A target-selected Apc-mutant rat kindred enhances
the modeling of familial human colon cancer
James M. Amos-Landgraf*, Lawrence N. Kwong*, Christina M. Kendziorski†, Mark Reichelderfer‡, Jose Torrealba§,
Jamey Weichert¶, Jill D. Haag*, Kai-Shun Chen*, Jordy L. Waller*, Michael N. Gould*, and William F. Dove*?**
Departments of†Biostatistics and Medical Informatics,§Pathology and Laboratory Medicine,¶Radiology, and‡Medicine, Section of Gastroenterology
and Hepatology, *McArdle Laboratory for Cancer Research, and?Laboratory of Genetics, University of Wisconsin School of Medicine and Public Health,
Madison, WI 53726
Contributed by William F. Dove, January 3, 2007 (sent for review December 19, 2006)
Progress toward the understanding and management of human
colon cancer can be significantly advanced if appropriate experi-
a rat model carrying a knockout allele in the gatekeeper gene
Adenomatous polyposis coli (Apc) recapitulates familial colon can-
cer of the human more closely than existing murine models. We
have established a mutagen-induced nonsense allele of the rat Apc
gene on an inbred F344/NTac (F344) genetic background. Carriers
of this mutant allele develop multiple neoplasms with a distribu-
tion between the colon and small intestine that closely simulates
that found in human familial adenomatous polyposis patients. To
been designated the polyposis in the rat colon (Pirc) kindred. The
Pirc rat kindred provides several unique and favorable features for
the study of colon cancer. Tumor-bearing Pirc rats can live at least
17 months, carrying a significant colonic tumor burden. These
tumors can be imaged both by micro computed tomography
scanning and by classical endoscopy, enabling longitudinal studies
of tumor genotype and phenotype as a function of response to
chemopreventive and therapeutic regimes. The metacentric char-
acter of the rat karyotype, like that of the human and unlike the
acrocentric mouse, has enabled us to demonstrate that the loss of
the wild-type Apc allele in tumors does not involve chromosome
loss. We believe that the Pirc rat kindred can address many of the
current gaps in the modeling of human colon cancer.
chromosome biology ? genomic instability ? Min mouse ?
virtual colonoscopy ? endoscopy
estimated for 2006 in the United States (1). Beyond early detection
and surgical removal of the adenomatous precursor lesions, ther-
apeutic approaches to colon cancer are currently inadequate.
The majority of human sporadic and familial adenomatous
polyposis (FAP) colonic tumors involve mutations that inactivate
the regulatory gene APC. Mouse strains carrying mutagen-induced
or targeted mutations in the ortholog Apc develop intestinal
adenomas. Thus, the APC/Apc gene has been designated a ‘‘gate-
keeper’’ tumor-suppressor gene (2).
Most mouse strains mutated in Apc, including the ApcMin(Min)
several months from multiple such adenomas (3). Derivatives in
which an Apc mutation is combined with mutations in the ho-
meobox gene Cdx2 (4) or the Tgf?-signaling element Smad3 (5)
shift the proportion of tumors toward the colon. But the former
construct creates a genomic instability and the phenotype of the
latter construct seems to depend on a contribution from the
commensal microbial flora, unlike the Min mouse (6).
Other platforms for the experimental investigation of colon
cancer have their own limitations. Carcinogen-induced colonic
neoplasms in the mouse and rat arise with a long latency at low
multiplicity and incomplete penetrance. Models in which
olon cancer is a major cause of morbidity and mortality in the
Western world: 148,610 new cases and 55,170 deaths are
human tumors are xenografted into immunodeficient mice
lack the microenvironment of the corresponding autochtho-
nous tumor (7, 8). The deficiencies in modeling human colon
cancer with animal and in vitro models have generated a
challenge recently summarized by Sjoblom et al. (9): ‘‘For
cancer biology, it is clear that no current animal or in vitro
model of cancer recapitulates the genetic landscape of an
actual human tumor.’’
The world of biomedical research on colon cancer is severely
hampered by the lack of a reliable preclinical experimental plat-
form. We have chosen to investigate whether a rat model carrying
a knockout allele in Apc can simulate more closely the human
disease. In pursuing this possibility, we have been emboldened by
of interest (10–13) and by the report that chemically induced
colonic neoplasms in the rat can metastasize to distant sites such as
the liver (14).
Our goals in seeking an enhanced experimental model for
familial human colon cancer include the following: (i) to incorpo-
rate a dependence on a mutation in Apc, thus matching genetically
the majority of FAP and sporadic colon cancers in the human; (ii)
to simulate the regional distribution between the colon and small
intestine of neoplasms found in the human; (iii) to enhance the
sizes; and (iv) to create a platform permitting prospective longitu-
dinal studies with enhanced statistical power through imaging and
report presents evidence that we have achieved important aspects
of each of these goals.
