Collagen telopeptides (cross-linking sites) play a role in collagen gel lattice contraction

Solubilized interstitial collagens will form a fibrillar, gel-like lattice when brought to physiologic conditions. In the presence of human dermal fibroblasts the collagen lattice will contract. The rate of contraction can be determined by computer-assisted planemetry. The mechanisms involved in contraction are as yet unknown. Using this system it was found that the rate of contraction was markedly decreased when collagen lacking telopeptides was substituted for native collagen. Histidinohydroxylysinonorleucine (HHL) is a major stable trifunctional collagen cross-link in mature skin that involves a carboxyl terminal, telopeptide site 16c, the sixteenth amino acid residue from the carboxy terminal of the telopeptide region of alpha 1 (I) in type I collagen. Little, if any, HHL was present in native, purified, reconstituted, soluble collagen fibrils from 1% acetic acid-extracted 2-year-old bovine skin. In contrast, HHL cross-links were present (0.22 moles of cross-link per mole of collagen) in lattices of the same collagen contracted by fibroblasts. However, rat tail tendon does not contain HHL cross-links, and collagen lattices made of rat tail tendon collagen are capable of contraction. This suggests that telopeptide sites, and not mature HHL cross-links per se, are essential for fibroblasts to contract collagen lattices. Beta-aminopropionitrile fumarate (BAPN), a potent lathyrogen that perturbs collagen cross-linking by inhibition of lysyl oxidase, also inhibited the rate of lattice cell contraction in lattices composed of native collagen. However, the concentrations of BAPN that were necessary to inhibit the contraction of collagen lattices also inhibited fibroblast growth suggestive of cellular toxicity. In accordance with other studies, we found no inhibition of the rate of lattice contraction when fibronectin-depleted serum was used. Electron microscopy of contracted gels revealed typical collagen fibers with a characteristic axial periodicity. The data provide evidence that collagen telopeptide sites play a role in collagen gel lattice contraction.     

Analysis of the age-related composition of human skin collagen and collagens synthesized by fibroblast culture

Age-related differences in the composition and the post-translational modifications of human skin collagens were examined in the present study. The data were compared with results of collagen synthesis from in vivo-aged fibroblasts in culture. Skin extracts and newly synthesized collagen from fibroblast cultures derived from both old and young donor groups showed the same ratio of collagen III to collagen I. Furthermore, no difference was noted in the degree of prolyl and lysyl hydroxylation of collagen I and collagen III. Young and old fibroblasts synthesized a similar quantity of collagen in vitro. The data suggest that fibroblasts maintain a uniform level of collagen production, composition and modification independent of the age of the donor.     

Collagen degradation in aged/photodamaged skin in vivo and after exposure to matrix metalloproteinase-1 in vitro

Biochemical and ultrastructural approaches were used to assess collagen changes in photodamaged skin. Extensive collagen fragmentation, clumping of the fragmented collagen, and interaction of fibroblasts with the damaged matrix were observed. Similar, though less extensive, collagen damage was also observed in sun-protected skin-individuals aged 80 y or older (naturally aged skin). In comparison, sun-protected skin from young individuals (18-29 y of age) demonstrated little damage. A uniform distribution of collagen fibrils was seen. Interstitial fibroblasts were embedded in the collagen matrix and in close apposition with intact collagen fibrils. In additional studies, three-dimensional lattices of type I collagen were exposed in vitro to matrix metalloproteinase-1 (interstitial collagenase), and examined for biochemical and ultrastructural alterations. Under conditions in which enzyme treatment produced fragmentation in 30-40% of the collagen molecules, the lattices demonstrated collagen fragmentation and clumping of the damaged matrix. Recent studies have demonstrated a loss of procollagen production by fibroblasts in contact with collagen fragments in vitro. This study demonstrates similar changes in collagen structure in vivo in aged and photodamaged skin. We suggest that collagen fragmentation in vivo could underlie the loss of collagen synthesis in photodamaged skin and, to a lesser extent perhaps, in aged skin.     

The influence of age and sex on skin thickness, skin collagen and density

Forearm skin collagen, dermal thickness and collagen density were measured in a large number of normal subjects as a standard reference for future studies. Skin collagen decreased with age and was less in the females at all ages. There is a direct relationship between skin collagen and dermal thickness but variations in collagen density in disease limit the use of dermal thickness as a guide to changes in its collagen content.     

Lipoid proteinosis: an inherited disorder of collagen metabolism?

