Novel medium‐term carcinogenesis model for lung squamous cell carcinoma induced by N‐nitroso‐tris‐chloroethylurea in mice

Abstract

Targeted treatments for lung cancer based on pathological diagnoses are required to enhance therapeutic efficacy. There are few well‐established animal models for lung squamous cell carcinoma although several highly reproducible mouse models for lung adenoma and adenocarcinoma are available. This study was carried out to establish a new lung squamous cell carcinoma mouse model. In the first experiment, female A/J mice were painted topically on back skin twice weekly with 75 μL 0.013 M N‐nitroso‐tris‐chloroethylurea for 2, 4, and 8 weeks (n = 15–20 per group) as initiation of lung lesions, and surviving mice were killed at 18 weeks. In the second experiment, mice were treated as above for 4 weeks and killed at 6, 12, or 18 weeks (n = 3 per group). Lung lobes were subjected to histopathological, immunohistochemical, immunoblotting, and ultrastructural analyses. In the case of treatment for 2, 4, and 8 weeks, incidences of lung squamous cell carcinoma were 25, 54, and 71%, respectively. Cytokeratin 5/6 and epidermal growth factor receptor were clearly expressed in dysplasia and squamous cell carcinoma. Desmosomes and tonofilaments developed in the squamous cell carcinoma. Considering the carcinogenesis model, we conclude that 2 or 4 weeks of N‐nitroso‐tris‐chloroethylurea treatment may be suitable for investigating new chemicals for promotional or suppressive effects on lung squamous cell carcinoma.

We have established a new mouse model for lung SCC. Our findings indicated that the NTCU treatment duration‐dependent increase in lung SCC is related to the up‐regulation of cytokeratin 5/6 and EGFR expression.

Histopathological classification of lung cancer is important as a prognostic factor and for the evaluation of treatment methods. The World Health Organization classifies lung cancer as non‐small‐cell lung cancer (NSCLC) and small‐cell lung cancer. Squamous cell carcinoma (SCC), adenocarcinoma, and large cell carcinoma are classified as NSCLC and account for approximately 85% of all lung cancers.1 Cumulative information suggests that SCC and adenocarcinoma progress through different carcinogenic pathways.2 But the genetic aberrations promoting these differences, particularly the molecular differences between the two subtypes, remain unclear.

The therapeutic efficacy of molecular‐targeted drugs can be predicted using specific histological types or genotypes. However, some clinical trials have revealed the existence of important risk factors related to these drugs. Pemetrexed and bevacizumab, which are anti‐folate and anti‐vascular endothelial growth factor drugs, respectively, have therapeutic effects against NSCLC,3, 4 but may be related to adverse effects such as serious pulmonary hemorrhagic events, which appear to be more common among patients diagnosed with lung SCC.5, 6

As a result of large clinical studies, the National Comprehensive Cancer Network issued guidelines for NSCLC in 2010 and called for selecting regimens most effective for histological subtypes. The US Food and Drug Administration has similarly restricted the use of pemetrexed and bevacizumab only for patients with a history of non‐SCC. This highlights the need for precise animal models that mimic the events occurring in human lung cancer in the coming era of histology‐guided and gene‐targeted therapies.

Many chemical‐induced mouse models for lung cancer have been developed. Treatment with cigarette smoke carcinogens, such as nicotine, benzo[a]pyrene (BaP), and 4‐(methylnitrosamino)‐1‐(3‐pyridyl)‐1‐butanone (NNK), increase the incidence of lung tumors.7, 8, 9 However, most tumors in these models develop as relatively benign adenomas and only rarely as more aggressive adenocarcinoma. In chronic inhalation studies, Mauderly et al.10 showed that whole‐body exposure to cigarette smoke induces adenoma and adenocarcinoma in female rats. However, cigarette smoke is considered as a major causal agent of lung SCC, with little association to human adenocarcinoma. Although several highly reproducible models for lung adenoma and adenocarcinoma have been developed, there are few well‐established animal models for lung SCC. Although intratracheal instillation of BaP and charcoal powder have been reported to induce lung SCC in mice,11, 12 there have been few follow‐up reports. Eight months painting the skin with 25 μL 0.04 M N‐nitroso‐tris‐chloroethylurea (NTCU) induces lung SCC in female Swiss mice.13 However, this model is not applicable for evaluating the effects of chemicals on lung carcinogenesis because of high mortality and long treatment duration.

