Search

Brown Norway rats injected with mercuric chloride (HgCl2) develop autoantibodies which immunolocalize along the glomerular basement membrane at first in a linear pattern and then in a granular pattern. The aim of this study was to characterize the specificity of these antibodies and to investigate the mechanisms responsible for the formation of granular immune deposits in the subepithelial zone of the glomerular basement membrane. The rats were found to develop circulating anti-laminin, anti-type IV collagen, anti-heparan sulfate proteoglycan, and anti-entactin antibodies. Antibodies against laminin and type IV collagen were found in relatively high titers in the sera and were specifically concentrated in the nephritic kidneys. Antibodies eluted from the nephritic kidneys with either linear or granular deposits reacted with basement membrane antigens synthesized and secreted by cultured rat glomerular visceral epithelial cells. Thus, in this model, the interaction of anti-laminin and type IV collagen antibodies with antigens secreted by glomerular visceral epithelial cells might, together with other mechanisms, contribute to the formation of granular immune deposits in the subepithelial part of the glomerular basement membrane.

Antibodies to type IV collagen obtained from the basement membrane of the mouse EHS tumor were incubated with sections of rat incisor teeth and other tissues for immunostaining by direct or indirect methods. In all locations, the immunostaining was pronounced in basement membranes in which it was restricted to the "basal lamina" layer, from which "bridges" often extended to nearby basal laminae. Usually no immunostaining was detectable in the cells associated with the basement membranes. However, examination of the capillaries at the posterior extremity of the rat incisor tooth, where tissues are at an early stage of development, showed immunostaining not only of the basement membrane, but also of the endothelial cells. The staining was localized in rough endoplasmic reticulum cisternae, some Golgi saccules and their peripheral distensions, and structures believed to be secretory granules. These findings suggest that the synthesis of type IV collagen proceeds along the classical secretory pathways through rough endoplasmic reticulum and Golgi apparatus. At the same time, immunostaining was usually lacking in the cells of the capillaries that had migrated about 2 mm away from the posterior end of the incisor tooth and also in the cells of most other tissues examined, even though the associated basal laminae were reactive. It is, therefore, presumed that the production of type IV collagen may be high in cells at an early stage of development and that any later production and turnover of basement membrane collagen can only be minimal.

In cartilage, type II collagen is present as thin, short, randomly oriented fibrils. In vitro, however, type II collagen forms fibrils of large diameter, indicating that additional factors may be involved in the regulation of collagen fibril formation. We have examined extracts of a cartilage-producing tumor for the presence of collagen-binding proteins. In addition to fibronectin and link protein, a Mr 54,000 protein was found to bind to collagen fibrils as well as to native and denatured type II collagen. Immunological studies using antibody against the protein indicate that it is a cartilage matrix protein, not present in bone or in several other tissues. In vitro studies show that the Mr 54,000 protein in combination with cartilage proteoglycan decreases the rate of type II fibril formation and causes the fibrils to be of small diameter (24 +/- 8 nm). These studies indicate that complexes between collagen and proteoglycans mediated by this protein may regulate the assembly of cartilage matrix.

Cells cultured from human skin synthesize precursor forms of types I and III collagens. After denaturation and reduction, the polypeptide chains obtained from these molecules had molecular weights estimated to be 140,000 and 120, 000. The larger chains, the pro alpha chains, are believed to be derived from the original precursor molecules. The smaller chains arise from altered forms, p-collagens, that may be normal intermediates in the conversion of both procollagens to collagens I and III.

Electron microscopic immunostaining of rat duodenum and incisor tooth was used to examine the location of four known components of the basement-membrane region: type IV collagen, laminin, heparan sulfate proteoglycan, and fibronectin. Antibodies or antisera against these substances were localized by direct or indirect peroxidase methods on 60-microns thick slices of formaldehyde-fixed tissues. In the basement-membrane region of the duodenal epithelium, enamel-organ epithelium, and blood-vessel endothelium, immunostaining for all four components was observed in the basal lamina (also called lamina densa). The bulk of the lamina lucida (rara) was unstained, but it was traversed by narrow projections of the basal lamina that were immunostained for all four components. In the subbasement-membrane fibrous elements or reticular lamina, immunostaining was confined to occasional "bridges" extending from the epithelial basal-lamina to that of adjacent capillaries. The joint presence of type IV collagen, laminin, heparan sulfate proteoglycan, and fibronectin in the basal lamina indicates that these substances do not occur in separate layers but are integrated into a common structure.

