Oncogene (2001) 20, 6551 ± 6558

ã 2001 Nature Publishing Group All rights reserved 0950 ± 9232/01 $15.00
www.nature.com/onc

Mouse model for lung tumorigenesis through Cre/lox controlled sporadic

activation of the K-Ras oncogene

Ralph Meuwissen1,4, Sabine C Linn1,4, Martin van der Valk2, Wolter J Mooi3 and Anton Berns*,1
1
Division of Molecular Genetics and Center of Biomedical Genetics, The Netherlands Cancer Institute, Plesmanlaan 121, 1066
CX Amsterdam, The Netherlands; 2Department of Experimental Animal Pathology, The Netherlands Cancer Institute,
Plesmanlaan 121, 1066 CX Amsterdam, The Netherlands; 3Division of Diagnostic Oncology, The Netherlands Cancer Institute,
Plesmanlaan 121, 1066 CX Amsterdam, The Netherlands

The onset of human lung cancer occurs through
sequential mutations in oncogenes and tumor suppressor
genes. Mutations in K-Ras play a prominent role in
human non-small cell lung cancer. We have developed a
mouse lung tumor model in which K-Ras can be
sporadically activated through Cre-lox mediated somatic
recombination. Adenoviral mediated delivery of Cre
recombinase in lung epithelial cells gave rise to rapid
onset of tumorigenesis, yielding pulmonary adenocarci-
nomas with 100% incidence after a short latency. The
lung tumor lesions shared many features with human
non-small cell lung cancer. Our data show that sporadic
expression of the K-Ras oncogene is sucient to elicit
lung tumorigenesis. Therefore this model has many
advantages over conventional transgenic models used
thus far. Oncogene (2001) 20, 6551 ± 6558.
Keywords: somatic activation; Cre-lox; conditional K-
Ras allele; lung epithelial tissue; lung tumors; non-
small cell lung cancer
Introduction
Lung cancer is the leading cause of cancer-related
death in man (Wingo et al., 1999). Because of
important di€erences in clinical behavior, lung cancer
is commonly subdivided into SCLC [small cell lung
cancer] and NSCLC [non-small cell lung cancer], the
latter consisting of three main histologic types:
adenocarcinoma (25 ± 40%), squamous cell carcinoma
(25 ± 40%) and large cell carcinoma (10 ± 20%) (World
Health Organization, 1981; Edwards, 1987). Pulmon-
ary adenocarcinomas are known to harbor multiple
cytogenetic and molecular abnormalities (Rom et al.,
2000), in full accordance to a multi-step model for
tumorigenesis (Hanahan and Weinberg, 2000).
One dominant genetic alteration often found in
*30% of pulmonary adenocarcinomas is an activating
mutation of K-Ras (Mills et al., 1995; Rodenhuis et al.,
1988; Huncharek et al., 1999). Mouse models for
pulmonary adenocarcinomas have been generated by
introducing K-Ras mutations in lung epithelium by
means of chemical carcinogens or by the use of
transgenic mouse strains harboring an activated H-
Ras gene driven from a lung-epithelial tissue speci®c
promoter (Tuveson and Jacks, 1999).
However, these approaches have major limitations.
Carcinogen-induced lung tumorigenesis may a€ect
many other genes besides the K-Ras oncogene. In
most of the transgenic systems supraphysiological
levels of H-Ras and N-Ras rather than K-Ras were
used. Most importantly, the classical transgenic tumor
model does not mimic the sporadic occurrence of Ras
mutations. There is a fundamental di€erence between a
transgenic model in which an oncoprotein is expressed
in every cell of a tissue or cell lineage, and a sporadic
cancer model in which normal wild-type cells surround
the incipient tumor cell. These normal cells could have
a prominent suppressing e€ect on the proliferative
capacity of cancer-prone cells. Alternatively, but not
mutually exclusive, a single mutant cell might fail to
support its own proliferation through autocrine
mechanisms as factors di€use. This inhibition might
be overcome once a single mutant cell has given rise to
an aggregate of descendants forming a small patch of
cells.
Here, we have circumvented this problem by
establishing a tumorigenesis model using sporadic
activation of K-Ras in mouse pulmonary epithelium.
We have chosen the Cre-lox mediated somatic
recombination (Sauer, 1998) to introduce activated
K-Ras expression in lung epithelium of adult mice.
Using adenoviral based delivery of Cre recombinase
into lung epithelial cells, we achieved sporadic K-RasV12
activation, thereby mimicking the human condition of
NSCLC in which K-Ras is often mutated and is linked
to aggressive tumor behavior (Rodenhuis and Slebos,
1992; Slebos et al., 1990). The model shows close
similarity with a recently described `hit and run'
sporadic lung tumor model (Johnson et al., 2001).