The dermal collagen of a patient with lipoid proteinosis was investigated by immunohistochemistry and biochemical analysis. The affected skin was found to contain significantly less collagen per unit dry weight than normal dermis but showed elevated levels of type 3 collagen with respect to type I. Purification of collagen types from affected skin after pepsin digestion showed no novel forms, but a doubling in the yield of type 5 collagen. These results correlated well with those of immunohistochemistry which showed a patchy, diffuse, widely distributed type 3 collagen and an increase in types 4 and 5 collagens associated with 'onion skin' endothelial basement membrane thickening. Estimation of collagen cross-links showed an abnormal pattern with a preponderance of the keto-imine form not normally associated with skin. These results strongly suggest that lipoid proteinosis involves a primary perturbation of collagen metabolism.     

Necrobiosis lipoidica: ultrastructural and biochemical demonstration of a collagen defect

Ten patients with necrobiosis lipoidica lesions were studied. Five patients had diabetes mellitus. The age of the patients varied from 15 to 73 years and the duration of the skin lesions was from 2 to 20 years. Histologically, the lesions were characterized by degeneration of collagen and elastin. In some lesions elastin fibers could be seen in areas devoid of normal-looking collagen. Electron microscopy revealed loss of cross-striation of collagen fibrils and a marked variation in the diameter of individual collagen fibrils. The concentration of collagen, measured by assay of hydroxy-proline, a collagen-specific amino acid, was markedly decreased in the lesional skin, but the ratio of type I/III collagen was unchanged in the affected skin. Fibroblasts established from affected skin synthesized less collagen than cells derived from healthy-looking skin. The decreased collagen synthesis was due to a decreased amount of messenger RNA for type I procollagen, measured by hybridization with a specific human cDNA clone. The production of collagenase by these fibroblasts was not increased. Our results thus indicate that in necrobiosis lipoidica lesions, collagen fibrils are defective and the amount of collagen is reduced, probably due to decreased synthesis of collagen by affected fibroblasts.     

Effect of UV on the susceptibility of acid-soluble Skh-1 hairless mouse collagen to collagenase

BACKGROUND/PURPOSE: Photoaging of the skin is a result of chronic exposure to environmental ultraviolet radiation (UV). The milieu provided by the extracellular matrix, which significantly influences the behaviour of resident fibroblasts, depends critically on the supermolecular collagen structure. We ask whether direct photochemical treatment of type I collagen with solar wavelengths capable of reaching the dermis can modify the substrate's susceptibility to collagenase in a model in vitro system. METHODS: Acid- extracted Skh-1 hairless mouse collagen samples were irradiated with 0-140 J/cm2 of radiation from bank of filtered FS lamp (UVB/UVA = 0.33, fluence rate = 0.81 mW/cm2). Subsequent to UV irradiation, collagen samples were coupled with fluorescein isothiocyanate (FITC) and assayed for susceptibility to bacterial collagenase by monitoring the appearance of supernatant FITC fluorescence (a measure of lysed collagen) over time of incubation. As a 'reference', unirradiated commercial FITC-labelled citrate-soluble collagen (Elastin Products, Owensville, MO 65066, USA) was similarly analysed. RESULTS: Unirradiated mouse collagen had a lower rate of cleavage than did the calfskin sample. Irradiation of unlabelled mouse collagen for 0-48 h (0-140 J/cm2 total UV) rendered the sample more soluble, with concomitant chain degradation, cross-linking and loss of intrinsic collagen fluorescence. At irradiation time's >/= 4 h (>/=11.7 J/cm2), the irradiated collagen was significantly more susceptible to bacterial collagenase digestion. DISCUSSION: It appears that the rate of cleavage depends on the superstructure of the collagen, since the kinetics of collagen cleavage differ for two collagen samples having essentially the same primary structure. Cleavage kinetics may depend on the 'maturity' (solubility) of the collagen. The observation that UV-damaged mouse collagen is a better substrate for collagenase than the intact sample may be illustrative of a mechanism whereby damaged collagen targets itself for selective attack by collagenase.     

