Distinct localization of collagenase and tissue inhibitor of metalloproteinases expression in wound healing associated with ulcerative pyogenic granuloma

Enhanced expression of interstitial collagenase, stromelysin-1, and urokinase plasminogen activator in lesions of dermatitis herpetiformis

Because dermatitis herpetiformis is characterized by neutrophilic inflammation and destructive changes in the basement membrane zone, we studied the in situ expression of interstitial collagenase and stromelysin-1 in 11 lesions. A prominent signal for collagenase mRNA was consistently detected in the basal keratinocytes of rete ridges surrounding the neutrophilic abscesses in 10 of 11 lesions, and the expression was independent of the age of the lesion and the migratory state of the basal keratinocytes. Expression of stromelysin-1 was detected in seven of 11 lesions and co-localized with collagenase. No expression of the 92-kDa gelatinase mRNA or matrilysin protein was found in the vicinity of neutrophilic accumulations or the damaged basement membrane. Urokinase-type plasminogen activator mRNA was found in basal keratinocytes in seven of nine samples. Collagenase, stromelysin-1, and urokinase-type plasminogen activator were not expressed in normal-appearing skin of patients with dermatitis herpetiformis. Our results suggest that in lesions of dermatitis herpetiformis, collagenase and stromelysin-1 may be induced in basal keratinocytes by neutrophil cytokines or by altered cell-matrix interactions through contact of keratinocytes with the matrix due to damaged basement membrane. Stromelysin-1, in particular, may contribute to formation of blisters by degrading basement membrane components.

Production and localization of 92-kilodalton gelatinase in abdominal aortic aneurysms. An elastolytic metalloproteinase expressed by aneurysm-infiltrating macrophages

Abdominal aortic aneurysms (AAA) are characterized by disruption and degradation of the elastic media, yet the elastolytic proteinases involved and their cellular sources are undefined. We examined if 92-kD gelatinase, an elastolytic matrix metalloproteinase, participates in the pathobiology of AAA. Gelatin zymography of conditioned medium from normal, atheroocclusive disease (AOD), or AAA tissues in organ culture showed that all tissues produced 72-kD gelatinase. AOD and AAA cultures also secreted 92-kD gelatinase, but significantly more enzyme was released from AAA tissues. ELISA confirmed that AAA tissues released approximately 2-fold more 92-kD gelatinase than AOD tissue and approximately 10-fold more than normal aorta. Phorbol ester induced a 5.3-fold increase in 92-kD gelatinase secretion by normal aorta and AOD and an 11.5-fold increase by AAA. By immunohistochemistry, 92-kD gelatinase was not detected in normal aorta and was only occasionally seen within the neointimal lesions of AOD tissue. In all AAA specimens, however, 92-kD gelatinase was readily localized to numerous macrophages in the media and at the adventitial-medial junction. The expression of 92-kD gelatinase mRNA by aneurysm-infiltrating macrophages was confirmed by in situ hybridization. These results demonstrate that diseased aortic tissues secrete greater amounts of gelatinolytic activity than normal aorta primarily due to increased production of 92-kD gelatinase. In addition, the localization of 92-kD gelatinase to macrophages in the damaged wall of aneurysmal aortas suggests that chronic release of this elastolytic metalloproteinase contributes to extracellular matrix degradation in AAA.

Interstitial collagenase is expressed by keratinocytes that are actively involved in reepithelialization in blistering skin disease

