A Porcine Model of Familial Adenomatous Polyposis
2012; Elsevier BV; Volume: 143; Issue: 5 Linguagem: Inglês
10.1053/j.gastro.2012.07.110
ISSN1528-0012
AutoresTatiana Flisikowska, Claudia Merkl, Martina Landmann, Stefan Eser, Nousin Rezaei, Xinxin Cui, Mayuko Kurome, Valeri Zakhartchenko, Barbara Keßler, Hagen Wieland, O. Rottmann, Roland M. Schmid, Günter Schneider, Alexander Kind, Eckhard Wolf, Dieter Saur, Angelika Schnieke,
Tópico(s)Cancer Genomics and Diagnostics
ResumoWe created gene-targeted pigs with mutations in the adenomatous polyposis coli (APC) gene (APC) that are orthologous to those responsible for human familial adenomatous polyposis (FAP). One-year-old pigs with the APC1311 mutation (orthologous to human APC1309) have aberrant crypt foci and low- and high-grade dysplastic adenomas in the large intestine, similar to the precancerous lesions that develop in patients with FAP. Dysplastic adenomas accumulate β-catenin and lose heterozygosity of APC. This large-animal, genetic model of FAP will be useful in the development of diagnostics and therapeutics for colorectal cancer. DNA sequence data: NCBI accession number GU951771. We created gene-targeted pigs with mutations in the adenomatous polyposis coli (APC) gene (APC) that are orthologous to those responsible for human familial adenomatous polyposis (FAP). One-year-old pigs with the APC1311 mutation (orthologous to human APC1309) have aberrant crypt foci and low- and high-grade dysplastic adenomas in the large intestine, similar to the precancerous lesions that develop in patients with FAP. Dysplastic adenomas accumulate β-catenin and lose heterozygosity of APC. This large-animal, genetic model of FAP will be useful in the development of diagnostics and therapeutics for colorectal cancer. DNA sequence data: NCBI accession number GU951771. See Covering the Cover synopsis on page 1125; see editorial on page 1133. See Covering the Cover synopsis on page 1125; see editorial on page 1133. Familial adenomatous polyposis (FAP) is characterized by foci of dysplastic growth in the colon and rectum that develop to adenomatous polyps and adenocarcinoma.1Fodde R. et al.Trends Mol Med. 2001; 7: 369-373Abstract Full Text Full Text PDF PubMed Scopus (77) Google Scholar Germline mutations in the adenomatous polyposis coli (APC) tumor-suppressor gene are responsible for FAP, and somatic APC mutations are found in most sporadic colorectal tumors at the earliest stages.2Fodde R. et al.Nat Rev Cancer. 2001; 1: 55-67Crossref PubMed Scopus (769) Google Scholar A representative, human-scale animal model of colorectal cancer will aid in the development of novel drug, endoscopic or surgical interventions, and investigations of dietary components and gut flora. No confirmed cases of polyps or spontaneous colon cancer in wild-type pigs have been reported.3Sievert Jr, C.E. et al.Gastrointest Endosc. 1981; 27: 12-16Abstract Full Text PDF PubMed Scopus (7) Google Scholar We generated gene-targeted cloned pigs carrying translational stop signals in the APC gene at codons 1061 and 1311, orthologous to common germline mutations (APC1061 and APC1309) in human FAP4Polakis P. Biochim Biophys Acta. 1997; 1332: F127-F147PubMed Google Scholar (Supplementary Figure 1). Mutation at human codon 1309 is associated with a particularly severe phenotype with early onset and prolific polyposis; mutation at codon 1061 causes less severe polyposis.5Newton K. et al.Clin Genet. 2012; 81: 521-531Crossref PubMed Scopus (42) Google Scholar The gastrointestinal tracts of a wild-type, an APC1061/+ mutant, and an APC1311/+ mutant pig were examined at 1 year. The wild-type animal was normal. The APC1061/+ founder pig also showed no evidence of polyposis, which was unsurprising at this young age given the human phenotype. Older APC1061/+ animals will be investigated for later onset of polyposis. Examination of the APC1311/+ pig, however, revealed more than 100 macroscopically visible lesions, including more than 60 sessile polyps, in the colon and rectum (Figure 1A and B, Supplementary Figure 2A–C). Polyps ranged from barely visible mucosal nodules of 1–2 mm, to flat polyps up to 1 cm. They were scattered along the whole large bowel, with most in the proximal colon (Figure 1D). As in young human FAP patients, no gastric polyps or duodenal adenomas were observed. In humans, colonic adenomas develop during childhood, but gastric polyps and duodenal adenomas occur later in adulthood. Again, examination of older animals will reveal whether the pig phenotype replicates this aspect of human FAP. This contrasts with murine Apc mutants such as Min (multiple intestinal neoplasia), in which polyps form almost exclusively in the duodenum and small bowel.6Boivin G.P. et al.Gastroenterology. 