Aims-To investigate the role of the matrix metalloproteinases (MMPs) in the connective tissue changes seen in the intestine in Crohn's disease. Methods-Indirect immunofluorescence microscopy using specific antibodies to the MMPs (collagenase, gelatinase A and B, and stromelysin) were used to assess the distribution of these enzymes in normal and diseased intestine. Results-In normal intestine the matrix metalloproteinases were confined to a few isolated inflammatory cells, but in Crohn's disease, the inflammatory infiltrate was associated with increased numbers of polymorphonuclear leucocytes which stained positive for gelatinase B. Stromelysin was also detected extracellularly on the connective tissue matrix in regions of smooth muscle cell proliferation and mucosal degradation. Interestingly, in ulcerative colitis, another inflammatory bowel disease, stromelysin was localised in the lamina propria in regions of mucosal loss. Conclusions-The increased numbers of inflammatory cells containing gelatinase B, and the localisation of extracellular stromelysin in regions of fibrosis and mucosal degradation, suggest that these enzymes have a role in the pathological changes seen in Crohn's disease. In cases of ulcerative colitis stromelysin was also detected on the laIina propria in regions of mucosal loss, and seems to be associated with the connective tissue changes that precede mucosal loss.

The collagen content of resected strictured intestine, with and without fistulas, from patients with Crohn's disease has been compared with that in macroscopically normal intestine removed from the same patients and from others without inflammatory bowel disease. Collagen content per unit wet or dry weight of tissue was significantly increased in all grossly diseased tissue whether fistulated or not. Although there was a significant increase in collagen types I, III, and V in diseased tissue, the relative proportions of major collagen types extracted by limited pepsin digestion were similar for both Crohn's and non-Crohn's intestine (type I, 65 to 70 percent; type III, 25 to 30 percent; type IV, 2 to 3 percent; and type V, 2.5 to 3 percent). CNBr digestion of pepsin insoluble material showed a similar relative abundance of types I and III, indicating no major change in collagen type distribution between older (insoluble) and more newly synthesized collagen. There was no evidence of the presence of type I trimer collagen. Type VI collagen, although not quantitated, was observed in 70 percent of intestinal specimens. The proportion of total collagen solubilized by pepsin treatment was significantly greater in both grossly diseased and macroscopically normal Crohn's bowel compared with non-inflammatory bowel disease bowel. These findings suggest that there are disturbances of collagen metabolism in Crohn's intestine, which account for the stricturing process and which may predate gross pathologic changes.

No abstract