Type I collagen degradation by invasive oral squamous cell carcinoma.

Expression of extracellular matrix-degrading proteases is required for tumor cell invasion. In the present study we examined the production of type I collagen-degrading matrix metalloproteinases (MMPs) in the invasive oral squamous cell carcinoma-derived cell line HSC-3. In the absence of serum or exogenous growth factors, HSC-3 cells displayed no collagen degradation activity. Addition of serum slightly increased collagen proteolysis. However, addition of epidermal growth factor (EGF) resulted in nearly complete degradation of the collagen matrix. Zymography showed that MMP-2 and -9 are secreted by HSC-3 cells. EGF stimulated secretion of an additional gelatinase with a molecular weight similar to that of MMP-1. Immunoblotting of conditioned medium confirmed that EGF and, to a lesser degree type I collagen, increased production of MMP-1. Finally, in situ hybridization revealed intense expression of MMP-1 in oral squamous cell carcinoma specimens. Together, these results indicate that MMP-1 is expressed, induced by EGF, and required for type I collagen degradation in oral squamous cell carcinoma.

Post-translational proteolytic processing of procollagen C-terminal proteinase enhancer releases a metalloproteinase inhibitor.

Activity of matrix metalloproteinases (MMP) is regulated by a family of proteins called tissue inhibitors of metalloproteinases (TIMP). Four TIMPs have been cloned, and their molecular weights range from 29,000 to 20,000. By reverse zymography, we have observed a metalloproteinase inhibitor with an apparent molecular weight of 16, 500 from medium conditioned by human brain tumor cells. Antibodies directed against TIMPs failed to react with the 16,500 molecular weight inhibitor, indicating that it was not a truncated form of a known TIMP. The inhibitor was isolated from conditioned medium using affinity and ion exchange chromatography. N-terminal sequences of the inhibitor matched amino acid sequences within the C-terminal domain of a protein known as procollagen C-terminal proteinase enhancer (PCPE). Thus, the inhibitor was named CT-PCPE. Comparison of the N-terminal domain of TIMP with CT-PCPE revealed that both contained six cysteine residues. As in the case of TIMP, reduction and alkylation abolished the inhibitory activity of CT-PCPE. Purified CT-PCPE inhibited MMP-2 with an IC(50) value much greater than that of TIMP-2. This implies that MMPs may not be the physiologic targets for CT-PCPE inhibition. However, these results suggest that CT-PCPE may constitute a new class of metalloproteinase inhibitor.

Impaired wound contraction in stromelysin-1-deficient mice.

OBJECTIVE: To determine whether the deletion of stromelysin-1, a single metalloproteinase gene product, will alter the time course and quality of dermal wound repair in mice. SUMMARY BACKGROUND DATA: After dermal injury, a highly coordinated program of events is initiated by formation of a fibrin clot, followed by migration of keratinocytes, contraction of the dermis, recruitment of inflammatory macrophages, formation of granulation tissue with angiogenesis, and finally tissue remodeling. Matrix metalloproteinases are rapidly induced in the dermis and granulation tissue and at the leading edge of the epidermis in the healing wounds. METHODS: Incisional and circular full-thickness wounds 2 to 10 mm were made in the dermis of stromelysin-1-deficient and wild-type mice. The wounds were analyzed for rate of cellular migration and epithelialization. The wound contraction was examined by immunohistochemical staining for alpha-smooth muscle actin and fluorescent staining for fibrillar actin. RESULTS: Independent of the age of the animal, excisional wounds in stromelysin-1-deficient mice failed to contract and healed more slowly than those in wild-type mice. Cellular migration and epithelialization were unaffected in the stromelysin-1-deficient animals. The functional defect in these mice is failure of contraction during the first phase of healing because of inadequate organization of actin-rich stromal fibroblasts. CONCLUSIONS: Excisional dermal wound healing is impaired in mice with a targeted deletion in the stromelysin-1 gene. Incisional wound healing is not affected. These data implicate stromelysin-1 proteolysis during early wound contraction and indicate that stromelysin-1 is crucial for the organization of a multicellular actin network.

Stromelysin-1-deficient fibroblasts display impaired contraction in vitro.

