GENES, CHROMOSOMES & CANCER 35:299 –310 (2002)

Application of Inter–Simple Sequence Repeat PCR to

Mouse Models: Assessment of Genetic Alterations in
Carcinogenesis
Fernando Benavides,1* Mónica Zamisch,1 Mónica Flores,1 Marcia R. Campbell,2 Susan E. Andrew,2 Joe M. Angel,1
Julien Licchesi,3 Gabriel Sternik,1 Ellen R. Richie,1 and Claudio J. Conti1
1
Department of Carcinogenesis, The University of Texas M.D. Anderson Cancer Center, Science Park–Research Division,
Smithville, Texas
2
Department of Medical Genetics, University of Alberta, Alberta, Canada
3
Cranfield BioMedical Centre, Cranfield University, Silsoe, Bedfordshire, United Kingdom

Genomic instability is believed to play a significant role in cancer development by facilitating tumor progression and tumor
heterogeneity. Inter–simple sequence repeat (inter-SSR) PCR has been proved to be a fast and reproducible technique for
quantitation of genomic instability (amplifications, deletions, translocations, and insertions) in human sporadic tumors.
However, the use of inter-SSR PCR in animal models of cancer has never been described. This new technique has been adapted
in our laboratory for the analysis of spontaneous and induced mouse tumors. We established the best PCR conditions for each
microsatellite-anchored primer and critically evaluated the reproducibility of the band patterns. We also studied the variation
of the fingerprints between and within various inbred mouse strains, including wild-derived lines. Tumor-specific alterations
were detected as gains, losses, or intensity changes in bands when compared with matched normal DNA. We quantitated the
extent of alterations by dividing the number of altered bands in the tumor by the total number of bands in normal DNA
(instability index). By means of inter-SSR PCR, we successfully analyzed genomic alterations in various mouse tumors, including
spontaneous thymic lymphomas developed in Msh2 knockout mice as well as chemically induced squamous cell carcinomas and
thymic lymphomas. Instability index values ranged between 0 and 9%, the highest levels observed in N-methyl-N-nitrosourea–
induced thymic lymphomas generated in Trp53 (p53) nullizygote (⫺/⫺) mice. We report here, for the first time, the use of
inter-SSR PCR to detect somatic mutations in mouse tumoral DNA, including laser-capture microdissected, methanol-fixed
tissues. These PCR-based fingerprints provide a novel approach to assessing the number and onset of mutational events in
mouse tumors and will help to understand better the mechanisms of carcinogenesis in mouse models.
© 2002 Wiley-Liss, Inc.

INTRODUCTION in three major forms: (1) aneuploidy, found in the

The major stages of carcinogenesis were de- majority of solid tumors, in which entire chromo-
duced over the past 50 years, primarily from animal somes are gained or lost; (2) chromosomal instabil-
model studies (particularly in the mouse skin), and ity (CIN), including insertions, deletions, translo-
are termed initiation, promotion, and progression. cations, amplifications, and others, all of which
Tumor initiation is characterized by the generation involve DNA breakage; and (3) point or oligobase
of a genetically altered cell resulting from exposure mutations, including microsatellite instability (MSI
to exogenous or endogenous carcinogens. The tu-
mor promotion stage involves the expansion of
initiated cells through epigenetic mechanisms, ac-
Abbreviations: CIN, chromosome instability; DMBA, 7,12-dim-
companied by alterations in the expression of crit- ethylbenz[a]anthracene; HNPCC, hereditary nonpolyposis colorec-
ical genes that regulate cell proliferation, differen- tal cancer; inter-SSR, inter–simple sequence repeat; LCM, laser-
captured microdissection; LOH, loss of heterozygosity; MMR,
tiation, and inflammation (Boland and Ricciardiello, mismatch repair; MNU, N-methyl-N-nitrosourea; MSI, microsatel-
1999; DePinho, 2000). During progression, pre- lite instability; PCR, polymerase chain reaction.
F. Benavides and M. Zamisch contributed equally to this work.
neoplastic cells become neoplastic as a result of Supported by: National Institutes of Health; Grant numbers:
progressive genomic instability and altered gene CA69146, CA90922, ES07784, CA16672.
expression (Loeb, 1996). *Correspondence to: Fernando Benavides, The University of
Texas M.D. Anderson Cancer Center, Department of Carcinogen-
Genomic instability reflects the propensity and esis–Science Park, Park Road 1C, P.O. Box 389, Smithville, TX
the susceptibility of the genome to acquire multi- 78957. E-mail: [email protected]
ple alterations and is believed to play a significant Received 18 February 2002; Accepted 24 April 2002
DOI 10.1002/gcc.10129
role in cancer development (Loeb, 1991; Loeb and Published online 15 September 2002 in
Loeb, 2000). In cancer, genomic instability appears Wiley InterScience (www.interscience.wiley.com).

© 2002 Wiley-Liss, Inc.