The Generation, Detection, and Molecular Characterization of a Rat
Knockout Allele of Apc. To identify a founder for a mutant Apc rat
Apc gene (Fig. 1A). A single male harbored a heterozygous point
Author contributions: J.M.A.-L. and L.N.K. contributed equally to this work; J.M.A.-L.,
J.L.W. performed research; J.M.A.-L., L.N.K., C.M.K., J.D.H., and K.-S.C. contributed new
reagents/analytic tools; J.M.A.-L., L.N.K., C.M.K., and J.T. analyzed data; and J.M.A.-L.,
L.N.K., and W.F.D. wrote the paper.
The authors declare no conflict of interest.
Freely available online through the PNAS open access option.
Abbreviations: CT, computed tomography; ENU, N-ethyl-N-nitrosourea; F344, F344/NTac;
FAP, familial adenomatous polyposis; LOH, loss of heterozygosity; Min, ApcMin; Pirc,
Data deposition: The sequence reported in this paper has been deposited in the Rat
Genome Database (accession no. 1554322).
**To whom correspondence should be addressed at: McArdle Laboratory for Cancer
This article contains supporting information online at www.pnas.org/cgi/content/full/
© 2007 by The National Academy of Sciences of the USA
March 6, 2007 ?
vol. 104 ?
change at codon 1137 (Fig. 1B). No other mutations were found
after sequencing the coding region and intron–exon junctions 5? of
the mutation. We therefore named this allele Apcam1137. The
predicted Apc protein would be truncated at the third amino acid
of the second 15-aa ?-catenin binding domain (Fig. 1C), which is
highly conserved among vertebrates.
Neoplasms in Heterozygotes and Embryonic Lethality of Homozygotes
for the Apcam1137Allele. We analyzed the phenotype of F344-
Apcam1137/?animals at the second and third backcross generations
mutation have survived ?17 months of age with no external signs
of disease. Dissections at ages ranging from 88 to 397 days revealed
multiple neoplasms in the intestinal tract of both male and female
heterozygotes (Table 1). Importantly, extensive polyposis was ob-
served in the colon, with a 100% incidence after 4 months of age.
The strain was thus designated the F344-polyposis in the rat colon
Mutant rats in the F344-Pirc kindred also developed a number
intestine (Table 1). Analysis of the molecular histopathology of
polyps and microadenomas will be discussed below. Full necropsies
of animals carrying the Apcam1137mutation also uncovered tumors
in a majority of animals, but no overt distant metastases were
The Pirc polyposis phenotype segregated perfectly with the
Apcam1137allele in 56 carriers and 23 noncarriers, supporting the
hypothesis that this mutation causes the disease. There is less than
a 5% chance that a separate mutation ?3.8 cM from Apc would
Scheme for the colorimetric yeast assay. Two thousand five
hundred and thirty bases of Apc exon 15 spanning codons
757–1,600 were amplified with primers chimeric for Apc se-
such chimeric amplicon (11). The amplicon was then gap-
yeast plate with half-red and half-white colonies, which is the
expected ratio for a heterozygous mutant. (B) Sequence trace
version at nucleotide 3409 of Apc (Upper) compared with a
wild-type littermate (Lower). (C) Structure of the human Apc
gene. Arrows indicate orthologous locations of mouse model
and Pirc truncating mutations and the two most common FAP
mutation sites. The color bar below indicates the genotype–
Isolation and identification of the Pirc line of rats. (A)
Table 1. Tumor multiplicities in Pirc rats
mean ? SD
Lesions in small intestine,
mean ? SD
352 ? 1
8 ? 3
14 ? 8
3 ? 2
5 ? 3
7 ? 5
79 ? 11
11 ? 12
7 ? 9
14 ? 5
22 ? 9
0 ? 0
2 ? 2
4 ? 5
57 ? 13
18 ? 8.5
21 ? 20
88 ? 64
178 ? 116
1 ? 2
19 ? 29
35 ? 44
665 ? 103
208 ? 223
Colonic microadenoma multiplicities could not be accurately measured without histopathological confirma-
tion and were excluded from these analyses. Neoplasms ?0.5 mm in size were classified as microadenomas.
*Controls were injected with phosphocitrate buffer plus ethanol without ENU.
Amos-Landgraf et al.
