Skip Navigation


Carcinogenesis Advance Access originally published online on May 17, 2007
Carcinogenesis 2007 28(11):2367-2374; doi:10.1093/carcin/bgm119
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
28/11/2367    most recent
bgm119v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by De Miglio, M. R.
Right arrow Articles by Feo, F.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by De Miglio, M. R.
Right arrow Articles by Feo, F.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author 2007. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org

Identification and chromosome mapping of loci predisposing to colorectal cancer that control Wnt/ß-catenin pathway and progression of early lesions in the rat

Maria R. De Miglio, Patrizia Virdis, Diego F. Calvisi, Daniela Mele, Maria R. Muroni, Maddalena Frau, Federico Pinna, Maria L. Tomasi, Maria M. Simile, Rosa M. Pascale and Francesco Feo*

Department of Biomedical Sciences, Division of Experimental Pathology and Oncology, University of Sassari, 07100 Sassari, Italy

* To whom correspondence should be addressed. Tel: +0039 079 228307; Fax: +0039 079 228485; Email: feo{at}uniss.it


   Abstract
Top
 Abstract
Introduction
 Materials and methods
 Results
 Discussion
 Supplementary material
 Funding
 References
 
Sporadic colorectal cancer (CRC) is a major health concern worldwide. Epidemiologic evidence suggests a polygenic predisposition to CRC, but the genes responsible remain unknown. Here, we performed genome-wide scanning of male (ACI/SegHsd x Wistar-Furth)F2 (AWF2) rats to map susceptibility genes influencing the evolution of early colorectal lesions to adenocarcinoma following 1,2-dimethylhydrazine administration. Phenotypic analysis revealed higher incidence/multiplicity and lower size of adenomas in ACI/SegHsd (ACI) and (ACI/SegHsd x Wistar-Furth)F1 (AWF1) than Wistar-Furth (WF) rats and higher incidence/multiplicity of poorly differentiated adenocarcinomas in WF than ACI rats, with intermediate values in AWF1 rats. Linkage analysis of 138 AWF2 rats identified three loci on chromosomes 4, 15 and 18 in significant linkage with lesion multiplicity that were identified as rat Colon cancer resistance (rCcr) 1, rCcr2 and rCcr3, respectively. Seven other loci on chromosomes 5, 6, 15, 17, 18 and 20 were in suggestive linkage with adenoma/adenocarcinoma multiplicity/surface area. Six of them were identified as rCcr4–9 and a locus on chromosome 5 was identified as a susceptibility locus, rCcs1. Significant interactions between rCcr3 and rCcr6, rCcr6 and rCcr8 and rCcr5 and rCcr9, and four novel epistatic loci controlling multiplicity/size of colorectal lesions were discovered. Apc, located at rCcr3, did not show functional promoter polymorphisms. However, influence of susceptibility/resistance genes on Wnt/ß-catenin pathway was shown by defective ß-catenin inactivation in WF but not in ACI and AWF1 rat adenocarcinomas. These data indicate that inheritance of predisposition to CRC depends on interplays of several genetic factors, and suggest a possible mechanism of polygenic control of CRC progression.

Abbreviations: ACF, aberrant crypt focus; ACI, ACI/SegHsd; ANOVA, analysis of variance; APC, adenomatous polyposis coli; AWF1, (ACI/SegHsd x Wistar-Furth)F1; AWF2, (ACI/SegHsd x Wistar-Furth)F2; CRC, colorectal cancer; DMH, 1,2-dimethylhydrazine; LoD, logarithm of differences; PCR, polymerase chain reaction; PD, poorly differentiated; QTL, quantitative trait locus; rCcr, rat colorectal cancer resistance; rCcs, rat colorectal cancer susceptibility; Scc, susceptibility to colon cancer; WD, well differentiated; WF, Wistar-Furth


   Introduction
Top
 Abstract
 Introduction
Materials and methods
 Results
 Discussion
 Supplementary material
 Funding
 References
 
Colorectal cancer (CRC) is a common disease, both in men and women, worldwide (1). Environmental risk factors include diet and lifestyle (2). Sporadic, inherited and familial CRC have been identified (2,3). Inherited syndromes, occurring in <10% of patients, include familial adenomatous polyposis and hamartomatous polyposis, hereditary non-polyposis CRC (Lynch syndrome) and cancer family syndrome (Lynch-like syndrome) (4).

Sporadic disease accounts for ~70% of CRCs. In some families, CRC develops too frequently to be considered sporadic, but not in a pattern consistent with an inherited syndrome (1). This envisages a role of low-penetrance predisposing genes, each partly contributing to CRC predisposition. Various genes, including adenomatous polyposis coli (APC), TGF-ßR1, H-RAS1, MTHFR, BLM, HFE, GSTT1 and CCND1 have been found to be involved in the pathogenesis of sporadic CRC (2,5). However, their relationship with polygenic predisposition to CRC is unclear.

Mapping of low-penetrance susceptibility genes is difficult in the human population due to its heterogeneity and the interactions between genes and environmental factors. Consequently, experimental models are currently used to map susceptibility-related genes in quantitative trait loci (QTLs). Various studies indicate that the genetic variants controlling susceptibility to complex diseases, including cancer, often map to orthologous regions of rodent and human genomes, and various diseases are caused by polymorphism of equivalent rodent and human genes (6,7). Thus, preliminary mapping of susceptibility-related genes in rodent models allows defining the genetic model, identifying candidate modifier genes and evaluating their phenotypic and molecular effects. This information is essential to attempt gene cloning and identifying the genetic mechanisms of the human disease.

Researches on mouse have led to the identification of Colon cancer susceptibility (Ccs) loci 1 and 2 on chromosomes 12 and 3, respectively, affecting the susceptibility to 1,2-dimethylhydrazine (DMH)-induced CRC (8,9). Various congenic mice strains exhibit different patterns of susceptibility to aberrant crypt foci (ACFs) and colon adenomas, indicating the involvement of different subsets of genes (10). The inheritance of the susceptible parental alleles susceptibility to colon cancer (Scc) 1 or Scc2, on chromosome 2, and the resistance allele Scc3, on chromosome 1, affects the susceptibility to DMH-induced CRC (11,12). Scc4 (chromosome 17) and Scc5 (Chromosome 18), not displaying individual phenotypic effects, show reciprocal interactions (12). Interestingly, Ptprj gene, located at Scc1, encoding a receptor-type protein tyrosine phosphatase, is a candidate modifier of colon carcinogenesis in mice (13) and is frequently deleted in human cancers, including CRC.

