Myh Deficiency Enhances Intestinal Tumorigenesis in Multiple Intestinal Neoplasia (ApcMin/+) Mice

[CANCER RESEARCH 64, 8876 – 8881, December 15, 2004]

Myh Deficiency Enhances Intestinal Tumorigenesis in Multiple Intestinal Neoplasia (ApcMin/⫹) Mice Oliver M. Sieber,1 Kimberley M. Howarth,1 Christina Thirlwell,1 Andrew Rowan,1 Nikki Mandir,2 Robert A. Goodlad,2 Ashfaq Gilkar,3 Bradley Spencer-Dene,3 Gordon Stamp,3 Victoria Johnson,4 Andrew Silver,4 Hanjing Yang,5 Jeffrey H. Miller,5 Mohammad Ilyas,6 and Ian P. M. Tomlinson1 1

Molecular and Population Genetics Laboratory, 2Histopathology Unit, and 3Experimental Pathology Laboratory, London Research Institute, Cancer Research United Kingdom, London, United Kingdom; 4Colorectal Cancer Unit, Cancer Research United Kingdom, St. Mark’s Hospital, Harrow, United Kingdom; 5Molecular Biology Institute, University of California, Los Angeles, California; and 6Academic Unit of Pathology, Leeds General Infirmary, Leeds, United Kingdom

ABSTRACT Monoallelic APC and biallelic MYH (homolog of Escherichia coli mutY) germ-line mutations are independently associated with a strong predisposition to colorectal adenomas and carcinoma in humans. Whereas mice heterozygous for mutant Apc develop intestinal tumors, mice homozygous for mutant Myh do not show increased tumor susceptibility. We analyzed the phenotype of ApcMin/⫹/Myh⫺/⫺ mice and found that they developed significantly more adenomas in the small intestine than did ApcMin/⫹/ Myh⫹/⫹ or ApcMin/⫹/Myh⫹/⫺ mice (median 231 versus 151 versus 152). In the large bowel, ApcMin/⫹/Myh⫺/⫺ mice showed significant increases in the number of aberrant crypt foci. In addition, ApcMin/⫹/Myh⫺/⫺ mice developed an increased number of mammary tumors. Molecular analyses suggested that at least 19% of intestinal tumors from ApcMin/⫹/Myh⫺/⫺ mice had acquired intragenic Apc mutations rather than allelic loss. Consistent with a defect in base excision repair, three intragenic Apc mutations in polyps without allelic loss from ApcMin/⫹/Myh⫺/⫺ mice were shown to be G:C to T:A transversions which resulted in termination codons; no such mutations were found in polyps from ApcMin/⫹/Myh⫹/⫹ or ApcMin/⫹/Myh⫹/⫺ mice. Tumors from ApcMin/⫹/Myh⫹/⫺ mice harbored neither somatic mutations nor allelic loss at Myh. Thus, homozygous, but not heterozygous, Myh deficiency enhanced intestinal tumorigenesis in ApcMin/⫹ mice. The excess small-bowel adenomas in ApcMin/⫹/Myh⫺/⫺ mice, therefore, appear to be a model of MYH-associated polyposis in humans.

INTRODUCTION Familial adenomatous polyposis is a rare, autosomal dominant disease characterized by hundreds or thousands of colorectal adenomas, resulting in a near 100% lifetime risk of colorectal cancer. In addition, familial adenomatous polyposis patients often develop extracolonic features, including upper gastrointestinal tumors, congenital hypertrophy of the retinal pigment epithelium (CHRPE), desmoid tumors and dental abnormalities. Familial adenomatous polyposis is caused by germ-line mutations in the adenomatous polyposis coli (APC) gene, and somatic APC mutations are found in up to 80% of sporadic colorectal cancers. Both familial adenomatous polyposis and sporadic tumors usually harbor biallelic APC mutations, resulting in constitutive activation of the Wnt signaling pathway by up-regulating the level of ␤-catenin (1, 2). Patients with MYH (homolog of Escherichia coli mutY)-associated polyposis have a phenotype similar to patients with mild familial adenomatous polyposis (3, 4). However, the disease shows autosomal Received 8/16/04; revised 10/8/04; accepted 10/19/04. Grant support: Supported by Cancer Research United Kingdom, the Medical Research Council (Clinical Research Fellowship; C. Thirlwell), and the National Cancer Institute (grant CA 85952; J. H. Miller). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Requests for reprints: Ian Tomlinson, Molecular and Population Genetics Laboratory, London Research Institute, Cancer Research United Kingdom, 44 Lincoln’s Inn Fields, London WCA 3PX, United Kingdom. Phone: 0044-0-207-269-2884; Fax: 00440-207-269-2885; E-mail: [email protected]. ©2004 American Association for Cancer Research.