2003; 124: 762-777Abstract Full Text Full Text PDF PubMed Scopus (416) Google Scholar To show that the porcine FAP model is suitable for evaluation of diagnostic procedures using human-sized equipment, we performed high-resolution magnification chromoendoscopic imaging of the small and large intestines. Methylene blue staining revealed more than 30 additional colonic lesions that were not visible by white light (Figure 2A). Some showed classic features of aberrant crypt foci (ACF).7Pretlow T.P. et al.Cancer Res. 1991; 51: 1564-1567PubMed Google Scholar Lesions were classified according to pit pattern8Huang Q. et al.Gastrointest Endosc. 2004; 60: 520-526Abstract Full Text Full Text PDF PubMed Scopus (85) Google Scholar as non-neoplastic (pit pattern types I and II), or neoplastic microadenomas (pit pattern type III/IV) (Figure 2A). Histologic evaluation of more than 30 such lesions confirmed endoscopic classification in approximately 80% of cases, consistent with the accuracy of human chromoendoscopy.8Huang Q. et al.Gastrointest Endosc. 2004; 60: 520-526Abstract Full Text Full Text PDF PubMed Scopus (85) Google Scholar Most ACF showed features of dysplastic crypts (unicryptal adenomas) and, similar to most pit pattern type III lesions, showed elongated and crowded cells with hyperchromatic nuclei that recapitulated organized glandular structures (Figure 2B). These were classified as adenomas with low-grade intraepithelial neoplasia.6Boivin G.P. et al.Gastroenterology. 2003; 124: 762-777Abstract Full Text Full Text PDF PubMed Scopus (416) Google Scholar Adenomas in human FAP also arise from a single dysplastic crypt.9Hamilton S.R. et al.Pathology and genetics of tumours of the digestive system. World Health Organization Press, Lyon2000Google Scholar Pit pattern type I and II lesions (Figure 2A) showed serrated morphology without dysplasia, reminiscent of human hyperplastic polyps (Supplementary Figure 3A–D). Such lesions are common in humans, but not in Apc mutant mice.6Boivin G.P. et al.Gastroenterology. 2003; 124: 762-777Abstract Full Text Full Text PDF PubMed Scopus (416) Google Scholar Once considered benign,10O'Brien M.J. Gastroenterol Clin North Am. 2007; 36 (viii): 947-968Abstract Full Text Full Text PDF PubMed Scopus (153) Google Scholar they now are thought to be associated with increased risk of colorectal cancer.11Hiraoka S. et al.Gastroenterology. 2010; 139 (e1–3): 1503-1510Abstract Full Text Full Text PDF PubMed Scopus (173) Google Scholar, 12Schreiner M.A. et al.Gastroenterology. 2010; 139: 1497-1502Abstract Full Text Full Text PDF PubMed Scopus (207) Google Scholar Our observations suggest this may apply in pigs. Some larger adenomas showed focal features of more advanced tumors: pronounced nuclear atypia and pleomorphism with aberrant chromatin pattern, large nuclei with loss of polarity with respect to the basement membrane, large nucleoli, and cribriform gland architecture. These were classified as adenomas with focal high-grade intraepithelial neoplasia6Boivin G.P. et al.Gastroenterology. 2003; 124: 762-777Abstract Full Text Full Text PDF PubMed Scopus (416) Google Scholar (Figure 2B). Porcine tumors appeared identical to human colonic adenomas with respect to surface involvement, with dysplastic cells on the superficial mucosal surface (Figures 1C and 2B and D, and Supplementary Figures 2A and 4). This contrasts with murine Apc mutant adenomas, which have a surface layer of nondysplastic epithelium (Supplementary Figure 2D).6Boivin G.P. et al.Gastroenterology. 