Effects of Mutant Human Ki-ras^G12C Gene Dosage on Murine Lung Tumorigenesis and Signaling to its Downstream Effectors

Abstract

Studies in cell culture have suggested that the level of RAS expression can influence the transformation of cells and the signaling pathways stimulated by mutant RAS expression. However, the levels of RAS expression in vivo appear to be subject to feedback regulation, limiting the total amount of RAS protein that can be expressed. We utilized a bitransgenic mouse lung tumor model that expressed the human Ki-ras^G12C allele in a tetracycline-inducible, lung-specific manner. Treatment for 12 months with 500 μg/ml of doxycycline (DOX) allowed for maximal expression of the human Ki-ras^G12C allele in the lung, and resulted in the development of focal hyperplasia and adenomas. We determined if different levels of mutant RAS expression would influence the phenotype of the lung lesions. Treatment with 25, 100 and 500 μg/ml of DOX resulted in dose-dependent increases in transgene expression and tumor multiplicity. Microscopic analysis of the lungs of mice treated with the 25 μg/ml dose of DOX revealed infrequent foci of hyperplasia, whereas mice treated with the 100 and 500 μg/ml doses exhibited numerous hyperplastic foci and also adenomas. Immunohistochemical and RNA analysis of the downstream effector pathways demonstrated that different levels of mutant RAS transgene expression resulted in differences in the expression and/or phosphorylation of specific signaling molecules. Our results suggest that the molecular alterations driving tumorigenesis may differ at different levels of mutant Ki-ras^G12C expression, and this should be taken into consideration when inducible transgene systems are utilized to promote tumorigenesis in mouse models.

Introduction

Lung cancer is the leading cause of cancer-related deaths in both men and women in the United States. To date, less than 15% of all lung cancer patients survive >5 years (Jemal et al., 2007). Non-small cell lung cancer (NSCLC), which occurs in ∼75-80% of lung cancer patients, is the most common histological type of lung cancer with its subtype, adenocarcinoma (AC), being the predominant form (Jemal et al., 2003). Activating mutations of Ki-ras occur in 30-50% of AC (Rodenhuis, 1992;Reynolds et al., 1992;Mitsudomi et al., 1991;Slebos et al., 1991;Mills et al., 1995). Thus, a better understanding of the role of Ki-ras in lung tumorigenesis could provide insight into the genetics, prevention, and ultimately development of novel therapies for treating this condition.

Mutations in Ki-ras have been shown to be an early event in both human and mouse lung tumors (Leone-Kabler et al., 1997;Li et al., 1994;Westra et al., 1993), although clearly it is not the sole cause of cancer development. Tumorigenesis occurs following multiple genetic lesions to genes involved in cell growth, differentiation, survival, immune surveillance, metastasis, and invasion (Hanahan et al., 2000), resulting from the increased expression of oncogenes and decreased expression of tumor suppressor genes. Epidemiological studies have shown NSCLC specimens contain deletions or mutations in p53 (∼50%) (Hainaut et al., 1997), deletion or reduced expression of p16^INK4a (∼60%) (Serrano, 1997), and deletion or reduced expression of pRb (∼30%) (Malkinson, 1998). Similar mutations are observed in murine lung tumors which exhibit deletion or reduced expression of p16^INK4a in approximately 50% of the tumors and reduced expression of Rb (Tuveson et al., 1999), although p53 is rarely affected in murine lung tumors, with only a few reports occurring in late-stage tumors (Horio et al., 1996;Miller, 1999).

In addition to qualitative alterations in gene structure and function resulting from mutations and deletions at key regulatory loci, quantitative changes such as amplification or overexpression of critical genes, such as ras, are equally important in the neoplastic process and are prevalent in a variety of human tumors (Pulciani et al., 1985;Alitalo, 1984;Varmus, 1984;Mo et al., 2007). Cell culture studies with NIH 3T3 cells have demonstrated that morphologic transformation of the cells occurred only when large amounts (≥10μg) of recombinant DNA molecules carrying the normal human H-ras-1 proto-oncogene were transfected into the cells (Pulciani et al., 1985). Similar studies addressed the levels of H-ras^wt on transformation by comparing the phenotype of cells expressing highly elevated levels of Ha-ras^wt with cells expressing more modest levels of Ha-ras^Val12 on rat cells (Ricketts et al., 1988). This work showed that markedly elevated levels of Ha-ras^wt did not form colonies, but varied levels or doses of H-ras^Val-12 resulted in the formation of a wide-ranging number of colonies depending on the gene dosage of the mutant RAS gene. Mo et al. (Mo et al., 2007) also demonstrated, utilizing a mouse bladder tumor model, that doubling the gene dosage of activated Ha-ras triggered early on-set, rapidly growing, and fully penetrant urothelial tumors throughout the urinary tract. Low level expression of constitutively active Ha-ras was insufficient in initiating urothelial development, even with deletion of the Ink4a/Arf gene locus.

