• murine colitis;
  • colon carcinogenesis;
  • DSS colitis;
  • ursodeoxycholic acid


  1. Top of page
  2. Abstract
  3. Material and method
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Bile acids in the intestinal lumen contribute to the homeostatic regulation of proliferation and death of the colonic epithelial cells: Deoxycholic acid (DCA) appears to enhance and ursodeoxycholic acid (UDCA) to attenuate the process of chemically induced carcinogenesis. We studied the effects of UDCA on colitis-related colorectal carcinogenesis. Three groups of 25 mice were given 0.7% dextran sulphate in drinking water for 7 days and pure water for 10 days and were fed a standard diet containing double iron concentration. In 2 groups, the diet was supplemented with 0.2% cholic acid (CA), the precursor of DCA, or with 0.4% UDCA. After 15 cycles, the histology, the expression of MUC2, β-catenin, p27 and p16 and the fecal water concentration of DCA and UDCA were investigated. All animals showed colitis with similar severity and histologic as well as immunophenotypic alterations, resembling those of human colitis. Among the animals fed the nonsupplemented diet, 46% developed colorectal adenocarcinomas and 54% anal-rectal squamous cell carcinomas. The prevalence of dysplasia and carcinomas did not change significantly in the animals given CA. Among the mice fed with UDCA, none developed adenocarcinomas and 20% squamous carcinomas. Dysplastic lesions were found in 88%, 67% and 40% of each group, respectively. The prevalence of dysplasia as well as of carcinoma showed an inverse relationship to the UDCA concentration in the fecal water. These data indicate that UDCA suppresses colitis-associated carcinogenesis. This model is suitable for investigation of the mechanism of the anticarcinogenic effect of UDCA in vivo. © 2005 Wiley-Liss, Inc.

Ulcerative colitis (UC) in patients and in experimental animals is associated with a high risk of intraepithelial neoplasia and colonic carcinoma. The cumulative risk for colorectal cancer in a patient with UC is 2% at 10 years, 8% at 20 years and 18% at 30 years of disease duration.1 Since the occurrence of dysplasia can be a strong indication for prophylactic proctocolectomy,2 the inhibition or slowing down of the colitis-associated carcinogenesis would essentially reduce the number of necessary operations.

Factors contributing to colon carcinogenesis are high epithelial proliferation rate in the inflamed region and the resulting increased accumulation of mutations as well as the disturbed regulation of cell death and proliferation. Intestinal bile acids contribute to the homeostatic regulation of proliferation and apoptosis of the colonic epithelium. Deoxycholic acid (DCA), whose fecal water concentration in patients with colitis is increased,3 stimulates the proliferation of the colonic epithelial cells4 and promotes colon carcinogenesis in an azoxymethane (AOM) rat model.5 The addition of ursodeoxycholic acid (UDCA) to the diet decreases the prevalence of AOM-induced rat colonic tumors6 as well as the prevalence of tumors in MIN mice.7 These data indicate that DCA enhances the chemical colon carcinogenesis in the animal model, while UDCA inhibits the chemical as well as the wnt-pathway-driven colon carcinogenesis.

A chemopreventive effect of UDCA was recently demonstrated also in patients with primary sclerosing cholangitis (PSC), which accompanies 2–7% cases of ulcerative colitis (UC) and increases the risk for developing colon cancer 3- to 5-fold.8, 9 Two retrospective studies showed that patients with PSC and UC, who were treated with UDCA to improve the liver function, have a significantly lower risk of colitis-associated cancer than the non-treated group.10, 11 Whether UDCA treatment has also a protective effect in patients with UC alone, in whom the treatment with UDCA is not a standard therapy, has not been investigated.

The lack of data on this important subject is due to the very slow development of colitis-associated carcinomas and to the dysplasia-related preventive colectomy, which in most cases is performed prior to carcinoma development. This makes an animal model necessary for the investigation of the mechanism of action of UDCA in vivo.

In different colitis models, colon carcinomas develop at different speeds. The DSS-induced intermittent inflammation leads to a slow development of carcinomas within 8 months.12 This time can be reduced to 5 months if colitis-associated colon carcinogenesis is accelerated by a single injection of the carcinogens AOM13 or dimethylhydrazine (DMH),14 both of which are known to activate the wnt pathway.14, 15

The objective of the present work was to test the effects of UDCA on carcinogenesis induced solely by colitis and to compare them with the concentration of bile acids in the colonic fecal water. For this purpose, the mice with colitis were given either UDCA at a dose previously shown to inhibit chemical carcinogenesis6 or–as a control–CA, which is metabolised to DCA and was previously shown to enhance chemical carcinogenesis.6, 16 The results clearly show the chemopreventive effect of UDCA, while the increase of fecal DCA concentration had no significant effect. Our findings indicate that the present model is suitable for investigation of the UDCA-mediated tumour suppression mechanism under controlled in vivo conditions.