Preparation of [3H]collagen for studies of the biologic fate of xenogenic collagen implants in vivo

Reduction of a commercially available, pepsin-solubilized, bovine dermal collagen (Vitrogen 100) with sodium [3H]borohydride provided radiolabeled collagen preparations with specific activities ranging from 7.1-12.0 muCi/mg collagen. These specific activities were 2-3 times greater than those obtained by reduction of intact rat tail tendon collagen under similar conditions. The alpha, beta, and higher aggregate components of type I collagen were radiolabeled as well as the alpha component of a small amount of type III collagen present in the samples. Fractionation of cyanogen bromide peptides showed that alpha 1(I)CB7, alpha 1(I)CB8, and alpha 2(I)CB3,5 were the predominant peptides labeled by this procedure. Amino acid analysis indicated that the majority of the radioactivity was in reducible cross-links, precursors of these cross-links, and in hexosyllysine residues. Reconstitution experiments comparing this radiolabeled collagen with nonlabeled collagen showed them to be indistinguishable. Bacterial collagenase digestion of this reconstituted fibrillar collagen in both a lightly cross-linked (glutaraldehyde 0.0075%) and noncross-linked form provided evidence that digestion of labeled and nonlabeled collagens proceeded at similar rates. Thus, labeling did not change the properties of the collagen. Cross-linking made the preparation refractory to proteolytic degradation. Injection of fibrillar collagen preparations, spiked with radiolabeled collagen, into the guinea pig dermis followed by quantitation of the amount of radioactivity recovered from implant sites as a function of time, indicated that the lightly cross-linked samples also were more resistant to degradation in vivo than the noncross-linked preparation. The half-life of noncross-linked collagen was about 4 days while that of the cross-linked collagen was about 25 days. These degradation rates were much faster than observed for similar, nonlabeled samples injected into the dermis of humans, presumably due to a higher metabolic activity in the guinea pig dermis.     

Factor XIII in scleroderma: in vitro studies

The administration of Factor XIII (FXIII) produces a beneficial effect on the skin lesions in about 50% of the treated patients with progressive systemic sclerosis (PSS). The effect of FXIII on various skin fibroblast functions (proliferation, attachment, biosynthetic activity and mechanical properties) was investigated in vitro using normal and PSS strains. In cell culture, most of the PSS fibroblast strains synthesized excessive amounts of collagen. Other cell functions such as adhesion to collagen I or III, to fibronectin, retraction of collagen lattices, proliferation in low serum concentration and degradation of newly synthesized collagen were not significantly different. The addition of FXIII (I U/ml) inhibited the synthesis of collagen by normal fibroblasts and reduced it in PSS fibroblasts to a level similar to that of normal fibroblasts. This effect was observed for cells cultured on plastic or in a collagen lattice. In the latter, an increased amount of collagen degradation was observed. No significant effect of FXIII on the other cell functions was noted. Excessive collagen production by PSS fibroblasts can be repressed by FXIII in vitro by at least two distinct mechanisms: a reduction of collagen synthesis and an increased degradation of the newly synthesized collagen.     

The carboxyl-terminal domain of type VII collagen is present at the basement membrane in recessive dystrophic epidermolysis bullosa

Recent studies showing that type VII collagen is a component of anchoring fibrils suggests that the absence of anchoring fibrils in recessive dystrophic epidermolysis bullosa may be due to a defect in the synthesis, secretion, and deposition of type VII collagen. That hypothesis is further supported by recent studies suggesting that monoclonal antibodies to type VII collagen do not react with the basement membrane in most patients. To investigate further, we examined skin from 12 patients by electron microscopy and by immunohistology and immunoelectron microscopy using a concentrated and purified monoclonal antibody to the carboxy-terminal domain of Type VII collagen. Although anchoring fibrils were not detected by electron microscopy, the results of immunohistology showed definite, but reduced, binding of the monoclonal antibody to type VII collagen at the basement membrane in a linear pattern in 11 of 12 patients. By immunoelectron microscopy, reduced deposition of anti-type VII collagen antibody was detected beneath the lamina densa. The results of this study show that the carboxyl-terminal domain of type VII collagen is synthesized, secreted, and deposited at the basement membrane zone in 11 of 12 patients with recessive dystrophic epidermolysis bullosa and suggest that the absence of anchoring fibrils may be due either to deposition of abnormal type VII collagen, reduced levels of normal type VII collagen, defective assembly of type VII collagen into anchoring fibrils, or destruction of the collagenous domain of type VII collagen.     