March 6, 2007 ?
vol. 104 ?
no. 10 ?
show this degree of association [see supporting information (SI)
Text). No significant differences in the multiplicity or phenotype of
intestinal tumors were seen between successive backcross genera-
tions of the F344-Pirc kindred on the inbred F344 background,
indicating that any effect of ENU-induced modifying alleles is
We investigated whether the Apcam1137mutation is homozygous-
or a [F344 ? WF/NHsd (WF)] F1background. No homozygous
mutants were obtained of 71 total progeny, with heterozygotes and
wild-type progeny exhibiting a 49:22 segregation, not significantly
different from the expected 2:1 Mendelian ratio (P ? 0.8, ?2test).
Thus, the Apcam1137allele is homozygous-lethal on two genetic
a second ENU-induced mutation, it would lie within 2.4 cM (95%
CI) of the Apcam1137mutation (SI Text).
Age, Gender, and Carcinogen (ENU) Effects on the Pirc Phenotype. A
significant dependence on gender was observed, with males devel-
oping more tumors throughout the intestinal tract. There was also
a monotonic increase in tumor multiplicity with age, reaching an
average of 14 colonic adenomas in males and 7 in females ? 8
months of age (Table 1 and SI Table 4). The presence of colonic
tumors in females at early ages, when tumors of the small intestine
are not detected (Table 1), implies that colonic tumors arise first.
To increase the tumor multiplicity and consequent statistical
injected with a single dose of 40 mg/kg of ENU at 2 weeks of age.
At 7 months of age, treated Pirc animals developed nearly 80
colonic tumors (Table 1) without major external signs of distress.
This finding represents a 7-fold increase over mock-treated Pirc
controls, whereas only a 3-fold increase was observed in macroad-
enomas of the small intestine. The wild-type littermate control rats
receiving the same dose of ENU did not develop any detectable
intestinal lesions. ENU-treated Pirc rats could therefore provide
enhanced statistical power for experimental studies of chemopre-
ventive and therapeutic regimens.
Long-Lived Pirc Animals Develop Adenocarcinomas with Local Inva-
sion. The histopathology and morphology of the tumors closely
resembled that of human tumors, with adenomatous changes
evident, including dysplasia, nuclear enlargement, an increased
columnar architecture (Fig. 2). Grossly, most colonic tumors were
peduncular, whereas adenomas in the small intestine had a flat
appearance. Each type of adenoma frequently reached 1 cm in
diameter in animals over 6 months of age and up to 2 cm in those
approaching 1 year. Immunofluorescent staining of tumors re-
vealed nuclear and cytoplasmic accumulation of ?-catenin within
dysplastic cells (Fig. 2 F and G) as well as up-regulation of the
proliferation marker Ki-67 (data not shown). By contrast, microad-
enomas of the colon and small intestine expressed Ki-67 but failed
to show a convincing nuclear accumulation of ?-catenin in any of
?5,000 cells assayed (Fig. 2 H–J and data not shown). Longitudinal
studies are needed to ascertain the neoplastic potential of these
lesions. Importantly, in animals at 6 months of age or greater, 3 of
14 histologically examined colonic tumors were shown to have
high-grade dysplasia accompanied by the local invasion of neoplas-
tic cells into the stalk, classifying the tumors as adenocarcinomas
human (Fig. 2 A–C).
Loss of Heterozygosity at the Apcam1137Site: Chromosome Loss Is Not
Involved. Loss of heterozygosity (LOH) at the Apcam1137site on the
inbred F344 background was quantitatively assayed by using Pyro-
sequencing technology. Pyrosequencing is a registered trademark
of Biotage (Uppsala, Sweden). Control assays of allele ratio were
performed on DNA from normal intestinal tissue from Apcam1137/?
heterozygotes. The allele ratios determined on tumor and control
DNA were then plotted together and analyzed by a Gaussian
mixture model (SI Fig. 5). The majority of F344 adenomas in the
colon (87%, 34 of 39) and small intestine (100%, 24 of 24) showed
LOH of the wild-type Apc allele at codon 1137.
Studies of LOH in the mouse face an obstacle in rigorously
ascertaining chromosome loss with or without reduplication.
The acrocentric character of the mouse karyotype prevents a
systematic survey of both sides of the centromere. When LOH
involves somatic recombination, only one arm of the chromo-
some is usually involved, in contrast to the chromosome-wide
pattern found for chromosome-loss events. The rat karyotype is
metacentric, by contrast, permitting us to test whether the LOH
event observed in tumors involved whole-chromosome loss or
loss followed by reduplication.
addressed whether the function of the wild-type WF allele of Apc
adenocarcinoma with high-grade dysplasia. (B) Enlargement of the larger rect-
rectangle in A (C), showing high-grade dysplasia compared with normal crypts
(red) and DAPI (blue) immunofluorescence of the same tumor. The dashed line
delineates dysplastic (above the line) and hyperplastic and normal tissue (below
the line). (G) Magnification of the rectangle shown in F. (H) H&E of a colonic
microadenoma (central crypt) surrounded by normal crypts, which is represen-
tative of all colonic microadenomas examined. (I) ?-catenin (red). (J) ?-catenin
(red) merged with DAPI (blue). (Scale bars: A and E, 1 mm; H, 0.1 mm.)
