Carcinogenesis vol.30 no.5 pp.799–807, 2009 doi:10.1093/carcin/bgn246 Advance Access publication November 20, 2008 Aldose reductase deficiency in mice prevents azoxymethane-induced colonic preneoplastic aberrant crypt foci formation Ravinder Tammali, Aramati B.M.Reddy, Kota V.Ramana, J.Mark Petrash1,2 and Satish K.Srivastava To whom correspondence should be addressed. Tel: þ1 409 772 3926; Fax: þ1 409 772 9679; Email: [email protected] Aldose reductase (AR; EC 220.127.116.11), an nicotinamide adenine dinucleotide phosphate-dependent aldo–keto reductase, has been shown to be involved in oxidative stress signaling initiated by inflammatory cytokines, chemokines and growth factors. Recently, we have shown that inhibition of this enzyme prevents the growth of colon cancer cells in vitro as well as in nude mice xenografts. Herein, we investigated the mediation of AR in the formation of colonic preneoplastic aberrant crypt foci (ACF) using azoxymethane (AOM)-induced colon cancer mice model. Male BALB/c mice were administrated with AOM without or with AR inhibitor, sorbinil and at the end of the protocol, all the mice were euthanized and colons were evaluated for ACF formation. Administration of sorbinil significantly lowered the number of AOM-induced ACF. Similarly, AR-null mice administered with AOM demonstrated significant resistance to ACF formation. Furthermore, inhibition of AR or knockout of AR gene in the mice significantly prevented AOM-induced expression of inducible nitric oxide synthase and cyclooxygenase-2 proteins as well as their messenger RNA. AR inhibition or knockdown also significantly decreased the phosphorylation of protein kinase C (PKC) b2 and nuclear factor kappa binding protein as well as expression of preneoplastic marker proteins such as cyclin D1 and b-catenin in mice colons. Our results suggest that AR mediates the formation of ACF in AOM-treated mice and thereby inhibition of AR could provide an effective chemopreventive approach for the treatment of colon cancer. Introduction Colon cancer is the third most common form of cancer and the second leading cause of cancer-related deaths in western countries, including the USA (1,2). Epidemiological and experimental studies indicate that colon cancer is usually mediated by dietary and environmental factors and is more pronounced in genetically predisposed subjects (3–6). Diet rich in fat, red meat, refined carbohydrates and animal proteins, along with low physical activity, obesity and hyperinsulinemia are risk factors, which can predispose colon cancer (3–7). Normal colonic mucosa is constantly renewed through controlled proliferation and differentiation of mitotically active cell populations in the lower third of colonic crypts (8). With normal maturation, the resultant cells lose their ability to proliferate and migrate to the surface, constantly replacing sloughed cells that have an approximate life span of 6 days (8). Conditions leading to increased proliferation and loss of differentiation capacity lead to the formation of preneoplastic lesions referred to as aberrant crypt foci (ACF) (9,10). ACF, abnormal crypts, Abbreviations: ACF, aberrant crypt foci; AOM, azoxymethane; AR, aldose reductase; ARKO, aldose reductase knockout; Cox-2, cyclooxygenase-2; GS-DHN, glutathionyl-1,4-dihydroxynonene; GS-HNE, glutathionyl-4-hydroxynonenal; HNE, 4-hydroxy-trans-2-nonenal; IHC, immunohistochemical; IL, interleukin; iNOS, inducible nitric oxide synthase; mRNA, messenger RNA; NF-jB, nuclear factor kappa binding protein; PCNA, proliferating cell nuclear antigen; PKC, protein kinase C; ROS, reactive oxygen species; WT, wild-type. Ó The Author 2008. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected] 799 Downloaded from http://carcin.oxfordjournals.org at UCHSC Denison Memorial Library Box A-003 on July 14, 2010 Department of Biochemistry and Molecular Biology, University of Texas Medical Branch, Galveston, TX 77555-0647, USA, 1Department of Ophthalmology and Visual Science and 2Department of Genetics, Washington University School of Medicine, St Louis, MO 63110, USA are two to three times larger than normal crypts and are characterized by hyperproliferation, increased and elevated size and expanded pericryptal zones (9). Pretlow et al. (11) have shown the presence of ACF in the colonic mucosa of colon cancer patients. Over period of time, they develop into adenomas and carcinomas. Development of ACF associates with mutations in the colon cancer biomarker genes such as APC and ras oncogene (12,13). The transition of normal epithelium to adenoma to carcinoma is associated with a variety of molecular and biochemical events such as genetic alterations, intestinal epithelial cell proliferation/differentiation and inflammation (13–15). The major initiators of carcinogenesis include (i) cells that suffered irreparable DNA damage due to increased free radicals that cause activation of specific nucleases and damage DNA, RNA, proteins and lipids; (ii) loss of extracellular stimulation that regulates expression of growth factors and their receptors as well as cell growth and (iii) autosomal dominant inheritance of oncogenes among the multiple family members (9,16–18). In addition, diet and environmental factors play an important role in predisposition of carcinogenesis in genetically inherited subjects (4). Furthermore, many chronic inflammatory diseases such as hepatitis, gastritis and ulcerative colitis are associated with an elevated risk of colon cancer (14,19,20). However, it is not clear how cancer is initiated in the setting of chronic inflammation, but accumulating evidence strongly supports the association between colon cancer and inflammation. Furthermore, upregulation of cytokines such as tumor necrosis factor-a, interleukin (IL)-6, growth factors such as insulin-like growth factor-II, hepatocyte growth factor, hepatocyte growth factor receptor, epidermal growth factor receptor, insulin growth factor receptor, vascular endothelial growth factor, fibroblast growth factor and platelet-derived growth factor and their receptors have been observed in various human colon cancer cells (17,19–21). Exposure of cells to inflammatory cytokines and growth factors triggers upregulation of prostaglandin E2 (PGE2) and nitric oxide (NO) production via cyclooxygenase-2 (Cox-2) and inducible nitric oxide synthase (iNOS), respectively (17,22–24). These local messenger molecules act further