Activation of Ras leads to the sequential activation of Raf, MEK, p42, and p44 MAP-ERK kinases (Finney et al., 1993;Howe et al., 1992). Studies in NIH 3T3 cells have shown that Raf-activated pathways are capable of eliciting distinctly different biological outcomes depending on the strength of the signal (Woods et al., 1997;Sewing et al., 1997). Low levels of Raf activity led to activation of cyclin D1-cdk4 and cyclin E-cdk2 complexes and to cell cycle progression, whereas higher activity levels elicited cell cycle arrest correlating with induction of p21^Cip1 and inhibition of cyclin-cdk activity. This data demonstrated the importance of evaluating not only structural but also quantitative aspects of oncogene activation when evaluating the tumorigenic potential of critical genes.

In a previous study focusing on the early events in lung AC formation, our laboratory developed and characterized a bitransgenic mouse lung tumor model that conditionally expresses the human Ki-ras^G12C allele in a doxycycline (DOX)-inducible, lung-specific manner (Floyd et al., 2005). This system makes use of the inducible tet-on system (Shockett et al., 1996). One strain of transgenic mice contain a reverse tetracycline (tet) trans-activator (rtTA) protein linked to the Clara cell secretory protein (CCSP) promoter and constitutively express the rtTA gene product specifically in the lungs (Hayashida et al., 2000). The second line of transgenic mice contain the cDNA of Ki-ras^G12C cloned into the tetO₇-CMV plasmid, placing the mutant human Ki-ras^G12C transgene downstream of a tet-inducible promoter (Floyd et al., 2005). When these two transgenic mouse strains are crossed, lung specific expression of the rtTA gene product occurs in the presence of the tet0₇-CMV-Ki-ras transgene. In the absence of DOX, the rtTA gene product was unable to recognize the tetO sequence and was thus unable to stimulate transcription. Treatment of the bitransgenic mice with DOX allowed binding of the rtTA protein to the tetO enhancer, resulting in activation of the CMV promoter and transcription of the Ki-ras gene specifically in the lung (Floyd et al., 2005). This model appears to recapitulate the earliest stages of lung tumorigenesis, as mice treated with DOX (500 μg/ml) for 12 months develop benign foci of hyperplasia and adenomas, showing little progression to adenocarcinomas at a dose of DOX inducing maximal transgene expression (500 μg/ml in the drinking water). Examination of the lung tissue by both quantitative real-time PCR and immunohistochemistry (IHC) demonstrated alterations in the expression and/or phosphorylation of cell cycle regulatory and signal transduction genes in the tumor tissues. These alterations included elevated levels of p19^ARF, survivin, and cyclin D1, decreased expression of pRb, elevated RAL-GDS activity (as assessed by pull down assays), and increased phosphorylation of erk and its downstream effectors, p38 and its downstream effectors, and p53. Our data suggested that the Ki-ras^G12C allele activates a specific subset of downstream effector pathways and that the combination of alterations in both proliferative and anti-proliferative signaling pathways determined the benign phenotype of the lung tumors in the Ki-ras^G12C expressing mice (Floyd et al., 2005;2006).

To analyze the effects of RAS gene dosage on lung tumor formation in an in vivo setting, we utilized the Ki-ras^G12C expressing mice and varied the levels of mutant RAS transgene expression by administering different concentrations of DOX in the drinking water. Our studies suggest that the molecular alterations driving tumorigenesis may differ at different levels of mutant RAS expression. These results, combined with earlier studies on the effects of different RAS mutations on the tumorigenic potential of the lesion, suggest that the type of mutation and the level of expression of the RAS gene determine which signaling pathways are activated or repressed during the neoplastic process and thus play a role in determining the pathological phenotype and severity of the tumor.