Researches on the genetic susceptibility of rats to CRC are scanty. No correlation was found between Apc and Pla2g2a gene polymorphisms and the susceptibility to ACFs development in rats treated with 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (14). On the other hand, a locus on chromosome 16, in significant linkage with ACF formation, was identified (15).

Colorectal carcinogenesis is a multistage process with a well-defined sequence of events from aberrant crypt proliferation to benign adenoma, well-differentiated (WD) adenocarcinoma and finally poorly differentiated (PD) adenocarcinoma. Available data do not allow identifying the stages of colorectal carcinogenesis controlled by the loci discovered so far. The analysis of the development of ACFs, commonly considered putative pre-cancerous lesions (16), may give insights on the genes controlling the initiation stage in rats. Only a relatively small subset of ACFs undergo further evolution to PD adenocarcinoma, and it is not clear whether the locus discovered on chromosome 16 (15) influences the evolution of initiated cells to cancer. To determine the map location of tumor susceptibility genes influencing the evolution of early colorectal lesions to PD adenocarcinoma and further enlighten the pathogenesis of sporadic CRC, we performed a genome-wide scanning on a group of (ACI/SegHsd x Wistar-Furth)F2 (AWF2) rats, in which the development of adenomas and adenocarcinomas has been induced with DMH. New QTLs and epistatic loci in linkage with the propensity of colorectal adenomas to progress to malignancy have been discovered. Also, we show that at least some of the loci identified affect the Wnt/ß-catenin pathway in pre-malignant and malignant colorectal lesions of susceptible rats.


   Materials and methods
Top
 Abstract
 Introduction
 Materials and methods
Results
 Discussion
 Supplementary material
 Funding
 References
 
Animals
Twenty ACI/SegHsd (ACI), Wistar-Furth (Charles-River Italia, Calco, Italy), (ACI/SegHsd x Wistar-Furth)F1 (AWF1) and 138 AWF2 rats (180-200 g) were fed a standard diet (type 48, Piccioni, Gessate, Italy) and tap water ad libitum. They were housed individually in a room with constant temperature (22°C) and humidity (55%) and with a 12 h light (6 a.m.–6 p.m.)–dark cycle. Study protocols were in compliance with our institution's guidelines for the use of laboratory animals.

Phenotyping
Colorectal adenomas, WD and PD adenocarcinomas were induced by injecting subcutaneously DMH (20 mg/kg body wt) weekly for 26 weeks. Control (untreated) and DMH-treated rats were killed 32 weeks after initiation by bleeding through thoracic aorta under ether anesthesia. Colorectal lesions present in the entire length of colon and rectum appeared as spherical or ovoid lesions with a velvety smooth or slightly bossy appearance. Lesions were counted and a caliper was used to determine their maximum and minimum diameter. Lesion thickness could not be accurately evaluated due to the sessile feature of the majority of lesions, and the surface area was taken as a measure of the lesion size. For microscopic evaluation, small pieces of isolated lesions were processed for hematoxylin–eosin staining and ß-catenin immunohistochemistry as published (17), using an anti-ß-catenin monoclonal antibody (Transduction Laboratories, Lexington, KY), whose specificity has been previously tested (18).

Genotyping
Genomic DNA from spleens of intercross rats was extracted from isolated nuclei and purified (19). Genotyping was performed using 156 polymorphic microsatellite markers (20; http://rgd.mcw.edu/GENOMESCANNER/). The markers (Roche Diagnostic S.p.A., Monza, Italy) were distributed throughout all autosomes at an average density of one marker/7.3 cM, leaving two gaps of 26 and 36.3 cM on chromosome 19 and one gap of 26.8 cM on chromosome 3, with all other gaps being smaller than 25 cM. Marker polymorphism was revealed by polymerase chain reaction (PCR) analysis as published (19).

Nucleotide sequence analysis
Primers were designed to amplify rat ß-catenin exon 2 (GenBank accession no. AF397179), Apc promoter (GenBank accession no. AB071148) and all Apc exons (GenBank accession no. NM012499). DNA from Wistar-Furth (WF), ACI and AWF1 rats was amplified by PCR as published (21). Aliquots of the PCR products were purified by the ‘High Pure PCR Product Purification Kit’ (Roche Diagnostics, Mannheim, Germany), sequenced using an Alfexpress automated sequencer (GE Healthcare, Milano, Italy) and aligned and compared with identify polymorphisms. Gene data base searches were performed at the National Center for Biotechnology Information.

Quantitative reverse transcription–PCR
Primers for Apc, c-myc and RNR-18 were chosen using the ‘Assays-on-DemandTM Products’ (Applied Biosystems, Foster City, CA). PCRs were performed with 75–300 ng of cDNA, using the ABI Prism 7000 Sequence Detection System and the TaqMan Universal PCR Master Mix (Applied Biosystems) as published (17). Quantitative values were calculated using the PE Biosystems Analysis software.

Immunoblot analysis
Tissue samples were homogenized in lysis buffer and protein concentrations were determined as published (17). Whole-cell lysates (350 µg) from normal and neoplastic colorectal mucosa tissues were processed as reported (17). Immunoprecipitations were carried out with 6 µg of rabbit polyclonal anti-antibodies anti-ß-catenin, Axin2, phosphorylated ß-catenin (Serine 33), and ß-actin (SC-7199, SC-8570, SC-22192 and SC-1615, respectively; Santa Cruz Biotechnology, Santa Cruz, CA). Immunoprecipitated proteins were subjected to western blotting and membranes were probed with horseradish peroxidase–secondary antibodies. For western blot analysis, 100 µg of proteins were extracted, denatured, separated by sodium dodecyl sulfate–polyacrylamide gel electrophoresis, and transferred onto nitrocellulose membranes by electroblotting. Membranes were probed with goat polyclonal antibodies anti-Apc (SC-896, C-terminus epitope), followed by incubation with horseradish peroxidase–secondary antibody and revealed with Luminol Reagent. Densities of the protein bands were normalized to ß-actin levels and calculated by ImageQuant software.