018 New source for evaluating PDT drugs in the laboratory

To test the hypothesis that UV could alter collagen susceptibility to interstitial collagenase, acid-extracted Skh-1 hairless mouse collagen samples were (un)irradiated with 0–140 J/cm² of radiation from bank of filtered FS lamp (UVB/UVA = 0.33, fluence rate = 0.81 mW/cm²). Subsequent to UV irradiation, collagen samples were coupled with fluorescein isothiocyanate (FITC) and assayed for its susceptibility to bacterial collagenase by monitoring the appearance of supernatant FITC fluorescence (a measure of lysed collagen) over time of incubation. As a reference. unirradiated commercial FITC – labeled collagen (Elastin Products) was similarly analyzed. Mouse collagen had a lower rate of cleavage than did the calf skin sample. Mouse collagen initial cleavage followed a quasi-linear time course over the first 5 h. Calf-skin collagen displayed a 'sigmoidal' time course, reminiscent of a cooperative mechanism. UV irradiation afforded no significant effect on the ability of collagenase to cleave mouse collagen, although a small effect could be discerned after 48 h (140 J/cm²). On the other hand, these samples exhibited significant chain degradation. cross-linking and loss of intrinsic collagen fluorescence on UV photolysis. It appears that the rate of cleavage depends on the superstructure of the collagen, and that the collagen fluorophores are not in proximity to the specific site of collagenase cleavage.
Supported in part by NIH/MBRS Grant #GM 08248 and RCMI Grant #RR 03034.     

Migration of a human keratinocyte cell line (HACAT) to interstitial collagen type I is mediated by the alpha 2 beta 1-integrin receptor.

The migratory response of the human keratinocyte cell line HaCaT to collagen type I and the molecular mechanism underlying collagen-mediated migration have been analyzed. The migratory response of HaCaT cells to collagen type I consisted of a dose-dependent migration to insoluble step gradients of substratum-bound collagen (haptotaxis) and to gradients of soluble collagen (chemotaxis). Checkerboard analysis demonstrated a minor chemokinetic component. Denatured collagen type I was less chemoattractive than the native triple-helical form. Pre-treatment of cells with 25-250 micrograms/ml of synthetic peptides containing the fibronectin cell-recognition sequence RGD (Arg-Gly-Asp) resulted in a concentration-dependent inhibition of fibronectin-mediated chemotaxis, whereas chemotaxis to collagen was not affected. We then investigated the role of VLA/collagen-receptors for collagen type I-induced chemotaxis. Monoclonal antibody (MoAb) 5E8, which selectively blocks function of the alpha 2 subunit of the VLA-2/collagen receptor, dose-dependently inhibited the chemotactic response of HaCaT cells to collagen. This effect was specific for collagen-mediated chemotaxis because the chemotactic response to fibronectin remained unaffected. In contrast, a function blocking MoAb directed to the alpha 3 subunit of the coexpressed VLA-3 receptor, which is also capable of binding collagen, had no effect. However, function blocking MoAb directed to the beta 1-chain of integrins completely inhibited chemotaxis to collagen type I. Based on our results, we propose that the chemotactic migration of the human keratinocyte cell line (HaCaT) to collagen type I is specifically mediated by the RGD independent VLA-2/collagen receptor (alpha 2 beta 1) of the integrin family.     

The molecular basis of glucocorticoid-induced skin atrophy: topical glucocorticoid apparently decreases both collagen synthesis and the corresponding collagen mRNA level in human skin in vivo

The effects of topical betamethasone-17-valerate on collagen propeptide levels, collagen mRNA level, lysyl oxidase mRNA and matrix metalloproteinase (MMP)-1 and MMP-2 mRNA levels were studied in human skin. Three days' treatment of healthy skin with topical betamethasone caused a 70–80% decrease in type I and III collagen propeptides in suction blister fluid. A similar decrease was found in type I collagen mRNA when assayed by either slot-blot hybridization or a quantitative polymerase chain reaction method, indicating that the decrease in collagen synthesis after topical glucocorticoid treatment is apparently due to a decrease in corresponding mRNA. mRNA of lysyl oxidase, which is an important enzyme catalysing the cross-linking of collagen chains, and collagen-degrading enzyme MMP-1 and MMP-2 mRNAs were not decreased in the same skin samples. This suggests that in vivo , glucocorticoids modulate variably the genes involved in collagen synthesis and degradation. Our study provides a solid molecular basis for glucocorticoid-induced dermal atrophy, which results from the decrease in functional collagen mRNA in the skin.     