Materials and Methods

Animal treatment and histological analysis of lungs

The CCSP-rtTA/Ki-ras^G12C bitransgenic mice, as well as Ki-ras^G12C monotransgenic controls, were either untreated or treated with 25 or 100 μg/ml of DOX in the drinking water, beginning 8 weeks after birth, for 12 months. Tumor multiplicity and staging data obtained from mice treated with 25 or 100 μg/ml of DOX were compared to the multiplicity and staging data obtained previously from studies utilizing 500 μg/ml of DOX (Floyd et al., 2005). The mice were euthanized by CO₂ asphyxiation and cervical dislocation. Following intratracheal cannulation, the thorax was opened and the lungs were removed en bloc and carefully examined for the presence of pulmonary masses with the aid of a dissecting microscope. All macroscopic pulmonary lesions were recorded. Because of the small size of the lung tumors, several tumors from the same animal were pooled and isolated from areas of the lung containing several small tumors that consisted primarily of tumor tissue, though some normal tissue was included as well. The remainder of the lung was processed for histopathology and IHC by fixation for 24 hr in 4% chilled paraformaldehyde fixative. Following fixation, the tissue was transferred and stored in 70% ethanol until the lungs were embedded in paraffin and prepared for routine microtomy (cut at 4 microns) and hematoxylin and eosin staining. The sections were examined by an ACVP Board certified veterinary pathologist, and all proliferative lesions examined were classified with respect to standard murine pulmonary tumor characteristics (Nikitin et al., 2004).

FISH Analysis

Cells derived from lung tissue of Ki-ras^G12C monotransgenic mice were grown in Amniomax culture medium (Gibco, Grand Isle NY) and harvested for metaphase spreads using standard tissue culture techniques. Cells were treated with 0.5 μg of Colcemid (Gibco) for 20 min and re-suspended in 0.075 M KCl for 8 min at 37°C, followed by 3:1 methanol:acetic acid. Slides were prepared by air-drying and were pretreated with 0.001% pepsin (Sigma, St. Louis MO) in 0.01 N HCl for 10 min., followed by fixation in 1% formaldehyde and dehydration through an ethanol series. The 1.4 kb Ki-ras probe was directly labeled with the Vysis Nick Translation kit (Downer's Grove, IL) in Spectrum Green according to the manufacturers' protocol. The labeled Ki-ras probe (200 ng) was combined with blocking probes, precipitated, and re-suspended according to the Vysis Nick Translation kit instructions. Metaphases were denatured for 2 min at 70°C in 70% formamide (Fluka, Switzerland) in 2xSSC and incubated with probe at 37°C overnight. Slides were washed according to Vysis instructions. For sequential hybridization, slides were dehydrated and co-denatured with mouse whole chromosome probes labeled with Cy3 (Pinkel et al., 1986) and the slides then washed as described above. The slides were counterstained with DAPI II (Vysis) and viewed with a Zeiss Axioplan 2 fluorescent microscope. The DAPI (367 nm), Spectrum Green (FITC: 509 nm) and Cy3 (588 nm) fluors were sequentially excited by a rotating filter wheel. Metaphases were captured with software (M-Fish 3.012) which accompanied the Quips imaging system (Vysis).

