Increased susceptibility of mice lacking Clara cell 10-kDa protein to lung tumorigenesis by 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone, a potent carcinogen in cigarette smoke.

Ninety percent of all human lung cancers are related to cigarette smoking. Both tobacco smoke and lung tumorigenesis are associated with drastically reduced levels of Clara cell 10-kDa protein (CC10), a multifunctional secreted protein, naturally produced by the airway epithelia of virtually all mammals. We previously reported that the expression of CC10 is markedly reduced in animals exposed to 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone, NNK, a potent carcinogen in tobacco smoke. Furthermore, it has been reported that CC10 expression, induced in certain tumor cells, reverses the transformed phenotype. We demonstrate here that NNK exposure of CC10-knock-out (CC10-KO) mice causes a significantly higher incidence of airway epithelial hyperplasia and lung adenomas compared with wild type (WT) littermates (30% CC10-KO versus 5% WT, p = 0.041). We also found that compared with NNK-treated WT mice, CC10-KO mice manifest increased frequency of K-ras mutation, elevated level of Fas ligand (FasL) expression, and increased MAPK/Erk phosphorylation, all of which are considered predisposing events in NNK-induced lung tumorigenesis. We propose that CC10 has a protective role against NNK-induced lung tumorigenesis mediated via down-regulation of the above-mentioned predisposing events.

Lung cancer is the leading cause of cancer deaths in both men and women in the United States, and 90% of all human lung cancers are related to cigarette smoking (1,2). Both tobacco smoke (3) and lung tumorigenesis (4,5) are associated with reduced levels of Clara cell 10-kDa (CC10) 1 protein, a steroid-inducible, multifunctional, secreted polypeptide that accounts for ϳ7% of the total protein in bronchioalveolar lavage fluid (6,7). CC10, first identified as blastokinin (8) or uteroglobin (9), is the founding member of a newly formed superfamily of proteins called Secretoglobins (10). Most of the proteins of this superfamily are tissue-specifically expressed in the secre-tory epithelia of virtually all organs. The human CC10 gene is mapped to chromosome 11q12.2-13.1 (11) and encodes a 16-kDa homodimeric protein in which the two identical 70-amino acid subunits are covalently linked by two disulfide bonds (6). The altered expression of CC10 (3) or single nucleotide polymorphism in the CC10 gene (9) is associated with a variety of pulmonary diseases in humans (9,(11)(12)(13). Previous studies have suggested a potential protective role of CC10 in suppressing inflammation or modulating the immune response in the lungs following pulmonary injury or infection (6,(11)(12)(13).
It has been reported that CC10 expression is rarely detectable in human non-small cell lung cancers, despite the fact that it is abundantly produced by the progenitor cells in normal airways (4,14). Its expression is drastically reduced in SV40induced carcinogenesis (15,16), and it has been reported that CC10 expression induced in certain cancer cells leads to diminished invasiveness and anchorage-independent growth, characteristic of these cells (17,18). Moreover, the overexpression of CC10 in immortalized bronchial epithelial cells delayed the induction of anchorage-independent growth in response to a potent carcinogen in cigarette smoke, 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK; Refs. 5 and 17-20). Thus, it is suggested that CC10 may have a protective role against lung tumorigenesis, induced by NNK. To test this hypothesis, we exposed wild type (WT) and CC10-knockout (KO) mice to NNK and compared the rate of lung adenoma formation in these two groups of animals. In addition, we determined whether compared with WT mice NNK treatment of the CC10-KO mice caused: (a) epithelial hyperproliferation, (b) a higher incidence of mutations in the proto-oncogene, K-ras, (c) a higher level of FasL expression, and (d) increased phosphorylation of MAPK/ Erk1, all of which are associated with lung tumorigenesis. Our results show that compared with the lungs of NNK-treated WT mice, those of the NNK-treated CC10-KO mice manifest: (i) a markedly higher incidence of airway epithelial hyperplasia and the formation of adenomas, (ii) a markedly increased frequency of K-ras mutations, (iii) a significantly higher level of FasL expression, and (iv) an increased phosphorylation of MAPK/ Erk1. We propose that CC10 plays a critical role in protecting the lungs against NNK-induced hyperplasia and adenoma formation, most likely by suppressing the events that are known to precede tumorigenesis in this organ.