*Correspondence: A Berns; E-mail: tberns@nki.nl
4
Both authors contributed equally to this work
Received 2 April 2001; revised 17 May 2001; accepted 16 July 2001
Results
Previous transgenic mouse models overexpressing
mutant H-Ras showed a stochastic onset of pulmonary
 Conditionally activated K-Ras induces murine lung tumors
R Meuwissen et al
6552
neoplasia (Saitoh et al., 1990; Maronpot et al., 1991;
Suda et al., 1987). We decided to use another approach
by generating transgenic animals harboring a condi-
tional K-RasV12 allele, which expresses K-RasV12 only
after Cre-mediated recombination. K-RasV12 is a potent
oncogene and similar K-Ras mutations have often been
found in pulmonary, pancreatic and colorectal adeno-
carcinomas (Bos, 1989; Mills et al., 1995; Pellegata et
al., 1994; Oudejans et al., 1991). K-Ras has an intrinsic
GTPase activity and the K-RasV12 mutation locks the
protein in its active, GTP-bound state, leading to
constitutive signaling to downstream targets (Manne et
al., 1985; McGrath et al., 1984; Shields et al., 2000).
The conditional transgenic K-RasV12 transgene
(Figure 1) used here consists of a broadly active b-
actin promoter, followed by a GFP [Green Fluores-
cence Protein] expression cassette ¯anked by two lox
sites and preceding a K-RasV12 cDNA combined with a
PLAP [human placenta-like alkaline phosphatase]
expression construct. Six founder lines were obtained
with this construct. Only one founder contained a
single copy, full-length transgene. Analysis of transgene
expression showed ubiquitous GFP expression in a
wide range of tissues (results not shown). A poly-
adenylation signal behind the GFP cassette prevented
read-through into the K-RasV12 gene. Therefore,
expression of K-RasV12 was dependent on Cre-lox
mediated deletion of the GFP fragment.
Since NSCLC, including pulmonary adenocarcino-
ma, arises from bronchopulmonary epithelium (Tsu-
chiya et al., 1995; Cooper et al., 1997), we targeted
pulmonary epithelial cells by administering recombi-
nant AdCre [adenoCrevirus] of sucient high titer (up
to 46109 pfu/ml) intratracheally. In order to suppress
perioperative infections, mice were treated with co-
trimoxazole 1 week before and after AdCre adminis-
tration. Perioperative mortality was less than 5%. Our
experiments using adenoLacZ virus showed b-galacto-
sidase activity in both bronchial epithelial cells as well
as in the more distal alveolar pneumocytes 48 h after
infection (data not shown). However, as has been
reported by others (Zsengeller et al., 1995; Otake et al.,
1998), we found an extensive clearance of infected cells
within 2 ± 4 weeks after infection, probably due to an
immune response against the infected cells. To suppress
clearance of the cells (Howell et al., 1998), mice were
treated with cyclosporin A for 4 ± 6 weeks following
adenovirus administration. The e€ective cyclosporin
concentration remained roughly constant at about
40 ng/ml blood plasma and led to a marked decrease
of infected cell clearance. To further validate the
model, we decided to use AdCre for intratracheal
infection of ACZL reporter mice (Akagi et al., 1997).
In these mice, a Lac Z reporter gene is transcribed after
Cre-mediated recombination and the recombined Lac
Z reporter locus can be detected by staining for the
resulting b-galactosidase activity. Immunostaining of
lung tissue sections obtained from ACZL mice 2 weeks
after AdCre showed the presence of Cre recombinase
mainly in bronchial epithelial cells but also in the more
distal alveolar epithelium (Figure 2a). Only sporadic
staining of cells was observed. Histological staining for
b-galactosidase activity in these lung tissue sections
con®rmed these results. Cre-mediated recombination
activity is only present in sporadicly infected bronchial
epithelial cells as well as alveolar cells (Figure 2b). No
Figure 2 Validation of the model. ACZL reporter mice had
been injected 107 pfu AdCre virus intratracheally. Two weeks
after administration mice had been sacri®ced. (a) Immunohisto-
chemical staining shows that Cre recombinase is expressed in
alveolar pneumocyte nuclei. (b) Staining for Lac Z. b-gal -positive
bronchial epithelial cells as well as alveolar type II pneumocytes
are observed

Figure 1 Schematic representation of the conditional K-Ras construct. Without Cre recombinase GFP mRNA is expressed and
the K-RASVal 12 oncogene behind the poly-A signal remains silent. Upon Cre mediated recombination of the loxP sites (") the K-
RASVal 12 oncogene is placed directly under control of the b-actin promoter. The K-RASVal 12 oncogene is transcribed together with
the PLAP cDNA and because of their linkage through the IRES [Internal Ribosomal Entry Site] this results in translation of two
independent proteins. Position and orientation of the PCR primers used for analysis are depicted (?)

Oncogene

recombination of the LacZ locus could be detected in
tissues other than the lung. The use of R26R reporter
mice (Soriano, 1999) gave similar results (data not
shown).
