Carcinogenesis vol.21 no.8 pp.1477–1484, 2000
3-Methylcholanthrene triggers the differentiation of alveolar
tumor cells from canine bronchial basal cells and an altered p53
gene promotes their clonal expansion
Ank A.W.TenHave-Opbroek3, Xu-Bao Shi1 and
Paul H.Gumerlock2
Department of Pulmonology, Leiden University Medical Center,
PO Box 96020, NL-2300 RC Leiden, The Netherlands and 1Department of
Urology and 2Department of Internal Medicine, University of California
Davis Medical Center, Sacramento, CA 95817, USA
3To
whom correspondence should be addressed
Email: [email protected]
Alveolar type II cells arising in canine bronchial autografts
following exposure to 3-methylcholanthrene (MCA) give
rise to carcinomas of varying glandular and squamous
growth patterns. To study the role of the tumor suppressor
gene p53 in this process, sections from progressive lesions
were immunostained for p53 protein; microdissected
regions were screened for p53 mutations. Adjacent sections
were examined for type II cell expression [cuboid cell
shape, large roundish nucleus, cytoplasmic staining for
surfactant protein A (SP-A)] and proliferating cell nuclear
antigen expression. Evidence for an altered p53 function
(nuclear staining, missense mutations) was found in most
carcinomas of all histologic types and in all grades of
bronchial dysplasia, but not in hyperplastic or normal
bronchial epithelium. It was primarily associated with the
hyperplastic type II cell populations present in the basal
zone of the lesions. In addition, we found SP-A staining in
hyperplastic (but not in normal) bronchial basal cells. These
data suggest that MCA initiates type II cell differentiation
through phenotypic selection (basal cells). Inactivation of
the p53 gene promotes the clonal expansion of the type II
cells into discernible populations of (squamous or glandular) alveolar tumor cells. This in vivo study is the first to
show that p53 is involved in a specific pathway leading to
bronchogenic carcinoma.
Introduction
The tumor suppressor gene, p53, is the most commonly
mutated gene in human lung and other cancers. Wild-type p53
protein is active as a transcription factor and able to initiate
cell-cycle arrest at the G1 checkpoint (1) and apoptosis of
DNA-damaged cells (2). Mutant p53 protein fails to function
as a transcription factor and accumulates in the nuclei because
of an extended half-life relative to the wild-type protein (3).
Studies using immunohistochemistry have detected high levels
of p53 protein in human lung carcinomas (4). The majority of
these tumors have mutations in the p53 gene (4,5). Several
authors report on increasing levels of p53 protein (6–8) and
p53 gene mutations (9) during early stages of bronchial
carcinogenesis. However, to date, it is not known whether
Abbreviations: H&E, hematoxylin and eosin; MCA, 3-methylcholanthrene;
PCNA, proliferating cell nuclear antigen; SBA, subcutaneous bronchial
autograft; SP-A, surfactant protein A; SSCP, single-strand conformation
polymorphism.
© Oxford University Press
such alterations are associated with specific tumor progenitor
cells or cellular pathways that lead to lung cancer. Information
about this question is important from pathogenetic and therapeutic standpoints.
Bronchogenic carcinoma is the most important type of
human lung cancer. Tobacco smoking is a high risk factor for
the disease. Among the many components of cigarette smoke,
polycyclic aromatic hydrocarbon compounds (10) are strongly
implicated as causative factors. Examples of these compounds,
such as 3-methylcholanthrene (MCA), are therefore used to
study the disease in animal models. The results indicate that
MCA is a potent and appropriate lung carcinogen (11–13).
Exposure of canine bronchial epithelium to MCA leads to
subsets of bronchogenic carcinoma as found in humans (14–
16). The canine model [subcutaneous bronchial autografts
(SBAs) exposed to MCA (15)] offers optimal conditions for
studying the cellular and molecular pathways involved in
bronchial carcinogenesis. The exact origin of the lesions is
known, i.e. the normal bronchial epithelium of the lobar or
segmental bronchi used for grafting. The easy access to the
lesions allows sampling during tumor progression, without
loss of the host (dog).
Previous studies of the canine MCA-SBA model have
identified a novel tumor progenitor cell for bronchogenic
carcinoma, namely alveolar type II cells (17–20). Alveolar
type II cells are the actively proliferating secretory cells of the
normal distal lung. Distinctive features which do not depend
upon cellular maturity, are the approximately cuboid shape,
large and roundish nucleus, and presence of the major surfactant
protein A (SP-A) and multilamellar bodies or their precursory
forms in the cytoplasm (21,22). These properties allow distinction from bronchial epithelial cells such as mucous, basal and
Clara (23) cells. Cells manifesting the type II phenotype have
been found in bronchioloalveolar and other adenocarcinomas
(17–20) and squamous cell carcinomas (18; unpublished data)
that arose from MCA-treated SBAs. They were also present
in bronchial dysplasia and in situ carcinoma (17–19) but not
in canine bronchial epithelium before carcinogen exposure
(17–20). The type II cells predominated in the basal epithelial
layers of the progressive lesions (17–20) where they were
associated with increased proliferation (19). In view of this,
we have postulated that type II cells arising in major bronchi
following carcinogen exposure may function as pluripotential
stem cells (see Definitions in Materials and methods) in
bronchial carcinogenesis and give rise to type II cell carcinomas
of varying glandular and squamous growth patterns (17–19).
