Aberrant transforming growth factor-beta signaling in azoxymethane-induced mouse colon tumors

Alterations in the transforming growth factor-beta (TGF-beta) pathway are implicated in the pathogenesis of colorectal cancer. We hypothesize that alterations in the TGF-beta pathway contribute to differential sensitivity of mice to the colon carcinogen azoxymethane (AOM). A/J (sensitive) and AKR/J (resistant) mice were injected intraperitoneally with AOM (10 mg/kg of body weight once a week for 6 wk). Twenty-four weeks after AOM exposure, mutational analysis of TGF-beta type II receptor (TbetaR-II) from normal colons and from tumors showed no AOM-induced alterations. A significant decrease (1.5-fold, P < 0.05) in TbetaR-II mRNA levels, however, was found in A/J tumors with the RNase protection assay. Immunofluorescence of TbetaR-II showed marked loss of staining in A/J tumors. The RNase protection assay and sequence analysis of the downstream signaling molecule Smad3 revealed no carcinogen-induced alterations in either strain. To gain further insight into the functionality of the pathway, expression of TGF-beta, TGF-beta type I receptor (TbetaR-I), and several downstream targets of TGF-beta signaling, including Smad7, c-myc, and p15, was examined. Although no alterations in TGF-beta, TbetaR-I, or Smad7 were found in tumors, a significant increase in c-myc expression (2.5-fold, P < 0.05 ) and a significant decrease in p15 expression (4.5-fold, P < 0.05 ) were noted. Concomitant repression of TbetaR-II and overexpression of c-myc may render epithelial cells insensitive to TGF-beta-mediated growth arrest, a possibility that also is suggested by this model. The significant decrease in p15 expression in tumors provides additional evidence that TGF-beta signaling may be markedly attenuated during colon tumorigenesis.     

Enhanced colon carcinogenesis induced by azoxymethane in min mice occurs via a mechanism independent of beta-catenin mutation

The multiple intestinal neoplasia (min) mouse is a well-established cancer model in which loss of a single copy of the APC protein predisposes mice to the development of numerous tumors in the intestine. We have developed a novel variation of the min mouse model by using azoxymethane (AOM) to cause an increase in tumor incidence, number and size. Thus, treatment of min mice with AOM resulted in 2.6-, 6.3- and 5.9-fold increases in overall tumor incidence, multiplicity and size, respectively, when compared to wild type C57BL/6J mice treated with AOM. Furthermore, adenocarcinomas of the colon, which are otherwise relatively rare in min mice, increased in incidence (P<0.004), multiplicity (P<0.005), and size (P<0.02) in the AOM-treated min mice when compared to control untreated min mice. Of these adenocarcinomas, the number of poorly plus moderately differentiated adenocarcinomas was also significantly higher in the AOM-treated min mice (P<0.008). Thirty-seven histopathologically verified colon tumors (eight adenomas, five carcinoma in situ and 24 adenocarcinomas) induced in min mice and in C57BL/6J mice after treatment with or without AOM were analyzed for mutations in the beta-catenin gene or de novo mutations in the Apc gene. No mutations in the beta-catenin gene were found in any of colon tumors in min mice with or without treatment with AOM. However, mutations in either the beta-catenin gene or the Apc gene were found in tumors induced in C57BL/6J mice by AOM. These results suggest that mutations in the beta-catenin gene are less contributory to tumor development in min mice, as is the case in familial adenomatous polyposis (FAP) in humans. However, de novo mutations in either the Apc or beta-catenin gene can play a role in tumor development in C57BL/6J mice treated with AOM. The differences in mutation status between min and C57BL/6J mice may indicate different genetic pathways for developing colon tumors. These two experimental systems may, therefore, be useful animal models of human colon carcinomas in patients with FAP and in patients with sporadic colon carcinomas.     

