulfate proteoglycan from bovine GBM (21) reacted as expected only with a 200,000 molecular weight glycoconjugate in the GBM guanidine HC1 extract, substantially larger components were visualized in the extracts from anterior and posterior lens capsules. The higher molecular weight of the lens capsule proteoglycan was further indicated by earlier elution during filtration on a CL- 6B column (anterior lens capsule, Kav = 0.087 cf. to GBM, K,, = 0.30). It is not known at this time whether the lens capsule proteoglycans differ from the GBM glycoconjugate in the size of their core proteins, in the number and length of their heparan sulfate chains, or in both parameters. However, previous analyses indicated that the glycosaminoglycan content of posterior lens capsule is greater than that of the anterior and that both contain more of the polysaccharide than GBM (43). Indeed these differences are in the same direction as the apparent molecular weights of the proteoglycans which is what would be anticipated if their core proteins are similar in size.

While DEAE-cellulose chromatography of the guanidine HC1 extracts of the GBM and lens capsules clearly separated the laminin, entactin, &d fibronectin from the proteoglycan, some of the Type IV collagen eluted with this latter glycocon-


jugate. This unexpected finding might be explained by the previous observation that a collagenous peptide appears to be linked by disulfide bonds to the core protein which contains the heparan sulfate chains (44); the alternate possibility that we are observing a noncovalent collagen/proteoglycan interaction seems less likely in view of the fact that the DEAE- cellulose chromatography was carried out in 8 M urea.

Our present studies helped to gain insight into the interaction of basement membranes with B. simplicifolia I, a lectin, which after conjugation to fluorescein, has been observed to outline these extracellular structures in various rodent tissues by virtue of its binding to terminal cu-D-galaCtOSyl groups (22). While the presence of such sugar residues in laminin (23) appeared to provide an appropriate basis for the B. simplicifolia I basement membrane interaction, we detected substantial lectin binding activities in extracts and DEAE-cellulose column fractions of both GBM and lens capsules, which could not be attributed to this glycoprotein alone. Indeed lectin transblots indicated that a number of glycoconjugates other than laminin appear to contain terminal a-D-galaCtOSyl residues. More direct evidence for the presence of such saccharide groups was obtained for Type IV collagen, and it can be presumed that these occur on the asparagine-linked carbo- hydrate units (25) of this protein.

In contrast to the bovine basement membrane, we were unable to detect B. simplicifolia I binding activity in human GBM extracts, and this finding is consistent with the report that fluorescein-conjugated B. simplicifolia I does not label basement membranes in human tissue sections (22). Indeed, such a species difference does not appear to be limited to basement membranes glycoconjugates as it has been reported that neither human thyroglobulin (28) nor thyrotropin-recep- tor (45), in contrast to the bovine components, contain terminal a-D-galaCtOSyl residues.

It is apparent that the guanidine HCl extract of both GBM and lens capsule contains a complex mixture of proteins all of which identified so far appear to be glycoconjugates; the electrophoretic profile of this fraction suggests that other components besides those characterized in the present study may occur.

It would be informative to be able to assign the complicated mixture of glycoconjugates extracted by guanidine HCl to an ultrastructurally defined region of the basement membranes. Although several of these components are believed to be involved in cell attachment (41, 46) and might logically be thought to occur in that portion of the basement membrane (lamina lucida or lamina rara) in close contact with the cell layer as several studies have suggested (12, 13, 17), such a localization of the noncollagenous proteins cannot be consid- ered established in light of other reports which view the basement membrane as an integrated rather than a layered entity (14-16). It is anticipated that the disposition of the various basement membrane macromolecules will be further clarified by a combination of approaches involving morpho- logical, biochemical, and physiological techniques.

REFERENCES 1. Heathcote, J. G., and Grant, M. E. (1981) Int. Rev. Connect.

Tissue Res. 9 , 191-264 2. Timpl, R., Martin, G. R., Bruckner, P., Wick, G., and Weide-

mann, H. (1978) Eur. J. Biochem. 84,43-52 3. Kleinman, H. K., McGarvey, M. L., Liotta, L. A., Robey, P. G.,

Tryggvason, K., and Martin, G. R. (1982) Biochemistry 21,

4. Alitalo, K., Vaheri, A., Krieg, T., and Timpl, R. (1980) Eur. J. 6188-6193

Biochem. 109,247-255 5. Crouch, E., Sage, H., and Bornstein, P. (1980) Proc. Natl. Acad.

Sci. U. S. A. 77 , 745-749

6. Carlin, B., Jaffe, R., Bender, B., and Chung, A. E. (1981) J. Biol.

7. Hayman, E. G., Oldberg, A., Martin, G. R., and Ruoslahti, E.

8. Oohira, A., Wight, T. N., and Bornstein, P. (1983) J. Bid. Chem.

9. Paul, J. I., and Hynes, R. 0. (1984) J. Bid. Chem. 259 , 13477-

10. Tyree, B., Horigan, E. A., Klippenstein, D. L., and Hassel, J. R.

Chem. 256,5209-5214

(1982) J. Cell Biol. 9 4 , 28-35

258,2014-2021

13487

(1984) Arch. Biochem. Biophys. 23 1,328-335 11.