The present study aims to optimize NTCU treatment duration to achieve lower toxicity and shorten the treatment period compared with those of previous reports, in order to establish a new lung SCC mouse model for evaluating the effects of chemicals on lung SCC development. Moreover, to determine if this new mouse model replicates human lung SCC, we also evaluated molecular and ultrastructural changes during lung carcinogenesis.

Materials and Methods

Chemicals

N‐nitroso‐tris‐chloroethylurea (purity >90%, CAS No. 69113‐01‐5) was purchased from Toronto Research Chemicals (North York, Canada). All other reagents were of the highest quality available from Sigma‐Aldrich (St. Louis, MO, USA) or Wako Pure Chemical Industries (Osaka, Japan).

Experimental designs

All procedures were approved by the Institutional Animal Care and Use committee of Osaka City University Graduate School of Medicine (Osaka, Japan). Female A/J mice (5 weeks old) were purchased from Japan SLC (Hamamatsu, Shizuoka, Japan), housed in an animal room with a 22 ± 3°C targeted temperature, 55 ± 5% relative humidity, and a 12‐h light/dark cycle, with free access to a basal diet (CE‐2 pellets; Oriental Yeast, Tokyo, Japan) and tap water.

In experiment 1 (Fig. 1a), 60 mice were randomly distributed into four groups of 10–20 mice each. Before initial treatment, the dorsal skin was shaved. Twice weekly at 2–3‐day intervals, mice were painted topically on back skin using a micropipette with 75 μL 0.013 M NTCU for 2 (Group 1, n = 20), 4 (Group 2, n = 15), and 8 (Group 3, n = 15) weeks. Control mice (Group 4, n = 10) were painted with 75 μL acetone for 4 weeks, the NTCU solvent. Dose selection was based on previous reports13 and modified on the basis of the results of our own preliminary test. The reason for using female mice in the present study is that female A/J mice have higher sensitivity to lung chemical carcinogenesis than do male mice.14 Surviving mice were killed 18 weeks after initial NTCU treatment. After the mice were killed, the lung lobes were placed in 10% phosphate‐buffered formalin for histopathological and immunohistochemical analysis.

Figure 1.

Treatment protocols for experiments 1 (a), 2 (b), and 3 (c). Mice from each group (except control group) were painted topically on back skin with 75 μL 0.013 M N‐nitroso‐tris‐chloroethylurea (black bar). Control group was painted with 75 μL acetone (gray bar). S, mice killed.

In experiment 2 (Fig. 1b), 12 mice of the same sex and age as those in experiment 1 were randomly distributed into four groups of three mice each. They were treated with NTCU or acetone as above for 4 weeks and killed 6 (Group 1, n = 3), 12 (Group 2, n = 3), or 18 (Group 3 and control Group 4, n = 3 each) weeks after initial treatment. After the mice were killed, the left lung lobe was used for histopathological analysis, the right upper and middle lobes were used for immunoblotting, and the right lower lobe was used for ultrastructural analysis.

In experiment 3 (Fig. 1c), 12 mice of the same sex and age as those in experiments 1 and 2 were randomly distributed into two groups of six mice each. They were treated with NTCU or acetone as above for 1 week and killed 7 days after initial treatment. After the mice were killed, the left lung lobe was used for gene expression analysis.

Histopathological analysis

Lung tissue sections (4‐μm thick) were prepared from formalin‐fixed lungs, stained with H&E, and histologically examined using light microscopy. Lesions, including hyperplasia and metaplasia, dysplasia, and SCC, were diagnosed from the H&E stained sections. For quantitative image analysis, SCC and total lung areas of the maximum section of each lung were measured and quantified with an Image Processor for Analytical Pathology (Sumika Technos, Hyogo, Japan).