Frequent mutations of the beta-catenin gene in mouse colon tumors induced by azoxymethane

The beta-catenin gene is frequently mutated at codons 33, 41 and 45 of the glycogen synthase kinase-3beta phosphorylation motif in human colon cancers in patients without APC mutations. Frequent mutations at codons 32 and 34, as well as 33 and 41, have been detected in rat colon tumors induced by azoxymethane (AOM), with the second G of CTGGA sequences being considered as a mutational hot-spot. In the present study, exon 3 of the beta-catenin gene in mouse colon tumors induced by AOM was amplified by PCR and mutations were detected by the single strand conformation polymorphism method, restriction enzyme fragment length polymorphism and direct sequencing. All 10 colon tumors tested were found to have beta-catenin mutations, four in codon 34, three in codon 33, two in codon 41 and one in codon 37, nine being G:C-->A:T transitions. However, no mutations were found in codon 32 of the mouse beta-catenin gene. On immmunostaining, beta-catenin was observed in the cytoplasm and nucleus of the tumor cells. The cytoplasmic staining was homogeneous, while both homogeneous and heterogeneous patterns were noted for the nuclei. Highly frequent mutations of the beta-catenin gene in AOM-induced mouse colon tumors suggest that consequent alterations in the stability and localization of the protein may play an important role in this colon carcinogenesis model.     

Coordinate regulation of cyclooxygenase-2 and TGF-beta1 in replication error-positive colon cancer and azoxymethane-induced rat colonic tumors

Evidence is accumulating which indicates that cyclooxygenase-2 (COX-2) is involved in the pathogenesis of colorectal cancer. We evaluated the expression of COX-2 in replication error-positive (RER) colon cancers, colon cancers metastatic to liver and azoxymethane (AOM)-induced rat colonic tumors. Immunohistochemistry showed that COX-2 was low to undetectable in normal human mucosa, but abundant in the RER adenocarcinomas we examined. COX-2 immunoreactivity in metastatic colon cancers was less abundant, but clearly detectable. In the colon of AOM-treated rats, COX-2 protein was not detectable in normal mucosa, but present in most of the epithelial cells comprising the tumors. The TGF-beta1 staining pattern in these human and rat tumors was similar to that observed for COX-2. The role of TGF-beta in RER adenocarcinomas is complex because of the increased mutation rate of TGF-beta type II receptors. Northern analysis showed abundant TGF-beta1 mRNA in AOM-induced tumors, but not in paired mucosa. TGF-beta1 induced the expression of COX-2 mRNA and protein in intestinal epithelial cells (IEC-6). Chronic TGF-beta1 treatment caused a TGF-beta-dependent overexpression of COX-2 in rat intestinal epithelial cells (RIE-1). TGF-beta1 may regulate COX-2 expression during the colonic adenoma to carcinoma sequence.     

Preliminary analysis of azoxymethane induced colon tumors in inbred mice commonly used as transgenic/knockout progenitors

Azoxymethane (AOM) is a colon carcinogen that is used to study the pathogenesis of sporadic colorectal cancer. We have evaluated differential susceptibility to AOM in inbred mice used as progenitors of recombinant/transgenic lines. In experiment 1, male FVB/N, 129/SvJ, C57Bl/6J mice were treated i.p. with 10 mg/kg AOM once per week for 4 weeks and sacrificed after 20 weeks. Only AOM-treated FVB/N mice developed tumors (3.6 tumors/mouse) in distal colon. In experiment 2, A/J, AKR/J, Balb/CJ mice were treated with AOM for 6 weeks and sacrificed after 24 weeks. AOM-treated A/J and Balb/CJ mice developed 9.2 and 1 tumor/mouse, respectively. Despite these differences, tumors had similar morphology regardless of strain. Immunohistochemistry with beta-catenin resulted in marked nuclear and cytoplasmic staining of tumor cells in FVB/N. However, fainter and heterogeneous beta-catenin staining was observed in A/J tumors, suggesting distinct pathways of tumorigenesis in different strains. Irrespective of cytological features of malignancy, tumor cells rarely breached the muscularis mucosa and showed no evidence of distant metastasis. Lack of invasiveness and metastasis in even the most sensitive strains provides a model system for studying the potential role of 'metastasis genes' in imparting a malignant phenotype.     