12.

13.

14.

15.

16.

17.

18. 19.

20.

21.

22.

23.

24. 25.

26. 27.

28.

29. 30.

31. 32.

33. 34.

35.

36.

37.

38.

Madri, J. A., Roll, F. J., Fukhmayr, H.; and Foidart, J.-M. (1980) J. Cell Bid. 8 6 , 682-687

Foidart, J. M., Bere, Jr., E. W., Yaar, M., Rennard, S. I., Gullino, M., Martin, G. R., and Katz, S. I. (1980) Lab. Inuest. 42 , 336- 342

Courtoy, P. J., Timpl, R., and Farquhar, M. G. (1982) J. Histo- chem. Cytochem. 30,874-886

Laurie, G. W., Leblond, C. P., and Martin, G. R. (1982) J. Cell Biol. 9 5 , 340-344

Mynderse, L. A., Hassel, J . R., Kleinman, H. K., Martin, G. R., and Martinez-Hernandez, A. (1983) Lab. Inuest. 48,292-302

Laurie, G. W., Leblond, C. P., Inoue, S., Martin, G. R., and Chung, A. (1984) Am. J. Anat. 169, 463481

Stow, J. L., Sawada, H., and Farquhar, M. G. (1985) Proc. Natl. Acad. Sci. U. S. A. 82,3296-3300

Spiro, R. G. (1978) Ann. N . Y. Acad. Sci. 312 , 106-121 Fukushi, S., and Spiro, R. G. (1969) J. Biol. Chem. 2 4 4 , 2041-

Cammarata, P. R., and Spiro, R. G. (1982) J. Cell. Physiol. 113 ,

Parthasarathy, N., and Spiro, R. G. (1984) J. Biol. Chem. 259 , 12749-12755

Peters, B. P., and Goldstein, I. J. (1979) Exp. Cell Res. 120,321- 334

Shibata, S., Peters, B. P., Roberts, D. D., Goldstein, I. J., and Liotta, L. A. (1982) FEBS Lett. 142 , 194-198

Spiro, R. G. (1967) J. Bid. Chem. 242 , 1915-1922 Levine, M. J., and Spiro, R. G. (1979) J. Biol. Chem. 254,8121-

Vaitukaitis, J. L. (1981) Methods Enzymol. 73 , 46-52 Greenwood, F. C., Hunter, W. M., and Glover, J. S. (1963)

Spiro, R. G., and Bhoyroo, V. D. (1984) J. Bid. Chem. 2 5 9 ,

Laemmli, U. K. (1970) Nature 227,680-685 Towbin, H., Staehelin, T., and Gordon, J. (1979) Proc. Natl. Acad.

Burnette, W. N. (1981) Anal. Biochem. 112, 195-203 Hancock, K., and Tsang, V. C. W. (1983) Anal. Biochem. 133,

Peterson, G. L. (1977) Anal. Biochem. 8 3 , 346-356 Cammarata, P. R., and Spiro, R. G. (1985) J. Cell. Physiol. 125 ,

393-402 Rao, C. N., Margulies, I. M. K., Tralka, T. S., Terranova, V. P.,

Madri, J. A., and Liotta, L. A. (1982) J. Bid. Chem. 257,9740- 9744

Hudson, B. G., and Spiro, R. G. (1972) J. Bwl. Chem. 247,4229- 4238

Heathcote, J. G., Bailey, A. J., and Grant, M. E. (1980) Biochem.

Dean, D. C., and Barr, J. F., Freytag, J. W., and Hudson, B. G.

2048

273-280

8124

Biochem. J. 89,114-123

9858-9866

Sci. U. S. A. 76,4350-4354

157-162

J. 190,229-237

(1983) J. Bid. Chem. 258. 590-596 39. Taylor, C. M., and Grant, M . E. (1985) Biochem. J. 226, 527-

40. Sato, T., and Spiro, R. G. (1976) J. Biol. Chem. 251,4062-4070 41. Yamada, K. M., Akiyama, S. K., Hasegawa, T., Hasegawa, E.,

H., Olden, K., and Chen, W-T. (1985) J. Cell. Biochem. 2 8 , Humphries, M. J., Kennedy, D. W., Nagata, K., Urushihara,

79-97 42. Wewer, U., Albrechtsen, R., Manthorpe, M., Varon, S., Engvall,

E., and Ruoslahti, E. (1983) J. Biol. Chem. 258 , 12654-12660 43. Parthasarathy, N., and Spiro, R. G. (1982) Arch. Biochem. Bio-

44. Parthasarathy, N., and Spiro, R. G. (1981) J. Biol. Chem. 256 ,

45. Kress, B. C., and Spiro, R. G. (1984) Fed. Proc. 43,1706 46. Kleinman, H. K., Klebe, R. J., and Martin, G. R. (1981) J. Cell

536

p h y ~ . 213, 504-511

507-513

Biol. 88,473-485

the of biological chemistry vol. no. march pp. in u.s.a ...€¦ · from the departments of...

Documents