Statistical analysis
Linkage maps were constructed using the MAPMAKER/EXP 3.0 program (22). The associations of tumor susceptibility (low responder versus high responder, as defined on the basis of multiplicity and surface area of lesions in parental strains) with alleles of rat microsatellite markers were evaluated by Logarithm of differences (LOD) score. Threshold LOD score values at 2.8 and 4.3 were considered for ‘suggestive’ and ‘significant’ linkage, respectively (23). The proportion of total intercross variability explained by the association between the marker and the trait (R2) was taken as an index of the importance of each locus. QTL analysis was carried out using parametric or non-parametric methods with MAPMAKER/QTL 1.1 (24). The ‘additive and dominant’ method of MAPMAKER/QTL and analysis of variance (ANOVA) (SAS Institute, Cary, NC) procedures were used to confirm QTL analysis and evaluate the allelic contribution to the phenotypes. Furthermore, genome-wide significance thresholds were established for each phenotype by 1000 permutation replicates by using the maximum likelihood method via an Expectation-maximization (EM) algorithm as implemented in the R/qtl mapping software (25). Empirical 5% significant thresholds were used. Potential interactions between genetic loci were evaluated by ANOVA and P values were corrected for multiple comparisons (22). Differences between parental strains and between homozygous and heterozygous intercross rats for phenotypic parameters were analyzed by ANOVA and multiple comparisons were made by the Tuckey–Kramer test using GraphPad InStat 3 (http://www.graphpad.com).


   Results
Top
 Abstract
 Introduction
 Materials and methods
 Results
Discussion
 Supplementary material
 Funding
 References
 
Inheritance of the susceptibility to colorectal carcinogenesis
WF rats are highly susceptible to CRC, which develops spontaneously in 38% of male rats at 6 months of age (26). ACI rats are phylogenetically distant from WF rats (27), and show resistance to the development of ACFs induced by 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (15).

Food intake and body weight did not show any interstrain difference throughout the experiment (data not shown). In all strains examined, tubular adenomas with columnar darkly stained dysplastic epithelium were detected, whereas villous adenomas were only rarely found. WD adenocarcinomas exhibited irregular glandular pattern with little evidence of mucous secretion and hyperchromic epithelial cells showing irregular, prominent nuclei. PD adenocarcinomas displayed low tendency to form glandular patterns, cell pleomorphism, large irregular nuclei and tendency to invade the submucosa.

We tested the AWF1 progeny and parental strains for the susceptibility to DMH-induced colorectal tumors. As reported in Table I, WF rats exhibited a higher incidence and multiplicity of PD adenocarcinomas and a lower incidence and multiplicity of adenomas and WD adenocarcinomas than ACI rats (although the differences regarding WD adenocarcinomas incidence and multiplicity did not reach significance threshold). This suggests the resistance of ACI rats to the progression of benign lesions to more advanced lesions, which was further substantiated by the relatively small surface area of ACI rat lesions, when compared with those of WF rats. In AWF1 rats, the pattern of colorectal adenoma development was similar to that of ACI rats, whereas WD and PD adenocarcinomas development resembled that of WF rats, excepting a lower surface area of the lesions.


View this table:
[in this window]
[in a new window]

  		Table I. Development of colorectal adenomas, WD adenocarcinomas (adenocarcinoma), and PD adenocarcinomas (carcinoma) in WF, ACI and AWF1 rats

 
Collectively, these findings suggest the existence of an intermediate susceptibility of AWF1 rats to the progression of benign lesions to WD and PD adenocarcinomas. As expected, the multiplicities and mean areas of adenomas, WD and PD adenocarcinomas of AWF2 rats were in the range of those found in parental strains (data not shown).

Linkage mapping of loci affecting the multiplicity and size of colorectal lesions
The multiplicity of lesions was normally distributed, whereas surface area values were not normally distributed in 138 AWF2 rats. Therefore, to obtain an improved normality (28), QTL analysis was performed using a parametric method for multiplicity of lesions and a non-parametric method for surface area. This analysis identified three QTLs in significant linkage with phenotypic parameters on chromosomes 4, 15 and 18 (Figure 1). The QTL on chromosome 4 exhibited a LOD score value of 4.3 in linkage with the surface area of adenomas. The QTL on chromosome 15 was in linkage with the multiplicity of total colorectal lesions (adenomas plus WD and PD adenocarcinomas) and showed a LOD score peak of 6.57, at 12 cM. LOD score was 4.3 when only PD adenocarcinoma multiplicity was considered. It should be noted that this QTL has been identified by using microsatellite markers relatively far from the QTL peak and reciprocally distant 23.7 cM, a distance which is analyzed by MapMaker/QTL. In addition, ANOVA gave still significant P values for these markers (Table II), although the degree of significance was not high due to the distance of the markers from the QTL peak. This situation makes us confident of the existence of this QTL, although not completely excluding the possibility of an artifact. Moreover, a LOD score peak of 2.9, suggestive of genetic linkage with PD adenocarcinoma multiplicity, was present on chromosome 15 at 63.8 cM. A significant QTL was identified on chromosome 18, with LOD score peaks of 4.3 and 5.02, at 0 cM for PD adenocarcinoma multiplicity and surface area, respectively. Another LOD score peak of 3.1 was found on the same chromosome at 19.5 cM in suggestive linkage with PD adenocarcinoma multiplicity. The presence of the QTLs on chromosomes 4, 15 and 18 was confirmed by the Expectation-maximization algorithm, which showed LOD score peaks equivalent to those calculated via MapMaker/QTL, and higher than the LOD thresholds calculated by permutation analysis, for each phenotype, for empirical 5% significance (Table II).


View this table:
[in this window]
[in a new window]

  		Table II. Summary of linkage analysis in intercross rats

 


Figure 1
View larger version (12K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 1. LOD score for susceptibility to colorectal neoplastic lesions in AWF2 rats with genetic markers spanning the length of chromosomes 4, 15 and 18. QTL analysis was performed using a parametric method for the multiplicity of lesions and a non-parametric method for surface area. On the abscissa, distance in cM from the first microsatellite marker to the markers used in each chromosome. Dashed horizontal lines, Lander–Kruglyak thresholds for significant (LOD score 4.3 in intercross) and suggestive (LOD score 2.8 for intercross) linkage.

 
Linkage analysis also showed other loci on chromosomes 5, 6, 17 and 20, with LOD score peak values suggestive of genetic linkage with multiplicity and/or average area of the lesions. The suggestivity of these QTLs was confirmed by the calculation of R/qtl values, for each phenotype, <5% significance threshold (Table II). Nevertheless, all these R/qtl values were higher than a 10% significance threshold (not shown).