recessive inheritance. MYH-associated polyposis accounts for ⬃30% of patients with 15 to 100 adenomas, and ⬃10% of APC mutationnegative patients with classical polyposis (⬎100 adenomas; ref. 4). Colorectal cancer associated with MYH-associated polyposis appears to have a prevalence similar to that associated with familial adenomatous polyposis (5). MYH-associated polyposis is caused by biallelic germ-line mutations in the base excision repair gene MYH (6). Base excision repair is important in repairing oxidative DNA damage originating as a byproduct of normal cellular metabolism or from extrinsic sources. The most frequent mutagenic lesion is 7,8-dihydro-8oxoguanine (8-oxo-G). 8-oxo-G can mispair with adenine during DNA replication and cause G:C to T:A transversions. The products of three human base excision repair genes, MTH1, OGG1, and MYH, act together to prevent 8-oxo-G–induced mutagenesis. In the nucleotide pool, MTH1 hydrolyzes 8-oxo-dGTP to 8-oxo-dGMP (7–9); OGG1 detects and removes 8-oxo-G generated in situ or incorporated into DNA during replication (10 –13); and MYH, an adenine-specific DNA glycosylase, scans the daughter strand after DNA replication and removes adenines mispaired with 8-oxo-G (14 –16). MYH deficiency leads to an increased frequency of G:C to T:A mutations in genes such as APC and K-ras in MYHassociated polyposis tumors (17, 18). In contrast to MYH-associated polyposis patients, Myh-deficient mice do not seem to show increased tumor susceptibility. Mice homozygous for a loss-of-function Myh allele (Myh⫺/⫺) showed no difference from wild-type littermates in terms of survival after 14 months or of tumor incidence after 15 to 17 months (19). However, Myh and Ogg1“double knockout” mice displayed a marked increase of lymphomas and lung and ovarian tumors. Myh⫺/⫺ mice have been shown to accumulate 8-oxo-G at an increased rate in the liver when compared with wild-type mice, although no age-dependent accumulation was detected in brain, small intestine, spleen, or kidney (20). We aimed at developing a mouse model for MYH-associated polyposis. We reasoned that, despite the data above, Myh deficiency might have some subtle effect on the intrinsic mutation rate that was insufficient to cause macroscopic tumor development on an otherwise wild-type genetic background. Because most colorectal tumors from MYH-associated polyposis patients seem to harbor biallelic APC mutations (17, 18), we crossed Myh-deficient mice with multipleintestinal-neoplasia (ApcMin/⫹) mice, which carry a nonsense mutation at codon 850 of the Apc gene. ApcMin/⫹ mice are predisposed to multiple adenomas, primarily in the small intestine. In addition, these mice have an increased risk of developing mammary tumors, desmoids, and epidermoid cysts (21). Because homozygosity for the Min allele is embryonic-lethal, we analyzed animals of the six remaining genotypes (ApcMin/⫹/Myh⫹/⫹, ApcMin/⫹/Myh⫹/⫺, ApcMin/⫹/Myh⫺/⫺, Apc⫹/⫹/Myh⫹/⫹, Apc⫹/⫹/Myh⫹/⫺, and Apc⫹/⫹/Myh⫺/⫺) for the development of intestinal tumors, cystic crypts of the small intestine and aberrant crypt foci (ACFs) of the colon.

8876

Downloaded from cancerres.aacrjournals.org on February 29, 2016. © 2004 American Association for Cancer Research.