Migrating keratinocytes actively involved in reepithelialization in dermal wounds acquire a collagenolytic phenotype upon contact with the dermal matrix. To determine whether this phenotype is associated with repair in other forms of wounds, we assessed collagenase expression in 50 specimens representing a variety of blistering skin diseases, including subtypes of epidermolysis bullosa, porphyria cutanea tarda, bullous pemphigoid, pemphigus, transient acantholytic dermatosis, and suction blisters. Distinct from that seen in chronic ulcers or in normal healing by second intention, reepithelialization in these blistering conditions was not necessarily associated with a complete loss of basement membrane, as determined by immunostaining for type IV collagen. Collagenase mRNA was detected in the basal keratinocytes of several specimens of epidermolysis bullosa simplex (six of 10) and of pemphigus (three of seven), as well as in one quarter of transient acantholytic dermatosis samples in the presence of an intact basement membrane. In contrast, three of nine porphyria cutanea tarda, one third of epidermolysis bullosa acquisita, and one of 10 bullous pemphigoid samples had collagenase-positive basal keratinocytes with the basement membrane disrupted. The collagenase-positive lesions generally represented older blisters with evidence of epithelial regeneration. Collagenase was also expressed in suction blisters at 2 and 5 d after induction of the blister, but was shut off when the epidermis had healed. Other metalloproteinases were expressed occasionally, if at all. Our results suggest that keratinocyte migration is associated with collagenase expression and that contact of keratinocytes with the dermal matrix is not necessarily needed for collagenase induction.

Collagenase expression is rapidly induced in wound-edge keratinocytes after acute injury in human skin, persists during healing, and stops at re-epithelialization

Collagenolytic activity has been reported previously in association with wounds. We used in situ hybridization and immunohistochemistry to localize cellular sites of interstitial collagenase production in acute wounds in human skin at days 1, 2, 4, 6, 9, and 14 after wounding. In vivo, collagenase expression peaked in migrating basal keratinocytes at the wound edge at day 1, then gradually decreased and was undetectable at day 9 when healing was complete. To minimize the effects of crust formation and inflammation, we examined the healing of wounds made with a 3-mm punch in organ-cultured skin. In these in vitro wounds, re-epithelialization occurred by 5-7 d in 10% serum, although remodeling of the connective tissue was minimal. Collagenase expression showed a similar pattern as in the in vivo wounds; it was detected in migrating keratinocytes already 4-6 h after wounding, peaked at 12-24 h, gradually decreased during the next few days, and subsided upon re-epithelialization. In dermal fibroblasts, on the other hand, expression of collagenase started considerably later, after 5-7 d in culture, and persisted after complete re-epithelialization, indicating that collagenase is differentially regulated in different cell types. Our findings also show that collagenase induction in keratinocytes does not require inflammation and occurs as a rapid response to wounding, suggesting that interstitial collagenase is not only necessary for remodeling of the extracellular matrix, but may also have a role in initiating migration of keratinocytes in wound healing.

Tissue inhibitor of metalloproteinases-1 is decreased and activated gelatinases are increased in chronic wounds

The balance between matrix deposition and tissue turnover is fundamental in wound healing. It is likely that the balance between proteolytic enzymes and their inhibitors contributes to this balance. Matrix metalloproteinases are clearly important in tissue turnover, but their roles in wound healing are poorly understood. To investigate this, fluid from healing wounds resulting from mastectomies was collected from 1 h to 10 d post-surgery, and was analyzed for tissue inhibitor of metalloproteinases-1 concentrations. In all cases, tissue inhibitor of metalloproteinases-1 levels were initially comparable to those in serum, but increased rapidly to significantly higher levels within two days, with a tenfold average increase for five patients. On the other hand, zymography revealed that gelatinase A (72 kDa) levels increased moderately, whereas gelatinase B levels (92 kDa) decreased an average of twofold within 4 d. In contrast, fluid from chronic wounds had significantly more gelatinolytic activity, including lower-molecular-weight proteinase species that may represent activated or superactivated gelatinase fragments, as suggested by immunoprecipitation with specific antibodies. Tissue inhibitor of metalloproteinases-1 levels were lower in chronic than in healing wounds. These data may indicate that excess proteolysis in chronic wounds retards successful healing, and results from an imbalance of proteinase and inhibitors, as well as the presence of higher levels of activated metalloproteinases.