This hypothesis does not exclude that other potential tumor
stem cells such as mucous or basal cells may contribute to
such pattern formation, although there is no evidence for this
possibility at present.
In this report, we demonstrate that p53 gene mutations are
associated with the development of type II cell carcinomas
from canine MCA-treated SBAs. We provide some evidence
that the type II cells derive from bronchial basal cells under
1477
A.A.W.TenHave-Opbroek, X.B.Shi and P.H.Gumerlock
the influence of MCA. Interestingly, p53 protein accumulation
takes place at stages similar to those found in human bronchial
carcinogenesis. Missense mutations are present in the evolutionarily conserved DNA-binding domain of the p53 gene
(exons 5–7) as has been reported for human lung cancers (1,5).
These data further validate the canine MCA-SBA bronchogenic
carcinoma model as an effective means for in vivo studies of
early cellular growth and growth control.
Materials and methods
Definitions
As defined (17,18), we use the term ‘stem cell’ to indicate the predominant
proliferating cell type that occupies the dividing layers of the (pre)neoplastic
lesions and constitutes the actively dividing and invading part of the neoplasm.
We reserve the term ‘cell of origin’ to indicate the very first undifferentiated
(primordial-like) tumor progenitor cell that appears during conversion (via
metaplasia) of normal bronchial epithelium to bronchogenic carcinoma.
Canine bronchial tissues: tumor induction protocol
To gain insight into the mechanisms that lead to bronchogenic carcinoma, we
developed a heterotopic model in dogs (SBAs exposed to MCA) (14–
16). The SBAs were prepared as described elsewhere (15). Briefly, after
pneumonectomy, the bronchial tree of the left lung was dissected free and
segments of the major bronchi were placed subcutaneously onto the dorsal
back through two short midline incisions (10–14 SBAs per dog). MCA in
sustained release implants was placed in the SBAs 6–8 weeks after preparation.
Bronchial carcinomas (non-small-cell varieties) developed in 77% (86/111)
of these autografts at 16–24 months thereafter; metastases occurred in two
dogs (16). Tumors were successfully transplanted in athymic nude mice (24).
For this study, we used SBA material that was excised after 3–24 months
exposure to MCA. Histologic typing identified abnormal epithelial areas with
mild to severe atypia (n ⬎ 100) and without atypia, interspersed with
hyperplastic or normal bronchial epithelium, and also micro- and macroinvasive carcinomas (n ⫽ 86). This material was used for further light
microscopic and immunohistochemical investigation. A limited number of
progressive lesions (two dysplastic lesions, two adenocarcinomas and two
squamous cell carcinomas) derived from different dogs was subjected to
microdissection-based mutational analysis of p53 (see below). Unexposed
SBAs and normal canine lungs were used as controls. The studies were
reviewed and approved by the Research Animal Care Committees of the
University of California at Davis, CA.
Light microscopy and immunohistochemistry
To study p53 expression during alveolar type II cell carcinoma development,
formalin-fixed and paraffin-embedded sections (5 µm) from tumor specimens
and controls were placed on poly-L-lysine coated slides. Three consecutive
sections were used for immunohistochemistry and every fourth section for
hematoxylin and eosin (H&E) staining. This allowed morphologic characterization of the immunoreactive cells and histologic typing of the lesions according
to international standards (25,26). We used the following antibodies: (i) mouse
monoclonal Pab240, isotype IgG1, which recognizes a conformation-specific
epitope present on mutant but not wild-type p53 protein in tissue sections
(Santa Cruz Biotechnology Inc.); (ii) mouse monoclonal to proliferating cell
nuclear antigen (PCNA), clone PC10 (Signet Laboratories, Dedham, MA);
and (iii) polyclonal to SP-A, called SALS-Hu. The latter antibody was used
in all previous studies of bronchogenic carcinoma development (17–20,24)
and normal lung morphogenesis (21,27,28) and structure (22). SALS-Hu was
prepared as reported (27,28). Briefly, rabbits were immunized with a surfactantenriched fraction of human bronchoalveolar lavage fluid. After absorption
with serum and cross-reacting organs, the SP-A specificity of the final
immune serum was assessed by immunohistochemistry, western blots, in vitro
translation of mRNA from human lungs and immunoprecipitation. SALS-Hu
recognizes recombinant human SP-A in western blots and can be depleted
with this substance (28). It cross-reacts with canine SP-A in western blots
(17) and lung sections (17–20).
The antibodies were applied to the sections according to the avidin–biotin
complex/peroxidase method using biotinylated horse anti-mouse or swine
anti-rabbit secondary antibodies (Vector Laboratories, Burlingame, CA),
streptavidin peroxidase (Dako Corp., Carpenteria, CA) and 3,3-diaminobenzidine as the chromogen, as reported previously (19,22). Prior to incubation,
endogenous peroxidase activity was quenched with a freshly prepared 3%
hydrogen peroxide solution in methanol. Antigen retrieval for p53 and PCNA
detection was performed as described (19). The sections were counterstained
with Mayer’s hematoxylin. Immunohistochemical controls were performed
1478
with SALS-Hu pre-incubated with SP-A and normal mouse or rabbit serum
as the primary antibody or omission of incubation steps.
Molecular analysis of p53
p53 stained slides were used to select relevant areas with and without p53
staining for DNA extraction from adjacent unstained slides. Using a dissecting
microscope, and with the stained slide as a guide, the selected areas were
dissected off using tiny pieces of razor blade fixed in fine surgical tweezers.