Myh DEFICIENCY IN MULTIPLE INTESTINAL NEOPLASIA MICE

MATERIALS AND METHODS Mouse Breeding. Animals were bred at Cancer Research United Kingdom Laboratories (South Mimms, United Kingdom). Experiments were conducted in full accordance with the United Kingdom Animal Procedures Act 1986. The C57BL/6J-ApcMin/⫹ mice were originally a gift from A. R. Moser (McArdle Laboratory, University of Wisconsin, Madison, WI). The 129/C57BL/6JMyh⫹/⫺ mice have been described previously (19). Male ApcMin/⫹ mice were bred with female Myh⫹/⫺ mice. Of the F1 generation, male ApcMin/⫹/Myh⫹/⫺ mice were intercrossed with Apc⫹/⫹/Myh⫹/⫺ females. Only the F2 offspring were included in the analyses because these are expected to share equivalent contributions of the genetic background from the two strains, with the only systematic difference between the groups of mice being Myh itself and, inevitably, small regions flanking this locus. Our three study groups were ApcMin/⫹/Myh⫹/⫹, ApcMin/⫹/Myh⫹/⫺, and ApcMin/⫹/Myh⫺/⫺ mice; our three control groups were Apc⫹/⫹/Myh⫹/⫹, Apc⫹/⫹/Myh⫹/⫺, and Apc⫹/⫹/Myh⫺/⫺ mice. Animals of the six groups arose at the expected Mendelian frequencies (P ⫽ 1, ␹2 test). Genotyping. DNA was extracted from tail snips by using standard methods. All of the animals were genotyped for Apc and Myh with three-primer, allele-specific PCR assays (19, 22); all founder mice were homozygous for the sensitive allele of Mom1 (22). Counts of Tumors, Cystic Crypts, and Aberrant Crypt Foci. Mice were sacrificed at 110 to 130 days by CO2 asphyxiation. The intestinal tract was removed and divided into four segments consisting of three equal lengths of small intestine (proximal, middle, and distal) and the whole of the large intestine. Each segment was flushed with PBS, opened longitudinally, and laid out on filter paper. The samples were fixed in Carnoy’s medium for 3 hours and stored in 70% EtOH. Before analysis, preparations were stained with 0.2% methylene blue for 3 minutes and were washed in PBS for 20 minutes. Tumors were counted with a dissecting microscope at ⫻3 magnification, and categorized according to size with an inoculating loop with an inner diameter of 1 mm and an outer diameter of 2 mm. Cystic crypts in the small intestine and Fig. 1. Apc allelic loss assay. Fluorescence-SSCP analysis allowed us to distinguish the ACFs in the large intestine were counted at ⫻3 and ⫻5 magnification, wild-type (Wt Normal) and Min (Min Normal) alleles of Apc. Allelic loss in adenomatous respectively. Criteria used to identify ACFs were dark-blue staining, increased polyps (Polyp 1, Polyp 2) from ApcMin/⫹/Myh⫹/⫹, ApcMin/⫹/Myh⫹/⫺, and ApcMin/⫹/ ⫺/⫺ mice, if present, exclusively targeted the wild-type Apc allele. crypt size, and flat appearance hidden in the surrounding mucosa. All of the Myh counts were done in duplicate by an observer (O. M. S.) blinded to the genotype of the mice. A subset of cases was counted by a second observer relative ratio of normal-to-tumor peak area was ⬍0.5 or ⬎2, thereby allowing (C. T.) to confirm the first observer’s results. for contaminating normal tissue. The validity of the F-SSCP-LOH assay had Histology. After tumor counting, the proximal small-bowel segment from been established with serial dilutions of Apc wild-type and Min alleles, and Min/⫹ ⫹/⫹ Min/⫹ ⫹/⫺ Min/⫹ five animals per study group (Apc /Myh , Apc /Myh , Apc / results were confirmed for selected samples by sequencing tumor DNA alongMyh⫺/⫺) was rolled up, embedded in paraffin, sectioned, and stained with side normal DNA (details not shown). Loss of the wild-type Myh allele was hematoxylin and eosin (H&E). In addition, selected large tumors of the small analyzed with the same allele-specific PCR assay as for the genotyping, in and large intestine were processed, as well as all detected mammary tumors. which a 262-bp product is diagnostic of the wild-type allele and a 376-bp To examine the histology of colonic ACFs, serial transverse sections were product is diagnostic of the targeted allele. The common forward primer was prepared at 4 ␮m. All samples were analyzed by a histopathologist (G. S.) fluorescently labeled. PCR products were run on the ABI 3100 and allelic loss blinded to the genotype of the mice. scored as above. DNA Extraction from Tumor and Normal Tissue. Tumors were disMutation Analysis. Primer pairs amplifying coding regions and exonsected from the methylene blue-stained, whole-mount preparations with a intron boundaries were designed for Apc exons 5 to 14, Myh exons 2 to 15, and dissecting microscope at ⫻3 magnification. A fresh scalpel was used for each K-ras exon 1. Apc codons 652 to 1648 (exon 15) were covered by 10 sample, and care was taken to avoid the surrounding normal epithelium. In overlapping primer pairs. Details and PCR conditions are available from the Min/⫹ ⫹/⫹ total, 134 small-bowel and 6 colonic tumors from 12 Apc /Myh mice, authors. Apc and Myh were screened by F-SSCP analysis. In addition, areas of Min/⫹ ⫹/⫺ 147 small-bowel and 21 colonic tumors from 16 Apc /Myh mice, and Apc rich in T/AGAA sequences that are particularly susceptible to G:C3 T:A Min/⫹ ⫺/⫺ 167 small-bowel and 16 colonic tumors from 13 Apc /Myh mice were mutation to chain termination codons in humans (codons 894-1357 and 1545– collected. Tail tissue and normal bowel epithelium were also sampled. DNA 1648 encompassing 44 such sites) were analyzed by direct DNA sequencing was extracted from these samples with standard methods. (17, 18). K-ras exon 1 was analyzed by direct DNA sequencing. For F-SSCP Loss of Heterozygosity (Allelic Loss) Analysis at Apc and Myh. Loss of analysis, PCR products were run at 18°C and 24°C on the ABI 3100 and were heterozygosity (LOH) at Apc was analyzed with a fluorescence–single-strand analyzed with Genotyper 2.5 software. For tumor samples showing aberrant conformational polymorphism (F-SSCP) assay that distinguished the wild-type SSCP bands, the relevant region of the gene was sequenced directly from a and Min alleles (Fig. 1) with dye-labeled primers: Apc LOH1 forward (5⬘new PCR product alongside normal DNA. TACGGTATTGCCCAGCTCTT-3⬘) and Apc LOH1 reverse (5⬘-TGCAGACCTCGTTTTGATGA-3⬘). Each 25-␮L PCR reaction contained 1⫻ PCR buffer (Perkin-Elmer, Boston, MA), 2 mmol/L MgCl2, 200 ␮mol/L dNTPs, 200 RESULTS nmol/L concentration of each primer, 30 ng of genomic DNA, and 1 unit of Myh Deficiency Enhances Tumor Formation in the Small InTaqDNA polymerase (Qiagen). PCR conditions consisted of 95°C for 5 testine of ApcMin/⫹ Mice. In the small intestine, ApcMin/⫹/Myh⫹/⫹, minutes, 35 cycles of 95°C/55°C/72°C for 1 minute each, and a final step at Min/⫹ Apc /Myh⫹/⫺, and ApcMin/⫹/Myh⫺/⫺ F2 offspring developed me72°C for 10 minutes. The resulting PCR products were run at 18°C through a 50-cm capillary on the ABI 3100 and analyzed with Genotyper 2.5 software dian numbers of 151, 152, and 231 polyps, respectively (Table 1; Fig. (Perkin-Elmer Applied Biosystems). Allelic loss was considered present if the 2). Thus, tumor multiplicity in the small intestine was increased by 8877

Downloaded from cancerres.aacrjournals.org on February 29, 2016. © 2004 American Association for Cancer Research.

Myh DEFICIENCY IN MULTIPLE INTESTINAL NEOPLASIA MICE

Table 1 Effect of heterozygous and homozygous Myh deficiency on adenoma, cystic crypt, and ACF numbers in the intestine of ApcMin/⫹ mice ApcMin/⫹/Myh⫹/⫹ (n ⫽ 15) Small bowel Adenomas Proximal Middle Distal Cystic crypts Large bowel Adenomas ACFs

151, 17, 65, 84, 13,

ApcMin/⫹/Myh⫹/⫺ (n ⫽ 25)

62–285 (161 ⫾ 58) 6–35 (21 ⫾ 8) 22–105 (62 ⫾ 23) 30–150 (78 ⫾ 33) 3–90 (20 ⫾ 24)

152, 22, 62, 68, 7,

1, 0–12 (2 ⫾ 3) 20, 4–52 (23 ⫾ 13)

25–345 (164 ⫾ 92) 6–57 (25 ⫾ 16) 5–128 (60 ⫾ 34) 13–184 (79 ⫾ 47) 0–116 (24 ⫾ 32)

1, 0–6 (2 ⫾ 2) 15, 3–48 (17 ⫾ 12)

ApcMin/⫹/Myh⫺/⫺ (n ⫽ 15) 231, 38, 101, 99, 28,

68–336 (228 ⫾ 84) 17–82 (44 ⫾ 21) 21–131 (85 ⫾ 35) 30–156 (99 ⫾ 36) 1–171 (46 ⫾ 50)

1, 0–5 (2 ⫾ 1) 43, 21–96 (48 ⫾ 22)

NOTE. Values are medians and ranges. Means ⫾ SD are given in parentheses. The three study groups had similar male-to-female ratios (8:7, 11:14, and 6:9 for ApcMin/⫹/Myh⫹/⫹,

ApcMin/⫹/Myh⫹/⫺, and ApcMin/⫹/Myh⫺/⫺ mice, respectively).