Type II procollagen messenger ribonucleic acid (mRNA) was isolated from chick sternum and rat chondrosarcoma cells and translated in a reticulocyte lysate cell-free system. A high molecular weight band was identified as type II procollagen by gel electrophoresis, collagenase digestion, and specific immunoprecipitation. The translation of type II mRNA was specifically inhibited by addition of type I procollagen amino-terminal extension peptide. When this peptide was added to the media of cultured fetal calf chondrocytes, chick sternal chondrocytes, or chick tendon fibroblasts, no inhibition of collagen synthesis was evident. These data suggest a general regulation of collagen biosynthesis by these peptides in the cell-free translation system. However, as indicated by the cell culture experiments, cellular characteristics and evolutionary divergence of animal species seem to restrict the effect of the peptides.

Two overlapping clones spanning 19 kilobase pairs (kb) of the 5' end of the alpha 1 (IV) collagen gene were isolated and found to contain a single exon which encoded the 5'-untranslated sequence and 84 base pairs of the signal peptide. The 5' end of this exon was determined to be the 5' end of the transcript by S1 nuclease protection and primer extension. The nucleotide sequence of 1 kb of the 5'-flanking DNA was extremely G + C-rich (greater than 70%) and contained two GC boxes and a putative cAMP regulatory sequence. The transcriptional regulation of the alpha 1 (IV) gene was studied with chimeric gene constructs utilizing 2.5 kb of the 5'-flanking sequence coupled to the gene for chloramphenicol acetyltransferase. Transfection of this construct into differentiating F9 cells resulted in low chloramphenicol acetyltransferase activity compared to beta-actin or Rous sarcoma virus long terminal repeat promoters, although these cells produce large amounts of collagen IV. Inclusion of a 2.7-kb sequence 2.3 kb downstream from the first exon in either orientation increased the transcription of the chloramphenicol acetyltransferase construct approximately 10-fold in F9 cells, but was not active in NIH 3T3 cells, which synthesize little collagen IV. These results indicate the presence of an enhancer within the first intron, which increases the expression of this gene.

Enzyme-linked immunoadsorbant assays have been developed for types I, II, III, and IV collagen and for laminin and fibronectin. These assays offer a specific, sensitive, and convenient method for the measurement of various connective tissue components either separately or simultaneously. Plastic microtiter wells are coated with purified antigen, then antibodies to the antigen are allowed to bind to the insolubilized antigen in each well. The amount of bound antibody is determined by incubation with a second antibody which is covalently linked to alkaline phosphatase or horseradish peroxidase. The amount of bound enzyme is assayed after adding an appropriate substrate. The level of protein in a sample is estimated from its ability to inhibit the binding of the first antibody to the antigen-coated well. Specificity of the assay depends on the purity of the adsorbed antigen and allows for highly specific tests. Under optimum conditions the sensitivity of the assay is determined by the KB of the antibody and allows 10–20 ng of a specific type of collagen or laminin and 1 ng fibronectin to be detected.

The nature of the collagenous protein synthesized from bone and cartilage mRNA in an in vitro reticulocyte translation system has been investigated. The properties of a significant portion of the collagenous protein are those of native rather than denatured procollagen. It appears that the polypeptide chains contain sufficient information to direct chain association.

The interaction of fibronectin with native collagen during collagen fibril formation was investigated. Fibronectin prepared from serum, or from the cell surface, bound to the forming collagen fibrils while less fibronectin bound to preformed fibers. Denatured collagen competed with native collagen in binding fibronectin. Fibronectin delayed the precipitation of collagen fibrils but did not alter the total amount of fibrils formed. Fibronectin which was heated to 30 degrees C for 30 min did not promote cell adhesion but still bound to native collagen and delayed fiber formation. The collagen-binding fragment of fibronectin produced by digestion either with chymotrypsin or with neutrophil elastase had a similar effect in delaying fibril formation, but the cell-binding fragment was not active. These studies indicate that fibronectin can bind to aggregating collagen fibers probably at the same site shown previously to bind to denatured collagen. Since fibronectin inhibits the rate of collagen fibrillogenesis, it may regulate the size of collagen fibers.