We subsequently administered AdCre to conditional
K-RasV12 mice. Table 1 summarizes the results. Three
weeks after infection we detected intrapulmonary
proliferative epithelial lesions with a reactive lympho-
cytic in®ltrate (Figure 3b). This intraparenchymal
rather than airway lesion strongly resembled the so-
called AAH [alveolar atypical hyperplasia] of the
human, a lesion that is thought to precede the
development of some pulmonary adenomas (Carey et
al., 1992). At 5 ± 6 weeks post-infection, all mice had
developed multiple lesions of this type (Figure 3a, c).
The lesions subsequently developed into larger,
papillary-like tumors (Figure 3d), seen at 8 weeks
post-infection. The last group of mice was analysed at
9 ± 13 weeks post-infection: all had multiple pulmonary
tumors causing cachexia and respiratory distress due to
mass e€ect and bronchial obstruction. The architecture
of these large nodules was again mainly papillary so
that the tumors resembled human papillary adenocar-
cinomas (Malkinson, 1998). The tumor cells were
positive for TTF-1 [thyroid transcription factor-1]
(Figure 3e), a speci®c immunohistochemical marker
of thyroid and pulmonary epithelium and tumors
thereof (Ordonez, 2000) and often used for the
identi®cation of pulmonary adenocarcinomas (Kauf-
mann and Dietel, 2000). The proliferative activity of
these solid nodules, as shown by BrdU staining (Figure
3f), was moderate but greater as compared to the
surrounding normal tissue. At 12 ± 13 weeks post-
infection, some mice showed sporadic pulmonary
tumors with marked nuclear enlargement, prominent
nucleoli and extensive mitotic activity. These histologic
indicators of tumor progression coincided with occa-
sional emergence of macroscopic metastases to lymph
nodes and kidney (Table 1). In normal control FVB/N
mice undergoing the same AdCre administration, no
pulmonary lesions were found.
The progression observed histologically, from AAH,
via small papillary tumors to full-blown adenocarcino-
mas, is reminiscent of that of some pulmonary
adenocarcinomas in man. As has been found for the
latter (Ten Have-Opbroek et al., 1997; Gazdar, 1994),
Table 1 Tumor development in conditional K-Rasv12 mice after AdCre administration
Time between p.f.u. of virus
No. of adenoCre administration particles
mice* and sacrifice (weeks) administered
6 2 107
5 3±4 107
5 5±6 46107
6 8 107746107
6 9 ± 13 106 ± 107
*Median age of mice at moment of AdenoCre administration is 9 weeks (range 8 ± 12 weeks)
Conditionally activated K-Ras induces murine lung tumors
R Meuwissen et al
6553
the lung tumors arising in the conditional K-RasV12
mice probably also arose from alveolar type II cells.
We investigated this by immunostaining for Sp-B and
C [surfactant apoproteins B and C], which are both
markers for alveolar type II cells (Bohinski et al., 1993;
Whitsett et al., 1995; Khoor et al., 1994). As shown in
Figure 3g, h, tumors were positive for both Sp-B and
Sp-C, attesting to the alveolar type II cell origin of the
tumors.
To demonstrate that the lung tumorigenesis requires
activation of the conditional K-RasV12 allele, we
performed a Northern analysis on total RNA from
AdCre treated, tumor prone conditional K-RasV12 mice
and controls. In all lung tumors we could detect a K-
RasV12 mRNA transcript of 5.2 kB, corresponding to
the K-RasV12-IRES-PLAP transcript (Figure 4a). No
transcript was detected in control normal lung tissue.
The inactive conditional K-RasV12 allele showed the
GFP containing transcript in all tissues tested (Figure
4b), even in the lung tumor samples, in accordance
with the presence of unrecombined wild-type, non-
neoplastic cells in the tumor mass. These results
indicate that lung tumors arose exclusively after AdCre
mediated activation of the K-RasV12 allele. Since we
already showed that AdCre infection occurred spor-
adically throughout the bronchopulmonary epithelium
(Figure 2), the tumors must have arisen from
sporadically switched cells. We then asked, whether
overexpression of K-RasV12 is sucient for the
formation of a patch of mutant cells or whether
additional mutations were required for their develop-
ment. In the latter situation the `®eld' of cells carrying
the activated K-RasV12 should extend outside the