Fig. 2. Adenomatous polyps and cystic crypts in the small intestine of ApcMin/⫹/ Myh⫺/⫺ mice. Methylene blue-stained whole-mount preparations showing adenomatous polyps (A) and cystic crypts (B, arrowheads). H&E-stained transverse sections showing cystic crypts with normal pathology (C) and dysplastic change (D).

⬃53% in ApcMin/⫹/Myh⫺/⫺ mice as compared with ApcMin/⫹/Myh⫹/⫹ mice (P ⫽ 0.016, Mann–Whitney U test), whereas heterozygous Myh deficiency had no effect (P ⫽ 0.83, Mann–Whitney U test). Subanalysis of the three segments of small intestine showed that the increase in tumor number in Myh-deficient mice was more pronounced in the proximal and middle segments than in the distal segment (P ⬍ 0.01, ANOVA for interaction; Table 1). When adenomas were grouped according to size, it was evident that neither heterozygous nor homozygous Myh deficiency had a significant effect on the proportions of lesions that were ⬍1 mm, 1 to 2 mm, and ⬎2 mm in size, being ⬃60, 35, and 5%, respectively (P ⫽ 0.10 for ApcMin/⫹/Myh⫹/⫹ versus ApcMin/⫹/Myh⫹/⫺; P ⫽ 1.0 for ApcMin/⫹/ Myh⫹/⫹ versus ApcMin/⫹/Myh⫺/⫺, ␹2 test). Cystic crypts are lesions of unknown neoplastic potential that predominantly occur in the proximal small intestine of ApcMin/⫹ mice and that increase in number when ApcMin/⫹ mice are treated with ethylnitrosourea or are bred to mismatch repair (MMR) deficient mice (23, 24). Cystic crypts were counted for the same samples that were also scored for intestinal adenomas. Median numbers of 13, 7, and 28 cystic crypts were found in ApcMin/⫹/Myh⫹/⫹, ApcMin/⫹/Myh⫹/⫺, and ApcMin/⫹/Myh⫺/⫺ mice, respectively (Table 1; Fig. 2). However, the increase in cystic crypt number in Myh-deficient mice as compared with Myh wild-type mice was not statistically significant (P ⫽ 0.17, Mann–Whitney U test). Histopathological analysis revealed that a subpopulation of cystic crypts displayed dysplastic changes (Fig. 2). In general, animals with a high polyp count in the small intestine tended to have a high cystic crypt count (R2 ⫽ 0.57; P ⬍ 0.001, linear regression analysis).

No adenomas or cystic crypts were found in the small intestine of any of 15 animals from each of the three control groups (Apc⫹/⫹/ Myh⫹/⫹, Apc⫹/⫹/Myh⫹/⫺, and Apc⫹/⫹/Myh⫺/⫺ mice). Myh Deficiency Increases ACF Number in the Colon of ApcMin/ⴙ Mice. ApcMin/⫹/Myh⫹/⫹, ApcMin/⫹/Myh⫹/⫺, and ApcMin/⫹/ Myh⫺/⫺ mice all developed a median of one polyp in the colon (P ⫽ 0.18 for ApcMin/⫹/Myh⫹/⫹ versus ApcMin/⫹/Myh⫹/⫺; P ⫽ 0.16 for ApcMin/⫹/Myh⫹/⫹ versus ApcMin/⫹/Myh⫺/⫺, Mann–Whitney U test; Table 1; Fig. 3). In addition, adenoma sizes did not differ significantly between ApcMin/⫹/Myh⫹/⫹, ApcMin/⫹/Myh⫹/⫺, and ApcMin/⫹/Myh⫺/⫺ mice, with 82% (19 of 23), 60% (28 of 47), and 88% (21 of 24), of lesions being ⬎2 mm in size, respectively (P ⫽ 0.06 for ApcMin/⫹/ Myh⫹/⫹ versus ApcMin/⫹/Myh⫹/⫺; P ⫽ 0.70 for ApcMin/⫹/Myh⫹/⫹ versus ApcMin/⫹/Myh⫺/⫺, Fisher’s Exact test). ACFs are putative premalignant lesions that develop in the colon of ApcMin/⫹ mice (25–27) and that can be induced by chemical carcinogens (28, 29). ACFs were counted for the same samples that were also scored for colonic adenomas. ApcMin/⫹/Myh⫹/⫹, ApcMin/⫹/ Myh⫹/⫺, and ApcMin/⫹/Myh⫺/⫺ had a median number of 20, 15, and 43 ACFs, respectively (Table 1). Thus, Myh-deficient mice showed an approximately 2-fold increase in ACF number compared with Myh wild-type mice (P ⫽ 0.001, Mann–Whitney U test), whereas heterozygous Myh deficiency had no effect (P ⫽ 0.10, Mann–Whitney U test). In all of the animals, ACFs occurred predominantly in the distal colon. Histopathological analysis of selected ACFs revealed mild dysplasia in all of the cases (Fig. 3). Surprisingly, ACF numbers were not associated with colonic tumor numbers, which suggested that

Fig. 3. Adenomatous polyps and ACFs in the colons of ApcMin/⫹/Myh⫺/⫺ mice. Methylene blue-stained, whole-mount preparations showing adenomatous polyps (A) and ACFs (B, arrowheads). H&E-stained transverse sections showing monocryptal (C) and polycryptal (D) ACFs. Scale bars, 500 ␮m.

8878

Downloaded from cancerres.aacrjournals.org on February 29, 2016. © 2004 American Association for Cancer Research.