We have studied the attachment of cultured Chinese hamster ovary cells to collagen substrates prepared in several ways. The attachment of these cells to collagen required under most conditions either serum or fibronectin purified from serum. Reconstituted collagen substrates required greater amounts of fibronectin than dishes coated by drying a collagen solution, but in each case the amount of fibronectin required was proportional to the amount of collagen on the dish. High levels of phosphate, 0.01 m and above, used as a buffer in heat-reconstituted collagen substrates allowed cell attachment without fibronectin. However, since the cells did not spread under these conditions and were not released from the substrate when incubated with trypsin, binding of cells with such levels of phosphate probably represents nonphysiological adhesion.

Extracellular matrices have diverse biological effects, including promoting the growth and differentiation of various cells of epithelial origin. The components of one of these matrices, the basement membrane, are discussed, as well as studies using these components alone or in combination with cells in culture. The particular response observed varies with the cell type examined and appears to be dependent on multiple interactions with components of the matrix. Potential uses for a basement membrane-derived matrix in vitro and in vivo are being developed.

CONTENTS INTRODUCTION.. .• .••.....• .•••.. 57 STRUCTURE 58 INTERACTIONS WITH OTHER BASEMENT MEMBRANE PROTEINS ......•• 65 BIOSYNTHESIS AND REGULATION •.•....• • • 69 CELL BINDING AND OTHER BIOLOGICAL PROPERTIES OF LAMININ 71 MOLECULAR MODELS OF BASEMENT MEMBRANE STRUCTURE •.. 76

Metastatic mouse melanoma cells have a high affinity for the basement membrane and the ability to degrade it; these properties may allow tumor cells to invade the membrane and disseminate. In this study it was found that the metastatic potential of mouse melanoma cells varied when the cells were exposed in culture to fibronectin or laminin. After removal of fibronectin or exposure to laminin, the cells had an increased affinity for basement membrane collagen, were more invasive of basement membranes in vitro, and produced more lung colonies in vivo. These changes are correlated with and may be due to an increase in the laminin-binding capacity of the tumor cell surface.

The invasive and metastatic characteristics of cloned cells derived from the K-1735 murine melanoma were investigated. Cell lines which are highly metastatic in mice were found to be invasive in vitro, and to show an enhanced attachment to, spreading on and migration toward laminin. As attachment, spreading and directional migration are thought to be receptor-mediated events, the binding of laminin to these cells was studied. Biotinylated laminin was used to evaluate receptor binding by fluorescence activated cell sorting (FACS) and this method was compared with that in which the binding of radioactive laminin is measured. Both studies revealed that metastatic K-1735 cells (a) have more receptors for laminin compared with non-metastatic cells and (b) exhibit a second population of low-affinity binding sites not present on the non-metastatic cells. The differences in receptor number and type may account for the greater interaction of metastatic cells with laminin and their invasive phenotype.

Inbred mice bearing certain alleles at the Mottled locus have defects in connective tissue which result in weakness of skin and of blood vessels. Previous studies have established that cross-links in collagen and elastin are decreased in these animals due to impaired formation of lysine-derived aldehydes. Lysyl oxidase activity in extracts of skin is markedly lower in those prepared from affected animals than control mice. An inhibitor of lysyl oxidase is present in equal amounts in affected and control skins and does not account for diminished activity found in affected animals.

The collagenous proteins synthesized by placenta in organ culture were characterized. Types I and III collagen accounted for about two-thirds of the collagenous protein produced by the tissue while type IV procollagen comprised around 10%. Type IV collagen contained two chains of 185,000 and 175,000 daltons which are genetically distinct from one another as determined by a peptide mapping procedure. The type IV procollagen was identical to that produced by other tissues based on ratios of 3- to 4-hydroxyproline and hydroxyproline to proline, and on the pattern produced upon polyacrylamide gel electrophoresis before and after peptide mapping procedures. About 20% of the collagen resembled type V collagen in the proportions of 3- and 4-hydroxyproline to proline and of hydroxylysine to lysine, in solubility, and in peptide maps. However, it contained disulfide linked chains larger than those found in the type V collagen solubilized by pepsin. Following pepsin treatment, the disulfide bonds were removed and the mobility of the chains of the labeled protein resembled those in type V collagen. It is likely that the disulfide linked protein represents the intact type V collagen molecule.