in an autocrine and paracrine fashion elevate reactive oxygen species (ROS). The ROS in turn activate various genes involved in the expansion of normal epithelial cells to dysplasia (precancer) and cancer (16). For example, the proinflammatory cytokines tumor necrosis factor-a, IL-1, IL-6 play an important role at initial stages of cell transformation and tumor development. Among the proinflammatory cytokines, tumor necrosis factor-a is recognized as a central mediator in the pathophysiology of chronic inflammatory bowel diseases such as Crohn’s and ulcerative colitis which cause increased risk of colon cancer (19–21). The murine azoxymethane (AOM) model of colon cancer shares many clinical, histopathological and molecular features with human tumors, but they have a low tendency to metastasize (11,25). AOMinduced form of ACF shows hyperproliferation of colonic mucosal cells and often mutated on K-ras and b-catenin genes and show microsatellite instability, like human tumors (11,25). This mechanism is associated with the overexpression of iNOS and Cox-2 enzymes with increased NO and PGE2 production (17,24). Compared with human tumors, accumulation of b-catenin in the nucleus suggests that Wnt/ b-catenin/Tcf pathway plays a major role in carcinogen-induced rat tumors (13,26). However, unlike human tumors, AOM-induced tumors rarely show mutation of APC gene and never mutate p53 gene (25). Our recent studies in human colon cancer Caco-2 cells, human lens epithelial cells , vascular endothelial cells and vascular smooth muscle cells suggest that the polyol pathway enzyme,aldose reductase (AR; AKR1B1 in human, AKR1B3 in mouse and AKR1B4 in rat), a member of aldo–keto reductase super family, is a regulator of ROS signals induced by growth factors, cytokines, chemokines and R.Tammali et al. Materials and methods Materials C57BL/6 ARþ/þ and AR/ mice were maintained in a pathogen-free barrier facility at Washington University School of Medicine, St Louis, MO. BALB/c mice (weight 25 g) were obtained from Jackson laboratory (Bar Harbor, ME) and housed in pathogen-free conditions with free access to food and water at the institutional animal care facility. The animals were maintained in accordance with the Guide for the Care and Use of Laboratory Animals published by the National Institutes of Health and in accordance with the Institute’s ‘Guideline of the Animal Care and Use Committee’. Mice were kept in suspended cages 10 cm above bedding trays with a 12 h light–dark cycle in the animal facility. Temperature and relative humidity were controlled at 21°C and 55%. All mice were acclimatized to the above conditions for 1 week with free access to standard laboratory rodent chow and drinking water until initiation of the experiment. AOM bought from the Sigma–Aldrich Chemical Company (St Louis, MO), and sorbinil was gift from Pfizer (Groton, CT). Antibodies against Cox-2, iNOS, cyclin D1 and phospho-protein kinase C (PKC) b2 were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). b-Catenin antibodies obtained from Cell Signaling (Danvers, MA). Mouse anti-rabbit glyceraldehyde-3-phosphate dehydrogenase antibodies were obtained from Research Diagnostics (Concord, MA). Polyclonal AR and protein adducts of HNE antibodies raised in rabbit. Other reagents for western blot analysis were obtained from Sigma-Aldrich. All other reagents used were of analytical grade. AOM-induced colon carcinogenesis and ACF analysis Approximately, 6-week-old mice were divided into three groups with six mice in each group. Mice in groups 2 and 3 were treated with AOM in sterile saline, at a dose of 10 mg/kg body wt intraperitoneally once a week, for 3 weeks. In group 3, mice were treated with AR inhibitor sorbinil (25 mg/kg body wt/day; intraperitoneally) for entire period, 10 weeks, after 24 h of first AOM injection. Mice in group 1 received equal volume of sterile saline. Similarly, ARKO mice groups also treated with AOM (10 mg/kg body wt; intraperitoneally) once a week for 3 weeks. After 5 and 10 weeks of first AOM injection, all mice were euthanized by exposure to CO2. The colons were removed, flushed with saline and opened from anus to cecum and fixed flat between two pieces of filter paper in 10% buffered formalin for 24 h. Colons were stained with 0.2% methylene blue dissolved in saline, and the numbers of ACF were counted under the microscope. Immunohistochemistry After microscopic evaluation of ACF, the colons were Swiss-rolled and embedded in paraffin. Immunohistochemical (IHC) analyses serial sections of 800 colon (5 lM) were performed as described elsewhere. Briefly, slides were warmed at 60°C for 1 h and deparaffinized in xylene and rehydrated in decreasing concentrations of ethanol. Antigen retrieval was performed by boiling slides in 10 mM sodium citrate (pH 6.0) for 10 min followed by peroxidase reaction was blocked with 3% H2O2. After that, sections were rinsed in phosphate-buffered saline twice and incubated with blocking buffer (2% bovine serum albumin, 0.1% Triton X-100 and 2% normal goat serum) for overnight at 4°C. The sections were incubated with primary antibodies against proliferating cell nuclear antigen (PCNA), Cox-2, iNOS, cyclin D1, AR, protein–HNE and b-catenin for 1 h at room temperature. Antigen–antibody binding was detected by using DakoCytomation LSABþSystem-HRP kit. Sections were examined by bright-field light microscopy (EPI-800 microscope; Nikon, Tokyo, Japan) and photographed with a camera (Nikon) fitted to the microscope. Photomicrographs of the stained sections were acquired using an EPI-800 microscope (bright-field) connected to a Nikon camera. Percent staining was determined by measuring positive immunoreactivity per unit area. Arrows represent the area for positive staining for an antigen. The intensity of antigen staining was quantified by digital image analysis and the values are represented as fold change versus control in arbitrary units. Determination of lipid peroxidation The levels of lipid peroxides (total a, b-unsaturated aldehydes) were estimated in colon tissues by using lipid peroxidation assay kit. Briefly, the colon tissue homogenized in phosphate-buffered saline in the presence of butylated hydroxytoluene (5 mM) and aldehydes were quantified colorimetrically using a lipid peroxidation kit (Bioxytech LPO-586TM) obtained from Oxford Biomedical Research, Oxford, MI, as per the supplier’s instructions. The determination