Determination of matrilysin activity in gastrointestinal neoplasia

Sir, Matrilysin (matrix metalloproteinase-7) is mainly produced by epithelial cells and is involved in extracellular matrix cleavage, and cytokine and growth factor activation [1]. Matrilysin is a target of the APC-Wnt pathway, often affected in early stages of colorectal carcinogenesis. Studies using mouse models confirm that matrilysin is crucial in intestinal tumorigenesis [2]. Enhanced mRNA levels and immunohistochemical matrilysin expression were associated with metastasis and survival of colorectal cancer patients, suggesting its clinical relevance [3,4]. Because protein expression or mRNA levels do not reflect functional proteolytic activity, clinically the most relevant form presumably, we developed an immunocapture bioactivity assay (BIA) to determine simultaneously activity and total matrilysin levels in samples from cancer patients. Maxisorb 96-well plates were coated with three different matrilysin-specific antibodies for 2 h at 37 ° C, blocked (StabilCoat, SurModics, Minneapolis, MN, USA) and washed three times (PBS/0·05% Tween-20). Human recombinant matrilysin (0–32 ng mL –1 , R&D Systems Europe, Abingdon, UK) or sample was allowed to bind overnight at 4 ° C in assay buffer (50 m Tris-HCl pH 7·6, 1·5 m NaCl, 0·5 m CaCl 2 , 1 μ ZnCl 2 and 0·01% Brij-35). After washing (four times), samples were incubated with 1 m 4-aminophenylmercuric acetate (APMA, SigmaAldrich, Zwijndrecht, The Netherlands) or assay buffer to determine total or endogenously active matrilysin levels, respectively. Detection of matrilysin was based on a previously described method [5] using 7·5 μ g mL –1 modified pro-urokinase with a MMP-recognized amino acid activation sequence and 0·96 m chromogenic urokinase substrate S-2444 (Chromogenix, Milan, Italy). Colour development was detected by OD 450 using a Molecular Devices Microplate Reader during 7 h. Optimal immunocapture with a low cross-reactivity (0–8%), as confirmed by spiking with 32 ng mL –1 active MMP-2, -3, -9, -14, was achieved by a goat polyclonal antibody (R&D Systems). The BIA was first validated on gastric cancer tissue homogenates selected for increasing matrilysin levels as determined by a commercial ELISA (R&D Systems) [6]. The total matrilysin amount (BIA, APMA-activated) correlated significantly with the ELISA results (Pearson’s r = 0·910; P < 0·0001, n = 11), whereas active matrilysin levels (BIA, no APMA) did not (Fig. 1). The two most pronounced BIA samples (majority proor majority active matrilysin) showed abundant proor active matrilysin bands on Western blot consistent with the BIA data (Fig. 1 insert). All samples were spiked with 4 ng of active recombinant matrilysin, leading to a consistent increase in all homogenates, except for the two samples with high endogenous matrilysin levels. The intra-assay coefficient of variation of the BIA was 3·2% with an interassay variation of 8·4%. As a pilot study, the BIA was applied to determine matrilysin levels in colonic tissue homogenates. In general, matrilysin levels were up to fivefold enhanced in adenomas and carcinomas, with marked variations in the percentage active matrilysin. Finally, the BIA was successfully applied for plasma and urine samples from colorectal cancer patients revealing both active and pro-matrilysin in plasma and mainly the active form in urine. In conclusion, this study introduces an assay to determine total as well as active matrilysin validated for biological samples from patients with gastrointestinal (pre)malignancies. We propose that measurement of matrilysin activity could be clinically most relevant because high activity represents up-regulated matrilysin expression as well as an enhanced matrilysin activating cascade. Department of Gastroenterology-Hepatology, Leiden University Medical Centre (L. J. A. C. Hawinkels, H. W. Verspaget, M. van den Berg, C. F. M. Sier); TNO Quality of Life, BioSciences (R. Hanemaaijer), Leiden, The Netherlands