Targeted disruption of the stromelysin-1 gene in mice causes a delay in excisional wound healing due to a failure in wound contraction. Therefore, we postulated that stromelysin-1 activity is responsible for initiating contraction. To test this hypothesis, we compared the contractile capacity of fibroblasts from stromelysin-1 knockout mice (strom-1 KO) with that of normal fibroblasts using a collagen gel contraction model. Fibroblast cultures were established from explants of skin and lung parenchyma from strom-1 KO and wild-type mice, then transferred to the surface of collagen gels. The extent of contraction was determined by measuring greatest gel diameter. Results demonstrated that (1) all fibroblasts contracted collagen gels in a uniform concentric fashion, (2) skin fibroblasts from both sets of mice exhibited greater gel contraction than did lung fibroblasts, and (3) strom-1 KO fibroblasts demonstrated significantly less contraction (21-23%) than wild-type fibroblasts. These data support the hypothesis that absence of stromelysin-1 results in defective fibroblast contraction that may contribute to delayed wound healing.

Transforming growth factor beta-1 decreases interstitial collagenase in healing human fetal skin.

Fetal dermal wounds heal without scarring. Because wound repair requires extracellular matrix turnover, the authors hypothesized that fetal skin would have increased levels of proteinases responsible for matrix degradation compared with adult skin. It was further hypothesized that transforming growth factor beta-1 (TGF-beta1) induces scarring in fetal skin by altering proteinase synthesis. A model of human fetal skin transplanted subcutaneously onto immunodeficient mice was used to study the role of matrix metalloproteinases in healing human fetal skin. In this model, transplanted second trimester fetal skin heals without scarring; addition of TGF-beta1 induces scarring. Proteinases were detected by immunohistochemistry in untransplanted fetal skin, untransplanted adult skin, TGF-beta1-treated fetal skin grafts, and sucrose-treated control grafts. In untransplanted fetal skin, interstitial collagenase, stromelysin-1, and gelatinase A were found in dermal cells and keratinocytes, and around vascular structures. Proteinases were detected in adult skin at similar locations but stained less intensely. Addition of TGF-beta1 decreased interstitial collagenase in fetal skin, but detection of gelatinase A and stromelysin-1 was unchanged. The authors conclude that matrix metalloproteinases are present in midgestation human fetal skin and that more proteinase-containing cells are found in fetal skin than in adult skin. Manipulation of fetal skin with TGF-beta1 is accompanied by a decrease in interstitial collagenase. These data suggest that the increased matrix metalloproteinases found in fetal skin contribute to scarless healing and that the fibrotic effects of TGF-beta1 on fetal skin may be mediated in part by decreasing the synthesis of interstitial collagenase.

Human acellular dermal matrix as a novel model of malignant epithelial cell invasion.

Cancer invasion and metastasis are associated with matrix degradation. We describe a novel in vivo model of invasion by squamous epithelial neoplastic cells derived from transgenic mice grown on acellular human dermis. Human dermis was subjected to multiple freeze-thaw cycles to render it acellular, maintaining the basement membrane of the former dermal-epidermal junction. Cells representing discrete stages of a multistep transgenic mouse model of epidermal carcinogenesis (neonatal transgenic keratinocytes, moderately/poorly differentiated squamous cell carcinoma, and lymph node metastasis) were seeded onto the basement membrane surface, grown in culture for 4 days, grafted in a subpannicular pocket of athymic mice, and harvested after 3 weeks. Histological analysis demonstrated that neonatal transgenic keratinocytes did not degrade the basement membrane or invade the underlying dermis. In contrast, malignant cells derived from both a moderately differentiated squamous carcinoma and a lymph node metastasis were highly invasive. Immunohistochemical analysis revealed collagenase only in nests of invading malignant cells in contact with the dermal matrix, but not in the tumor mass remaining above the basement membrane, suggesting that this proteinase may be required for stromal invasion. This novel model recapitulates the events seen in malignant invasion: transgenic keratinocytes are unable to penetrate the dermis while cells from a moderately differentiated carcinoma and from lymph node metastasis consistently invade.

Matrix metalloproteinases in immunity.