The blades with adhering tissues were placed in microfuge tubes. The contents
of the tubes were deparaffinized by two serial washes in xylene and two
washes in ethanol. After removal of the razor blade fragments, the tissue was
placed in 15 µl TE buffer (pH 8.0) and digested using proteinase-K (0.5 mg/
ml) for 8 h. The proteinase-K was inactivated at 95°C for 5 min. All 15 µl
were used for PCR amplification.
PCR, single-strand conformation polymorphism (SSCP) and sequence analysis
PCR and SSCP were performed essentially as reported (29,30) using primers
designed on the canine p53 sequence. Four segments that correspond to the
evolutionarily conserved regions of p53 in humans (exons 5–8) were amplified
by a multi-step PCR approach with the primers positioned in the flanking
introns. Amplification of the exons was assessed in an ethidium bromide
stained agarose gel using a 123 bp DNA ladder as the marker. In the final
step of PCR the amplified products were labeled with 32P. DNA from a normal
canine lung was used as a control. Specimens with abnormal SSCP patterns
compared with the normal lung were sequenced to confirm mutations.
Sequencing was done as described either by sequencing the PCR products
directly or by cloning prior to sequencing (31). For the cloned samples, three
colonies were selected for DNA sequencing and the same base changes in at
least two colonies were scored as the true mutation.
Results
Immunoreactivity for p53, SP-A and PCNA in lesions from
canine MCA-treated SBAs
Positive staining for p53 was found in most carcinomas
(60%) that arose from MCA-treated SBAs. Light microscopy
performed in adjacent H&E-stained sections classified these
p53-positive tumors as squamous cell carcinomas and various
types of adenocarcinomas (bronchioloalveolar, adenoid cystic
and acinar). Closer inspection of the p53-stained sections
(Figure 1A, adenoid-cystic adenocarcinoma: specimen 89-D595) showed that the nuclear p53 staining predominated in the
basal epithelial layers of the lesions near the connective tissue.
Studies in adjacent H&E-stained sections revealed that these
layers were occupied by approximately cuboid cells with a
large and roundish nucleus. All, or nearly all, these cuboid
cells stained for SP-A (Figure 1B). The cytoplasmic staining
was more intense near the cell borders (see also Figure 3C).
SP-A staining was also found locally in more upper epithelial
layers. PCNA-positive nuclei (Figure 1C) showed a predominately basal distribution in the lesions. All the adenocarcinomas
and squamous cell carcinomas displayed a similar preferential
localization of the p53, SP-A and PCNA staining (i.e. the
cuboid cells with large roundish nuclei present in the basal
zone of the lesions), although the cuboid cells were often less
abundant among squamous lesions compared with glandular
lesions.
Positive staining for p53 was also found in abnormal areas
in the bronchial epithelial lining (Figure 2, specimen 89-R08E). Studies in adjacent H&E-stained sections revealed that
such p53-positive areas were dysplastic (histology: basal hyperplasia and squamous metaplasia with mild to severe atypia; in
situ carcinoma). The p53-positive nuclei (Figure 2) predominated in the basal epithelial layers of the dysplastic lesions but
often expanded into more upper layers in the more severe
lesions. All the dysplastic lesions identified by H&E staining
were p53 positive. Other bronchial lesions, ones not yet
showing signs of potential malignancy, were always p53
negative. There was no p53 staining in the normal or hyperplastic bronchial epithelium present between the lesions.
p53 and alveolar stem cells in bronchial carcinogenesis
Fig. 2. Immunohistochemical staining for p53 in bronchial dysplasia from
canine MCA-treated SBAs (specimen 89-R-08E). Hematoxylin
counterstaining. The p53 nuclear staining locates to the basal zone of the
dysplastic epithelium but may expand into the upper zone in the more
severe (arrow) lesions. Bronchial epithelium between lesions is p53
negative. Magnification 42⫻.
Fig. 1. Immunohistochemical staining for p53 (A), SP-A (B) and PCNA (C)
in carcinomas from canine MCA-treated SBAs (specimen 89-D-595).
Hematoxylin counterstaining. Nuclear staining for p53 predominates in the
basal epithelial layers near the connective tissue. Nearly all the cuboid cells
present in the basal epithelial layers stain for SP-A (cytoplasmic staining).
PCNA nuclear staining shows a predominately basal distribution in the
lesions. Magnification 104⫻.
Further examination of the dysplastic bronchial epithelium
revealed (Figure 3, specimen 92-D-54) that both the p53(Figure 3A) and PCNA-positive nuclei (Figure 3B) predominated in the basal epithelial layers. All, or nearly all, the cuboid
cells occupying these layers stained for SP-A (Figure 3C).
The SP-A staining was cytoplasmic and more intense near the
cell borders. It was also found in larger cells with more
abundant cytoplasm present in adjacent upper layers. Figure
3D illustrates the histology of specimen 92-D-54 (i.e. basal
hyperplasia and squamous metaplasia with mild to moderate
atypia). Areas with regular squamous metaplasia (Figure 3D,
far left) stained for PCNA (Figure 3B) and SP-A (Figure 3C)
but not for p53 (Figure 3A).
Positive staining for SP-A was first detectable in hyperplastic
but otherwise normal bronchial epithelium where it was present
in the basal cells (Figure 4A, specimen 89-R-08E). Such SPA-positive basal cells, which did not stain for p53 (compare
Figure 4A and B), were seen both in conjunction with dysplastic
lesions (Figure 4A) and clearly remote from such lesions.