Histological and gross appearance of Pirc tumors. (A) H&E of a focal
ss48531727 (43 Mb) on the q arm, were all heterozygous in the normal tissue.
The centromere (open circles) lies at approximately the 38-Mb position. LOH
status at each SNP was determined by using a quantitative Pyrosequencing
assay. Four possible tumor genotypes are given (left to right): LOH involving
only the two loci on the p arm, LOH involving only Apcam1137, maintenance of
heterozygosity (MOH) at all three loci, and LOH for all three loci. We have
diagrammed homozygosity; it must be noted that these Pyrosequencing
www.pnas.org?cgi?doi?10.1073?pnas.0611690104Amos-Landgraf et al.
is lost through elimination of the entire WF chromosome. Poly-
morphic SNPs on the p and q arms of rat chromosome 18 were
3). None of 22 colonic or 18 small intestinal tumors showed loss of
both arms of the WF homolog. Rather, 16 colonic and 12 small
intestinal tumors showed single-arm LOH, with all loss events
involving the Apc locus on the p arm and often extending at least
10 Mb distal of Apc. Each case of LOH entailed loss of the WF
allele and maintenance of the F344 allele. Thus, most tumors in the
Pirc rat involve LOH at the Apc locus either by somatic recombi-
nation or by extended deletion. The remaining 12 tumors main-
tained heterozygosity over the entire chromosome 18. These tu-
mors may have either silenced the wild-type Apc allele or
inactivated it by an intragenic mutation (15, 16).
The Chemopreventive Action of Celecoxib. The Min mouse strain
has enabled the analysis of the chemoprevention of intestinal
adenomagenesis. The nonsteroidal antiinflammatory agents pi-
roxicam (17), sulindac (18), and the clinically used celecoxib (19)
have been reported to show significant efficacy in the Min mouse
(for a review, see ref. 20). Statistically significant evidence for
these effects depended on the high multiplicity of adenomas in
the small intestine. By contrast, colonic tumor multiplicities in
the range of two compromised the power of tests for an effect
in the colon. Study designs involving large numbers of animals,
enhanced colonic tumor multiplicities, or longitudinal analysis of
individual imaged tumors would permit the analysis of response
for colonic neoplasms.
The F344-Pirc rat kindred overcomes this obstacle. We have
carried out an investigation of the action of celecoxib to prove this
principle (Table 2). Treated mutant rats were fed celecoxib in their
chow from 40 days of age and were killed at 6–7 months of age.
for both male and female rats and in the small intestine for males.
Only a small number of tumors were found in the small intestine of
females (also see Table 1), preventing a rigorous statistical analysis
of this class of tumor.
The Rat Permits both Classical Endoscopy and Virtual Colonoscopy.To
determine whether longitudinal in vivo studies of individual intes-
tinal tumors can be carried out in trials with agents such as
celecoxib, an 11-month-old F344-Pirc rat was anesthetized and its
tumors visualized by endoscopy. As shown in Fig. 4B, a 6-mm-
diameters 5.3, 5.7, and 6.8 mm. The same tumors were identified in
three-dimensional micro computed tomography (CT) images (Fig.
4A) and confirmed upon dissection (Fig. 4C).
It is widely believed that extant animal models of colon cancer do
not fully recapitulate the human disease. Specifically, a major
practical consideration is the lack of a significant colonic tumor
burden for genetic, therapeutic, or diet studies. Either the abun-
dance of small intestinal tumors prevents long-term analysis or the
low absolute colonic tumor multiplicity hinders statistical analysis.
In Table 3, we summarize the salient considerations in seeking an
experimental model for human colon cancer. Importantly, the Pirc
rat addresses several of the major drawbacks of all Apc-based
models of familial intestinal cancer in the mouse.
tract. In measuring this component, we took into account the
nuclear and cytoplasmic translocation of ?-catenin, which is con-
sidered a hallmark of Apc-loss-associated neoplasia. Because no
microadenomas showed such accumulation, only macroadenomas
were considered. The ratio of macroadenoma multiplicities in the
colon to that in the small intestine in Pirc rats averages 1:1. By
contrast, most lines of C57BL/6-Min mice have an average ratio of
(4), and Smad3?/?ApcMin/?mice develop polyps only in the colon
that of the FAP human. Furthermore, the short lifespans of both
models make them less than ideal for chemopreventive studies. In
comparing the mouse and rat with humans, it is important to note
that human FAP patients do not develop polyps solely in the colon.