is based on the reaction of the chromogenic reagent, methanesulfonic acid with a, b-unsaturated aldehydes such as HNE at 45°C. One molecule of aldehyde reacts with two molecules of reagent to yield a stable chromophore with maximal absorbance at 586 nm. Reverse transcription–polymerase chain reaction Total RNA from colon samples was isolated using the RNeasy kit as per supplier’s instructions (Qiagen, Valencia, CA). Equal aliquots of RNA (1.0 lg) isolated from each sample were reverse transcribed with Omniscript and Sensiscript reverse transcriptase One-Step Reverse Transcriptase–PCR system with HotStar Taq DNA polymerase at 55°C for 30 min followed by polymerase chain reaction amplification. The oligonucleotide primer sequences were as follows: 5#-GCATTGCCTCTGAATTCAACACAC-3# (sense) and 5#-GGACACCCCTTCACATTATTGCAG-3# (antisense) for Cox-2, 5#-CTGCAGGTCTTTGACGCTCG-3# (sense) and 5#-GTGGAACACAGGGGTGATGC-3# (antisense) for iNOS and 5#-GTGGGCCGCTCTAGGCACCAA3# (sense) and 5#-CTTTAGCACGCACTGTAGTTTCTC-3# (antisense) for b-actin. Polymerase chain reaction was carried out in a GeneAmp 2700 thermocycler (Applied Biosystems, Foster City, CA) under the following conditions: initial denaturation at 95°C for 15 min; 35 cycles of 94°C for 45 s, 61°C for 45 s, 72°C for 1 min and then 72°C for 5 min for final extension. Equal amounts of polymerase chain reaction products were electrophoresed with 2% agarose, 1 tris acetate ethylene diamine tetra acetic acid gels containing 0.5 lg/ml ethidium bromide. Bands were quantified using Kodak Image Station 2000R loaded with Kodak one-dimensional image analysis software and the average fold change intensities were calculated. Western blot analysis Colon extracts were prepared in RIPA cell lysis buffer and an equal amount of protein was separated on 12% sodium dodecyl sulfate–polyacrylamide gel electrophoresis, electroblotted on nitrocellulose membranes and probed with specific antibodies against AR, Cox-2, phospho-PKCb2, cyclin D1, b-catenin and glyceraldehyde-3-phosphate dehydrogenase. Antibody binding was detected by enhanced pico chemiluminescence (Pierce, Rockford, IL). Immunopositive bands were quantified using Kodak Image Station 2000R loaded with Kodak one-dimensional image analysis software and the average fold change intensities were calculated. Statistical analysis Data are presented as mean ± SE and the P values were determined using the Wilcoxon rank-sum test. P values ,0.05 were considered as statistically significant. Results Inhibition of AR prevents AOM-induced ACF formation in BALB/c mice The impetus for studying the effect of AR inhibition on AOM-induced ACF formation comes from our recent observation that AR inhibitors Downloaded from http://carcin.oxfordjournals.org at UCHSC Denison Memorial Library Box A-003 on July 14, 2010 lipopolysaccharide (27–30). AR efficiently reduces one of the most abundant and toxic lipid aldehyde, 4-hydroxy-trans-2-nonenal (HNE), to 1,4-dihydroxynonene and its glutathione conjugate, glutathionyl-4hydroxynonenal (GS-HNE), to glutathionyl-1,4-dihydroxynonene (GS-DHN) (31). Recently, we have demonstrated that the reduced glutathione-lipid aldehydes such as GS-DHN may be one of the regulators involved in nuclear factor kappa binding protein (NF-jB) activation (19,31). Inhibition of AR prevented protein kinase C, NFjB and activator protein (AP)-1 activation and the increase in cell growth caused by HNE and GS-HNE, but not GS-DHN suggesting that the already reduced form of GS-DHN is insensitive to AR inhibition (31). Further, in in vivo rat model, we have shown that inhibition of AR significantly decreases neointima formation in balloon-injured carotid arteries and in situ activation of NF-jB during restenosis (29,32). In human colon cancer Caco-2 cells, we have shown that inhibition of AR prevents the cytokines- and growth factors-induced Cox-2 expression, activation of NF-jB and PGE2 production (17,19). In addition, in in vivo nude mice xenograft model, we have shown that inhibition of AR by AR–small interfering RNA (siRNA) prevents the human colon adenocarcinoma SW480 cells-induced tumor growth (17). Therefore, to further understand the role of AR in the pathogenesis of colon cancer, we investigated the role of AR in AOM-induced ACF formation and inflammatory markers such as Cox-2 and iNOS expression by using BALB/c mice. The results obtained from BALB/c mice were further confirmed by using aldose reductase knockout (ARKO) C57BL/6 mice and its wild-type (WT) littermates. Our studies indicate that inhibition or knockout of AR prevents AOM-induced ACF formation and expression of inflammatory markers in mice. Aldose reductase in ACF formation Inhibition of AR prevents AOM-induced inflammatory markers expression in BALB/c mice Since both Cox-2 and iNOS are known to be involved in chronic inflammation, which creates a microenvironment contributing to the development of preneoplastic lesions in the colon carcinogenesis and inhibitors of Cox-2 and iNOS have been shown to reduce ACF formation in rodents (11,14,24), we next investigated if inhibition of AR could prevent AOM-induced Cox-2 and iNOS expression in mice colons. As shown in Figure 2A, AOM-induced expression of Cox-2 and iNOS was significantly (68% and 86%) prevented by AR inhibition. These results were further confirmed by immunohistochemistry using antibodies against Cox-2 and iNOS (Figure 3A and B). The colon from AOM-treated mice showed a significant intensity of antibody staining, whereas treatment of mice with sorbinil showed a marked decrease in antibody staining, suggesting that inhibition of AR prevented Cox-2 and iNOS expression. We next determined the effect of AR inhibition on Cox-2 and iNOS expression at messenger RNA (mRNA) level (Figure 2B). Treatment of mice with AOM caused a significant increase in the expression of both Cox-2 and iNOS mRNA and inhibition of AR prevented it by 60 and 42%, respectively, suggesting that AR could regulate transcriptional activation of Cox-2 and iNOS. Inhibition of AR by sorbinil prevents AOM-induced phosphorylation of PKCb2 and expression of AR in BALB/c mice Carcinogen-induced preneoplastic lesions in the colonic epithelium are known to be associated with the activation of PKCb2 that could cause hyperproliferation (33,34). We therefore examined the effect of AR inhibition on the activation of AOM-induced PKCb2 in mice colons. A dramatic increase in PKCb2 