Material and method

  1. Top of page
  2. Abstract
  3. Material and method
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Animals and treatment

Three groups of 25 female C57BL/6J mice were given water with 0.7% dextran sulfate sodium (DSS, molecular weight 40,000, MP Biomedicals, Irvine, CA, USA) for 7 days, and for the next 10 days water without any additives as described.12 They were fed with a standard basal diet AIN-76 (for composition see 2-fold enriched with iron (final concentration 90 mg/kg iron, “basal diet”) or with a basal diet containing 0.2% cholic acid (“basal + CA”) or with a basal diet containing 0.4% ursodeoxycholic acid (“basal + UDCA”). After 15 cycles, the surviving mice were weighed, killed and the colons were investigated by histology and immunohistochemistry.

Histopathological examination and scoring of inflammation

Murine colon was washed with PBS and fixed in 4% formaldehyde in PBS, partitioned in 3 segments (proximal, middle, distal) and embedded in paraffin. Serial sections (5 μm) were stained with hematoxylin and eosin. The degree of inflammation was assessed by 2 investigators, using a modified scoring system described by Neurath et al.17: 0 – no signs of inflammation; 1 – mild (inflammatory infiltrates restricted to lamina propria); 2 – moderate (infiltration of mucosa and submucosa, focal erosions); 3 – severe (transmural infiltration, multiple erosions/ulcerations). The extent of inflammation was given as the percentage length of the involved colon segment. The product of the degree- and the extent of inflammation was termed colitis index; the total colitis index was the sum of the colitis indices of the 3 segments.

Immunohistochemical staining and evaluation

Sections (5 μm) of paraffin-embedded tissues were used for expression analysis of β-catenin, p16, p27 and MUC2 proteins. For β-catenin staining, the antigen was retrieved by boiling in 10 mM citric acid, pH 6.0, in a pressure cooker for 10 min and the slides were incubated with the mouse monoclonal anti-β-catenin antibody (Clone 14, BD Pharmingen, 0.08 μg/ml). For detection, the alkaline phosphatase-anti alkaline phosphatase complex (APAAP) method was used. Alkaline phosphatase was revealed by Fast Red as chromogen. Before p16 staining, antigen was retrieved as mentioned earlier, the slides were rinsed in cold running water, washed in Tris-buffered saline (pH 7.4), blocked using a commercial peroxidase-blocking reagent (DakoCytomation, Glostrup, Denmark) and incubated for 60 min with the monoclonal mouse antibody against p16 (F12, Santa Cruz, CA, USA, 1 μg/ml) followed by the EnVision peroxidase kit (DakoCytomation). For p27 staining, the sections were treated with 0.6% H2O2 for 10 min prior to antigen retrieval as mentioned earlier, followed by 1 hr incubation with the mouse monoclonal anti-p27 antibody (Clone 57, BD Transduction Laboratories, Heidelberg, Germany, 2,5 μg/ml). For detection, a ChemMate™ Detection kit (DakoCytomation, Hamburg, Germany) with a biotinylated second antibody and horse radish peroxidase conjugated streptavidin was used. The staining was developed with DAB (3,3′-diaminobenzidine) and counterstained with hematoxylin (Merck, Darmstadt, Germany). Microscopic evaluation was done on an Olympus BX60F5 (Optical Co. GmbH, Hamburg, Germany) using analysis imaging software (Soft Imaging System, Münster, Germany). The expression intensity of p16 or p27 in the epithelial cells was scored as 0 (no staining), 1 (weak staining) or 2 (strong, dark brown staining) and the mean value in the normal epithelium and the various lesions was calculated for each group of mice. MUC2 staining was carried out after antigen retrieval with a rabbit anti-MUC2 antiserum MCM (diluted 1:1000), followed by anti-rabbit biotin and streptavidin-peroxidase detection.

Determination of the bile acid concentration in the fecal water

Murine faeces were collected overnight in 12–17 day intervals and kept frozen at −20°C. Before preparation, they were lyophilised to constant weight and pulverized. Lyophilising of the freshly collected murine faeces showed that water represents 70% of the faeces weight of the colitis mice. To 500 mg of the pulverised lyophilisate, 3.48 ml distilled water were added (fecal water dilution factor 3), shaken at 37°C for 1 hr and centrifuged at 20,000 g for 20 min. The bile acid (BA) analysis was carried out as described previously,18 with minor modifications. In short, 800 μl of the water extract were hydrolysed with ethanolic NaOH and the neutral sterols were extracted with cyclohexane. Hyodeoxycholic acid (HDCA) was added to the remaining aqueous phase as an internal standard at a final concentration of 20 μM. The BAs were saponified with NaOH, then acidified to pH 1.0 with HCl, extracted with diethyl ether and dried. The residue was methylated with dimethoxypropane, evaporated to dryness and silylated with hexamethyl disilazane/chlorotrimethyl silane in pyridine. After drying, the silylated BAs were dissolved in 200 μl decane and subjected to GC-MS analysis. For detection of the highly concentrated BAs, the sample was diluted correspondingly. The analysis was performed with GC17-QP5000 instrument (Shimadzu, Kyoto, Japan), equipped with a capillary column (ZB5; length 30 m; inner diameter 0.25 mm; film thickness 0.25 μm; Phenomenex, Torrance, CA, USA). The injection of 2.5 μl sample solution was carried out at 280°C in split mode (1:50) and the separation followed with a nonlinear temperature programme from 150°C to 270°C.