We have studied a murine tumor previously classified as a poorly differentiated chondrosarcoma. Although the cells in this tumor are surrounded by large quantities of extracellular matrix material, the matrix fails to react with stains specific for the sulfated glycosaminoglycans present in normal cartilage. Here we show at the ultrastructural level that the tumor matrix is a homogeneous, nonfibrillar material, resembling basement membrane. Neither the proteoglycan matrix granules nor collagen fibrils characteristic of cartilage are present in the tumor matrix. Amino acid analyses of whole tumor tissue, enzyme-solubilized tumor components, and the protein extracted from lathyritic tumors confirmed that the tumor matrix is a basement membrane collagen. The collagenous protein extracted from the tumor by nonenzymatic means contains three unique polypeptides larger than the alpha-chain components of the other types of collagen. These studies indicate that the tumor is not a type of chondrosarcoma but a basement membrane producing tumor.

We have found that the collagen from a patient with the Ehlers-Danlos syndrome type VII contained a polypeptide chain, pN alpha 2, not present in collagen prepared from normal tissue. Fibroblasts cultured from the patient's skin produced type I procollagen in which the NH2-terminal propeptide of pro alpha 2 was cleaved to about half of normal values by chick procollagen neutral protease which removes the NH2-terminal propeptides from procollagen (N-protease). The NH2-terminal propeptide on the pro alpha 2 chain of the patient's procollagen was also more resistant than procollagen from control fibroblasts to digestion by pepsin or alpha-chymotrypsin. assays for procollagen N-protease indicated that the patient's fibroblsts contained about the same level of enzymic activity as normal fibroblasts. These results suggest that the patient's fibroblasts synthesize both an abnormal pro alpha 2 chain and a normal pro alpha 2 chain. The abnormality probably consists of a structural mutation in or near the site at which procollagen N-protease cleaves the pro alpha 2 chain. The results presented here appear to provide the first example of a mutation in a structural gene for collagen. Since equal amounts of pN alpha 2 and alpha 2 are found in the protein in neutral salt extracts of the patient's tissue, as well as in newly synthesized collagen produced by cultured skin fibroblasts, and since both parents are phenotypically normal and express exclusively normal collagen chains, the patient is likely to be a sporadic heterozygote, arisen by new mutation, with one normal and one abnormal gene coding for pro alpha 2.

Methods are described for screening for teratogenic compounds using differentiating neural crest and prechondrogenic limb bud mesenchyme cells in culture. Substances to be tested are either added directly to the culture medium or are combined in a dialysis bag with the postmitochondrial fraction from rat liver and certain cofactors. In the latter case, the compound and its metabolites are gradually released into the medium from the dialysis bag. The results obtained with 14 compounds demonstrate a positive relationship between teratogenicity in vivo and alterations in the growth or the differentiation of the cultured cells.

A recent report suggests that fluoride has mutagenic activity in mice. To examine the potential clastogenic effect of ingested fluoride, we examined the frequencies of baseline SCE and mitomycin C induced SCE as well as baseline chromosomal aberrations and cell-cycle kinetics in mice raised on high and low fluoride diets. The lack of significant differences in any of these parameters between the two groups of animals indicates that dietary fluoride is not clastogenic and supports the continued use of water fluoridation.

Immunization of rabbits with guinea pig skin collagen elicited antibody with specificity for α2-CB1, a cyanogen bromide derived peptide from the amino-terminal region of α2, whereas, the α1 chain and peptides derived from it by cleavage with cyanogen bromide were essentially devoid of antigenic activity. Immunization with gelatin (denatured collagen) resulted in the formation of antibodies with specificities for α2-CB1 as well as to a portion of the α1 chain. Antiserum prepared against native collagen contained antibodies which reacted with antigens detected only in gelatin. The denaturation accompnying the transformation of collagen to gelatin is believed to expose immunogenic and antigenic sites previously hidden in the native molecule.

In recent years our understanding of the macromolecules in connective tissue has increased considerably. Much has been learned about the chemistry and manner of synthesis of these substances, but the determination of their normal functions and their involvement in disease states has been less successful. In this paper, I summarize recent studies on collagen. Certain controversial or little noted questions will be emphasized along with recent developments in methodology. Wound healing and the removal of collagen are discussed elsewhere in this conference.