Immunohistochemistry

Expression of clara cell marker (CC10), SSC marker (cytokeratin 5/6), proliferating cell nuclear antigen (PCNA), and epidermal growth factor receptor (EGFR) was examined in mice lungs treated with NTCU for 4 weeks (Group 2). Anti‐CC10 goat polyclonal (dilution 1:1000, clone T‐18; Santa Cruz Biotechnology, Santa Cruz, CA, USA), anti‐cytokeratin 5/6 mouse monoclonal (dilution 1:200, clone D5/16B4; Millipore, Bedford, MA, USA), anti‐PCNA rabbit polyclonal (dilution 1:1000, clone FL‐26; Santa Cruz Biotechnology), and anti‐EGFR rabbit monoclonal (dilution 1:15, clone E114; Epitomics, Burlingame, CA, USA) IgGs were used as primary antibodies. After deparaffinization, lung sections were treated in a microwave oven with 10 mM citrate buffer (pH 6.0), sequentially with 3% H₂O₂, and incubated with specific antibodies at appropriate dilutions. The avidin–biotin complex method was used to determine protein expression (Vectastain ABC Elite Kit; Vector Laboratories, Burlingame, CA, USA). Tissue sections were lightly counterstained with hematoxylin.

Immunoblotting

Tissue samples from right upper and middle lung lobes were homogenized in ice‐cold CelLytic MT and protease inhibitor cocktail (Sigma‐Aldrich). Homogenates were centrifuged at 14 000g for 25 min at 4°C, and 30 μg pooled protein aliquots from three mice in each group were resolved in 10% polyacrylamide gels, transferred to nitrocellulose membranes blocked with blocking buffer comprising 5% non‐fat dried milk and 0.1% Tween 20 for 1 h at room temperature, and incubated with anti‐cytokeratin 5/6 for 1.5 h at room temperature and with anti‐EGFR overnight at 4°C with gentle shaking, followed by incubation with HRP‐conjugated secondary antibodies. Proteins were detected by standard chemiluminescence Western blotting protocol using ECL Plus Western Blotting Detection Reagents (GE Healthcare, Piscataway, NJ, USA). The HRP‐conjugated anti‐β‐actin antibody (clone AC‐15; Abcam, Cambridge, UK) was used as an internal control to confirm equal loading.

Transmission electron microscopy

Portions of right lower lung lobe were minced into 1‐mm³ pieces and fixed in phosphate buffer containing 2% glutaraldehyde and 2% paraformaldehyde for 1.5 h, treated twice with 7% sucrose in phosphate buffer for 7 min, and postfixed in 2% osmium tetroxide. Subsequently, samples were dehydrated through ascending concentrations of ethanol and propylene oxide, embedded in Durcupan (TAAB Laboratories, Aldermaston, UK) and polymerized at 60°C overnight. Ultrathin sections showing a pale gold interference color from selected blocks were cut with a diamond knife, mounted on grids, and stained with ethanol uranyl acetate and lead citrate for transmission electron microscopy (JEM‐1200 EX2; JEOL, Tokyo, Japan).

Real‐time quantitative PCR

Real‐time quantitative PCR was carried out essentially as described previously.15 Briefly, total lung RNAs were isolated using an Isogen RNA Isolation Kit (Wako Pure Chemical Industries) according to the manufacturer's instructions. cDNA synthesis was carried out using an Advantage RT‐for‐PCR kit (Takara Bio, Shiga, Japan). We examined expression levels of the following cytochrome P450 (CYP) genes involved in xenometabolism, including CYP1A1, CYP1A2, CYP2A4, CYP2B10, CYP2C29, CYP2E1, CYP2G1, CYP3A11, CYP4A10, CYP4A14, and CYP7B1. Sequence‐specific primers and probes from the TaqMan Gene Expression Assay were purchased from Life Technologies (Carlsbad, CA USA). β‐actin was used as an internal control.

Statistics

Statistical analysis was carried out using EXSAS (CAC, Tokyo, Japan). The incidence of lung SCC was analyzed by Fisher's exact test. Mean values for body weight and SCC/lung area among the NTCU‐treated groups were evaluated by Student's t‐test or the Wilcoxon test after analysis of homogeneity of variance by Bartlett's test.

