Interaction of a 59-kDa connective tissue matrix protein with collagen I and collagen II

We have studied binding to collagen of the 59-kDa protein present in most connective tissues. Collagen fibril formation, measured as increasing turbidity, was markedly retarded and reduced by the presence of small amounts of this protein. This was true for both collagen I and collagen II. The effect was also observed when pepsin-treated collagens were used, indicating that interaction with the telopeptides is not involved. The proportion of collagen precipitated in the assay was not or only marginally reduced. Thus, the altered optical properties indicate that structurally different fibrils are formed in the presence of the 59-kDa protein. The 59-kDa protein bound to collagen I or collagen II that had been insolubilized on polystyrene 96-well microtiter plates, as measured by enzyme-linked immunosorbent assay. Analogously, binding to the collagens was demonstrated for the PG-S2 low Mr proteoglycan, previously shown only to inhibit collagen fibrillogenesis. The two matrix components showed similar strength of binding, i.e. Kd 35 nM for the 59-kDa protein and 16 nM for PG-S2 at 20 degrees C. The results do not reveal if the collagen interaction site of the 59-kDa protein is different from that of PG-S2. Our observations do, however, suggest that the 59-kDa protein, as well as PG-S2, have functions related to the regulation of collagen organization in tissues.     

Interaction of vitronectin with collagen

Purified human plasma vitronectin was demonstrated to bind to type I collagen immobilized on plastic as measured by enzyme-linked immunosorbent assay and by binding of 125I-radiolabeled vitronectin to a collagen-coated plastic surface. Vitronectin did not bind to immobilized laminin, fibronectin, or albumin in these assays. Vitronectin showed similar interaction with all types of collagen (I, II, III, IV, V, and VI) tested. Collagen unfolded by heat treatment bound vitronectin less efficiently than native collagen. Vitronectin-coated colloidal gold particles bound to type I collagen fibrils as shown by electron microscopy. Salt concentrations higher than physiological interfered with the binding of vitronectin to collagen, suggesting an ionic interaction between the two proteins. Binding studies conducted in the presence of plasma showed that purified vitronectin added to plasma bound to immobilized collagen, whereas the endogenous plasma vitronectin bound to collagen less well. Although fibronectin did not interfere with the binding of vitronectin to native collagen, vitronectin inhibited the binding of fibronectin to collagen. These results show that vitronectin has a collagen-binding site(s) which, unlike that of fibronectin, preferentially recognizes triple-helical collagen and that the binding between vitronectin and collagen has characteristics compatible with the occurrence of such an interaction in vivo.     

Interaction of water with native collagen

The interaction of water with collagenous tissue was investigated using dynamic mechanical spectroscopy and cryogenic X-ray techniques. The loss spectrum was found to be very sensitive to water which is highly associated with the macromolecule. Two water-sensitive loss peaks were observed below 0°C: the β₂ or “water dispersion” at 150°K and the β₁ at 200°K which is attributed to motion of polar side chains. Changes in peak temperature and intensity were not continuous with water content, but exhibited regimes in behavior which were associated with two types of nonfreezable water, structural and bound water. In cryogenic X-ray experiments, specimens which contained some freezable water exhibited reflections identified with the cubic form of ice. These ice crystals underwent an irreversible transition to the more common hexagonal form when warmed above 200°K. On the basis of these experiments, a model for the hydration of native collagenous tissue was proposed.     

Interaction of osteogenin, a heparin binding bone morphogenetic protein, with type IV collagen

Osteogenin, an extracellular matrix component of bone, is a heparin binding differentiation factor that initiates endochondral bone formation in rats when implanted subcutaneously with an insoluble collagenous matrix. We have examined the interaction of osteogenin with various extracellular matrix components including basement membranes. Osteogenin, purified from bovine bone, binds avidly to type IV collagen and to a lesser extent to both type I and IX collagens. Osteogenin binds equally well to both native and denatured type IV collagen. Both alpha 1 and alpha 2 chains of type IV collagen are recognized by osteogenin. Osteogenin binds to a collagen IV affinity column, and is eluted by 6.0 M urea with 1 M NaCl, pH 7.4, and the eluate contained the osteogenic activity as demonstrated in vivo. Binding of osteogenin to collagen IV is not influenced by either laminin or fibronectin. These results imply that osteogenin binding to extracellular matrix components including collagens I and IV and heparin may have physiological relevance, and such interactions may modulate its local action.     