l -Ascorbic acid-1-C 14 slowly equilibrates in various tissues when injected into rats, whereas dehydroascorbie acid-1-C 14 equilibrates more rapidly. Dehydroascorbic acid was found to penetrate into brain, eye, and erythrocytes to a greater extent than ascorbic acid. Intravenous administration of large doses of ascorbic acid (20 mg.) to nephrectomized rats or the induction of accelerated ascorbic acid biosynthesis in such animals resulted in a marked accumulation of the vitamin in serum without appreciable tissue transfer. A similar amount of exogenous dehydroascorbic acid, the oxidized nonionic form of the vitamin, was largely removed from the serum. The ascorbic acid in serum of blood drawn from the renal vein of dogs was primarily in an oxidized form, whereas serum from the carotid artery and from the femoral and jugular veins was largely in the reduced form. These findings suggest that ascorbic; acid enters tissues as dehydroascorbic acid and that the kidney may play a role in the distribution of the vitamin by oxidizing it to the un-ionized, easily penetrable form.

Summary β-Aminopropionitrile binds to and inactivates lysyl oxidase, the enzyme carrying out the first step in the cross-linking of collagen and elastin.

After a lag period, rat aortas incubated in rat serum in vitro accumulated substantial quantities of calcium and phosphate. Examination of the tissue by x-ray diffraction, microradiography, electron diffraction, and electron microscopy indicated that the calcium-phosphate phase which formed was hydroxyapatite and that the crystals were localized almost exclusively in elastin. Selective elimination of various components of the aorta with proteolytic enzymes indicated that the presence of elastin was required for mineralization. Collagen fibers did not appear to be required for the initiation of calcification, nor did they seem to undergo appreciable calcification in the time periods studied. Analysis of the initial lag period suggested that at least two changes occurred in serum prior to the mineralization of this tissue. Inhibitors of the reaction were destroyed, and the level of dialyzable calcium was elevated owing to its release from serum protein.

The function of the aorta in aortic calcification in vitro has been investigated. In this system calcification does not depend on metabolic processes in the aorta but on the nucleating activity of elastin fibers. Homogenates of aorta calcify, but fractions isolated by centrifugation from the homogenate do not. However, the 1000 g sediments from aortas incubated in serum for a relatively short time prior to homogenization do calcify. These results suggest that more than one component in the aorta is necessary for mineral induction. Treatment of aortas and nuchal ligaments, a tissue consisting chiefly of elastin, with group-blocking reagents indicates that both sulfhydryl and amino groups are involved in the calcification process. Amino groups appear to participate at a later stage in the process than sulfhydryl groups.

The incorporation of [ 14 C]glycine into the subunits of skin collagen was followed in normal and lathyritic rats. Label appears first and at the same rate in the primary subunits, α1 and α2, and later in β1 and β2. This is explained by the fact that β1 and β2 are derived from α1 and α2 by intramolecular crosslinking. In lathyritic animals the formation of β1 and β2 is markedly reduced though the synthesis of collagen proceeds. This defect is evident long before the appearance of the gross symptoms of lathyrism. The results are consistent with the suggestion that the lack of intramolecular crosslinks affects the integrity of connective tissue through a loss of fibril stability resulting in the gross symptoms characteristic of lathyrism.

The enzyme-dependent production of cross-linked collagen and the lysine-derived cross-link precursor allysine (α-aminoadipic δ-semialdehyde) has been demonstrated. The enzyme, lysyl oxidase, was extracted from embryonic chick cartilage and was inhibited by the lathyrogen β-aminopropionitrile. Cross-linking of collagen was indicated by production of a β12 component of collagen and insoluble collagen in the samples incubated with lysyl oxidase. Allysine production was observed in both α1 and α2 chains. Approximately three times as much was present in the α1 as in the α2 chain. The only lysyl residue that was found to be converted to allysine in the α1 chain was Residue 9 from the amino-terminal end. Since this is the residue that is converted in vivo and β-aminopropionitrile inhibits the conversion in vivo, it is concluded that lysyl oxidase is the physiologically active enzyme necessary for the first step in the cross-linking of collagen.