phosphorylation was observed in the AOM-treated mice, whereas inhibition of AR in AOM-treated mice showed a significant decrease in the phosphorylation of PKCb2 (Figure 2C). Since AR is an oxidative stress response protein that is elevated in various pathological conditions including cancer (29,35), we next examined the expression of AR in AOM-treated mice colons. As shown in Figure 2C, treatment of mice with AOM increased the expression of AR that was prevented by AR inhibition. IHC staining of colon sections with anti-AR antibodies further confirmed the increased expression of AR in AOM-treated mice as evidenced by a strong staining of AR in colon epithelium compared with colon from untreated controls (Figure 3C). Fig 1. Inhibition of AR prevents AOM-induced ACF formation in BALB/c mice. (A) After 10 weeks of AOM treatment, mice were killed and colons were fixed in formalin and stained with 0.2% methylene blue for 5 min. ACF were identified under light microscope (400 magnification) with colon mucosal side up; 1, singlet; 2, doublet; 3, triplet and 4, multiaberrant crypts with well-defined eye were identified and counted. (B) Balb/c mice were divided into three groups: (i) control; (ii) AOM (10 mg/kg body wt, weekly for 3 weeks) and (iii) received sorbinil (25 mg/kg body wt, intraperitoneally per day) throughout the study after 24 h of first AOM injection. Mice were euthanized 10 weeks after first AOM injection and colons were stained and examined for ACF formation. For each group of animals, n 5 6; P , 0.016 as compared with WT þ AOM alone. Wilcoxon rank-sum test was performed for the statistical analysis of ACF data obtained from AOM versus AOM þ sorbinil-treated animals. (C) Colon sections were stained with PCNA antibodies and total intensity of sections was measured using metamorph software. Bars represent mean ± SE (n 5 3); P , 0.001 as compared with control and P , 0.01 as compared with AOM alone. Open circle (s) indicates number of ACF and closed circle indicates (d) average number of ACF. 801 Downloaded from http://carcin.oxfordjournals.org at UCHSC Denison Memorial Library Box A-003 on July 14, 2010 prevent colon cancer cell growth in culture as well as nude mice xenografts (17). Since ACF formation in animals is recognized as early preneoplastic lesions in the colon and mimics putative precursor lesions from which adenomas and carcinomas are formed in humans (11), we first examined the effect of an AR inhibitor on AOM-induced ACF formation in mice. The body weights of AOM and vehicle or AR inhibitor-treated groups were comparable and no significant changes were observed throughout the study (data not shown). At the early preneoplastic stage (10 weeks after first AOM injection), mice were killed and their colons were removed and analyzed microscopically for the presence of ACF. ACF were distinguished from the surrounding normal crypts by increased thickening of the crypt walls and aberrant change in the shape of the crypt lumen (Figure 1A). The parameters used to assess the ACF were their occurrence/colon and the number of aberrant crypts/foci. All the colons were scored by three observers who were masked to the identity of the samples. In mice, AOM treatment induced ACF, 9–17/colon (Figure 1B). No evidence of ACF was observed in vehicle-treated control animals. Administration of sorbinil to AOM mice significantly suppressed the formation of ACF, 2–4/colon, indicating that inhibition of AR prevents preneoplastic lesion formation in mice. Since inhibition of AR prevents the AOM-induced ACF formation, we next measured the proliferation index by staining colon sections with anti-PCNA antibodies. Significant expression of PCNA was observed in mice treated with AOM alone (Figure 1C). Treatment of mice with AR inhibitor significantly (70%) reduced the expression of PCNA. A good correlation was observed between the ability of AR inhibitor to prevent ACF formation and decreased PCNA-labeling index. R.Tammali et al. Collectively, these results suggest that AR inhibition prevents its signaling events responsible for its own gene expression. We next measured the effect of AR inhibition on unsaturated lipid aldehydes to determine the status of lipid peroxidation in AOM-treated colons. As shown in Figure 2D, increased levels of fluorescence in AOM-treated mice indicated increased levels of a, b-unsaturated aldehydes in colons and inhibition of AR significantly prevented the increase in the formation of unsaturated aldehydes. Since lipid aldehydes are highly electrophilic and can form conjugates with cellular proteins (16,32,36), we next measured the effect of AR inhibition on the levels of protein–HNE adducts in colon sections. Treatment of mice with AOM increased the formation of protein–HNE adducts as evidenced by strong staining with HNE antibodies (Figure 3D). Inhibition of AR significantly prevented it. These results suggest that inhibition of AR prevents AOM-induced lipid peroxidation, protein–HNE adducts, activation of PKCb2 and expression of Cox-2 and iNOS, which could be considered as major factors for the development of ACF. Inhibition of AR prevents AOM-induced cyclin D1 and b-catenin expression in BALB/c mice We next measured the expression of key colon cancer-related markers such as cyclin D1 and b-catenin in mice colon. Cyclin D1 is an important cell cycle-regulated protein involved in G1–S phase transition. This cyclin is overexpressed in the presence of growth factors and carcinogens and causes epithelial cell proliferation in various malignancies, including colon cancer (37). b-Catenin is involved in Wnt-signaling pathway. During normal conditions, it binds to axin– GSK-3–APC complex and undergoes proteolytic degradation, whereas during carcinogenesis b-catenin stabilizes in the cytoplasm by deregulated Wnt-signaling pathway and translocates into the nucleus and interacts with T-cell factor/lymphoid enhancer factor family of transcription factors to promote expression of various oncogenes 802 (26,38). IHC examination of cyclin D1 and b-catenin showed that the abundance of these two enzymes was markedly elevated when mice were treated with AOM alone as evident by dark brown staining using antibodies (Figure 3E and F). Inhibition of AR by sorbinil resulted in lower levels of AOM-induced expression of cyclin D1 and b-catenin, indicating that AR regulates the expression of these proteins. AR-deficient mice are resistant to