Interaction of type-I collagen with phospholipid monolayer.

The effects of type-I collagen on dipalmitoyl phosphatidylcholine (DPPC) and dimyristoyl phosphatidylcholine (DMPC) monolayer films with different compositions were studied using monolayer technique. The addition of collagen in the subphase of different monolayer films induced a considerable shift towards larger area/molecule in the compression-isotherm curves. This is either referred to the insertion of collagen into the monolayer by its hydrophobic residues or to an adsorption process causing a protein layer to be located parallel to the lipid monolayer [1]. The variation of collagen interaction with different lipid compositions was also verified through the penetration-kinetics experiment. Comparing our results to the results of Pajean et al. [2] and Pajean and Herbage [3] on the effect of collagen on the stability of lipid vesicles implies that the collagen induced stability could be explained on the basis of collagen-lipid monolayer interaction.     

Interaction of bilirubin with reconstituted collagen fibrils.

1. Interaction of bilirubin with collagen fibrils was explored in a two-phase system where collagen was present as an opaque rigid gel composed of striated fibrils, and bilirubin as an aqueous solution. 2. The Ka value of the binding of bilirubin to collagen fibrils is 5.4 X 10(3)M-1. The interaction of bilirubin with collagen fibrils depends on temperature. Below 5 degrees C, the binding is greatly diminished and denaturation of collagen fibril aggregates at 52--53 degrees C into a dissolution state abolishes binding of bilirubin. 3. Salicylate and sulphanilamide do not affect the binding of bilirubin to reconstituted collagen fibrils. 4. Serum albumin (40--80mM), known to reverse the binding of bilirubin to lipids, dissociates only 50% of the bilirubin bound to collagen fibrils. This suggests that sites located on collagen participate in some tight binding of bilirubin and the corresponding binding sites on albumin do not compete with them. 5. Urea (4M) abolishes more than 70% of the binding of bilirubin to collagen. Urea and thermal denaturation studies indicate the importance of conformation and organization of collagen fibrillar aggregates for the binding of bilirubin.     

Interaction of collagen with hydrophobic protein granules in the egg capsule of the dogfish scyliorhinus canicula

The egg capsule of the dogfish is a composite material containing collagenous fibrils and 2 mum spherical hydrophobic protein granules. The latter appear to owe much of their hydrophobicity to an exceptionally high tyrosine content (approximately 20% of total amino acid residues). The hydrophobic component appears to form as an emulsion in the secretory granules of the D and E zone gland cells of the nidamental gland. Droplets of the hydrophobic material appear to become coated with remarkably regular layers of radially-arranged collagen molecules which form a series of concentric, evenly spaced layers around each hydrophobic granule. Numerous disclinations were seen where the layers around adjacent granules interfered with one another. The layers are thought to represent a lamellar liquid crystalline phase previously described for this collagen (Knight et al., 1993). The fine structural appearance of the concentric layers and evidence for radial arrangement of collagen molecules within them is compatible with the suggestion that the layers are built from a dumbbell-shaped unit approximately 35 nm long with hydrophobic groups concentrated at the ends. This unit may represent a dumbbell-shaped molecule or an oligomer of two or more molecules lying parallel with one another in a head-to-tail arrangement. Such a unit can be readily incorporated into models for the micellar, hexagonal columnar and final fibrillar phases previously described for this collagen (Knight et al., 1993). Evidence from the TEM study of stretched egg capsule wall suggests that there is a mechanical interaction between the hydrophobic granules and the collagen fibrils in the fully formed material. We suggest that the radial, concentric layered arrangement of collagen molecules is established by hydrophobic interactions within the liquid crystalline material and locked into place by oxidative covalent cross-linking to give a 3-dimensional cross-linked meshwork of collagen fibrils and hydrophobic granules. The latter arrangement helps to account for the high tensilestrength and toughness of this material.     