EXPERIMENTAL PROCEDURES
Animals, Exposure, and Tissue Collection-CC10-KO mice (21) were generated by gene targeting in embryonic stem cells. The strain-and age-matched C57BL/6 WT mice (Jackson Laboratory, Bar Harbor, ME) were housed under standard conditions in a National Institutes of Health animal facility. All procedures were approved by the NCI Animal Care and Use Committee. Starting at 8 weeks of age, the mice received NNK intraperitoneally (104 mg/kg of body weight, three times, given every other month) or physiological saline, respectively. The * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. animals were serially sacrificed at 5 and 8 months and at 11 and 12 months. These time periods were chosen because it has been reported that aging B6 and 129 mouse strains spontaneously develop cancers in various organs, including the lung, at a high frequency (22). Two hours prior to sacrifice, mice received 100 g/g 5-bromo-2Ј-deoxyuridine (Br-dUrd, Sigma) intraperitoneally. Representative tissue specimens were fixed in 4% phosphate-buffered paraformaldehyde, embedded in paraffin, sectioned and stained, and embedded in Tissue-Tek® O.C.T. compound (Miles Laboratories Inc., Elkhart, IN) before freezing or snapfrozen in dry ice-cooled 2-methylbutane and stored at Ϫ140°C for molecular analysis as described below.
Immunoprecipitation and Western Blot Analysis-Snap-frozen lung tissues were homogenized by ultrasonication in lysis buffer composed of PBS, 0.1% Nonidet P-40, 0.5% sodium deoxycholate, 1% proteinase and phosphatase inhibitor mixtures (Sigma). The levels of CC10 in the lung lysates were analyzed by immunoprecipitation as described previously (17). The levels of Erk1/2 were directly immunoblotted with a rabbit anti-ERK1 antibody (diluted 1:500; Santa Cruz Biotechnology, Inc.), and the levels of phospho-MAPK and phospho-Elk1 were assessed, respectively, using a rabbit anti-phospho-MAPK polyclonal antibody (diluted 1:1500; New England Biolabs, Inc., Beverly, MA) and a Phospho Plus Elk1 Antibody Kit (New England Biolabs, Inc.) as described in the manufacturer's procedures. The levels of CC10, Erk1, phospho-MAPK and phospho-Elk1 were quantified with a Densitometer (Pharmaceutical Biotech).
DNA Isolation and Laser Capture Microdissection-About 500 cells from adenomas and airway epithelia were acquired from deparaffinized sections using PixCell I (Arcturus Engineering, Mountain View, CA) laser capture microdissection. DNA was purified according to manufacturer's procedures and used for amplification of K-Ras DNA by polymerase chain reaction (PCR) as described below.