Results

General observations of NTCU‐treated mice (experiment 1)

Survival rates and body weight curves are shown in Figure 2. The final survival rates were 100%, 87%, 47%, and 100% for Groups 1, 2, 3, and 4, respectively. Body weights of mice in Groups 1, 2, and 3 were clearly decreased compared with that of Group 4. There was significant loss (P < 0.05) of body weight in Group 3 compared with that of other treatment groups. These data indicate that NTCU treatment duration correlates with reductions in survival rate and body weight. Physical deconditioning may have occurred because of lung disorder with severe inflammation, characterized by abnormal breathing and histopathological changes such as epithelial degeneration and detachment. In addition, no tumorigenic effects of NTCU were observed in any organs other than the lung. No toxic lesions in the liver, kidney, or spleen were observed by histopathological examination.

Figure 2.

Survival rates (a) and body weights (b) of mice from each group (G1–G4) in experiment 1. Two mice in Group 2 were found dead at 11 and 16 weeks. Eight mice in Group 3 were found dead or moribund at 12, 14, 16, and 17 weeks. Mice from Group 3 revealed a significant decrease in body weight compared with Groups 1 and 2 from 7 to 18 weeks. n = 20, 15, 15, and 10 for Groups 1, 2, 3, and 4. 2W, 2 weeks; 4W, 4 weeks; 8W, 8 weeks. *Significant difference among the N‐nitroso‐tris‐chloroethylurea (NTCU) treated groups (P < 0.05 for each comparison).

Treatment duration‐dependent development of lung lesions in NTCU‐treated mice (experiment 1 and 2)

Diffuse pulmonary disorders were characteristically observed in a number of NTCU‐treated mice (Fig. 3a). Diagnostic criteria for lung lesions were based on those published by Renne et al.16 with minor modification. For hyperplasia and metaplasia (Fig. 3b), a single layer of normal bronchiolar epithelial cells became multiple layers, and normal columnar epithelium was replaced by flattened squamous epithelium. For dysplasia (Fig. 3c), atypical cells, including those with irregular cytoplasmic and nuclear morphologies and high nucleus/cytoplasm ratio, were present compared with those of hyperplasia and metaplasia, while maintaining parenchymal architecture. Irregular tumor nests comprising squamous cells with prominent intercellular bridges developed eventually and were diagnosed as SCC (Fig. 3d). Keratin pearl, multiple nuclei, and an increasing mitotic index were observed in well‐differentiated SCC. Tumor cell necrosis and foam cell aggregation were observed. Normal lung architecture was disrupted. The shapes of ciliated and clara cells in trachea were changed from columnar to cubic, but no hyperplastic lesions were observed. Because few dysplastic cells were observed in the bronchi, these abnormalities were considered to have originated in the terminal bronchiole due to initial direct damage to ciliated and clara cells.

Figure 3.

Photomicrographs of lung lesions induced by N‐nitroso‐tris‐chloroethylurea. (a) Lung lesions at low magnification. (b) Metaplasia. (c) Dysplasia. (d) Squamous cell carcinoma. Scale bar = 500 μm (A) and 50 μm (b–d).

Histopathological data for experiment 1 were based on the above criteria and analyzed by different methods: the mice survived until they were killed and those who survived until at least 14 weeks after initial treatment, the time‐point at which proliferative lesions were first observed. Incidences of hyperplasia and metaplasia, dysplasia, and SCC were: 90%, 70%, and 25% in Group 1; 92%, 85%, and 54% in Group 2; and 100%, 100%, and 71% in Group 3, respectively, among the mice that survived until termination (Table 1, upper panel). Among mice surviving until at least 14 weeks after initial treatment, incidences of hyperplasia and metaplasia, dysplasia, and SCC were: 90%, 70%, and 25% in Group 1; 86%, 79%, and 50% in Group 2; and 85%, 85%, and 69% in Group 3, respectively (Table 1, lower panel). These data indicated treatment duration‐dependent development of lung lesions with NTCU treatment. Incidence of SCC in Groups 2 and 3 increased significantly compared with that of Group 4 (P < 0.05). Quantitative image analysis showed that the lung/SCC percentage increased significantly (P < 0.05) and gradually with prolongation of the treatment period, and was >10‐fold higher in Group 3 than in Group 1 (Table 1). In experiment 2, damaged epithelial cells in Group 1, hyperplasia and metaplasia or dysplasia in Group 2, dysplasia or SCC in Group 3, and no lesion in Group 4 were seen on histopathological analysis, respectively (data not shown).