con®nes of the microscopic lesions. This permitted us
to determine whether `®eld cancerization' (Sozzi et al.,
1995; Park et al., 2000; Garcia et al., 1999) does play a
role in this sporadic model. One might argue that in
regular transgenic models `®eld cancerization' is
inherent to the system, since most if not all cells are
predisposed by transgene expression or tumor suppres-
sor gene inactivation. This point has often been raised
as an argument against the suitability of transgenic
mice as a model for sporadic cancer. A `®eld' of pre-
neoplastic cells could obviously serve as a basis for
tumor progression, as any additional mutation might
e€ectively synergize with the lesion present. In
Histopathology lungs
(No. of mice with described pathology/No. of mice analysed)
Only reactive infiltrate (6/6)
Epithelial focal neoplasia (4/5)
Multifocal glandular neoplasia/adenomatous tumors (5/5)
Epithelial focal neoplasia (2/5)
Papillary neoplastic lesions (6/6)
Epithelial focal neoplasia (0/6)
Mixed multifocal papillary neoplasias and papillary carcinomas (6/6)
Lymph node metastases (3/6), kidney metastasis (1/6)
Oncogene
 Conditionally activated K-Ras induces murine lung tumors
R Meuwissen et al
6554
Figure 3 Histopathological analysis of lung tumor development
in activated conditional K-RasV12 transgenic mice. After intra-
tracheal AdCre administration, mice were sacri®ced at various
time points. (a) After 5 weeks, multiple foci of developing alveolar
adenomas were found throughout normal lung parenchyma
(H&E, power®eld 206). (b) Mild dysplastic lesions along the
alveolar lining were found after 3 weeks (upper and lower
arrowheads, respectively; H&E, power®eld 1006). No lesions
were found at earlier time-points (c) Alveolar dysplastic lesion
started to form papillary structures after 5 weeks (H&E,
power®eld 2006). (d) More progressive, papillary adenocarcino-
ma-like lesions were found at 8 weeks post-infection. (H&E,
power®eld 1006). (e). Immuno-staining of papillary adenocarci-
noma nodule for TTF-1 (9 weeks post-infection, power®eld
1006). (f) Positive staining for BrdU (black stain) in a large
adenoma nodule (6 weeks post-infection, power®eld 1006). (g)
Staining for surfactant B (power®eld 4006) and (h) surfactant C
(power®eld 1006) in cytoplasm of adenocarcinoma nodules (8
weeks post-infection)
addition, the `®eld' of mutant cells might also have
di€erent characteristics as it might create a micro-
Oncogene
Figure 4 Conditional K-RasV12 transgene expression in tumor
tissue. (a) Northern blot analysis of K-RasV12 expression on total
RNA from control FVB, various organs from untreated
conditional K-RasV12 transgenic mice and lung tumor tissues
from four independent mice treated with AdCre. Lung tumor
tissues were isolated 8 weeks after AdCre administration. A
5.2 kb mRNA corresponding to the predicted transgenic K-
RasV12 transcript was only detected in the lung tumor lanes. (b)
Identical to panel A, the same samples and amount of total RNA
was hybridized with a GFP probe. A 1.5 kb GFP mRNA was
detected in all tissues from the conditional K-RasV12 transgenic
mouse. The transgene expression level varies between each tissue
and even some GFP expression was found in the tumor RNA
samples. (c) Control hybridization with an actin probe shows the
relative amount of total RNA loaded in each lane
environment in which cell ± cell contacts and paracrine
e€ects promote cell proliferation or transformation, as
has also been observed in vitro (Land et al., 1986). We
performed laser dissection and PCR analysis on the
early dysplastic lesions or AAH, and on the histolo-
gically normal, surrounding tissue (Figure 5a). We
detected the 240 bp PCR fragment characteristic for
the mutant K-RasV12 allele in AAH lesions and late-
stage adenocarcinomas (Figure 5b), but not in
surrounding tissue. This surrounding tissue only
showed the 480 bp GFP fragment corresponding to
the unrecombined conditional allele (Figure 5a, b). No
GFP fragment was seen in AAH or adenocarcinoma
lesions, suggesting that all neoplastic cells harbored the
activated K-RasV12 alleles. These results indicate that
activating K-RasV12 itself is sucient for the formation
of the early lesions such as AAH, which then
subsequently develop into papillary adenocarcinomas.