Myh DEFICIENCY IN MULTIPLE INTESTINAL NEOPLASIA MICE

there is no simple stochastic progression of ACF to adenoma (R2 ⫽ 0.0054, P ⫽ 0.61, linear regression analysis). No tumors or ACFs were found in the colons of any of 15 animals from each of the three control groups (Apc⫹/⫹/Myh⫹/⫹, Apc⫹/⫹/ Myh⫹/⫺, and Apc⫹/⫹/Myh⫺/⫺). Adenomas from Myh-Deficient ApcMin/⫹ Mice Show No Evidence of Progression to Carcinoma. Histopathological analysis of 8 small-bowel and 8 large-bowel tumors from ApcMin/⫹/Myh⫹/⫹, 21 small-bowel and 17 large-bowel tumors from ApcMin/⫹/Myh⫹/⫺, and 30 small-bowel and 17 large-bowel tumors from ApcMin/⫹/Myh⫺/⫺ mice revealed that all were adenomatous polyps and none displayed evidence of malignant change. Thus, similar to tumors from Myh wild-type and heterozygous mice, tumors from Myh-deficient ApcMin/⫹ mice did not appear to progress to carcinoma by 120 to 130 days of age. In addition, no differences were apparent in the histology of normal intestinal mucosa between the three study groups. Myh-Deficient ApcMin/⫹ Mice Develop an Increased Frequency of Mammary Gland Tumors. Mammary gland tumors were detected in 0 (0%) of 16 ApcMin/⫹/Myh⫹/⫹, 3 (⬃7%) of 43 ApcMin/⫹/ Myh⫹/⫺, and 4 (⬃27%) of 15 ApcMin/⫹/Myh⫺/⫺ mice. The increase in mammary gland tumors was significant in My-deficient mice as compared with Myh wild-type and heterozygous mice (P ⫽ 0.028, Fisher’s Exact test). All mammary gland tumors showed squamous cell differentiation, and one, from a Myh-deficient mouse, had invaded into the surrounding fat. Small-Bowel Tumors from ApcMin/⫹/Myh⫺/⫺ but Not from ApcMin/⫹/Myh⫹/⫺ Mice Show a Reduced Frequency of Allelic Loss at Apc. Tumor formation in ApcMin/⫹ mice has been shown to involve somatic inactivation of the wild-type Apc allele (30, 31). On a C57BL/6J background, this generally occurs by allelic loss. If the increased number of small-bowel tumors in ApcMin/⫹/Myh⫺/⫺ mice reflected a failure to repair oxidative DNA damage, a corresponding proportion of the tumors from these mice were expected to harbor truncating somatic mutations rather than allelic loss at Apc. We found that 118 (93%) of 127 tumors from ApcMin/⫹/Myh⫹/⫹ mice and 87 (89%) of 98 tumors from ApcMin/⫹/Myh⫹/⫺ mice had lost the wildtype Apc allele, compared with 121 (81%) of 150 tumors from ApcMin/⫹/Myh⫺/⫺ mice (Fig. 1; P ⬍ 0.001, for Myh-deficient mice versus Myh wild-type and heterozygous mice, Fisher’s Exact test). These results are consistent with the notion that whereas the growth of most polyps in ApcMin/⫹/Myh⫺/⫺ mice is initiated because of loss of wild-type Apc, as in ApcMin/⫹/Myh⫹/⫹ mice, the significant excess of tumors occurs because absent Myh activity leads to “second hits” that are distinct from allelic loss. The fraction of tumors that retained the Apc wild-type allele in Myh-deficient mice (⬃19%) was, however, lower than the observed increase in tumor number (⬃53%). This difference may be the result of chance but may also reflect a requirement for mutation in one or more additional genes for macroscopic polyp development. Alternatively, Myh deficiency may have some other effect in addition to increasing the intrinsic mutation rate. Analysis of large-bowel polyps from each genotypic class revealed no significant difference in the frequency of allelic loss at Apc. We found that 6 (100%) of 6 tumors from ApcMin/⫹/Myh⫹/⫹, 18 (100%) of 18 tumors from ApcMin/⫹/Myh⫹/⫺, and 15 (94%) of 16 tumors from ApcMin/⫹/Myh⫺/⫺ harbored loss of the wild-type Apc allele (P ⫽ 1.0 for ApcMin/⫹/Myh⫹/⫹ versus ApcMin/⫹/Myh⫺/⫺, Fisher’s Exact test). G:C3 T:A Truncating Somatic Mutations Occur in Adenomas from Myh-Deficient ApcMin/⫹ Mice. All 29 small-bowel tumors without allelic loss at Apc from ApcMin/⫹/Myh⫺/⫺ mice were examined for somatic mutations in Apc. Mutations were detected in 3 (10%) polyps. All three changes were G:C to T:A transversions [2⫻ nucleotide (nt) 3919G⬎T, nt4754C⬎A] resulting in “stop” codons (2⫻ E1307X, S1585X). The sequence context around the E1307X