Matrix metalloproteinases (MMPs) are a family of zinc-containing endo-proteinases that share structural domains but differ in substrate specificity, cellular sources, and inducibility. Macrophage production and secretion of large quantities of many MMPs, after contact with matrix proteins, is enhanced by surface determinants on activated T cells and suppressed by cytokines from Th1 and Th2 cells. T cells secrete predominantly the gelatinases MMP-2 and -9, after beta 1, integrin- or vascular cell adhesion molecule (VCAM)-1-dependent stimulation by cytokines and inflammatory mediators. MMPs of both T cells and macrophages facilitate secretion of TNF-alpha, by cleavage of the membrane-bound form. T cell MMPs prepare connective tissue matrices for T cell chemotaxis across basement membranes and through tissues. The greater amounts of diverse MMPs from macrophages are capable of degrading connective tissues, which may release stored growth factors. In limited studies of animal models of autoimmunity, specific MMP inhibitors have significantly decreased edema and inflammatory tissue damage, suggesting possible therapeutic benefits.

Neutrophil elastase processing of gelatinase A is mediated by extracellular matrix.

Gelatinase A (72-kDa type IV collagenase) is a metalloproteinase that is expressed by many cells in culture and is overexpressed by some tumor cells. It has been suggested that the serine proteinase neutrophil elastase might play a role in the posttranslational processing of gelatinase A and that noncatalytic interactions between gelatinase A and components of the extracellular matrix might alter potential processing pathways. These questions were addressed with the use of gelatin substrate zymography, gelatinolytic activity assays, and amino acid sequence analysis. We found that neutrophil elastase does proteolytically modify gelatinase A by cleaving at a number of sites within gelatinase A. Sequential treatment of gelatinase A with 4-aminophenylmercuric acetate (APMA) and neutrophil elastase yielded an active gelatinase with a 4-fold increase in gelatinolytic activity. The increased gelatinolytic activity correlated with that of a 40-kDa fragment of gelatinase A. Matrix components altered the proteolytic modifications in gelatinase A that were mediated by neutrophil elastase. In the absence of gelatin, neutrophil elastase destructively degraded gelatinase A by hydrolyzing at least two bonds within the fibronectin-like gelatin-binding domain of gelatinase A. In the presence of gelatin, these two inactivating cleavage sites were protected, and cleavage at a site within the hemopexin-like carboxyl-terminal domain resulted in a truncated yet active gelatinase. The results suggest a regulatory role for extracellular matrix molecules in stabilizing gelatinase A fragments and in altering the availability of sites susceptible to destructive proteolysis by neutrophil elastase.

Wound healing in the fetus. Possible role for inflammatory macrophages and transforming growth factor-beta isoforms.

Macrophages are believed to play a crucial role in wound healing by synthesizing and secreting numerous cytokines. Some of these cytokines, such as transforming growth factor-beta and tumor necrosis factor-alpha, promote fibrosis and repair. We have shown that macrophages are recruited to sterile fetal wounds and have the potential to regulate repair by synthesizing transforming growth factor-beta(1), transforming growth factor-beta(2), and tumor necrosis factor-alpha. Transforming growth factor-beta was present in fetal lamb wounds in higher amounts than in adult sheep wounds. Furthermore, the concentrations and ratios of the transforming growth factor-beta isoforms in wounds that healed without scarring were different from those in wounds that scarred; transforming growth factor-beta(2) was highest in fetal wounds that did not scar and lowest in adult wounds. These data suggest that concentrations of transforming growth factor-beta isoforms rather than total transforming growth factor-beta concentration may be important in the regulation of fibrosis in prenatal and postnatal wound healing.

Large induction of keratinocyte growth factor expression in the dermis during wound healing.

Recent studies have shown that application of basic fibroblast growth factor (basic FGF) to a wound has a beneficial effect. However, it has not been assessed whether endogenous FGF also plays a role in tissue repair. In this study we found a 160-fold induction of mRNA encoding keratinocyte growth factor (KGF) 1 day after skin injury. This large induction was unique within the family of FGFs, since mRNA levels of acidic FGF, basic FGF, and FGF-5 were only slightly induced (2- to 10-fold) during wound healing, and there was no expression of FGF-3, FGF-4, and FGF-6 detected in normal and wounded skin. High levels of FGF receptor 1 and FGF receptor 2 mRNA and low levels of FGF receptor 3 mRNA were found in both normal and wounded skin. No change in the levels of these transcripts was detected during wound healing. In situ hybridization studies revealed highest levels of KGF mRNA expression in the dermis at the wound edge and in the hypodermis below the wound. In contrast, mRNA encoding the receptor of this growth factor (a splice variant of FGF receptor 2) was predominantly expressed in the epidermis. These results suggest that basal keratinocytes are stimulated by dermally derived KGF during wound healing and implicate a unique role of this member of the FGF family in wound repair.