AOM-induced ACF formation Our results with AR inhibitor strongly suggest that AR mediates AOMinduced ACF formation. If this is true, the AR-deficient mice should be resistant to AOM-induced ACF formation. To investigate this, we examined the effect of AOM on ACF formation in ARKO C57BL/6 mice and their WT controls. The results as shown in Figure 4A demonstrate that ARKO mice had a significant decrease in the total number of ACF per colon at both 5 and 10 weeks as compared with WT controls. The low numbers of ACF were found in ARKO at both 5 weeks (3–5 ACF/ colon) and at 10 weeks (5–6 ACF/colon) as compared with WT (8–10 and 12–16 ACF/colon, respectively). These results suggest that AR-null mice are less susceptible to AOM-induced ACF formation. Expression of proliferation index marker, PCNA was significantly lowered in AOM-treated ARKO mice compared with WT mice (Figure 4B). These results suggest that ARKO mice are less susceptible to AOM-induced growth of colonic cells. AR-deficient mice are resistant to AOM-induced inflammatory markers expression and formation of protein–HNE adducts We have also examined the effect of AOM on inflammatory markers expression in AR-null mice. Our results show a significant decrease in Cox-2 and iNOS expression in ARKO mice as compared with WT mice at 5 and 10 weeks after AOM injection (Figure 4C). Similar results were observed in IHC analysis of colon sections with Cox-2 and iNOS antibodies (Figure 5A and B). Downloaded from http://carcin.oxfordjournals.org at UCHSC Denison Memorial Library Box A-003 on July 14, 2010 Fig. 2. Inhibition of AR prevents expression of inflammatory markers in AOM-treated BALB/c mice. (A and C) Ten weeks after first AOM injection, colons were removed and homogenized. Equal amounts of cell lysates were subjected to western blot analysis using antibodies against Cox-2, iNOS, phospho-PKCb2, AR and glyceraldehyde-3-phosphate dehydrogenase (GAPDH). (B) Reverse transcription–polymerase chain reaction analysis of colon samples obtained 10 weeks after AOM treatment. Densitometry analysis was performed using Kodak 1D image analysis software. A representative blot is shown (n 5 3). (D) The levels of a, b-unsaturated aldehydes were measured as described in Material and Methods. Bars represent mean ± SE (n 5 3); P , 0.001 as compared with control and P , 0.01 as compared with AOM alone. Aldose reductase in ACF formation Consistent with the protein levels, the mRNA levels of Cox-2 and iNOS were significantly increased in WT mice treated with AOM alone (Figure 4D). However, AOM-treated ARKO mice show a significantly (50%) reduced expression of Cox-2 and iNOS mRNA suggesting that AR regulates the expression of these inflammatory proteins at transcriptional level. Further, as shown in Figure 5C, the levels of protein–HNE adducts are significantly increased in AOM WT mice but not in AOM AR-null mice, suggesting that AR deficiency reduces the formation of protein–HNE adducts. Further, we also examined the effect of AOM on the expression of cyclin D1 and b-catenin in WT and ARKO mice. As shown in Figure 5D and E, treatment of WT mice with AOM significantly increased the expression of cyclin D1 and b-catenin as compared with ARKO mice, suggesting that AR deficiency prevents AOM-induced expression of cyclin D1 and b-catenin in mice. We next measured the extent of AOM-induced inflammation in colon sections. WT and ARKO control mice displayed normal crypts and colonic architecture with no signs of apparent abnormality (Figure 5F). WT mice treated with AOM alone showed enlarged nuclei and thickened mucosal layer with densely packed inflammatory cell infiltrations, whereas no infiltration of inflammatory cells was observed in ARKO mice treated with AOM. These results suggest that ARKO mice are resistant to AOM-induced inflammation. AR-deficient mice are resistant to AOM-induced phosphorylation of PKCb2 and p65 and general inflammation Since the expression of Cox-2 and iNOS depends on the activation of NF-jB and its upstream signals, we next measured the phosphorylation of PKCb2 and p65 in ARKO and WT mice treated with AOM. As shown in Figure 6A, WT mice at 5 and 10 weeks after first AOM injection showed increased phosphorylation of PKCb2 and p65, whereas ARKO mice showed a significant resistance to AOM-induced activation of PKCb2 and p65. Similar results were observed with IHC analysis (data not shown). Discussion Colon cancer is a multistep process that involves sequential transformation of normal epithelial cells into malignant cells through various progressive stages such as preneoplastic, neoplastic and malignant (17,37). This multistep carcinogenesis is referred as a progressive disorder in signal transduction mechanism. Altered regulations of redox stress signal transduction pathway intermediates such as PKCb2 and NF-jB, whichpromote the expression of various inflammatory markers, are early events in the colon carcinogenesis (17,33,34). In this mechanism, identification of the earliest detectable preneoplastic lesions may provide appropriate way of screening and prevention 803 Downloaded from http://carcin.oxfordjournals.org at UCHSC Denison Memorial Library Box A-003 on July 14, 2010 Fig. 3. Inhibition of AR prevents the expression of inflammatory and preneoplastic markers in AOM-treated BALB/c mice. (A–F) Colon sections were stained with antibodies against Cox-2, iNOS, AR, protein–HNE, cyclin D1 and b-catenin. Immunoreactivity of the antibody was assessed by quantifying the evident as a dark brown stain in the cells, whereas the non-reactive areas displayed only the background color. Photomicrographs of the stained sections were acquired using an EPI-800 microscope (bright-field) connected to a Nikon camera (400 magnification). Percent staining was determined by measuring positive immunoreactivity per unit area. Arrows represent the area for positive staining for an antigen. The intensity of antigen staining was quantified by digital image analysis and the values are represented as fold change versus control in arbitrary units. R.Tammali et al. method for colon cancer. In mice model, AOM-induced formation of ACF was first described by Bird (10). ACF are described as enlarged, thicker epithelial lining, slightly elevated from the surrounding mucosa and darkly stained with methylene blue. In humans and experimental animal models, ACF are considered putative preneoplastic lesions (11). Formation