Extracellular matrix organization modulates fibroblast growth and growth factor responsiveness

To learn more about the relationship between extracellular matrix organization, cell shape, and cell growth control, we studied DNA synthesis by fibroblasts in collagen gels that were either attached to culture dishes or floating in culture medium during gel contraction. After 4 days of contraction, the collagen density (initially 1.5 mg/ml) reached 22 mg/ml in attached gels and 55 mg/ml in floating gels. After contraction, attached collagen gels were well organized; collagen fibrils were aligned in the plane of cell spreading; and fibroblasts had an elongated, bipolar morphology. Floating collagen gels, however, were unorganized; collagen fibrils were arranged randomly; and fibroblasts had a stellate morphology. DNA synthesis by fibroblasts in contracted collagen gels was suppressed if the gels were floating in medium but not if the gels were attached, and inhibition was independent of the extent of gel contraction. Therefore, growth of fibroblasts in contracted collagen gels could be regulated by differences in extracellular matrix organization and cell shape independently of extracellular matrix density. We also compared the responses of fibroblasts in contracted collagen gels and monolayer culture to peptide growth factors including fibroblast growth factor, platelet-derived growth factor, transforming growth factor-beta, and interleukin 1. Cells in floating collagen gels were generally unresponsive to any of the growth factors. Cells in attached collagen gels and monolayer culture were affected similarly by fibroblast growth factor but not by the others. Our results indicate that extracellular matrix organization influenced not only cell growth, but also fibroblast responsiveness to peptide growth factors.     

Beta 1 integrin-mediated collagen gel contraction is stimulated by PDGF

The attachment of primary rat hepatocytes and fibroblasts to collagen type I is mediated by non-RGD-dependent beta 1 integrin matrix receptors. In this report we describe a novel 96-well microtiter plate assay for the quantification of fibroblast-mediated contraction of floating collagen type I gels. Fetal calf serum and platelet-derived growth factor (PDGF), but not transforming growth factor-beta 1, stimulated primary rat heart fibroblasts and normal human diploid fibroblasts (AG 1518) to contract collagen gels to less than 10% of the initial gel volume within a 24-h incubation period. Rabbit polyclonal antibodies directed to the rat hepatocyte integrin beta 1-chain inhibited the PDGF-stimulated collagen gel contraction. The inhibitory activity on contraction of the anti-beta 1 integrin IgG could be overcome by adding higher doses of PDGF. The contraction process was not blocked by anti-fibronectin IgG nor by synthetic peptides containing the tripeptide Arg-Gly-Asp (RGD), in concentrations that readily blocked fibroblast attachment to fibronectin-coated planar substrates. Autologous fibronectin or control peptides containing the tripeptide Arg-Gly-Glu were without effect. Immunofluorescence microscopy on fibroblasts grown within collagen gels revealed a punctate distribution of the beta 1 integrin and a lack of detectable levels of endogenously produced fibronectin. Collectively these data suggest a role for integrin collagen receptors with affinity for collagen fibers, distinct from the previously described RGD-dependent fibronectin receptors, in the fibronectin-independent PDGF-stimulated collagen gel contraction process.     

Fibronectin and fibrinolysis are not required for fibrin gel contraction by human skin fibroblasts

Human skin fibroblasts contracted fibrin gels in a time- and cell-dependent manner. Under optimal conditions, gel contraction amounted to more than 50% in 2 hr. Fibronectin did not promote contraction, and fibrinolysis was not required for contraction, although gels contracted without serum or aprotinin were lysed. Before contraction, fibrin was present in loosely packed, randomly organized fibrils. After contraction, the fibrils were more densely packed and aligned in the plane of cell spreading. Cycloheximide treatment of fibroblasts inhibited gel contraction in serum-free medium but not in serum-containing medium. Fibronectin could not substitute for serum in overcoming the cycloheximide effect. Binding sites for fibrin were distributed randomly over the cells' surfaces based on electron microscopic observations. Often small groups of fibrils were localized in indentations at the cell surface. Finally, peptides containing the arg-gly-asp-ser sequence inhibited gel contraction.     