300 BENAVIDES ET AL.

or MIN), observed in DNA replication error inher- amounts of DNA and time-consuming Southern
ited syndromes and a small fraction of sporadic blot techniques.
cancers (Lengauer et al., 1997; Dasika et al., 1999; AP-PCR, also known as random amplified poly-
Luo et al., 2000; Richards, 2001). morphic DNA, is a PCR-based DNA fingerprinting
It has been proposed recently that the different technique that provides patterns of amplified frag-
forms of genetic instability in tumor cells reflect ments under low-specificity priming conditions
the selection pressures exerted by specific carcin- (Welsh and McClelland, 1990; Williams et al.,
ogens (Breivik and Gaudernack, 1999). Bardelli 1990; Welsh et al., 1991). Primer annealing at mul-
and colleagues (2001) demonstrated that exposure tiple sites generates many PCR products that form
to specific carcinogens can select for tumor cells a DNA fingerprint specific for each primer–DNA
with distinct types of genetic instability. Although template combination. However, AP-PCR can ex-
CIN is believed to be a common feature of most hibit low reproducibility and spurious bands. In
human tumors (Cahill et al., 1998), the stage at AP-PCR, the relative amplification levels of the
which such instability begins has not been defined. amplified DNA fragments are quantitative (reflect-
By analyzing single-nucleotide polymorphism al- ing differences in the relative abundance of the
lelic imbalance (as a marker for CIN) in human templates), allowing the detection of amplifications
colorectal tumors, Shih et al. (2001) demonstrated and deletions in cancer cells (Ionov et al., 1993;
this type of imbalance to be a common event, even Malkhosyan et al., 1998; Navarro and Jorcano,
in incipient small tumors. Since the identification 1999; Anami et al., 2000). Using a new approach
of several DNA mismatch repair (MMR) genes as called simultaneous hybridization of arbitrarily
the site of germline mutations associated with the primed PCR DNA fingerprinting products
rather uncommon human hereditary nonpolyposis (SHARP), Perucho and collaborators could deter-
colorectal cancer syndrome (HNPCC), it was clear mine the chromosomal origin of major bands from
that the DNA mismatch repair pathway is essential human AP-PCR fingerprints (Yasuda et al., 1996).
for the correct maintenance of the genetic material However, AP-PCR fingerprinting and SHARP
(Atkin, 2001). Over the past decade, it was shown have never been applied to genomic instability
that where there is an inactivating germline muta- studies in laboratory animal tumors.
tion in one of the MMR genes, most of the tumors CGH is a recently developed cytogenetics tech-
show MSI (Perucho, 1996; Boland et al., 1998). nique that allows the detection of changes in the
The MSI phenomenon was originally associated chromosomal copy number in a single experiment
with hereditary colorectal cancer but there are non- (Hermsen et al., 1996; Forozan et al., 1997). CGH
colonic and non-HNPCC tumors that display ele- has been used extensively in human cancer pathol-
vated frequencies of MSI. Genomewide measure- ogy because it provides a large-scale outline of
ment of the different types of instability has been chromosomal gains and losses throughout the ge-
made by a variety of techniques, including DNA nome of tumoral DNA (Knuutila et al., 1998; Weiss
fingerprinting, arbitrarily primed PCR (AP-PCR), et al., 1999). However, CGH suffers from limita-
microsatellite analysis, and comparative genomic tions in the sensitivity (only alterations larger than
hybridization (CGH). 2 megabases are detected), and its application has
In 1985, Jeffreys and collaborators described the not been extended to mouse tumor studies.
use of minisatellite multilocus probes, which si- Inter-SSR PCR was developed by Zietkiewicz
multaneously detect a large number of dispersed and colleagues as a new fingerprinting approach
hypervariable loci, to generate individual-specific utilizing (CA)n repeats as primer sites for inter-
DNA fingerprints (Jeffreys et al., 1985, 1987; Arm- microsatellite amplification (Zietkiewicz et al.,
our et al., 1989). This approach (the original “DNA 1994). This relatively new technique has several
fingerprinting”) was applied to the detection of advantages: (1) no sequencing is required to design
somatic changes in tumors of diverse type and the primers; (2) primers are anchored at the termini
origin. Specific genomic alterations have been de- of the (CA)n repeats and extend into the flanking
scribed in human rectal carcinomas (Thein et al., sequence by two nucleotide residues; (3) complex,
1987), B-lymphoproliferative disorders (de Jong et species-specific patterns are obtained from a vari-
al., 1988), Burkitt lymphoma and Wilms tumor ety of eukaryotic taxa; and (4) intraspecies poly-
(Hayward et al., 1990), ovarian cancer (Boltz et al., morphism is observed, and it segregates as Men-
1990), and DMBA-induced mouse liver tumors delian markers. Inter-SSR PCR has been proved to
(Ledwith et al., 1990). This classical DNA finger- be a fast and reproducible technique for quantita-
printing is a sensitive assay for genomic rearrange- tion of amplifications, deletions, translocations, and
ments, but depends on the availability of large insertions in human sporadic colorectal cancer
 INTER-SSR PCR IN MOUSE MODELS 301
(Basik et al., 1997; Stoler et al., 1999; Anderson et C58/J, C57BLKS, C3H/HeJ, DBA/1J, FVB/NJ,
al., 2001). The genomic instability detected by this NOD/LtJ NON/LtJ, NOR/LtJ, NZB/J, NZW/J,
technique is independent of the replication error and SWR/J), Mus musculus musculus (PWK/Pas and
phenomenon (MSI) because only inter-microsatel- MAI/Pas), Mus musculus castaneus (CAST/Ei), and
lite DNA segments are analyzed. Mus spretus (SPRET/Ei).
Although several mouse models are currently For the tumor/normal pair comparisons, DNA
used for the study of tumor development and the was extracted from fresh or snap-frozen tumor and
genomic instability associated with tumor progres- normal (tail or brain) tissues with the phenol/chlo-
sion, the use of inter-SSR PCR in mouse models roform/isoamyl alcohol method and purified by
has never been described. The aims of this study means of the DNeasy Tissue System kit (Qiagen,
were (1) to apply inter-SSR PCR to the mouse and Chatsworth, CA). Thymic lymphomas arose spon-
evaluate the reproducibility as well as the inter- taneously in the Msh2 ⫺/⫺ mice (BALBcJ;129P2
and intrastrain variability of the fingerprints and (2) mixed genetic background) between 2 and 8
to evaluate the use of inter-SSR fingerprints for the months of age (de Wind et al., 1995; Reitmair et al.,
analysis of genomic instability in spontaneous and 1995; Andrew et al., 2000; Campbell et al., 2000).
induced mouse tumors. Inter-SSR PCR has been Mice were euthanized when moribund, and the
successfully adapted in our laboratory for the anal- thymic origin of the tumors was established by
ysis of spontaneous and chemically induced mouse necropsy and histologic examination. Msh2-defi-
tumors by use of several single microsatellite-an- cient mice were kindly provided by Drs. A. Reit-
chored primers. We established the best PCR con- mair and T. Mak (AMGEN, Toronto, ON, Canada).
ditions for each primer and critically evaluated the N-Methyl-N-nitrosourea (MNU)–induced thymic
reproducibility of the band patterns and the ab- lymphomas were obtained from 4- to 6-week-old
sence of spurious bands. Tumor-specific alterations AKR/J ⫻ (AKR⫻C57L)F1 backcross mice and B6;
were detected as gains, losses, or intensity changes 129S-Trp53 ⫺/⫺ mice after treatment with 75
in bands when compared with matched normal mg/kg MNU, as previously described (Richie et al.,
DNA (Stoler et al., 1999). As proposed by Kahlen- 1991; Angel et al., 1996). For the thymic lympho-
berg et al. (1996), we quantitated the extent of mas, tissues used for DNA extraction were not
alterations by dividing the number of altered bands microdissected.
in the tumor by the total number of bands in SENCAR mice (outbred population) were ob-
normal DNA, obtaining an “instability index.” Ex- tained from the University of Texas M.D. Ander-
amples of genomic alterations analyzed by inter- son Cancer Center, Science Park–Veterinary Divi-
SSR PCR of a variety of mouse tumors are pre- sion (Bastrop, TX). Six-week-old mice were shaved
sented, including laser-captured microdissection 1–2 days before initiation with 10 nmol DMBA and
(LCM) tumors. These PCR-based multilocus fin- promoted 2 weeks later with twice-weekly appli-
gerprints provide a novel approach to assessing the cations of 0.5 g of 12-O-tetradecanoylphorbol-13-
number and onset of mutational events during tu- acetate (TPA) for 20 weeks. Tumors were fixed in
mor development in spontaneous and induced formalin and confirmed histologically with hema-
mouse tumors. toxylin and eosin (H&E) staining (Stern et al.,
1995). In this case, we obtained DNA from non-
MATERIALS AND METHODS microdissected as well as LCM tissues.
To validate the inter-SSR PCR protocol, we
Mice, Tumor Samples, and Genomic DNA used a cell line (PAM 27) established from a pri-
Extraction mary epidermal culture from newborn BALB/c
Mouse genomic DNA from 22 different mouse mice that underwent spontaneous malignant trans-
inbred strains and substrains was obtained from tail formation in culture (Aldaz et al., 1986). This cell
clips [The University of Texas M.D. Anderson line produces squamous cell carcinomas (SCCs)
Cancer Center (Smithville, TX) and Institut Pas- when injected into nude mice and is known to
teur (Paris, France)] by use of standard methods of present chromosome numbers in the hypotet-
proteinase K digestion and phenol/chloroform/ raploid range, double minutes, and homogeneously
isoamyl alcohol extractions (Wolff and Gemmill, staining regions.
1997) or purchased (ready for PCR) from The Jack-
son Laboratory (Bar Harbor, ME) and Taconic LCM Samples
Farms (Germantown, NY). The mouse strains an- For the LCM samples, chemically induced pap-
alyzed were Mus musculus domesticus (129/J, AKR/J, illomas (two-stage carcinogenesis protocol) were
A/J, BALB/cJ, C57BL/6J, C57BL/10J, C57L/J, collected and fixed in 100% methanol (to avoid
302 BENAVIDES ET AL.