Rather, the polyp incidence in the small intestine ranges from 58%
to 74% at first endoscopy and approaches 100% by 70 years (21),
with multiplicities exceeding 80 in 17% of patients monitored with
video capsule endoscopy (22). Thus, the distribution of intestinal
tumors in the Pirc rat resembles that in human FAP patients.
The Pirc rat also improves on another aspect: the absolute
incidence and multiplicity of colonic tumors, which are higher in
F344-Pirc rats than in carcinogen-treated wild-type F344 rats or in
Min mice (Table 3). The ENU treatment of Pirc rats provides an
even greater advantage in the statistical power of studies of
chemopreventive and therapeutic protocols. Additionally, Pirc co-
lonic tumors are capable of reaching a diameter in excess of 1 cm,
enhancing the analysis of the volume and cellular composition of a
tumor. Finally, the capacity for longitudinal studies in long-lived
Pirc rats provides further qualitative advantages by using either
specialized microCT instrumentation or, more generally, the clin-
ical bronchoscope to bypass dissection as the necessary endpoint.
Our proof of principle of in vivo tumor imaging paves the way for
investigation into the sensitivity and specificity of endoscopy versus
bar: 1 cm.)
In vivo imaging of Pirc tumors. MicroCT (A), endoscopic (B), and dissection (C) views of three colonic tumors in an 11-month-old F344 Pirc male. (Scale
Table 2. The effect of celecoxib on the multiplicity of intestinal
tumors >1 mm in diameter in F344 Pirc rats
Tumor multiplicity, mean ? SD
(no. of rats)
1.3 ? 1.2 (12)
1.2 ? 0.9 (12)
0.8 ? 0.9 (12)
0.3 ? 0.5 (12)
7.6 ? 4.3 (11)
3.6 ? 2.7 (11)
0.6 ? 0.8 (15)
1.3 ? 0.7 (15)
Animals were treated from 40 days of age with 1,200 ppm celecoxib in
Teklad 8604 chow and euthanized at 6–7 months of age. Tumors were
counted on freshly dissected tissue without using a dissecting microscope. P
values were determined by using the Wilcoxon rank sum test.
Amos-Landgraf et al.
March 6, 2007 ?
vol. 104 ?
no. 10 ?
virtual colonoscopy (23). Endoscopy also enhances the statistical
power of pharmacologic and genetic investigations, minimizing the
numbers of tumor-bearing rats required in these studies. We note
with interest that the chemopreventive action of celecoxib is not
complete (Table 3). Thus, such tumor resistance, as well as longi-
tudinal studies of tumor progression, regression, and recurrence,
can be related prospectively to molecular profiles of endoscopic
portion to progress represents a unique opportunity unavailable in
humans or in mice.
one major genetic advantage: the metacentric rat karyotype per-
mits a direct test of the mechanism of LOH that is unavailable in
acrocentric mouse models. In future detailed studies of genomic
stability versus instability during the initiation and progression of
colonic neoplasms, the analysis will benefit significantly not only
from the genetic homogeneity of inbred strains, but from the fact
that the metacentric karyotype can distinguish between whole
chromosome loss and somatic recombination or deletion in vivo.
Finally, the familial colonic tumors we have observed to date
to human stage T1 lesions (Fig. 2 A and B). In developing the Pirc
kindred in the rat, we are encouraged by the report of metastasis
to the liver of chemically induced colonic adenocarcinomas in
wild-type rats (14). We are also encouraged by the progress of the
molecular genetics in the rat, including a rapidly closing genome
project (24) and a growing set of genetic resources including large
recombinant–inbred sets (25), transgenic strains, and ENU-
induced mutant alleles in genes of interest (10, 12). Any of these
resources could help fulfill the need for an animal model of
significantly progressed adenocarcinomas.
Intriguingly, the Pirc rat model shows a significant gender bias in
tumor multiplicity and distribution (Table 1). Females are strongly
protected from intestinal tumor development and more often
become moribund due to the development of jaw tumors and
related teeth abnormalities, rather than tumor-related anemia.
Women do have a slightly reduced incidence of colorectal cancer
compared with men, and recent epidemiologic studies indicate that
women have a delayed onset of advanced disease (26, 27). This
gender effect may reflect the protective effect of hormones on
colon tumor development that has been observed in women on
hormone replacement therapy (28).