Collagen gel contraction by fibroblasts requires cellular fibronectin but not plasma fibronectin

Fibroblasts embedded in three-dimensional lattices of collagen fibrils have been known to require serum constituents to induce a cell-mediated contraction of collagen gels. The gel contraction was studied with human skin fibroblasts cultured in the presence of fetal bovine serum (FBS). Removal of bovine serum fibronectin (sFN) from FBS did not affect the extent of gel contraction. Gel contraction occurred in serum-free defined media. Therefore, it is concluded that sFN is not required for gel contraction. That cellular FN (cFN) synthesized and secreted by fibroblasts plays a crucial role in gel contraction was suggested by the following experiments: (1) We obtained monoclonal antibodies (mAb A3A5) against fibroblast surface antigens, which suppressed the fibroblast-mediated gel contraction. Immunoblot analyses showed that mAb A3A5 recognizes cFN secreted by human fibroblasts and human plasma FN (pFN), but not bovine sFN in FBS used for culture. (2) Addition of rabbit antisera, which recognize human cFN, to a serum-free gel culture inhibited contraction. Uninvolvement of human pFN in gel contraction was further confirmed by the fact that neither pretreatment of fibroblasts with excess amounts of human pFN nor the presence of excess amounts of human pFN in gels affected the extent of gel contraction. This study seems to be the first demonstration of functional difference between cFN and pFN (or sFN) and proposes a novel mode of binding of fibroblasts with collagen fibrils via cFN during cell-mediated collagen morphogenesis.     

β1 Integrin-mediated collagen gel contraction is stimulated by PDGF

The attachment of primary rat hepatocytes and fibroblasts to collagen type I is mediated by non-RGD-dependent β₁ integrin matrix receptors. In this report we describe a novel 96-well microtiter plate assay for the quantification of fibroblast-mediated contraction of floating collagen type I gels. Fetal calf serum and platelet-derived growth factor (PDGF), but not transforming growth factor-β1, stimulated primary rat heart fibroblasts and normal human diploid fibroblasts (AG 1518) to contract collagen gels to less than 10% of the initial gel volume within a 24-h incubation period. Rabbit polyclonal antibodies directed to the rat hepatocyte integrin β₁-chain inhibited the PDGF-stimulated collagen gel contraction. The inhibitory activity on contraction of the anti-β₁ integrin IgG could be overcome by adding higher doses of PDGF. The contraction process was not blocked by anti-fibronectin IgG nor by synthetic peptides containing the tripeptide Arg-Gly-Asp (RGD), in concentrations that readily blocked fibroblast attachment to fibronectin-coated planar substrates. Autologous fibronectin or control peptides containing the tripeptide Arg-Gly-Glu were without effect. Immunofluorescence microscopy on fibroblasts grown within collagen gels revealed a punctate distribution of the β₁ integrin and a lack of detectable levels of endogenously produced fibronectin. Collectively these data suggest a role for integrin collagen receptors with affinity for collagen fibers, distinct from the previously described RGD-dependent fibronectin receptors, in the fibronectin-independent PDGF-stimulated collagen gel contraction process.     