An N-terminal peptide from link protein can stimulate biosynthesis of collagen by human articular cartilage

Previous studies have shown that a peptide identical in sequence to the N-terminal of link protein can function as a growth factor and up-regulate proteoglycan synthesis by human articular cartilage in explant culture (L. A. McKenna et al., Arthritis Rheum. 41, 157-162, 1998). The present study has extended these investigations to determine the effects of this peptide on the synthesis of collagen, another essential component of normal cartilage matrix. Explants from normal adult knee cartilage were maintained for periods of up to 8 days in medium with or without serum. Peptides were added during each day of culture. Synthesis of collagen was determined by the incorporation of [3H]proline into hydroxyproline and proteoglycans by incorporation of [35S]sulfate. The type of newly synthesized collagen was measured by SDS-polyacrylamide gel electrophoresis, fluorography, and immunoblotting. The link protein peptide stimulated synthesis of type II collagen in cartilage from a number of different subjects. Maximum up-regulation of synthesis was attained at a concentration of 100 ng/ml, similar to that observed previously for up-regulation of proteoglycan. Synthesis was up-regulated in both the presence and the absence of serum, although the overall rate of synthesis was greater when serum was added. The findings that this link peptide growth factor stimulated synthesis of proteins, including collagen, in a manner analogous to that shown previously for proteoglycans support the hypothesis that this peptide may have an important role in the feedback control of cartilage matrix synthesis.     

Interaction of intact type VI collagen with hyaluronan.

The capacity of non-pepsinyzed type VI collagen to bind to hyaluronan was investigated. Type VI collagen was extracted from bovine meniscal cartilage with 6 M GuHCl and purified by extraction of PEG precipitates and dissociative Sephacryl S-500 HR chromatography. Type VI collagen, detected with a monoclonal antibody, bound in 0.5 M NaCl to hyaluronan-coated micro-wells, the degree of binding being higher at 37 degrees C than 23 degrees C and 4 degrees C. Incubation of type VI collagen in competitive inhibition assays with testicular hyaluronidase digests of hyaluronan in liquid phase, reduced binding of the protein to hyaluronan-coated microwells to background levels. Thus, non-pepsinyzed type VI collagen binds to hyaluronan in vitro.     

Aggrecan and link protein affect cell adhesion to culture plates and to type II collagen

Cartilage is a hypocellular tissue in which a balance of matrix molecules, especially aggrecan and link protein, play a critical role in maintaining structural integrity. To study the role of aggrecan and link protein in mediating cell activities, we have stably expressed them in NIH/3T3 fibroblasts and observed the effect on cell-substratum interactions. Overexpression of either protein destabilized the cell-substratum interaction. However, when both were co-expressed, the interaction between cell and substratum was less impaired. Similar results were obtained on type II collagen-coated plates. The addition of exogenous gene products into fibroblast cell lines and chondrocyte culture had the same effect as expression of the genes. The addition of exogenous hyaluronan to the growth medium or treatment of cells with hyaluronidase also decreased cell adhesion, indicating that hyaluronan also plays a role in the cell-substratum adhesion. The presence of aggrecan seems to increase the amount of link protein on the cell surface. Chondrocytes expressing high concentrations of aggrecan and link protein were maintained within a matrix network and were able to survive in suspended culture. Imbalances in aggrecan or link protein concentrations, or degradation of hyaluronan, disrupted the network and caused the chondrocytes to aggregate or adhere to the plates.     