On the basis of statistical considerations suggestive linkage may be wrong (23). Thus, we analyzed by ANOVA these loci together with the significant ones to evaluate whether the phenotypic features of the lesions in AWF2 rats are accounted for by the presence of susceptibility/resistance genes in the QTLs with suggestive linkage. We also analyzed the allelic distribution pattern in homozygous and heterozygous progeny to identify the contribution of parental strains to the phenotypic behavior of F2s. The results in Table II show that the presence of W alleles at the QTL on chromosome 4 (Figure 1) was significantly and dominantly associated with a decrease in adenoma surface area and was denominated rCcr1 for rat Colorectal cancer resistance. The presence of ACI alleles at the QTL on chromosome 15 was associated with average multiplicity values, for total lesions and PD adenocarcinomas, significantly lower that those of rats homozygous for W allele. An analogous behavior was found for PD adenocarcinoma multiplicity, for the QTL with suggestive LOD score values, on telomeric side of chromosome 15. We define these QTLs as rat rCcr2 and rCcr6, respectively. Rats bearing a W allele at the centromeric QTL on chromosome 18 showed lower PD adenocarcinoma multiplicity and surface area than rats homozygous for A allele. This QTL was tentatively named rat rCcr3. A second QTL on the chromosome 18 had a LOD score peak at 19.5 cM, suggestive of genetic linkage with PD adenocarcinoma multiplicity. A lower PD adenocarcinoma multiplicity occurred in rats bearing at least one W allele, suggesting the existence of a resistance gene and the locus was denominated rCcr8. Additional QTLs, in suggestive linkage with surface area or multiplicity of adenoma, and surface area of WD adenocarcinomas, on chromosomes 6, 17 and 20, were identified on the basis of the origin of the dominant allele and its phenotypic effects, as resistance loci and denominated rCcr4, rCcr5, rCcr7 and rCcr9 (Table II). Finally, a locus on chromosome 5 was classified as a susceptibility locus affecting adenoma surface area, and denominated rat Colorectal cancer resistance (rCcs) 1.

Interactions between loci
All QTLs identified, except rCcr2 whose R2 value was around 58%, have R2 values between 10.5 and 19.5% (Table II), indicating the presence of a main locus and several low-penetrance genes predisposing to CRC. Since additive interactions may modify the phenotypic effects of QTLs, we evaluated the interactions between QTLs exhibiting analogous effects in relation to the allele distribution pattern in AWF2 rats for the markers closest to LOD score peaks. ANOVA showed significant variations among allelic combinations for the interactions between rCcr6 and rCcr3, rCcr6 and rCcr8 and rCcr5 and rCcr9 (P < 0.0001) (Figure 2). Low values of PD adenocarcinoma multiplicity were found in rats bearing at least one A and W allele at markers D15Rat63, on chromosome 15, and D18Rat46, on chromosome 18, closest to the LOD score peaks of rCcr6 and rCcr3, respectively. The lowest values occurred in rats homozygous for the A allele at D15Rat63 and W allele at D18Rat46. In contrast, the inverse allelic association (WW at D15Rat63 and AA at D18Rat46) produced PD adenocarcinoma multiplicity values 1.9–24 times higher than all other combinations (1.2 ± 0.21 versus 0.05 ± 0.001 to 0.62 ± 0.08; means ± SDs, at least P < 0.01). An analogous situation was found for the interaction between rCcr6 and rCcr8. Relatively low PD adenocarcinoma multiplicity occurred in AWF2 rats homozygous or heterozygous for the A allele at marker D15Rat63 and for W allele at marker D18Rat57, whereas PD adenocarcinoma multiplicity of 1.1 ± 0.16, produced by the combination WW at D15Rat63 and AA at D18Rat57, was 3.1–27 times higher than all other combinations (at least P < 0.05). Finally, the presence of homozygous W allele at D6Rat101 and A allele at D20Rat13, closest to the LOD score peaks of rCcr5 and rCcr9, respectively, was associated with adenoma multiplicity of 2.6 ± 0.53, significantly higher than the multiplicities at all other allelic combinations (at least P < 0.005). Analogous evaluation of the interactions between rCcr2 and rCcr6, rCcr3 or rCcr8 and rCcr1 and rCcr7 showed the absence of additive interactions.


Figure 2
View larger version (13K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 2. Additive interactions between rCcr6/rCcr3, rCcr6/rCcr8 and rCcr5/rCcr9 loci. On the abscissa, allelic combinations in AWF2 rats: upper line, alleles at markers D15Rat63, for rCcr6, and D6Rat101, for rCcr5; lower line, alleles at markers D18Rat46, D18Rat57 and D20Rat13, for rCcr3, rCcr8and rCcr9, respectively. Numbers on the columns represent the number of animals per each allelic combination. Data are means ±SEs. Multiple comparisons of the means were made by the Tuckey–Kramer test: AA/AW, AA/WW, AW/AW and AW/WW, different from AA/AA, WW/WW, WW/AW, AW/AA and WW/AA for at least P<0.05.

 
Strong genetic interactions leading to synergistic effects in conferring cancer resistance (epistatic interactions) allowed identifying several tumor modifier loci in mouse and rat models of human cancer (29–31). The evaluation of two-way interactions inducing phenotypic effects that were undetectable as the sum of single effects led to the identification of four novel loci with highly significant interactions: multiple comparisons showed significant increases in adenoma multiplicity for the allelic combination WW/AA at D6Rat1 and D11Rat38 (P < 1.1–7; corrected as in ref. 21, P < 5.6–6), and AA/WW at D1Rat165 and D6Rat101 (P < 4-5; corrected P < 2.1–3), and significant increase in adenoma size for the allelic combination WW/AA at D1Rat86 and D10Rat25 (P < 7.3–7; corrected P < 3.8–5). Another locus affecting the multiplicity of WD and PD adenocarcinomas was detected by interaction between D13Rat11 and D18Rat57 for the allelic combination WW/AA (P < 7–5; corrected P < 2.9–3).

