Development of pulmonary bronchiolo-alveolar adenocarcinomas in transgenic mice overexpressing murine c- myc and epidermal growth factor in alveolar type II pneumocytes

Transgenic mouse models were established to study tumorigenesis of bronchiolo-alveolar adenocarcinomas derived from alveolar type II pneumocytes (AT-II cells). Transgenic lines expressing the murine oncogene c- myc under the control of the lung-specific surfactant protein C promoter developed multifocal bronchiolo-alveolar hyperplasias, adenomas and carcinomas respectively, whereas transgenic lines expressing a secretable form of the epidermal growth factor (IgEGF), a structural and functional homologue of transforming growth factor α (TGFα), developed hyperplasias of the alveolar epithelium. Since the oncogenes c- myc and TGFα are frequently overexpressed in human lung bronchiolo-alveolar adenocarcinomas, these mouse lines are useful as models for human lung bronchiolo-alveolar adenocarcinomas. The average life expectancies of hemizygous and homozygous c- myc transgenics were 14.25 months and 9.2 months, respectively, suggesting that a dosage effect of c- myc caused an accelerated bronchiolo-alveolar adenocarcinoma formation. First analyses of double transgenics, hemizygous for both c- myc and IgEGF, show that these mice develop bronchiolo-alveolar adenocarcinomas at the average age of 9 months, indicating that these oncogenes cooperate during the lung cancer formation. Our results demonstrate that c- myc and EGF are directly involved and cooperate with one another during formation of bronchiolo-alveolar adenocarcinomas in the lung. © 2001 Cancer Research Campaign http://www.bjcancer.com

genes are appropriate candidates for use in the construction of lung specific gene constructs.
Several proto-oncogenes, including c-myc and the transforming growth factor α (TGFα) as well as its homologue, epidermal growth factor (EGF), are frequently found to be overexpressed in human pulmonary carcinoids and adenocarcinomas (Battista et al, 1993;Broers et al, 1993;Lorenz et al, 1994;Moody, 1996), suggesting that they may be directly involved in lung carcinoma formation. c-myc is a member of a group of regulatory proteins which are involved in controlling cell cycle entry, progression and differentiation (reviewed in Facchini and Penn, 1998).
The EGF family includes EGF, TGFα and heparin-binding EGF. Both EGF and TGFα bind to and activate the EGF-receptor (EGFR) (Yeh and Yeh, 1989). Bronchiolo-alveolar adenocarcinomas often show constitutive overexpression of EGFR as well as of TGFα, which indicates that the resulting autocrine loop promotes loss of cell cycle control (Tateishi et al, 1990). The oncogenic potential of these growth factors is supported by the observation that overexpression of TGFα or EGF in the liver of different transgenic mouse strains cause hepatocellular carcinomas (Sandgren et al, 1993;Tönjes et al, 1995).
In this work we used the AT-II cell specific SP-C promoter to generate transgenic mouse lines constitutively overexpressing the oncogene c-myc and a secretable form of EGF (IgEGF) (Tönjes et al, 1995) in the lung. Transgenics expressing c-myc developed multifocal bronchiolo-alveolar adenomas and carcinomas respectively, those expressing IgEGF developed multifocal alveolar hyperplasias. Cooperation in lung tumour formation of both transgenes was demonstrated in IgEGF/myc double transgenic mouse lines. The established transgenics will provide useful animal models to test targeted gene therapy protocols, in which the expression of potentially cytotoxic gene products can be targeted to cancer cells by the SP-C promoter.

Cloning procedures and production of transgenic mouse lines
The ApaI-HindIII mouse c-myc DNA fragment from the plasmid pTG2948 (Dalemans et al, 1990) was subcloned into the corresponding restriction sites in pBSKS II (+/-) (Stratagene). ApaI was converted into a SalI restriction site. The 2.7 kb SalI/EcoRI c-myc DNA fragment was ligated to the SalI/EcoRI site of the vector pUC18/3.7SP-C downstream of the human SP-C promoter 5′-flanking region . The BamHI site of the BamHI-SalI IgEGF fragment (nucleotides 0 to 360, including the Ig signal sequence and a synthetic EGF gene) derived from the plasmid alb-DS4 (Tönjes et al, 1995) was converted to a SalI restriction site. The new IgEGF SalI fragment was cloned into the SalI restriction site 3′ of the promoter of the human SP-C gene of pUC18/3.7SP-C. Both gene constructs were cleaved with NdeI and NotI and the fusion gene fragments were purified by the Qiagen gel extraction kit and microinjected into male pronuclei of fertilized oocytes from hybrid CD2/F1 (DBA/2 × Balb/C) mice (Hogan et al, 1994). Viable oocytes were transferred into the oviduct of pseudopregnant CD2F1 recipient mice. Transgenic founder mice were mated with CD2F1 for propagation as hemizygous transgenics.

