Mucin-depleted foci show strong activation of inflammatory markers in 1,2-dimethylhydrazine-induced carcinogenesis and are promoted by the inflammatory agent sodium dextran sulfate Angelo Pietro Femia, Piero Dolara, Cristina Luceri, Maddalena Salvadori and Giovanna Caderni * Department of Pharmacology, University of Florence, Florence, Italy Mucin-depleted foci (MDF), formed by dysplastic crypts devoid of mucins, have been identified in the colon of carcinogen-treated rodents and in humans at high risk for colon cancer. The lack of the protective layer of mucus may cause inflammation which has been linked to colon carcinogenesis, therefore, the expression of markers such as cyclooxygenase-2 (COX-2), inducible nitric oxide synthase (i-NOS) and macrophage infiltration was studied with immunohistochemistry (IH) in MDF harvested from F344 rats treated with the colon carcinogen 1,2-dimethylhydrazine (DMH). The same determinations were performed in aberrant crypt foci (ACF) and, at a later time point, in tumours. A dramatic increase in COX-2, i-NOS and macrophage infiltration was observed in MDF but ACF showed a moderate increase compared with the paired normal mucosa. Tumours were positive for all the markers. RT-PCR experiments demonstrated that i-NOS RNA expression was increased in a set of MDF confirming the results obtained with immunohistochemistry. In an inflammation-cancer experi- mental model [mice treated with azoxymethane (AOM) and dex- tran sodium sulphate (DSS)], we observed that DSS-induced inflammation promoted MDF in a dose-dependent manner, whereas ACF were not affected. In conclusion, we report here for the first time a strong activation of the inflammatory process in MDF, which may contribute to the further progression of MDF to tumours. ' 2009 UICC Key words: colon carcinogenesis; preneoplastic lesions; mucin- depleted foci; COX-2; i-NOS Mucin-depleted foci (MDF), formed by dysplastic crypts devoid of mucins, have been identified by our group in the colon of carcinogen-treated rodents and, recently, also in humans at high risk of colon cancer. 1,2 MDF harbour mutations in genes affecting colon carcinogenesis and, like colonic tumours, show Wnt signal- ling activation, features suggesting that these lesions are precan- cerous. 3,4 Recently, we also reported that the expression of MUC2, the most abundant mucin in the colon, is dramatically reduced in MDF. 5 This last result, led us to speculate that the focal loss of the protective mucous layer might activate local inflammation, since MDF would be more exposed to noxious agents present in the co- lonic lumen. This local inflammation might drive carcinogenesis to more advanced stages. Inflammation has been linked in fact to increased colon cancer risk as demonstrated by higher cyclooxy- genase 2 (COX-2) expression in colon tumours and by reduced carcinogenesis in humans and rodents given anti-inflammatory drugs. 6–9 Interestingly, it has also been demonstrated that Muc2- deficient mice are prone to intestinal carcinogenesis and that the introduction of a mutant Muc2 allele into Apc- mutated mice greatly increases the transformation induced by Apc mutation. 10,11 Given these considerations and the fact that MDF represent one of the earliest stages of colon carcinogenesis, we thought it inter- esting to study inflammation in MDF. Therefore, we studied with immunohistochemistry (IH) the expression of COX-2, inducible nitric oxide synthase (i-NOS), NO-tyrosine (NO-tyr) and macro- phage infiltration in MDF induced in rats by the colon carcinogen 1,2 dimethylhydrazine (DMH). To contextualize the MDF data within the various steps of colon carcinogenesis, the same deter- minations were performed in tumours harvested at a later time as well in ACF, which are well-characterized preneoplastic lesions. 12 We also set up conditions to isolate RNA from single MDF and determined i-NOS expression with RT-PCR experiments in a small separate set of MDF. To verify whether inflammation is an important cofactor in MDF progression, we exploited a mouse model of colitis-induced carcinogenesis in which a subcarcinogen dose of azoxymethane (AOM) is followed by a frank inflammatory stimulus (administra- tion of dextran sodium sulphate, DSS). 13 Although it has been reported that DSS promotes dysplasia after a subcarcinogenic dose of AOM 14,15 and that combined treatment of a carcinogenic dose of AOM and DSS induces MDF in rats, 16 it still remains to be demonstrated whether a subcarcinogenic AOM dose followed by DSS induces MDF. Therefore, we evaluated the presence of MDF, ACF and tumours induced by a subcarcinogenic dosage of AOM followed by DSS. Material and methods Carcinogenesis induction in rats Male F344 rats, 4–5 weeks old, were obtained from Nossan (Milan, Italy) and fed an AIN-76 based diet modified to contain high-fat content (23% corn oil w/w) as described previously. 17 At 6–7 weeks of age rats were treated twice, 1 week apart, with sub- cutaneous injections of DMH (150 mg/kg 3 2). Fifteen weeks af- ter carcinogen induction, a group of rats (n 5 9) intended for IH experiments was killed, whereas one rat intended for RT-PCR experiments was killed 2 weeks later. No macroscopic tumours were detected while killing. The colons intended for IH experi- ments were excised, washed with saline, opened longitudinally, fixed with 70% ethanol and then stained with Alcian Blue-Neutral Red for MDF and ACF visualization as previously described. 4 The colon intended for the RT-PCR experiment was processed as described in the next section. Twenty-eight weeks after carcinogen induction, the remaining group of rats (n 5 10) was killed to har- vest tumours. Colons were opened longitudinally while killing, and those lesions which were clearly visible with the naked eye were considered tumours, following subsequent histopathological examination as described. 5 Tumours were fixed in formalin and embedded in paraffin after no more than 24 hr. Expression of COX-2, i-NOS, NO-tyr and macrophages in ACF, MDF and tumours induced in rats by DMH-Immunohistochemistry experiments Expression of COX-2, i-NOS, NO-tyr and macrophages was determined in ACF, MDF and tumours. ACF and MDF were marked as described, 5 positioning the permanent ink laterally to the lesions as not to interfere with IH procedures. Longitudinal sections (4 lm) were mounted on electrostatic-treated slides Grant sponsor: American Institute for Cancer Research; Grant number: 05A019-REV; Grant sponsor: Fondo Ateneo ex-60% of the University of FlorenceGrant sponsor: ITT (Istituto Toscano Tumori). *Correspondence to: Department of Pharmacology, University of Florence, 6 Viale Pieraccini, 50139 Florence, Italy. Fax: 139-55-427- 1280. E-mail: giovanna.caderni@unifi.it Received 22 January 2009; Accepted after revision 2 March 2009 DOI 10.1002/ijc.24417 Published online 11 March 2009 in Wiley InterScience (www.interscience. wiley.com). Int. J. Cancer: 125, 541–547 (2009) ' 2009 UICC Publication of the International Union Against Cancer