Fibroblast behavior on gels of type I, III, and IV human placental collagens

Various collagens were extracted and purified from human placenta after partial pepsin digestion. We prepared type III + I (57:43), enriched type I, type III, and type IV collagens on an industrial level, and studied their biological properties with MRC5 fibroblast cells. Using the process of contraction of a hydrated collagen lattice described by Bell, we found tha the contraction rate was dependent on collagen type composition. The contraction was faster and more pronounced with pepsinized type I collagen than with pepsinized type III + I (57:43) collagen; the lowest rate was obtained with the pepsinized type III collagen. Using a new technique of collagen cross-linking, a gel was made with type IV collagen. This cross-linking procedure, based on partial oxidation of sugar residues and hydroxylysine by periodic acid, followed by neutralization, resulted in an increased number of natural cross-link bridges between oxidized and nonoxidized collagen molecules, without internal toxic residues. The fibroblasts were unable to contract type IV/IVox collagen gels. The type IV/IVox collagen gel was transparent and its amorphous ultrastructure lacked any visible striated fibrils. Fibroblast cells exhibited atypical behavior in these type IV/IVox collagen gels as evidenced by optical and electron microscopy. The penetration of fibroblasts could be measured. Fibroblasts penetrated faster in type IV/IVox collagen gels than in untreated type III + I collagen gels. The lowest rate of penetration was obtained with cross-linked type III + I gels. Fibroblast proliferation was similar on untreated or cross-linked type III + I collagen gels and slightly increased on type IV/IVox collagen gels, suggesting that this cross-linking procedure was not toxic.     

The behaviour of BHK cells and neutrophil leukocytes on collagen gels of defined mechanical strength

This study demonstrates how the mechanical strength of a series of collagen/composite gels can be measured using a penetrometer. It was found that the presence of fibrin in collagen gels resulted in increased gel strength. Similarly hyaluronic acid was found to increase the strength of collagen gels. Addition of heparin weakened collagen gels as did chondroitin-6-sulphate. Neutrophil migration into collagen gels was found to be inversely proportional to gel strength. Fibrin and hyaluronic acid containing gels inhibited neutrophil migration while the presence of heparin and chondroitin sulphate increased neutrophil migration. BHK gel contraction experiments demonstrated how the presence of fibrin prevents gel contraction. Despite increasing gel strength the presence of hyaluronic acid appeared to have no effect on BHK contraction of collagen gels. Similarly the presence of heparin or chondroitin sulphate had no effect on gel contraction by BHK cells.     

Collagen type V enhances matrix contraction by human periodontal ligament fibroblasts seeded in three-dimensional collagen gels

Extracellular matrix components play an important role in modulating cellular activity. To study such capacities of the matrix, fibroblasts are frequently cultured in a three-dimensional gel and contraction is assessed as a measure of cellular activity. Since a connective tissue contains several types of collagen, we investigated the effect of gels composed of collagen I alone or in combination with 10% collagen III and/or 5% collagen V on contraction by human periodontal ligament fibroblasts. Gels containing collagen V contracted much faster than those without this type of collagen. Blocking of the integrin beta1-subunit with an activity-blocking antibody delayed (gels with collagen V) or almost completely blocked (gels without collagen V) contraction. Use of an antibody directed against integrin alpha2beta1 resulted in delay of gel contraction for gels both with and without collagen V. Anti-integrin alpha v beta3 or RGD peptides partially blocked contraction of gels containing collagen V, but had no effect on gels consisting of collagen I alone. The beta1-containing integrins are involved in the basal contraction by fibroblasts that bind to collagens I and III. The enhanced contraction, stimulated by collagen V, appears to be mediated by integrin alpha v beta3. We conclude that collagen V may play an important modulating role in connective tissue contraction. Such a modulation may occur during the initial stages of wound healing and/or tissue regeneration.     

Albumin Synthesis by Rat Hepatocytes Cultured on Collagen Gels Is Sustained Specifically by Heparin

We investigated the influence of various kinds of glycosaminoglycans (GAGs) in collagen gels on the maintenance of albumin synthesis in primary culture of rat hepatocytes. Among the GAGs examined (heparin, heparan sulfate, keratan sulfate, chondroitin sulfate A, dermatan sulfate, and hyaluronic acid), only heparin-containing collagen gel cultures could significantly sustain albumin synthesis. However, other GAGs, such as heparan sulfate and keratan sulfate, had almost no effect on the maintenance of albumin synthesis. Heparin in collagen gels exhibited a dose-dependent effect on albumin synthesis: heparin at 400 μg/ml-collagen solution maintained albumin synthesis for over 3 weeks. On the other hand, when an equivalent amount of heparin was added directly to the collagen gel culture medium, it prolonged albumin synthesis for only 10 days. The results demonstrate that specific regulation of albumin synthesis by heparin was significantly promoted by coincubating it with collagen, suggesting that some specific interaction between heparin and collagen might be of importance for the maintenance of hepatocyte functions.     