We never found SP-A or p53 staining in normal bronchial
epithelium, irrespective of whether such epithelium was present
in MCA-treated SBAs (Figure 4A), untreated SBA controls or
normal canine lung.
Tumor sections treated with SALS-Hu preincubated with
SP-A showed extinction of the SP-A immunostaining. Other
immunohistochemical controls were also negative.
The relationship between SP-A and p53 immunoreactivity
during the conversion of normal bronchial epithelium to
bronchogenic carcinoma is summarized in Table I.
p53-positive lesions from canine MCA-treated SBAs contain
p53 gene mutations
To investigate the cause of the p53 protein accumulation in
the canine lesions, initially four carcinomas positive for p53
as identified by immunohistochemistry were examined for p53
mutations. DNA was extracted and, following amplification
of exons 5–8 of p53, the PCR products were subjected to
SSCP analyses as a screen for sequence changes. Two of the
tumors showed abnormal SSCP patterns compared with DNA
from normal canine lung. These abnormal SSCP patterns of
p53 exon 5 of the adenoid-cystic adenocarcinoma 89-D-595
and exon 6 of squamous cell carcinoma 91-D-31 are shown
in Figure 5A and C, respectively. Sequencing of these regions
revealed three point mutations in exon 5 from the 89-D-595
tumor: a missense point mutation at the codon corresponding
to the human codon 153 (CCC→CAC, Pro→His) (Figure 5B)
and silent point mutations at codon 145 (CTG→TTG) and
codon 182 (TGT→TGC). Tumor 91-D-31 showed a single
missense point mutation in exon 6 at codon 197 (GTG→ATG,
Val→Met) (Figure 5D). The other two tumors (bronchioloalveolar adenocarcinoma 89-D-38 and squamous cell carcinoma 92-D-58) revealed no abnormal SSCP patterns from any
of exons 5–8 and were not examined further.
1479
A.A.W.TenHave-Opbroek, X.B.Shi and P.H.Gumerlock
Fig. 3. Immunohistochemical staining for p53 (A), PCNA (B) and SP-A (C) in bronchial dysplasia from canine MCA-treated SBAs (specimen 92-D-54).
Hematoxylin counterstaining. Both the p53 and the PCNA nuclear stainings predominate in the basal epithelial layers. All the cuboid cells present in the basal
epithelial layers stain for SP-A. The membrane pattern of SP-A staining is striking for a secretory protein with intra-cytoplasmic storage. It is also found in
larger cells in adjacent upper layers. (D) Adjacent section stained with H&E. Areas with regular squamous metaplasia (on the far left) stain for SP-A and
PCNA but not for p53 (A–C). Magnification 208⫻.
To investigate the cause of the p53 protein accumulation in
dysplastic bronchial lesions, mutational analysis of p53 was
performed on microdissected regions from tissue sections
adjacent to the sections stained for p53 (specimens 92-D-54
and 89-R-08E). Different regions of epithelial cells, corresponding to cells positive and negative for p53 staining, were
removed from slides and the DNA extracted. The PCR products
from the p53-positive staining dysplastic areas of two specimens showed abnormal migration patterns in SSCP analysis
compared with those from the p53-negative staining normal
bronchial epithelium, connective tissue, hyperplastic bronchial
epithelium and the normal canine lung described above. The
specimen 92-D-54 with mild to moderate atypia showed an
abnormality of exon 6, whereas the other specimen (89-R08E) with mild to severe atypia displayed abnormal migration
of exon 7 (Figure 5E and F). Sequencing confirmed missense
mutations in both specimens, both encoding substitutions of
leucine for proline at codons 223 and 250, respectively (Figure
5G). The results are summarized in Table I.
Discussion
p53-associated development of type II cell carcinomas from
canine major bronchi
The data in this paper demonstrate that the development of
squamous cell carcinomas and adenocarcinomas from canine
MCA-treated SBAs is accompanied by p53 abnormalities.
Nuclear staining for p53 is found in most of the tumors and
in all cases of bronchial dysplasia and in situ carcinoma, but
not in hyperplastic and normal bronchial epithelium. There
are reasons to assume that this staining is due to p53 gene
mutations. Firstly, positive staining for p53 usually results
1480
from accumulation of the mutant form of p53 protein (3).
Secondly, the antibody used for immunohistochemistry
(Pab240) is directed against mutant p53 protein. Finally,
molecular analysis of p53-positive cell samples (obtained from
lesions by microdissection) using PCR, SSCP, cloning of p53
PCR products and DNA sequencing detects missense mutations
in the dysplastic lesions and in two of four end-stage tumors
(exons 5, 6 and 7 of p53; Table I). The p53 protein accumulation
found in the other two tumors may result from mutations
inside exons 5–8 of p53 or outside the examined region, which
were not detected by the present study. Unfortunately, no
material remained for further sequencing.
The canine carcinomas investigated originate from a novel
type of tumor progenitor cells (i.e. alveolar type II tumor stem
cells) that appear in the basal layer of the MCA-exposed
bronchial epithelium (17,18). Clonal expansion of these type
II tumor stem cells, followed by differentiation to squamous
or glandular alveolar tumor cells, are key mechanisms in the
formation of the tumors (17,19) (Introduction). The present
studies of marker expressions in consecutive sections demonstrate that both the p53 and the PCNA nuclear stainings
localize to the type II cell populations present in the basal
zone of the preneoplastic and neoplastic lesions. This strongly
suggests that the p53 gene mutations identified above are
associated with type II tumor stem cells and, thus, with the
genesis of the type II cell carcinomas. However, we cannot
exclude that some mutations are related to other (unknown)
tumor stem cells which may contribute to the genesis of
the tumors.