Degradation of the COL1 domain of type XIV collagen by 92-kDa gelatinase

Type XIV collagen is a newly described member of the fibril-associated collagens with interrupted triple helices (FACITs). Expression of this collagen has been localized to various embryonic tissues, suggesting that it has a functional role in development. All FACITs thus far described (types IX, XII, XIV, and XVI) contain a highly homologous carboxyl-terminal triple helical domain designated COL1. We have studied the capacity of various matrix metalloproteinases (interstitial collagenase, stromelysin, matrilysin, and 92-kDa gelatinase) to degrade the COL1 domain of collagen XIV. We found that only 92-kDa gelatinase cleaves COL1. Furthermore, digestion of whole native collagen XIV by the 92-kDa gelatinase indicates that this enzyme specifically attacks the carboxyl-terminal triple helix-containing region of the molecule. COL1 is cleaved by 92-kDa gelatinase at 30 degrees C, a full 5-6 degrees C below the melting temperature (Tm) of this domain; native collagen XIV is also degraded at 30 degrees C. In comparison to interstitial collagenase degradation of its physiologic native type I collagen substrate, the 92-kDa enzyme cleaved COL1 (XIV) with comparable catalytic efficacy. Interestingly, following thermal denaturation of the COL1 fragment, its susceptibility to 92-kDa gelatinase increases, but only to a degree that leaves it several orders of magnitude less sensitive to degradation than denatured collagens I and III. These data indicate that native COL1 and collagen XIV are readily and specifically cleaved by 92-kDa gelatinase. They also suggest a role for 92-kDa gelatinase activity in the structural tissue remodeling of the developing embryo.

Regulation of collagenase gene expression by IL-1 beta requires transcriptional and post-transcriptional mechanisms

Interleukin-1 beta is believed to contribute to the pathophysiology of rheumatoid arthritis by activating collagenase gene expression. We have used a cell culture model of rabbit synovial fibroblasts to examine the molecular mechanisms of IL-1 beta-mediated collagenase gene expression. Stimulation of rabbit synovial fibroblasts with 10 ng/ml recombinant human IL-1 beta resulted in a 20-fold increase in collagenase mRNA by 12 h. Transient transfection studies using collagenase promoter-CAT constructs demonstrated that proximal sequences responded poorly to IL-1 beta, possibly due to insufficient activation of AP-1 by this cytokine. More distal sequences were required for IL-1 beta responsiveness, with a 4700 bp construct showing approximately 5-fold induction above control. To examine post-transcriptional mechanisms, transcript from a human collagenase cDNA was constitutively produced by the simian virus 40 early promoter. IL-1 beta stabilized the constitutively expressed human transcript. Furthermore, mutation of the ATTTA motifs in the 3' untranslated region of the human gene also stabilized the transcript. Finally, the rabbit collagenase 3' untranslated region destabilized a constitutively transcribed chloramphenicol acetyltransferase transcript. These data indicate that in addition to activating transcription, IL-1 beta increases collagenase transcript stability by reversing the destabilizing effects of sequences in the 3' untranslated region.

Distinct populations of basal keratinocytes express stromelysin-1 and stromelysin-2 in chronic wounds

Wound repair involves cell migration and tissue remodeling, and these ordered and regulated processes are facilitated by matrix-degrading proteases. We reported that interstitial collagenase is invariantly expressed by basal keratinocytes at the migrating front of healing epidermis (Saarialho-Kere, U. K., E. S. Chang, H. G. Welgus, and W. C. Parks, 1992. J. Clin. Invest. 90:1952-1957). Because of the limited substrate specificity of collagenase, principally for interstitial fibrillar collagens, other enzymes must also be produced in the wound environment to effectively restructure tissues with a complex matrix composition. Stromelysins-1 and -2 are closely related, yet distinct metalloproteinases, and both can degrade many noncollagenous connective tissue macromolecules. Using in situ hybridization and immunohistochemistry, we found that both stromelysins are produced by distinct populations of keratinocytes in a variety of chronic ulcers. Stromelysin-1 mRNA and protein were detected in basal keratinocytes adjacent to but distal from the wound edge in what probably represents the sites of proliferating epidermis. In contrast, stromelysin-2 mRNA was seen only in basal keratinocytes at the migrating front, in the same epidermal cell population that expresses collagenase. Stromelysin-1-producing keratinocytes resided on the basement membrane, whereas stromelysin-2-producing keratinocytes were in contact with the dermal matrix. Furthermore, stromelysin-1 expression was prominent in dermal fibroblasts, whereas no signal for stromelysin-2 was seen in any dermal cell. These findings demonstrate that stromelysins-1 and -2 are produced by different populations of basal keratinocytes in response to wounding and suggest that these two matrix metalloproteinases serve distinct roles in tissue repair.