Secreted inhibitors of metalloproteinases (IMPs) that are distinct from TIMP.

The tissue inhibitor of metalloproteinases (TIMP, M(r) 30,000) is secreted by many cell and tissue types and has been shown to inhibit most secreted mammalian metalloproteinases. In matrix and tissue invasion assays, the inactivation or removal of TIMP enhances invasiveness. However, many of the cells that secrete TIMP also secrete other metalloproteinase inhibitors. By analysis of medium conditioned by various endothelial, mesenchymal, and neural cells on SDS-.substrate-polyacrylamide-inhibitor gels (reverse zymograms), we have detected at least three other distinct inhibitors of metalloproteinases (IMPs). Some or all of these IMPs have been detected in secretions of mouse, rabbit, sheep, and human cells and are all smaller in apparent molecular size than TIMP (IMP-1, M(r) 26,000; IMP-2, M(r) 21,000; IMP-3, M(r) 18,000). These IMPs are not proteolytic degradation products of TIMP nor do they represent nonglycosylated TIMP. The IMPs do not cross-react in the native or denatured state with any of several anti-TIMP antibodies. The IMPs appear to be regulated independently of each other and of TIMP. In vitro, the complex consisting of one of the IMPs, or TIMP, and a metalloproteinase can be dissociated into functional inhibitor and metalloproteinase. Whether this characteristic is significant in vivo is not known. IMP-2 has been purified from several sources and shares sequence homology with TIMP, suggesting that the IMPs and TIMP may constitute a gene family. The most significant characteristic of IMP-2 is that it appears to preferentially inhibit, on a mole:mole basis, the M(r) 68,000 gelatinase rather than collagenase or stromelysin.(ABSTRACT TRUNCATED AT 250 WORDS)

Binding of tissue inhibitor of metalloproteinases 2 to two distinct sites on human 72-kDa gelatinase. Identification of a stabilization site.

We have identified a binding site for tissue inhibitors of metalloproteinases 2 (TIMP-2) on human 72-kDa gelatinase that is distinct from the active site. 72-kDa progelatinase is found in a complex with TIMP-2 in the medium of cultured cells and can be activated with organomercurial compounds to yield a gelatinolytic proteinase that remains bound to TIMP-2. Removal of TIMP-2 from 72-kDa progelatinase by reverse-phase high performance liquid chromatography, followed by reconstitution of the progelatinase in neutral pH buffer, results in autocatalytic activation. When samples of autoactivated gelatinase were blotted onto nitrocellulose, then probed with 125I-TIMP-2, we found a 29-kDa peptide that was capable of binding TIMP-2. We isolated this fragment and identified it as the region of gelatinase from amino acid 414 to the carboxyl terminus in the primary amino acid sequence of progelatinase. This portion of the molecule does not contain the putative zinc- or gelatin-binding sites and is proteolytically inactive. Incubation of 125I-TIMP-2 with 72-kDa progelatinase-TIMP-2 complexes resulted in a concentration-dependent exchange of labeled TIMP-2 with unlabeled TIMP-2, in both the presence and absence of the metalloproteinase inhibitor 1,10-phenanthroline. Saturation binding kinetics for the active site of 72-kDa gelatinase were measured in pools of the 43-kDa active fragment that results from the autoactivation of 72-kDa progelatinase; this fragment has no carboxyl-terminal TIMP-2 binding capability. Binding of 125I-TIMP-2 to the active site was completely inhibited by 1,10-phenanthroline. Binding kinetics for the putative stabilization site were determined with isolated 72-kDa progelatinase. In the presence of 1,10-phenanthroline, 72-kDa progelatinase bound 125I-TIMP-2 but not 125I-TIMP-1. Scatchard analysis yielded an approximate dissociation constant (Kd) of 0.72 nM for the active site and 0.42 nM for the stabilization site.

Regulation of the autoactivation of human 72-kDa progelatinase by tissue inhibitor of metalloproteinases-2.