2003; 124: 762-777Abstract Full Text Full Text PDF PubMed Scopus (416) Google Scholar Wnt pathway activation caused by inactivation of the second (wild-type) APC allele is a hallmark of human FAP and sporadic disease.13Albuquerque C. et al.Hum Mol Genet. 2002; 11: 1549-1560Crossref PubMed Google Scholar We assessed the wild-type APC allele in 5 porcine adenomas (0.4–1 cm), 3 with low-grade and 2 with low and focal high-grade dysplasia. The wild-type APC allele was lost in each case (Figure 2C). ACF and colonic tumors also showed strong nuclear and cytoplasmic β-catenin staining outside the proliferative zone, suggesting that tumor initiation and progression occurs through Wnt pathway activation (Figure 2D and Supplementary Figure 4). Consistent with human FAP, there was no increase in epithelial proliferation of histologically normal intestinal mucosa, as assessed by Ki67 immunohistochemistry (Figure 2D and Supplementary Figure 4). However, adenomatous epithelium was highly proliferative, with almost all dysplastic cells Ki67 positive. Adenomas expressed cytokeratin-19 and CDX2 (caudal type homeobox2) markers of colonic epithelium, validating their origin (Supplementary Figure 5). Adenomas also showed strong expression of the β-catenin target c-MYC, and frequent phosphorylation of extracellular signal-related kinase (ERK) 1 and 2, markers of mitogen-activated protein kinase (MAPK) pathway activation, a known driver of intestinal tumorigenesis14Lee S.H. et al.Nat Med. 2010; 16: 665-670Crossref PubMed Scopus (170) Google Scholar (Supplementary Figure 5). These data confirm that in pigs, unlike mice, a single heterozygous germline APC mutation is sufficient to initiate the well-characterized precancer sequence leading to adenomas in the large intestine, replicating early stage human FAP. F1 generation APC1311/+ piglets were born recently, enabling further characterization of the FAP model including progression to cancer. The molecular pathogenesis of colorectal cancer is marked by multiple genetic alterations of proto-oncogenes and tumor-suppressor genes. As with mice, gene targeting in pigs now allows the generation of animals with multiple, precisely defined oncogenic mutations. We recently produced pigs with knockout and conditional dominant-negative mutation of the TP53 gene,15Leuchs S. et al.PLoS One. 2012; (In press)PubMed Google Scholar which can be combined with the APC mutations described here. These pigs will contribute to the study of human colorectal cancer and improve prospects for those affected or at risk. The authors thank Margret Bahnweg and Peggy Mueller-Fliedner for technical assistance with molecular biology, Steffen and Viola Loebnitz for animal husbandry, Tatiana Schmid for assistance with histology, Roland Rad and Barbara Seidler for technical advice, Sabine Langer for help with microdissection, and Anna-Lisa Boettcher for help generating the APC1061 targeting vector. T.F. and C.M. contributed equally to this work. Animal experiments were approved by the Government of Upper Bavaria (permit number 55.2-1-54-2531-92-08) and performed according to the German Animal Welfare Act and European Union Normative for Care and Use of Experimental Animals. Chemicals were obtained from Sigma-Aldrich Chemie GmbH (Taufkirchen, Germany), cell culture media and supplements were from PAA Laboratories GmbH (Cölbe, Germany) unless specified. A 17.13-kb porcine APC gene fragment spanning exons 13–17 was isolated; the DNA sequence was determined (NCBI accession number, GU951771) and used to generate 2 promoter trap gene targeting vectors. Vector APC1061 comprised a 9.36-kb, 5′ homologous arm from exon 15 to codon 1061 in exon 17; a 1.24-kb mutation cassette including a translational sop codon at position 1061, an internal ribosome entry site (IRES) linked to a blasticidin selectable gene (bsr) and polyadenylation signal; and a 3.1-kb 3′ short arm. Vector APC1311 comprised a 10.14-kb, 5′ homologous arm from exon 15 to codon 1311 in exon 17; a mutation cassette including a translational top codon at position 1311 and IRES-bsr marker; and a 2.67-kb 3′ short arm. Mesenchymal stem cells were isolated by standard methods from leg bones or subcutaneous fat of 6- to 7-month-old male Landrace/Pietrain or Landrace pigs obtained at slaughter. Mesenchymal stem cells were cultured