Interaction during fibril formation of soluble collagen with cartilage proteinpolysaccharide

Precipitation of soluble forms of collagen from solutions containing the soluble protein-polysaccharide (PP-L) of bovine nasal cartilage, followed by centrifugation at 100,000 g, resulted in the formation of coherent elastic pellets whose wet weights increased with the concentration of PP-L in the initial solution. Dry weights and uronic acid contents of these pellets showed that the amount of water held in the wet pellet was nearly constant for any one kind and concentration of collagen, and ranged from 20 to 100 mg./mg. PP-L in the pellet. Soluble collagens from four different sources and PP-L from three kinds of cartilage showed similar effects. Precipitation of soluble collagen in the presence of hyaluronate or dextran yielded pellets of much smaller size than those formed in the presence of PP-L. The presence of chondroitin sulfate had only a slight effect on wet pellet weights. Wet weights of pellets formed in the presence of PP-L decreased with increasing ionic strength. A model involving entanglement between insoluble collagen fibrils and the relatively stiff chondroitin sulfate chains of branched PP-L seems qualitatively capable of accounting for these results.     

Interaction of human platelet membrane glycoproteins with collagen and lectins

The binding of platelets to collagen is the first step in hemostasis. We attempted three approaches for elucidation of the chemical nature of receptors of human platelets for collagen. First, we examined the effect of platelet surface alteration by chymotrypsin treatment. On increasing the concentration of chymotrypsin, collagen-induced platelet aggregation and the release reaction decreased, and in parallel with this change, remarkable decrease of membrane glycoproteins IIb and V, as well as 400 kDa and 300 kDa membrane proteins, was observed. Secondly, effects of several lectins on the platelet-collagen interaction were examined. Lens culinaris agglutinin was found to specifically inhibit the platelet aggregation and release reaction induced by collagen. This inhibition appeared to be caused mainly by blocking of the collagen receptors on platelets by Lens culinaris agglutinin. Furthermore, Lens culinaris agglutinin was found to bind preferentially to glycoprotein IIb as identified by sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis of platelet membranes followed by staining with ¹²⁵I-Lens culinaris agglutinin. In addition, a polymerized preparation of Lens culinaris agglutinin induced platelet aggregation. Thirdly, the membrane component which could bind to collagen-Sepharose 4B was determined. Analysis by SDS-polyacrylamide gel electrophoresis combined with autoradiography or fluorography revealed that glycoprotein IIb was most enriched in the bound fraction to collagen. From these results, glycoprotein IIb is most likely a receptor for collagen on human platelet membranes.     

Interaction of discoidin domain receptor 1 with collagen type 1

Discoidin domain receptor 1 (DDR1) is a widely expressed tyrosine kinase receptor which binds to and gets activated by collagens including collagen type 1. Little is understood about the interaction of DDR1 with collagen and its possible functional implications. Here, we elucidate the binding pattern of the DDR1 extracellular domain (ECD) to collagen type 1 and its impact on collagen fibrillogenesis. Our in vitro assays utilized DDR1-Fc fusion proteins, which contain only the ECD of DDR1. Using surface plasmon resonance, we confirmed that further oligomerization of DDR1-Fc (by means of anti-Fc antibody) greatly enhances its binding to immobilized collagen type 1. Single-molecule imaging by means of atomic force microscopy revealed that DDR1 oligomers bound at overlapping or adjacent collagen molecules and were nearly absent on isolated collagen molecules. Interaction of DDR1 oligomers with collagen was found to modulate collagen fibrillogenesis both in vitro and in cell-based assays. Collagen fibers formed in the presence of DDR1 had a larger average diameter, were more cross-linked and lacked the native banded structure. The presence of DDR1 ECD resulted in "locking" of collagen molecules in an incomplete fibrillar state both in vitro and on surfaces of cells overexpressing DDR1. Our results signify an important functional role of the DDR1 ECD, which occurs naturally in kinase-dead isoforms of DDR1 and as a shedded soluble protein. The modulation of collagen fibrillogenesis by the DDR1 ECD elucidates a novel mechanism of collagen regulation by DDR1.     