Discussion
Conventional H-Ras transgenic lung tumor models
have previously underlined the importance of Ras
expression for the development of some lung tumors
(Saitoh et al., 1990; Maronpot et al., 1991; Tuveson
and Jacks, 1999). We report here that sporadic
expression of a K-RasV12 transgene is already enough
for the rapid onset of lung tumorigenesis. One might
 Conditionally activated K-Ras induces murine lung tumors
R Meuwissen et al
6555

Figure 5 Analysis of conditional K-RasV12 allele recombination in neoplastic and surrounding tissue. PCR analysis was performed
on genomic DNA derived from laser dissected neoplastic lesions or normal cells directly surrounding these lesions. (a) Laser
dissected cells from four independent AAH lesions and their respective Sur.Tis. [surrounding tissue] cells were used for PCR analysis
either of the Ras Rec. [recombined K-RasV12 allele] or GFP [unrecombined allele, still containing the GFP fragment]. A 240 bp
PCR product representing the recombined conditional K-RasV12 allele (primers AG2 and Ras-5-ANTI, depicted in Figure 1) was
only found in AAH cells, not in the surrounding cells. The unrecombined allele, however, represented by a 480 bp fragment
(primers GFP-5 and GFP-3) was only present in surrounding cells and not in AAH cells. (b) Similar experiment performed as in (a)
but now laser dissected cells from four independent Ad.Car. [adenocarcinomas] and cells from their respective surrounding normal
tissue were used. Identical to AAH cells, the 240 bp PCR product for a recombined allele, is only present in Ad.Car. cells and the
480 bp PCR product for the unrecombined allele can only be detected in surrounding tissue

argue that in this model the K-RasV12 transgene is
expressed at elevated levels and therefore the K-RasV12
activation or `switch' represents in fact two events that
usually will take place sequentially in normal tumor
development: mutation and upregulation or deletion of
the wt allele. This could explain the absence of `®eld
cancerization' as an intermediate step and the rapid
development of full-¯edged adenocarcinomas (9 ± 13
weeks post-infection) in our model. Also, the suscept-
ibility of FVB/N mice to adenocarcinoma formation
could play a role in the rapid development of these
tumors and serve itself as a mechanism similar to that
of ®eld cancerization (Shimkin and Stoner, 1975;
Mahler et al., 1996). However, a recent new model
for sporadic mutations in the endogenous K-Ras allele
shows lesions very reminiscent to ours (Johnson et al.,
2001). Therefore, it is more likely that sporadic
activation of K-Ras in alveolar type II cells in mice
is sucient to initiate this process. Interestingly, the
susceptibility is not restricted to any particular
developmental window as used in lung tumor induction
with carcinogens (Oomen et al., 1991; Fijneman et al.,
1995). Activation of the K-RasV12 allele in adult mice
leads to the synchronous development of multiple
tumor lesions that progress in time. It is likely that
additional mutations did occur in the transition to late-
stage adenocarcinomas. The nature of these acquired
mutations will be investigated in further studies.
Another interesting observation in our model is the
preferential occurrence of parenchymatous, pneumo-
cyte-type adenocarcinomas, and the absence of bron-
chial-type adenocarcinomas, although bronchial
epithelial cells were e€ectively infected with AdCre
(Figure 2). Probably the cellular context in which the
K-RasV12 expression arises is important. This suggests
that alveolar type II cells are more sensitive to K-Ras
mutations than broncho-epithelial cells, or are more
prone to undergo mutant ras-driven neoplastic trans-
formation. This is in line with the observation that no
K-Ras mutations were found in bronchial adenocarci-
nomas in another study (Cooper et al., 1997).
Similarly, K-Ras mutations are distinctly uncommon
in human pulmonary squamous cell carcinomas (Slebos
et al., 1990), and likewise, in our model we did not
encounter squamous cell carcinomas, thus further
enhancing the resemblance of the murine model to
the human situation. Although the lung tumors
described here resemble those found in the previous
H-Ras transgenic lung tumor models (Saitoh et al.,
1990; Maronpot et al., 1991; Tuveson and Jacks, 1999),
our model has important advantages over classical
transgenic or knockout models. The moment of tumor
initiation can be chosen, the frequency in which de®ned
lesions are induced can be altered, and multiple lesions
can be induced at the same time. This not only permits
establishment of clear genotype-phenotype relation-
ships, but also enables longitudinal monitoring of the
tumorigenic process. The model is also particularly
suited to test prevention and intervention strategies as
oncogenic activation events can be limited in number
Oncogene
 Conditionally activated K-Ras induces murine lung tumors
R Meuwissen et al
6556
and restricted to a small time interval, comparable to
that of carcinogen exposure, except that now a de®ned
lesion is produced with distinct oncogenic conse-
quences.
One word of caution should be added. Recently, it
has been shown that Cre activity can lead to
chromosomal aberrations in cells in vitro (Loonstra et
al., 2001) as well as in vivo (Schmidt et al., 2000). We
cannot exclude that temporal Cre expression has served
as a co-factor in the onset of tumorigenesis. However,
in the elegant study of a sporadic model for K-Ras
induced mutations, cited above, tumors very similar to
those described here were observed (Johnson et al.,
2001), arguing against a major autonomous role of Cre
in tumorigenesis in this setting. The Cre/lox-K-RasV12
model should be particularly suited to test intervention
protocols directed against the K-Ras pathway. In this
respect it will be worthwhile to re-evaluate the ecacy
of previously tested farnesyl-transferase inhibitors and
other drugs directed against downstream targets of K-
Ras (Omer et al., 2000; Gibbs, 2000) in this more
critical model. To further improve the utility of this
model for this particular application we are currently
combining this sporadic lung tumor model with
sensitive in vivo imaging using (Contag et al., 2000) a
¯oxed luciferase expression cassette. This will permit
non-invasive quantitative assessment of tumor growth,
metastasis, and regression.
