Purification of an endogenous activator of procollagenase from rabbit synovial fibroblast culture medium

A personal journey with matrix metalloproteinases

I was given the honor of delivering the 2015 Lifetime Membership Award lecture at the International Proteolysis Society's annual meeting held in Penang, Malaysia in October 2015. It gave me an opportunity to look back on how I started my research on matrix metalloproteinases (MMPs) and how I continued to work on these proteinases for the next 42 years. This is a series of sketches from the personal journey that I took with MMPs, starting from the purification of metalloproteinases, cloning, structural studies, then to a more recent encounter, endocytic regulation of matrix-degrading metalloproteinases.

Analysis of gene expression profiles in fatal hepatic failure after hepatectomy in mice

BACKGROUND

We developed 90%-hepatectomized mice that were the fatal model, and analyzed the gene expression profiles using a complementary DNA (cDNA) microarray to clarify the mechanisms of hepatic failure after excessive hepatectomy.

MATERIALS AND METHODS

Ribonucleic acid (RNA)s from the remnant hepatic tissue of 70%- and 90%-hepatectomized mice were labeled with fluorescent dyes, and hybridized to the Riken set of 39,168 full-length enriched mouse cDNA arrays. The gene expression profiles in 90%- and 70%-hepatectomized mice were analyzed by scanning date for fluorescent dye signals.

RESULTS

The down-regulated genes in 90%-hepatectomized mice were genes activating extracellular matrix (ECM) remodeling (matrix metalloproteinases, laminins, and integrins), genes related to cytokines (tumor necrosis factor α converting enzyme, and Janus kinase 3) that were related to the priming, genes related to growth factor (heparin-binding epidermal growth factor-like growth factor and others), and genes promoting cell cycle progression (cyclin D1, D2, and E2) that were related to the progression of hepatocytes. The up-regulated genes were genes inhibiting ECM remodeling [plasminogen activator inhibitors (PAIs)].

CONCLUSIONS

Hepatic failure after hepatectomy was characterized by the inhibition of hepatic cell cycle priming and progression both induced by ECM remodeling in liver regeneration. Particularly, the overexpression of PAIs was thought to play the major role in the first step of inhibition of ECM remodeling.

Hepatocyte growth factor gene transfer with naked plasmid DNA ameliorates dimethylnitrosamine-induced liver fibrosis in rats

AIM

Hepatocyte growth factor (HGF) has various biological properties, including antifibrogenic activity. In the present study, we tested the efficacy of HGF gene therapy using naked plasmid DNA in dimethylnitrosamine (DMN)-induced liver fibrosis in a rat model.

METHODS

Naked plasmid DNA encoding human HGF was injected once, together with a hypertonic solution, into the hepatic artery after DMN treatment on three consecutive days per week for 3 weeks. Naked plasmid DNA encoding beta-galactosidase was injected similarly in the DMN-treated control rats. DMN treatment was continued once weekly after gene transfer for additional 3 weeks.

RESULTS

The human HGF protein expression was detected in livers transfected with human HGF naked plasmid DNA, gradually decreasing by day 21. The expression of the endogenous rat HGF protein was also upregulated after human HGF gene transfer. Phosphorylation of c-Met, a HGF receptor, was detected only in livers transfected with human HGF plasmid DNA. Fibrosis was attenuated significantly in livers transfected with the human HGF plasmid. Attenuation wasaccompanied by decreased expression of alpha-smooth muscle actin. Increased portal vein pressure after treatment with DMN was suppressed significantly by HGF gene transfer. The upregulated hepatic protein expression of transforming growth factor-beta (TGF-beta) in response to DMN was markedly attenuated by HGF gene transfer accompanied by the increased protein expression for matrix metalloproteinases (MMP)-3 and -13.

CONCLUSION

The hepatic arterial injection of human naked plasmid HGF DNA was effective in suppressing liver fibrosis induced in rats by DMN. The mechanisms by which HGF expression attenuated liver fibrosis may include the suppression of hepatic TGF-beta expression and the induction of MMP expression.

The turnover and degradation of collagen

The interstitial collagens are degraded predominantly extracellularly, by specific collagenases (metalloproteinases) capable of cleaving the helical region across the three chains at a similar locus, solubilizing the cleaved products from the fibril. Other neutral proteinases may also function in this role by cleaving near cross-links in the fibril. Collagen type, molecular aggregation and small changes in temperature all markedly affect rates of collagenolysis in the fibril. Regulation of collagenolysis is also modulated at the levels of (1) cellular production of latent collagenase (procollagenase), (2) activation of latent collagenase, and (3) production of collagenase inhibitors. Fibroblastic cells and certain macrophages are probably the predominant sources of collagenases in inflammation; an enzyme in polymorphonuclear leucocytes (neutrophils) is distinct from the tissue enzyme. Molecules such as mononuclear cell factor (MCF), homologous with interleukin 1, which augment cellular collagenase production in inflammation, are derived from monocytes. The mechanisms of augmented collagenase production involve new protein synthesis and, if this augmentation is analogous to that produced by urate crystals, it is probably associated with increased levels of procollagenase mRNA. MCF production is itself controlled by products of lymphocytes as well as by interactions of monocytes with the Fc portion of immunoglobulins and components of the extracellular matrix. Activation of latent (pro)collagenase probably occurs in vivo through the action of neutral proteinases such as plasmin (through plasminogen activator). These effects may be indirect and exerted through proteolytic activation of a procollagenase activator. Tissue inhibitors act to regulate the active collagenase.

Oncogene induction of metastases

A metastatic colony is the end result of a complex series of steps involving multiple gene products. In some cases, the augmented metastatic potential of certain tumour cells may be due to the increased expression of specific gene products which confer a selective advantage. Transfection of the c-Ha-ras oncogene into suitable recipient cells constitutes a powerful experimental model with which to identify putative gene products augmented in highly metastatic tumour cells compared to their non-metastatic counterparts. Transfection of the activated ras oncogene into 3T3 and 10T1/2 embryo fibroblasts, and adult rat fibroblasts, results in transformants which produce high numbers of spontaneous metastases in nude mice or syngeneic recipients. The ras oncogene will also increase the metastatic aggressiveness of murine tumours with low metastatic potential. However, the ras oncogene will not induce the metastatic phenotype in all recipient cells. Furthermore, specific genes such as adenovirus 2 E1A suppress the ability of ras to induce the metastatic phenotype. Natural 'suppressor' gene products may exist which render certain cells resistant to the induction of metastases by ras. Ras oncogene transfection induces the production of type IV collagenase, motility factors and growth factors. The ras oncogene therefore induces a cascade of gene functions leading to rapid progression to the metastatic phenotype. The mechanism of the induction probably involves complex interactions between the ras p21 product and multiple cellular gene products.