Inverse association between phospholipase A2 and COX-2 expression during mouse colon tumorigenesis

Cytosolic phospholipase A(2) (cPLA(2)) releases arachidonic acid (AA) from intracellular phospholipids. We evaluated the status of cPLA(2) in azoxymethane (AOM)-induced mouse colon tumors. Despite increased expression of cyclooxygenase 2 (3.7-fold) and PGE(2) (3.4-fold) production in tumors, cPLA(2) mRNA levels and enzyme activity were significantly reduced (3.6- and 3-fold, respectively). Reduced levels of cPLA(2) were also observed in pre-neoplastic aberrant crypt foci (ACF), a distinct morphological alteration that represents an early stage in the pathogenesis of colon tumors. Furthermore, the reciprocal expression patterns of these two genes were found to occur in human colorectal cancers (CRC). Examination of the activity of the secretory phospholipases A(2) (sPLA(2)) and expression of the groups V and X sPLA(2)s revealed no compensatory increase in tumor tissue. As cPLA(2) has been shown to be involved in TNF-alpha-induced apoptosis in certain cell types, and TNF-alpha expression is significantly enhanced in AOM-induced tumors (2.8-fold), we examined the role of cPLA(2) in TNF-alpha-induced apoptosis of cultured mouse colonocytes (YAMC). The specific cPLA(2) inhibitor, AACOCF(3) (arachidonoyl trifluoromethyl ketone), was able to protect colonocytes from TNF-alpha-induced apoptosis in vitro. In summary, our data demonstrate an inverse relationship between COX-2 and cPLA(2) expression in both AOM-induced mouse colon tumors and human CRC and suggest that down regulation of cPLA(2) may attenuate TNF-alpha mediated apoptosis during tumorigenesis and facilitate tumor progression.     

Hemizygous mice for the angiotensin II type 2 receptor gene have attenuated susceptibility to azoxymethane-induced colon tumorigenesis

Evidence suggests that the use of angiotensin-converting enzyme inhibitors potentially reduces the risk of cancer, though the mechanism is unclear. To clarify a potential involvement of angiotensin II (Ang II) signaling in cancer risk, we have examined the effect of Ang II receptor deficiency on azoxymethane (AOM)-induced colon tumorigenesis. Male Ang II type 2 receptor gene-disrupted (AT(2)-null) mice with a 129/Ola and C57BL/6J genetic background, AT(2)-null mice with an SWR/J genetic background, and their corresponding control wild type mice were treated once a week with AOM (10 mg/kg, i.p., 4 consecutive weeks) or saline vehicle. All mice were killed 23-26 weeks after the initial injection of AOM, and tumor burdens were examined. AOM treatment caused the development of colon tumors in all wild type control mice regardless of genetic background (100% tumor prevalence), but only one tumor was present in AT(2)-null mice with a 129/Ola and C57BL/6J genetic background (11.1% tumor prevalence). Although the introduction of the AOMsusceptible SWR/J genetic background induced AOM susceptibility in AT(2) null mice, the tumor multiplicity (6.3) and tumor size (19.8 +/- 3.0 mm(3)) were significantly smaller than those in wild type mice (multiplicity, 12.0 and size, 36.8 +/- 3.2 mm(3)). AOM efficiently downregulated cytochrome P450 2E1 (CYP2E1) in the liver of wild type mice significantly more than in AT(2)-null mice. The levels of DNA methyl adducts formed in wild type mouse colon epithelium by AOM treatment were also significantly higher than in AT(2)-null mice. These results imply that the AT(2) receptor functions to augment AOM-induced downregulation of CYP2E1 expression in the liver, and thus increases AOM-induced tumorigenesis in the colon. The AT(2) receptor function in the liver may be a potential determinant of tumor susceptibility in chemical carcinogen-induced colon tumorigenesis.     