The simultaneous mass spectrometric detection of the derivatised BAs was performed in multi ion current (MIC) mode, using for DCA: m/z = 255.3 arbitrary mass units (amu), for HDCA: 81.15 amu and for UDCA: m/z = 460.0 amu. Quantification was carried out in MIC mode by using internal standard method and peak areas were calculated from the chromatograms generated by data-handling software Class 5000 (Shimadzu, Kyoto, Japan).The calibrated linear detection range was between 5 μM and 200 μM for each bile acid. Component identification was based on fragmentation and comparison of the retention times with those of the standards (HDCA, DCA and UDCA from Sigma, Munich, Germany). The recovery of the internal standard was between 81 and 92%.

Statistical analysis

The t-test was used for evaluation of the observed differences of p16 and p27 expression as well as for the weight differences between the untreated and the treated groups and the z-test for comparison of the percentage of the identified tumours in each group. The correlation between the tumor prevalence and the fecal water concentration was tested by the Pearson Product Moment Correlation test. The differences or correlations were considered significant if the p-value was p < 0.05.


  1. Top of page
  2. Abstract
  3. Material and method
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Phenotypic characterisation of the DSS colitis-associatedalterations

The inflammatory lesions with erosions and ulcerations showed an appearance characteristic also for human colitis (Fig. 1a) with the occurrence of inflammatory pseudopolyps.

thumbnail image

Figure 1. Histology of the colitis-associated lesions and the immunohistochemical detection of the alterations in β-catenin and p16-expression. β-catenin is localised in the membrane of (b) inflammatory epithelium and in (e): the cytosol and nucleus in limited areas of adenomas (arrows) and (h) mucinous carcinomas (arrow). p16 is strongly expressed in the nuclei of (f) adenoma and weakly in (c) the epithelium of the inflammatory regions (arrow) and (i) adenocarcinomas. Bar: μm.

Download figure to PowerPoint

In the basal diet group, the prevalence of the inflammatory pseudopolyps, typically occurring in the regenerating areas, was 100% and the multiplicity 2.1 ± 1.1 (mean ± SD). The prevalence of tubular adenomas was 17%, while villous adenomas were absent. The prevalence of dysplasia was 88%. Fifty four percent of the mice had rectal squamous carcinomas (supplementary Fig. S1 A) and 46% had adenocarcinomas (Fig. 1g). One third of the adenocarcinomas was localised in the middle and two thirds in the distal colon. They had the appearance of mucinous adenocarcinomas and abundantly expressed MUC2 mucin (supplementary Fig. S2 A).

The expression of β-catenin, p16 and p27 proteins, which exhibit characteristic alterations in the epitelium of patients with ulcerative colitis, was further investigated in the observed lesions.

β-catenin, one of the central molecules of the wnt pathway, was detectable in the plasma membrane of the normal epithelial cells (not shown), in the inflamed mucosa (Fig. 1b), dysplasia (not shown) and in the squamous carcinomas (supplementary Fig. S1b), while in some areas of all adenomas (Fig. 1e, arrows) and adenocarcinomas (Fig. 1h, arrow) it was translocated into the cytoplasm and the nucleus. The translocation of β-catenin, indicating its transcriptional activity, was described previously in colitis-associated human and murine carcinomas.19, 20

The gene coding for cyclin-dependent kinase inhibitor p16 is frequently silenced in ulcerative colitis-associated dysplasia through promoter methylation.21 We observe on the protein level that the strong nuclear expression, visible in the normal colonic mucosa (not shown), and in adenomas (Fig. 1f) is partly or—in about 1/3 of the cases—completely suppressed in the epithelial cells in the inflamed areas (Fig. 1c), in the adenocarcinomas (Fig. 1i) and squamous carcinomas (supplementary Fig. S1 C). The decrease of the mean intensity of p16 expression (Table I) was in all 3 tissues significant (p < 0.01). The expression of p16 in dysplasia, however, corresponded to that of the normal mucosa (Table I).

Table I. Summary of The Alterations in Localisation of β-Catenin and in The Intensity of Expression of p16 and p27 Proteins
Groupβ-catenin localisation (% cases)Intensity of nuclear epithelial p16 expressionIntensity of nuclear epithelial p27 expression
  1. NOR, Normal; INF, inflammatory; PSE, pseudopolyp; DYS, dysplasia; ADE, adenoma; ADE-CA, adenocarcinoma; SQ-CA, squamous carcinoma epithelium in the three animal groups. M, membrane; C, cytosol; N, nucleus. The intensity of expression was evaluated in all available tissues and the mean intensity in each type of tissue was calculated for every group.