DNA fragmentation) as previously described PCR products were analyzed by electrophoresis

(Noguchi et al., 1997; Anami et al., 2000). Briefly, through non-denaturing 8 –10% polyacrylamide
6-␮m sections of methanol-fixed, paraffin-embed- gels using Sequi-Gen Sequencing Cell (BioRad)
ded tissue were cut and mounted onto plain glass and Criterion precast 10% polyacrylamide gels
slides. Paraffin tissue was deparaffinized by incu- (BioRad) and visualized by silver staining (Amer-
bation of the slides twice in xylene for 5 min and sham-Pharmacia Biotech, Piscataway, NJ) and with
rehydration in absolute ethanol, 95% ethanol, 70% Vistra Green staining (Amersham International,
ethanol, and distilled water for 30 sec each. All the Buckinghamshire, UK). Gels were scanned or cap-
slides were stained with H&E and examined under tured with a digital camera. Intensity changes were
a light microscope. Stained sections were microdis- scored by means of fluor imaging with Vistra Green
sected using a PixCell II LCM system (Arcturus with a FluorImager apparatus (Molecular Dynam-
Engineering, Mountain View, CA). Tumor cells ics, Sunnyvale, CA) and IQ Mac 1.2 software (Bio-
and adjacent non-tumor stromal cells were visual- Image, Ann Arbor, MI). A change was considered
ized under the microscope and selectively procured significant when at least a twofold increase or de-
by activation of the laser. Approximately 5 ⫻ 103 crease in band intensity was observed in the tumor
tumor and non-tumor stromal cells were dissected. as compared to normal tissue DNA. For the anal-
Dissected cells were collected into 50 ␮l of cell ysis of variability within inbred strains, the differ-
lysis buffer containing 10 mM Tris–HCl, 1 mM ence value (D) of the fingerprints was calculated as
EDTA, 1% Tween-20, and 0.04% proteinase K the number of fragments that differed between two
(Sigma Chemical, St. Louis, MO) and incubated strains divided by the total number of fragments
for 72 hr at 55°C and 24 hr with Instagene Matrix
(BioRad, Hercules, CA). The proteinase K was
inactivated by heating at 95°C for 10 min. Samples
were purified by means of the DNeasy Tissue
System kit (Qiagen) before PCR.

