Search

Type VIII collagen was isolated from bovine Descemet's membranes by pepsin treatment and salt fractionation, as described by Kapoor et al. [(1986) Biochemistry 25, 3930-3937]. Contaminating type IV collagen was removed by ion-exchange chromatography. Purified type VIII collagen consisted of two different polypeptide chains and, compared to the fiber forming collagens, showed a higher thermal stability. Corresponding fractions isolated from pepsinized human Ewing's sarcoma and fetal calf aorta reacted immunologically with a protein of similar molecular mass. After extraction of Descemet's membranes with guanidine hydrochloride, a peptide of about 60 kDa was obtained. This seems to be the tissue form of type VIII collagen.

Various forms of heparan sulfate proteoglycan were solubilized from the mouse Engelbreth-Holm-Swarm (EHS) sarcoma by extraction with 0.5 M NaCl, collagenase digestion and extraction with 4 M guanidine. They could be separated into high (greater than or equal to 1.65 g/ml) and low (1.38 g/ml) buoyant density variants. The high-density form from the NaCl extract and collagenase digest had Mr = 130000 and So20,W = 4.5 S and contained 4-10% protein, indicating Mr = 5 000-12 000 for the protein core. This proteoglycan exhibited polydispersity as shown by rotary shadowing electron microscopy and ultracentrifugation. An average molecule consisted of four heparan sulfate chains (Mr = 29 000) each with a length of 32 +/- 10 nm. The low-density form (Mr about 400 000) could not be completely purified and contained about 50% protein. As shown by radioimmunoassay, the various proteoglycans shared similar protein cores. Labeling of the tumor in vivo or in vitro demonstrated preferential incorporation of radioactive sulfate in the high-density form. The high-density proteoglycan interacted in affinity chromatography by virtue of its heparan sulfate chains with laminin, fibronectin, the globular domain NC1 and the triple helix of collagen IV. These interactions were abolished at moderate concentrations of NaCl (0.1-0.2 M) and in the presence of heparin, chondroitin sulfate or dextran sulfate. Interactions with the globule NC1 could also be demonstrated by velocity band centrifugation in sucrose gradients and a binding constant of about 10(6) M-1 was derived.

Native type IV collagen was isolated from human placenta using pepsin solubilisation followed by fractional salt precipitation and chromatogarphic purification. The native preparation was characterised using amino acid analyses, disc gel electrophoresis, segment-long-spacing crystallites and immunological methods. Two component alpha chains were isolated with molecular weights of approximately 95000 and 70000. Cyanogen bromide digests of these chains indicated that they are not related to any of the known alpha chains of interstitial collagens or to the recently described collagen containing alphaA and alphaB chains. They are also not related to one another and are therefore probably fragments of two genetically distinct type IV collagen alpha chains.

The globular domain of collagen IV was solubilized by collagenase digestion from a mouse tumor, human placenta and bovine aorta and was purified by chromatographic methods. The materials show a unique, mainly non-collagenous amino acid composition and contain small amounts of glucosamine and galactosamine. The globular structures with Mr = 170 000 appear as a hexameric assembly originating from two collagen IV molecules. Subunits of this assembly are two different dimers Da and Db (Mr about 56 000) and monomers (Mr = 28 000). Their N-terminal amino acid sequences start with short triple-helical sequences, which overlap with the C-terminal triple helix of the alpha 1(IV) and alpha 2(IV) chain, demonstrating that the globule originates from the C terminus of collagen IV. Dimers arise from monomers by disulfide cross-linking (form Db) and/or formation of non-reducible cross-links (form Da). Reduction under non-denaturing conditions causes partial dissociation of the globule and of collagen IV dimers, indicating that reducible cross-links are formed between monomers of two different collagen IV molecules. Dissociation of the hexamer into the subunits can be achieved with 8 M urea, sodium dodecyl sulfate or in the pH range 2.5-4. The latter indicates that carboxyl groups are essential for association. Mixtures of the subunits (monomers and dimers) or purified dimers reassemble in neutral buffer into hexamers as shown by ultracentrifugation and electron microscopy. Reconstituted hexamers, however, dissociate in a much broader pH range than the native globules. Circular dichroic spectra indicate that the structure is more completely refolded from acid-treated than from urea-treated material. These data suggest that globules originating from monomers (as existing in single collagen IV molecules) are stabilized by the adjacent triple helix. Covalent cross-link formation stabilizes the globular structure and allows reconstitution in stoichiometric proportions.