BasalM (100)M (100)M (100)M (100)M (100)C + N (100)C+N (71)22220.70.81221.
Basal + CAM (100)M (100)M (100)M (100)M (100)C + N (100)C+N (100)22221.21.512221.710.51.5
Basal + UDCAM (100)M (100)M (100)M (100)M (100)C + N (100)22221.01.5221.5222.0

The expression of the cyclin-dependent kinase inhibitor p27 was strongest in the superficial, terminally differentiated cells in the normal epithelium (Fig. 2a). In about 80% of the dysplastic areas, a strong expression was visible, the remaining lesions exhibited partial or complete loss of p27 (Fig. 2e). In adenomas, it was homogeneously distributed throughout the epithelium (not shown). In the epithelium of the inflamed mucosa (Fig. 2b, arrow), in adenocarcinomas (Fig. 2d) and in squamous carcinomas (supplementary Fig. S3 A) the p27 protein was expressed weakly or not at all. The decrease of the mean p27 expression intensity (Table I) from 2 to 0.9 (in INF), 0.8 (in ADE-CA) and 1.3 (in SQ-CA) was significant in all 3 types of tissues (p < 0.03). Similar changes in p27 expression pattern were previously shown in human inflammatory bowel disease-associated dysplasia and carcinomas.22

thumbnail image

Figure 2. Expression of p27 protein is affected by treatment with UDCA. p27 protein expression is strong in (a) the normal mucosa and weak in (b) the inflamed regions (arrow) and (d): carcinoma. In (e) dysplasia a strong as well as a weak (arrow) expression was observed. In UDCA-treated mice the expression of p27 in (c) the inflamed regions and frequently in (f) dysplasia (white arrow) remains at the normal level. Bar: μm.

Download figure to PowerPoint

The observed alterations in all 3 groups are summarized in Table I.

The expression and distribution of β-catenin was membranous in the normal epithelium, the inflamed areas, dysplasia, the pseudopolyps and squamous carcinomas, while the epithelium of adenomas and adenocarcinomas exhibited translocation into the cytosol and the nucleus. Strong nuclear expression of p16 and p27 proteins was present in the normal epithelium, pseudopolyps and adenomas, while a weak staining pattern in the non-UDCA supplemented groups was noted in the epithelium of the inflammatory regions, adenocarcinomas and squamous carcinomas.

Together, these data indicate that the phenotype of the observed colitis-associated adenocarcinomas in mice corresponds to that reported in human ulcerative colitis.

UDCA treatment does not affect chronic inflammation orhyperplasia

Colitis was evident in all 3 groups, with the colitis index increasing from 12 to about 50 from proximal to distal segments. Colonic prolaps was not observed in any group. Neither the treatment with CA nor with UDCA had an effect on the grade or on the extent of inflammation, although the weight loss at the end of the experiment was significant in both CA and UDCA groups. This was associated with a decreased total food intake, which in both groups amounted to about 70% of the intake of the basal diet not supplemented with bile acids. The food intake and the average weight did not significantly differ between the CA and UDCA groups. The multiplicity of postinflammatory pseudopolyps, which do not represent and do not lead to premalignant growth, also did not significantly differ among the groups (Table II).

Table II. Effect of CA- or UDCA-Addition to The Diet of DSS-Colitis Mice on Their Survival, Colitis Index, Weight Development and The Prevalence of Pseudopolyps1
DietSurviving mice (%)p vs. basalColitis indexTotal colitisindexWeight(g) ± SDp vs. basalPseudopolyp multiplicity (±SD)
  • 1

    p Values were obtained through comparison with the basal group in z-test (survival) or t-test (weight).

Basal24 (96)12424610023.9 ± 1.82.1 ± 1.1
Basal + CA18 (72)NS16474711021.9 ± 2.80.0112.3 ± 0.9
Basal + UDCA20 (80)NS10514610720.4 ± 2.6<0.0012.0 ± 0.8

UDCA treatment decreases the prevalence of carcinomasand affects the tumour phenotype

The prevalence of dysplasia in the basal, basal + CA and basal + UDCA diet group was 88%, 67% and 40%, respectively. All adenomas were of the tubular type and appeared with a similar prevalence (17–20%) in all 3 groups (Table III). Adenocarcinomas, of predominantly mucinous phenotype, occurred in 46% of the mice fed with the basal diet and in 28% of the basal + CA diet group (p > 0.05) but were completely absent in the basal + UDCA group. The multiplicity of neoplasia (adenoma + carcinoma) was similar in the basal (1.25) and basal + CA (1.11) groups, and was significantly lower in the basal + UDCA diet group (0.45, p < 0.001)). Additionally, squamous carcinoma were found in 54% of the basal diet group and 50% of the basal + CA diet group, but only in 20% of the basal + UDCA diet group.

Table III. Effect of CA- or UDCA- Addition to The Diet of DSS-Colitis Mice on The Prevalence and Themultiplicity of Dysplasia, Adenomas and Carcinomas1
DietNo ofmice (%)Mice withDYS (%)Mice withtubularADE (%)GradeMice withADE-CA (%)p vs. basalMice withSQ-CA (%)p vs. basalNeoplasiamultiplicity
  • 1

    The p values were obtained through comparison with the basal group in z-test. NS, not significant.

Basal24 (100)21 (88)4 (17)1.811 (46)13 (54)1.25
Basal + CA18 (100)12 (67)4 (22)1.45 (28)n.s9 (50)NS1.11
Basal + UDCA20 (100)8 (40)4 (20)2.00 (0)<0.0014 (20)0.0400.45

Thus UDCA significantly suppressed the development of dysplasia and of squamous carcinomas and completely inhibited the emergence of adenocarcinomas. Surprisingly, the diet supplementation with CA which, based on the data on its cocarcinogenic effect in chemical carcinogenesis,16 was expected to increase the prevalence of carcinomas, had no significant effect on the prevalence of carcinomas.