Southern and Northern analysis
Transgenic mouse lines were identified by Southern analysis of DNA extracted from biopsied mouse tails (Hogan et al, 1994). Restricted DNA was separated through 0.8% agarose and transferred to nylon membrane (Amersham Life Sciences) according to standard protocols. Hybridization was performed in Church buffer (0.25 M NaHPO 4 , 7.0% SDS, 10 mM EDTA, pH 7.2) at 65˚C with the randomly labelled transgene.
Total RNA from various tissues was isolated by the Qiagen RNA extraction kit after homogenization using a Polytron homogenizer and blotted according to standard protocols.

Histopathology
Tissues were fixed in 4% paraformaldehyde in PBS for approximately 20 h, dehydrated and embedded in paraffin (Roti®-Plast, Roth). Tissue sections were stained with haematoxylin & eosin according to standard protocols. The mouse tumours were classified according to the International Agency for Research on Cancer (IARC) - WHO (2000).

Generation of transgenic mouse lines and their phenotypes
The gene constructs SP-C/myc and SP-C/IgEGF ( Figure 1A, B) consisted of the murine c-myc gene and a secretable form of EGF (IgEGF), whose expression were controlled by the human SP-C promoter. One SP-C/IgEGF and 5 SP-C/myc founder mice were identified by the generation of diagnostic fragments upon restriction enzyme digestion of mouse tail DNA and subsequent Southern analysis as shown representative for SPC/myc transgenic mice in Figure 3. Transgenic mouse lines were established from the SP-C/IgEGF and two SP-C/myc transgenic founder mice. All other founder mice were not germ line transgenic and did not transfer the transgene to their descendants. 2 of the SP-C/myc founder mice showed hyperplasias in the lung alveolar epithelium (not shown). In contrast, the founder SP-C/myc 3.2 as well as all established transgenic mouse lines, e.g SP-C/myc 8.2 and 13.0, developed multifocal pulmonary bronchiolo-alveolar adenocarcinomas originating from the alveolar epithelium. Littermates of the SP-C/IgEGF transgenic mouse line 6.2 showed no bronchioloalveolar adenocarcinomas, but they developed hyperplasias derived from the alveolar epithelium. The observed phenotypes of all founder mice and their offspring are summarized in Table 1. Transgene expression could be detected in the lung from all shown founder animals and the copy number of the transgene c-myc was 1-2 copies for the established transgenic lines SPC/myc 8.2 and 13.0 and 2-3 copies for the founder animals SPC/myc 3.2 and SPC/myc 13.0 ( Table 1). The following work is focused on the transgenic mouse lines SP-C/myc 8.2 and SP-C/IgEGF 6.2.
to Northern analysis. c-myc-and IgEGF-specific mRNAs were detected only in the lung of both transgenic mice ( Figure 1C, D) and not in any other tissue including salivary gland, liver, pancreas and ovary (not shown). Non-transgenic mice showed no signal in the lung for both transgenes, respectively ( Figure 1C, D).

Development of hyperplasias in SP-C/IgEGF transgenics and development of bronchiolo-alveolar adenocarcinomas in SP-C/myc transgenics
Expression of the SP-C/IgEGF transgene induced the development of alveolar hyperplasias in the alveolar epithelium ( Figure 2A) when compared to non-transgenic mice ( Figure 2B). Alveolar hyperplasias in analysed SPC/IgEGF individuals occurred at the average of 19 months. In SP-C/myc transgenics different stages of tumour development in the alveoli were frequently observed. Large bronchiolo-alveolar adenocarcinoma developed only in the lung of SP-C/myc transgenics. Early stages of tumour development were characterized by multifocal hyperplasias originating in the alveolar epithelium ( Figure 2C). Adenomas, which developed in the alveolar septae were observed in lung sections of SP-C/myc transgenics at the age of 6-7 months ( Figure 2D). Advanced stages of carcinogenesis consisted of multifocal bronchiolo-alveolar adenocarcinomas ( Figure 2E) were detected in SP-C/myc transgenics at the average age of 14.25 months, whereas the bronchiolar epithelium was not affected. Figure 2F demonstrates a lung of a non-transgenic and of a SP-C/myc transgenic mouse, both of 14 months of age. One lobe of the lungs was completely transformed to a bronchiolo-alveolar adenocarcinoma.