To have in hand a second experimental mammalian species with
tumors in the human, the mouse, and the rat, we can learn what
aspects of the transcriptome, proteome, and biology are species-
specific and what aspects are broadly conserved. By comparing
autochthonous colonic tumors in the rat with their properties as
xenografts in immunodeficient mice, investigators can become
tal platforms when studying the biology, chemoprevention, and
therapeutics of human colon cancer.
Materials and Methods
Animal Maintenance. Ratsweremaintainedinstandardcagesunder
a university-approved animal protocol in an American Association
of Laboratory Animal Care-approved facility. Rats were fed either
WI), with access to an automatic supply of acidified water. Tumor
counts were assessed as described with mice in ref. 16, except as
noted in Table 2. The following inbred rat strains were used in this
study: F344/NTac (Taconic, Hudson, NY) and WF/NHsd (Harlan,
Colorimetric Yeast Assay. The ENU treatment of male rats for the
mutagenesis screen was performed as described in ref. 10. F344
males were injected with 60 mg/kg of ENU once per week for 2
weeks. Long-range PCR from genomic DNA was performed by
using the following chimeric universal vector primers: forward,
GGC CAT CGA TAG CTC GAT GTA ACG TGC AGT TAA
CGC CCA TGT CTC CTG GCT CAA GTT TGC; and reverse,
CCT ACT AAC AGA TAC GCT ATG CAG GAC TCT GGA
TTG CCC TGT TGG CAT GGC TGA AAT AA. The final PCR
concentration mix was 1x Herculase Hotstart buffer (Stratagene,
Hotstart High-Fidelity DNA Polymerase (Stratagene), 2 ?l of
2 min, then 35 cycles of 94°C for 45 sec, 61°C for 45 sec, and 72°C
PCR products were confirmed by gel electrophoresis and cotrans-
formed into yeast along with the universal vector (11).
were developed to genotype the mutant SNP site of the Apcam1137
allele by restriction enzyme digestion. In one amplicon, the mutant
allele was cut by NheI; in the other amplicon, the wild-type allele
was cut by HindIII. These complementary methods control for
incomplete digestion. The primer sequences were as follows: (for
the NheI amplicon) forward, GGA AGA CGA CTA TGA AGA
TGG and reverse, TGC CCT GTA CTG ATG GAG; and (for
the HindIII amplicon) forward, AAT AAC GTT CAC TGT
AGT TGG TAA GCT and reverse, AGG CAA TCA AGA
AGC CAG AA.
MicroCT and Endoscopy. Normal chow was replaced with a nonsolid
diet the night before microCT imaging. Anesthesia was adminis-
tered through the regulated flow of isoflurane vapor (1–2%)
through a nose cone. The colon was flushed with a warm PBS
enema (40 ml). The colon was insufflated with air (20–40 ml), and
Digital Images. Images for Fig. 2 A–E and H were taken by using
automatic exposure with a Spot digital camera (Diagnostic Instru-
ments, Sterling Heights, MI) mounted on a Zeiss (Oberkochen,
Germany) Axiophot microscope, with ?2.5, ?20, ?40, ?40, ?5,
taken with a Zeiss Axiocam HRm mounted on a Zeiss Axiovert
Table 3. Comparison of FAP human, Min mouse, and Pirc rat colorectal cancer phenotypes.
Human FAP/animal model
Average colonic tumor count
Average colonic tumor
tumor count ratio
Human FAP (codon 700-1500)
C57BL/6J-Min mouse (codon 850)
F344 rat treated with carcinogen ?typically
?-catenin mutations (ref. 31)?
F344-Pirc rat (codon 1137)
?100 (?12 years)
2–4 (?2 months)
1–2 (?10 months)
Metacentric80–90 (ref. 32)
10 (?4 months)1001:1Metacentric
Ranges are for strain averages, not for individual animals. The majority of data for carcinogen-treated rats uses time points later than 10 months of age.
www.pnas.org?cgi?doi?10.1073?pnas.0611690104 Amos-Landgraf et al.
200M microscope with ?10, ?63, ?40, and ?40 objectives, re-
Axiovision software (version 22.214.171.124). Fig. 2F is a composite of
several overlapping images. The microCT image in Fig. 4A was
acquired on a Siemens (Iselin, NJ) MicroCAT-2 scanner. Acqui-
sition proceeded for 8 min (80 kVp, 500 ?A, 400 steps, one frame
per view, 360° rotation, 93 ? 93 ? 100 ?m voxel size) and
reconstruction was done by using a Shepp–Logan filter with back
projection. Amira software (version 4.1; TGS, San Diego, CA) was
production. Accordingly, isosurface images with appropriate den-
sity threshold levels and down-sampled two times along each axis
as well as volume-rendered images were created by using the
appropriate Amira functions. No further image manipulation was
performed. The endoscopy image in Fig. 4B was taken with an
EG-1870K 6.0-mm color CCD chip video gastroscope and an EPK
1000 video processor (Pentax Medical, Montvale, NJ). The image
was captured on the workstation and a hard copy image generated
with a Sony Mavigraph video printer (Sony, New York, NY). Fig.