Low protein diet blunts the rise in glomerular gene expression in focal glomerulosclerosis

The present study was designed to assess whether expression of mRNA for extracellular matrix (ECM) components, metalloproteinases (MMP) and tissue inhibitor of metalloproteinases (TIMP) in glomeruli is affected by a low protein diet during the course of focal glomerulosclerosis (FGS). Puromycin aminonucleoside (PAN) was injected intraperitoneally in rats and the right kidney was removed on day 22. Nephrotic rats received successive intraperitoneal injections of PAN on days 27, 34, and 41. Control rats were subjected to a nephrectomy or a sham operation on day 22. Animals were divided into six groups. In group 1, the PAN-injected rats were fed a standard diet containing 22% protein. In group 2, the PAN-injected rats were fed a low protein diet containing 6% protein, starting on the same day as the first PAN injection. In group 3, the nephrectomized rats without PAN were fed a standard diet. In group 4, the nephrectomized rats without PAN were fed a low protein diet for the same period. In group 5, the sham operated rats were fed a standard diet. In group 6, the sham operated rats were fed a low protein diet for the same period. Rats were sacrificed on days 0, 60 or 80 after the initial PAN or saline injection. The percentage of sclerotic glomeruli in group 1 rats increased markedly with time, reaching 77% on day 80. The mRNA levels encoding for alpha 1(I), alpha 1(III), alpha 1(IV) collagen chains, laminin B1 and B2 chains, heparan sulfate proteoglycan (HSPG), MMP-2, TIMP-1 and TIMP-2 increased significantly as glomerulosclerosis progressed, whereas MMP-1 and MMP-3 mRNA levels were unchanged, and no MMP-9 mRNA was detected throughout the experiments. In group 2, the low protein diet reduced the prevalence of glomerulosclerosis and attenuated the increased mRNA expression for ECM components, MMP-2, TIMP-1 and TIMP-2 in FGS glomeruli. In groups 3 through 6, mRNA levels for ECM components decreased with age, whereas those for MMPs and TIMPs changed little throughout the experiments. Immunofluorescence studies revealed the accumulation of types I, III and IV collagens, laminin, and HSPG in the sclerotic area and low protein diet attenuated the accumulation of these proteins. These data suggest that glomerulosclerosis may result from an imbalance among ECM components, MMPs and TIMPs and that a low protein diet attenuates the otherwise increased levels of mRNA for ECM components, MMP-2, TIMP-1 and TIMP-2 in glomerulosclerosis.

92-kD gelatinase is produced by eosinophils at the site of blister formation in bullous pemphigoid and cleaves the extracellular domain of recombinant 180-kD bullous pemphigoid autoantigen

Eosinophils are prominent in bullous pemphigoid (BP), and proteases secreted from these and other inflammatory cells may induce disruption of the basement membrane. We used in situ hybridization and immunohistochemistry to localize the sites of 92-kD gelatinase expression in BP lesions. In all samples (20/20), a strong signal for gelatinase mRNA was detected only in eosinophils and was most pronounced where these cells accumulated at the floor of forming blisters. No other cells were positive for enzyme mRNA. Both eosinophils and neutrophils, however, contained immunoreactive 92-kD gelatinase indicating that active expression occurred only in eosinophils. Degranulated eosinophils were also seen near blisters, and as demonstrated by gelatin zymography, immunoblotting, and ELISA, 92-kD gelatinase protein was prominent in BP blister fluid. No other gelatinolytic activity was specifically detected in BP fluid, and only small amounts of 92-kD gelatinase were present in suction blister fluids. As demonstrated in vitro, 92-kD gelatinase cleaved the extracellular, collagenous domain of recombinant 180-kD BP autoantigen (BP180, BPAG2, HD4, type XVII collagen), a transmembrane molecule of the epidermal hemidesmosome. Our results suggest that production and release 92-kD gelatinase by eosinophils contributes significantly to tissue damage in BP.