mutations was AAAGAGA (Fig. 4), the reverse of the consensus motif (underlined) identified in human MYH-associated polyposis tumors. No somatic Apc mutations were detected in eight and nine small-bowel tumors without allelic loss from ApcMin/⫹/Myh⫹/⫺ and ApcMin/⫹/Myh⫹/⫹ mice, respectively. Adenomas from Myh-Deficient ApcMin/⫹ Mice Do Not Harbor K-ras Mutations. Activating K-ras mutations occur in ⬃50% of sporadic colorectal adenomas in humans (32, 33), and tumors from MYH-associated polyposis patients show an increased frequency of K-ras mutations as compared with tumors from familial adenomatous polyposis patients (18, 34). In addition, about 75% of lung tumors from Myh and Ogg1 double-knockout mice harbor G to T mutations in K-ras codon 12 (19). However, we found no K-ras mutation in 20 intestinal tumors from ApcMin/⫹/Myh⫹/⫹ nor in 40 tumors from ApcMin/⫹/Myh⫺/⫺ mice (35). Adenomas from Myh-Heterozygous ApcMin/⫹ Mice Do Not Show Somatic Mutations or Allelic Loss at Myh. In our study, ApcMin/⫹/Myh⫹/⫺ mice did not show an increase in polyp number or size as compared with ApcMin/⫹/Myh⫹/⫹ mice, providing little evidence for a role of Myh heterozygosity in tumor development. However, it has been suggested in humans that inactivation of the wildtype Myh allele in heterozygotes is selected in intestinal tumorigenesis (36). We, therefore, screened adenomas from ApcMin/⫹/Myh⫹/⫺ mice for somatic mutations and allelic loss at Myh. Only 2 (1%) of 137 tumors showed allelic loss at Myh, but in one of these cases, LOH targeted the inherited mutant rather than the wild-type allele. A total of 82 tumors were subsequently screened for somatic mutations in Myh exons 2 to 15, but no changes were detected. Taken together, these data do not support the hypothesis that inactivation of the wild-type Myh allele is selected in tumors from ApcMin/⫹/Myh⫹/⫺ mice. DISCUSSION In this study, we have demonstrated that Myh deficiency has a significant effect on both intestinal and extraintestinal tumor development in ApcMin/⫹ mice. ApcMin/⫹/Myh⫺/⫺ mice showed an in-

Fig. 4. Somatic G to T transversion mutation in Apc resulting in a termination codon (E1307X) identified in two small-bowel adenomas from ApcMin/⫹/Myh⫺/⫺ mice. Forward sequence from tumor DNA (A) and cloned mutant allele (B).

8879

Downloaded from cancerres.aacrjournals.org on February 29, 2016. © 2004 American Association for Cancer Research.

Myh DEFICIENCY IN MULTIPLE INTESTINAL NEOPLASIA MICE

creased adenomatous polyp number in the small bowel. In addition, Myh deficiency increased the frequency of mammary tumors. The magnitude of the increase in tumor number in the small intestine was correlated with the fraction of tumors that failed to show loss of the wild-type Apc allele. This association is consistent with studies on tumors from ApcMin/⫹ mice with deficiencies in the base excision repair protein Mbd4 or the MMR proteins Mlh1, Msh2, or Pms2 (24, 37– 40). In addition, G:C⬎T:A somatic Apc mutations were found in polyps from ApcMin/⫹/Myh⫺/⫺ mice, as expected for defective base excision repair. Taken together, these data strongly suggest that the loss of Myh function enhances tumorigenesis in ApcMin/⫹ mice. Our finding of exclusively G:C⬎T:A somatic mutations in polyps from ApcMin/⫹/Myh⫺/⫺ mice (and no such mutations in polyps from ApcMin/⫹/Myh⫹/⫺ and ApcMin/⫹/Myh⫹/⫹ mice) provides good evidence to show that this is the mechanism that causes the excess of tumors, as it is in humans (6, 17). Previous studies have shown somatic G:C⬎T:A Apc mutations to be very rare in tumors from ApcMin/⫹ mice (40, 41); we detected truncating somatic Apc mutations in a minority of ApcMin/⫹/Myh⫺/⫺ tumors without allelic loss, a similarly low mutation frequency to that reported in almost all previous studies in which ApcMin/⫹ mice have been crossed with animals defective in DNA repair (24, 37, 40). A highly plausible explanation for the low frequency of somatic, truncating Apc mutations is that adenomas from ApcMin/⫹ mice are polyclonal, the contribution of each mutant allele to the tumor DNA too small to be detected in most cases (42). Detailed examination of both tumor size and histology failed to reveal any evidence of tumor progression in the small bowel of ApcMin/⫹/Myh⫺/⫺ mice by 120 to 130 days of age. These data are consistent with studies on tumors from ApcMin/⫹ mice with deficiencies in Mlh1, Msh2, or Pms2 (24, 38, 39) and suggest that Myh deficiency has a larger effect on tumor initiation than on progression in the small bowel. ApcMin/⫹/Myh⫺/⫺ mice showed a more pronounced increase in tumor number in the proximal and middle segments than in the distal segment of the small bowel. A similar pattern is observed in MYHassociated polyposis patients who sometimes develop duodenal polyps (4). MMR-deficient ApcMin/⫹ mice, by contrast, show a uniform increase in the number of small-bowel polyps (24). The finding in ApcMin/⫹/Myh⫺/⫺ mice and MYH-associated polyposis patients may reflect variation in the level of oxidative stress throughout the small bowel, although it could also reflect variation in the level or activity of Myh protein or site-specific differences in functional redundancy. Heterozygous Myh deficiency had no detectable effect on tumor development in ApcMin/⫹ mice or the frequency of allelic loss at Apc. Furthermore, inactivation of the wild-type Myh allele was very rare in these mice. Although these data must be interpreted with some caution because of the possible polyclonal origin of polyps in the mouse small intestine, they are consistent with our findings, in sporadic colorectal cancers from humans, that failed to detect somatic mutations or allelic loss at MYH (43). Besides intestinal adenomas, ApcMin/⫹ mice are predisposed to two types of lesions of unknown neoplastic potential, cystic crypts of the small intestine and ACFs of the colon. Cystic crypt multiplicity is increased in ApcMin/⫹ mice deficient in Mlh1 or treated with chemical carcinogens (23, 24). Although our data were suggestive of a similar effect in Myh-deficient ApcMin/⫹ mice, the increase did not reach statistical significance. In contrast, ApcMin/⫹/Myh⫺/⫺ mice showed a significant, ⬃2-fold, increase in the number of ACFs, supporting a role of Myh deficiency in promoting the development of these lesions. In conclusion, our data suggest that Myh deficiency leads to an increase in the intrinsic mutation rate. This increase is insufficient to cause intestinal tumorigenesis in our Myh-deficient mice on an oth-

