Differences in the levels and pattern of DNA-adduct labelling in human cell lines MCL-5 and CCRF, proficient or deficient in carcinogen-metabolism, treated in vitro with bile from familial adenomatous polyposis patients and from unaffected controls.

In patients with familial adenomatous polyposis (FAP), duodenal adenomas cluster around the ampulla and their distribution closely resembles mucosal exposure to bile, suggesting a role for bile in their development. Previous studies using 32P-postlabeling to detect DNA adducts, have provided evidence to support this hypothesis. We have now investigated the role of metabolic activation in influencing the levels and patterns of adduct formation by incubating precolectomy gallbladder bile from FAP patients and bile from unaffected controls with human lymphoblastoid cell lines that are metabolically proficient (MCL-5), or deficient (CCRF). 32P-Postlabeling assays showed that MCL-5 cells (genetically engineered to express five human cytochromes P450 and microsomal epoxide hydrolase) formed characteristic adduct spots with benz[a]pyrene, benzo[g]chrysene, 7,12-dimethylbenz[a]anthracene, benzidine, sterigmatocystin and 3-methylcholanthrene, whereas CCRF cells did not. Accordingly, we assayed the ability of bile from FAP patients and controls to form DNA adducts in MCl-5 and in CCRF cells. Relative adduct labelling (RAL) in MCL-5 cells treated with FAP bile (12 patients, median 10, range 1-74) was significantly higher than in cells treated with control bile (12 patients, median 4, range 0-9; P = 0.0007) as was RAL for the two major adduct spots. These two major adduct spots were not observed when bile was incubated with CCRF cells. The adduct spots in CCRF DNA appeared in positions similar to some of the minor adduct spots produced by bile in MCL-5 DNA and to some of the adduct spots seen previously when bile was incubated with salmon sperm DNA in vitro. RAL for CCRF cells incubated with FAP bile (seven patients, median 23.0, range 0-49) was significantly higher than in cells treated with control bile (seven patients, median 2.0, range 0-26; P = 0.0034). These results indicate that the bile obtained from FAP and control patients contains adduct-forming substances, some of which are direct acting and some of which require metabolic activation. In both cell lines, FAP bile produced significantly higher adduct labelling than control bile, adding to the evidence that bile can induce DNA damage in vitro and plays a role in neoplastic development in the FAP foregut.