Immunolocalization of matrix metalloproteinases and TIMP-1 (tissue inhibitor of metalloproteinases) in human gingival tissues from periodontitis patients

The matrix metalloproteinases (MMPs) collagenase, gelatinase A (72 kDa gelatinase), stromelysin, and their specific inhibitor TIMP-1 (tissue inhibitor of metalloproteinases), were immunolocalized using specific polyclonal antisera in gingival tissues from 21 patients with chronic inflammatory periodontal disease. Monoclonal antibodies against macrophages (Leu-M5), B cells (Leu-14), helper T cells (OKT4), suppressor T cells (OKT8) and the HLA-DR epitope were also used to identify leukocyte subsets. MMPs were observed in connective tissues at sites that histologically showed signs of remodelling. The number and distribution of positive cells varied widely, however, not only between individual biopsy specimens, but also within the same specimen. The same was true for the composition and distribution of the inflammatory cell infiltrate. Moreover, although there was a positive correlation between the number of MMP-producing cells and the severity of inflammation in some specimens, for others with comparable leukocyte subset scoring the number was reduced and sometimes absent altogether. Cells secreting MMPs were fibroblasts, macrophages and epithelial cells. It was not possible to determine unequivocally whether a MMP-positive cell within the connective tissue was a fibroblast or a macrophage, since the antisera recognise both fibroblast and macrophage MMPs and the different fixation requirements for MMPs (4% paraformaldehyde) and Leu-M5 (acetone) precluded co-localization on the same section. TIMP-1 was immunolocalized within connective tissue cells at sites of tissue remodelling. Our results support the hypothesis that tissue-derived MMPs may be involved in tissue remodelling in periodontal disease and conclusively demonstrate that epithelial cells may be involved as well as connective tissue cells.

Cell-matrix interactions modulate interstitial collagenase expression by human keratinocytes actively involved in wound healing

We reported that interstitial collagenase is produced by keratinocytes at the edge of ulcers in pyogenic granuloma, and in this report, we assessed if production of this metalloproteinase is a common feature of the epidermal response in a variety of wounds. In all samples of chronic ulcers, regardless of etiology, and in incision wounds, collagenase mRNA, localized by in situ hybridization, was prominently expressed by basal keratinocytes bordering the sites of active re-epithelialization indicating that collagenolytic activity is a characteristic response of the epidermis to wounding. No expression of mRNAs for 72- and 92-kD gelatinases or matrilysin was seen in keratinocytes, and no signal for any metalloproteinase was detected in normal epidermis. Immunostaining for type IV collagen showed that collagenase-positive keratinocytes were not in contact with an intact basement membrane and, unlike normal keratinocytes, expressed alpha 5 beta 1 receptors. These observations suggest that cell-matrix interactions influence collagenase expression by epidermal cells. Indeed, as determined by ELISA, primary cultures of human keratinocytes grown on basement membrane proteins (Matrigel; Collaborative Research Inc., Bedford, MA) did not express significant levels of collagenase, whereas cells grown on type I collagen produced markedly increased levels. These results suggest that migrating keratinocytes actively involved in re-epithelialization acquire a collagenolytic phenotype upon contact with the dermal matrix.