To study the activation of human 72-kDa gelatinase, and its relation to tissue inhibitor of metalloproteinases 2 (TIMP-2), we purified human 72-kDa progelatinase both as a complex with TIMP-2 and as a free proteinase. Activation of progelatinase-TIMP-2 complexes with 4-aminophenylmercuric acetate yielded gelatinolytically active enzyme migrating at 62 kDa. TIMP-2 remained bound to the active enzyme. Removal of TIMP-2 from progelatinase by reverse-phase high performance liquid chromatography in the presence of trifluoroacetic acid, followed by complete dialysis in neutral pH buffer, resulted in multiple fragments. These fragments were formed as a result of the cleavage of 72-kDa progelatinase at several locations. Cleavage at the amino terminus was restricted to the removal of the propeptide, except in the case of degradation leading to inactive fragments. Two active species autocatalytically evolved upon removal of TIMP-2 from progelatinase. The 62 kDa-activated gelatinase lacked the amino-terminal propeptide, which is known to be removed upon treatment with 4-aminophenylmercuric acetate. In addition, an active 42.5-kDa fragment lacking both the propeptide and a portion of the carboxyl terminus was formed. This low-molecular-weight active form of 72-kDa progelatinase retained its ability to bind and degrade gelatin. Self-activation and degradation of 72-kDa progelatinase can be prevented by agents that inhibit metalloproteinases, including 1,10-phenanthroline. Evidence presented here suggests that TIMP-2 binds to a stabilization site that is independent of the active site. This stabilization site does not bind TIMP-1 (TIMP). Occupation of this site by TIMP-2 prevents autocatalytic activation and degradation but does not prevent gelatinolysis by the enzyme-inhibitor complex.

Preferential inhibition of 72- and 92-kDa gelatinases by tissue inhibitor of metalloproteinases-2.

Transformed human fibroblasts secrete two structurally and functionally related inhibitors of matrix metalloproteinases, tissue inhibitor of metalloproteinases (TIMP) 1 and 2. In assays measuring the relative inhibitory capability of TIMP-1 and TIMP-2 against autoactivated 72-kDa gelatinase, which consists of two major active peptides and several inactive fragments, TIMP-2 was more effective than TIMP-1. The isolated 42.5-kDa active fragment that formed as a result of the autoactivation of 72-kDa gelatinase showed the greatest preference for TIMP-2; at half-maximal inhibition, TIMP-2 was greater than 10-fold more effective than TIMP-1. TIMP-2 was also greater than 2-fold more effective than TIMP-1 at inhibiting 72-kDa gelatinase-TIMP-2 complexes activated with 4-aminophenylmercuric acetate, and greater than 7-fold more effective than TIMP-1 at inhibiting 92-kDa gelatinase activated with 4-aminophenylmercuric acetate. Furthermore, these active gelatinases preferentially bound 125I-TIMP-2 when incubated with equal amounts of radiolabeled TIMP-1 and TIMP-2. The ratios of 125I-TIMP-2/125I-TIMP-1 binding to 92-kDa gelatinase, autoactivated 72-kDa gelatinase, and 42.5-kDa fragment were 4.4, 10, and 33, respectively. On the other hand, interstitial collagenase was inhibited by TIMP-1 greater than 2-fold more effectively than TIMP-2 in assays measuring cleavage of loose collagen fibrils.

Expression of metalloproteinases and metalloproteinase inhibitors by fetal astrocytes and glioma cells.

Metalloproteinases have been implicated as important factors mediating the tissue migration of a variety of normal and transformed cells. The conditioned medium (CM) of fetal human astrocytes and five glioma cell lines did not degrade azocoll in suspension, but several proteolytic activities, inhibitable by 1,10-phenanthroline, were detected on sodium dodecyl sulfate-polyacrylamide gels containing gelatin. Both cell types secreted three major proteolytic species (Mr 65,000, 57,000, and 52,000). Two of the glioma lines secreted an additional proteinase (Mr 92,000). After treatment with 12-O-tetradecanoylphorbol-13-acetate, the secretion of the Mr 92,000, 57,000, and 52,000 proteinases was induced or enhanced in all of the cells. The Mr 92,000 and 65,000 proteinases bound specifically to a gelatin affinity column. When purified by preparative gel electrophoresis, the Mr 65,000 proteinase was found to degrade type IV procollagen. The Mr 57,000 and 52,000 species were precipitated by anticollagenase IgG. Tissue inhibitor of metalloproteinases was detected in the CM of all of the cells by substrate gel analysis and immunoprecipitation of [35S]methionine-labeled proteins with anti-tissue inhibitor of metalloproteinases IgG. The glioma lines also secreted various amounts of two smaller inhibitors of metalloproteinases (IMPs), also seen in rabbit brain capillary endothelial cell CM (IMP-1 at Mr 22,000 and IMP-2 at Mr 19,000), and an inhibitor not previously identified (IMP-3 at Mr 16,500). 12-O-Tetradecanoylphorbol-13-acetate stimulated the secretion of tissue inhibitor of metalloproteinases in all of the cells and induced IMPs in some of the glioma lines. When gel filtration chromatography of concentrated CM was used to resolve inhibitors from proteinases, the isolated proteinases had activity against azocoll and the glycoprotein and collagen components of an in vitro model of the extracellular matrix. The secretion of a battery of metalloproteinases by astrocytes may be important in facilitating astrocytic migration during development and in pathological conditions such as inflammation or local invasion of astrocytic neoplasms.