Inter-SSR PCR
Genomic DNA from various inbred strains and
from tumors and matched normal tissue was ana-
lyzed by means of inter-SSR PCR by use of
(CA)8RY, (CA)8RG, (AAC)6Y, and (AGC)4Y prim-
ers (R ⫽ purine; Y ⫽ pyrimidine). Thus, in the case
of (CA)8RY, we used an equimolar mixture of four
different primers. Amplifications were carried out
in 25-␮l reactions containing 50 –100 ng of genomic
DNA, 2.5 ␮l 10⫻ PCR buffer (15 or 20 mM
MgCl2), 2% formamide, 0.2 mM dNTP, 1 ␮M
primer, and 1 U Taq polymerase. Testing PCR
conditions were as follows: initial denaturation 3
min at 94°C, 30 cycles of 30 sec at 94°C/45 sec at
52– 60°C/2 min at 72°C, and final extension 7 min
at 72°C. We tested two thermal cyclers: a Perkin–
Elmer Model 9600 DNA thermal cycler (Perkin
Elmer Cetus Instruments, Norwalk, CT) and a
RoboCycler 96 Temperature Cycler (Stratagene,
La Jolla, CA), and two DNA Taq polymerases:
AmpliTaq (PE Applied Biosystems, Foster City,
CA) and FastStart Taq (Roche, Indianapolis, IN).
Both machines and enzymes yield similar but not
identical band patterns under the same conditions.
Because of a better reproducibility, we decided to
Figure 1. Reproducibility analysis for the inter-SSR PCR technique
continue the assessment of genetic alterations with relative to DNA final concentration and source of DNA (tissues). A:
the RoboCycler 96 Temperature Cycler (at 54°C of Different final DNA concentrations (duplicated) from the same sample
amplified with primer (CA)8RY. B: Different tissues (duplicated) from
annealing temperature) and FastStart Taq polymer- the same mouse amplified with primer (AGC)4Y. See Materials and
ase (hot-start PCR). Methods for PCR and electrophoresis conditions.
 INTER-SSR PCR IN MOUSE MODELS 303

Figure 2. Reproducibility studies for the inter-SSR PCR

technique between individual mice from the same inbred strain.
Fingerprints of nine individual AKR mice (The Jackson Labora-
tory) amplified with primers (CA)8RG (upper left), (CA)8RY
(upper right), (AAC)6Y (lower left), and (AGC)4Y (lower right).
Even though fingerprints are identical, changes in intensity of
some bands (F) could be a concern for the genomic instability
analysis. See Materials and Methods for PCR and electrophore-
sis conditions.

present in the two strains. The average percentage scattered PCR products is what allows the wide
difference (APD) is simply the average of all D scanning of the genome. Each individual primer
values for the group of strains (Gilbert et al., 1990). generates a unique set of PCR products.
Assays were repeated six times for each tumor/ DNA fingerprints obtained from mouse DNA
normal pair to ensure reproducibility. Tumor-spe- with each individual primer typically exhibited a
cific alterations were detected as gains, losses, or set of 15–25 easily distinguishable bands (from 200
intensity changes in bands, and only those alter- to 1,000 bp) that can be used for the analysis. After
ations that recurred in at least four of the six assays initial evaluation with different final template
were counted in the estimation of the instability DNA concentrations, PCR was carried out with
index. We quantitated the amount of alterations by 50 –100 ng of total DNA. Figure 1A shows that no
dividing the number of altered bands in the tumor changes in the band patterns were detected in the
by the total number of bands in matched normal wide range of 100 ng to 2 ␮g of template DNA with
DNA, generating a genomic instability index
primer (CA)8RY. Moreover, we found no changes
(Kahlenberg et al., 1996; Basik et al., 1997; Stoler et
in the band patterns between tissues from the same
al., 1999).
individual, ruling out the interference of epigenetic
RESULTS factors, such as DNA methylation (Fig. 1B).
Figure 2 shows fingerprints originated from nine
Inter-SSR DNA Fingerprinting Tune-Up individual AKR mice from The Jackson Laboratory
Inter-SSR PCR consists of the amplification of amplified with primers (CA)8RY, (CA)8RG, (AAC)6Y,
DNA segments between simple sequence repeats and (AGC)4Y. The fingerprints were identical, al-
[such as (CA)n] by use of 3⬘-anchored primers. though changes in intensity of some bands among
Adjacent repeat sequences (less than 2 kb apart different individuals could be a concern for the
and arranged tail to tail) are required for the suc- genomic instability analysis. Therefore, tumor DNA
cessful PCR amplification that yields a large num- was always compared to normal DNA from the same
ber of products. The availability of these random animal to assess somatic changes.
304 BENAVIDES ET AL.

Figure 3. Interassay reproducibility studies for the inter-SSR PCR

and validation of the technique for the detection of somatic changes. A:
The proportion of irreproducible bands (including band intensity) in six
separate experiments [primer (CA)8RG] between MNU-induced lym- Figure 4. Inter-SSR fingerprints obtained from tumor cells (T) and
phoma and normal tissue was below 1%. B: Duplicated inter-SSR PCR adjacent non-tumor (N) stromal cells (approximately 5 ⫻ 103 cells)
fingerprints obtained from the PAM 27 mouse skin cell line (BALB/c) from LCM papilloma with primer (CA)8RG. Two-stage carcinogenesis
and normal DNA pairs with primers (AGC)4Y (left) and (CA)8RG (right). papillomas were collected and fixed in 100% methanol. See Materials
PAM 27 cell line showed four altered bands (F) with each of the and Methods for PCR and electrophoresis conditions.
primers when compared with BALB/c normal DNA. See Materials and
Methods for PCR and electrophoresis conditions. N, non-tumor; T,
tumor.
LCM samples were successfully amplified when
fixed in 100% methanol, as previously described for
the AP-PCR procedure (Anami et al., 2000). Figure
Interassay reproducibility studies were con-
4 shows fingerprints obtained from tumor cells and
ducted in six independent PCR reactions by use of adjacent non-tumor stromal cells (approximately
the same DNA template (MNU-induced lym- 5 ⫻ 103 cells were dissected from a papilloma) with
phoma and normal tissue) with primer (CA)8RG. primer (CA)8RG. LCM samples obtained from for-
As shown in Figure 3A, the proportion of irrepro- malin-, ethanol-, and acetone-fixed tissues failed to
ducible bands (including band intensity) in the six amplify by inter-SSR PCR (not shown).
separate experiments, under the same conditions,
was very low (less than 1%). Inter-SSR DNA Fingerprinting of Mouse Strains
The spontaneously transformed PAM 27 mouse Because unique DNA fragments could be useful
skin cell line (established from BALB/c mice), strain-specific molecular markers, we analyzed dif-
used to validate our protocol for the detection of ferences in the fingerprints obtained by inter-SSR
somatic changes in tumor samples, showed four PCR within various inbred strains. Polymorphism
altered bands with each primer when compared analysis using 22 inbred strains and substrains, am-
with BALB/c normal DNA controls (Fig. 3B). In- plified with primers (CA)8RY and (CA)8RG, re-
stability indices were 18% for the (CA)8RG primer vealed low levels of polymorphism within classical
and 16% for the (AGC)4Y primer. strains and almost identical band patterns for many
 INTER-SSR PCR IN MOUSE MODELS 305