erwise wild-type genetic background but is made manifest when mice who already harbor the first step of tumorigenesis (Apc mutation) are studied. MYH deficiency may have a more profound effect on the mutation rate in humans, who are subjected to greater levels of oxidative stress in the intestine than are the laboratory mouse. Although we did not detect somatic K-ras mutations, we believe that the excess adenomas in ApcMin/⫹/Myh⫺/⫺ mice are a useful model of MYH-associated polyposis, allowing studies of molecular mechanisms or the evaluation of therapeutic agents. Furthermore, suppression of loss of the wild-type Apc allele, for example, by crossing ApcMin/⫹/Myh⫺/⫺ mice onto the AKR genetic background (44), may produce a mouse model of MYH-associated polyposis in which most polyps develop as a result of defective base excision repair. ACKNOWLEDGMENTS We thank staff from the Cancer Research United Kingdom Clare Hall Laboratories and Equipment Park for excellent technical support. Jackie Haines (Radiation Effects Department, National Radiological Protection Board, Oxon, UK) provided the DNA from ApcMin/Min embryos, and Tomas Lindahl kindly gave advice on the experimental analysis.

REFERENCES 1. Sieber OM, Tomlinson IP, Lamlum H. The adenomatous polyposis coli (APC) tumour suppressor— genetics, function and disease. Mol Med Today 2000;6:462–9. 2. Fearnhead NS, Britton MP, Bodmer WF. The ABC of APC. Hum Mol Genet 2001;10:721–33. 3. Sampson JR, Dolwani S, Jones S, et al. Autosomal recessive colorectal adenomatous polyposis due to inherited mutations of MYH. Lancet 2003;362:39 – 41. 4. Sieber OM, Lipton L, Crabtree M, et al. Multiple colorectal adenomas, classic adenomatous polyposis, and germ-line mutations in MYH. N Engl J Med 2003;348: 791–9. 5. Enholm S, Hienonen T, Suomalainen A, et al. Proportion and phenotype of MYHassociated colorectal neoplasia in a population-based series of Finnish colorectal cancer patients. Am J Pathol. 2003;163:827–32. 6. Al-Tassan N, Chmiel NH, Maynard J, et al. Inherited variants of MYH associated with somatic G:C3 A mutations in colorectal tumors. Nat Genet 2002;30:227–32. 7. Mo JY, Maki H, Sekiguchi M. Hydrolytic elimination of a mutagenic nucleotide, 8-oxodGTP, by human 18-kilodalton protein: sanitization of nucleotide pool. Proc Natl Acad Sci USA 1992;89:11021–5. 8. Sakumi K, Furuichi M, Tsuzuki T, et al. Cloning and expression of cDNA for a human enzyme that hydrolyzes 8-oxo- dGTP, a mutagenic substrate for DNA synthesis. J Biol Chem 1993;268:23524 –30. 9. Oda H, Nakabeppu Y, Furuichi M, et al. Regulation of expression of the human MTH1 gene encoding 8-oxo-dGTPase. Alternative splicing of transcription products. J Biol Chem 1997;272:17843–50. 10. Lu R, Nash HM, Verdine GL. A mammalian DNA repair enzyme that excises oxidatively damaged guanines maps to a locus frequently lost in lung cancer. Curr Biol 1997;7:397– 407. 11. Radicella JP, Dherin C, Desmaze C, et al. Cloning and characterization of hOGG1, a human homolog of the OGG1 gene of Saccharomyces cerevisiae. Proc Natl Acad Sci USA 1997;94:8010 –5. 12. Roldan-Arjona T, Wei YF, Carter KC, et al. Molecular cloning and functional expression of a human cDNA encoding the antimutator enzyme 8-hydroxyguanineDNA glycosylase. Proc Natl Acad Sci USA 1997;94:8016 –20. 13. Shinmura K, Kasai H, Sasaki A, et al. 8-hydroxyguanine (7,8-dihydro-8-oxoguanine) DNA glycosylase and AP lyase activities of hOGG1 protein and their substrate specificity. Mutat Res 1997;385:75– 82. 14. Slupska MM, Baikalov C, Luther WM, et al. Cloning and sequencing a human homolog (hMYH) of the Escherichia coli mutY gene whose function is required for the repair of oxidative DNA damage. J Bacteriol 1996;178:3885–92. 15. Slupska MM, Luther WM, Chiang JH, et al. Functional expression of hMYH, a human homolog of the Escherichia coli MutY protein. J Bacteriol 1999;181:6210 –3. 16. Takao M, Zhang QM, Yonei S, et al. Differential subcellular localization of human MutY homolog (hMYH) and the functional activity of adenine:8-oxoguanine DNA glycosylase. Nucleic Acids Res 1999;27:3638 – 44. 17. Jones S, Emmerson P, Maynard J, et al. Biallelic germline mutations in MYH predispose to multiple colorectal adenoma and somatic G:C3 T:A mutations. Hum Mol Genet 2002;11:2961–7. 18. Lipton L, Halford SE, Johnson V, et al. Carcinogenesis in MYH-associated polyposis follows a distinct genetic pathway. Cancer Res 2003;63:7595–9. 19. Xie Y, Yang H, Cunanan C, et al. Deficiencies in mouse Myh and Ogg1 result in tumor predisposition and G to T mutations in codon 12 of the K-ras oncogene in lung tumors. Cancer Res 2004;64:3096 –102. 20. Russo MT, De Luca G, Degan P, et al. Accumulation of the oxidative base lesion 8-hydroxyguanine in DNA of tumor-prone mice defective in both the Myh and Ogg1 DNA glycosylases. Cancer Res 2004;64:4411– 4.

8880

Downloaded from cancerres.aacrjournals.org on February 29, 2016. © 2004 American Association for Cancer Research.