The interaction of human papillary and reticular fibroblasts and human keratinocytes in the contraction of three-dimensional floating collagen lattices

Fibroblasts derived from the papillary and reticular dermis of human skin and human keratinocytes show differences in their abilities to contract floating three-dimensional gels constructed from type I collagen. Reticular fibroblasts produce greater gel contraction than papillary fibroblasts. When equal numbers of papillary and reticular fibroblasts are mixed in the gels, papillary fibroblasts consistently inhibit gel contraction by reticular fibroblasts indicating interaction between these cell types in the contraction process. Surprisingly, keratinocytes alone produce greater gel contraction than that produced by either fibroblast type. Cooperativity in the gel contraction process is observed when fibroblasts are incorporated into the collagen matrix and keratinocytes are seeded onto the gel surface. Keratinocytes and dermal fibroblasts adhere to the collagen fibril to induce gel contraction by different mechanisms. Fibroblast contraction of collagen gels does not require fibronectin but is a serum-dependent reaction. In contrast, keratinocyte contraction of collagen gels occurs in a serum-free environment. Polyclonal, affinity-purified antibodies to human plasma fibronectin at high concentrations do not inhibit gel contraction by keratinocytes, making unlikely the possibility that fibronectin synthesized by the keratinocyte is a significant factor in the gel contraction process. We are currently examining the possibilities either that keratinocytes are synthesizing other adhesion proteins or that receptors on the cell surface can interact directly with the collagen fiber.     

Transforming growth factor-beta1 stimulates collagen matrix remodeling through increased adhesive and contractive potential by human renal fibroblasts

Renal tubulointerstitial fibrosis is the common final pathway leading to end-stage renal failure. Tubulointerstitial fibrosis is characterized by fibroblast proliferation and excessive matrix accumulation. Transforming growth factor-beta1 (TGF-beta1) has been implicated in the development of renal fibrosis accompanied by alpha-smooth muscle actin (alpha-SMA) expression in renal fibroblasts. To investigate the molecular and cellular mechanisms involved in tubulointerstitial fibrosis, we examined the effect of TGF-beta1 on collagen type I (collagen) gel contraction, an in vitro model of scar collagen remodeling. TGF-beta1 enhanced collagen gel contraction by human renal fibroblasts in a dose- and time-dependent manner. Function-blocking anti-alpha1 or anti-alpha2 integrin subunit antibodies significantly suppressed TGF-beta1-stimulated collagen gel contraction. Scanning electron microscopy showed that TGF-beta1 enhanced the formation of the collagen fibrils by cell attachment to collagen via alpha1beta1 and alpha2beta1 integrins. Flow cytometry and cell adhesion analyses revealed that the stimulation of renal fibroblasts with TGF-beta1 enhanced cell adhesion to collagen via the increased expression of alpha1 and alpha2 integrin subunits within collagen gels. Fibroblast migration to collagen was not up-regulated by TGF-beta1. Furthermore, TGF-beta1 increased the expression of a putative contractile protein, alpha-SMA, by human renal fibroblasts in collagen gels. These results suggest that TGF-beta1 stimulates fibroblast-collagen matrix remodeling by increasing both integrin-mediated cell attachment to collagen and alpha-SMA expression, thereby contributing to pathological tubulointerstitial collagen matrix reorganization in renal fibrosis.     