The expression of β-catenin and the frequency and extent of its translocation to the cytosol and the nucleus were not affected by treatment with CA or with UDCA in all tissues investigated (Table I). Similarly, the expression of p16 protein and its downregulation in the inflamed regions and in the neoplastic tissue were not significantly altered by CA or UDCA (Table I).

By contrast, the loss of expression of the cyclin-dependent kinase inhibitor p27 in the inflamed regions and in squamous carcinomas was prevented by UDCA (but not by CA) treatment: i.e., in the UDCA-treated mice there was no loss of p27 expression neither in the inflamed tissue (Fig. 2c) nor in the squamous carcinomas (supplementary Fig. S3 B).

Fecal UDCA concentration shows an inverse relationshipto tumour prevalence

The water-soluble fraction of the bile acids is most likely to be responsible for the cellular effects in the colonic epithelium. The time course of DCA and UDCA concentration in the fecal water during the entire observation period was therefore investigated in each group (Fig. 3a3c). As an integral measure of the fecal water concentration during the whole experiment, a mean value of the concentrations at all time points was calculated (Table IV).

thumbnail image

Figure 3. Time course of the concentration of DCA (squares) and UDCA (dots) in (a) the basal, (b) basal + CA and (c) basal + UDCA group of mice in the fecal water of pooled faeces of each group of animals.

Download figure to PowerPoint

Table IV. Mean Values (±SD) of The Fecal Water Concentration (μM) of DCA and UDCA in The 3 Animal Groups During The Observation Period
GroupDCA (μM)UDCA (μM)
Basal6 ± 39 ± 4
Basal + CA131 ± 4212 ± 5
Basal + UDCA1.1 ± 0.46585 ± 1797

The fecal water concentration of DCA and of UDCA during the whole experiment was relatively stable in each group, with S.D. not exceeding 50% of the mean (Table IV), notwithstanding the variation of individual measurements. In the basal group, the mean values of DCA and UDCA were within the range previously reported for non-treated mice.23, 24, 25 There was no significant difference in the mean fecal water concentration of UDCA in the basal (9 ± 4 μM) and the basal + CA (12 ± 5 μM) group over the entire time span of the experiment. By contrast, the mean concentration of UDCA in the basal + UDCA group was 6585 ± 1797 μM.

The prevalence of dysplasia, adenocarcinomas as well as squamous carcinomas and the mean fecal water concentration of UDCA showed an inverse relationship in the 3 groups (Fig. 4a). Indeed, there was a significant correlation between the logarithm of the fecal water concentration and the prevalence of dysplasia (Pearson correlation coefficient −0.920, p = 0.035) or the prevalence of the squamous carcinoma (Pearson correlation coefficient −0.998, p = 0,037). There was no straightforward relationship between the lesion prevalence and the concentration of DCA or the DCA/UDCA concentration ratio (data not shown).

thumbnail image

Figure 4. (a) The prevalence of dysplasia (white bars), adenocarcinomas (black bars) and squamous carcinomas (hatched bars) decreases as the fecal water concentration of UDCA increases. (b) The increase of DCA concentration in fecal water above the normal level (6 μM) does not significantly affect the prevalence of adenocarcinomas or squamous carcinomas.

Download figure to PowerPoint

After the increase of DCA concentration from 6 to 131 μM (Fig. 4b), there was no change in the prevalence of squamous carcinomas and a statistically not significant decrease of adenocarcinomas, indicating that under the conditions of experimental colitis DCA has no cocarcinogenic effect.


  1. Top of page
  2. Abstract
  3. Material and method
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

The present work demonstrates for the first time that the development of the murine colitis-associated colorectal adenocarcinomas can be inhibited by oral treatment with UDCA. It supports the previous retrospective studies on patients with PSC, who showed a lower risk of developing colon cancer when treated with UDCA, and offers a model for a detailed investigation of the mechanism of UDCA action.

The present model of DSS colitis-associated colon carcinogenesis exhibits several features of the etiology, tumor type, distribution and gene expression, similar to that of human colitis-associated carcinomas. The development of tumors in this model is slow, caused by the chronic inflammation alone, with no additional activation of signaling pathways by a high dose of a chemical carcinogen. According to the original report, the 2-fold iron-enriched diet enhances colitis and increases carcinoma prevalence.12 The frequent occurrence of adenocarcinomas of mucinous type observed in the present work is also found among human tumours associated with ulcerative or radiation colitis (30–50%) (reviewed in ref. 26). The prevalence of squamous carcinomas (54%) was, however, higher than in patients with colitis, in whom squamous carcinomas of the anal region are rare.27 These data indicate that in the present model, the prevalence of carcinoma types varies from that in the human colitis.

The altered localisation of β-catenin and the suppression of p16 and p27, observed in our study suggests a similar gene alteration profile as previously reported in colitis-associated carcinomas in humans,19, 21, 22 and we were able to demonstrate downregulation of p16 on the protein level.

Furthermore, the tumours in the present model appear to exhibit morphogenesis analogous to that of the human colitis-associated neoplasms. About 70% of the human sporadic colon carcinomas are assumed to develop from the adenoma-carcinoma sequence and 30% de novo, without an polypoid precursor.28, 29 In colitis, polypoid growth of adenomas is rare and the transition steps from flat colitis-associated intramucosal dysplasia to invasive carcinoma are not sharply delineated.