The presence of type II tumor stem cells in the basal zone
of the preneoplastic and neoplastic lesions was assessed based
upon phenotypic criteria (cuboid cell shape and large roundish
p53 and alveolar stem cells in bronchial carcinogenesis
nucleus) in combination with immunostaining for SP-A. These
criteria allow for type II cell recognition irrespective of cellular
maturity (21,22). The membrane pattern of SP-A staining
found in the present as in the previous studies (17–20) is
Fig. 4. Immunohistochemical staining for SP-A (A) and p53 (B) in
bronchial hyperplasia from canine MCA-treated SBAs (specimen 89-R08E). Hematoxylin counterstaining. Hyperplastic bronchial basal cells stain
for SP-A [shafted arrow in (A)] but not for p53 (B). There is no SP-A
immunoreactivity in normal bronchial epithelium [area marked by
arrowhead in (A)]. The dysplastic lesion present [arrows in (A)] stains for
SP-A (A) and p53 (B). Magnification 130⫻.
striking for a secretory protein with intra-cytoplasmic storage
(21,22). As shown in this paper and previously (17–20), the
type II progeny present in the upper zone of the preneoplastic
and neoplastic lesions often has more abundant cytoplasm
which is (almost) devoid of SP-A staining. This strongly
suggests that type II cell metaplasia, irrespective of the stage
of bronchial carcinogenesis, is accompanied by loss of SP-A
expression. At the ultrastructural level, however, such cells
may still contain (im)mature multilamellar bodies (17,24) as
found in normal fetal or adult alveolar type II cells (32,33).
The present study already detects SP-A staining in hyperplastic
bronchial epithelium (i.e. basal cells). The significance of this
finding will be discussed below.
Role of MCA and p53 in the transformation of bronchial
epithelium to alveolar type II cell carcinoma, as identified in
the canine MCA-SBA model
The data in this paper show that basal cell hyperplasia is
among the first distinct morphologic changes to occur in
canine bronchial epithelium following MCA exposure. The
phenomenon is marked by positive staining for PCNA and
Ki-67 (present study) (19) but not by the presence of p53 gene
mutations. This favors the view that the abnormal proliferation
of the basal cells is not caused by p53 inactivation (1) but
rather is triggered by MCA, possibly through auto/paracrine
loops of growth stimulation (4). As presently shown, the basal
cell hyperplasia is coincident with novel protein expression
(SP-A). This strongly suggests that the type II tumor stem
cells that generate the type II carcinomas (see above) differentiate from normal bronchial basal cells under the influence of
MCA. This process is apparently based upon phenotypic
selection (basal cells).
Previous work in the canine bronchogenic carcinoma model
(17,18) and in nude mice (17,20) indicates that type II tumor
stem cells may give rise to an actively proliferating yet stable
type II cell population. This means that alterations in the genetic
make-up induced by MCA may get fixed in the DNA and can
be inherited. The biologic impact of this event depends on the
status of the genes that control cellular proliferation and survival.
As mentioned above, the p53 gene is not mutated at the time that
potential type II tumor stem cells differentiate from bronchial
Table I. SP-A/p53 immunoreactivity and p53 gene mutations during the conversion of normal bronchial epithelium to bronchogenic carcinoma as identified in
the canine MCA-SBA model
Specimen no. and histology
IHC
SSCP exon no.a
Base pair change
Codonb
Amino acid change
NS
NS
NS
CCC→CTC
CCC→CTC
CTG→TTG
CCC→CAC
TGT→TGC
NS
GTG→ATG
NS
223
250
145
153
182
Proline→leucine
Proline→leucine
None (silent)
Proline→histidine
None (silent)
197
Valine→methionine
SP-A p53
Normal bronchial epithelium
Connective tissue
Hyperplastic bronchial epitheliumc
92-D-54 dysplastic bronchial epitheliumd
89-R-08E dysplastic bronchial epitheliume
89-D-595 adenocarcinoma
–
–
⫹
⫹
⫹
⫹
–
–
–
⫹
⫹
⫹
normal
normal
normal
6
7
5
89-D-38 adenocarcinoma
91-D-31 squamous cell carcinoma
92-D-58 squamous cell carcinoma
⫹
⫹
⫹
⫹
⫹
⫹
normal
6
normal
IHC, immunohistochemistry (graded as: ⫹, present; –, absent); NS, not sequenced.
aAbnormal migration pattern in SSCP analysis of p53 (exons 5–8).
bRelative to human p53 codon.
cRegular basal hyperplasia.
dBasal hyperplasia and squamous metaplasia with mild to moderate atypia.
eBasal hyperplasia and squamous metaplasia with mild to severe atypia and in situ carcinoma.
1481
A.A.W.TenHave-Opbroek, X.B.Shi and P.H.Gumerlock
Fig. 5. Microdissection-based mutational analysis of p53 in adenocarcinoma and squamous cell carcinoma (A–D) and dysplastic bronchial epithelium (E–G)
from canine MCA-treated SBAs. NL, normal canine lung DNA; 89-D-595, and 89-D38, p53 positive canine lung adenocarcinoma DNA; 91-D-31, p53
positive canine squamous cell carcinoma DNA, 92-D-54 and 89-R-08E, p53 positive canine dysplastic bronchial epithelium DNA. (A) SSCP of p53 exon 5.