Generation of homozygous SP-C/myc transgenics and hemizygous double transgenic mice expressing c-myc and IgEGF
The medial survival times of hemizygous SP-C/myc transgenics is 14.25 months (Table 2), whereas the medial age of death of homozygous SP-C/myc transgenic is 9.2 months (Table 2). At the age of 14.25 months and 9.2 months respectively, 75% of all hemizygous and 80% of homozygous mice were diagnosed with bronchiolo-alveolar adenocarcinomas transforming both lung lobes (Table 2 and Figure 2F). These findings suggest that a gene dosage effect of c-myc expression contributed to the accelerated tumor development as compared to hemizygous transgenics. Hemizygous and homozygous mice were distinguished by Southern analysis (Figure 3). A summary of homozygous and hemizygous SP-C/myc transgenics, their phenotype, and their life expectancies are shown in Table 2. These results demonstrated that c-myc overexpression was causally involved in bronchioloalveolar adenocarcinoma formation. A gene dosage effect was also observed in another transgenic mouse, who expressed SV40 Tag under the control of the fetal globin promoter. In this transgenic mouse strain prostate tumours were induced in 75% of the male hemizygous for the transgene but in 100% of all male homozygous mice (Perez-Stable et al, 1997). ( n o n t r a n s g e n ic ) lu n g S P -C / m y c 8 . 2 lu n g ( n o n t r a n s g e n ic ) lu n g S P -C / I g E G  To generate hemizygous double transgenics expressing c-myc and IgEGF, offspring of transgenic mouse lines SP-C/myc 8.2 and SP-C/IgEGF 6.2 were cross-bred. Littermates were analysed for the presence of both transgenes by Southern analyses. The life expectancy of SP-C/myc/IgEGF double transgenic individuals analysed so far was 9 months, which was clearly reduced compared to the medial survival times of hemizygous SP-C/myc or SP-C/IgEGF transgenics, i.e. 14.25 and 19 months, respectively ( Table 2). 100% of examined SP-C/myc/IgEGF double transgenics were diagnosed with bronchiolo-alveolar adenocarcinomas ( Table  2). From these results we conclude, that c-myc and IgEGF cooperated during lung tumour formation. Histological analysis confirmed that lung tumours in SP-C/myc/IgEGF double transgenics originated from AT-II cells and thus were classified as bronchiolo-alveolar adenocarcinomas (not shown). The number of lesions in the 4 examined double transgenics is macroscopically lower but the lesion size is enlarged in comparison to SPC/myc transgenics at the age of 9 months. We speculate that additional randomly occurring genetic changes in each lesion have to take place for tumour induction.

Gene expression in lung tumours of transgenic mice
Expression of one transgene in a target tissue is usually not sufficient for tumour development. Therefore we analysed tumours for abnormal expression of selected proto-oncogenes. We also checked expression patterns of genes which are known to be typically expressed in AT-II cells, in order to obtain information about the extent of dedifferentiation of the tumour cells as compared with AT-II cells from which they were derived. For this purpose we used the reverse Northern slot blot hybridization technique. Lungs of non-transgenic CD2F1 littermates (14 months old), from tumour nodules of SP-C/myc transgenics (14 months old) and from tumour nodules of SP-C/myc/IgEGF double transgenics (9 months old) were investigated. The intensity of mRNA expression of three surfactant proteins, which are known to be expressed in AT-II cells, differed moderately among the analysed tissues in comparison to those of non-transgenic littermates (Figure 4). The expression levels also differed among tissues from mice expressing c-myc or both, c-myc and IgEGF, which may indicate various stages of dedifferentiation of the lung tumour tissue.
To analyse the expression levels of genes involved in regulating cell proliferation we analysed the expression of selected cell cycle regulating genes including cyclin D1, cdc2 and c-jun. The  expression of these genes was distinctly increased in tumour nodules of SP-C/myc/IgEGF double transgenics, but not in tumours of SP-C/myc transgenics or in normal lung tissue ( Figure  4). These preliminary results indicated increasing deregulation of the cell cycle at various stages of lung tumour development in the SP-C/myc/IgEGF double transgenics. It can be speculated that the deregulation of cell cycle regulating genes might be the reason for the decreased medial survival times in these mice (Figure 4).