In the present model, adenomas most frequently had a sessile or pedunculated polypoid form and a tubular growth pattern. Their prevalence did not correlate with the prevalence of carcinomas. Furthermore, the expression of p16 and p27 proteins in the inflamed tissue showed a pattern similar to the adenocarcinomas rather than to the adenomas. The prevalence of dysplasia was, however, decreasing with the increase of UDCA concentration in fecal water. Together, these data are compatible with the hypothesis of a frequent transition from the inflamed and regenerating mucosa to intramucosal dysplasia and then to adenocarcinoma, without a polypoid adenoma intermediate.

On the other hand, the activation of the wnt pathway, evident through the translocation of β-catenin from the membrane to the cytosol and to the nucleus, was common in carcinomas and adenomas but was not detected in the epithelium of the inflamed tissue. This may be due to the comparatively small number of epithelial cells accessible to evaluation in the inflamed lesions.

The squamous carcinomas always retained the membranous β-catenin localisation. Together, these data indicate that β-catenin activation was not associated with the development of squamous carcinomas and its potential contribution to the development of adenocarcinomas could have occurred at a postinflammatory stage.

The cellular and molecular targets of the chemopreventive action of UDCA remain obscure. UDCA was previously shown to attenuate acute inflammation in TNBS colitis.30 In the present work, it did not affect the chronic inflammation, i.e. the modulation of inflammation was not the basis of the chemopreventive effect.

In fact, UDCA was shown to inhibit inflammation-independent tumor development in MIN mice,31 as well as in rats treated with AOM5, 6, 32, 33 or DMH,34 all of which are associated with the activation of the wnt pathway.34, 35 It is therefore possible, that the chemoprotection observed in the present work was related to inhibition of a downstream target of this pathway. This hypothesis is supported by a less efficient suppression of squamous carcinomas, in which the wnt pathway is not activated.

The inflammation-related suppression of the cyclin dependent kinase inhibitor p27 was reversed by UDCA treatment. In the dysplastic areas, the expression of p27 and the effects of UDCA were not consistent; this is compatible with the hypothesis that the dysplasias would further develop along different pathways. The reduced p27 expression was previously reported to enhance proliferation of gastrointestinal tumors in DMH-treated and in MIN mice and to cooperate with the activated wnt pathway.36 Whether a similar mechanism is affected by UDCA treatment during ulcerative colitis warrants further study.

The supplementation of the diet with the bile acids resulted, probably due to their bitter taste, in a decreased food intake and weight loss. This unspecific effect was observed in both supplemented groups, while the chemoprevention was specific for UDCA.

The lack of procarcinogenic action of CA was not expected, since the addition of 0.2% of CA to the diet has been previously shown to increase the prevalence of AOM-induced aberrant crypt foci37 as well as MNU-induced colon tumours.16 Other authors, however, observed the significant increase of AOM-induced tumours only at 0.4% but not 0.2% CA.6 These data suggests that the cocarcinogenic effect of 0.2% CA is weak and if the DSS treatment is a less potent activator of carcinogenic pathways than AOM or MNU, the applied CA dose was not sufficient to enhance inflammation-induced carcinogenesis. (Fig. 4b).

UDCA was suggested to prevent chemical carcinogenesis by decreasing the proportion of the cocarcinogenic DCA in the fecal water.25 In the present work, we did not observe any correlation between the decrease of DCA/UDCA concentration ratio and tumour prevalence (data not shown), which further underpins the difference between the chemical and the colitis-associated carcinogenesis.

In conclusion, the present model demonstrates the complete inhibition of colitis-associated colon adenocarcinomas by UDCA and offers a suitable tool for an in-depth analysis of the mechanistic aspects of anticarcinogenic effects of UDCA in vivo.


  1. Top of page
  2. Abstract
  3. Material and method
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

This project has been supported by Broad Foundation, Los Angeles, USA (to C.H.) and the Z1SFB633 (to H.S. and M.Z.); UDCA was kindly donated by the Falk Foundation, Freiburg, Germany. The authors thank Dr. Ingrid B. Renes, Laboratory of Pediatrics, Erasmus MC Sophia, Rotterdam, The Netherlands, for the kind donation of the rabbit anti-MUC2 antiserum and Simone Spieckermann for excellent technical assistance.