Note the abnormal migration pattern seen from 89-D-595. (B) DNA sequence of 89-D-595 p53 exon 5. Codon 153 shows a missense point mutation of
CCC→CAC encoding a proline→histidine substitution. (C) SSCP of p53 exon 6. Note the abnormal migration pattern seen from 91-D-31. (D) DNA sequence
of 91-D-31 p53 exon 6. Codon 197 shows a missense point mutation of GTG→ATG encoding a valine→methionine substitution. (E) SSCP of p53 exon 6.
Note the abnormal migration pattern seen from 92-D-54. (F) SSCP of p53 exon 7. Note the abnormal migration pattern seen from 89-R-08E. (G) DNA
sequence of 89-R-08E p53 exon 7. Codon 250 shows a missense point mutation of CCC→CTC encoding a proline→leucine substitution.
basal cells. Most of these cells may therefore undergo cell-cycle
arrest (1) or apoptosis (2). However, those that go through may
have a high proliferation potential (19). The present data suggest
that p53 gene inactivation from MCA may play an important
role in propagating type II tumor stem cells into discernible
populations of squamous or glandular tumor cells.
General discussion
The present observations that p53 protein accumulation and
missense mutations are common findings in canine bronchial
carcinogenesis and already occur at preneoplastic stages similar
to those found in human bronchial carcinogenesis (6–9) validate
the strength of our canine bronchogenic carcinoma model. The
1482
presence of the base substitution mutations in the evolutionary
conserved midregion of the p53 gene is in complete agreement
with findings in human lung and other cancers (1,5) and in
canine osteosarcomas (34). Further extrapolation of our findings to the human situation is not feasible at the present time.
Firstly, information about p53 mutation patterns in similar
(alveolar) subsets of bronchial carcinomas in humans is not
yet available. Secondly, sequence analysis of canine p53 in
the regions of exons 3–8 shows differences in nucleotide
sequences, although there is a strong homology between the
predicted amino acid sequence of the dog and that of other
species including humans, especially in the evolutionary conserved domains II–V (almost 100%) (31). Finally, bronchial
p53 and alveolar stem cells in bronchial carcinogenesis
Fig. 6. Diagram illustrating the novel pathway in canine bronchial carcinogenesis leading to alveolar type II cell carcinomas (17,18) and the role of MCA,
p53 and other factors therein. According to the present study, alveolar type II tumor stem cells may differentiate from normal bronchial basal cells under the
influence of the carcinogen. Clonal expansion into discernible populations of squamous or glandular alveolar tumor cells may result from p53 gene
inactivation besides other growth, genetic or environmental factors.
carcinomas in humans are not induced by one single carcinogen
but result from the impact of multiple endogenous and exogenous factors (4,18).
Our observation that alveolar type II cell differentiation can
be induced in the epithelium of the larger airways of the dog
(present study; 17,18) is in line with observations in culture
studies in the fetal rat (35,36). At present, there is no evidence
that a similar process takes place in normal fetal or adult
lungs. In the light of our findings in the canine MCA-SBA
model (17–20) and in view of present insights into embryonic
lung differentiation (21,37,38), we have postulated that the
unusual differentiation in carcinogen-exposed bronchial epithelium may take place through an oncofetal mechanism (17,18).
In the embryo, both the bronchial and the alveolar epithelial
cells originate from the undifferentiated primordial epithelium
of the lung anlage. Very likely, chemical carcinogenesis in
adult bronchi results in a biological modulation (retrodifferentiation) of existing bronchial basal cells to more pluripotent
primordial-like cells. Acquisition of the type II phenotype may
come from (simultaneous or subsequent) activation of genes
that produce alveolar type II cell expression but were repressed
in utero in earmarked bronchial epithelial cells; this phenomenon is not unlike the appearance of α-fetoprotein in hepatocellular carcinoma. Recent studies performed by us and other
investigators using immunohistochemistry and electron microscopy, detect the alveolar type II cell phenotype in dysplastic
lesions of the human bronchus (39) and in human squamous
cell carcinomas and adenocarcinomas (39,40) (reviewed in ref.
18). Therefore, it seems not unlikely that similar mechanisms
as described for the canine MCA-SBA model above may play
a role in human bronchogenic carcinoma development.
Conclusions
We have investigated the role of the p53 gene with respect to
tumor stem cell development and expansion using a bronchogenic carcinoma model in dogs (s.c. bronchial autografts
exposed to MCA). Our data suggest (Figure 6) that carcinogens
like MCA stimulate normal bronchial basal cells to differentiate
into alveolar type II tumor stem cells (see Definitions in
Materials and methods). Clonal expansion of these tumor stem
cells, followed by squamous or glandular differentiation, may
result from inactivation of the p53 gene besides other growth,
genetic and environmental factors. This process may lead to
alveolar type II cell carcinomas of varying glandular and
squamous growth patterns. It is evident that this tumor classification is based upon the predominant proliferating cell type
in the lesions. It does not exclude that also other pathways,
e.g. those established by potential tumor stem cells such as
mucous and basal cells, could contribute to the tumor formation,
although there is no evidence for this possibility at present.
This in vivo study of bronchial carcinogenesis is the first to
show that an altered p53 gene is involved in a specific
(alveolar) pathway leading to bronchogenic carcinoma.