Single-strand Conformational Polymorphism (SSCP)-The SSCP analysis was based on the methods described previously (23). Briefly, a 143-base pair K-ras exon 1 DNA fragment was yielded by PCR using primers RasE1F (5Ј-TTA TTG TAA GGC CTG CTG AA-3Ј) and RasE1R (5Ј-GCA GCG TTA CCT CTA TCG TA-3Ј). In addition, a 192-base pair K-ras exon 2 DNA fragment was generated with primers RasE2F (5Ј-TTCTCAGGACTCCTACAGGA-3Ј) and RasE2R (5Ј-ACC CAC CTA TAA TGG TGA AT-3Ј). PCR was carried out in a 20-l PCR reaction mixture containing 3 l of DNA prepared by laser capture microdissection, 2 l of 10ϫ PCR buffer, 0.4 l of 10 mM dNTP, 0.4 l of a 10 M concentration of each primer, 0.4 l of [ 32 P]dCTP (20 Ci/l) (PerkinElmer Life Sciences) and 1 unit of Taq polymerase (Invitrogen). Each sample was subjected to 35 cycles, and each cycle consisted of denaturing at 94°C for 45 s, annealing at 58°C for 45 s, and extending at 72°C for 90 s, with a final step at 72°C for 8 min. An equal amount of PCR product was mixed with 2ϫ SSCP loading buffer (98% formamide, 20 mM EDTA, pH 8, 0.1% bromphenol blue, 0.1% xylene cyanol) and denatured at 95°C for 5 min. Four l of each sample was loaded onto the SSCP gel (FMC BioProducts, Rockland, ME) and run at 6 watts for 16 h in 0.6ϫ TBE at room temperature. The SSCP gels were dried and exposed to autoradiography film.
DNA Sequencing-Mutant DNA derived from a variant SSCP band was amplified by PCR with the primers RasE1F and RasE1R. The amplified DNA was then purified with QIAquick PCR Purification Kit (Qiagen, Valencia, CA) and sequenced with a BigDye Terminator Cycle Sequencing Kit and an ABI PRISM DNA sequencer (PE Applied Biosystems, Foster City, CA).

Morphology of Epithelial Cells in the Airways of WT and
CC10-KO Mice before and after NNK Treatment-To assess the effects of NNK on WT and CC10-KO mouse lungs, we first examined the morphology of airway epithelium and CC10 level (Fig. 1). As expected, while high levels of CC10 expression in the lungs of WT mice are readily detectable (Fig. 1A), none of the airway epithelial cells of CC10-KO mice were positive for CC10 (Fig. 1B). The Clara cells of CC10-KO mice showed a marked reduction in apical cytoplasm, normally the storage site for CC10 protein (Fig. 1, C-F). Most importantly, the NNK treatment of WT mice leads to a precipitous reduction in the levels of lung CC10 as assessed by immunohistochemistry (Fig.  1G) and by immunoblot analysis (Fig. 1H). Morphologically, the apical cytoplasm of bronchial epithelial cells in NNKtreated CC10Ϫ/Ϫ mice remain flat. These results indicate that CC10 deficiency leads to morphological changes in lung epithe-  ). B, lack of CC10 in the bronchiolar epithelium of a CC10-KO mouse (immunoperoxidase staining). C, photomicrograph of a small bronchiolus in a WT mouse lung with non-ciliated secretory (Clara) cells (arrows) with prominent apical cytoplasm protruding to the lumen (hematoxylin-eosin staining). D, flattened appearance of bronchiolar epithelial cells (arrow) in a CC10-KO mouse, which demonstrates reduction in the apical cytoplasm (hematoxylin-eosin staining). E and F, high magnification (ϫ200) of C and D. Following 5 months of NNK exposure of WT mice the airway epithelia show markedly reduced levels of CC10 immunoreactivity (G), which was further confirmed by immunoblot analysis (H). Note the reduction of CC10 protein in the lung lysates of NNK-treated WT mice (lanes 3 and 4). Photomicrographs A-D and G are of the same magnification (ϫ100). lial cells of CC10-KO mice and that NNK treatment of WT mice markedly reduces the production of this protein in the lung.
At 5-8 months after NNK treatment, histopathology revealed focal airway epithelial hyperplasia in 33% (4 of 12 mice) and solitary lung adenomas in 25% (3 of 12 mice) of the CC10-KO animals (Table I). At 11-12 months after NNK exposure, epithelial hyperplasia and small solitary lung adenomas were induced in 38% (3 of 8) of the CC10-KO mice ( Table  I). All NNK-induced lung adenomas in CC10-KO mice were solitary, less than 1 mm in diameter, and of alveolar type II cell lineage (Fig. 2, A-C). Moreover, compared with the cells surrounding the alveoli proliferating cells in adenomas are markedly increased as suggested by an elevated BrdUrd index (Fig.  2, D and E). Only one out of 22 NNK-treated WT mice developed a solitary lung adenoma (p ϭ 0.041). These results indicate that CC10-KO mice are more prone to developing hyperplasia and lung adenomas than their WT counterparts following exposure to NNK.
