Toxic effects of non-steroidal anti-inflammatory drugs in a human intestinal epithelial cell line (HCT-8), as assessed by the MTT and neutral red assays

Abstract

The human ileocaecal adenocarcinoma cell line HCT-8 was characterized for its potential as an in vitro organ-specific model for gastro-intestinal toxicity. HCT-8 cells showed typical epithelial cell morphology, with microvilli and intercellular junctional complexes, and formed domes, consistent with transepithelial fluid secretion. The cells express three intestinal brush-border enzyme activities (alkaline phosphatase, leucine aminopeptidase and α-glucosidase), and adenylate cyclase can be stimulated with vasoactive intestinal peptide. The toxicity of eight non-steroidal anti-inflammatory drugs (NSAID; indomethacin, mefenamic acid, ketoprofen, ibuprofen, sulindac, aspirin, phenylbutazone and naproxen) were assessed using the MTT and neutral red uptake assays. The MTT assay was consistently a more sensitive measure of NSAID-induced toxicity, which suggests that perturbation of mitochondrial function may be an early event in NSAID-induced cellular damage. Comparing the rankings observed in acute studies in the rat in vivo with those observed with HCT-8 cells, there are some general agreements. Indomethacin, a potent ulcerogen in vivo, was consistently among the most toxic in vitro, while aspirin and phenylbutazone have comparatively low rankings in vitro and in vivo. In man, with chronic administration, indomethacin is again ranked as a potent ulcerogen, as is aspirin, in contrast to the in vitro data with HCT-8. Therefore, NSAID-induced toxicity in HCT-8 cells assessed by the MTT or neutral red assays, can only partially predict toxicity in vivo, which suggests that local gastro-intestinal environmental factors, such as luminal acidity, may play a role in vivo. The ability of HCT-8 cells to reconstitute intact epithelial layers, thereby allowing such environmental factors to be mimicked, allows further development of these cells as a model for the in vitro prediction of in vivo gastro-intestinal toxicity.