Acknowledgements
We wish to acknowledge Joost Boex (Pulmonology, Leiden University Medical
Center) and Jeremy Johnson (Cardiothoracic Surgery, University of California
1483
A.A.W.TenHave-Opbroek, X.B.Shi and P.H.Gumerlock
Davis Medical Center) for expert technical asssistance, and thank Dr Derick
Lau (Hematology and Oncology, University of California Davis Medical
Center) for supplying primers for canine p53 amplification.
References
1. Levine,A.J., Momand,J. and Finlay,C.A. (1991) The p53 tumour suppressor
gene. Nature, 351, 453–456.
2. Clarke,A.R., Purdie,C.A., Harrison,D.J., Morris,R.G., Bird,C.C.,
Hooper,M.L. and Wyllie,A.H. (1993) Thymocyte apoptosis induced by
p53-dependent and independent pathways. Nature, 362, 849–852.
3. Oren,M., Reich,N.C. and Levine,A.J. (1982) Regulation of the cellular
p53 tumor antigen in teratocarcinoma cells and their differentiated progeny.
Mol. Cell Biol., 2, 443–449.
4. Brambilla,E. and Brambilla,Ch. (1997) p53 and lung cancer. Pathol. Biol.,
45, 852–863.
5. Hollstein,M., Sidranski,D., Vogelstein,B. and Harris,C.C. (1991) p53
mutations in human cancers. Science, 253, 49–53.
6. Brambilla,E., Gazzeri,S., Lantuejoul,S., Coll,J.L., Moro,D., Negoescu,A.
and Brambilla,Ch. (1998) p53 mutant immunophenotype and deregulation
of p53 transcription pathway (Bcl2, Bax and Waf1) in precursor bronchial
lesions of lung cancer. Clin. Cancer Res., 4, 1609–1618.
7. Nuorva,K., Soini,Y., Kamel,D., Autio-Harmainen,H., Risteli,L., Risteli,J.,
Vähäkangas,K. and Pääkkö,P. (1993) Concurrent p53 expression in
bronchial dysplasias and squamous cell lung carcinomas. Am. J. Pathol.,
142, 725–732.
8. Boers,J.E., Ten Velde,G.P.M. and Thunnissen,F.B.J.M. (1996) p53 in
squamous metaplasia: A marker for risk of respiratory tract carcinoma.
Am. J. Respir. Crit. Care Med., 153, 411–416.
9. Sozzi,G., Miozzo,M., Donghi,R., Pilotti,S., Cariani,C.T., Pastorino,U.,
Della Porta,G. and Pierotti,M.A. (1992) Deletions of 17p and p53 mutations
in preneoplastic lesions of the lung. Cancer Res., 52, 6079–6082.
10. Shopland,D.R., Eyre,H.J. and Pechacek,T.F. (1991) Smoking attributable
cancer mortality in 1991: is lung cancer now the leading cause of death
among smokers in the United States? J. Natl Cancer Inst., 83, 1142–1148.
11. Eastman,A. and Bresnick,E. (1979) Persistent binding of 3methylcholanthrene to mouse lung DNA and its correlation with
susceptibility to pulmonary neoplasia. Cancer Res., 39, 2400–2405.
12. Stewart,B.W. and Haski,R. (1984) Relationships between carcinogen
metabolism, adduct binding and DNA damage in 3-methylcholanthreneexposed lung. Chem. Biol. Interact., 52, 111–128.
13. Douriez,E.,
Kermanach,P.,
Fritsch,P.,
Bisson,M.,
Morlier,J.P.,
Monchaux,G., Morin,M. and Laurent,P. (1994) Cocarcinogenic effect of
cytochrome P-450 1A1 inducers for epidermoid lung tumour induction in
rats previously exposed to radon. Radiat. Protect. Dosim., 56, 105–108.
14. Benfield,J.R. and Hammond,W.G. (1992) Bronchial and pulmonary
carcinogenesis at focal sites in dogs and hamsters. Cancer Res., 52 (suppl.),
2687s–2693s.
15. Hammond,W.G., Benfield,J.R., Paladugu,R.R., Azumi,N., Pak,H.Y. and
Teplitz,R.L. (1986) Carcinogenesis in heterotopic respiratory epithelium
in canine subcutaneous bronchial autografts. Cancer Res., 46, 2995–2999.
16. Derrick,M.J., Hammond,W.G., Pak,H.Y., Azumi,N., Smith,S.S. and
Benfield,J.R. (1988) Non-small cell lung cancer in autogenous
subcutaneous bronchial grafts in dogs. J. Thorac. Cardiovasc. Surg., 95,
562–571.
17. Ten Have-Opbroek,A.A.W., Hammond,W.G., Benfield,J.R., Teplitz,R.L.
and Dijkman,J.H. (1993) Expression of alveolar type II cell markers
in acinar adenocarcinomas and adenoid-cystic carcinomas arising from
segmental bronchi. A study in a heterotopic bronchogenic carcinoma
model in dogs. Am. J. Pathol., 142, 1251–1264.
18. Ten Have-Opbroek,A.A.W., Benfield,J.R., Hammond,W.G., Teplitz,R.L.
and Dijkman,J.H. (1994) Invited review: In favour of an oncofoetal
concept of bronchogenic carcinoma development. Histol. Histopathol., 9,
375–384.