NNK-treated CC10-KO Manifest a Markedly Elevated Level of Proliferation of Airway Epithelial Cells-We next determined whether lack of CC10 protein had changed the cellular dynamics in the pulmonary epithelia following NNK treatment by studying cell proliferation. Again, at 5-6 months after treatment, about 1.5 and 0.91% of BrdUrd-positive epithelial cells appeared in the airways and alveoli, respectively, of the NNKtreated KO mice (Fig. 3, A and B). This reflected a 3-10-fold increase in the bronchiolar epithelium, as compared with 0.45% of BrdUrd-positive cells in NNK-treated WT mice (p ϭ 0.016), 0.13% in PBS-treated KO mice (p ϭ 0.036), and 0.37% in PBS-treated WT mice (data not shown). These results were confirmed by immunostaining using the cellular proliferation marker Ki-67. There was a 2-6-fold increase in the Ki-67 labeling index in the lungs of NNK-treated CC10-KO, as compared with NNK-treated WT mice, PBS-treated CC10-KO, and WT mice (data not shown). These data suggest that hyperproliferation induced by NNK in KO mice may be responsible for higher incidence of lung adenomas.
NNK Induces Markedly Elevated FasL Expression in CC10-KO Lungs-Since Clara cells are a source of FasL, the expression of which is associated with enhanced tumorigenesis (23, 24), we determined whether CC10-KO mice have an abnormal expression of FasL by immunohistochemistry. While Fas expression appeared similar among the four treatment groups (data not shown), markedly enhanced FasL immunoreactivity occurred in lungs of NNK-treated CC10-KO mice as compared with a minimal to moderate increase in the lungs of NNK-treated WT mice. Using a terminal deoxynucleotidyltransferase-mediated dUTP nick end labeling assay we determined the level of apoptosis in bronchial epithelia of NNKtreated CC10-KO and WT mice and found that NNK treatment markedly increased the level of apoptosis in CC10-KO mice. 2 Moreover, enhanced immunoreactivity for FasL was also detected within alveolar type II cells and lymphocytes in the hyperplasia and adjacent to tumors in NNK-treated KO mice (Fig. 3, C and D). The results indicate that CC10 protein may have a protective role on cellular stability against NNK-induced cellular biochemical alteration such as FasL overexpression.
Increased K-ras Mutation in CC10-KO Lungs-Recent reports indicate that most chemically induced lung tumors (ϳ90%) in A/J mice carry an activating point mutation in the K-ras proto-oncogene (25,26). To determine whether there is an increase in K-ras mutations in the lung tumors of NNKtreated KO mice, we microdissected the tissues and carried out PCR-SSCP and sequencing assay to detect mutations in exons 1 and 2 of this gene. This area contains clusters of the known transformation mutations. We found that three out of six lung adenomas (50%) in NNK-treated CC10-KO mice had K-ras mutations in codon 12 of exon 1 (supplemental Fig. 1), in which GGT was transversed to GCT, resulting in an amino acid change from glycine to alanine. A K-ras mutation at codon 32 was found in a lung adenoma of an NNK-treated WT mouse, in which TAC was transitioned to TAT, resulting in a silent mutation. All three K-ras mutations in NNK-exposed KO mice were early phase alveolar adenomas, suggesting that CC10 deficiency confers susceptibility to NNK-mediated genomic instability such as K-ras mutations, which may contribute to carcinogenesis.