of a large number of ACF with increased cell proliferation, genetic alteration in K-ras, APC and p53 and microsatellite instability is observed in patients with familial adenomatous polyposis as well as in patients with sporadic colon cancer (25,39). Numerous reports show that treatment of colon cancer patients with non-steroidal anti-inflammatory drugs effectively reduces the formation of ACF (24,40). Most drugs with antioxidant property prevent the generation of ROS thereby inhibit the formation of ACF (4,16,40,41). However, antioxidant drugs, which have good solubility in both hydrophilic and hydrophobic milieu and whichdo not become pro-oxidants when given in therapeutic doses, are currently unavailable. We have shown previously that inhibition of AR can prevent growth factors-, cytokines-, chemokines- and hyperglycemia-induced ROS signals (27–30). Indeed, our present report on AOM-induced ACF formation in mouse model shows that inhibition of AR by pharmacological agents could prevent AOM-induced ROS generation as well as its dependent activation of NF-jB. In human colon cancer cells, in- 804 hibition of AR prevented growth factors such as fibroblast growth factor- and platelet-derived growth factor-induced ROS formation (17). The ROS generated during growth factors and cytokines signaling could induce formation of a wide range of cytotoxic, lipid peroxidation-derived aldehydes such as HNE (31). These aldehydes could readily conjugate with reduced glutathione and form glutathionyl aldehydes such as GS-HNE (31). We have reported that AR efficiently reduces GS-HNE to GS-DHN, which in turn activates transcription factors such as NF-jB and AP-1 to induce expression of various inflammatory markers such as Cox-2 and iNOS via signaling cascade axis of PKC/phosphotidyl inositol 3-kinase/mitogen-activated protein kinase/inhibitory kappa kinase (17,31). Uncontrolled production of these inflammatory markers would lead to cellular cytotoxicity leading to ACF formation (Figure 6B). To further understand the role of AR in colonic epithelial cell proliferation and colon carcinogenesis, we have used AOM-induced ACF formation in BALB/c mouse model. Treatment of mice with AR inhibitors significantly prevented the AOM-induced ACF formation. Various reports show that AR is a growth-responsive enzyme and is overexpressed in various forms of cancer such as hepatocarcinogenesis, lung, breast and colon cancers (29,35,42). We are first to report the role of AR in AOM-induced ACF formation. Consistent with Downloaded from http://carcin.oxfordjournals.org at UCHSC Denison Memorial Library Box A-003 on July 14, 2010 Fig. 4. AR-deficient mice are resistant to AOM-induced ACF formation and exhibit decreased expression of AOM-induced inflammatory markers. (A) Six-weekold ARKO and WT mice were injected with AOM (10 mg/kg body wt) weekly for 3 weeks. After 5 and 10 weeks of first AOM injection, mice were killed, colons were fixed in formalin and stained with 0.2% methylene blue for 5 min. Colons were analyzed for the presence of ACF counted under light microscope (400 magnification) with colon mucosal side up. For each group of animals, n 5 6; P , 0.015 as compared with WT þ AOM alone. Wilcoxon rank-sum test was performed to statistically analyze ACF data obtained from WT þ AOM versus ARKO þ AOM. (B) Colon sections, obtained after 10 weeks after first AOM, were stained with PCNA antibodies and total intensity of sections was measured using metamorph software. Bars represent mean ± SE (n 5 3); P , 0.001 as compared with control and P , 0.01 as compared with WT þ AOM alone. (C) Total cell lysate of colons subjected to western blot analysis by using antibodies against Cox-2, iNOS and glyceraldehyde-3-phosphate dehydrogenase (GAPDH). (D) Reverse transcription–polymerase chain reaction analysis of colon samples. Bands were quantified by densitometry analysis using Kodak 1D image analysis software. A representative blot is shown (n 5 3). Open circle (s) indicates number of ACF and closed circle (d) indicates average number of ACF. Aldose reductase in ACF formation prevention of AOM-induced ACF formation by AR inhibitor, we have shown protection of AR-null mice from AOM-induced ACF formation. Kawamura et al. (43) using a mice model showed that inhibition of AR by ponalrestat prevents the cachexia syndrome induced by colon 26 adenocarcinoma cells. Further, they showed that inhibition of AR prevents the lipopolysaccharide-induced IL-1 secretion in monocytes. In nude mice xenograft model, we have shown that inhibition of AR by AR–siRNA prevented the progression of human adenocarcinoma SW480 cells-induced tumor growth (17). In addition, our results further confirm that inhibition or deletion of AR prevents AOM-induced expression of Cox-2, iNOS, phospho-PKCb2 and phospho-p65. The inducible form of Cox-2 enzyme plays an important role in the pathogenesis of colon cancer by generating hyperalgesic, proinflammatory prostaglandins and leukotrienes. There are numerous non-steroidal anti-inflammatory drugs, designed based on the structure and function of the Cox-2 enzyme for the treatment of colon carcinogenesis (41). The de novo synthesis of Cox-2 is triggered by the exposure of epithelial cells to endotoxin, cytokines and growth factors, whereas constitutive Cox-1 enzyme is involved in eicosanoid metabolism and homeostasis of tissues (17). Moeckel et al. (44) in Cox-2 knockout mice demonstrated that expression of AR mRNA is dramatically reduced during hypertonic stress indicating that AR is involved in the signal transduction path- way of Cox-2 enzyme. In addition, other inflammatory markers such as iNOS, PKCb2 and p65 play important roles in the progression of colon carcinogenesis (17,24,33). Overexpressing PKCb2 transgenic mice have been shown to be more sensitive to AOM-induced ACF formation than its WT littermates and overexpression of PKCb2 is early step to promote colon carcinogenesis (33,34). Various reports have demonstrated that AR-deficient mice are protected from increased c-Jun-N-terminal kinase activation, depletion of reduced glutathione, increased superoxide accumulation and DNA damage during oxidative stress conditions such as hyperglycemia (45). Similar to diabetic ARKO mice, inhibition of AR in diabetic and WT mice showed decreased activation of mitogen-activated protein kinase suggesting that AR is involved in oxidative stress (46). Our results in vascular