Table 1.

Histopathological examination of lung lesions in mice induced by N‐nitroso‐tris‐chloroethylurea (NTCU)

Immunohistochemistry of NTCU‐induced lung SCC (experiment 1)

Protein expressions of CC10, cytokeratin 5/6, EGFR, and PCNA in lung SCC of Group 2 mice are shown in Figure 4. CC10 was expressed only in non‐ciliated clara cells (Fig. 4a). Compared with normal regions, the number of CC10‐positive cells decreased in lung dysplasia and SCC areas. Cytokeratin 5/6 was expressed in SSCs (Fig. 4b) as well as in early lesions including dysplasia. Overexpression of EGFR was detected, particularly in well‐differentiated SCC (Fig. 4c), but not in hyperplasia, metaplasia, or dysplasia. The cell proliferation marker, PCNA, stained SCC nuclei (Fig. 4d); PCNA‐positive cells were markedly higher in SCC than in other lung lesions.

Figure 4.

Immunohistochemistry of lung squamous cell carcinoma induced by N‐nitroso‐tris‐chloroethylurea in experiment 1. Tissue sections of lung squamous cell carcinoma in Group 2 mice were immunostained with CC10 (a), cytokeratin 5/6 (b), epidermal growth factor receptor (c), and proliferating cell nuclear antigen (d) antibodies. The results of immunohistochemical analysis revealed a similar trend in staining in all animals. Scale bar = 200 μm.

Protein expression of lung with NTCU treatment (experiment 2)

Immunoblotting was carried out to further investigate changes in protein expression from early stages of the 4‐week NTCU treatment in mice (Fig. 5). High‐level expression of cytokeratin 5/6 was detected in mice in Groups 2 and 3, but no expression was detected in those in Group 4. Low‐level cytokeratin 5/6 expression was detected in Group 1 mice. Expression of EGFR was higher in Group 3 mice than in Group 2 mice. Low‐level EGFR expression was detected in Groups 1 and 4 mice. This result indicated a correlation between increased cytokeratin 5/6 and EGFR expression and development of SCC in lungs.

Figure 5.

Immunoblotting of lung treated with N‐nitroso‐tris‐chloroethylurea (NTCU) in experiment 2. Protein prepared from lung lobes of three mice in each group. Immunoblotting was carried out using cytokeratin 5/6 and epidermal growth factor receptor (EGFR) antibodies. Expression of β‐actin was compared as an internal control.

Lung ultrastructural changes with NTCU treatment (experiment 2)

Transmission electron microscopy was carried out to confirm ultrastructural changes in lungs of 4‐week NTCU treated mice between early and late stage. In Group 4 control mice, most clara cells were found in large numbers in terminal bronchiole, with large dense mitochondria (Fig. 6a). In contrast, there were increased intercellular spaces and an elevated number of abnormal cells in Group 1 mice (Fig. 6b). Damaged epithelial cells were stratified and lacked cytoplasmic organelles, with prominent decrease in mitochondria number compared with intact clara cells. Squamous cells with atypia were widespread in the mice of Groups 2 and 3 (Fig. 6c,d). Tumor cells had intercellular bridges, were rich in tonofilaments, and were connected to each other by desmosomes. A large number of desmosomes and tonofilaments were observed and the number increased in the Group 3 mice compared with other group mice.

Figure 6.