Real-Time RT-PCR Analysis

Eight week old CCSP/Ki-ras bitransgenic mice, as well as monotransgenic Ki-ras mice, were either untreated or administered drinking water containing 25, 100, or 500 μg/ml of DOX for 2 weeks (transgene expression analysis) or 12 months (tumor study). To determine expression of the Ki-ras transgene, 30 mg of whole lung tissue was homogenized with a Polytron homogenizer in RLT lysis buffer supplied in the RNeasy Mini Kit at speed 6. For analysis of lung tumors, tumor tissue was excised as described above. Total RNA was extracted using the RNeasy Mini Kit (Qiagen, Valencia, CA). For RT–PCR, cDNA was initially generated from 1 μg of RNA using the iScript cDNA Synthesis Kit (Bio-Rad). One tenth of the cDNA (2 μl) was used to amplify the Ki-ras transgene, p19^ARF, survivin, and GAPDH using the iCycler (Bio-Rad) for 40 cycles with SYBR Green Supermix (Bio-Rad). To generate a standard curve, each PCR product was first cloned into the pCR2.1 TA cloning vector and verified by gene sequencing. The plasmids were then utilized to generate standard curves for each of the genes using the same primers used for the unknown samples in order to quantify the copy number of each gene by the relative standard curve method. The Ki-ras transgene, p19^ARF, survivin, and GAPDH primers are listed in Table I. Cycling conditions included an initial denaturation step at 94°C for 3 min followed by 40 cycles of den aturation at 94°C for 30 s, annealing at 66.5°C (transgene and GAPDH), 59°C (survivin), or 61°C (p19^ARF) for 1 min, and elongation at 72°C for 1 min. A melt curve was determined following each real-time PCR run to ensure the presence of a single product. The iCycler software determined a copy number for each of the genes for each unknown sample based on the gene-specific standard curve. The values obtained for each gene of interest were normalized to GAPDH by dividing each sample by the respective GAPDH copy number. These normalized values were averaged and the average ± standard error. PCR efficiencies and R² values for the standard curves ranged from 98-103% and 0.997-0.999, respectively. Significant difference between all doses was determined by ANOVA. Significant differences between individual doses were determined by the standard t-test.

Table 1.

Primer sets used for real-time RT-PCR analysis

Gene	Oligonucleotide
ARF	5′ primer 5′-GGCTAGAGAGGATCTTGAGAAGAGG-3′ 3′ primer 5′-GCCCATCATCATCACCTGGTCCAGG-3′
Survivin	5′ primer 5′-GATGACAACCCGATAGAGGAGCAT-3′ 3′ primer 5′-CTCCTTTGCAATTTTGTTCTTGGCTC-3′
Ki-ras	5′ primer 5′ -CTGCAGAATTCGCCCTTATGACTGA-3′ 3′ primer 5′ -TAGCTGTATCGTCAAGGCACTCTTGC-3′
GAPDH	5′ primer 5′-TCTCCCTCACAATTTCCATCCCAG-3′ 3′ primer 5′-GGGTGCAGCGAACTTTATTGATGG-3′

Immunohistochemistry

Slides were deparaffinized with 3x 5 min washes with xylene. The slides were hydrated through graded alcohols 3x 1 min each. They were then placed in distilled water for 5 min. Antigen retrieval was performed using 10 mM citrate buffer (pH 6.0) for 35 min using a steamer. The slides were cooled for 20 min and placed in distilled water for 5 min. The following primary antibodies were used: anti-cyclin D1 1:100 (Abcam Inc., Cambridge, MA); anti-p19^ARF 1:100 (Abcam Inc. Cambridge, MA ); anti-survivin 1:500 (Abcam Inc., Cambridge, MA); anti-phospho-p53 (Ser³⁸⁹) 1:50 (Cell Signaling Technology); anti-caspase-3 (Cell Signaling Technology, Beverly, MA); anti-phospho-SAPK/JNK (Cell Signaling Technology}, anti-phospho-p44/42 MAPK (Thr²⁰²/Tyr²⁰⁴) 1:60 (Cell Signaling Technology Beverly, MA); anti-phospho-p38 MAPK (Thr¹⁸⁰/Tyr¹⁸²; 12F) 1:100 (Cell Signaling Technology Beverly, MA), anti-phospho-AKT (Ser⁴⁷³; 736E11) 1:50 (Cell Signaling Technology, Beverly, MD); and anti-Ki-67 (Abcam Inc. Cambridge, MA) Samples as well as negative controls (which lacked the primary antibody) were incubated overnight at 4°C. Slides were then washed 3x 1 min with 1x Tris buffer (Biomeda, Foster City, CA) containing 0.5% casein. A ready-to-use secondary biotinylated anti-rabbit antibody (BioGenex, San Ramon, CA) was added and the slides were incubated for 30 min at 32°C. The slides were washed 3x 1 min in Tris buffer followed by a 1:20 dilution of Streptavidin-alkaline phosphatase conjugate (BioGenex) for 30 min at 32°C. The slides were then washed 3x 1 min in Tris buffer, Vector Red substrate (Vector Laboratories, Burlingham, CA) was added for 5 min, and the slides washed 2x in 0.1 M Tris–HCl buffer (pH 8.2–8.4) and for 5 min with distilled water. The slides were counterstained with Mayer's hematoxylin for 2 min, then washed under running water for 5 min. Slides were dehydrated through graded alcohol and cleared through several changes of p-xylene.