Figure 5. Inter-SSR PCR analysis [primer

(CA)8RY] of various inbred strains revealed low levels
of polymorphism within classical strains, whereas wild-
derived inbred strains showed higher levels of poly-
morphism. The intra-group APD value for the group
of classical strains was 15.9 and 44.7% for the wild-
derived inbred strains. See Materials and Methods for
PCR and electrophoresis conditions. APD, average
percentage difference.

groups of substrains. The intra-group APD ranged strains generated by the inter-SSR PCR limits its
from 3.1% for the C57BL substrains to 15.9% for application to mouse strain identification (genetic
the group of classical strains. As expected, variabil- monitoring) and genetic mapping.
ity between wild-derived inbred strains (M. m. mus-
culus, M. m. castaneous, and M. spretus origin) and Inter-SSR DNA Fingerprinting of Mouse Tumors
classical strains was higher (Bonhomme and A comparison of PCR products from tumor/nor-
Guénet, 1996), as indicated by the D value of mal DNA pairs (examined in six separate, parallel
38.2% between BALB/c and SPRET/Ei strains. assays) from mouse models reveals a reproducible
The intra-group APD for the four wild-derived pattern of bands in which most products are iden-
inbred strains was 44.7%, with the highest D value tical, although tumor-specific rearrangements are
(60%) between CAST/Ei and SPRET/Ei strains also present. We primarily observed intensity
(Fig. 5). changes, but also a small number of band shifts,
Even though some of the classical inbred mice new bands, and band losses. These band alter-
have common ancestral origins (Atchley and Fitch, ations need not be limited to insertions or deletions
1991; Beck et al., 2000), many could be discrimi- between primer binding sites, nor to the deletion
nated by unique DNA patterns by use of classical of binding sites itself. Other large-scale processes
DNA fingerprinting, AP-PCR, and other repeat- such as whole-chromosome losses or large deletions
sequence PCR techniques (Jeffreys et al., 1987; can eliminate one or more products from the fin-
Welsh et al., 1991; Wood and Hamm, 1995). The gerprint. Also, gene amplifications can increase the
low level of polymorphism among the classical intensity of a band, and partial losses might pro-
306 BENAVIDES ET AL.

Figure 6. Inter-SSR PCR analysis of MNU-induced thymic lymphomas obtained from AKR ⫻
(AKR⫻C57L)F1 mice. Normal (N) and tumor (T) pairs were examined with primer (CA)8RG and showed
no evidence of genomic instability by means of inter-SSR PCR.

duce a reduction in the signal intensity. These

altered PCR products are valuable, in that they can
reveal where and when genetic instability events
have taken place during the carcinogenesis process
(Stoler et al., 1999; Anderson et al., 2001).
An example of MNU-induced thymic lymphoma
fingerprint analysis with primer (CA)8RG is shown
in Figure 6. No evidence of genomic instability was
found after the analysis of 25 lymphomas/normal
pairs with primers (CA)8RY and (CA)8RG. How-
ever, all the MNU-induced thymic lymphomas
(n ⫽ 5) generated in Trp53 nullizygote mice re-
vealed at least one altered band with each individ-
ual primer, with genomic instability indices be-
tween 4 and 9% (Fig. 7). In this case, inter-SSR
PCR proved to be useful in identifying nonrandom
changes (altered bands that reappear in different
tumors) that occur in response to the exogenous
carcinogen. Three of the 12 spontaneous thymic
lymphomas from Msh2 ⫺/⫺ mice analyzed with
primers (CA)8RY, (CA)8RG, and (AGC)4Y (as tu-
mor/normal DNA pairs) showed evidence of
genomic instability as measured by inter-SSR
PCR, with instability indices between 4 and 8.3%
(Fig. 8).
Papillomas and SCCs generated in SENCAR Figure 7. Inter-SSR PCR analysis of MNU-induced thymic lympho-
mice by the two-stage carcinogenesis protocol were mas in p53 nullizygote mice (B6;129S-Trp53tm1Tyj). Duplicate normal (N)
and tumor (T) pairs examined with primers (CA)8RG are shown.
analyzed by means of inter-SSR PCR. PCR prod- Genomic instability was detected, and instability indices (as proposed by
ucts from 12 papilloma/normal DNA and 12 SCC/ Anderson and co-workers) were calculated (shown under duplicate
tumor samples). Dots (F) placed alongside lanes indicate bands with
normal DNA pairs were examined with primers twofold increase or decrease intensity. These alterations are likely to
(CA)8RY, (CA)8RG, and (AGC)4Y for reproducible result from genomic lesions.
 INTER-SSR PCR IN MOUSE MODELS 307

Figure 8. Inter-SSR PCR analysis of spontaneous thymic lymphoma from Msh2 ⫺/⫺ mice. Duplicate
normal (N) and tumor (T) pairs examined with primers (CA)8RG are shown. Genomic instability (%) is
shown under duplicate tumor samples. Dots (F) placed alongside lanes indicate bands with twofold
increases or decreases in intensity.