smooth muscle cells, human lens epithelial cells, vascular endothelial cells and Caco-2 cells also show that inhibition of AR prevents the growth factors-, cytokines-, chemokines- and hyperglycemia-induced oxidative signals such as generation of ROS, activation of poly (ADP) – ribose polymerase and mitogenactivated protein kinase via NF-jB (17,27–30). Since AR gene promoter has binding site for NF-jB, AR is regulated by NF-jB (28). Our results show that inhibition of AR by sorbinil prevents AOMinduced colonic AR expression suggesting that AR inhibition responsible for signaling events regulates its own gene expression. Our 805 Downloaded from http://carcin.oxfordjournals.org at UCHSC Denison Memorial Library Box A-003 on July 14, 2010 Fig. 5. AR-deficient mice exhibit decreased expression of AOM-induced preneoplastic markers. (A–E) Colon sections were stained with antibodies against Cox-2, iNOS and protein–HNE, cyclin D1 and b-catenin. Immunoreactivity of the antibody was assessed by quantifying the evident as a dark brown stain in the cells, whereas the non-reactive areas displayed only the background color. Photomicrographs of the stained sections were acquired using an EPI-800 microscope (brightfield) connected to a Nikon camera (400 magnification). Percent staining was determined by measuring positive immunoreactivity per unit area. Arrows represent the area for positive staining for an antigen. The intensity of antigen staining was quantified by digital image analysis and the values are presented as fold change versus control in arbitrary units. (F) Histopathological examination of colon sections stained with hematoxylin and eosin (H&E). Photomicrographs of the stained sections were acquired using an EPI-800 microscope (bright-field) connected to a Nikon camera (400 magnification). Arrows in black and pink color indicate densely packed inflammatory cell infiltration and enlarged nuclei in AOM-treated mice. R.Tammali et al. results further suggest that inhibition or deletion of AR prevents AOM-induced expression and nuclear translocation of b-catenin and decreases the expression of cyclin D1 and PCNA. Altered expression of b-catenin seems to be one of the earliest genetic events critical in colonic hyperproliferation through regulation of cyclin D1 (13,37). Further, number of reports shows that targeted downregulation of bcatenin or its downstream regulator, cyclin D1, plays an important role in chemoprevention of colon carcinogenesis (38). In addition, it has been shown that numerous non-steroidal anti-inflammatory drugs have chemopreventive effect by suppressing the expression of b-catenin and cyclin D1 (38). In summary, our results show for the first time that inhibition or deletion of AR significantly prevents AOM-induced ACF formation by inhibiting the expression of important inflammatory markers and that the use of AR inhibitors may provide an effective chemopreventive approach for the treatment of colon cancer. Funding Pearle Vision Research Foundation; Washington University Department of Ophthalmology and Visual Sciences from Research to Prevent Blindness, Inc.(CA 129383, DK 36118 to S.K.S., GM 71036 to K.V.R.); EY05856, EY02687 to J.M.P. Acknowledgements The authors thank Dr Stephen Chung (University of Hong Kong, Hong Kong, China) for providing the ARKO mice and Theresa Harter (Washington University School of Medicine) for assistance with animal studies. The authors also thank Dr Heidi Weiss (Biostatistics Core, Department of Comprehensive Cancer Center, University of Texas Medical Branch, Galveston, TX) for helping in performing statistical analysis. Conflict of Interest Statement: None declared. References 1. Majer,M. et al. (2007) Oncologists’ current opinion on the treatment of colon carcinoma. Anticancer Agents Med. Chem., 7, 492–503. 2. Jemal,A. et al. (2005) Cancer statistics 2005. CA Cancer J. Clin., 55, 10–30. 806 3. Heavey,P.M. et al. (2004) Colorectal cancer and the relationship between genes and the environment. Nutr. Cancer, 48, 124–141. 4. Yang,K. et al. (2005) Dietary components modify gene expression: implications for carcinogenesis. J. Nutr., 135, 2710–2714. 5. Raju,J. et al. (2005) Low doses of beta-carotene and lutein inhibit AOMinduced rat colonic ACF formation but high doses augment ACF incidence. Int. J. Cancer, 113, 798–802. 6. Singh,S.V. et al. (2004) Sulforaphane-induced G2/M phase cell cycle arrest involves checkpoint kinase 2-mediated phosphorylation of cell division cycle 25C. J. Biol. Chem., 279, 25813–25822. 7. Durai,R. et al. (2006) Biology of insulin-like growth factor binding protein4 and its role in cancer. Int. J. Oncol., 28, 1317–1325. 8. MacDonald,T. et al. (1999) T cells orchestrate intestinal mucosal shape and integrity. Immunol. Today, 20, 505–510. 9. Cohen,G. et al. (2006) Epidermal growth factor receptor signaling is upregulated in human colonic aberrant crypt foci. Cancer Res., 66, 5656–5664. 10. Bird,R.P. (1987) Observation and quantification of aberrant crypts in the murine colon treated with a colon carcinogen: preliminary findings. Cancer Lett., 37, 147–151. 11. Pretlow,T.P. et al. (1991) Aberrant crypts: putative preneoplastic foci in human colonic mucosa. Cancer Res., 51, 1564–1577. 12. Samowitz,W.S. et al. (2007) APC mutations and other genetic and epigenetic changes in colon cancer. Mol. Cancer Res., 5, 165–170. 13. Takahashi,M. et al. (2004) Gene mutations and altered gene expression in azoxymethane-induced colon carcinogenesis in rodents. Cancer Sci., 95, 475–480. 14. Jackson,L. et al. (2006) Chronic inflammation and pathogenesis of GI and pancreatic cancers. Cancer Treat. Res., 130, 39–65. 15. Itzkowitz,S.H. et al. (2004) Inflammation and cancer IV. Colorectal cancer in inflammatory bowel disease: the role of inflammation. Am. J. Physiol. Gastrointest. Liver Physiol., 287, G7–G17. 16. Valko,M. et al. (2006) Free radicals, metals and antioxidants in oxidative stress-induced cancer. Chem. Biol. Interact., 160, 1–40. 17. Tammali,R. et al. (2006) Aldose reductase regulates growth factor-induced cyclooxygenase-2 expression and prostaglandin E2 production in human colon cancer cells. Cancer Res., 6, 9705–9713. 18. Romero-Giménez,J. et al. (2008) Germ line hypermethylation of the APC promoter is not a frequent cause of familial adenomatous polyposis in APC/ MUTYH mutation negative families. Int. J. Cancer, 122, 1422–1425. 