in advanced Dulbecco's modified Eagle's medium (Gibco, Darmstadt, Germany), 2 mmol/L GlutaMAX, 1× non-essential amino acids, 10% fetal calf serum, 5 ng/mL fibroblast growth factor-2 (PromoKine, Heidelberg, Germany) at 37°C, 5% CO2, and passaged using Accutase. Samples of 1 × 106 mesenchymal stem cells were electroporated with 13 μg linearized vector DNA and selected with 8 μg/mL blasticidin (Invivogen, Toulouse, France). Targeted cell clones were identified by polymerase chain reaction (PCR) amplification across the shorter arm of each targeting vector using primers BSf (5′ GAGCAACGGCTACAATCA 3′) and APCr (5′ TCCGAACTCCTGGAATGTGA 3′) with the PCR Extender System (5 PRIME GmbH, Hamburg, Germany). Thermal cycling parameters were as follows: 2 minutes at 94°C; then 40 cycles of 30 seconds at 94°C; 30 seconds at 60°C; 4 minutes at 72°C; followed by 7 minutes at 72°C. Bicistronic messenger RNA from targeted APC alleles was detected by reverse-transcription (RT)-PCR from exon 13 to bsr using primers APCf (5′ GACTACAGGCCATTGCAGAA 3′) and BSr (5′ GGCAGCAATTCACGAATC 3′). Wild-type APC messenger RNA was detected by RT-PCR from exons 13–17 using primers APCf (sequence described earlier) and APCr2 (5′ TGAGCACCACTCTTTGATGG 3′). RT-PCR used SuperscriptIII One-Step RT-PCR with Platinum Taq (Invitrogen, Darmstadt, Germany) with thermal cycling parameters of 30 minutes at 55°C; 2 minutes at 94°C; then 40 cycles of 15 seconds at 94°C; 30 seconds at 55°C; 3 minutes at 68°C; followed by 7 minutes at 68°C. BglII (New England Biolabs, Frankfurt am Main, Germany) digested genomic DNA from cell clones or piglet ear tip samples was electrophoresed, bound to membrane, hybridized, and probe-detected with antidigoxigenin antibody Fab fragments conjugated with alkaline phosphatase (Roche, Penzberg, Germany) by standard methods. The 1206-bp APC probe was generated by PCR using primers APCSf (5′ AAGTCAGGCGGCTACCACTT 3′) and APCSr (5′ GGACAGTCCTCGATTCTCAC 3′), incorporating alkali labile digoxigenin-11-dUTP (Roche). Nuclear transfer and embryo transfer were performed as previously described.1Klymiuk N. et al.FASEB J. 2012; 26: 1086-1099Crossref PubMed Scopus (51) Google Scholar For histopathologic analysis, specimens were fixed in 4% buffered paraformaldehyde, embedded in paraffin, sectioned (3 μm), and stained with H&E. For immunohistochemistry, formalin-fixed, paraffin-embedded tissue sections were dewaxed, microwaved (10 min, 600 watt), and incubated with primary antibodies, as follows: c-MYC (1:250; N-262; sc-764; Santa Cruz Biotechnology, Heidelberg, Germany), β-catenin (1:100; #610154; BD Biosciences, Heidelberg, Germany), phospho-p44/42 MAPK (Erk1/2 Thr202/Tyr204) (1:300; D13.14.4E XP; Cell Signaling, Danvers, MA), Ki-67 (1:100; SP6; DCS Hamburg, Germany), cytokeratin 19 (1:300; TROMAIII; Developmental Studies Hybridoma Bank, University of Iowa), CDX2 (1:20; ATM28; #ABIN152122; antikoerper-online.de, Aachen, Germany). Detection was by biotinylated secondary antibodies (Vector Laboratories, Peterborough, UK) peroxidase-conjugated streptavidin with 3,3′-diaminobenzidine tetrahydrochloride as a chromogen, as previously described.2Eser S. et al.Proc Natl Acad Sci U S A. 2011; 108: 9945-9950Crossref PubMed Scopus (72) Google Scholar Macroscopic pictures of colonic lesions were taken with a STORZ (Tuttlingen, Germany) endoscopy system and a Zeiss (Munich, Germany) Stemi 11 stereomicroscope. Colonic mucosa was stained with 1% methylene blue in phosphate-buffered saline for chromoendoscopy. Adenomas and morphologically normal mucosa were microdissected from 8-μm–thick, dewaxed, H&E stained, formalin-fixed, paraffin-embedded tissue sections using a P.A.L.M. laser microdissection system (Zeiss) according to the manufacturer's recommendations. Approximately 50 cells from each were dissected and DNA was isolated using a QIAamp DNA Micro Kit (Qiagen, Hilden, Germany). The wild-type APC allele was detected by PCR using primers APC38f (5′ GCCACAGACATCCCTTCTTC 3′) and APC9r (5′ TGAGCACCACTCTTTGATGG 3′) using Phire Hot Start DNA polymerase (Finnzymes, Schwerte, Germany). Thermal cycling parameters were as follows: 30 seconds at 98°C; then 40 cycles of 5 seconds at 98°C; 15 seconds