Isolation, characterization, and expression of feline stromelysin-1 in naturally developing tumors in cats

OBJECTIVE

To detect, isolate, and characterize feline stromelysin-1 (ie, matrix metalloproteinase [MMP]-3) in naturally developing tumors in cats.

SAMPLE POPULATION

31 tissue samples obtained from primary tumors and 6 samples of normal tissues from cats.

PROCEDURE

Biopsy specimens were obtained from primary tumors. Primers were designed on the basis of known sequences. The sequence of stromelysin-1 was cloned and analyzed. An additional primer set was used as a screening tool. Samples were assayed in duplicate or triplicate, when possible. Data obtained were analyzed for differences in expression of stromelysin-1 with regard to overall survival among cats of various sex, age, and disease status.

RESULTS

A 1,181-bp cDNA nucleotide sequence was amplified. The open reading frame encoded 393 amino acids. This amino acid sequence shared 70% to 85% sequence homology with sequences of other species. In addition, samples were screened for stromelysin-1. Of the 31 tumor samples tested, 16 (51.6%) had positive results for expression of stromelysin-1. Total RNA expression was detected in a diverse group of tumor types. Prognostic factors associated with a shorter duration of survival included evidence of metastasis and metastasis associated with expression of stromelysin-1.

CONCLUSIONS AND CLINICAL RELEVANCE

Feline stromelysin-1 contains all the conserved regions typically found in members of the MMP family. Activity of stromelysin-1 has been implicated in a wide number of physiologic and pathologic processes. Identification of this gene may lead to the development of useful reagents to assist with diagnosis and management of neoplastic diseases in cats.

Pharmacokinetics, safety, and tolerability of BAY 12-9566 and nonsteroidal anti-inflammatory agents (naproxen, ibuprofen) during coadministration in patients with osteoarthritis

The pharmacokinetic interactions between BAY 12-9566 and two nonsteroidal anti-inflammatory drugs (NSAIDs), naproxen and ibuprofen, were investigated in osteoarthritis (OA) patients. The study comprised six groups: two NSAID groups with three levels of treatment (BAY 12-9566 400 mg, BAY 12-9566 100 mg, and placebo). Plasma pharmacokinetic parameters (AUC(0-tau), Cmax, and tmax) were determined for each treatment group following 5 days of NSAID administration, 14 days of BAY 12-9566 administration, and 14 days of concurrent NSAID and BAY 12-9566 administration. For most conditions, the total plasma drug concentrations of both NSAID and BAY 12-9566 were diminished by coadministration; total plasma BAY 12-9566 was not affected by ibuprofen treatment. Importantly, the free drug concentrations were largely unaffected by coadministration. Most side effects were mild or moderate in intensity, and all events, with the exception of headache, were reported in both NSAID groups and in both placebo and BAY 12-9566 groups.

Furin-independent pathway of membrane type 1-matrix metalloproteinase activation in rabbit dermal fibroblasts

We investigated the gene expression and intracellular activity of processing protease furin and its involvement in the process of membrane type 1-matrix metalloproteinase (MT1-MMP) activation in rabbit dermal fibroblasts. When the rabbit fibroblasts were treated with concanavalin A (ConA), pro-MMP-2 was converted to an active 62-kDa MMP-2 through the appearance of a 64-kDa intermediate MMP-2. The ConA-induced pro-MMP-2 activation resulted from increasing the gene expression and production of MT1-MMP in the rabbit fibroblasts. Reverse transcriptase-polymerase chain reaction demonstrated that in rabbit dermal fibroblasts furin mRNA was detected and, unlike MT1-MMP, was not increased by ConA. These findings are further supported by the fact that the intracellular furin activity also was constitutively detected and was unchanged by the ConA treatment. Very similar phenomena were also observed in human uterine cervical fibroblasts, which are known to produce MT1-MMP by ConA stimulation. These results suggest that the expression of the furin gene and the intracellular activity are not regulated by ConA. On the other hand, neither a synthetic furin inhibitor, decanoyl-RVKR-CH(2)Cl (25-100 microM) nor a furin antisense oligonucleotide (40 microM) inhibited the MT1-MMP-mediated pro-MMP-2 activation in ConA-treated rabbit dermal fibroblasts, whereas these compounds interfered with pro-MMP-2 activation in ConA-treated human uterine cervical fibroblasts. Nonetheless, the furin antisense oligonucleotide completely suppressed furin gene expression in both rabbit and human fibroblasts. These results suggest that furin does not participate in the process of MT1-MMP activation induced by ConA in rabbit dermal fibroblasts.

The modulation of corneal keratocyte and epithelial cell responses to poly(2-hydroxyethyl methacrylate) hydrogel surfaces: phosphorylation decreases collagenase production in vitro

We examined the regulation of collagenase production by rabbit keratocyte, epithelial and mixed keratocyte/epithelial cell cultures which were exposed to poly(2-hydroxyethyl methacrylate) (PHEMA) hydrogel surfaces with different chemistries and morphologies (sponge and homogeneous gels). Tissue culture modified polystyrene (TCP), used as a control surface, induced the maximum collagenase response with all cell culture types. Copolymer homogeneous gels containing 2-ethoxyethyl methacrylate (EEMA) or methyl methacrylate (MMA) induced a high response in keratocyte cultures, whilst PHEMA hydrogels induced a moderate response and the phosphorylated PHEMA (phos-PHEMA) hydrogel induced no response. Epithelial cells cultured on PHEMA, copolymer and phos-PHEMA hydrogels produced less collagenase activity than the keratocyte cells. The profile of collagenases produced by epithelial cells in response to phos-PHEMA was different to that for the other hydrogels. Co-cultured cells produced higher levels of collagenase (relative to the TCP) in response to hydrogels than did either the keratocytes or epithelial cells alone, but the response of phos-PHEMA was still the lowest. The overall enzyme response to the sponge hydrogels was lower than that to the homogeneous hydrogels, although this effect was less prominent in the keratocyte cultures. The markedly reduced and alternative collagenase responses to phosphorylated surfaces was not a consequence of cell death, and may be a phenomenon related to changes in cell surface charge and morphology.