Wnt/ß-catenin pathway
Among the putative candidate genes sited in the identified QTLs was Apc, located at rCcr3. Thus, we evaluated the eventual deregulation of Wnt/ß-catenin pathway by determining the subcellular localization of ß-catenin in colorectal lesions. ß-Catenin immunostaining was located exclusively in cell membranes of normal colorectal mucosa of both ACI and WF rats (data not shown). Evident interstrain differences in ß-catenin localization occurred in colorectal lesions (Figure 3). In ACI rat lesions, diffuse cytoplasmic/nuclear ß-catenin positivity occurred only in one of three PD adenocarcinomas, with the other two PD adenocarcinomas and three of four WD adenocarcinomas exhibiting positivity in <50% of cells (Figure 3C), and all adenomas (n = 5) showing exclusively membranous ß-catenin immunoreactivity (Figure 3A). A strikingly different ß-catenin pattern was detected in colorectal lesions from WF rats. Indeed, strong membranous, cytoplasmic and/or nuclear staining (affecting at least 50% of tumor cells) was found in three of five adenomas (Figure 3B) and in all 19 WD adenocarcinomas (Figure 3D) developed in these rats. In more advanced adenocarcinomas (Figure 3E) and all 16 PD adenocarcinomas (Figure 3F and G) of the same strain, strong cytoplasmic positivity was associated with ß-catenin nuclear translocation. In AWF1 rats, diffuse positivity only occurred in 9 of 13 PD adenocarcinomas, whereas adenomas (n = 10) and WD adenocarcinomas (n = 3) were negative. In agreement with these findings, the ß-catenin targets Axin2 and c-myc mRNA levels were 6- and 4-fold higher in WF CRCs than normal mucosa, respectively (P < 00.05), whereas moderate over-expression occurred in AWF1 and, at lower level, in ACI rats (supplementary Figure 1 is available at Carcinogenesis Online).


Figure 3
View larger version (170K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 3. Representative immunohistochemical analysis for ß-catenin expression in adenomas, WD adenocarcinomas and PD adenocarcinomas from ACI and WF rats. In ACI rats, ß-catenin positive immunostaining exclusively occurs at cell membrane in adenoma (A), only relatively few cells show cytoplasmic positivity of ß-catenin immunolabeling in WD adenocarcinoma (C). In WF rats, membrane and diffuse cytoplasmic immunoreactivity is present in adenoma (B) and WD adenocarcinoma (D). Strong cytoplasmic and nuclear positivity occurs in more advanced WD adenocarcinoma (E) and PD adenocarcinoma (Fand G). Immunostained cells of an infiltrating PD adenocarcinoma of WF rats are scattered among inflammatory cells of the submucosa (G). Magnification: A, B and F x240; C and D x12; E and G x360; insets: A and D x480; E and F x800.

 
The activation of Wnt/ß-catenin pathway was also assessed by immunoblot analysis after immunoprecipitation (Figure 4). No interstrain differences in ß-catenin and Axin2 expression occurred in normal colon mucosa. ß-Catenin increased in CRC of WF and ACI rats, whereas Axin2 was up-regulated in all strains, with highest values in WF rats. Furthermore, the levels of phosphorylated ß-catenin (targeted for ubiquitination) were significantly more elevated in CRC of ACI rats than in WF corresponding lesions. The levels of the above described proteins in CRC from AWF1 were intermediate between those of WF and ACI strains.


Figure 4
View larger version (58K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 4. Detection by immunoblot analysis after immunoprecipitation of ß-catenin, Axin2, and phosphorylated (p) ß-catenin in normal colorectal mucosa (control) and CRC (WD plus PD adenocarcinomas) of WF, ACI and AWF1 rats. Sample proteins and blocking peptides were co-immunoprecipitated with antibodies against ß-catenin, Axin2 and p-ß-catenin and separated by sodium dodecyl sulfate–polyacrylamide gel electrophoresis. Enhanced chemiluminescence was used to visualize the immune complexes. Upper panel: Representative immunoprecipitation analysis. Lower panel: Chemiluminescence analysis showing mean ±SD of three and nine rats for CRC and control values. Tuckey–Kramer test: (dot) CRC versus control, at least P<0.01; (asterisk) A or AW versus W, at least P>0.01; (double dagger) AW versus A, P<0.001.

 
Experiments aimed at evaluating the role of Apc polymorphism and mutations in colorectal neoplastic lesions showed the absence of interstrain polymorphisms in Apc promoter of normal colon mucosa, whereas single-nucleotide polymorphisms were found at exon 11 (nt 1505: C in WF and T in ACI rats) and exon 15 (nt 2777, 6440, 7634, T, G, T in WF and C, A, C, in ACI, respectively). However, these polymorphisms did not result in changes of the predicted aminoacid sequence of the protein. No mutations in exons 12 and 14 of Apc gene, and mutation cluster region (between 2074 and 4778 bp) of exon 15, were found (data not shown). Mutations were also absent in ß-catenin exon 2, containing the binding sites for GSK-3ß, of CRCs from WF and ACI strains. No interstrain differences in Apc mRNA levels from normal colon mucosa were detected, and approximately the same decrease in Apc expression occurred in CRC of all rat strains (supplementary Figure 1 is available at Carcinogenesis Online). These data were further confirmed by western blotting, which also excluded the presence of Apc-truncating mutations in all tumor samples (data not shown).


   Discussion
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
Supplementary material
 Funding
 References
 
Our results show that the resistance of ACI rats to CRC induction by DMH is associated with a low capacity of adenomas and WD adenocarcinomas to evolve to PD adenocarcinomas, whereas the high susceptibility of the WF strain rests on the rapid progression of adenomas to full malignancy. AWF1 rats exhibit an intermediate phenotype that suggests a global semi-dominant effect on the phenotypic trait of multiple A and W alleles and of the reciprocal interactions involved.

Genomic scanning of AWF2 intercrosses revealed three loci on chromosomes 4, 15 and 18, named rCcr1 to 3 on the basis of the phenotypic effects of the dominant allele in significant linkage with the adenoma surface area, multiplicity of total lesions or multiplicity and surface area of PD adenocarcinomas. Various other QTLs were in suggestive linkage with the phenotypic parameters. However, since their phenotypic effects were confirmed by ANOVA and some of them exhibited additive effects, these QTLs were not considered artifacts. We also discovered three epistatic interactions, affecting the multiplicity or size of adenomas, and a forth interaction affecting the multiplicity of WD and PD adenocarcinomas. Total variability explained by the association between the markers closest to the LOD score peaks and the character (R2) was in the range of 10.5–19.5 for all QTLs except rCcr2, whose variability was ~58% indicating that this locus gives a major contribution to lesion multiplicity, sufficient to elicit per se a resistant phenotype (32). All other CRC modifiers behaved as low-penetrance genes, but additive interaction of rCcr6 with rCcr3 and rCcr8 and of rCcr5 with rCcr9, produced phenotypic effects amounting between 23 and ~30%.