Myh DEFICIENCY IN MULTIPLE INTESTINAL NEOPLASIA MICE

21. Shoemaker AR, Gould KA, Luongo C, et al. Studies of neoplasia in the Min mouse. Biochim Biophys Acta 1997;1332:F25– 48. 22. Wilding J, Straub J, Bee J, et al. Cyclin D1 is not an essential target of beta-catenin signaling during intestinal tumorigenesis, but it may act as a modifier of disease severity in multiple intestinal neoplasia (Min) mice. Cancer Res 2002;62:4562–5. 23. Shoemaker AR, Moser AR, Dove WF. N-ethyl-N-nitrosourea treatment of multiple intestinal neoplasia (Min) mice: age-related effects on the formation of intestinal adenomas, cystic crypts, and epidermoid cysts. Cancer Res 1995;55:4479 – 85. 24. Shoemaker AR, Haigis KM, Baker SM, et al. Mlh1 deficiency enhances several phenotypes of Apc(Min)/⫹ mice. Oncogene 2000;19:2774 –9. 25. Paulsen JE, Namork E, Steffensen IL, et al. Identification and quantification of aberrant crypt foci in the colon of Min mice—a murine model of familial adenomatous polyposis. Scand J Gastroenterol 2000;35:534 –9. 26. Paulsen JE, Steffensen IL, Loberg EM, et al. Qualitative and quantitative relationship between dysplastic aberrant crypt foci and tumorigenesis in the Min/⫹ mouse colon. Cancer Res 2001;61:5010 –5. 27. Yamada Y, Hata K, Hirose Y, et al. Microadenomatous lesions involving loss of Apc heterozygosity in the colon of adult Apc(Min/⫹) mice. Cancer Res 2002;62:6367–70. 28. Bird RP. Observation and quantification of aberrant crypts in the murine colon treated with a colon carcinogen: preliminary findings. Cancer Lett 1987;37:147–51. 29. Wargovich MJ, Jimenez A, McKee K, et al. Efficacy of potential chemopreventive agents on rat colon aberrant crypt formation and progression. Carcinogenesis (Lond) 2000;21:1149 –55. 30. Levy DB, Smith KJ, Beazer-Barclay Y, et al. Inactivation of both APC alleles in human and mouse tumors. Cancer Res 1994;54:5953– 8. 31. Luongo C, Moser AR, Gledhill S, et al. Loss of Apc⫹ in intestinal adenomas from Min mice. Cancer Res 1994;54:5947–52. 32. Bos JL. The ras gene family and human carcinogenesis. Mutat Res 1988;195:255–71. 33. Forrester K, Almoguera C, Han K, et al. Detection of high incidence of K-ras oncogenes during human colon tumorigenesis. Nature (Lond) 1987;327:298 –303.

34. Lamlum H, Papadopoulou A, Ilyas M, et al. APC mutations are sufficient for the growth of early colorectal adenomas. Proc Natl Acad Sci USA 2000;97:2225–28. 35. Smits R, Kartheuser A, Jagmohan-Changur S, et al. Loss of Apc and the entire chromosome 18 but absence of mutations at the Ras and Tp53 genes in intestinal tumors from Apc1638N, a mouse model for Apc-driven carcinogenesis. Carcinogenesis (Lond) 1997;18:321–7. 36. Kambara T, Whitehall VL, Spring KJ, et al. Role of inherited defects of MYH in the development of sporadic colorectal cancer. Genes Chromosomes Cancer 2004;40: 1–9. 37. Wong E, Yang K, Kuraguchi M, et al. Mbd4 inactivation increases C3 T transition mutations and promotes gastrointestinal tumor formation. Proc Natl Acad Sci USA 2002;99:14937– 42. 38. Reitmair AH, Cai JC, Bjerknes M, et al. MSH2 deficiency contributes to accelerated APC-mediated intestinal tumorigenesis. Cancer Res 1996;56:2922– 6. 39. Baker SM, Harris AC, Tsao JL, et al. Enhanced intestinal adenomatous polyp formation in Pms2⫺/⫺;Min mice. Cancer Res 1998;58:1087–9. 40. Smits R, Hofland N, Edelmann W, et al. Somatic Apc mutations are selected upon their capacity to inactivate the beta-catenin downregulating activity. Genes Chromosomes Cancer 2000;29:229 –39. 41. Sohn KJ, Choi M, Song J, et al. Msh2 deficiency enhances somatic Apc and p53 mutations in Apc⫹/⫺Msh2⫺/⫺ mice. Carcinogenesis (Lond) 2003;24:217–24. 42. Merritt AJ, Gould KA, Dove WF. Polyclonal structure of intestinal adenomas in ApcMin/⫹ mice with concomitant loss of Apc⫹ from all tumor lineages. Proc Natl Acad Sci USA 1997;94:13927–31. 43. Halford SE, Rowan AJ, Lipton L, et al. Germline mutations but not somatic changes at the MYH locus contribute to the pathogenesis of unselected colorectal cancers. Am J Pathol. 2003;162:1545– 8. 44. Shoemaker AR, Moser AR, Midgley CA, et al. A resistant genetic background leading to incomplete penetrance of intestinal neoplasia and reduced loss of heterozygosity in ApcMin/⫹ mice. Proc Natl Acad Sci USA 1998;95:10826 –31.

8881

Downloaded from cancerres.aacrjournals.org on February 29, 2016. © 2004 American Association for Cancer Research.

Myh Deficiency Enhances Intestinal Tumorigenesis in Multiple Intestinal Neoplasia ( ApcMin/+) Mice Oliver M. Sieber, Kimberley M. Howarth, Christina Thirlwell, et al. Cancer Res 2004;64:8876-8881.

Updated version

Cited articles Citing articles

E-mail alerts Reprints and Subscriptions Permissions

Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/64/24/8876

This article cites 44 articles, 28 of which you can access for free at: http://cancerres.aacrjournals.org/content/64/24/8876.full.html#ref-list-1 This article has been cited by 8 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/64/24/8876.full.html#related-urls

Sign up to receive free email-alerts related to this article or journal. To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at [email protected]. To request permission to re-use all or part of this article, contact the AACR Publications Department at [email protected].

Downloaded from cancerres.aacrjournals.org on February 29, 2016. © 2004 American Association for Cancer Research.