Cell-mediated contraction of collagen lattices in serum-free medium: Effect of serum and nonserum factors

Summary This study was conducted to identify a defined, serum-free culture medium that supports cell dependent contraction of a collagen lattice. Collagen lattices were found to contract in cultures containing human foreskin fibroblasts (HFF) or rabbit aortic smooth muscle (RASM) cells incubated in serum-free medium. HFF and RASM cells required different supplements to contract the collagen gels. HFF cultured in Dulbecco’s modified Eagle’s (DME) medium supplemented with bovine serum albumin (BSA) and either endothelial cell growth supplement (EnGS), insulin (In), or platelet derived growth factor (PDGF) supported collagen lattice contraction. Replacement of BSA with casein without the addition of other supplements improved contraction. In contrast, RASM cells supplemented with BSA, EnGS, In, and PDGF were able to contract collagen gels only minimally. Similar to HFF, RASM cells cultured in DME medium supplemented with casein, but without the addition of other supplements, contracted collagen lattices. HFF-mediated collagen contraction was inhibited by prostaglandins E₁ or E₂, fibronectin, or ascorbic acid. The reported serum-free model provides a useful in vitro method to investigate the role of serum and nonserum factors regulating cell mediated-contraction of insoluble collagen fibrils. Presented in part as abstract 1963 in the 1985 Federation of American Societies for Experimental Biology, April 21–26, Anaheim, California, and published in Fed. Proc. (44):747; 1985.     

The collagen recognition sequence for fibroblasts depends on collagen topography

We found that the peptide Gly-Arg-Gly-Glu-Ser-Pro (GRGESP) inhibited spreading of human fibroblasts inside collagen gels and markedly decreased gel contraction, but this peptide had no effect on cell spreading on collagen-coated surfaces. On the other hand, the peptide Gly-Arg-Gly-Asp-Ser-Pro (GRGDSP), which inhibited cell spreading on collagen-coated surfaces, did not inhibit cell spreading within collagen gels and was a less effective inhibitor of collagen gel contraction than GRGESP. Based on these findings, we conclude that human fibroblasts can interact with different collagen cell recognition sequences depending upon topographical organization of the collagen.     

Temporal relationships of F-actin bundle formation, collagen and fibronectin matrix assembly, and fibronectin receptor expression to wound contraction

Wound contraction can substantially reduce the amount of new tissue needed to reestablish organ integrity after tissue loss. Fibroblasts, rich in F-actin bundles, generate the force of wound contraction. Fibronectin-containing microfibrils link fibroblasts to each other and to collagen bundles and thereby provide transduction cables across the wound for contraction. The temporal relationships of F-actin bundle formation, collagen and fibronectin matrix assembly, and fibronectin receptor expression to wound contraction have not been determined. To establish these relationships, we used a cutaneous gaping wound model in outbred Yorkshire pigs. Granulation tissue filled approximately 80% of the wound space by day 5 after injury while wound contraction was first apparent at day 10. Neither actin bundles nor fibronectin receptors were observed in 5-d wound fibroblasts. Although fibronectin fibrils were assembled on the surfaces of 5-d fibroblasts, few fibrils coursed between cells. Day-7 fibroblasts stained strongly for nonmuscle-type F-actin bundles consistent with a contractile fibroblast phenotype. These cells expressed fibronectin receptors, were embedded in a fibronectin matrix that appeared to connect fibroblasts to the matrix and to each other, and were coaligned across the wound. Transmission EM confirmed the presence of microfilament bundles, cell-cell and cell-matrix linkages at day 7. Fibroblast coalignment, matrix interconnections, and actin bundles became more pronounced at days 10 and 14 coinciding with tissue contraction. These findings demonstrate that granulation tissue formation, F-actin bundle and fibronectin receptor expression in wound fibroblasts, and fibroblast-matrix linkage precede wound contraction.     

Inhibition of the contraction of collagen gels by extracts from human dermis.

Extracts from the human dermis were prepared and evaluated with respect to their ability to influence fibroblasts to contract collagen gels in vitro. The extract which had the most inhibitory effect on fibroblasts to cause contraction of collagen gels was extract D. It also inhibited fibroblast growth. Inhibition of contraction was not simply related to fibroblast cell numbers and the data suggests a specific effect upon the ability of fibroblasts to cause contraction. Other extracts were without significant effect.     