Interaction of embryonic corneal epithelium with exogenous collagen,laminin, and fibronectin: Role of endogenous protein synthesis

Previously, we have shown that the embryonic corneal epithelium is capable of interacting with exogenous collagen, laminin, and fibronectin in soluble form, each of which causes isolated epithelium cultured on Millipore filter to stop blebbing, reorganize the basal cytoskeleton, and flatten. Here we examine the involvement of endogenously derived extracellular matrix (ECM) molecules in the interaction of the basal epithelial cell surface with the added ECM molecules. We demonstrate here that the isolated avian corneal epithelium cultured on Millipore filter is capable of synthesizing collagens and laminin, but not fibronectin. To examine whether the epithelium is capable of interacting directly with exogenous ECM components or if there is the necessity for production of a linker molecule, epithelial protein synthesis was inhibited with cycloheximide (CHX). The blebbing epithelium in the presence of CHX was then confronted with soluble ECM molecules added to the medium under the filter; such epithelia are able to interact with, and flatten in response to, both collagen and laminin. However, such inhibited epithelia continue to bled in the presence of fibronectin. We next used l-azetidine-4-carboxylic acid (LACA) to interfere with collagen secretion. Epithelia exposed to LACA are still capable of interacting with collagen and laminin, but not fibronectin, indicating a dependence on collagen secretion. These results suggest that fibronectin requires a linker protein, probably collagen, to interact with the basal epithelial surface, whereas both collagen and laminin may interact directly with the cell surface to transform the basal cytoskeleton into the cortical mat typical of differentiating corneal epithelium in situ.     

Interaction of PDZRhoGEF with microtubule-associated protein 1 light chains: link between microtubules, actin cytoskeleton, and neuronal polarity

Rat (r) PDZRhoGEF, initially identified as a glutamate transporter EAAT4-associated protein, is a member of a novel RhoGEF subfamily. The N terminus of the protein contains a PDZ and a proline-rich domain, two motifs known to be involved in protein-protein interactions. By using the yeast two-hybrid approach, we screened for proteins that interact with the N terminus of rPDZRhoGEF. The light chain 2 of microtubule-associated protein 1 (LC2) was the only protein identified from the screen that does not contain a type I PDZ-binding motif at its extreme C terminus (-(S/T)Xphi-COOH, where phi is a hydrophobic amino acid). However, the C terminus does conform to a type II-binding motif (-phiXphi). We report here that rPDZRhoGEF interacts with LC2 via the PDZ domain, and the interaction is abolished by mutations in the carboxylate-binding loop. The specificity of the interaction was confirmed using GST fusion protein pull-down assays and coimmunoprecipitations. Expression of rPDZRhoGEF mutants that are unable to interact with proteins via the carboxylate-binding loop induced changes in cell morphology and actin organization. These mutants alter the activation of RhoGTPases, and coexpression of dominant-negative RhoGTPases prevent the morphological changes. Furthermore, in cells expressing wild type rPDZRhoGEF, drug-induced microtubule depolymerization produces changes in cell morphology that are similar to those induced by rPDZRhoGEF mutants. These results indicate that modulation of the guanine nucleotide exchange activity of rPDZRhoGEF through interaction with microtubule-associated protein light chains may coordinate microtubule integrity and the reorganization of actin cytoskeleton. This coordinated action of the actin and microtubular cytoskeletons is essential for the development and maintenance of neuronal polarity.     

Interaction of rat sertoli cells with a collagen lattice in vitro

Summary Sertoli cells from rats aged 15, 20, and 25 d were subcultured onto collagen-coated, plastic dishes. If the collagen was released from the plastic surface by rimming, the floating rats of collagen showed uniform shrinkage. If the collagen was allowed to remain attached to the plastic, holes appeared in the collagen with cells from rats aged 25 d but not with those of 15 d. Cells from rats aged 20 d caused fewer and smaller holes to appear. The holes were associated with the formation of clumps of spherical cells from which elongated Sertoli cells extended into the surrounding collagen to end near holes. Rhodamine-phalloidin revealed diffusely distributed actin in the spherical cells in contrast to well-developed microfilaments in the peripheral elongated cells. Addition of cytochalasin B (5 μg/ml) to the medium prevented contraction of the floating rats and the development of holes in the attached collagen. In addition, cytochalasin B caused the peripheral cells to become spherical and to separate from the clumps. Moreover, rhodamine-phalloidin revealed that actin in the peripheral cells occurred as clumps without microfilaments when cytochalasin B was present. When Sertoli cells were subcultured onto silicone rubber films, the cells produced wrinkling of the rubber surface within 4 h of plating. These observations were interpreted to mean that Sertoli cells exert local tractional forces on various substrata. These forces require actin, presumably acting by a contractile mechanism. When the collagen is attached to plastic and the cells are organized into clumps with radiating elongated cells (cells from rats aged 25 d), the tractional forces in the elongated cells reorganize the collagen fibers to produce holes. When cells are uniformly distributed (cells from rats aged 15 d), holes are not formed. When the collagen is released from the plastic surface, tractional forces cause the floating rafts to shrink. These interactions of the cells with collagen are likely to be important in determining the shape of the Sertoli cell in vivo, the polarity of the cell, and its biochemical differentiation. This investigation was supported by grants HD 16525, AM 32236, and GM 32705 from the National Institutes of Health, and from the Shriners of North America.     