A novel acidic glycoprotein, BM-40, with Mr = 40,000, was purified from the basement-membrane-producing mouse EHS tumor and characterized with regard to its unique chemical and antigenic properties. It was obtained from the tumor in a neutral salt-soluble form or as a component requiring extraction with 6M guanidine X HCl. This protein could also be identified in many other tissue extracts and cell and tissue cultures. The most intact form of BM-40 consists of a single polypeptide chain which undergoes limited proteolysis during extraction and purification. BM-40 exists in most tissues in stoichiometric amounts compared to other basement membrane proteins (laminin, nidogen) and is secreted by various teratocarcinoma and epithelial cells. It can be visualized by immunofluorescence in the extracellular matrix of the EHS tumor and Reichert's membrane. Other tissues which contain extractable BM-40 were negative in immunofluorescence.

A monoclonal antibody was produced against purified nidogen extracted from a mouse basement-membrane-producing tumor. This antibody reacted with a determinant on Nd-40, a rod which separates the globular domains of nidogen. Antigenicity depends on intrachain disulfide bonds within this rod. The monoclonal antibody was used to detect nidogen fragments after proteolytic cleavage of isolated nidogen, and nidogen complexed to laminin. The data indicate that thrombin and thermolysin generated very different patterns of degradation, but in both cases no differences were found between isolated and complexed nidogen. In contrast, nidogen in the laminin-nidogen complex was much less degraded by trypsin than isolated nidogen, indicating that an interaction between these basement membrane components reduces the susceptibility of nidogen to trypsin digestion. Immunofluorescent studies, using the monoclonal antibody on sections of the EHS tumor after proteolytic digestion, showed that the retention or disappearance of the Nd-40 determinant correlated with the in vitro digestion pattern of the laminin-nidogen complex.

The complete amino acid sequence of the 914-residue-long pepsin fragment alpha 1 (IV)95 from the alpha 1 chain of human placental basement membrane (type IV) collagen is presented. This sequence contains 12 interruptions of the collagenous triplet sequence Gly-Xaa-Yaa which varied in length from 1 to 11 residues. The distribution of amino acids between the Xaa and Yaa position was similar to that found in interstitial collagens but the extent of proline and lysine hydroxylation differed. Computer comparisons of the alpha 1 (IV)95 sequence with those of the interstitial collagen chains did not reveal any homology, whereas a comparison with the partial sequences of mouse tumor and bovine lens capsule alpha 1 (IV) showed an approximately 85% identity. The unique sequence characteristics of type IV collagen are discussed in relation to its macromolecular structure and to the interstitial collagens.

Collagen type IV, laminin, heparan sulfate proteoglycans, nidogen (entactin) and BM-40 (osteonectin, SPARC) represent major structural proteins of basement membranes. They are well-characterized in their domain structures, amino acid sequences and potentials for molecular interactions. Such interactions include self-assembly processes and heterotypic binding between individual constituents, as well as binding of calcium (laminin, BM-40) and are likely to be used for basement membrane assembly. Laminin, collagen IV and nidogen also possess several cell-binding sites which interact with distinct cellular receptors. Some evidence exists that those interactions are involved in the control of cell behaviour. These observations have provided a more defined understanding of basement membrane function and the definition of new research goals in the future.

Large quantities of intact laminin-nidogen complex could be extracted from a mouse tumor basement membrane with a physiological buffer containing EDTA. Analysis of the purified complex demonstrated that the two proteins occur in an equimolar ratio and that anchoring of these complexes to the extracellular matrix requires divalent cations. Reversible dissociation of the complex was achieved with 2 M guanidine X HCl and has been used for purification of the individual components. Electron microscopy and binding studies using laminin fragments demonstrated that nidogen interacts specifically with the center of the cross-shaped laminin molecule as represented by the short-arm structure fragment 1. The complex was also useful to confirm and refine a previously proposed dumb-bell structure of nidogen and to prepare and characterize the cell-binding fragment 8 from the long arm of laminin.