alterations in tumor-derived bands (Fig. 9). No template (two orders of magnitude). It also has an
evidence of tumor-specific bands was found, as advantage over AP-PCR because of the high strin-
measured by inter-SSR PCR. In summary, in the gency conditions used in the PCR amplification.
variety of tumors analyzed, instability index values These stringent PCR conditions avoid the prob-
ranged between 0 and 9%, with the highest levels lems of primer competition that are probably re-
observed in MNU-induced thymic lymphomas sponsible for the appearance of spurious bands
generated in Trp53 nullizygote (⫺/⫺) mice. observed in the AP-PCR procedure. However, be-
cause reproducibility of the DNA patterns is also a
DISCUSSION concern for inter-SSR PCR, repeated (parallel) am-
We conclude that inter-SSR PCR is a suitable plifications and the use of tumoral and normal
technique for the assessment of the degree of DNA from the same animal are highly recom-
genomic damage in mouse models by unbiased mended.
DNA fingerprinting. We have shown using inter- Cloning and sequencing of altered bands open
SSR PCR that 25% of thymic lymphomas from the possibility to characterize and verify the inter-
mismatch repair-deficient mice, 100% of chemi- SSR PCR products and trace the origin of the
cally induced lymphomas from Trp53-deficient sequence in the GenBank. It is also possible to
mice, and 0% of chemically induced lymphomas in estimate the total number of genomic events in the
wild-type mice demonstrate genomic instability. tumor cells by calculating how much of the entire
Therefore, the differences in the amount of genome is being sampled in each fingerprint and
genomic instability in spontaneous and induced how many events have occurred within the sample
tumors derived on different genetic backgrounds (Stoler et al., 1999).
suggest that this assay can be useful for investigat- By means of inter-SSR PCR, we successfully
ing the underlying mechanistic differences in dis- analyzed genomic alterations in various mouse tu-
tinct tumorigenic pathways. In comparison with mors, including spontaneous thymic lymphomas
classic DNA fingerprinting techniques, inter-SSR developed in Msh2 ⫺/⫺ mice, MNU-induced thy-
PCR is much simpler and requires less DNA as mic lymphomas, and papillomas and SCCs ob-
308 BENAVIDES ET AL.

Figure 9. Inter-SSR PCR analysis of papilloma (P) and normal (N) pairs; SCC (S) and normal pairs; and
normal, papillomas, and SCC trios (from the same animal) with primer (CA)8RG. No evidence of genomic
instability was found by means of inter-SSR PCR.