Transmission electron microscopy of lung lesions induced by N‐nitroso‐tris‐chloroethylurea (NTCU) in experiment 2. (a) Intact terminal bronchiole epithelium with clara cells (CL) including mitochondria (M) and secretory granule (S) of a control Group 4 mouse. (b) Widening of the intercellular space (IS) and cubic‐ and flat‐shaped cells damaged by NTCU treatment of a Group 1 mouse. (c,d) Squamous cells with atypia from a Groups 2 mouse and a Group 3 mouse. Desmosomes (arrows) and tonofilaments (arrowheads) are relatively well developed. Scale bar = 4 μm (a,b) and 0.4 μm (c,d).

Gene expression of lung with NTCU treatment (experiment 3)

Quantitative PCR was carried out to confirm the role of CYP genes in the metabolic activation of NTCU. CYP1A1, CYP2A4, CYP2B10, and CYP2E1 were expressed in the lungs, but there were no significant changes in expression level between NTCU treatment group and control group (data not shown). CYP1A2, CYP2C29, CYP2G1, CYP3A11, CYP4A10, CYP4A14, and CYP7B1 were not expressed in the lungs in either group.

Discussion

The origin of human lung SCC is considered to be related to injury of the bronchiole epithelium. Most cases are mediated by exposure to cigarette smoke, and SCC is more common in men than in women.17 Lung SCC is divided into central and peripheral types according to the primary site. Human lung SCC caused by cigarette smoking generally originates from the central portion of the lung including the bronchi. However, the clinicopathological characteristics of peripheral‐ type lung SCC remain unclarified and elucidation of epidemiological relationships is required urgently. Funai et al. indicated that peripheral‐type lung SCC is classified in humans as alveolar space‐filling, expanding, and combined types.18 Furthermore, the alveolar space‐filling type was confirmed to be an important clinical factor related to good prognoses.19

In the present study, short‐ to medium‐period NTCU treatments of 2, 4, and 8 weeks led to lung SCC, which commonly arose primarily from the peripheral portion of the lung. Severely changed squamous lesions, including SCC, expanded with extended NTCU treatment. Atypical cells, probably originating from terminal bronchiole epithelium, proliferated to fill the alveolar space and expanded to solid tumors. These results indicated that this new medium‐term carcinogenesis model for lung SCC in mice clinically resembled peripheral‐type human lung SCC.

Squamous differentiation is a multistage process that involves sequential expression of several specific genes.20, 21 Cytokeratin 5/6 and P63 are useful markers for differentiating SCC from adenocarcinoma and small‐cell carcinoma in human lung biopsies.22, 23 In experiment 2, the histopathological appearance of squamous cells in Group 2 mice coincided with cytokeratin 5/6 expression by immunoblotting. These results showed that cytokeratin 5/6 was expressed from early squamous differentiation stages, not only in lung SCC.

Epidermal growth factor receptor is critical to the control of cellular proliferation, differentiation, and survival.24 Abnormalities in EGF–EGFR signaling are found in a broad range of cancers including lung, breast, and colon.25, 26, 27 Mutation and amplification of EGFR are associated with EGFR–tyrosine kinase inhibitor responses, which are molecular‐targeted drugs for NSCLC, particularly adenocarcinoma.28, 29 However, it was not possible to predict EGFR–tyrosine kinase inhibitor sensitivity by levels of EGFR overexpression. On the contrary, EGFR expression in early stage SCC is strongly associated with hazard risk for metastasis in lung SCC patients.30 Immunohistochemistry and immunoblotting indicated increased EGFR expression in experiments 1 and 2, suggesting NTCU‐induced SCC is similar to human SCC in EGFR expression, and NTCU may alter functional EGFR expression in A/J mice lungs. Although there were discrepancies in EGFR expression between the results of immunohistochemical analysis and immunoblotting, this was considered to cause lower sensitivity of anti‐EGFR antibody for immunohistochemistry, so that we could not detect early changes in EGFR expression in dysplastic cells. One could speculate that the fixation procedure directly affects the preservation of EGFR epitope.

A modulator of EGFR expression, SOX2, was recently reported to be a key upregulated transcriptional factor in human lung SCC.31 In addition to EGFR interaction, SOX2 correlates with the squamous markers p63 and cytokeratin 6A,32 indicating that EGFR has a different role in modifying lung cancers, including adenocarcinoma and SCC. It is interesting to note that NFE2L2 mutations, which are exclusively related to EGFR mutations and may play a role in tumor prognosis for human lung SCC, tend to be more frequent among patients with advanced stage disease.33 However, information about the relationship between SOX2 and NFE2L2 is presently unavailable. In our next study, we plan to analyze the expression of SOX2 and NFE2L2 in our model, and elucidate their relationships with EGFR function.