Materials and methods
Generation of conditional transgenic K-RasV12 mice
A 1.2 kb human cDNA fragment encoding K-RasV12 was
placed under the control of the 1.8 kb chicken b-actin
promoter (Akagi et al., 1997). In between the promoter and
K-RasV12 a 1.3 kb GFP [Green Fluorescent Protein] cassette
(Clonetech) with a poly-A linker was placed, ¯anked by two
loxP sites. This GFP cassette served as a marker for b-actin
promoter activity. Behind the K-RasV12 a 2.9 kb PLAP
[human placenta-like alkaline phosphatase] was placed as a
reporter of K-RasV12 expression. The construct, ¯anked by
SalI sites, was puri®ed by agarose gel electrophoresis and
electroelution, and injected into fertilized FVB/N mouse
oocytes.
Screening of transgenic mice
Transgenic conditional K-RasV12 founder mice were identi®ed
by Southern blot analysis. Genomic DNA was extracted from
tails according to standard protocols, digested with EcoRV
and hybridized with a 500 bp GFP probe or 1 kb PLAP
probe. Genotyping of o€spring was performed by PCR on
tail tip DNA, that was ampli®ed with the primers RAS-5' (5'-
CAGTGCAATGAGGGCCAG-3') and RAS-3' (5'-CACCC-
TGTCTTGTCTTTGCTG-3'), yielding a 280 bp product.
Thermocycling conditions consisted of 30 cycles of 30 s at
948C, 30 s at 598C, and 50 s at 728C. The PCR mix consisted
of 16PCR bu€er (Gibco), 200 mM dNTPs, 2.5 mM MgCl2,
0.4 mM primers, 0.5 units of Taq polymerase and 1 ml of a
dilution of extracted tail DNA, in a 25 ml-volume. ACZL and
R26R reporter mice were screened by PCR using the same
Oncogene
conditions as described above, only the annealing tempera-
ture was changed into 608C. The primers used for
ampli®cation were LZ1 (5'-CGTCACACTACGTCTGAA-
CG-3') and LZ2 (5'-CGACCAGATGATCACACTCG-3'),
for both reporter mice.
Analysis of transgene transcription
Expression of the transgene was analysed by Northern
blotting. Total RNA was prepared from snap-frozen tissue
using Trizol isolation reagent (Gibco). Northern blotting was
performed according to standard protocols using 10 ± 20 mg
of total RNA. The probes used were a 480 bp GFP fragment
and a 280 bp K-Ras fragment.
PCR analysis of recombination
Microdissection of tumors and adjacent normal lung tissue
was performed using a laser capture microdissection system
(Arcturus). Ten-micrometer thick sections from formalin
®xed paran-embedded material were used after dewaxing
in xylene and drying. The dissected material was transferred
to 50 ml of proteinase K bu€er and incubated at 378C
overnight, after which the proteinase K was inactivated for
10 min at 968C. Five microliters were used for subsequent
PCR ampli®cation using primers AG2 (5'-CTGCTAACCA-
TGTTCATGCC-3') and Ras 5-anti (5'-CCTACGCCACAA-
GCTCCAACTAC-3') or GFP-5 (5'-GACCACATGAAGCA-
GCACGAC-3') and GFP-3 (5'-CGAACTCCAGCAGGAC-
CATG-3'). The recombined Ras allele yielded a 240 bp
product while the unrecombined, GFP containing, allele
yielded a 480 bp product. The PCR cycling pro®le consisted
of 35 cycles of 30 s at 948C, 30 s at 598C, and 50 s at 728C.
The PCR mix consisted of 16PCR bu€er (Gibco), 100 mM
dNTPs, 2.5 mM MgCl2, 0.5 mM primers, 2.5 units of Taq
polymerase in a 25 ml-volume.
Immunohistochemistry
BrdU [bromodeoxyuridine] 100 mg/g was administered in-
traperitoneally to selected mice, 2 ± 3 h before they were
sacri®ced. Mice were sacri®ced via CO2 inhalation. Para-
formaldehyde 4% in PBS was instilled through the trachea.
Lungs were excised, ®xed in formalin and embedded in
paran. Four mm thick slides were stained for BrdU
(monoclonal, 470 mg/ml; DAKO), Cre (polyclonal,
1 : 15 000; Novagen), pro-Sp-B (polyclonal, 1 : 1000; Chemi-
con International), pro-SP-C (polyclonal, 1 : 500; Chemicon
International) and TTF-1 (DAKO clone 8G7G3/1, 1 : 40 000)
Antigen retrieval with microwave heating in citric acid at
pH=6 was performed for BrdU, Cre and TTF-1. As
secondary antibodies biotinylated goat anti-mouse (1 : 600;
DAKO), biotinylated goat anti-rabbit (1 : 800; DAKO) or
horseradish peroxidase labeled goat anti-rabbit (1 : 50;
DAKO) were used. Visualization was achieved using biotin/
avidin-peroxidase (Vector) (BrdU, pro-Sp-B and pro-Sp-C)
and diaminobenzidine as chromogen.
Staining for b-galactosidase activity in ACZL reporter mice
was performed according to Akagi et al. (1997).