19. Tammali,R. et al. (2007) Aldose reductase regulates TNF-alpha-induced PGE2 production in human colon cancer cells. Cancer Lett., 252, 299–306. 20. Weaver,S.A. et al. (2001) Regulatory role of phosphatidylinositol 3-kinase on TNF-a-induced cyclooxygenase 2 expression in colonic epithelial cells. Gastroenterology, 120, 1117–1127. Downloaded from http://carcin.oxfordjournals.org at UCHSC Denison Memorial Library Box A-003 on July 14, 2010 Fig. 6. AR-deficient mice exhibit decreased phosphorylation of PKCb2 and p65 in the presence of AOM compared with WT. (A) Five and 10 weeks after first AOM treatment, colons were removed and homogenized. Total cell lysate was subjected to western blot analysis using antibodies against phospho-PKCb2, phospho-p65 and glyceraldehyde-3-phosphate dehydrogenase (GAPDH). Band intensity was quantified by densitometry analysis using Kodak 1D image analysis software. A representative blot is shown (n 5 3). (B) Molecular mechanism by which AR inhibition prevents ACF formation. Oxidative stress causes peroxidation of lipids resulting in the generation of toxic lipid aldehydes such as HNE and GS-HNE. The AR catalyzed reduced forms of glutathione–aldehyde conjugates mediate the oxidative stress signals (17,31) that cause activation of NF-jB and expression of iNOS and Cox-2, which are responsible for DNA damage, inflammation leading to ACF formation in colons. Aldose reductase in ACF formation 35. Saraswat,M. et al. (2006) Overexpression of aldose reductase in human cancer tissues. Med. Sci. Monit., 12, CR525–CR529. 36. Kawanishi,S. et al. (2006) Oxidative and nitrative DNA damage in animals and patients with inflammatory diseases in relation to inflammation-related carcinogenesis. Biol. Chem., 387, 365–372. 37. Arber,N. et al. (1996) Increased expression of cyclin D1 is an early event in multistage colorectal carcinogenesis. Gastroenterology, 110, 669–674. 38. Nath,N. et al. (2003) Nitric oxide-donating aspirin inhibits beta-catenin/ T cell factor (TCF) signaling in SW480 colon cancer cells by disrupting the nuclear beta-catenin-TCF association. Proc. Natl Acad. Sci., 100, 12584–12589. 39. Rao,C.V. et al. (2005) Colonic tumorigenesis in BubR1þ/-ApcMin/þ compound mutant mice is linked to premature separation of sister chromatids and enhanced genomic instability. Proc. Natl Acad. Sci., 102, 4365–4370. 40. Sakoguchi-Okada,N. et al. (2007) Celecoxib inhibits the expression of survivin via the suppression of promoter activity in human colon cancer cells. Biochem. Pharmacol., 73, 1318–1329. 41. Swamy,M.V. et al. (2004) Modulation of cyclooxygenase-2 activities by the combined action of celecoxib and decosahexaenoic acid: novel strategies for colon cancer prevention and treatment. Mol. Cancer Ther., 3, 215–221. 42. Jin,J. et al. (2006) Role of aldo-keto reductases in development of prostate and breast cancer. Front. Biosci., 11, 2767–2773. 43. Kawamura,I. et al. (1999) Ponalrestat, an aldose reductase inhibitor, inhibits cachexia syndrome induced by colon26 adenocarcinoma in mice. Anticancer Res., 19, 4105–4111. 44. Moeckel,G.W. et al. (2003) COX2 activity promotes organic osmolyte accumulation and adaptation of renal medullary interstitial cells to hypertonic stress. J. Biol. Chem., 278, 19352–19357. 45. Ho,E.C. et al. (2006) Aldose reductase-deficient mice are protected from delayed motor nerve conduction velocity, increased c-Jun NH2-terminal kinase activation, depletion of reduced glutathione, increased superoxide accumulation, and DNA damage. Diabetes, 55, 1946–1953. 46. Price,S.A. et al. (2004) Mitogen-activated protein kinase p38 mediates reduced nerve conduction velocity in experimental diabetic neuropathy: interactions with aldose reductase. Diabetes, 53, 1851–1856. Received June 23, 2008; revised September 26, 2008; accepted October 21, 2008 807 Downloaded from http://carcin.oxfordjournals.org at UCHSC Denison Memorial Library Box A-003 on July 14, 2010 21. Atreya,R. et al. (2005) Involvement of IL-6 in the pathogenesis of inflammatory bowel disease and colon cancer. Clin. Rev. Allergy Immunol., 28, 187–196. 22. Duque,J. et al. (2006) Up-regulation of cyclooxygenase-2 by interleukin1beta in colon carcinoma cells. Cell. Signal., 18, 1262–1269. 23. Panja,A. et al. (1998) The regulation and functional consequence of proinflammatory cytokine binding on human intestinal epithelial cells. J. Immunol., 161, 3675–3684. 24. Rao,C.V. et al. (2002) Chemopreventive properties of a selective inducible nitric oxide synthase inhibitor in colon carcinogenesis, administered alone or in combination with celecoxib, a selective cyclooxygenase-2 inhibitor. Cancer Res., 62, 65–70. 25. Ochiai,M. et al. (2003) Characterization of dysplastic aberrant crypt foci in the rat colon induced by 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine. Am. J. Pathol., 163, 1607–1614. 26. Takahashi,M. et al. (2000) Altered expression of beta-catenin, inducible nitric oxide synthase and cyclooxygenase-2 in azoxymethane-induced rat colon carcinogenesis. Carcinogenesis, 21, 1319–1327. 27. Ramana,K.V. et al. (2003) Aldose reductase mediates cytotoxic signals of hyperglycemia and TNF-a in human lens epithelial cells. FASEB J., 17, 15–17. 28. Srivastava,S.K. et al. (2005) Role of aldose reductase and oxidative damage in diabetes and the consequent potential for therapeutic options. Endocr. Rev., 26, 380–392. 29. Ramana,K.V. et al. (2002) Aldose reductase mediates mitogenic signaling in vascular smooth muscle cells. J. Biol. Chem., 277, 32063–32070. 30. Ramana,K.V. et al. (2006) Endotoxin-induced cardiomyopathy and systemic inflammation in mice is prevented by aldose reductase inhibition. Circulation, 114, 1838–1846. 31. Ramana,K.V. et al. (2006) Mitogenic responses of vascular smooth muscle cells to lipid peroxidation-derived aldehyde 4-hydroxy-trans-2-nonenal (HNE): role of aldose reductase-catalyzed reduction of the HNE-glutathione conjugates in regulating cell growth. J. Biol. Chem., 281, 17652–17660. 32. Srivastava,S. et al. (2006) Contribution of aldose reductase to diabetic hyperproliferation of vascular smooth muscle cells. Diabetes, 55, 901–910. 33. Murray,N.R. et al. (1999) Overexpression of protein kinase C betaII induces colonic hyperproliferation and increased sensitivity to colon carcinogenesis. J. Cell Biol., 145, 699–711. 34. Gökmen-Polar,Y. et al. (2001) Elevated protein kinase C betaII is an early promotive event in colon carcinogenesis. Cancer Res., 61, 1375–1381.
© Copyright 2019