4C was photographed with an Olympus (Melville, NY) Camedia
C-5050ZOOM digital camera by using automatic exposure.
Somatic ENU Treatment. ENUinjectionwasperformedasdescribed
in ref. 29.
as described in ref. 30, and images were acquired as described
Pyrosequencing Assay. Formalin-fixed tumors were excised under a
dissecting microscope to minimize contamination from nontumor
and any surrounding hyperplastic villi were excluded to enrich for
Excised tissue was incubated overnight in distilled water at 65°C to
reverse formalin cross-links. NaOH (100 mM) was added to a final
concentration of 50 mM, and samples were incubated at 95°C for
at least 4 hours. Tris?HCl (pH 5.5) was then added to neutralize the
solution, and the samples were briefly centrifuged. The final PCR
MgCl2, 0.2 mM dNTP, 264 pM of each primer, 0.6 units of GoTaq
Flexi (Promega, Madison, WI), 8 ?l of DNA, and ddH2O to 50 ?l.
The PCR cycling profile was as follows: 94°C for 3 min, followed by
50 cycles of 94°C for 15 sec, 57°C for 1 min and 30 sec, and 72°C for
2 min, with a final elongation step of 72°C for 10 min. Pyrose-
quencing assays were performed according to the manufacturer’s
protocols with Pyro Gold Reagents on a PSQ96MA machine and
PSQ 96MA version 2.1 software (Biotage). Forty microliters of
base peak heights of ?120 units were included. Primer sequences
were as follows: (for the Pirc SNP) forward, ATG TGA ACC AGT
CTT TGT GTC AG; biotinylated, reverse, ATG CTG TTC TTC
CTC AGA ATA ACG; sequencing, GGA AGA CGA CTA TGA
AGA T; [for the p arm, SNP (dbSNP ss48531311)], forward, GTG
GAA ACG AAG CAT CAT TCT GA; biotinylated, reverse, TGC
TGT TCT AAA TTG CAC GTT TAC; sequencing, CGT ATT
GGG TTG TGA GA; [for the q arm SNP (dbSNP ss48531727)],
biotinylated, forward, TCA AAC AGA AGG CAG TTT ATT
CAG; reverse, GGG GGT AAA ATA ATA TGC CGA GA;
sequencing, TCT TAG TAA TGT ACC AGA TG. An LOH/
maintenance of heterozygosity cutoff value of 32.8% was deter-
mined by adapting the normal mixture technique from Shoemaker
et al. (15) (SI Fig. 5).
Celecoxib Treatment. Ratsweretreatedwith1,200ppmcelecoxibin
Teklad 8604 chow given ad libitum, were housed, maintained, and
dissected, and their tumors were counted independently of those in
We thank Michael Newton for the analysis of LOH data; Linda Clipson for
critical reading of the manuscript and assistance with its preparation;
Norman Drinkwater, Alexandra Shedlovsky, and Richard Halberg for
critical reading; Henry Pitot and Ruth Sullivan for assistance with histo-
pathology; Jane L. Remfert and Yunhong Zan for technical assistance with
rat husbandry and yeast screening; and Ben Durkee for assistance with
microCT imaging. J.M.A.-L and L.N.K. were supported in part by the
National Cancer Institute Institutional Training Grants CA009681 and
CA009135, respectively. This work was supported in part by National
Institutes of Health Grants CA106216 (to M.N.G.) and CA63677 (to
W.F.D.) and by the University of Wisconsin School of Medicine and Public
Health. This is publication no. 3633 from the Laboratory of Genetics,
University of Wisconsin, Madison.
1. American Cancer Society (2006) Cancer Facts & Figures 2006 (Am Cancer Soc,
2. Kinzler KW, Vogelstein B (1998) in The Genetic Basis of Human Cancer, eds
Vogelstein B, Kinzler KW (McGraw–Hill, New York), pp 241–242.
3. Moser AR, Pitot HC, Dove WF (1990) Science 247:322–324.
4. Aoki K, Tamai Y, Horiike S, Oshima M, Taketo MM (2003) Nat Genet 35:323–330.
5. Sodir NM, Chen X, Park R, Nickel AE, Conti PS, Moats R, Bading JR, Shibata D,
Laird PW (2006) Cancer Res 66:8430–8438.