Contraction of fibroblast-containing collagen gels: Initial collagen concentration regulates the degree of contraction and cell survival

Remodeling of extracellular matrix involves a number of steps including the recruitment, accumulation, and eventual apoptosis of parenchymal cells as well as the production, organization, and rearrangement of extracellular matrix produced by these cells. The culture of fibroblasts in three-dimensional gels made of type I collagen has been used as a model of tissue contraction which characterizes both wound repair and fibrosis. The current study was designed to determine the effect of initial collagen concentration on the ability of fibroblasts to contract collagen gels and on cell survival. Native type I collagen was extracted from rat tail tendons and used to prepare collagen gels with varying collagen concentrations (0.75-2.0 mg/ml). Human lung fibroblasts (HFL-1) were cast into the gels and cultured in Dulbecco modified Eagle medium with 0.1% fetal calf serum for 2 wk. The gel size, collagen content, and deoxyribonucleic acid (DNA) content were determined. Gels prepared with an initial concentration of 0.75 mg/ml contracted more rapidly and to a smaller final size than gels prepared from 2 mg/ml initial collagen concentration (final size 7.1 versus 36.4% of initial size, P < 0.01). There was no significant degradation of the collagen in the gels under either condition. Hence, the dramatically increased contraction of the lower density gels resulted in a higher final density (P < 0.01). Cell density was estimated from DNA content. In low initial density gels, the final DNA content was significantly less than that in higher initial density gels (0.73 versus 1.88 microg/gel, P < 0.05). This was accompanied by an increased percentage of apoptotic cells at day 14 (43.3 versus 34.1%, P < 0.05). If the gels were maintained in the attached state which largely prevents contraction, apoptosis was significantly reduced, suggesting that contraction rather than matrix composition was a requirement for the increased apoptosis. In summary, these findings indicate that the initial matrix composition can lead to differing outcomes during fibroblast-mediated wound contraction.     

Smad3 mediates the TGF-beta-induced contraction of type I collagen gels by mouse embryo fibroblasts

TGF-beta signals through TGF-beta receptors and Smad proteins. TGF-beta also augments fibroblast-mediated collagen gel contraction, an in vitro model of connective tissue remodeling. To investigate the importance of Smad2 or Smad3 in this augmentation process, embryo-derived fibroblasts from mice lacking expression of Smad2 or Smad3 genes were cast into native type I collagen gels. Fibroblast-populated gels were then released into 0.2% FCS-DMEM alone or with recombinant human TGF-beta1, beta2, beta3, or recombinant rat PDGF-BB. Gel contraction was determined using an image analyzer. All three isoforms of TGF-beta significantly augmented contraction of collagen gels mediated by fibroblasts with genotypes of Smad2 knockout (S2KO), Smad2 wildtype (S2WT), and Smad3 wildtype (S3WT), but not Smad3 knockout (S3KO) mice. PDGF-BB augmented collagen gel contraction by all fibroblast types. These results suggest that expression of Smad3 but not Smad2 may be critical in TGF-beta augmentation of fibroblast-mediated collagen gel contraction. Thus, the Smad3 gene could be a target for blocking contraction of fibrotic tissue induced by TGF-beta.     