As already observed for the polygenic inheritance of tumors in other rodent tissues (31), the linkage analysis of AWF2 rats indicates that some susceptibility alleles controlling colorectal carcinogenesis of AWF1 rats are contributed by the resistant strain. This observation suggests that the generation of the susceptible WF strain from a common presumably resistant feral ancestor (27) depends on the selective mutation of resistance alleles, with the production of allelic variants that are not activated by carcinogen treatment. The maintenance of unaltered resistance alleles in ACI rats inactivates susceptibility alleles that, however, may be active in a different genetic background, i.e. in subgroups of intercross rats. In the latter, epistatic interactions between recessive W and A alleles, in different chromosomes, contribute to the determination of a susceptible phenotype. Our data are in favor of a model of polygenic inheritance of CRC characterized by multiple low-penetrance genes and a main locus, in which different subsets of the population carry different allelic combinations conferring them various degrees of intermediate susceptibility to CRC, whereas individuals with very low or high risk should be rare. A similar model seems to occur for rat (32) and human (33) liver carcinogenesis, and is compatible with the epidemiological evidence of at least a subgroup of sporadic human CRCs (1).

Modifier genes implicated in rat colorectal carcinogenesis are not yet known. Nevertheless, a plausible candidacy could be suggested for various genes on the basis of their location at the QTLs identified and their involvement in colorectal carcinogenesis. Apc gene mutations have a well-known role on intestinal tumorigenesis of Min-1 mice and human familial adenomatous polyposis inheritance. A pathogenetic role in sporadic CRC has also been postulated (3). Our data deny its candidacy as a modifier gene in chemically induced CRC of rat, due to the absence of functional interstrain polymorphisms and interstrain differences in Apc mRNA levels. Nonetheless, deregulation of the Wnt/ß-catenin pathway, as shown by diffuse nuclear/cytoplasmic localization of ß-catenin, occurred in almost all lesions of WF rats. In ACI and AWF1 rats only 33 and 69% of PD adenocarcinomas, respectively, exhibited diffuse nuclear/cytoplasmic ß-catenin immunostaining. Other adenocarcinomas showed ß-catenin immunostaining in <50% of cell, whereas adenomas of ACI and AWF1 rats were always negative. These observations link the deregulation of Wnt/ß-catenin pathway to the capacity of mucosal lesions to progress to CRC, taking into account the higher propensity of colorectal lesions of susceptible rats to evolve to more malignant stages. Indeed, ß-catenin accumulation in the cytoplasm and nuclei results in the activation of a number of genes, including growth-related genes, such as c-myc, Cyclin D1, Cox2, Fra1 and Id2 (34), and active growth is a prerequisite of tumor progression. APC mutation occurs in early stages of ~60% of sporadic human CRC (35). Different from the human disease, Wnt/ß-catenin pathway activation does not depend on both Apc and ß-catenin mutations during CRC development in the rat strains investigated. Nonetheless, we provide evidence that susceptibility genes control this pathway by regulating ß-catenin phosphorylation (inactivation). This suggests that proteasomal degradation of phosphorylated ß-catenin is critical for the expression of a phenotype susceptible or resistant to CRC. Since a susceptible phenotype is characterized by the development of more advanced colorectal neoplastic lesions, the expression of phosphorylated ß-catenin may be a marker of CRC progression. The relationships of other components of the complex controlling ß-catenin phosphorylation (i.e. Axin1 and Gsk-3ß) with polygenic predisposition in CRC are currently under investigation.

Other possible candidate modifiers or genes somehow influenced by modifiers are present in the loci discovered. rCcr1 harbors Nos3, over-expressed early during azoxymethane-induced CRC of rat (36), and CDK5 up-regulated in human CRC (37). The Perokisome proliferator activated receptor gamma (PPAR-{gamma}) ligand ciglitazone induces CDK5 down-regulation and inhibits in vitro proliferation of HT-29 colon PD adenocarcinoma cells (37). Other genes located at rCcr1, include CD36, Il6 and Hepatocyte growth factor (HGF). CD36 polymorphism, interacting with moderate–high meat consumptions (38) and Il6 polymorphism are associated with increased CRC risk (39). Molecular co-expression of Mesenchymal epithelial transition factor (c-MET) and HGF correlates with a metastatic phenotype of human CRC (40). rCcr4 harbors Odc and rCcr9 harbors S-adenosylmethionine decarboxylase, two genes encoding key enzymes in polyamine synthesis. These genes are up-regulated in human CRC and ODC gene polymorphisms represent a genetic marker for sporadic human CRC risk (41). N-myc, located at rCcr4 regulates NDRG1 (N-myc downstream-regulated gene 1), which plays an important role in human CRC progression and presumably acts as a tumor metastasis promoter gene (42). At rCcr5 maps AKT-1, a component of Phosphatidylinositol 3-kinase (PI3K) pathway, whose up-regulation in human CRC correlates with clinical stage (43). Another interesting candidate gene is Tlp1 (telomerase protein component 1), located at rCcr2, on chromosome 15. Telomerase is up-regulated early during human colorectal tumorigenesis (44) and telomerase protein component 1 activity is strongly implicated in the immortalization of colorectal cells. Other genes involved in colorectal carcinogenesis, located in the discovered QTLs, include Early growth response protein 1 (EGR1) (45) and Fibroblast growth factor 1 (FGF1) (46), at rCcr3 and CtsL (cathepsin L) at rCcr7 (47). The deregulation in human CRC of several genes located at rat QTLs suggests the commonalty of some genetic mechanisms in rodent and human colorectal carcinogenesis. Moreover, rat chromosomal segments where rCcr1, rCcr3, rCcr4 and rCcr7 are located, syntenic to human microsatellite loci at 5q21, 18q11–q12, 2p16, 5q12 and 3p21, frequent sites of allelic imbalance in CRCs and, less frequently, in ACFs (48). The human 15q21.1 subband, syntenic to rCcr6, harbors a putative oncosuppressor gene (49), suggesting a possible location of QTLs related to CRC in the human genome. This hypothesis is reinforced by the location of mouse Ccs1 and Scc5 loci on chromosomes 12 and 18, respectively, at sites syntenic to rat rCcr4 and rCcr3, suggesting the existence of some interspecies correspondences in the genetic predisposition to CRC. However, the modifier of mouse intestinal tumorigenesis, Ptprj gene located at Scc1, maps to chromosome 3 of rat, where no known QTLs were found. In accordance with previous observations (2,5), the present results strongly support the idea that rat models of colorectal carcinogenesis, despite the different etiology from that of human CRC, mimic some phenotypic and molecular features controlling the development and progression of human pre-neoplastic and neoplastic colorectal lesions.