  1. Top of page
  2. Abstract
  3. Material and method
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information
  • 1
    Eaden JA, Abrams KR, Mayberry JF. The risk of colorectal cancer in ulcerative colitis: a meta-analysis. Gut 2001; 48: 52635.
  • 2
    Itzkowitz SH, Harpaz N. Diagnosis and management of dysplasia in patients with inflammatory bowel diseases. Gastroenterology 2004; 126: 163448.
  • 3
    Ejderhamn J, Rafter JJ, Strandvik B. Faecal bile acid excretion in children with inflammatory bowel disease. Gut 1991; 32: 134651.
  • 4
    Milovic V, Teller IC, Faust D, Caspary WF, Stein J. Effects of deoxycholate on human colon cancer cells: apoptosis or proliferation. Eur J Clin Invest 2002; 32: 2934.
  • 5
    Wali RK, Frawley BP,Jr, Hartmann S, Roy HK, Khare S, Scaglione-Sewell BA, Earnest DL, Sitrin MD, Brasitus TA, Bissonnette M. Mechanism of action of chemoprotective ursodeoxycholate in the azoxymethane model of rat colonic carcinogenesis: potential roles of protein kinase C-alpha, -beta II, and -zeta. Cancer Res 1995; 55: 525764.
  • 6
    Earnest DL, Holubec H, Wali RK, Jolley CS, Bissonette M, Bhattacharyya AK, Roy H, Khare S, Brasitus TA. Chemoprevention of azoxymethane-induced colonic carcinogenesis by supplemental dietary ursodeoxycholic acid. Cancer Res 1994; 54: 50714.
  • 7
    Jacoby RF, Cole CE, Hawk ET, Lubet RA. Ursodeoxycholate/Sulindac combination treatment effectively prevents intestinal adenomas in a mouse model of polyposis. Gastroenterology 2004; 127: 83844.
  • 8
    Balan V, LaRusso NF. Hepatobiliary disease in inflammatory bowel disease. Gastroenterol Clin North Am 1995; 24: 64769.
  • 9
    Schurmann G, Ochman S, Neurath MF. [ Colonic carcinoma associated with ulcerative colitis. Risk factors, molecular pathogenesis and surveillance]. Dtsch Med Wochenschr 2000; 125: 104550.
  • 10
    Pardi DS, Loftus EV,Jr, Kremers WK, Keach J, Lindor KD. Ursodeoxycholic acid as a chemopreventive agent in patients with ulcerative colitis and primary sclerosing cholangitis. Gastroenterology 2003; 124: 88993.
  • 11
    Tung BY, Emond MJ, Haggitt RC, Bronner MP, Kimmey MB, Kowdley KV, Brentnall TA. Ursodiol use is associated with lower prevalence of colonic neoplasia in patients with ulcerative colitis and primary sclerosing cholangitis. Ann Intern Med 2001; 134: 8995.
  • 12
    Seril DN, Liao J, Ho KL, Warsi A, Yang CS, Yang GY. Dietary iron supplementation enhances DSS-induced colitis and associated colorectal carcinoma development in mice. Dig Dis Sci 2002; 47: 126678.
  • 13
    Tanaka T, Kohno H, Suzuki R, Yamada Y, Sugie S, Mori H. A novel inflammation-related mouse colon carcinogenesis model induced by azoxymethane and dextran sodium sulfate. Cancer Sci 2003; 94: 96573.
  • 14
    Kohno H, Suzuki R, Sugie S, Tanaka T, Yamada Y, Mori H. Beta-Catenin mutations in a mouse model of inflammation-related colon carcinogenesis induced by 1,2-dimethylhydrazine and dextran sodium sulfate: a novel inflammation-related mouse colon carcinogenesis model induced by azoxymethane and dextran sodium sulfate. Cancer Sci 2005; 96: 6976.
  • 15
    Takahashi M, Fukuda K, Sugimura T, Wakabayashi K. Beta-catenin is frequently mutated and demonstrates altered cellular location in azoxymethane-induced rat colon tumors. Cancer Res 1998; 58: 426.
  • 16
    McSherry CK, Cohen BI, Bokkenheuser VD, Mosbach EH, Winter J, Matoba N, Scholes J. Effects of calcium and bile acid feeding on colon tumors in the rat. Cancer Res 1989; 49: 603943.
  • 17
    Neurath MF, Fuss I, Kelsall BL, Stuber E, Strober W. Antibodies to interleukin 12 abrogate established experimental colitis in mice. J Exp Med 1995; 182: 128190.
  • 18
    Keller S, Jahreis G. Determination of underivatised sterols and bile acid trimethyl silyl ether methyl esters by gas chromatography-mass spectrometry-single ion monitoring in faeces. J Chromatogr B Analyt Technol Biomed Life Sci 2004; 813: 199207.
  • 19
    Aust DE, Terdiman JP, Willenbucher RF, Chew K, Ferrell L, Florendo C, Molinaro-Clark A, Baretton GB, Lohrs U, Waldman FM. Altered distribution of beta-catenin, and its binding proteins E-cadherin and APC, in ulcerative colitis-related colorectal cancers. Mod Pathol 2001; 14: 2939.
  • 20
    Cooper HS, Murthy S, Kido K, Yoshitake H, Flanigan A. Dysplasia and cancer in the dextran sulfate sodium mouse colitis model. Relevance to colitis-associated neoplasia in the human: a study of histopathology, B-catenin and p53 expression and the role of inflammation. Carcinogenesis 2000; 21: 75768.
  • 21