DISCUSSION
Constitutive overexpression of c-myc under the transcriptional control of the SP-C promoter is frequently associated with the development of bronchiolo-alveolar adenocarcinomas, adenomas or hyperplasias in transgenic mice. Hemizygous SPC/myc and SPC/IgEGF and homozygous SPC/myc transgenic mice examined in this study had a life span of between 9 and 14.25 months. Not all analysed transgenic mice developed bronchiolo-alveolar adenocarcinomas. We speculate that additional genetic changes have to occur for tumour induction, e.g. knock out of tumour suppressor and/or activation of proto-oncogenes. These events occur randomly and may explain that not all offspring develop tumours. However, death inducing spontaneous bronchiolo-alveolar adenocarcinomas are uncommon at this age, but it should considered, that spontaneous lung tumours are not a rare event in mice. Reported data are related to the age 24 month, are not specified for bronchiolo-alveolar entities and are not available for the hybrid strain CD2F1 (compare overview in Rittinghausen et al, 1997). It should be emphasized, that nontransgenic control mice of the breed and age used for transgenic studies, did not display any lung tumours. Therefore, it is evident, that the bronchiolo-alveolar neoplasias or hyperplasias were indeed caused by the overexpression of the c-myc transgene. The role of c-myc overexpression as a first step in the process of tumour formation was further confirmed by the gene dosage effect observed in homozygous transgenics, which showed accelerated tumour development in the lung. These findings support the hypothesis that this gene is causally involved in the development of human alveolar lung bronchiolo-alveolar adenocarcinomas, where overexpression of c-myc is frequently observed (Broers et al, 1993;Lorenz et al, 1994). Overexpression of IgEGF under the control of the SP-C promoter led to the formation of hyperplasias of the alveolar epithelium in the lung of transgenic mice, whereas overexpression of TGFα in the lung of transgenic mice has been shown to induce enlarged parenchymal airspace and pulmonary fibrosis (Hardie et al, 1997). The induction of different phenotypes by IgEGF and TGFα might be due to the fact that EGF -but not TGFα -binds to other receptor subunits of the EGF receptor family; e.g. erbB2, 3 and/or 4 (Alimandi et al, 1997;Wang et al, 1998). A similar cooperation of c-myc and IgEGF, which led to accelerated bronchiolo-alveolar adenocarcinomas formation in SP-C/myc/IgEGF double transgenics was also demonstrated for hepatocarcinogenesis in transgenic mouse lines, which overexpress these oncogenes in hepatocytes (Tönjes et al, 1995).
First results indicate that other genes may be involved in the accelerated growth of tumors in SP-C/myc/IgEGF double transgenics ( Figure 4). The expression level of cyclin D1 was shown to be strongly increased in tumour nodules of SP-C/myc/IgEGF double transgenic mice but not in lungs of non-transgenics or SP-C/myc transgenics. It is known that EGF induces cyclin D1 Figure 3 Southern analysis of homozygous and hemizygous SP-C/myc 8.2 transgenic mice. DNA, isolated from biopsied tails of SP-C/myc transgenic mice, was digested with BamHI. A diagnostic 3.5 kb transgene specific fragment was detected in the DNA of all transgenic mice but not in non transgenic mice. An additional 6.0 kb BamHI fragment represents the endogenous c-myc DNA fragment, which was found in transgenic and non transgenic mice. Homozygous transgenic mice are characterized by a stronger signal of the transgene specific c-myc DNA fragment in comparison to the endogenous c-myc DNA fragment. myc/0, hemizygous SP-C/myc transgenic mouse; myc/myc, homozygous SP-C/myc transgenic mouse; 0/0, non transgenic mouse Gene expression profiles in the lungs of non transgenic littermates, in SP-C/myc transgenic and in SP-C/myc/IgEGF double transgenic mice. A subset of plasmids containing c-DNA sequences of the indicated genes were denatured and immobilized on a nylon membrane. The filters were hybridized with a [ 32 P]-labeled cDNA, which was generated by reverse transcription of poly A+ mRNA from lung tissue of a non transgenic littermate, from a tumor nodule of the lung of a SP-C/myc transgenic and from a tumor nodule of the lung of a SP-C/myc/IgEGF double transgenic mouse. cycDI, cyclin DI expression (Ravitz et al, 1996;Ramljak et al, 1998), which is one of the most frequently overexpressed oncogenes in human bronchiolo-alveolar adenocarcinomas (Marchetti et al, 1998). Also, cdc2 and c-jun were overexpressed in lung tumours of SP-C/myc/IgEGF double transgenics. cdc2 is an important cell cycle controlling gene, which binds to and activates cyclin B1. Upregulation of c-jun was also observed in human cell lines established from NSCLC when stimulated by growth factors (Szabo et al, 1996). These observations indicate that the tumours in the transgenic mice are excellent models for human lung adenocarcinomas, which will be useful for understanding the molecular basis for the development of human lung cancer. Future experiments, involving gene expression profiles of developing tumours, that include a broader spectrum of tumour suppressor and oncogenes, will provide a more detailed view, which genes become involved during tumour progression in developing lung carcinomas in SP-C/myc as well as in SP-C/myc/IgEGF double transgenic mice. The surfactant proteins SP-A, SP-B and SP-C were expressed moderately reduced or at similar levels in tumours of SP-C/myc and SP-C/myc/IgEGF transgenics as compared to lungs of nontransgenics indicating that bronchiolo-alveolar adenocarcinomas were derived from AT-II cells. Expression of SP-C was also shown to occur in human bronchiolo-alveolar adenocarcinomas  suggesting similarities between human bronchiolo-alveolar adenocarcinomas and the homologous tumour type in SP-C/myc transgenics. In summary we present a new model for bronchiolo-alveolar adenocarcinomas, which will be useful to address several questions about lung tumour formation.