Overall, we identified three novel loci, in significant linkage with chemically induced CRC of rat, and three epistatic loci controlling the progression of early benign lesions to full malignancy. The existence of a locus with relatively high phenotypic effect, and various additive and epistatic genetic interactions, envisages a complex genetic model, controlling sporadic colorectal carcinogenesis, with various low-penetrance genes and a main gene. Furthermore, we demonstrated the influence of susceptibility genes on the Wnt/ß-catenin signaling. The transition from locus to gene is a complex and long-term task (50). Future development of the present research include narrowing of the QTL regions, by the production of congenic rat strains and gene cloning, to better define candidate modifier genes and discover eventual interstrain polymorphisms consistent with interstrain differences in phenotypic behavior. This will allow identifying these genes as modifiers or targets of modifiers, whose role must be evaluated in human colorectal carcinogenesis. The identification of the molecular mechanisms affected by predisposition genes can further clarify the pathogenesis of human CRC.


   Supplementary material
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Supplementary material
Funding
 References
 
Supplementary Figure 1 can be found at http://carin.oxfordjournals.org/


   Funding
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Supplementary material
 Funding
References
 
Associazione Italiana Ricerche sul Cancro; Ministero dell'Istruzione; Università e Ricerca; Assessorato Igiene e Sanità RAS.


   Acknowledgments
 
Conflict of Interest statement: None declared.


   References
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Supplementary material
 Funding
 References
 