6. Dove WF, Clipson L, Gould KA, Luongo C, Marshall DJ, Moser AR, Newton MA,
Jacoby RF (1997) Cancer Res 57:812–814.
7. O’Brien CA, Pollett A, Gallinger S, Dick JE (2007) Nature 445:106–110.
8. Ricci-Vitiani L, Lombardi DG, Pilozzi E, Biffoni M, Todaro M, Peschle C, De Maria
R (2007) Nature 445:111–115
9. Sjoblom T, Jones S, Wood LD, Parsons DW, Lin J, Barber TD, Mandelker D, Leary
RJ, Ptak J, Silliman N, et al. (2006) Science 314:268–274.
10. Zan Y, Haag JD, Chen KS, Shepel LA, Wigington D, Wang YR, Hu R, Lopez-
Guajardo CC, Brose HL, Porter KI, et al. (2003) Nat Biotechnol 21:645–651.
11. Chen KS, Gould MN (2004) Biotechniques 37:383–388.
12. Smits BM, Mudde JB, van de BJ, Verheul M, Olivier J, Homberg J, Guryev V, Cools
AR, Ellenbroek BA, Plasterk RH, et al. (2006) Pharmacogenet Genomics 16:159–169.
13. Smits BM, Guryev V, Zeegers D, Wedekind D, Hedrich HJ, Cuppen E (2005) BMC
15. Shoemaker AR, Moser AR, Midgley CA, Clipson L, Newton MA, Dove WF (1998)
Proc Natl Acad Sci USA 95:10826–10831.
16. Haigis KM, Dove WF (2003) Nat Genet 33:33–39.
17. Jacoby RF, Marshall DJ, Newton MA, Novakovic K, Tutsch K, Cole CE, Lubet RA,
Kelloff GJ, Verma A, Moser AR, et al. (1996) Cancer Res 56:710–714.
18. Beazer-Barclay Y, Levy DB, Moser AR, Dove WF, Hamilton SR, Vogelstein B,
Kinzler KW (1996) Carcinogenesis 17:1757–1760.
19. Jacoby RF, Seibert K, Cole CE, Kelloff G, Lubet RA (2000) Cancer Res 60:5040–
20. Corpet DE, Pierre F (2003) Cancer Epidemiol Biomarkers Prev 12:391–400.
21. Bulow S, Bjork J, Christensen IJ, Fausa O, Jarvinen H, Moesgaard F, Vasen HF
(2004) Gut 53:381–386.
22. Schulmann K, Hollerbach S, Kraus K, Willert J, Vogel T, Moslein G, Pox C, Reiser
M, Reinacher-Schick A, Schmiegel W (2005) Am J Gastroenterol 100:27–37.
23. Pickhardt PJ, Choi JR, Hwang I, Butler JA, Puckett ML, Hildebrandt HA, Wong RK,
Nugent PA, Mysliwiec PA, Schindler WR (2003) N Engl J Med 349:2191–2200.
24. Gibbs RA, Weinstock GM, Metzker ML, Muzny DM, Sodergren EJ, Scherer S, Scott
G, Steffen D, Worley KC, Burch PE, et al. (2004) Nature 428:493–521.
25. Shisa H, Lu L, Katoh H, Kawarai A, Tanuma J, Matsushima Y, Hiai H (1997) Mamm
26. Jackson-Thompson J, Ahmed F, German RR, Lai SM, Friedman C (2006) Cancer
27. Regula J, Rupinski M, Kraszewska E, Polkowski M, Pachlewski J, Orlowska J,
Nowacki MP, Butruk E (2006) N Engl J Med 355:1863–1872.
28. Rossouw JE, Anderson GL, Prentice RL, LaCroix AZ, Kooperberg C, Stefanick ML,
Jackson RD, Beresford SA, Howard BV, Johnson KC et al. (2002) J Am Med Assoc
29. Moser AR, Mattes EM, Dove WF, Lindstrom MJ, Haag JD, Gould MN (1993) Proc
Natl Acad Sci USA 90:8977–8981.
30. Haigis KM, Caya JG, Reichelderfer M, Dove WF (2002) Proc Natl Acad Sci USA
31. Corpet DE, Pierre F (2005) Eur J Cancer 41:1911–1922.
32. Kohno H, Suzuki R, Yasui Y, Hosokawa M, Miyashita K, Tanaka T (2004) Cancer
Amos-Landgraf et al.
March 6, 2007 ?
vol. 104 ?
no. 10 ?