Immunohistochemical Scoring and Statistical Analysis

Stained sections of lung were quantified using a digital camera attached to the image J software at 40x magnification. Magnification was scored based on the “H score” (Umemura et al., 2004). Fields from each slide were randomly chosen, a grid applied, and cells that fell on the intersection of two grid lines were quantified based on their staining intensity for each selected antibody. In the modified H-score, a score of “zero” designates no stain, and a score of “three” defines the darkest stain, with scores of “one” and “two” having intermediate staining (a score of “one” being the least and “two” being the greater). For each image, 100 cells with nuclei were assessed for staining intensity and assigned a value from 0 to 3. Tissue blocks from 6 to 8 individual mice were assessed at each of the different doses of DOX (each level of transgene expression). The slides were assessed by one individual (STD-B) who was blinded to the identity of the slides to assure consistency in assigning scores.

A mixed model approach was performed to determine the statistical significance of the difference between normal and proliferative tissue, as well as any statistical significance between proliferative tissue derived from mice treated with the 25, 100, or 500 μg/ml dose of DOX. Scored proliferative lesions were either hyperplastic lesions or adenomas. For each tissue block, two fields of proliferative lung tissue or surrounding normal tissue were quantified and averaged. A mean count ± SE was calculated which represented the average of the scores for the proliferative tissue minus the average of the scores from the surrounding normal tissue obtained from each of the quantified sections. In this calculation, 0 would represent no difference between proliferative and normal tissue, values greater than 0 would indicate an increase in staining intensity in proliferative lesions compared to normal tissue, and negative values would indicate a decrease in staining intensity in proliferative lesions compared to normal tissue. The higher the absolute number, the greater the difference between the proliferative and surrounding normal tissue Preliminary experiments found no differences in staining intensity of normal surrounding tissue compared to tissue isolated from control mice that had not received DOX.

Results

Effects of mutant RAS gene dosage on tumor phenotype

Previous studies from our laboratory, which described the benign tumor phenotype and altered signaling pathways resulting from maximal induction of mutant RAS gene expression, demonstrated that 22 copies of the transgene had integrated into the mouse genome (Floyd et al., 2005). Thus, we sought to further characterize the site of transgene integration. Chromosomes from the Ki-ras^G12C bitransgenic mice were screened using fluorescent in situ hybridization to determine the integration site of the of the Ki-ras^G12C transgene. The results shown in Figure 1 demonstrated that the transgene integrated into chromosome 6, the same chromosome on which the endogenous Ki-ras gene is found. Interestingly, our previous studies also demonstrated that the transgene was expressed in the testes and prostate of monotransgenic mice that contained only the mutant Ki-ras allele and lacked the tet transactivator protein, although there were no morphological changes observed in either organ. The human homologs of two testes-expressed loci found on human chromosome 17 (Bibbins et al., 1989), as well as an additional gene that plays a role in spermatogenesis (Bryan, 1977), have been shown to map to opposite arms of mouse chromosome 6, which may explain these earlier observations (Floyd et al., 2005).

Figure 1.

FISH analysis and identification of the integration site of the Ki-ras^G12C transgene in metaphase spreads from cells derived from lung tissue of Ki-ras^G12C monotransgenic mice. →, denotes integration site of Ki-ras^G12C on chromosome 6.

The tet-inducible system allows for the regulation of transgene expression by exposing the mice to varying concentrations of DOX. Quantification by real-time RTPCR of RNAs isolated from lung tissue of bitransgenic mice treated for 2 weeks with 25, 100, or 500 μg/ml of DOX demonstrated a dose dependent increase in transgene expression, with average copy numbers of 1.42×10⁻⁵, 2.43×10⁻⁴, and 4.95×10⁻³, respectively (Figure 2A). This dose dependent increase in transgene expression resulted in a dose dependent increase in tumor multiplicity (Figure 2B). Bitransgenic mice that expressed lower levels of mutant RAS exhibited tumor multiplicities of 1.7±0.4 and 2.7±0.7 compared to the historical value of 34±0.5 observed in mice that expressed maximally obtainable levels of mutant RAS expression (Floyd et al., 2005), and these differences were statistically significant between the 3 groups of mice (Figure 2; p<0.001). Monotransgenic control mice that were treated with 25 or 100 μg/ml of DOX exhibited tumor multiplicities ≤0.5 tumors/mouse, consistent with results obtained previously with 500 μg/ml of DOX.