Transforming growth factor beta receptor type II inactivation induces the malignant transformation of intestinal neoplasms initiated by Apc mutation

The transforming growth factor-beta (TGF-beta) signaling pathway is a tumor-suppressor pathway that is commonly inactivated in colon cancer. TGF-beta is a secreted ligand that mediates its effects through a transmembrane heteromeric receptor complex, which consists of type I (TGFBR1) and type II subunits (TGFBR2). Approximately 30% of colon cancers carry TGFBR2 mutations, demonstrating that it is a common target for mutational inactivation in this cancer. To assess the functional role of TGFBR2 inactivation in the multistep progression sequence of colon cancer, we generated a mouse model that recapitulates two common genetic events observed in human colon cancer by mating Apc(1638N/wt) mice with mice that are null for Tgfbr2 in the intestinal epithelium, Villin-Cre;Tgfbr2(E2flx/E2flx) mice. In this model, we observed a dramatic increase in the number of intestinal adenocarcinomas in the Apc(1638N/wt);Villin-Cre;Tgfbr2(E2flx/E2flx) mice (called Apc(1638N/wt);Tgfbr2(IEKO)) compared with those mice with intact Tgfbr2 (Apc(1638N/wt);Tgfbr2(E2flx/E2flx)). Additionally, in vitro analyses of epithelial tumor cells derived from the Apc(1638N/wt);Tgfbr2(IEKO) mice showed enhanced expression and activity of matrix metalloproteinase MMP-2 and MMP-9, as well as increased TGF-beta1 secretion in the conditioned medium. Similarly, primary tumor tissues from the Apc(1638N/wt);Tgfbr2(IEKO) mice also showed elevated amounts of TGF-beta1 as well as higher MMP-2 activity in comparison with Apc(1638N/wt);Tgfbr2(E2flx/E2flx)-derived tumors. Thus, loss of TGFBR2 in intestinal epithelial cells promotes the invasion and malignant transformation of tumors initiated by Apc mutation, providing evidence that Wnt signaling deregulation and TGF-beta signaling inactivation cooperate to drive the initiation and progression, respectively, of intestinal cancers in vivo.     

Testosterone strongly enhances azoxymethane/dextran sulfate sodium-induced colorectal cancer development in C57BL/6 mice

Colorectal cancer (CRC) is known to occur more frequently in males than in females, with sex hormones reportedly influencing the development. The purpose of the study was to investigate whether orchiectomy in C57BL/6 male mice reduces colorectal tumorigenesis and whether testosterone administration increases tumorigenesis after orchiectomy in an azoxymethane (AOM)/dextran sulfate sodium (DSS) mouse model. Clinical symptoms, including colitis and tumor incidence, were evaluated in the absence or presence of testosterone in AOM/DSS-treated male, as well as orchiectomized (ORX) male and female mice. The levels of serum testosterone and colonic myeloperoxidase, interleukin (IL)-1β, and IL-6 were measured by ELISA. Target mRNA expression was assessed by quantitative real-time PCR. Orchiectomy significantly diminished the AOM/DSS-induced colitis indices, including disease activity index, colon shortening, and histological severity at week 2, and decreased tumor numbers and incidence rates in the distal part of the colon increased following AOM/DSS administration at week 13; this reduction was reversed by testosterone supplementation. Furthermore, it was confirmed that the ELISA level (MPO and IL-1β) and the mRNA expression of the inflammatory mediators (COX-2 and iNOS) were maintained at high levels in the tumors of the testosterone-treated group compared with AOM/DSS groups. Interestingly, both endogenous and exogenous testosterone administrations were associated with tumor development (> 2 mm in size) and submucosal invasive cancer. Based on multivariate logistic regression analysis, testosterone was identified as a reasonable hazard factor for the progression of submucosal invasive cancer of the distal colon. In conclusion, endogenous and exogenous testosterone presented a stimulating effect on AOM/DSS-induced colitis and carcinogenicity.     