19. Ten Have-Opbroek,A.A.W., Benfield,J.R., Hammond,W.G. and
Dijkman,J.H. (1996) Alveolar stem cells in canine bronchial carcinogenesis.
Cancer Lett., 101, 211–217.
20. Ten Have-Opbroek,A.A.W., Hammond,W.G. and Benfield,J.R. (1990)
1484
Bronchiolo-alveolar regions in adenocarcinoma arising from canine
segmental bronchus. Cancer Lett., 55, 177–182.
21. Ten Have-Opbroek,A.A.W. and Plopper,C.G. (1992) Morphogenetic and
functional activity of type II cells in early fetal Rhesus monkey lungs. A
comparison between primates and rodents. Anat. Rec., 234, 93–104.
22. Ten Have-Opbroek,A.A.W., Otto-Verberne,C.J.M., Dubbeldam,J.A. and
Dijkman,J.H. (1991) The proximal border of the human respiratory unit,
as shown by scanning and transmission electron microscopy and light
microscopical cytochemistry. Anat. Rec., 229, 339–354.
23. Clara,M. (1937) Zur Histobiologie des Bronchalepithels. Z. Mikrosk. Anat.
Forsch., 41, 321–347.
24. Hammond,W.G., Benfield,J.R. and Teplitz,R.L. (1991) Metastatic
behaviour of canine lung carcinoma in autochthonous and xenotransplant
hosts. Clin. Exp. Metastasis, 9, 567–577.
25. Stuenzi,H., Head,K.W. and Nielsen,S.W. (1974) International histological
classification of tumours of domestic animals. I. Tumors of the lung. Bull.
WHO, 50, 9–19.
26. The World Health Organization (1982) Histologic typing of lung tumours.
2nd edn. Am. J. Clin. Pathol., 77, 123–136.
27. Otto-Verberne,C.J.M., Ten Have-Opbroek,A.A.W., Balkema,J.J. and
Franken,C. (1988) Detection of the type II cell or its precursor before
week 20 of human gestation, using antibodies against surfactant-associated
proteins. Anat. Embryol., 178, 29–39.
28. Otto-Verberne,C.J.M., Ten Have-Opbroek,A.A.W. and De Vries,E.C.P.
(1990) Expression of the major surfactant-associated protein, SP-A, in
type II cells of human lung before 20 weeks of gestation. Eur. J. Cell.
Biol., 53, 13–19.
29. Shi,X.B., Bodner,S.M., DeVere White,R.W. and Gumerlock,P.H. (1996)
Identification of p53 mutations in archival prostate tumors. Sensitivity of
an optimized single-strand conformational polymorphism (SSCP) assay.
Diagn. Mol. Pathol., 5, 271–278.
30. Kraegel,S.A., Pazzi,K.A. and Madewell,B.R. (1995) Sequence analysis of
canine p53 in the region of exons 3–8. Cancer Lett., 92, 181–186.
31. Gumerlock,P.H., Chi,S.G., Shi,X.B., Voeller,H.J., Jacobson,J.W.,
Gelmann,E.P. and deVere White,R.W. (1997) The Cooperative Prostate
Network: p53 abnormalities in primary prostate cancer: single-strand
conformation polymorphism analysis of complementary DNA in
comparison with genomic DNA. J. Natl Cancer Inst., 89, 66–71.
32. Ten Have-Opbroek,A.A.W., Dubbeldam,J.A. and Otto-Verberne,C.J.M.
(1988) Ultrastructural features of type II alveolar epithelial cells in early
embryonic mouse lung. Anat. Rec., 221, 846–853.
33. Ten Have-Opbroek,A.A.W., Otto-Verberne,C.J.M. and Dubbeldam,J.A.
(1990) Ultrastructural characteristics of inclusion bodies of type II cells
in late embryonic mouse lung. Anat. Embryol., 181, 317–323.
34. Johnson,A.S., Couto,C.G. and Weghorst,C.M. (1998) Mutation of the p53
tumor suppressor gene in spontaneously occurring osteosarcomas of the
dog. Carcinogenesis, 19, 213–217.
35. Shannon,J.M. (1994) Induction of alveolar type II cell differentiation in
fetal tracheal epithelium by grafted distal lung mesenchyme. Dev. Biol.,
166, 600–614.
36. Shannon,J.M., Gebb,S.A. and Nielsen,L.D. (1999) Induction of alveolar
type II cell differentiation in embryonic tracheal epithelium in
mesenchyme-free culture. Development, 126, 1675–1688.
37. Ten Have-Opbroek,A.A.W. (1981) The development of the lung in
mammals: An analysis of concepts and findings. Am. J. Anat., 162,
201–219.
38. Ten Have-Opbroek,A.A.W. (1991) Lung development in the mouse embryo.
Exp. Lung Res., 17, 111–130.
39. Ten Have-Opbroek,A.A.W., Benfield,J.R., Van Krieken,J.H.J.H. and
Dijkman,J.H. (1997) The alveolar type II cell is a pluripotential stem cell
in the genesis of human adenocarcinomas and squamous cell carcinomas.
Histol. Histopathol., 12, 319–336.
40. Linnoila,R.I., Mulshine,J.L., Steinberg,S.M. and Gazdar,A.F. (1992)
Expression of surfactant-associated protein in non-small-cell lung cancer:
a discriminant between biologic subsets. J. Natl Cancer Inst. Monogr., 13,
61–66.
Received October 19, 1999; revised and accepted April 25, 2000