tained after a two-stage carcinogenesis protocol. 49 chemically induced thymic lymphomas, papillo-
Our preliminary results in spontaneous and chem- mas, and SCCs analyzed. On the other hand, 3 out
ically induced mouse tumors show a variable de- of 12 (25%) of the spontaneous thymic lymphomas
gree of genetic alterations, from stability (MNU- from Msh2 ⫺/⫺ mice and 100% (n ⫽ 5) of the
induced thymic lymphomas and two-stage MNU-induced thymic lymphomas in Trp53-defi-
papillomas and SCCs) to moderate instability cient mice showed genomic instability. Although
(MNU-induced thymic lymphomas in p53-defi- the number of cases analyzed is low, the last find-
cient mice and spontaneous thymic lymphomas ing supports the idea of the genome instability
from Msh2 ⫺/⫺ mice), as measured by inter-SSR associated with Trp53 disruption. The loss or mu-
PCR. tation of this tumor-suppressor gene has been
In previous studies with human cancer speci- shown to correlate with CIN (Tarapore and Fuka-
mens, more than 80% (n ⫽ 59) of the sporadic sawa, 2000).
colorectal carcinomas evaluated by inter-SSR PCR
Because chromosomal alterations in Msh2 ⫺/⫺
showed at least one altered band. The median
mice have not been thoroughly studied to date, this
genomic instability index detected was 3.3% (that
technique can be used to analyze further the pres-
is, 3.3% of all bands generated for normal colonic
ence of genomic instability in those tumors. Our
mucosa DNA were altered in tumor DNA), and the
preliminary findings provide support for the pres-
most highly unstable colorectal tumor from this
study showed an index of 13%. Based on these ence of genomic instability in the absence of
results, Stoler and colleagues (1999) reported that MMR. Overall, our results suggest the existence of
typical sporadic colorectal cancers contain at least different forms of genomic instability driving tu-
11,000 genomic alterations per cell and that this is morigenesis in mouse models and human neopla-
an early event in tumor development. Perucho and sias and the possible involvement of carcinogen-
colleagues reported similar results with human specific genetic alterations. To address these
colorectal cancer by means of AP-PCR, the so- questions, a comprehensive analysis of genomic
called amplotype (Malkhosyan et al., 1998). instability by means of inter-SSR PCR (including
Our preliminary results through the use of inter- the determination of the genomic origin of altered
SSR PCR show low percentages of mouse tumors bands), by use of a variety of mouse tumors and
with genomic instability. The most interesting tumor cell lines, is now in progress in our labora-
finding is the absence of genetic instability in the tory.
 INTER-SSR PCR IN MOUSE MODELS 309
ACKNOWLEDGMENTS mutator genes in mismatch repair-deficient thymic lymphomas:
no evidence of mutations in the DNA polymerase delta gene.
We thank Dr. Joe Angel, Dr. Marcelo Rodriguez- Carcinogenesis 21:2281–2285.
Dasika GK, Lin SC, Zhao S, Sung P, Tomkinson A, Lee EY. 1999.
Puebla, and Dr. Marcela Franco for their help in DNA damage-induced cell cycle checkpoints and DNA strand
providing mouse tumor tissues. We also thank Dr. break repair in development and tumorigenesis. Oncogene 18:
7883–7899.
Marcelo Aldaz for providing the PAM 27 cell line. de Jong D, Voetdijk BM, Kluin-Nelemans JC, van Ommen GJ,
We are indebted to Melissa Bracher for secretarial Kluin PM. 1988. Somatic changes in B-lymphoproliferative disor-
ders (B-LPD) detected by DNA-fingerprinting. Br J Cancer 58:
assistance and Sharon Stockman for editing the 773–775.
manuscript. This work was supported in part by DePinho RA. 2000. The age of cancer. Nature 408:248 –254.
de Wind N, Dekker M, Berns A, Radman M, te Riele H. 1995.
grants CA69146 and CA90922 from the National Inactivation of the mouse Msh2 gene results in mismatch repair
Institutes of Health (to C.J.C.). deficiency, methylation tolerance, hyperrecombination, and pre-
disposition to cancer. Cell 82:321–330.
Forozan F, Karhu R, Kononen J, Kallioniemi A, Kallioniemi OP.
REFERENCES 1997. Genome screening by comparative genomic hybridization.
Trends Genet 13:405– 409.
Aldaz CM, Conti CJ, O’Connell J, Yuspa SH, Klein-Szanto AJ, Slaga Gilbert D, Lehman N, O’Brien S, Wayne R. 1990. Genetic finger-
TJ. 1986. Cytogenetic evidence for gene amplification in mouse printing reflects population differentiation in the California Chan-
skin carcinogenesis. Cancer Res 46:3565–3568. nel Island fox. Nature 344:764 –766.
Anami Y, Takeuchi T, Mase K, Yasuda J, Hirohashi S, Perucho M, Hayward N, Chen P, Nancarrow D, Kearsley J, Smith P, Kidson C,
Noguchi M. 2000. Amplotyping of microdissected, methanol- Ellem K. 1990. Detection of somatic mutations in tumors of
fixed lung carcinoma by arbitrarily primed polymerase chain re- diverse types by DNA fingerprinting with M13 phage DNA. Int J
action. Int J Cancer 89:19 –25. Cancer 45:687– 690.
Anderson GR, Brenner BM, Swede H, Chen N, Henry WM, Conroy Hermsen MA, Meijer GA, Baak JP, Joenje H, Walboomers JJ. 1996.
JM, Karpenko MJ, Issa JP, Bartos JD, Brunelle JK, Jahreis GP, Comparative genomic hybridization: a new tool in cancer pathol-
Kahlenberg MS, Basik M, Sait S, Rodriguez-Bigas MA, Nowak ogy. Hum Pathol 27:342–349.
NJ, Petrelli NJ, Shows TB, Stoler DL. 2001. Intrachromosomal Ionov Y, Peinado M, Malkhosyan S, Shibata D, Perucho M. 1993.
genomic instability in human sporadic colorectal cancer measured Ubiquitous somatic mutations in simple repeated sequences re-
by genome-wide allelotyping and inter-(simple sequence repeat) veal a new mechanism for colonic carcinogenesis. Nature 363:
PCR. Cancer Res 61:8274 – 8283. 558 –561.
Andrew SE, Xu XS, Baross-Francis A, Narayanan L, Milhausen K, Jeffreys AJ, Wilson V, Thein SL. 1985. Individual-specific “finger-
Liskay RM, Jirik FR, Glazer PM. 2000. Mutagenesis in PMS2- prints” of human DNA. Nature 316:76 –79.
and MSH2-deficient mice indicates differential protection from Jeffreys AJ, Wilson V, Kelly R, Taylor BA, Bulfield G. 1987. Mouse
transversions and frameshifts. Carcinogenesis 21:1291–1295. DNA “fingerprints”: analysis of chromosome localization and
Angel JM, Morizot DC, Richie ER. 1996. Localization of a novel germ-line stability of hypervariable loci in recombinant inbred
chromosome 7 locus that suppresses development of N-methyl- strains. Nucleic Acids Res 15:2823–2836.
N-nitrosourea-induced murine thymic lymphomas. Mol Carcinog Kahlenberg MS, Stoler DL, Basik M, Petrelli NJ, Rodriguez-Bigas
7:151–156. M, Anderson GR. 1996. p53 tumor suppressor gene status and the
Armour JA, Patel I, Thein SL, Fey MF, Jeffreys AJ. 1989. Analysis degree of genomic instability in sporadic colorectal cancers. J Natl
of somatic mutations at human minisatellite loci in tumors and Cancer Inst 88:1665–1670.
cell lines. Genomics 4:328 –334. Knuutila S, Bjorkqvist AM, Autio K, Tarkkanen M, Wolf M, Monni
Atchley WR, Fitch WM. 1991. Gene trees and the origins of inbred O, Szymanska J, Larramendy ML, Tapper J, Pere H, El-Rifai W,
strains of mice. Science 254:554 –558. Hemmer S, Wasenius VM, Vidgren V, Zhu Y. 1998. DNA copy
Atkin NB. 2001. Microsatellite instability. Cytogenet Cell Genet number amplifications in human neoplasms: review of compara-
92:177–181. tive genomic hybridization studies. Am J Pathol 152:1107–1123.
Bardelli A, Cahill DP, Lederer G, Speicher MR, Kinzler KW, Vo- Ledwith BJ, Storer RD, Prahalada S, Manam S, Leander KR, van
gelstein B, Lengauer C. 2001. Carcinogen-specific induction of Zwieten MJ, Nichols WW, Bradley MO. 1990. DNA fingerprint-
genetic instability. Proc Natl Acad Sci USA 98:5770 –5775. ing of 7,12-dimethylbenz[a]anthracene-induced and spontaneous
Basik M, Stoler DL, Kontzoglou KC, Rodriguez-Bigas MA, Petrelli CD-1 mouse liver tumors. Cancer Res 50:5245–5249.
NJ, Anderson GR. 1997. Genomic instability in sporadic colorectal Lengauer C, Kinzler KW, Vogelstein B. 1997. Genetic instability in
cancer quantitated by inter–simple sequence repeat PCR analysis. colorectal cancers. Nature 386:623– 627.
Genes Chromosomes Cancer 18:19 –29. Loeb K, Loeb L. 2000. Significance of multiple mutations in cancer.
Beck JA, Lloyd S, Hafezparast M, Lennon-Pierce M, Eppig JT, Carcinogenesis 2:379 –385.
Festing MF, Fisher EM. 2000. Genealogies of mouse inbred Loeb LA. 1991. Mutator phenotype may be required for multistage
strains. Nat Genet 24:23–25. carcinogenesis. Cancer Res 51:3075–3079.
Boland CR, Ricciardiello L. 1999. How many mutations does it take Loeb LA. 1996. Cancer surveys. In: Lindahl T, editor. Genetic
to make a tumor? Proc Natl Acad Sci USA 96:14675–14677. instability in cancer. Plainview, NY: Cold Spring Harbor Labora-
Boland CR, Thibodeau SN, Hamilton SR, Sidransky D, Eshleman tory Press. p 329 –342.
JR, Burt RW, Meltzer SJ, Rodriguez-Bigas MA, Fodde R, Ranzani Luo G, Santoro IM, McDaniel LD, Nishijima I, Mills M, Youssou-
GN, Srivastava S. 1998. A National Cancer Institute Workshop on fian H, Vogel H, Schultz RA, Bradley A. 2000. Cancer predispo-
Microsatellite Instability for cancer detection and familial predis- sition caused by elevated mitotic recombination in Bloom mice.
position: development of international criteria for the determina- Nat Genet 26:424 – 429.
tion of microsatellite instability in colorectal cancer. Cancer Res Malkhosyan S, Yasuda J, Soto JL, Sekiya T, Yokota J, Perucho M.
58:5248 –5257. 1998. Molecular karyotype (amplotype) of metastatic colorectal
Boltz EM, Harnett P, Leary J, Houghton R, Kefford RF, Fried- cancer by unbiased arbitrarily primed PCR DNA fingerprinting.
lander ML. 1990. Demonstration of somatic rearrangements and Proc Natl Acad Sci USA 95:10170 –10175.
genomic heterogeneity in human ovarian cancer by DNA finger- Navarro JM, Jorcano JL. 1999. The use of arbitrarily primed poly-
printing. Br J Cancer 62:23–27. merase chain reaction in cancer research. Electrophoresis 20:283–
Bonhomme F, Guénet J-L. 1996. The wild house mouse and its 290.
relatives. In: Lyon MF, Rastan S, Brown S, editors. Genetic Noguchi M, Furuya S, Takeuchi T, Hirohashi S. 1997. Modified
variants and strains of the laboratory mouse. Oxford, UK: Oxford formalin and methanol fixation methods for molecular biological
University Press. p 649 – 662. and morphological analysis. Pathol Int 47:685– 691.
Breivik J, Gaudernack G. 1999. Genomic instability, DNA methyl- Perucho M. 1996. Microsatellite instability: the mutator that mutates
ation, and natural selection in colorectal carcinogenesis. Semin the other mutator. Nat Med 2:630 – 631.
Cancer Biol 9:245–254. Reitmair AH, Schmits R, Ewel A, Bapat BV, Redston MS, Mitri A,
Cahill DP, Lengauer C, Yu J, Riggins GJ, Willson JK, Markowitz Waterhouse P, Mittrucker HW, Wakeham A, Liu B, Thomason A,
SD, Kinzler KW, Vogelstein B. 1998. Mutations of mitotic check- Griesser H, Gallinger S, Ballhausen WG, Fishel R, Mak TW.
point genes in human cancers. Nature 392:300 –303. 1995. Msh2 deficient mice are viable and susceptible to lymphoid
Campbell MR, Thang TY, Jirik FR, Andrew SE. 2000. Candidate tumors. Nat Genet 11:64 –70.
310 BENAVIDES ET AL.