Differential regulation of extracellular matrix synthesis during liver regeneration after partial hepatectomy in rats

Little is known about the modulation of the extracellular matrix (ECM) during liver regeneration. We studied the temporospatial expression of procollagens and of matrix metalloproteinases (MMPs) and their physiological antagonists, the tissue inhibitors of metalloproteinases (TIMPs) after two-thirds partial hepatectomy (PH) by Northern blot analysis and in situ hybridization. The entry of hepatocytes into the S-phase at 24 hours after PH was accompanied by a peak (sixfold induction) of hepatic TIMP-1 RNA levels that steadily declined thereafter to reach normal levels 144 hours after PH. Moderate MMP-2 and TIMP-2 RNA levels remained constant up to 144 hours after PH, and MMP-1 and -13 RNA were always undetectable. In situ hybridization showed a dramatic upregulation of TIMP-1 RNA transcripts in mesenchymal cells of portal, perisinusoidal and, to a lesser extent, pericentral areas. In contrast, scattered hepatocytes represented only a minor fraction (below 10%) of TIMP-1 RNA positive cells. When hepatocytes stopped DNA synthesis at 72 hours after PH, an upregulation of procollagen alpha1(I) and alpha2(III) transcripts was observed paralleled by threefold increased PIIINP levels in the sera. Our data reveal a tightly regulated program of de novo matrix synthesis after PH. Whereas interstitial procollagens appear to participate in the induction and maintenance of the quiescent hepatocyte phenotype, the early and localized expression of TIMP-1 indicates a role unrelated to its function as a general MMP-antagonist, e.g., as a growth promoting agent for hepatocytes.

Cyclooxygenase inhibitors augment the production of pro-matrix metalloproteinase 9 (progelatinase B) in rabbit articular chondrocytes

Matrix metalloproteinase 9 (MMP-9/gelatinase B) has recently been proposed to participate in the destruction of articular cartilage. Here, we report that interleukin 1 (IL-1) enhances the production of the precursor of MMP-9 in rabbit articular chondrocytes in primary culture, and this IL-1-mediated production of proMMP-9 is greatly augmented by cyclooxygenase inhibitors such as diclofenac and indomethacin, whereas the constitutive production of proMMP-2 (progelatinase A) is not modulated by IL-1 and/or cyclooxygenase inhibitors. Exogenous prostaglandin (PG) E1 and PGE2 suppress the proMMP-9 production in a dose-dependent manner. Similar results are also obtained with cultured rabbit synoviocytes. These results provide the first evidence that PGE down-regulates the production of proMMP-9 in chondrocytes and synoviocytes. Thus, cyclooxygenase inhibitors probably exert undesirable catabolic actions on the maintenance of articular cartilage under inflammatory conditions.

Transforming growth factor beta (TGF-beta) and dexamethasone have direct opposing effects on collagen metabolism in low passage human dermal fibroblasts in vitro

Collagen metabolism is a balance between synthesis and lysis. Both are under tight regulatory control. TGF-beta reverse the impairment of healing seen after glucocorticoid treatment in vivo. Both TGF-beta and glucocorticoids are known to regulate collagen metabolism directly. We have examined the effect of dexamethasone and of TGF-beta individually and in combination on the regulation of procollagen type 1, interstitial collagenase and tissue inhibitor of metallo-proteinase-1 (TIMP-1) synthesis at both the protein and mRNA levels in low passage human dermal fibroblasts. Dexamethasone treatment decrease synthesis of procollagen and caused a dose dependent down-regulation of TIMP-1 synthesis. Interstitial collagenase synthesis by fibroblasts was detectable but low. Thus, glucocorticoid treatment of fibroblasts tilts the balance of collagen metabolism away from accumulation. TGF-beta had opposing effects, stimulating both procollagen and TIMP-1 synthesis at the protein and mRNA levels. TGF-beta was able to cause a dose-dependent reversal of the glucocorticoid induced decrease in procollagen and TIMP-1 synthesis. Stimulation of healing in glucocorticoid treated animals by TGF-beta may be by the direct action of this agent upon fibroblast collagen metabolism.

Myocardial matrix metalloproteinase(s): localization and activation

Matrix metalloproteinases (MMPs) and neutrophil elastase (NE) may each contribute to fibrillar collagen degradation in various disease states. Little, however, is known about the activation and localization of MMP in the heart. Accordingly, we extracted MMP and examined mechanisms of proMMP activation in whole tissue extracts of the adult rat myocardium. Incubation of extracts with serine proteases (i.e., trypsin or neutrophil elastase) at 37 degrees C resulted in a time-dependent activation of proMMPs. Based on immunoblot and measurements of MMP activity by zymography, the molecular weight of active MMP was deduced to be 52 kDa. The second-order rate constant for activation of proMMP by serine protease was 5.5 +/- 0.2 x 10(5) M-1min-1 and for oxidized glutathione (GSSG) 1.5 +/- 0.1 M-1min-1. Incubation of the extract with both serine protease and GSSG increased the rate of activation 30-fold. Based on reverse zymographic analysis of collagenase inhibition, tissue inhibitors of metalloproteinases were identified. Indirect immunofluorescence localized proMMPs/MMPs to the endothelium and subendothelial space of the endocardium and throughout the interstitial space found between groups of muscle fibers. These results suggest that the mechanism of activation of MMPs by either a serine protease and by oxidizing, thiol-modifying reagents are mechanistically different and the presence of either a serine protease or GSSG synergistically increase the rate of activation of proMMPs. Our results also suggest that MMPs may be regulated by its own endogenous inhibitors. The contribution of this proteolytic enzyme to tissue remodeling and wound healing responses that occur in various diseases states remains to be established.