Wound macrophages express TGF-alpha and other growth factors in vivo: analysis by mRNA phenotyping.

The presence of macrophages is required for the regeneration of many cell types during wound healing. Macrophages have been reported to express a wide range of mitogenic factors and cytokines, but none of these factors has been shown in vivo to sustain all the wound-healing processes. It has been suggested that transforming growth factor-alpha (TGF-alpha) may mediate angiogenesis, epidermal regrowth, and formation of granulation tissue in vivo. Macrophages isolated from a wound site, and not exposed to cell culture conditions, expressed messenger RNA transcripts for TGF-alpha, TGF-beta, platelet-derived growth factor A-chain, and insulin-like growth factor-1. The expression of these transcripts was determined by a novel method for RNA analysis in which low numbers of mouse macrophages were isolated from wound cylinders, their RNA was purified and reverse-transcribed, and the complementary DNA was amplified in a polymerase chain reaction primed with growth factor sequence-specific primers. This single-cell RNA phenotyping procedure is rapid and has the potential for quantification, and mRNA transcripts from a single cell or a few cells can be unambiguously demonstrated, with the simultaneous analysis of several mRNA species. Macrophages from wounds expressed TGF-alpha antigen, and wound fluids contained TGF-alpha. Elicited macrophages in culture also expressed TGF-alpha transcripts and polypeptide in a time-dependent manner after stimulation with modified low-density lipoproteins and lipopolysaccharide endotoxin, which are characteristic of the activators found in injured tissues.

The inhibitory complex of human alpha 1-proteinase inhibitor and human leukocyte elastase is a neutrophil chemoattractant.

An inhibitor-proteinase complex consisting of human alpha 1-PI and human leukocyte elastase is chemotactic for human neutrophils. The chemotactic activity is optimal at 1 nM and is associated only with the alpha 1-PI portion of the complex. Neither HLE in the complex, free HLE, nor native alpha 1-PI possesses chemotactic activity for human neutrophils. alpha 1-PI in complex is hydrolyzed at the Met-358-Ser-359 bond. The chemotactic activity is associated with the Mr 4,200 fragment of alpha 1-PI that has Ser-359 as its NH2 terminus. The region of the HLE-alpha 1-PI complex that stimulates chemotaxis appears to be the same as that of the Mr 4,200 fragment generated by hydrolysis of the Pro-357-Met-358 bond during proteolytic inactivation of alpha 1-PI. The data suggest the presence of a neutrophil surface receptor bound by alpha 1-PI after the formation of a complex with HLE or after proteolytic degradation. This receptor may play a role in clearance of these modified alpha 1-PI molecules.

Alpha 1-proteinase inhibitor is a neutrophil chemoattractant after proteolytic inactivation by macrophage elastase.

Mouse macrophage elastase, a metalloproteinase, catalytically inactivates human alpha 1-proteinase inhibitor (alpha 1-PI) by attacking a single peptide bond between Pro357 and Met358, resulting in Mr = 4,200 and 47,800 fragments. We show here that this proteolytically inactivated alpha 1-PI is a potent chemotactic factor for human neutrophils at a concentration of 1 nM. The chemotactic response is equivalent to that elicited by formyl-methionyl-leucyl-phenylalanine. Native alpha 1-PI does not stimulate chemotaxis. Purification of the two fragments of alpha 1-PI that result from proteolysis by macrophage elastase indicated that the Mr = 4,200 fragment is responsible for the chemotactic activity. However, the two proteolysis fragments do not dissociate from each other under physiologic conditions. Therefore, the ability of proteolytically inactivated alpha 1-PI to act as a mediator of inflammation is due to rearrangement of the alpha 1-PI molecule rather than to release of a cleavage fragment.