Persistence of TGF-β1 induction of increased fibroblast contractility

Fibroblast contraction of collagen gels is regarded as a model of wound contraction. Transforming growth factor (TGF)-beta added to such gels can augment contraction consistent with its suggested role as a mediator of fibrotic repair. Since fibroblasts isolated from fibrotic tissues have been suggested to express a "fibrotic phenotype," we hypothesized that TGF-beta exposure may lead to a persistent increase in fibroblasts' contractility. To evaluate this question, confluent human fetal lung fibroblasts were treated with serum-free Dulbecco modified Eagle medium (DMEM), with or without 100 pM [corrected] TGF-beta1, TGF-beta2, or TGF-beta3 for 48 h. Fibroblasts were then trypsinized and cast into gels composed of native type I collagen isolated from rat tail tendons. After 20 min for gelation, the gels were released and maintained in serum-free DMEM. TGF-beta-pretreated fibroblasts caused significantly more rapid gel contraction (52.5+/-0.6, 50.9+/-0.2, and 50.3+/-0.5% by TGF-beta1, -beta2, and -beta3 pretreated fibroblasts, respectively) than control fibroblasts (74.0+/-0.3%, P < 0.01). This effect is concentration dependent (50-200 nM), and all three isoforms had equal activity. The effect of TGF-beta1, however, persisted for only a short period of time following the removal of TGF-beta, and was lost with sequential passage. These observations suggest that the persistent increase in collagen-gel contractility, mediated by fibroblasts from fibrotic tissues, would not appear to be solely due to previous exposure of these cells to TGF-beta.     

Cyclic AMP-induced inhibition of collagen lattice contraction by fibroblasts may be attenuated by both cyclic AMP dependent and independent mechanisms

The contraction of collagen lattices made with forskin fibroblasts in medium containing 1% fetal bovine serum was inhibited by intracelluar cyclic AMP raising drugs including cholera toxin (CT), forskolin, and dibutyryl-cAMP. The inhibition by CT was attenuated by insulin, acidic fibroblast growth factor (aFGF), and transforming growth factor-β (TGF-β). All three peptide factors have previously been reported to promote collagen lattice contraction by arterial smooth muscle cells and/or fibroblasts. Incubation of cells suspended in collagen gels with CT and forskolin resulted in a transient rise of the intracellular cyclic AMP levels, which peaked at 2 hr and 30 min, respectively, after drug exposure. Cholera toxin-induced intracellular cyclic AMP increase was attenuated by TGF-β, but not by aFGF and insulin, when added simultaneously. Thus, TGF-β may attenuate CT's inhibition on collagen lattice contraction by attenuating CTinduced intracellualr cyclic AMP increse, whereas the attenuation by insulin and aFGF on the inhibition of lattice contraction may be mediated by a cyclic AMPindependent mechannism. © 1993 Wiley-Liss, Inc.     

Deposition and ultrastructural organization of collagen and proteoglycans in the extracellular matrix of gel-cultured fibroblasts

Human skin fibroblasts were cultivated within the three-dimensional space of polymerized alginate and collagen, respectively. The in vitro synthesis of collagens and proteoglycans was measured during the first 3 days of culture, and the deposition as well as the ultrastructural organization of newly synthesized extracellular matrix components were examined by electron microscopy. The amount of collagens and proteoglycans synthesized by fibroblasts, embedded in calcium alginate gels as well as in collagen lattices, was lowered as compared to monolayer cultures. Furthermore, it was found that collagen synthesis was reduced to a greater extent in alginate gels than in collagen lattices. On the contrary, total proteoglycan biosynthesis was similarly reduced either in alginate gels or in collagen lattices. At the end of a 3-day-culture period, filamentous material as well as cross-striated banded structures were found extracellularly in the alginate gel. According to their periodicity, their banding pattern, their association with polyanionic matrix components and their sensitivity towards glycosaminoglycan-degrading enzymes we could distinguish (1) sheets of amorphous non-banded material consisting of irregularly arranged filaments and containing dermatan sulfate-rich proteoglycans (type I structures), (2) sheets of long-spacing fibrils consisting of parallel orientated filaments and containing chondroitin sulfate-rich proteoglycans (= zebra bodies; type II structures), and (3) fibrillar structures with a complex banding pattern different from that of native collagen fibrils (type III structures). In fibroblasts cultured in collagen lattices, we only sporadically found depositions which are identified as type I structures. Using indirect immunoelectron microscopy and monospecific polyclonal antibodies, we localized type VI collagen in type I structures and type II structures. Type III structures can be identified as type I collagen derived as becomes obvious by comparison with segment long spacing crystallites of type I collagen.