Collagen extracted from bovine anterior lens capsule basement membrane under minimally degradative conditions can be precipitated in the form of segment-long-spacing crystallites. Positive and negative stain preparations of these precipitates, using uranyl acetate, show the crystallites to have an asymmetric band pattern, indicating unidirectional orientation of all molecules in each crystallite. Non-banded globular masses appear at both ends of most of the crystallites, giving the segment-long-spacing aggregates the form of dumbells, with a total length of 300 nm. Repeated pepsin digestion removes first one then both globular regions leaving the helical region intact but with a length still equal to 300 nm. It thus appears that the pepsin-susceptibile protein is folded back over and covers the ends of the helical regions. Preparations with the most intact end regions tend to aggregate in fibrous segment-long-spacing form, that is, in end-interacted strings or nets. The very prominent molecular end-region interaction suggests a similar filamentous but non-periodic organization within native basement membrane.

Laminin was recently characterized as being a major non-collageneous protein of the basement membrane matrix produced by a mouse tumor. It was extracted from the tumor matrix with neutral buffer and purified under non-denaturing conditions. Rabbit and guinea pig antisera raised against laminin or a large pepsin fragment P1 of laminin showed strong binding to both laminin and peptide P1 but only a weak reaction with reduced and alkylated laminin. The major antigenic determinants of laminin were located in a disulfide knot, comprising one third of the molecule, which resisted degradation by pepsin or cyanogen bromide. Minor antigenic determinants shared by the native and reduced protein could also be identified. The data were interpreted as showing that laminin consists of conformationally rigid as well as more flexible domains. Absorption studies with mouse kidney homogenate indicated that authenic basement membranes contain a protein immunologically identical to laminin. Tissues from other species contain a related protein which exhibits partial cross-reaction with mouse laminin. Together with immunofluorescence data the findings demonstrate that laminin, like type IV collagen, occurs in most basement membranes of the body.

Extraction of a basement-membrane-producing mouse tumor with 6 M guanidine/HCl in the presence of protease inhibitors allowed the purification of the genuine form of the matrix protein nidogen (Mr = 150,000) and, in addition, two defined fragments (Mr = 130,000 and 100,000). Smaller fragments (Mr = 80,000 and 40,000) were obtained under conditions with less stringent control of endogenous proteolysis. Intact nidogen and the larger fragments were similar in amino acid and carbohydrate (about 5%) composition, the presence of a single polypeptide chain, conformational features as revealed by CD spectroscopy and all shared major epitopes located on the Mr = 80,000 fragment. Additional epitopes were found on intact nidogen and the Mr = 130,000 fragment. Nidogen and the various fragments possess different N-terminal amino acid sequences indicating a stepwise degradation from the N-terminal end of the molecule. Electron microscopical and hydrodynamic studies of the Mr = 80,000 fragment demonstrated a structure consisting of a globular head connected to a thin tail. Intact nidogen appears to contain a somewhat larger globule but the same tail, which is terminated at its opposite end by a second, smaller globular structure. The data suggest a multidomain structure for nidogen containing sites highly susceptible to proteolytic cleavage.

Nidogen was purified from a mouse tumor basement membrane where it accounted for 2-3% of the total proteins. It was isolated as two forms (A and B) of a monomer (Mr = 80000) each consisting of a single polypeptide chain folded into a globular head connected to a small tail. The B form of the monomer was shown to be capable of aggregating into a nest-like structure (Mr greater than 250000). A smaller form (Mr = 45000) was observed in some of the extracts. The amino acid composition of nidogen was different to that of other basement membrane proteins. It contained about 10% carbohydrate, with N-linked and O-linked oligosaccharide chains in similar proportions. Isoelectrofocussing demonstrated a limited heterogeneity of nidogen with pI in the range 6.5 - 7. Monomeric nidogen failed to interact with other basement membrane components and heparin. Aggregation could be induced by limited proteolysis and was reversed by detergents or high salt concentrations. Together with the observation that most of the nidogen could be solubilized only after destroying the collagenous matrix, the data indicate that aggregation of nidogen reflects an activity involved in matrix assembly. Specific antibodies raised against nidogen did not distinguish between the monomeric and aggregated form of the protein but showed that the fragment was antigenically deficient. These antibodies did not cross-react with collagen type IV, laminin, entactin and heparansulfate proteoglycan. Immunofluorescence staining and absorption studies demonstrated that nidogen is a common component of authentic basement membranes. Larger forms of nidogen (Mr about 100000 and 150000) were found in organ cultures of Reichert's membrane suggesting that it is synthesized in precursor forms.

