Mast cell activation in vivo is often associated with areas of oedema and connective-tissue degradation. Tryptase and chymase are the major serine proteinases released by mast cells, but they appear to have little activity on most components of the extracellular matrix. The matrix metalloproteinases (MMP) are purported to degrade almost all connective tissue elements and are secreted by cells in the form of inactive precursors. Since the mechanisms of MMP activation in vivo are poorly understood we have examined the potential of mast cell proteinases to activate the precursor forms of human collagenase (MMP-l), stromelysin (MMP-3), gelatinase A (MMP-2) and gelatinase B (MMP-9).Mast cell proteinases prepared from purified dog mastocytoma cells were shown to process and activate purified precursor forms of both MMP-1 and MMP-3. Using antipain and chymostatin, inhibitors for tryptase and chymase, respectively, it was demonstrated that both pMMP-1 and pMMP-3 were effectively processed and activated by the chymase component. By contrast, tryptase activated only pMMP-3. The mast cell proteinases were unable to process or activate purified precursor forms of MMP-2 and MMP-9. However, MMP-3 previously activated by mast cell proteinases was shown to activate pMMP-9, but not pMMP-2. Since we have no evidence that mast cells express these four metalloenzymes, the release of mast cell serine proteinases following activatioddegranulation could contribute to local metalloproteinase activation and subsequent matrix degradation.Although numerous enzymes may contribute to the degradation of connective tissues, the family of metalloproteinases is of particular interest since they are purported to degrade almost all components of the extracellular matrix [l, 21. These enzymes are grouped into three main subclasses : interstitial [matrix metalloproteinase (MMP)-11 and polymorphonuclear (MMP-8) collagenases which degrade type I, I1 and I11 collagens; gelatinases A and B [MMP-2 (72 kDa) and MMP-9 (92 m a ) ] which degrade basement-membrane type IV collagen and gelatin; and stromelysins (MMP-3, MMP-10) with activity on a broad spectrum of substrates including proteoglycans, laminin, fibronectin and some collagen species [3]. While expressing different substrate specificities these metalloenzymes share some conserved sequence similarity and activation mechanisms [2]. All the metalloproteinases are secreted as precursor forms requiring extracellular activation prior to substrate attack. Activation of the pro- ~-enzymes has been demonstrated in vitro by proteolytic cleavage of their propeptide domains and also by organomercurial compounds such as aminophenyl mercuric acetate (H,NPhHgAc) [2, 4-61, but the activation mechanisms that function in vivo remain unclear.Since mast cells are commonly associated with sites of connective-tissue lysis, for example at sites of tumour invasion in melanoma and breast carcinoma [7, 81 and at cartilage erosion sites in rheumatoid arthritis [9], a potential role for this cell in matrix degradation has be...

Histological studies have previously demonstrated an association between mast-cell activation/degranulation and areas of connective-tissue lysis in vivo; in addition, mast-cell extracts have been shown to activate latent forms of collagenase and stromelysin. In the present study we have examined the potential roles of rat mast-cell proteinase (RMCP) I and RMCP II as activators of the precursors of matrix metalloproteinase (MMP)-1 (interstitial collagenase), MMP-2 (gelatinase A) and MMP-3 (stromelysin 1). Both RMCPs I and II activated proMMP-3 by converting the 57 kDa precursor into a 45 kDa polypeptide. The N-terminal amino acid of 45 kDa MMP-3 activated by RMCP II was identified as Phe83. By contrast, only RMCP II activated the 52 kDa proMMP-1 by converting it into a 41 kDa protein and generating the new N-termini, namely Gln80 and Val82. The collagenolytic activity which resulted from this cleavage was only 35% of the full activity, but this could not be augmented by subsequent treatment with MMP-3, the latter being a crucial enzyme for the generation of the fully active MMP-1 with Phe81 at the N-terminus, in conjunction with other serine proteinases. Thus RMCP II activates proMMP-1 via a mechanism different from that reported for the stepwise processing by combinations of other trypsin-like enzymes and MMP-3. ProMMP-2 (pro-gelatinase A) was not activated by either RMCP I or RMCP II, despite processing to smaller products.

Primary cultures of adherent rheumatoid synovial cells (ASC) are comprised of variable proportions of fibroblasts, macrophages and stellate cells (activated fibroblasts). These cultures were shown to produce the metalloproteinases stromelysin-1 (MMP-3), gelatinase A (MMP-2) and gelatinase B (MMP-9) by Western blotting and zymography techniques. Immunolocalisation studies showed that MMP-3 was mainly produced by the fibroblastic cells whereas MMP-9 was restricted to macrophages (CD68 positive). Subcultured synovial fibroblasts, devoid of macrophages, did not produce MMP-9 as judged by zymography and immunolocalisation; but when stimulated with phorbol myristate acetate and interleukin-1 alpha both MMP-9 and MMP-3 were co-expressed. These 'activated' fibroblasts assumed a dendritic or stellate morphology, which in localisation studies was usually associated with enhanced enzyme production. Immunolocalisation studies of rheumatoid synovial tissue showed that relatively few cells were positive for MMP-3 and MMP-9. Localisation of MMP-9 corresponded to a proportion of macrophages positive for the CD68 marker throughout the synovial tissue. MMP-3 localisation was not associated with the macrophage marker, but was observed in both the synovial lining layer and deeper stromal locations. Widespread distribution of both enzymes was not observed in fresh tissues, but this increased in tissues subjected to short-term explant cultures. Thus, both in vitro and in vivo studies indicated that synovial fibroblasts or B-cells are effective producers of MMP-3 whereas macrophages elaborate MMP-9, observations that demonstrated different metalloproteinase phenotypes under similar environmental conditions.

Matrix metalloproteinases such as collagenase and stromelysin are recognised as important cartilage-degrading enzymes in the pathophysiology of rheumatoid arthritis. Synovial fibroblasts and macrophages are the major cellular components of rheumatoid synovium, but the regulation and relative expression of collagenase and stromelysin by these two cell types remains uncertain. Using in vitro cultures of adherent rheumatoid synovial cells we have examined the coordinate or separate expression of collagenase and stromelysin-1 by dual immunolocalisation and Western blotting techniques. Synovial fibroblasts, when activated by macrophage-derived products in primary culture or by interleukin-1/phorbol myristate acetate in subcultures, released significant quantities of collagenase and stromelysin in their inactive, precursor forms. The ratio of released procollagenase: prostromelysin varied between different synovial cell preparations. Dual immunolocalisation studies demonstrated both coordinate and separate expression of the two enzymes by single cells. Approximately 80% of the activated fibroblasts, especially those with stellate morphology, showed co-expression of both enzymes. By contrast synovial macrophages had a modest or negligible capacity to elaborate either enzyme under the same in vitro conditions. In many fibroblastic cells both collagenase and stromelysin were co-localised to the perinuclear Golgi region and the same cytoplasmic compartments. Vesicular structures appear to provide intracellular transport for both enzymes to sites of secretion. Both enzymes showed preferential pericellular binding to a collagenous substratum rather than any association with the plasma membrane/cell surface.