Past studies on injury and vitamin A deficiency have suggested that clara cells are pivotal for regeneration of terminal bronchiole epithelium.34, 35, 36 Cells formed with inadequately developed secretory‐cell features were reported in ultrastructural studies of a chemically induced hamster model.37 In this study, we observed damaged cubic and flat epithelial (presumed clara) cells with degenerative mitochondria at an early stage. Furthermore, because tumor cells did not have small granules, they may be transformed from clara cells. Decreased expression of desmosomal proteins is associated with poor prognosis, and is particularly related to increased metastasis.38 On the contrary, an increased number of desmosomes and lack of metastasis in this study might be connected with a good prognosis in human peripheral‐type lung SCC. Our mouse model provides a new approach for evaluating ultrastructural changes in early stage development of SCC and may also facilitate the identification of SCC origin.

Our study differs from previous NTCU treatment studies. First, we could induce lung SCC approximately 50% earlier than previous NTCU studies.13, 39, 40 Major differences in NTCU treatment protocols between the present and previous studies were the liquid volumes (75 μL and 25 μL, respectively) and concentrations (0.013 M and 0.04 M, respectively), although 1 μmol NTCU was given. We hypothesize that skin permeability is increased because of the one‐third concentration and three‐fold volume of NTCU. Second, half of SCCs arose from the central portion of the lung with an 8‐month NTCU induction study in mice.39 In contrast, in our mouse model, most SCCs arose from the peripheral portion. It is known that the poor clearance of inhaled carcinogens reach the peripheral portion due to lack the defenses present in the central portion, such as ciliated and mucous cells.41 The carcinogenic effect of NTCU might be exerted under the same conditions. In experiment 3, we could not confirm the expression changes of CYP genes of lungs treated with NTCU. Lack of significant changes in any of the genes investigated can be explained at least in part by the short treatment duration. To understand the metabolism of NTCU, further studies to identify NTCU‐responsive genes by extension of the treatment period are necessary.

It has been indicated that NNK may act as a lung cancer initiator and promoter.42 Activation of the PI3K/Akt and ERK pathways are related to the survival and growth of NNK‐induced lung cancer cells.43, 44 The promotional effect of NTCU in in vitro culture studies using lung SCC and adenocarcinoma cell lines was not confirmed by us (e.g., Tago Y, 2013, unpublished data). Wang et al.39 carried out a whole‐genome linkage disequilibrium analysis and showed that NTCU was implicated in cell proliferation and transformation. Based on these results, we suspect that NTCU is a strong initiating carcinogen. Chemical‐induced lung cancer models provide invaluable insights for discovering new compounds with promotional or suppressive effects.45, 46, 47, 48 Khan et al.49 showed that pomegranate fruit extract inhibits NTCU‐induced mouse lung SCC, and indicated that pomegranate fruit extract reduced expression of cell proliferation and angiogenesis markers such as MAPKs, nuclear factor‐κB, PI3K/Akt, and mTOR. However, continuous NTCU treatment for 8 months complicated the search. Our new method is appropriate for investigating novel compounds with promotional or suppressive effects toward lung SCC because of the high survival rate and reduced research time.

In conclusion, we have established a new mouse model for lung SCC that dramatically reduces the induction time compared with previous reports. Considering the carcinogenesis model, 2‐week and 4‐week NTCU treatment may be suitable for investigating new chemicals with promotional or suppressive effects toward lung SCC. Our findings indicate that the treatment duration‐dependent increase in lung SCC is related to the upregulation of cytokeratin 5/6 and EGFR expression. Further studies are required to elucidate the carcinogenic mechanism of NTCU‐induced lung SCC.

Disclosure Statement

The authors have no conflict of interest.

(Cancer Sci 2013; 104: 1560–1566)