Fragment P1 (Mr = 55 000) which is located in the carboxy-terminal portion of the triple-helical segment of the alpha 1(IV) chain was purified from a pepsin digest of a mouse tumor basement membrane. Peptides produced from P1 by cleavage with cyanogen bromide and trypsin were purified and characterised with respect to their size, composition and partial amino acid sequence. Fragment patterns and overlapping sequences allowed the ordering of these peptides within the P1 segment. About 70% of the sequence was determined by Edman degradation. Segments of seven or eight amino acid residues, which lacked the triple-helical sequence Gly-Xaa-Yaa, were found at both ends of fragment P1, explaining the susceptibility of native type IV collagen to pepsin. Two further interruptions of the triple helix were indicated by single deletions of GLy or Yaa positions in the triplet structure (Gly-Xaa-Yaa)n. The two 3-hydroxyproline residues were found in position Xaa and are surrounded by homologous sequences.

The amino acid sequence of the 212-residues-long N-terminal aggregation and cross-linking region of the alpha 2(IV) chain of human basement membrane collagen is presented. Comparing this with the primary structure of alpha 1(IV)7S [Glanville et al. (1985) Eur. J. Biochem. 152, 213-219] revealed a high degree of similar subdivisions in three functional regions. These are the 21-residue-long N-terminal non-triple-helical regions (NH1) containing cysteine and lysine residues which are putative cross-linking sites, a 117-residue-long triple-helical region (TH1) responsible for the aggregation of four molecules to form the 7S domain and which also possess cross-linking sites, and finally a 10-residue-long non-triple-helical region (NH2) which introduces the first of many flexible areas into the triple helical body of the molecule [Hofmann et al. (1984) J. Mol. Biol. 172, 325-343]. Computer calculations of interaction scores between parallel and antiparallelly aligned triple-helical regions (TH1) of the 7S domain allowed the prediction of a detailed model for the structure of the 7S complex which agreed well with models based primarily on electron micrographs of rotary shadowed type IV collagen tetramers. The results indicated that the assembly of the 7S domain is directed by hydrophobic interactions and is self-limiting to a tetramer. The most favourable chain configuration is alpha 2-alpha 1-alpha 1.

The hexameric complex of globular domains of type IV collagen was isolated after collagenase digestion of human placenta and the different monomers and dimers present were chromatographically separated. The ratio of alpha 1(IV)NC1 to alpha 2(IV)NC1 was 2:1. About 50% of the NC1 domains were connected to dimers. Predominantly alpha 1-alpha 1 dimers were found. Only 12% were alpha 2-alpha 2 dimers and no alpha 1-alpha 2 dimers could be detected. The majority (88%) of the intermolecular bonds was found to be disulfide bridges. The remainder could not be cleaved by reduction. To elucidate the arrangement of the disulfide bonds, the unreduced alpha 1(IV)NC1 monomers were treated with cyanogen bromide, the disulfide-bridged peptides isolated and characterized by Edman degradation. Each of the two homologous subdomains within a monomer is stabilized by an identical set of three disulfide bonds. In subdomain I, cysteines at positions 20 and 53 are connected with the C-terminal cysteine pair 108 and 111. Thus formed, the disulfide knot stabilizes two interconnected loops of 32 and 54 residues, respectively. A smaller loop of five residues occurs due to a disulfide bond between the cysteines 65 and 71. A similar disulfide arrangement is indicated for subdomain II which is separated from subdomain I by a segment of 20 amino acid residues. The same arrangement of disulfide bonds has been strongly suggested for the alpha 2(IV)NC1 monomer by the isolation and characterization of its disulfide-bridged tryptic fragments. Similar investigations on the dimeric alpha 1(IV)NC1 domain established the arrangement of the intermolecular disulfide bonds. They are formed by a complete disulfide exchange between corresponding disulfide knots of two monomeric NC1 domains.