Elevated Levels of MAPK/Erk1 Phosphorylation in NNKtreated CC10-KO Mouse Lungs-Since the MAPK cascade, situated downstream from activating ras mutations, plays a critical role in regulating cell growth, proliferation, and responses to extracellular signals (27,28), we examined the activation  2. Photomicrographs of NNK-induced lung tumors in CC10-KO mice. A, a small solitary adenoma in the peripheral lung of a NNK-treated CC10-KO mouse. All of the NNK-induced adenomas in KO mice were solitary and less than 1 mm in diameter. Note the location in the alveolar compartment without any apparent connection with the airways (hematoxylin-eosin stain, ϫ15). B, higher magnification (ϫ200) reveals a tubular adenoma with minimal atypia. C, tumor cells are positive for a type II pneumocyte marker SP-C, suggesting alveolar type II cell lineage of the tumor (immunoperoxidase stain, ϫ700). D, BrdUrd index in the adenomas was increased 5-fold, compared with the surrounding alveoli, while there was no significant difference in tumors or surrounding alveoli between NNK-treated KO and WT mice (immunoperoxidase staining, ϫ100). E, numerous tumor cells displayed Ki-67 immunoreactivity (immunoperoxidase staining, ϫ100). Throughout the lung tissue including adenomas, Ki-67-positive cells were observed at higher numbers than BrdUrd-positive cells. stages of MAPK by immunobloting analysis (Fig. 4). We found elevated levels of Erk1 in NNK-treated lungs from both KO and WT mice (Fig. 4A). Moreover, we also found dramatically increased MAPK activity in NNK-treated CC10-KO mouse lungs with adenomas as compared with those in PBS-treated lungs (Fig. 4A). The relative levels of phospho-MAPK in PBS-treated lungs from KO animals were twice as high as from WT mice and enhanced up to 3-fold following the exposure to NNK (Fig.  4B). Western blot analysis of phosphorylated Elk1 also showed similar results. Thus it appears that the lungs of CC10-deficient mice are more sensitive to NNK-induced activation of the MAPK that mediates increased cellular proliferation and cell growth, leading to tumorigenesis. DISCUSSION In this study, we provided evidence that CC10-KO mice are significantly more susceptible to developing hyperproliferation of airway epithelial cells and the formation of pulmonary adenomas in response to NNK-treatment. Smoking remains a significant health hazard throughout the world, and NNK is a potent carcinogen in cigarette smoke. We discovered that mice exposed to this carcinogen express drastically lower levels of CC10 as does cigarette smoking (3). The results of previous studies (17)(18)(19)(20) suggested an anti-carcinogenic role of this protein in vitro. More specifically, forced overexpression of CC10 in lung cancer cells led to diminished invasiveness and anchorage-independent growth, while the overexpression of CC10 in immortalized bronchial epithelial cells delayed the induction of anchorage-independent growth in response to NNK (17,19). Using CC10-KO mice, we now demonstrate that CC10 possesses a protective role against NNK-induced lung tumorigen-esis. Aging CC10-KO mice often develop multiorgan tumors, but NNK-induced lung tumors occurred much earlier (5-12 months after exposure). Structural abnormalities as well as susceptibility to oxygen toxicity of CC10-KO mice have been reported previously (29,30), and we have noted some of these abnormalities in the present study. Since structural specialization is related to cellular functions, the observed microscopic differences in CC10-KO mice may provide the basis for functional changes where the lack of CC10 protein hampers proper physiological activities.
Our current results further indicate that NNK exposure of CC10-KO mice manifests: (i) increase in cellular proliferation, (ii) elevated expression of FasL, (iii) increased mutation in proto-oncogene K-ras, and (iv) activation of MAPK signaling pathway, all of which are associated with lung tumorigenesis. Our results under score the necessity of further studies to establish a cause and effect relationship of these observations. It has been reported that nitrosamine-induced lung tumors frequently reveal activated K-ras in A/J mice (25,26,31), while NNK-exposed C57BL6 mice, the genetic background of our CC10-KO, have a very low incidence of K-ras mutations (32).
