Comparison of Mouse and Human Colon Tumors with Regard

Since human colorectal tumors are insensitive to most chemotherapeutic agents, there is a need for the discovery of new drugs that would show activity against this disease. In an attempt to better appreciate the relevance of a widely used mouse colon tumor (colon adenocarcinoma Co38) as a screening model for human colorectal tumors, we compared the main phase I and phase II drug-metabolizing enzyme systems in both tumoral and nontumoral colon tissues. The following enzymes were assayed by Western blot: cytochromes P-450 (IAI/A2, 2B1/B2, 2C, 2El, and 3A), epoxide hydrolase, and glutathione-S-transferases (GST-a, -g, and -a-). The activities of the following enzymes or cofactors were determined by spectrophotometric or fluorometric assays: total cytochrome P-450, l-chloro-2,4-dinitrobenzene-GST, selenium-independent glutathione peroxidase, 3,4-dichloronitrobenzene-GST, ethacrynic acid-GST, total glutathione, epoxide hydrolase, UDP-glucuronosyltransferase, ~/-glucuronidase, sulfotransferase, and sulfatase. Results obtained by Western blot showed that mouse colon adenocarcinoma Co38 did not express any of the probed cytochromes P-450, whereas human colorectal tumors expressed only low levels of cytochrome P-450 3A. GST-a and GST-w were detected in all tumoral and nontumoral tissues of both species. The neutral GST-9 was expressed in all murine tissues investigated and was found to be polymorphic in human tissues. For human peritumoral and tumoral colorectal tissues there was no significant difference between GST isoenzyme levels, whereas mouse colon adenocarcinoma Co38 had a lower expression of GST-~t and GST-w, compared to normal mouse colon. Enzymatic activities for glutathione peroxidase, 3,4-dichloronitrobenzene-GST, and ethacrynic acid-GST confirmed the Western blot results for GST-a, GST-~, and GST-w, respectively. Total GSH levels were similar between murine and human tumors but were 3-fold higher in human tumors than in peritumoral tissues, whereas they were 7-fold lower in mouse colon tumor Co38, compared to normal mouse colon. Epoxide hydrolase was not expressed in either mouse colon adenocarcinoma Co38 or normal mouse colon tissues, whereas it was expressed in human colon peritumoral and tumoral tissues at similar levels. No significant difference was observed between human tumors and peritumoral tissues for UDP-glucuronosyltransferase, ~-glucuronidase, sulfotransferase, and sulfatase. For murine colon tissues, the conjugation pathways (UDP-glucuronosyltransferase and sulfotransferase) were lower in colon adenocarcinoma Co38, whereas the converse was observed for the corresponding hydrolytic enzymes (~/-glucuronidase and sulfatase). In conclusion, similarities were noted between mouse colon adenocarcinoma Co38 and human colon tumors with regard to the absence of most forms of cytochromes P-450 except cytochrome P-450 3A, which was detected in human tumors only. However, for most phase II drug-metabolizing pathways the mouse model was qualitatively and quantitatively different from human colon tissues. These noteworthy interspecies differences may have implications with regard to drug-screening methodologies and preclinical Received 4/8/92; accepted 9/22/92. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 This work was supported by the Institut National de la Sant6 et de la Recherche M6dicale (INSERM), the Centre National de la Recherche Scientifique (CNRS), the Fondation pour la Recherche M6dicale, and the Association pour la Recherche sur le Cancer (Grants 6604 and 6917). 2 To whom requests for reprints should be addressed, at Institut GustaveRoussy, Laboratoire de pharmacologie clinique, Pavilion de recherche 2, 94805 Villejuif Cedex, France. evaluation of candidate anticancer drugs for the chemotherapy of human colorectal tumors. I N T R O D U C T I O N Unresectable or metastatic colorectal cancer is most often insensitive to chemotherapy and remains the second leading cause of all cancer deaths in the United States and in the European Communi ty (1, 2). 5-Fluorouracil, the reference drug for this disease, produces a response rate of approximately 15%, and al though its combinat ion with leucovorin increases the response rate its impact on long term survival seems unchanged (3-5). Clearly, there is a need for novel anticancer drugs that would change the survival in this disease. We believe that one of the reasons for the paucity of chemotherapeutic agents active against colon tumors could be that screening strategies did not use colon tumors in pr imary screening until the mid-1980s. Screening practices are now changing and rely on the use of either mouse or human tumor models of specific organ systems. Some of the mouse colon tumor models used have been characterized in terms of biological properties and drug responsiveness, and their resemblance to human colon tumors has strengthened their use as models for human disease (6, 7). Drug-metabolizing enzymes play an impor tant role in determining the susceptibility of organs or tissues to the toxic effects of drugs or other xenobiotics, and they may also influence tumor response to anticancer agents in vivo (8, 9). Although progress has recently been made in characterizing some enzymes involved in drug metabolism in human colorectal tumors (9-12), little informat ion is presently available concerning many other impor tant enzymatic systems of human tumors and, moreover, no information is available concerning the most frequently used mouse colon tumor models. It was, therefore, of interest to assess in these tumors the main phase I (P-450s 3) and phase II enzyme systems involved in drug metabolism (GST isoenzymes, EH, UDP-glucuronosyltransferase, and sulfotransferase), which correspond to functionalization and conjugation reactions, respectively. Moreover, in addit ion to GSH, the hydrolytic enzymes ~/-glucuronidase and sulfatase were assayed. In this report, we present a comparison of the main xenobiotic-metabolizing enzyme systems found in human colorectal tumors and in the widely used mouse co lon adenocarcinoma Co38 model (13). We also compared the human colorectal tumor to its corresponding peri tumoral tissue and compared the mouse colon adenocarcinoma Co38 to normal mouse colon. Our results showed similarities between colon tumors of the two species for some enzymatic pathways but also pointed to noteworthy qualitative and quantitative interspecies differ3 The abbreviations used are: P-450, cytochrome P-450; GST, glutathione-Stransferase; CDNB, l-chloro-2,4-dinitrobenzene; GPX, selenium-independent glutathione peroxidase; DCNB, 3,4-dichloronitrobenzene; EA, ethacrynic acid; GSH, glutathione; EH, epoxide hydrolase.