The wounded dermal equivalent offers a simplified model for studying wound repair in vitro

Dermal equivalents (DEs), fabricated by seeding fibroblasts into a collagen lattice, may be used as in vitro models for studying wounding and the remodelling processes observed in vivo. We investigated fibroblast responses to a wound stimulus in vitro by making small wounds in DEs. We found that, following wounding, fibroblasts appeared to migrate towards and into the wound defect and appeared to initiate the closure of the wound by bringing together the cut collagen surfaces over a period of 5-11 days. Fibroblast movement into the wound defects was significantly stimulated in the presence of EGF and PDGF (2.8- and 3.5-fold respectively) but not TGF-beta. Cell proliferation in wounded DEs was up to 21% greater than in non-wounded DEs and cell numbers were stimulated further by the addition of TGF-beta, EGF and PDGF (1.5-, 1.7- and 1.8-fold respectively). Wounded DEs also displayed a 2.1-fold increase in latent collagenase production followed by a 1.3-fold increase in active collagenase levels compared to non-wounded DEs. Staining actin fibers within fibroblasts using rhodamine-phalloidin showed that fibroblasts in DEs were under tension, but that this tension was lost upon wounding. Subsequently the stress fibers reappeared concomitantly with the observed "healing" process. Additionally a continuous cell-cell actin cable purse-string developed around the entire wound edge which may be involved in wound closure. The findings suggest that the wounded dermal equivalent offers a valuable model for studying wound healing in vitro.

Localization of mRNAs representing collagenase and TIMP in sections of healing human burn wounds

Interstitial collagenase, a matrix metalloproteinase, is known to be actively involved in remodeling of cutaneous tissues including those affected by trauma, neoplasia, and inflammation. Conversely, collagenase activity is blocked by tissue inhibitor of metalloproteinases (TIMP). Because both collagenase and TIMP are rapidly secreted into the extracellular matrix, their sites of synthesis remain ambiguous. To determine the site and sequence of collagenase and TIMP expression in cutaneous wound repair, we examined partial and full thickness excisions of human burn wounds representing days 2 to 34 postinjury. Prominent labeling for collagenase and TIMP was detected in epithelial cells at the burn margin and at the edges of surviving hair follicles and eccrine sweat structures in the wound bed. Within the dermis, cells expressing collagenase and TIMP were at first perivascular in location and later appeared at the interface zone between viable and nonviable dermis. A diversity of cell types including macrophages, fibroblasts, endothelial cells, and keratinocytes appeared to express mRNAs for collagenase and TIMP. Little if any labeling was detected in necrotic regions, in adjacent nonwounded dermis, or epidermis. Our data indicate that collagenase and TIMP are temporally and spatially regulated during cutaneous wound repair.

Expression of interstitial collagenase, 92-kDa gelatinase, and tissue inhibitor of metalloproteinases-1 in granuloma annulare and necrobiosis lipoidica diabeticorum

Granuloma annular (GA) and necrobiosis lipoidica diabeticorum (NLD) are disorders characterized by granulomatous inflammation and degenerative changes in collagen and elastic fibers. Because these disorders have often been described as being associated with altered extracellular matrix deposition, we studied the in situ expression of interstitial collagenase, 92-kDa gelatinase, and tissue inhibitor of metalloproteinases (TIMP)-1. Twelve lesions each of GA and NLD of different histopathologic types and durations were examined. Interstitial collagenase mRNA was seen in histiocyte-like cells in one-third of the cases of both diseases, typically in younger lesions. In GA, collagenase mRNA was only detected in lesions of the palisading type. Signal for 92-kDa gelatinase mRNA was observed in eosinophils, which were present in low numbers in five of 12 GA and three of 12 NLD samples. The signal for this enzyme and the presence of eosinophils did not correlate with the age of lesion. TIMP-1 mRNA was consistently expressed by histiocyte-like cells in both disorders. In GA, TIMP-1 mRNA was detected at the outer edge of the palisading granulomas, but in NLD, inhibitor expression was seen in the perivascular and periadnexal accumulation of inflammatory cells. Our data indicate that collagenase and TIMP are expressed early in these disorders and that these proteins may contribute to stromal remodeling associated with necrobiotic lesions. Our results further indicate that the localization of TIMP-1 production may provide a distinction between the two disorders, whereas metalloproteinase expression is not sufficiently specific to aid in the differential diagnosis of GA and NLD.