Richards RI. 2001. Fragile and unstable chromosomes in cancer: Welsh J, McClelland M. 1990. Fingerprinting genomes using arbi-
causes and consequences. Trends Genet 17:339 –345. trary primers. Nucleic Acids Res 18:7213–7218.
Richie ER, McEntire BB, Coghlan L, Poenie M. 1991. Murine Welsh J, Petersen C, McClelland M. 1991. Polymorphisms gener-
T-lymphomas corresponding to the immature CD4 – 8⫹ thymo- ated by arbitrarily primed PCR in the mouse: application to strain
cyte subset. Dev Immunol 1:255–263. identification and genetic mapping. Nucleic Acids Res 19:303–
Shih IM, Zhou W, Goodman SN, Lengauer C, Kinzler KW, Vo- 306.
gelstein B. 2001. Evidence that genetic instability occurs at an Williams JG, Kubelik AR, Livak KJ, Rafalski JA, Tingey SV. 1990.
early stage of colorectal tumorigenesis. Cancer Res 61:818 – 822. DNA polymorphisms amplified by arbitrary primers are useful as
Stern MC, Gimenez-Conti IB, Conti CJ. 1995. Genetic susceptibil- genetic markers. Nucleic Acids Res 18:6531– 6535.
ity to papilloma progression in the SENCAR mice. Carcinogene- Wolff R, Gemmill R. 1997. Purifying and analyzing genomic DNA.
sis 16:1947–1953. In: Birren B, Green E, Klapholz S, Myers R, Roskams J (eds.)
Stoler DL, Chen N, Basik M, Kahlenberg MS, Rodriguez-Bigas MA, Genome analysis: a laboratory manual (Vol. 1). Plainview, NY:
Petrelli NJ, Anderson GR. 1999. The onset and extent of genomic Cold Spring Harbor Laboratory Press. p 1– 82.
instability in sporadic colorectal tumor progression. Proc Natl Wood PA, Hamm DA. 1995. Survey of genomic repeat sequence-
Acad Sci USA 96:15121–15126. PCRs that detect differences between inbred mouse strains.
Tarapore P, Fukasawa K. 2000. p53 mutation and mitotic infidelity. Genet Res 65:151–155.
Cancer Invest 18:148 –155. Yasuda J, Navarro M, Malkhosyan S, Velazquez A, Arribas R, Sekiya
Thein SL, Jeffreys AJ, Gooi HC, Cotter F, Flint J, O’Connor NT, T, Perucho M. 1996. Chromosomal assignment of human DNA
Weatherall DJ, Wainscoat JS. 1987. Detection of somatic changes fingerprint sequences by simultaneous hybridization to arbitrarily
in human cancer DNA by DNA fingerprint analysis. Br J Cancer primed PCR products from human/rodent monochromosome cell
55:353–356. hybrids. Genomics 34:1– 8.
Weiss MM, Hermsen MA, Meijer GA, van Grieken NC, Baak JP, Zietkiewicz E, Rafalski A, Labuda D. 1994. Genome fingerprinting
Kuipers EJ, van Diest PJ. 1999. Comparative genomic hybridisa- by simple sequence repeat (SSR)-anchored polymerase chain re-
tion. Mol Pathol 52:243–251. action amplification. Genomics 20:176 –183.