The acute phase reactant serum amyloid A (SAA3) is a novel substrate for degradation by the metalloproteinases collagenase and stromelysin

We found that the matrix metalloproteinases collagenase (MMP-1) and stromelysin (MMP-3) each has the ability to degrade a novel substrate, serum amyloid A (SAA3). SAA3 is a product of rabbit synovial fibroblasts stimulated with phorbol esters or interleukin-1, and it acts in an autocrine or paracrine manner to induce collagenase in both rabbit and human fibroblasts. Recombinant rabbit fibroblast procollagenase and human fibroblast prostromelysin were produced by baby hamster kidney (BHK) cells stably transfected with these genes, and latent enzyme was activated with aminophenylmercuric acetate (APMA). The Km for both enzymes was approximately 10 microM, and the Vmax for collagenase was approximately 6 pmol/minute/100 ng enzyme, while that for stromelysin was about 3-fold faster. Treatment of SAA3 with either enzyme generated a fragment of approx. 6 kDa that has the same amino terminus as the parent molecule, but this fragment was rapidly degraded. We have been unable to isolate C-terminal fragments, suggesting that the mature protein is cleaved at multiple sites and/or that the initial cleavage fragment is readily digested. The amino acid composition of the 6 kDa fragment suggests that the 14 kDa protein is cleaved at residues 50-57, a hydrophobic region that is conserved between rabbit SAA3 and human SAA1. We conclude that the ability of collagenase and stromelysin to degrade SAA3 broadens the repertoire of substrates for these matrix degrading enzymes, and we speculate that the presence of a feedback mechanism that can subvert the autocrine/paracrine stimulation of matrix-degrading enzymes may play a role in limiting matrix degradation during inflammatory conditions.

High pressure gel-permeation assay for the proteolysis of human aggrecan by human stromelysin-1: kinetic constants for aggrecan hydrolysis

The adaptation of an analytical procedure for aggrecan based upon gel-permeation chromatography to an FPLC-based protocol has significantly sped up the analysis. The faster assay has permitted determination of the kinetic constants for digestion of human aggrecan by human stromelysin-1. Monomeric aggrecan appeared to be hydrolyzed by stromelysin-1 to multiple forms with lower molecular weight. The disappearance of high-molecular-weight aggrecan was first-order, showing Km much larger than 2 microM and kc/Km = 4000 M-1 s-1 at pH 7.5. The disappearance of high-molecular-weight aggrecan upon hydrolysis by stromelysin-1 at pH 5.5 was also first-order, with kc/Km = 10,700 M-1 s-1. The disappearance of high-molecular-weight aggrecan at pH 7.5 was first-order for digestion by human leukocyte elastase with kc/Km = 230,000 M-1 s-1, by human cathepsin G with kc/Km = 4200 M-1 S-1, and by human plasma plasmin with kc/Km = 2800 M-1 s-1, all with Km much larger than 2 microM.

Production of both 92- and 72-kDa gelatinases by bone cells

We investigated the ability of murine bone organ cultures and osteoblast-like bone cells to produce 72- and 92-kDa gelatinase. 4-6 day newborn mouse calvaria cultures were found to release gelatinase activity into their conditioned medium (CM). This activity was increased by four stimulators of resorption, tumor necrosis factor alpha (TNF), interleukin-1 alpha (IL-1), parathyroid hormone (PTH) and the active phorbol ester, 12-O-tetradecanoylphorbol-13-acetate (TPA). Both the 72- and 92-kDa forms of gelatinase were produced by murine bone cultures. In unstimulated bones 72-kDa gelatinase activity was approximately equal to that of the 92-kDa enzyme. IL-1, TNF, PTH and TPA all increased 92-kDa gelatinase activity in the CM of the bone cultures by about 2- to 2.5-fold. In addition TPA and IL-1 also increased 72-kDa gelatinase activity. In unstimulated osteoblast-like MC3T3-E1 cell cultures 72-kDa gelatinase enzyme activity was much greater than 92-kDa activity and was not substantially regulated (less than 40% change) by IL-1, TNF or PTH. In contrast, these agents stimulated 92-kDa gelatinase activity by 2- to 5-fold. As with the MC3T3-E1 cells, primary cells constitutively produced both 72-kDa and 92-kDa gelatinase. This was true for cells with both the most differentiated osteoblast-like phenotype (populations 3 and 4) and the least osteoblast-like phenotype (populations 1 and 2). In unstimulated cultures of all 4-primary populations, 92-kDa gelatinase production was less than 72-kDa and IL-1, TNF and PTH had only small effects on 72-kDa production in any of the populations (less than 60% change).(ABSTRACT TRUNCATED AT 250 WORDS)

Sequential degradation of interstitial collagen by metalloproteinases extracted from tumors of murine ascites hepatomas

Gelatinolytic and collagenolytic proteinases were separately isolated by different extraction methods from the mouse ascites hepatoma MH134, and from rat ascites hepatoma AH109A. The activities of two proteinases in each extract showed no significant differences, but after trypsin activation the activities of proteinases from the highly metastatic MH134 were significantly increased compared to the enzyme activities in AH109A, which has low metastatic potential. The total activities of collagenase and gelatinase were increased 7.2- and 5.1-fold; their specific activities were increased 5.2- and 4.8-fold, respectively. Gelatinase and collagenase from MH134 were characterized on gelatin zymography. The gelatinase had a molecular weight of 99 and activation by 4-aminophenylmercuric acetate (APMA) or trypsin resulted in its conversion to 79 or 79-95 kD, respectively. The collagenase revealed a major gelatinolytic band at 89 kD, which was converted to 85 and 70 kD by APMA-activation, and a minor gelatinolytic band at 60 kD. These proteinases could degrade native type I collagen to small fragments in a cooperative manner. Trypsin inhibitor, which affects the trypsin activation of latent gelatinase, was extracted together with gelatinase. The inhibitory activity of the enzyme from AH109A showed a 4.1-fold higher specific activity and 3.7-fold greater total activity than that from MH134. The proteinase(s) capable of activating the gelatinase was also extracted from MH134.

Inhibition of human type IV collagenase by a highly conserved peptide sequence derived from its prosegment

The proenzyme fragment of the 72 kDa type IV collagenase contains a conserved amino acid sequence, MRKPRCGN(V)PDV, that is shared with other members of the matrix metalloproteinase family, such as interstitial collagenase and stromelysin. This sequence is lost upon the autocatalytic removal of the 80-84 amino acids from the amino terminus of these proenzymes following enzyme activation. The loss of this profragment converts the latent proenzyme species into a stable active enzyme species. In the present study, we demonstrate that this conserved prosegment sequence is an inhibitor of these enzymes and plays a critical role in maintenance of the latent state of the matrix metalloproteinases. Peptides containing the conserved sequence, MRKPRCGNPDV, were capable of inhibiting activated enzyme. Free cysteine was also an effective inhibitor, whereas reduced glutathione was a less effective inhibitor. Oxidized glutathione was not inhibitory. The 72 kDa type IV collagenase holoproenzyme preparations did not contain a free cysteinyl side chain that reacted with the sulfhydryl substitution reagent 5,5'-dithiobis(2-nitrobenzoic acid) (Ellman's reagent). However, addition of ethylenediaminetetraacetic acid to the reaction mixture to generate the apoenzyme form resulted in the detection of titrable sulfhydryl side chains. Based on these data, we postulate that in the latent enzyme state the conserved profragment sequence interacts with the metal atom at the active site through a sulfhydryl-metal atom coordination that is further stabilized by the amino acyl residues surrounding the essential 73Cys residue. Disturbance of this interaction results in enzyme activation.(ABSTRACT TRUNCATED AT 250 WORDS)