Collagen extracted from the diaphyseal region of the tibiae of normal and lathyritic chicks with 0.5 m acetic acid and 5 m guanidine hydrochloride at neutral pH has been studied with regard to chain organization and amino acid composition. Decalcification by dialysis against large quantities of 0.5 m acetic acid and subsequent exposure of the decalcified bones to fresh 0.5 m acetic acid followed by 5 m guanidine resulted in the solubilization of 85% of the total collagen in the bones from lathyritic animals. Collagen in the bones of control animals was almost entirely resistant to extraction with 0.5 m acetic acid, and less than 20% of the total collagen could be extracted in 5 m guanidine. These observations suggest a high degree of intermolecular cross-linking. Chromatography of the collagen extracted by both solvents from the bones of lathyritic animals on columns of carboxymethyl cellulose indicated that the extracts were composed almost entirely of single- and double-chain components. Carboxymethyl cellulose chromatography of the collagen extracted with 5 m guanidine from the bones of normal chicks revealed that only 30% of the dissolved collagen (5% of the total collagen) could be recovered as these components. Chromatography on an agarose molecular sieve column showed that the remainder could be accounted for in part by high molecular weight aggregates and in part by low molecular weight, degraded material. These results suggest that bone collagen fibrils are stabilized through the formation of intra- and intermolecular cross-links and that lathyrogens exert their effects by inhibiting cross-link formation, as previously shown for soft tissue collagens. The extractability of normal bone collagen in guanidine appears to result from degradation in the presence of the denaturing agent to yield a highly heterogeneous mixture. Carboxymethyl cellulose chromatography of acid-extracted collagen from the bones of lathyritic animals indicated that the bone collagen molecule contains two chains (α1) that are identical or similar and one chain (α2) that is unlike the other two. Their compositions are analogous to those of other vertebrate collagens. Amino acid analyses of the chromatographically separated α chains in successive extracts of collagen from the lathyritic bones indicated that during development there was a progressive decrease in hydroxylysine content, accompanied by an equivalent increase in lysine content, with no significant change in the content of any other amino acid. This observation was verified by amino acid analyses of the total bone collagen of normal chicks of various ages. The hydroxylysine content decreased by nearly one-half, from 11 to 6 residues per 1000, between the 1st and the 49th day after hatching. This result suggests that the hydroxylation of bone collagen lysyl residues is dependent in part upon factors extrinsic to the collagen molecule.

In order to determine whether a specific subpopulation of epidermal cells selectively attaches to collagen substrates in vitro, epidermal cell suspensions, obtained by trypsinization of guinea pig skin, were incubated on type I or type IV collagen-coated glass cover slips. It was noted, morphologically and by electronic volume measurements, that small round cells, as opposed to the larger angulated flat cells, adhered to the collagen substrates. To further characterize the attached cells, the percentage of basal cells was determined in the attached cell population and in the initial epidermal cell suspension. Basal cells were identified by indirect immunofluorescence in 2 ways: (1) by the presence of pemphigoid antigen and (2) by the absence of upper cytoplasmic antigen, which is present in all keratinocytes except the basal cells. Whereas in the initial guinea pig epidermal cell suspensions about 50% of the cells were basal cells using either of these 2 criteria, 86-97% of the cells which adhered to the collagen substrates were basal cells. Human basal cells, as defined by pemphigoid antigen, also selectively adhered to the collagen substrates.

The members of the collagen family have diverse molecular structures and functions, and form a variety of histologically distinct matrix elements.

The invasion of tumor cells through basement membranes is a critical step in the formation of metastases. The binding of the malignant cells to laminin in the basement membranes allows their attachment and activates their invasiveness. Recently a synthetic nonapeptide from the B1 chain sequence of laminin was identified as a major site for cell binding. A pentapeptide within the nonapeptide sequence was found to reduce the formation of lung colonies in mice injected with melanoma cells and also to inhibit the invasiveness of the cells in vitro.

One of the genetically distinct collagens (type III) normally found in skin, aorta, and intestine is missing from the tissues of patients with the Ehlers-Danlos syndrome type IV. While skin fibroblasts from other individuals synthesize both types I and III collagen. Ehlers-Danlos syndrome IV cells synthesize only type I. These results suggest that the fragile skin, blood vessels, and intestines of Ehlers-Danlos syndrome IV patients result from an absence of type III collagen.

Macrophages incubated with products (lymphokines) secreted by stimulated spleen cells produced collagenase. Active lymphokines were obtained both from mitogen- and antigen-stimulated lymphocytes. These observations suggest that the degradation of collagen in chronic inflammatory lesions may be caused by macrophage collagenase.

Laminin is a large, cross-shaped protein composed of several domains with distinct activities. It promotes the binding of cells to basement membranes and links various macromolecules in the matrix. The structure--function relationship of laminin and its role in embryogenesis and various diseases is under investigation. Basement membranes are extracellular, sheet-like structures which separate epithelium from undedying connective tissue and surround nerve, muscle and fat