FIG. 3. CC10-KO mice manifest hyperproliferation of airway epithelium and increased FasL expression. Bronchiolar epithelia of mice exposed to 5 months of NNK are shown. A, WT mouse with a single labeled cell (arrow). It should be noted that no statistically significant differences between WT and CC10-KO mice were detected in PBS-treated control groups, although the lung BrdUrd labeling index was 3-fold lower in the airways of CC10-KO mice compared with those in WT mice. B, NNK-exposed KO mouse with numerous labeled bronchiolar cells (arrows), resulting in a significantly higher (up to 10-fold) labeling index than in WT mice. (A and B, immunoperoxidase staining for BrdUrd, ϫ100.) In contrast, no statistically significant difference of BrdUrd labeling index in the lung was found in WT animals between NNK treatment and PBS treatment. In the alveolar compartment, following NNK exposure there was a 2-3-fold increase in the labeling index of KO mice, as compared with 0.51% of BrdUrd-positive cells in WT mice (p ϭ 0.029). No statistically significant difference in the alveolar BrdUrd labeling index was found between CC10-KO and WT mice in PBS-treated control groups. C, WT control animal reveals minimal immunoreactivity in the bronchiolar epithelial cells for FasL. D, at 5 months of NNK exposure, there is increased staining in the airway epithelium in CC10-KO animals. (C and D, immunoperoxidase staining, ϫ300).

FIG. 4.
Activation of MAPK kinase pathway associated with NNK exposure in CC10-deficient mice. The levels of Erk1 and phosphorylated MAPK (phospho-MAPK) were analyzed by immunoblotting of lung lysates using state-specific antibodies. A representative gel from three CC10-KO and three WT mice is shown in A. Although the levels were variable, the average levels of Erk1 were unchanged between NNK-and PBS-treated CC10-KO and WT mice. B, the level of phospho-MAPK compared with the total kinase was determined by densitometric analysis and expressed in relative units of phospho-MAPK/ERK1. The level in KO control lung (bar 1) was 2-fold as compared with that of WT control (bar 4), and the exposure to NNK at 8 months was associated with a 3-fold increase in the CC10-KO tumor lung (bar 2) and lung (bar 3) lysates, when compared with the lysates from corresponding WT NNK-treated animals (bars 5 and 6, respectively).
Recent studies also suggest that wild type K-ras may have a tumor-suppressive role in resistant mouse strains and MAPK may be activated in only those lung neoplasms that have both K-ras mutations and the loss of the wild type allele (33). On the other hand, activated ERK1/2 has been associated with increased proliferation and lung tumorigenesis (34).
How CC10 deficiency might confer NNK-induced hyperproliferation of airway epithelial cells, formation of adenomas, elevated FasL expression, and MAPK/Erk1 phosphorylation is not yet clear. However, it has been reported that CC10 binds to cell surface proteins (putative receptors) with high affinity and specificity and regulates several biological functions of this protein (18,35). Moreover, we have recently demonstrated that CC10 suppresses allergen-induced phosphorylation of MAPK and inhibits NF-B activation and the expression of cyclooxygenases-2 (COX-2) in CC10-KO mice (36). Characterizations of its receptor(s) and the signaling pathways may delineate how CC10 exerts its inhibitory effects on NNK-induced lung tumorigenesis.
In summary, our data for the first time clearly demonstrate a protective role of CC10 against NNK-induced lung tumorigenesis and provide some insight into the potential role of CC10 in suppressing some of the concomitants of NNK-induced lung carcinogenesis such as increased Ras mutation, FasL expression, and MAPK phosphorylation. It is likely that the mechanism by which CC10 protects against NNK-induced cellular hyperproliferation and tumorigenesis involves the maintenance of cellular integrity and inhibition of Ras/MAPK signaling pathways.