Collagen stability and collagenolytic activity in the normal and aneurysmal human abdominal aorta

Two issues were addressed in this study. The first was whether a bona fide collagenase exists within the wall of the aorta. The second was whether the activity of this putative collagenase is greater in aneurysmal tissue. No collagenase could be extracted from the wall of the aorta under nondenaturing conditions. However, hydroxyproline was liberated from tissue samples allowed to autolyze at neutral pH under reaction conditions favoring the activity of collagenase. Such an activity was probably enzymic as it occurred in a time- and temperature-dependent fashion and was suppressed by chelators. In normal and stenotic tissue, activity was increased by adding aminophenylmercuric acetate, an activator of latent collagenase. Examination of the blanks revealed that the collagen of aneurysmal aorta was more soluble than normal. Furthermore, its digestion kinetics differed in a way that suggested that aneurysmal aorta possessed a labile component that was absent from normal tissue. Although the activity of the putative aortic collagenase was higher than normal in the aneurysmal tissue, our assays do not distinguish between changes in the amount or activity of the enzyme and alterations in the collagen.

Purification of a procollagenase-activator present in medium of cultured guinea pig carrageenin granuloma

Activation of procollagenase constitutes a crucial event in collagenolytic activity regulation. In this study we have purified by DEAE-cellulose, Ultrogel AcA-44, and zinc chelate sepharose chromatographies, a procollagenase-activator from the culture medium of the guinea pig carrageenin granuloma model. On SDS-PAGE, the activator migrates as a principal band of Mr approximately 44,000. The molecule activates procollagenase from human lung fibroblasts in a concentration dependent manner and an enhancement of collagenase activity of trypsin-treated crude culture medium was observed. A loss of about 50% of its activity occurs after heating. In addition, this activator degrades gelatin and casein. All these data suggest that this procollagenase-activator might be stromelysin. The activator was found in both phases of the granuloma, at 7 days when collagen is actively deposited and an important proportion of the collagenolytic activity remains in latent form; and at 14 days, when this enzymatic activity is fully expressed.

Identification of rabbit uterine cervical procollagenase activator as rabbit matrix metalloproteinase 3 (stromelysin)

1. We previously reported an endogenous activator of procollagenase from the culture medium of rabbit uterine cervical fibroblasts (Ishibashi et al. (1987) Biochem. J. 241, 527-534). 2. Similar activator was also purified and characterized from rabbit synovial fibroblasts (Vater et al. (1983) J. biol. Chem. 258, 9374-9382), but its mode of activation of procollagenase was reported to be different from that of purified activator from uterine cervical fibroblasts. 3. Here we report the comparative studies of the two activators of procollagenase and demonstrate that they are identified as matrix metalloproteinase 3 (stromelysin) by their immunological and functional criteria. The specific role of the activator in procollagenase activation is also described.

Matrix metalloproteinase 2 from human rheumatoid synovial fibroblasts. Purification and activation of the precursor and enzymic properties

Human rheumatoid synovial cells in culture secrete at least three related metalloproteinases that digest extracellular matrix macromolecules. One of them, termed matrix metalloproteinase 2 (MMP-2), has been purified as an inactive zymogen (proMMP-2). The final product is homogeneous on SDS/PAGE with Mr = 72,000 under reducing conditions. The NH2-terminal sequence of proMMP-2 is Ala-Pro-Ser-Pro-Ile-Ile-Lys-Phe-Pro-Gly-Asp-Val-Ala-Pro-Lys-Thr, which is identical to that of the so-called '72-kDa type IV collagenase/gelatinase'. The zymogen can be rapidly activated by 4-aminophenylmercuric acetate to an active form of MMP-2 with Mr = 67,000, and the new NH2-terminal generated is Tyr-Asn-Phe-Phe-Pro-Arg-Lys-Pro-Lys-Trp-Asp-Lys-Asn-Gln-Ile. However, following 4-aminophenylmercuric acetate activation, MMP-2 is gradually inactivated by autolysis. Nine endopeptidases (trypsin, chymotrypsin, plasmin, plasma kallikrein, thrombin, neutrophil elastase, cathepsin G, matrix metalloproteinase 3, and thermolysin) were tested for their abilities to activate proMMP-2, but none had this ability. This contrasts with the proteolytic activation of proMMP-1 (procollagenase) and proMMP-3 (prostromelysin). The optimal activity of MMP-2 against azocoll is around pH 8.5, but about 50% of activity is retained at pH 6.5. Enzymic activity is inhibited by EDTA, 1,10-phenanthroline or tissue inhibitor of metalloproteinases, but not by inhibitors of serine, cysteine or aspartic proteinases. MMP-2 digests gelatin, fibronectin, laminin, and collagen type V, and to a lesser extent type IV collagen, cartilage proteoglycan and elastin. Comparative studies on digestion of collagen types IV and V by MMP-2 and MMP-3 (stromelysin) indicate that MMP-3 degrades type IV collagen more readily than MMP-2, while MMP-2 digests type V collagen effectively. Biosynthetic studies of MMPs using cultured human rheumatoid synovial fibroblasts indicated that the production of both proMMP-1 and proMMP-3 is negligible but it is greatly enhanced by the treatment with rabbit-macrophage-conditioned medium, whereas the synthesis of proMMP-2 is constitutively expressed by these cells and is not significantly affected by the treatment. This suggests that the physiological and/or pathological role of MMP-2 and its site of action may be different from those of MMP-1 and MMP-3.