at 59°C; 30 seconds at 72°C; followed by 1 minute at 72°C. The wild-type APC amplified fragment corresponds to position 11359 to 11915 in the sequence (GU951771). The APC1311 allele was detected using bsr-specific primer BSf (5′ GAGCAACGGCTACAATCA 3′) and APC9r (sequence described earlier) with the PCR Extender System (5′). Thermal cycling parameters were as follows: 2 minutes at 94°C; then 40 cycles of 20 seconds at 94°C; 15 seconds at 59°C; 45 seconds at 72°C; followed by 1 minute at 72°C.Supplementary Figure 2Comparison of intestinal lesions in APC1311/+ mutant pig and Apcmin/+ mutant mouse. (A) Macroscopic endoscopic view (upper panel) and microscopic H&E-stained sections (lower panel) of adenomas in the colon of APC1311/+ mutant pig. (B and C) Age-matched (B) APC1061/+ mutant pig and (C) wild-type control pig showed no intestinal lesions by macroscopic endoscopy (upper left), high-magnification chromoendoscopy (upper right and lower left) or microscopy (lower right). (D) Upper panel: Macroscopic endoscopic (upper left) and stereomicroscopic (upper right) view of adenoma in the small intestine of Apcmin/+ mutant mouse. Lower panel: H&E-stained section of Apcmin mutant polyp showing adenoma covered by a surface layer of nondysplastic epithelium.View Large Image Figure ViewerDownload Hi-res image Download (PPT)Supplementary Figure 3Hyperplastic polyps in the colon of APC1311/+ mutant pig. (A) Macroscopic view of hyperplastic polyp in the colon of APC1311/+ mutant pig. (B) H&E-stained section of lesion shown in panel A. Note the elongated and serrated colonic crypts without dysplasia reminiscent of human hyperplastic polyps. (C and D) Ki67 staining of serial sections reveals expansion of the crypt proliferative zone. (E and F) Adjacent normal mucosa shows no signs of hyperplasia (E), or expansion of the proliferative zone (F).View Large Image Figure ViewerDownload Hi-res image Download (PPT)Supplementary Figure 4Wnt pathway activation induces cell proliferation in APC1311 mutant colonic adenomas. H&E, β-catenin, and Ki67 staining of serial sections of dysplastic crypt with low-grade (LG) intraepithelial neoplasia (IEN) (upper panel); adenoma with LG IEN (middle panel), and adenoma with focal high-grade (HG) IEN (lower panel). Dysplastic cells show strong nuclear and cytoplasmic β-catenin staining and expression of the Ki67 proliferation marker.View Large Image Figure ViewerDownload Hi-res image Download (PPT)Supplementary Figure 5Intestinal origin, mitogen-activated protein kinase (MAPK) and c-MYC activation in porcine APC1311 mutant adenomas. Cytokeratin-19 (CK19), CDX2 phospho ERK1/2 (phospho-p44/42 MAPK [Thr202/Tyr204]), and c-MYC staining of colonic adenoma with low-grade intraepithelial neoplasia (LG IEN) (left panel) and normal colonic mucosa (right panel). Dysplastic cells show strong staining of the colonic epithelial cell markers CK19 and CDX2, activation of the MAPK pathway, and induction of c-MYC expression. Normal colonic mucosa shows MAPK activation and c-MYC expression only in the proliferative zone.View Large Image Figure ViewerDownload Hi-res image Download (PPT) Covering the CoverGastroenterologyVol. 143Issue 5PreviewFamilial adenomatosis polyposis (FAP) syndrome is a genetic disorder characterized by numerous adenomatous polyps that first appear in the large intestine and in later years in the stomach and duodenum. About 90% of all patients with classic FAP develop colon cancer by age 45. Full-Text PDF Genotypes and Phenotypes: Animal Models of Familial Adenomatous Polyposis ColiGastroenterologyVol. 143Issue 5PreviewThe report of a porcine model of familial adenomatous polyposis coli (FAP) by Flisikowska et al1 in this issue of Gastroenterology provides a new animal model of adenoma formation that will permit the study of allelic variation at the pAPC locus and the polyposis phenotype. It adds to the valuable panel of animal models that allow investigation of the disrupted genetic and signaling mechanisms that lead to colon tumor formation in the human. It also provides a novel model in its replication of the human colonic phenotype that may be very useful to evaluate therapeutic interventions. Full-Text PDF
Referência(s)