Interaction of hydroxylated collagen IV with the von hippel-lindau tumor suppressor

The von Hippel-Lindau tumor suppressor (pVHL) targets hydroxylated alpha-subunits of hypoxia-inducible factor (HIF) for ubiquitin-mediated proteasomal destruction through direct interaction with the hydroxyproline binding pocket in its beta-domain. Although disruption of this process may contribute to VHL-associated tumor predisposition by up-regulation of HIF target genes, genetic and biochemical analyses support the existence of additional functions, including a role in the assembly of extracellular matrix. In an attempt to delineate these pathways, we searched for novel pVHL-binding proteins. Here we report a direct, hydroxylation-dependent interaction with alpha-chains of collagen IV. Interaction with pVHL was also observed with fibrillar collagen chains, but not the folded collagen triple helix. The interaction was suppressed by a wide range of tumor-associated mutations, including those that do not disturb the regulation of HIF, supporting a role in HIF-independent tumor suppressor functions.     

Interaction of calf skin collagen with glycerol: linked function analysis

Glycerol stabilizes the triple-helical structure of solubilized calf skin collagen. The equilibrium melting temperature of the protein increased linearly from 38.0 degrees C in AS buffer (0.01 M NaOAc and 0.02 M NaCl, pH 4.0) to 43.0 degrees C in AS and 6 M glycerol buffer. To understand the thermodynamic basis of this effect on the equilibrium melting temperature and the glycerol inhibition of collagen self-association, the preferential interactions of native and denatured calf skin collagens in AS buffer containing 1.5, 3, and 4.5 M glycerol were measured with a precision densimeter. The results indicated that native collagen binds glycerol preferentially whereas denatured collagen neither binds nor repels glycerol. The preferential binding of glycerol by native collagen, when interpreted in terms of the three-component solution thermodynamics, suggests that the surface interaction of native collagen with glycerol is energetically more favorable than its interaction with water. By use of the Wyman linked function, the negative chemical potential change of collagen derived from its preferential binding of glycerol can account for both the glycerol stabilization of the triple-helical structure of collagen and the inhibition of in vitro self-association of monomers into fibrils.     

Interaction of gene 5 protein with DNA

Gene 5 protein bound to both linear and circular single-stranded DNA and saturated the DNA at a protein-to-DNA weight ratio of 7–8. The viscosity of a complex of the protein with single-stranded DNA was initially less than that of the DNA and slowly increased with time suggesting that the complex adopts its final hydrodynamic shape very slowly. This shape change was confirmed by gradient centrifugation. The complex has a more extended structure than DNA alone accounting for its high viscosity and low S value. Gene 5 protein also bound to linear double-stranded DNA though not as strongly as to single-stranded DNA. The protein decreased the transition temperature, T_m, for viscosity loss of double-stranded DNA by 20 °C in 1 and 10 mm salt at a protein-to-DNA ratio of 2.2. At these low ratios there was no decrease in the hyperchromic T_m at 260 nm. At higher ratios of protein to DNA, the hyperchromic T_m was decreased to a constant value and not by a constant amount. Under no conditions was gene 5 protein able to completely separate the complementary strands of double-stranded DNA or to renature denatured DNA.