Cycloheximide induces stromelysin mRNA in cultured human fibroblasts

Stromelysin is a metalloproteinase that degrades extracellular matrix macromolecules including fibronectin, laminin, collagen IV and proteoglycans. We now report that cycloheximide, an inhibitor of protein synthesis, induces human stromelysin mRNA in fibroblast cultures in a time- and dose-dependent fashion. As determined by Northern hybridization, a 24-h treatment with cycloheximide increased stromelysin mRNA about 20-fold over the control level. In vitro translation or translation in cells after removal of cycloheximide resulted in increased levels of immunoprecipitable stromelysin suggesting that the cycloheximide-induced stromelysin mRNA was functional. Analysis of mRNA stability suggested that the cycloheximide effect is in part due to the increased activation of the stromelysin gene. In contrast to these results, cycloheximide did not induce collagenase mRNA but, rather, prevented its induction by interleukin-1 beta. These data provide evidence for discoordinate regulation of collagenase and stromelysin genes and suggest that a short-lived repressor protein may play a role in the stromelysin gene expression.

Effects of capsaicin on the metabolism of rheumatoid arthritis synoviocytes in vitro

The effects of capsaicin, the ingredient of hot pepper, on rheumatoid arthritis synoviocytes have been studied. Capsaicin was shown to have a direct action on the metabolism of synovial cells. Thus at 10(-6) mol/l and at higher doses DNA synthesis was restored to the control level. Capsaicin at both doses induced an increase in the synthesis of collagenase and at the lower concentration (10(-8) mol/l) only of prostaglandins. These results indicate that the different effects of capsaicin on cellular proliferation and on metabolic activities are dependent on dose. The responses seen in rheumatoid arthritis synoviocytes in vitro might not be mediated by tachykinins if the synovial tissue is still able to produce neuropeptides in the absence of neuronal afferents. These results suggest that capsaicin, in addition to its direct action on the afferent nervous fibres and the consequent release of tachykinins, may also have a direct action on the cells. The mechanisms by which capsaicin stimulates DNA synthesis and production of collagenase and prostaglandin E2, in a manner dependent on dose, remain to be determined.

Secretion and glycosylation of rabbit macrophage type V collagenase

Cultures of freshly isolated rabbit alveolar macrophages were used to study the synthesis, secretion, and glycosylation of type V collagenase. Cells were pulse-labeled with [35S-]methionine for 15 minutes followed by a chase with cold methionine for various time periods. Type V collagenase was identified in the culture supernatants and cell lysates by immunoprecipitation with a specific antiserum. Within 10 minutes of chase, an 82-kDa protein was found in the cell lysates. This protein was subsequently processed to a 92-kDa protein without identifiable intermediate forms. By 60 minutes of chase, intracellular radioactivity was no longer detectable. The larger protein could be detected within 20 minutes in the culture supernatants and accumulated in the medium for 60 minutes of chase time. Only the 92-kDa form was seen in the supernatants and the proteinase was secreted without intracellular storage or membrane association. Treatment of the 92-kDa proteinase with an enzyme which specifically removes N-linked carbohydrates resulted in an apparent reduction in molecular mass of approximately 10 kDa. Deglycosylation of the proteinase did not result in an apparent loss of activity. Thus, it was concluded that macrophage type V collagenase is synthesized as an 82-kDa polypeptide which is glycosylated by N-linkage and secreted.

Characterization and activation of procollagenase from human polymorphonuclear leucocytes. N-terminal sequence determination of the proenzyme and various proteolytically activated forms

Procollagenase of human polymorphonuclear leucocytes was purified to homogeneity using a rapid and reproducible method. The purification procedure included affinity chromatography on zinc chelate Sepharose, ion exchange chromatography on Q-Sepharose fast flow, followed by affinity chromatography on orange Sepharose and finally a gel-permeation step on Sephacryl S-300. It was shown by SDS/PAGE, under reducing conditions, that the latent collagenase of human polymorphonuclear leucocytes consists of a single polypeptide chain with an apparent relative molecular mass of 85,000. Upon deglycosylation by endoglycosidase F digestion, the apparent relative molecular mass of the procollagenase was reduced to 53,000 which is similar to that of the fibroblast enzyme, and indicates a close relationship between both enzymes. Sequence data were determined by direct automated Edman degradation of the purified polymorphonuclear leucocyte procollagenase. The complete sequence of the propeptide region (residue 1-120) was thereby established. The proteolytic activation of the polymorphonuclear leucocyte procollagenase by various enzymes was investigated by determining the N-terminal sequences of the intermediate and final activated forms. Activation by chymotrypsin and cathepsin G led to the active form (Mr 64,000) by cleaving 79 N-terminal residues from the proenzyme. Trypsin activates in a two-step process. Cleavage of 48 N-terminal residues led to a still latent Mr 70,000 species. The final active form (Mr 65,000) was obtained by splitting off 20 additional N-terminal residues.

The effect of phenytoin on collagenase and gelatinase activities in UMR 106-01 rat osteoblastic osteosarcoma cells

Phenytoin (PHT), a widely used anticonvulsant, has been shown to inhibit bone resorption in rodent organ cultures. The drug also has complex effects on bone metabolism including chronic clinical symptoms of osteomalacia. However, the precise mechanism of PHT action in bone is still unclear. Neutral collagenases that specifically cleave native collagen have been implicated in the turnover of connective tissue. The effect of PHT was assessed on collagenase and gelatinase activities from UMR 106-01 rat osteoblastic osteosarcoma cells. Semiconfluent cells were treated with PHT (50 and 10 micrograms/ml) in the presence of bovine parathyroid hormone, b-PTH-(1-34), at 10(-7) M for 24, 48, 72 and 96 h. The media were assayed following concentration, APMA activation, and incubation with native or denatured [3H]-methyl collagen substrate (approximately 100,000 dpm) at 27 degrees C for 18 h and 35 degrees C for 2 h, respectively. Enzyme activities were presented as primary counts per minute for each time point and calculated as % activity of PTH at 10(-7) M. Parathyroid hormone (10(-7) M) stimulated collagenase activity (approximately 65-fold) and gelatinase activity (approximately 400-fold). PHT (50 micrograms/ml) reduced the PTH-stimulated collagenase activity by 18-53% and the gelatinase activity by 58-72%. SDS PAGE and fluorography following PHT treatment indicated a PHT-induced partial inhibition of PTH-stimulated degradation to alpha A chains of Type I collagen. Phenytoin may inhibit bone resorption through its action on the transcription, synthesis, and/or secretion of the collagenolytic enzymes, collagenase and gelatinase.