Strain-specific homeostatic responses during early stages of Azoxymethane-induced colon tumorigenesis in mice

Identifying molecular changes that predict the risk for developing colon cancer is critical for designing effective prevention strategies. In the present study, we determined early-stage molecular alterations within the colonic epithelium of A/J and AKR/J mice that are sensitive and resistant to Azoxymethane (AOM)-initiated tumor development, respectively. Six week-old male mice were injected intraperitoneally with AOM (10 mg/kg body weight) once a week for six weeks. One week after the last injection, distal colons from both strains were analyzed for cell proliferation using a proliferating cell nuclear antigen (PCNA) assay. Unlike AKR/J, a significant increase (2.5-fold, p<0.05) in the number of PCNA-positive cells within the upper third of the crypt compartment was observed in the A/J colons. This proliferative response was associated with a sizeable increase in the levels of c-myc mRNA, quantified by RNase protection assay. cDNA sequencing, protein expression and localization of beta-catenin, an upstream activator of c-myc, however, showed no aberrant changes within AOM-exposed A/J colons. Interestingly, TdT-mediated dUTP nick-end labeling assay revealed a significant increase (4-fold) in the number of apoptotic colonocytes in A/J mice following AOM treatment. Consistent with this finding, a modest increase in the expression of pro-apoptotic Bak was limited to the sensitive A/J colons. In summary, the current study suggests that a significant alteration in the rate of cell turnover in the normal appearing colonic mucosa, as observed in susceptible A/J mice, may be one of the earliest events predisposing the colon to neoplastic growth.     

Indole-3-carbinol inhibits the growth of human colon carcinoma cells but enhances the tumor multiplicity and volume of azoxymethane-induced rat colon carcinogenesis

Indole-3-carbinol (I3C) is a naturally occurring phytochemical which exerts a broad range of biological activities. The purpose of this study was to examine the effects of I3C on colon carcinogenesis, cell proliferation, cell-cycle progression and apoptosis, and on the levels of expression of several cell-cycle control molecules. We used a long-term rat model by using azoxymethane (AOM) to induce tumors (adenomas and adenocarcinomas) in the colon. In the present study, we found that after AOM injection, the treatment of male F344 rats with 0.01 and 0.05% I3C caused a significant increase in the tumor multiplicity of adenocarcinomas by 2.2- (P<0.05 for 0.01% I3C) and 2.1-fold (P<0.0002 for 0.05% I3C) respectively, when compared to the control rats. In addition, the tumor multiplicity of adenoma plus adenocarcinoma and the volume of adenocarcinoma were also increased by 2.0- (P<0.00001) and 2.1-fold (P<0.05) respectively, compared to the control. I3C significantly increased the proliferating cell nuclear antigen labeling index (PCNA LI) (P<0.008) and decreased the apoptotic index (P<0.05) of the colon adenocarcinoma. In contrast, in HCT 116 and HT29 human colon carcinoma cells, I3C inhibited growth and induced G1-phase cell-cycle arrest and apoptosis. Furthermore, I3C caused approximately a 2- to 4-fold increase in the cellular levels of p27KIP1 and p21CIP1 mRNA. These results suggest that I3C inhibits the growth of human colon carcinoma cells, at least in part, by inducing p27KIP1 and p21CIP1-mediated G1 cell-cycle arrest but dietary I3C promotes AOM-induced rat colon carcinogenesis by inhibiting the apoptosis of colon tumors. Therefore, the present study may provide further evidence for the ambivalent modulatory activity of I3C and this information may be useful when including I3C in cancer chemoprevention and/or extensive clinical therapy trials.