  1. Calvert PM, et al. The genetics of colorectal cancer. Ann. Intern. Med. (2002) 137:603–612.[Abstract/Free Full Text]
  2. Potter JD. Colorectal cancer: molecules and populations. J. Natl Cancer Inst. (1999) 91:916–932.[Abstract/Free Full Text]
  3. De la Chapelle A. Genetic predisposition to colorectal cancer. Nat. Rev. Cancer (2004) 4:769–780.[CrossRef][Web of Science][Medline]
  4. Cormier RT, et al. Secretory phospholipase Pla2g2a confers resistance to intestinal tumorigenesis. Nat. Genet. (1997) 17:88–91.[CrossRef][Web of Science][Medline]
  5. Deakin M, et al. Glutathione S-transferase GSTT1 genotypes and susceptibility to cancer: studies of interactions with GSTM1 in lung, oral, gastric and colorectal cancers. Carcinogenesis (1996) 1:881–884.
  6. Korstanje R, et al. From QTL to gene: the harvest begins. Nat. Genet. (2002) 31:235–236.[CrossRef][Web of Science][Medline]
  7. Cowley AW, et al. The genetic dissection of essential hypertension. Nat. Rev. Genet. (2006) 7:829–840.[Web of Science][Medline]
  8. Jacoby RF, et al. Genetic analysis of colon cancer susceptibility in mice. Genomics (1994) 15:381–387.
  9. Angel JM, et al. A locus that influences susceptibility to 1, 2-dimethylhydrazine-induced colon tumors maps to the distal end of mouse chromosome 3. Mol. Carcinog. (2000) 27:47–54.[CrossRef][Web of Science][Medline]
  10. Moen CJ, et al. Different genetic susceptibility to aberrant crypts and colon adenomas in mice. Cancer Res. (1996) 56:2382–2386.[Abstract/Free Full Text]
  11. Moen CJ, et al. Scc-1, a novel colon cancer susceptibility gene in the mouse: linkage to CD44 (Ly-24, Pgp-1) on chromosome 2. Oncogene (1992) 7:563–566.[Web of Science][Medline]
  12. van Wezel T, et al. Gene interaction and single gene effects in colon tumour susceptibility in mice. Nat. Genet. (1996) 14:468–470.[CrossRef][Web of Science][Medline]
  13. Ruivenkamp CA, et al. Ptprj is a candidate for the mouse colon-cancer susceptibility locus Scc1 and is frequently deleted in human cancers. Nat. Genet. (2002) 31:295–300.[CrossRef][Web of Science][Medline]
  14. Ishiguro Y, et al. Strain differences of rats in the susceptibility to aberrant crypt foci formation by 2-amino-1-methyl-6-phenylimidazo- [4,5-b]pyridine: no implication of Apc and Pla2g2a genetic polymorphisms in differential susceptibility. Carcinogenesis (1999) 20:1063–1068.[Abstract/Free Full Text]
  15. Nakagama H, et al. Genetic analysis of the susceptibility in rats to aberrant crypt foci formation by 2-amino-1-methyl-6-phenylimidazo [4,5-b]pyridine, PhIP. Cancer Lett. (1999) 14:205–209.[Medline]
  16. Cheng L, et al. Aberrant crypt foci as microscopic precursors of colorectal cancer. World J. Gastroenterol. (2003) 9:2642–2649.[Web of Science][Medline]
  17. Pascale RM, et al. Role of HSP90, CDC37 and CRM1 as modulators of P16INK4A activity in rat liver carcinogenesis and human liver cancer. Hepatology (2005) 42:1310–1319.[CrossRef][Web of Science][Medline]
  18. Calvisi DF, et al. Activation of beta-catenin during hepatocarcinogenesis in transgenic mouse models: relationship to phenotype and tumor grade. Cancer Res. (2001) 61:2085–2091.[Abstract/Free Full Text]
  19. De Miglio MR, et al. Chromosome mapping of multiple loci affecting the genetic predisposition to rat liver carcinogenesis. Cancer Res. (2002) 62:4459–4463.[Abstract/Free Full Text]
  20. Szpirer C, et al. Gene-based anchoring of the rat genetic linkage and cytogenetic maps: new regional localizations, orientation of the linkage groups, and insights into mammalian chromosome evolution. Mamm. Genome (1998) 9:721–734.[CrossRef][Web of Science][Medline]
  21. Pascale RM, et al. Transferrin and transferrin receptor gene expression and iron uptake in hepatocellular PD adenocarcinoma in the rat. Hepatology (1998) 27:452–461.[CrossRef][Web of Science][Medline]
  22. Lander ES, et al. Genetic dissection of complex trait. Science (1994) 265:2037–2048.[Abstract/Free Full Text]
  23. Lander ES, et al. Genetic dissection of complex traits: guidelines for interpreting and reporting linkage results. Nat. Genet. (1995) 1:241–247.
  24. Kruglyak L, et al. A nonparametric approach for mapping quantitative trait loci. Genetics (1995) 139:1421–1428.[Abstract]
  25. Bowman KG, et al. R/qtl: QTL mapping in experimental crosses. Bioinformatics (2003) 19:889–890.[Abstract/Free Full Text]
  26. Myamoto M, et al. Colon carcinoma of highly inbred rats. J. Natl Cancer Inst. (1975) 55:1471–1472.[Web of Science][Medline]
  27. Canzian F, et al. Construction of a phylogenetic tree for inbred strains of rat by arbitrarily primed polymerase chain reaction (AP-PCR). Mamm. Genome (1995) 6:231–235.[CrossRef][Web of Science][Medline]
  28. Conover WJ, et al. Rank transformations as a bridge between parametric and nonparametric statistics. Am. Stat. (1981) 35:124–129.[CrossRef]
  29. Nagase H, et al. Epistatic interactions between skin tumors modifier loci in interspecific (spretus/musculus) backcross mice. Cancer Res. (2001) 61:1305–1308.[Abstract/Free Full Text]
  30. Mao JH, et al. Genomic approaches to identification of tumor-susceptbility genes using mouse models. Curr. Opin. Genet. Dev. (2003) 13:14–19.[CrossRef][Web of Science][Medline]
  31. De Miglio MR, et al. Polygenic control of hepatocarcinogenesis in copenhagen x F344 rats. Int. J. Cancer (2004) 111:9–16.[CrossRef][Web of Science][Medline]
  32. Feo F, et al. Hepatocellular carcinoma as a complex polygenic disease. Interpretive analysis of recent developments on genetic predisposition. BBA Cancer Rev. (2006) 1765:126–147.
  33. Hemminki K, et al. Familial risks of cancer as a guide to gene identification and mode of inheritance. Int. J. Cancer (2004) 110:291–294.[CrossRef][Web of Science][Medline]
  34. Gregorieff A, et al. Wnt signaling in the intestinal epithelium: from endoderm to cancer. Genes Dev. (2005) 19:877–890.[Abstract/Free Full Text]
  35. Powell SM, et al. APC mutations occur early during colorectal tumorigenesis. Nature (1992) 359:235–237.[CrossRef][Medline]
  36. Escribano M, et al. Aspirin inhibits endothelial nitric oxide synthase (eNOS) and Flk-1 (vascular endothelial growth factor receptor-2) prior to rat colon tumour development. Clin. Sci. (Lond) (2004) 106:83–91.[Medline]
  37. Kim E, et al. CDK5 is a novel regulatory protein in PPARgamma ligand-induced antiproliferation. Int. J. Oncol. (2006) 28:191–194.[Web of Science][Medline]
  38. Kuriki K, et al. Increased risk of colorectal cancer due to interactions between meat consumption and the CD36 gene A52C polymorphism among Japanese. Nutr. Cancer (2005) 51:170–177.[CrossRef][Web of Science][Medline]
  39. Landi S, et al. Association of common polymorphisms in inflammatory genes interleukin (IL)6, IL8, tumor necrosis factor alpha, NFKB1, and peroxisome proliferator-activated receptor gamma with colorectal cancer. Cancer Res. (2003) 63:3560–3566.[Abstract/Free Full Text]
  40. Kammula US, et al. Molecular co-expression of the c-Met oncogene and hepatocyte growth factor in primary colon cancer predicts tumor stage and clinical outcome. Cancer Lett. (2007) 248:219–228.[CrossRef][Web of Science][Medline]
  41. Martinez ME, et al. Pronounced reduction in adenoma recurrence associated with aspirin use and a polymorphism in the ornithine decarboxylase gene. Proc. Natl Acad. Sci. USA (2003) 100:7859–7864.[Abstract/Free Full Text]
  42. Wang Z, et al. Correlation of N-myc downstream-regulated gene 1 overexpression with progressive growth of colorectal neoplasm. World J. Gastroenterol. (2004) 10:550–554.[Medline]
  43. Rychahou PG, et al. Targeted molecular therapy of the PI3K pathway: therapeutic significance of PI3K subunit targeting in colorectal PD adenocarcinoma. Ann. Surg. (2006) 243:833–842.[CrossRef][Web of Science][Medline]
  44. Kolquist KA, et al. Expression of TERT in early premalignant lesions and a subset of cells in normal tissues. Nat. Genet. (1998) 19:103–104.[CrossRef][Web of Science][Medline]
  45. Tsunoda T, et al. Gene expression profiles and identification of the genes relevant to clinicopathologic factors in colorectal cancer selected by cDNA array method in combination with principal component analysis. Int. J. Oncol. (2003) 23:49–59.[Web of Science][Medline]
  46. Netzer P, et al. Inhibition of human colon cancer cell growth by antisense oligodeoxynucleotides targeted at basic fibroblast growth factor. Aliment. Pharmacol. Ther. (2001) 15:1673–1679.[CrossRef][Web of Science][Medline]
  47. Troy AM, et al. Expression of cathepsin B and L antigen and activity is associated with early colorectal cancer progression. Eur. J. Cancer (2004) 40:1610–1616.[CrossRef][Web of Science][Medline]
  48. Luo L, et al. Loss of heterozygosity in human aberrant crypt foci (ACF), a putative precursor of colon cancer. Carcinogenesis (2006) 27:1153–1159.[Abstract/Free Full Text]
  49. Park WS, et al. A distinct tumor suppressor gene locus on chromosome 15q21.1 in sporadic form of colorectal cancer. Cancer Res. (2000) 60:70–73.[Abstract/Free Full Text]
  50. Flint J, et al. Strategies for mapping and cloning quantitative trait genes in rodents. Nat. Rev. Genet. (2005) 6:271–276.[CrossRef][Web of Science][Medline]
Received February 1, 2007; revised May 8, 2007; accepted May 12, 2007.


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?



This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
28/11/2367    most recent
bgm119v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by De Miglio, M. R.
Right arrow Articles by Feo, F.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by De Miglio, M. R.
Right arrow Articles by Feo, F.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?