Secreted proteases. Regulation of their activity and their possible role in metastasis

Extracellular matrix metalloproteases are secreted by the resident cells of the tissue in a proenzyme form, and their extracellular activity is regulated at the level of gene expression, proenzyme activation, and interaction with inhibitors. To understand the molecular mechanisms that control the activity of ECM metalloproteases and their effect on the cellular phenotype, we have established cell lines in which the transcription of the protease genes is repressed. We also have undertaken a detailed study of the pathway of extracellular activation of interstitial procollagenase. Stable transfection of three human tumor cell lines--H-ras-transformed bronchial epithelial cells TBE-1, fibrosarcoma cells HT1080, and melanoma cells A2058--with the adenovirus E1A gene dramatically repressed the expression of the secreted proteases, type IV and interstitial collagenases, and urokinase-type plasminogen activator. Concomitantly, E1A-expressing cells showed reduced metastatic activity in vivo and reduced ability to traverse a reconstituted basement membrane in vitro. Monospecific anti-type IV collagenase antibody inhibited the invasive activity of parental tumor cell lines in the in vitro system, suggesting a possible causal relationship between the effect of E1A on the expression of secreted proteases and the reduced metastatic potential of the E1A-expressing transformants. We have also studied the mechanism of regulation of metalloprotease activity at the level of extracellular activation by investigating the cascade of proteolytic events that results in the activation of interstitial procollagenase. Cocultivation of the major cellular components of skin, dermal fibroblasts, and epidermal keratinocytes induces activation of interstitial procollagenase and prostromelysin in the presence of plasminogen. This activation occurs through a uPA-plasmin-dependent pathway in which plasmin catalyzes the first step in activation of both collagenase and stromelysin by amino-terminal processing. Activated stromelysin can in turn convert plasmin-activated collagenase into a fully active enzyme by removal of approximately 15 amino acid residues from the carboxyl end of the enzyme. This second step of activation results in a 5-8-fold further increase in specific activity of collagenase. This cascade of proteolytic events may constitute a major physiologic pathway of collagenase activation.

A human cartilage metalloproteinase with elastolytic activity

A metalloproteinase with elastolytic properties found in human fetal and osteoarthritic cartilage could not be detected in normal adult cartilage. During extraction the enzyme appeared to be mostly associated with cartilage proteoglycans, from which it can be separated by ion exchange chromatography. This enzyme migrated during electrophoresis with an apparent molecular weight of 62,000 daltons and was found to be fully activated in the tissue under the study conditions. The enzyme showed a preference for substrates rich in non-polar amino acid residues and was capable of breaking down elastin and casein at neutral pH. The enzyme activity can be inhibited by chelating agents and specific affinity reagents, chloroketones, and is not inhibited by other proteinase inhibitors such as PMSF, aprotinin and alpha-1-antitrypsin. This enzyme may play a significant role in conditions demanding rapid cartilage matrix turnover and/or remodeling, such as normal embryonic development or osteoarthritis.

The partial degradation of osteonectin by a bone-derived metalloprotease enhances binding to type I collagen

Cultured neonatal rat calvaria produce latent metalloproteases capable of degrading collagen, gelatin, and osteonectin. The osteonectin-degrading activity was further characterized and found to be optimally active between pH 6 and 8 and inhibited with EDTA and 1, 10-phenanthroline but not phenylmethylsulfonyl fluoride. Analysis of the degradation products of osteonectin by SDS-PAGE in the presence of dithiothreitol showed the generation of a somewhat stable 32,000 mw cleavage product. Comparison of the binding properties of this cleavage product with intact osteonectin indicated that the fragment retained its ability to bind hydroxyapatite in the presence of high salt (2 M NaCl). Importantly, the binding of osteonectin to type I collagen fibrils was enhanced by limited proteolysis.

Type I collagen degradation by mouse calvarial osteoblasts stimulated with 1,25-dihydroxyvitamin D-3: evidence for a plasminogen-plasmin-metalloproteinase activation cascade

To understand the mechanisms regulating osteoid removal by osteoblasts, mouse calvarial osteoblasts were grown on 14C-labelled type I collagen films and stimulated with 1,25-dihydroxyvitamin D-3 (2.5.10(-8) M) for 48-72 h. In the presence of 5% non-inhibitory rabbit serum this resulted in a 2-3-fold increase in collagen degradation and a dramatic change in osteoblast morphology, when compared with untreated osteoblasts. Collagenolysis was accompanied by increased synthesis and release of latent collagenase, gelatinase and stromelysin and a concomitant decrease in their specific inhibitor, TIMP (tissue inhibitor of metalloproteinases). In serum-free medium, osteoblasts failed to degrade collagen, but their ability to lyse collagen could be restored by adding plasminogen (5 micrograms/ml) to the cultures. Plasminogen-dependent collagenolysis was inhibited by human recombinant TIMP (5 units/ml), demonstrating that plasmin, derived from plasminogen, activated latent collagenase and did not itself degrade collagen. Plasminogen activator production was confirmed by culturing osteoblasts on 125I-labelled fibrin plates. Comparison with urokinase-type and tissue-type plasminogen activator standards suggested that osteoblast plasminogen activator was predominantly cell-associated and likely to be of the urokinase type. Immunocytochemistry indicated that osteoblasts also constitutively produce plasminogen activator inhibitor-1. These findings provide evidence for the involvement of a plasminogen-plasmin-latent metalloproteinase activation cascade in type I collagen degradation by osteoblasts, and for its regulation by TIMP and plasminogen activator inhibitor-1.

Identification, partial purification and characterization of high-molecular-weight gelatin-degrading metalloproteinases produced by a rat mammary carcinoma cell line

BC1 rat mammary carcinoma cells were found to secrete a unique profile of metalloproteinases, distinguished by two gelatin-degrading metalloproteinases of Mr greater than 220.10(3) and Mr much greater than 220.10(3). These enzymes were each partially purified by gel-filtration chromatography, and inhibitor studies showed them to be metalloproteinases. Under conditions where denatured collagen types I, II, and V were completely degraded, native collagen types I, II, IV and V, fibronectin, fibrinogen, C1q, casein, and denatured transferrin were not degraded significantly by these enzymes. The relationship of these enzymes to other extracellular matrix-degrading metalloproteinases and their possible significance in tumour invasion and metastasis is discussed.