    Hsieh CJ, Klump B, Holzmann K, Borchard F, Gregor M, Porschen R. Hypermethylation of the p16INK4a promoter in colectomy specimens of patients with long-standing and extensive ulcerative colitis. Cancer Res 1998; 58: 39425.
  • 22
    Walsh S, Murphy M, Silverman M, Odze R, Antonioli D, Goldman H, Loda M. p27 expression in inflammatory bowel disease-associated neoplasia. Further evidence of a unique molecular pathogenesis. Am J Pathol 1999; 155: 15118.
  • 23
    Kasbo J, Saleem M, Perwaiz S, Mignault D, Lamireau T, Tuchweber B, Yousef I. Biliary, fecal and plasma deoxycholic acid in rabbit, hamster, guinea pig, and rat: comparative study and implication in colon cancer. Biol Pharm Bull 2002; 25: 13814.
  • 24
    Barone M, Berloco P, Ladisa R, Ierardi E, Caruso ML, Valentini AM, Notarnicola M, Di LA, Francavilla A. Demonstration of a direct stimulatory effect of bile salts on rat colonic epithelial cell proliferation. Scand J Gastroenterol 2002; 37: 8894.
  • 25
    Batta AK, Salen G, Holubec H, Brasitus TA, Alberts D, Earnest DL. Enrichment of the more hydrophilic bile acid ursodeoxycholic acid in the fecal water-soluble fraction after feeding to rats with colon polyps. Cancer Res 1998; 58: 16847.
  • 26
    Hanski C. Is mucinous carcinoma of the colorectum a distinct genetic entity? Br J Cancer 1995; 72: 13506.
  • 27
    Frisch M, Johansen C. Anal carcinoma in inflammatory bowel disease. Br J Cancer 2000; 83: 8990.
  • 28
    Bedenne L, Faivre J, Boutron MC, Piard F, Cauvin JM, Hillon P. Adenoma–carcinoma sequence or “de novo” carcinogenesis? A study of adenomatous remnants in a population-based series of large bowel cancers. Cancer 1992; 69: 8838.
  • 29
    Chen CD, Yen MF, Wang WM, Wong JM, Chen TH. A case-cohort study for the disease natural history of adenoma-carcinoma and de novo carcinoma and surveillance of colon and rectum after polypectomy: implication for efficacy of colonoscopy. Br J Cancer 2003; 88: 186673.
  • 30
    Kullmann F, Arndt H, Gross V, Ruschoff J, Scholmerich J. Beneficial effect of ursodeoxycholic acid on mucosal damage in trinitrobenzene sulphonic acid-induced colitis. Eur J Gastroenterol Hepatol 1997; 9: 120511.
  • 31
    Cooper HS, Everley L, Chang WC, Pfeiffer G, Lee B, Murthy S, Clapper ML. The role of mutant Apc in the development of dysplasia and cancer in the mouse model of dextran sulfate sodium-induced colitis. Gastroenterology 2001; 121: 140716.
  • 32
    Khare S, Cerda S, Wali RK, von Lintig FC, Tretiakova M, Joseph L, Stoiber D, Cohen G, Nimmagadda K, Hart J, Sitrin MD, Boss GR, et al. Ursodeoxycholic acid inhibits Ras mutations, wild-type Ras activation, and cyclooxygenase-2 expression in colon cancer PKC-delta inhibits anchorage-dependent and -independent growth, enhances differentiation, and increases apoptosis in CaCo-2 cells. Cancer Res 2003; 63: 351723.
  • 33
    Wali RK, Khare S, Tretiakova M, Cohen G, Nguyen L, Hart J, Wang J, Wen M, Ramaswamy A, Joseph L, Sitrin M, Brasitus T, et al. Ursodeoxycholic acid and F(6)-D(3) inhibit aberrant crypt proliferation in the rat azoxymethane model of colon cancer: roles of cyclin D1 and E-cadherin. Cancer Epidemiol. Biomarkers Prev 2002; 11: 165362.
  • 34
    Koesters R, Hans MA, Benner A, Prosst R, Boehm J, Gahlen J, Doeberitz MK. Predominant mutation of codon 41 of the beta-catenin proto-oncogene in rat colon tumors induced by 1,2-dimethylhydrazine using a complete carcinogenic protocol. Carcinogenesis 2001; 22: 188590.
  • 35
    Aoki H, Ohnishi K, Wang X, Takahashi A, Ohnishi T, Nakamura M, Sakaki T. p53-independent WAF1 induction by ACNU in human glioblastoma cells. Mol Carcinog 1998; 21: 1716.
  • 36
    Philipp-Staheli J, Kim KH, Payne SR, Gurley KE, Liggitt D, Longton G, Kemp CJ. Pathway-specific tumor suppression. Reduction of p27 accelerates gastrointestinal tumorigenesis in Apc mutant mice, but not in Smad3 mutant mice. Cancer Cell 2002; 1: 35568.
  • 37
    Baijal PK, Clow EP, Fitzpatrick DW, Bird RP. Tumor-enhancing effects of cholic acid are exerted on the early stages of colon carcinogenesis via induction of aberrant crypt foci with an enhanced growth phenotype. Can J Physiol Pharmacol 1998; 76: 1095102.

Supporting Information

  1. Top of page
  2. Abstract
  3. Material and method
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

This article contains supplementary material available via the Internet at .

jws-ijc.21729.fig1.jpg1442KSupporting Information file jws-ijc.21729.fig1.jpg
jws-ijc.21729.fig2.jpg1604KSupporting Information file jws-ijc.21729.fig2.jpg
jws-ijc.21729.fig3.jpg1660KSupporting Information file jws-ijc.21729.fig3.jpg

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.