A major proteolytic fragment (E1/E1-4) of the basement membrane protein laminin, comprising the three short arms with some terminal globules missing, was isolated by elastase digestion, and partial protein sequence data were determined for several tryptic peptides. Sequences which corresponded to A-chain structures were used to synthesize oligonucleotides for the construction and screening of a primer-extended cDNA library from mouse PYS-2 cells. A clone of 1.1 kb was obtained and shown by sequencing to correspond to the 5' end of the 10-kb mRNA of the A chain of laminin. The clone contains 77 nucleotides of 5' untranslated sequence and a region coding for 334 amino acids, including a presumptive signal peptide of 24 amino acids. The sequence is 30% homologous to the corresponding N-terminal part of the B1 chain of laminin, suggesting the same structure for both domains. The data present further evidence for a recent structural model which postulates that each of the three laminin polypeptide chains forms a distinct short arm.

The cDNA and protein sequences of the N-terminal half of human basement membrane collagen (type IV) have been determined. Overlapping cDNA clones were constructed by repeated primer extension with synthetic oligonucleotides. They cover 2953 bp, beginning at the 5' end of the corresponding mRNA. At the protein level, the sequence of the cyanogen bromide peptide CB6 adjacent to the 7S domain has been additionally elucidated. The data presented here complete the protein sequence and nearly the entire cDNA sequence of the human alpha 1(IV) chain. The amino-terminal half of the alpha 1(IV) chain contains 8 cysteine residues involved in intramolecular and intermolecular cross-links. The entire triple-helical domain of alpha 1(IV) is interrupted by 21 non-triplet regions.

NC1, the C-terminal non-collagenous globular domain of collagen IV, represents one of the two end regions responsible for the assembly and cross-linking of the extracellular network of basement membrane collagen. Several cDNA clones for the NC1 domain of the alpha 1(IV) collagen chain of mouse have been isolated by using synthetic oligonucleotides as screening probes for mouse libraries. The oligonucleotides were synthesized according to known stretches of the corresponding protein sequence. Sequencing of the overlapping cDNA clones allowed the complete amino acid sequence of the NC1 domain to be deduced as well as the C-terminal 165 amino acid residues of the triple helix. It consists of 229 amino acid residues which comprise two homologous regions with a high content of cysteine. These DNA and protein sequences are compared to the corresponding sequences of other collagens and discussed with respect to their structural and biological significance.

The cDNA and protein sequences of the N-terminal 60% of the alpha 2(IV) chain of human basement membrane collagen have been determined. By repeated primer extension with synthetic oligodeoxynucleotides and mRNA from either HT1080 cells or human placenta overlapping clones were obtained which cover 3414 bp. The derived protein sequence allows for the first time a comparison and alignment of both alpha chains of type IV collagen from the N terminus. This alignment reveals an additional 43 amino acid residues in the alpha 2(IV) chain as compared to the alpha 1(IV) chain. 21 of these additional residues form a disulfide-bridged loop within the triple helix which is unique among all known collagens.

The paper describes further characterization of the 55-kDa short-chain collagen from lens capsule. Lens capsules were extracted with 5.5 M guanidine.HCl and the extracted material was fractionated on agarose A-5M followed by high-pressure liquid chromatography (HPLC). By amino acid composition, the major fraction obtained from HPLC was found to be different than type-IV collagen fragments. The 55-kDa short-chain collagen on pepsin digestion produced a 45-kDa pepsin-resistant fragment. The undifferentiated embryonal carcinoma (F-9) cells were found to synthesize increased amounts of 55-kDa short-chain collagen. The identity of this biosynthesized molecule with 55-kDa short-chain collagen from lens capsules was established by immunoprecipitation experiments. The results indicated a close similarity or identical nature of the short-chain collagens from these two sources.