Immunolocalization of matrix metalloproteinase 3 (stromelysin) in rheumatoid synovioblasts (B cells): correlation with rheumatoid arthritis

Metalloproteinases produced by connective tissue cells may play a key part in the destruction of joints in rheumatoid arthritis. Matrix metalloproteinase 3 (MMP-3; stromelysin) capable of degrading cartilage proteoglycans and type IX collagen and of activating procollagenase was immunolocalised in hyperplastic synovial lining cells in rheumatoid synovium, but not in the cells of normal synovium. Cells responsible for synthesis of MMP-3 have the phenotype of synovioblasts (B cells) by immunoelectron microscopy, but not of phagocytic synovial macrophages (A cells). Cultured monolayer of rheumatoid synovial cells synthesises MMP-3 only under treatment with macrophage conditioned medium. Immunolocalisation of MMP-3 in rheumatoid synovium and cultured synovial cells was possible when the specimens were treated with a monovalent ionophore, monensin. These results suggest that MMP-3 is synthesised and secreted continuously without storage from hyperplastic synovioblasts stimulated by factor(s) derived from activated macrophages present in the synovium.

Gingival fibroblasts degrade type I collagen films when stimulated with tumor necrosis factor and interleukin 1: evidence that breakdown is mediated by metalloproteinases

We previously suggested that periodontal pathogens might mediate connective tissue degradation in periodontal diseases through the ability of antigens from their cell walls to stimulate cytokine production by circulating mononuclear cells. Such cytokines would then induce metalloproteinase (MP) synthesis by resident gingival cells and thus initiate matrix degradation. In the present investigation human gingival fibroblasts (HGFs) were grown on [14C]-labelled type I collagen films and stimulated with either tumor necrosis factor (TNF) or interleukin-1 (IL-1) for 48 h. Collagenolysis occurred in a dose-dependent manner; the optimal dose for human rTNF alpha was 100 ng/ml and for rIL-1 alpha and rIL-1 beta, 1 ng/ml. Collagen degradation was accompanied by increased synthesis and release of the MPs collagenase, gelatinase and stromelysin, and there was a reduction in free TIMP (tissue inhibitor of metalloproteinases): collagenase and stromelysin were detected in both active and latent forms. Cytokine-stimulated collagenolysis was abolished by the addition of exogenous human rTIMP (5 units/ml). We also measured collagenase and TIMP by ELISAs which recognize all forms of collagenase (latent, active or complexed) and TIMP (free or complexed). These showed that while collagenase activity (0.6-1.2 microgram/ml) correlated with lysis, total TIMP levels remained unchanged at approximately 0.2 microgram/ml. These results demonstrate important roles for MPs and TIMP in regulating type I collagen degradation by HGFs, and support the hypothesis that connective tissue destruction during inflammatory diseases may be initiated, at least in part, by TNF and IL-1.

Identification of a metalloproteinase co-purifying with rat tumour collagenase and the characteristics of fragments of both enzymes

A metalloproteinase similar or identical to stromelysin was shown to co-purify with interstitial collagenase from the rat mammary carcinoma cell line, BC1. The mixture of BC1 metalloproteinase and collagenase degraded casein, gelatin, fibronectin, fibrinogen, laminin, proteoglycan and type IV collagen, in addition to types I and II collagen. Using SDS-PAGE and zymography, the Mr of both enzymes was 51.10(3). During storage, the 51.10(3) protein converted to fragments of Mr 34.10(3) and 24.10(3), and isoelectric points of 4.6-5.3 and 5.7-6.0, respectively. The fragments were separated from the intact (Mr 51.10(3) enzymes by DEAE-Sepharose chromatography, but intact metalloproteinase and collagenase activities resisted separation by a range of chromatographic methods. The Mr 34.10(3) fragment retained the proteinolytic activities of the intact enzymes, excepting collagenase cleavage of collagen types I and II. The Mr 24.10(3) fragment had no proteinolytic activity, showed an increase in Mr of 6.10(3) upon reduction, in common with the intact enzymes, and also had similar chromatographic properties to the intact enzymes. The data presented are consistent with a pattern of breakdown which is common to both collagenase and the metalloproteinase, and suggest that both enzymes are comprised of two protein domains.

Regulation of collagenase and collagenase mRNA production in early- and late-passage human diploid fibroblasts

The levels of collagenase and collagenase mRNA produced by early-passage (less than 40% of lifespan completed) and late-passage (greater than 80% of lifespan completed) cultures of human fibroblasts were analyzed. The constitutive levels of collagenase and collagenase mRNA produced by the late-passage cultures were 10-30 x greater than the levels observed in similarly treated early-passage cultures. Immunofluorescence analysis established that the percentage of collagenase-positive cells was also greater (77% vs. 4%) in the late-passage cultures. To determine whether the difference in collagenase production resulted from cell-derived regulatory factors, collagenase production was examined in cultures plated onto substrates coated with fibroblast extracellular matrix (ECM). Collagenase and collagenase mRNA production was enhanced in both types of cultures, although amounts produced by ECM-induced early-passage cultures was significantly less than that produced by similarly treated late-passage cultures. Collagen-coated substrates also induced collagenase synthesis.

Collagenase activity in the normal rat myocardium. An immunohistochemical method

Fibrillar collagen in the myocardium provides a supportive framework for myocytes and capillaries. Disruption of this organized framework has been observed in certain pathological states. Collagen degradation is primarily mediated by the specific enzyme collagenase, which has been found to exist in various tissues including the myocardium. In this report we describe a method that detects collagenase activity in sections of cardiac tissue. This method is on the basis of degradation of collagen by collagenase on one hand and the visualization of disrupted collagen fibers by immunofluorescence on the other. Frozen rat heart sections were incubated under optimal conditions for collagenase activity (37 degrees C in the presence of 0.1 M calcium at pH 7.4) for 24 h and 48 h. Subsequently, immunofluorescence staining with antibody to type I collagen was performed and the collagenous structures were visualized by immunofluorescence light microscopy. As control, untreated rat heart sections and sections incubated in the absence of calcium were similarly treated with antibody. After the 24 h of incubation, we found no change in the structural integrity of collagen fibers. Marked disruption of the type I collagen fibers was observed 48 h after incubation. No evidence of collagen fiber disruption was found in control sections. Experiments with exogenous collagenase resulted in similar collagen fiber disruption in the frozen rat heart sections. We conclude that the disruption of collagen type I fibers after 48 h of incubation, under optimal conditions for collagenolytic digestion, is the result of collagen degradation by intrinsic collagenase of the myocardium.