Intratracheal Adeno-Cre virus administration
Adeno-Cre virus was constructed and produced according to
Akagi et al. (1997). Two to 4 month old mice were treated
with 9.6 mg/day co-trimoxazole orally (480 mg co-trimox-
azole in 250 ml drinking water, presumed intake 5 ml/day), 1
week prior to virus administration. Co-trimoxazole was

administered to reduce perioperative mortality due to
infections. One day before AdCre administration mice
received 15 mg/kg cyclosporin A intraperitoneally. From
the day of AdCre administration until 4 ± 6 weeks later, mice
were administered cyclosporin A orally, 20 mg/kg/day.
Cyclosporin A from mouse plasma samples was measured
using a ¯uorescence polarization immuno assay (Chan et al.,
1992). Mice were anesthetized with a mixture of 200 mg
tribromoethanol and 100 ml amyl alcohol in 7.8 ml normal
saline, ®ltered through a 0.22 mm ®lter, and brought to a
temperature of 378C. The trachea was reached through a
cranio-caudal incision in the neck. Ten ml containing 106 to
107 pfu [plaque-forming units] were injected intratracheally.
Subsequently the incision was sutured.
References
Akagi K, Sandig V, Vooijs M, van de Valk M, Giovannini
M, Strauss M and Berns A. (1997). Nucleic Acids Res., 25,
1766 ± 1773.
Bohinski RJ, Hu€man JA, Whitsett JA and Lattier DL.
(1993). J. Biol. Chem., 268, 11160 ± 11166.
Bos JL. (1989). Cancer Res., 49, 4682 ± 4689.
Carey FA, Wallace WA, Fergusson RJ, Kerr KM and Lamb
D. (1992). Thorax, 47, 1041 ± 1043.
Chan GL, Weinstein SS, LeFor WW, Spoto E, Kahana L
and Shires DL. (1992). Ther. Drug Monit., 14, 42 ± 47.
Contag CH, Jenkins D, Contag PR and Negrin RS. (2000).
Neoplasia, 2, 41 ± 52.
Cooper CA, Carby FA, Bubb VJ, Lamb D, Kerr KM and
Wyllie AH. (1997). J. Pathol., 181, 401 ± 404.
Edwards CW. (1987). J. Clin. Pathol., 40, 125 ± 135.
Fijneman RJ, Oomen LC, Snoek M and Demant P. (1995).
Immunogenetics, 41, 106 ± 109.
Garcia SB, Park HS, Novelli M and Wright NA. (1999). J.
Pathol., 187, 61 ± 81.
Gazdar AF. (1994). Anticancer Res., 14, 261 ± 267.
Gibbs JB. (2000). Science, 287, 1969 ± 1973.
Hanahan D and Weinberg RA. (2000). Cell, 100, 57 ± 70.
Howell JM, Lochmuller H, O'Hara A, Fletcher S, Kakulas
BA, Massie B, Nalbantoglu J and Karpati G. (1998). Hum.
Gene Ther., 9, 629 ± 634.
Huncharek M, Muscat J and Geschwind JF. (1999).
Carcinogenesis, 20, 1507 ± 1510.
Johnson L, Mercer K, Greenbaum D, Bronson RT, Crowley
D, Tuveson DA and Jacks T. (2001). Nature, 410, 1111 ±
1116.
Kaufmann O and Dietel M. (2000). Histopathology, 36, 8 ±
16.
Khoor A, Stahlman MT, Gray ME and Whitsett JA. (1994).
J. Histochem. Cytochem., 42, 1187 ± 1199.
Land H, Chen AC, Morgenstern JP, Parada LF and
Weinberg RA. (1986). Mol. Cell Biol., 6, 1917 ± 1925.
Loonstra A, Vooijs M, Berna Beverloo H, Al Allak B, Van
Drunen E, Kanaar R, Berns A and Jonkers J. (2001). Proc.
Natl. Acad. Sci. USA, 98, 9209 ± 9214.
Mahler JF, Stokes W, Mann PC, Takaoka M and Maronpot
RR. (1996). Toxicol. Pathol., 24, 710 ± 716.
Malkinson AM. (1998). Exp. Lung Res., 24, 541 ± 555.
Manne V, Bekesi E and Kung HF. (1985). Proc. Natl. Acad.
Sci. USA, 82, 376 ± 380.
Maronpot RR, Palmiter RD, Brinster RL and Sandgren EP.
(1991). Exp. Lung Res., 17, 305 ± 320.
McGrath JP, Capon DJ, Goeddel DV and Levinson AD.
(1984). Nature, 310, 644 ± 649.
Conditionally activated K-Ras induces murine lung tumors
R Meuwissen et al
6557
Acknowledgments
We thank Loes Rijswijk and Fina van de Ahe for expert
assistance with the microsurgical procedures, Nell Bosnie
for supervising the daily care of the mice, Jurjen Bulthuis
for preparation of the histological slides, Paul Krimpenfort
for injecting fertilized oocytes and Olaf van Tellingen for
determination of cyclosporin A blood levels. This work
was supported by a grant of the Dutch Cancer Society.