Immune response to gamma-irradiated injectable human amnion and human skin collagens in the rat

The immune response in rats to gamma-irradiated human amnion and human skin collagen was characterized through histologic and immunologic methods. Pepsin-extracted human amnion collagen and skin collagen were purified and reconstituted. Implants of amnion collagen demonstrated greater persistence than skin collagen. For amnion collagen implants, no significant inflammatory response was found. Fibroblast and adipocyte ingrowth and neovascularization were present. Conversely, obvious inflammatory infiltration was evident in the skin collagen implants. Enzyme-linked immunosorbent assay results showed that anti-amnion collagen antibody levels were significantly lower than anti-skin collagen antibody levels against their respective implant materials. The ratios of type I to type III collagen are 56:44 and 95:5 for amnion collagen and skin collagen, respectively. These findings suggest that in this heterologous type system, type III collagen-rich amnion collagen preparations appear superior to skin collagen for soft-tissue augmentation.     

Effects of a collagen matrix containing prostaglandin E(1) on wound contraction

In this study, we evaluated the effectiveness of prostaglandin (PG) E(1) in inhibiting wound contraction, both alone and in combination with collagen matrix, using a in vivo full thickness skin defect model. To clarify the mechanisms involved in this inhibition we also used a fibroblast-populated collagen gel contraction in vitro model. We demonstrated that collagen matrix alone significantly inhibited wound contraction PG E(1) alone did not. Interestingly, their combination was much more effective than either collagen matrix or PG E(1) alone, a finding which was also supported by histopathological examination. Wounds treated with collagen matrix, but not control wounds, showed horizontal rearrangement of collagen fibers in the dermal part as well as evidence of active fibroblast proliferation which was not observed in scar regions surrounded by normal dermis. With the fibroblast-populated collagen gel contraction in vitro model, we found that PG E(1) significantly inhibited contraction at a high dose. It was concluded that collagen matrix combined with PG E(1) is effective for preventing contracture producing so called neodermis than collagen matrix alone, which remains one of the most challenging problems in treating full thickness type wounds.     

Inhibitory effects of bFGF,VEGF and minoxidil on collagen synthesis by cultured hair dermal papilla cells

Dermal papilla cells of rat vibrissa follicles cultivated in monolayers and in three-dimensional collagen gels show a different morphology in these culture systems. Dermal papilla cells cultured in lattices tend to express morphological features resembling those seen in vivo. Quantification of total collagen by incorporation of³H-proline in monolayer cultures and in collagen lattices show that the amount of collagen found in dermal papilla cells is higher than that secreted. Moreover, collagen synthesis measured in lattices is reduced to about 50% of that found in monolayer cultures. The influence of growth factors on collagen synthesis by hair dermal papilla cells was investigated. We studied the effects of basic fibroblast growth factor (bFGF), vascular endothelial growth factor (VEGF) and minoxidil on collagen synthesis in monolayers and in lattices. VEGF, bFGF and minoxidil significantly decreased the total amount of collagen. In monolayer cultures, there was approximately a 30% inhibition of collagen production with 5ng/ml bFGF, 0.1 ng/ml VEGF and 100 ng/ml minoxidil. However, in the lattices this inhibition was reduced to about half. These results suggest that both culture substrate and growth factors influence collagen production by rat hair dermal papilla cells.     

Coenzyme A stimulates collagen production in cultured fibroblasts; possible mechanisms in enzymatic and gene expression

Coenzyme A, a cofactor in enzymatic acetyl transfer reactions, stimulates collagen production in cultured fibroblasts. The mechanisms involved in this collagen stimulation were investigated. Enzymatic studies using radiolabeled procollagen as substrate revealed that coenzyme A enhanced prolyl hydroxylase activity. Prolyl hydroxylase is a key enzyme in collagen synthesis acting by hydroxylation of proline residues in procollagen peptide, which is necessary for stabilizing collagen. Northern blot analysis demonstrated that coenzyme A also enhanced mRNA levels of both the alpha subunit of prolyl hydroxylase and the alpha1 chain of type I collagen. The levels of protein production of prolyl hydroxylase and type I collagen were also increased in cultured fibroblasts by coenzyme A, which correlated well with observations from Northern blotting. On the other hand, coenzyme A did not stimulate the activity nor the gene expression of two other processing enzymes: lysyl hydroxylase, which provides the sites for glycosylation and crosslinking between collagen peptides, and lysyl oxidase, a fundamental enzyme in intermolecular crosslinking. These results indicate that coenzyme A stimulates collagen production by at least two separate mechanisms: by enhancing prolyl hydroxylase activity as well as stimulating gene expression of the alpha subunit of this enzyme, and by stimulating gene expression of alpha1 chain of type I collagen.