SC Linn was a recipient of a KWF fellowship.
Mills NE, Fishman CL, Rom WN, Dubin N and Jacobson
DR. (1995). Cancer Res., 55, 1444 ± 1447.
Omer CA, Chen Z, Diehl RE, Conner MW, Chen HY,
Trumbauer ME, Gopal-Truter S, Seeburger G, Bhim-
nathwala H, Abrams MT, Davide JP, Ellis MS, Gibbs JB,
Greenberg I, Koblan KS, Kral AM, Liu D, Lobell RB,
Miller PJ, Mosser SD, O'Neill TJ, Rands E, Schaber MD,
Senderak ET, Oli€ A and Kohl NE. (2000). Cancer Res.,
60, 2680 ± 2688.
Oomen LC, van der Valk MA and Demant P. (1991). Exp.
Lung Res., 17, 283 ± 304.
Ordonez NG. (2000). Adv. Anat. Pathol., 7, 123 ± 127.
Otake K, Ennist DL, Harrod K and Trapnell BC. (1998).
Hum. Gene Ther., 9, 2207 ± 2222.
Oudejans JJ, Slebos RJ, Zoetmulder FA, Mooi WJ and
Rodenhuis S. (1991). Int. J. Cancer, 49, 875 ± 879.
Park JY, Jeon HS, Park SH, Park TI, Son JW, Kim CH, Park
JH, Kim IS, Jung TH and Jun SH. (2000). Lung Cancer,
30, 83 ± 89.
Pellegata NS, Sessa F, Renault B, Bonato M, Leone BE,
Solcia E and Ranzani GN. (1994). Cancer Res., 54, 1556 ±
1560.
Rodenhuis S and Slebos RJ. (1992). Cancer Res., 52, 2665s ±
2669s.
Rodenhuis S, Slebos RJ, Boot AJ, Evers SG, Mooi WJ,
Wagenaar SS, van Bodegom PC and Bos JL. (1988).
Cancer Res., 48, 5738 ± 5741.
Rom WN, Hay JG, Lee TC, Jiang Y and Tchou-Wong KM.
(2000). Am. J. Respir. Crit Care Med., 161, 1355 ± 1367.
Saitoh A, Kimura M, Takahashi R, Yokoyama M, Nomura
T, Izawa M, Sekiya T, Nishimura S and Katsuki M.
(1990). Oncogene, 5, 1195 ± 1200.
Sauer B. (1998). Methods, 14, 381 ± 392.
Schmidt EE, Taylor DS, Prigge JR, Barnett S and Capecchi
MR. (2000). Proc. Natl. Acad. Sci. USA, 97, 13702 ±
13707.
Shields JM, Pruitt K, McFall A, Shaub A and Der CJ.
(2000). Trends Cell Biol., 10, 147 ± 154.
Shimkin MB and Stoner GD. (1975). Adv. Cancer Res., 21,
1 ± 58.
Slebos RJ, Kibbelaar RE, Dalesio O, Kooistra A, Stam J,
Meijer CJ, Wagenaar SS, Vanderschueren RG, van
Zandwijk N and Mooi WJ. (1990). N. Engl. J. Med.,
323, 561 ± 565.
Soriano P. (1999). Nat. Genet., 21, 70 ± 71.
Sozzi G, Miozzo M, Pastorino U, Pilotti S, Donghi R,
Giarola M, De Gregorio L, Manenti G, Radice P and
Minoletti F. (1995). Cancer Res., 55, 135 ± 140.
Oncogene
 Conditionally activated K-Ras induces murine lung tumors
R Meuwissen et al
6558
Suda Y, Aizawa S, Hirai S, Inoue T, Furuta Y, Suzuki M,
Hirohashi S and Ikawa Y. (1987). EMBO J., 6, 4055 ±
4065.
Ten Have-Opbroek AA, Ben®eld JR, van Krieken JH and
Dijkman JH. (1997). Histol. Histopathol., 12, 319 ± 336.
Tsuchiya E, Furuta R, Wada N, Nakagawa K, Ishikawa Y,
Kawabuchi B, Nakamura Y and Sugano H. (1995). J.
Cancer Res. Clin. Oncol., 121, 577 ± 581.
Tuveson DA and Jacks T. (1999). Oncogene, 18, 5318 ± 5324.
Whitsett JA, Nogee LM, Weaver TE and Horowitz AD.
(1995). Physiol Rev., 75, 749 ± 757.
Oncogene
Wingo PA, Ries LA, Giovino GA, Miller DS, Rosenberg
HM, Shopland DR, Thun MJ and Edwards BK. (1999). J.
Natl. Cancer Inst., 91, 675 ± 690.
World Health Organization: Histological Typing of Lung
Tumours, 2nd ed. Geneva: World Health Organization,
1981.
Zsengeller ZK, Wert SE, Hull WM, Hu X, Yei S, Trapnell
BC and Whitsett JA. (1995). Hum. Gene Ther., 6, 457 ±
467.