Figure 2. Transgene expression levels correlate with tumor multiplicity.

(A)Total RNA was extracted from lung tissue of CCSP/Ki-ras mice. Real-time RT-PCR was performed to detect Ki-ras expression based on average copy number. Average copy number was obtained by normalizing transgene copy number to GAPDH copy number and is expressed as copy number ± SE. *, denotes s tatistical significance with p=0.0078 as determined by ANOVA and t-test. (B) Gene dosage dependent increase in tumor multiplicity, expressed as number of tumors per mouse/total number of mice with tumors ± SE. *, denotes statistical significance with p<0.001 as determined by ANOVA and t-test. C) Pulmonary lesions in CCSP/Ki-ras mice following treatment with DOX. a) Photomicrograph of a small hyperplastic focus in the lung periphery in a mouse treated with 100 μg/ml of DOX. H and E 2X; b) Higher magnification of focal epithelial hyperplasia in the lung of a mouse treated with 100 μg/ml of DOX. The cuboidal cells are uniform, without evidence of atypia, have round deeply basophilic nuclei, abundant eosinophilic cytoplasm, and maintain normal alveolar architecture. H and E 40X; c) An adenoma in the periphery of the lung of a mouse treated with 100 μg/ml of DOX. This lesion is well circumscribed, unencapsulated, quite solid, and compresses the adjacent parenchyma. H and E 2X; d) Higher magnification of an adenoma with papillary features in the lung treated with 100 μg/ml of DOX. Connective tissue fronds are covered by one to several layers of cuboidal or columnar cells which demonstrate mild anisocytosis. Nuclei are generally deeply basophilic and cytoplasm is abundant and eosinophilic. Occasionally mitosis present (arrow). H and E 40X.

Our previous studies demonstrated that the total levels of Ki-ras expression appeared to be tightly regulated; as transgene expression increased over time, the levels of the endogenous Ki-ras gene decreased (Floyd et al., 2005). We thus evaluated the transcript levels of endogenous Ki-ras and found that as transgene expression was induced to higher levels, endogenous Ki-ras RNA levels decreased, with copy numbers (average from two individual lung preparations) of 1.58×10⁻², 4.75×10⁻³, and 3.92×10⁻⁴ at the 25, 100, and 500 μg/ml doses of DOX, respectively. As wild type Ki-ras has been shown to act as a tumor suppressor gene (Zhang et al., 2001), the results suggest that reduced expression of the endogenous gene may be needed to allow the full proliferative potential of the mutant transgene to be observed (see Discussion section).

As noted above, macroscopic analysis of the lungs demonstrated significantly fewer lesions with lower levels of mutant RAS expression (Figure 2B). While the average lesion size at all three doses was <1 mm, 22% and 26% of the lesions that exhibited higher levels of transgene expression (treatment with 100 or 500 μg/ml of DOX, respectively) were larger, ranging from 1-4 mm in size. While none of the tumors reached the carcinoma stage, there were dose dependent differences noted in the frequency of tumor types between the different levels of transgene expression. Mice exhibiting the lowest levels of mutant RAS expression (treated with 25 μg/ml of DOX) had very few microscopic lesions within the lung, and these were very small hyperplasic foci. At the two higher levels of mutant transgene expression, larger hyperplastic foci (often close to being classified as adenoma) and adenomas were observed. Figure 2C shows representative sections of the hyperplastic lesions and adenomas.

Effects of mutant RAS gene dosage on downstream signaling pathways

Because of the small size of the tumors, we were unable to obtain sufficient amounts of tumor tissue for Western blot analysis. However, immunohistochemical (IHC) analyses of lung tissue from CCSP/Ki-ras mice expressing different levels of mutant RAS transgene demonstrated variations between treatment groups in the expression of oncogenes and tumor suppressor genes, as well as activation of signaling pathways involved in lung tumorigenesis (Figure 3A and 3B). There was a significant increase in staining intensity for phospo-p53 and caspase-3 in lung proliferative lesions derived from mice that expressed lower levels of mutant RAS transgene (treated with either the 25 or 100 μg/ml dose of DOX). This increased staining intensity was represented by mean counts of 23.3±18 (p=0.03) and 38.3±7.3 (p=0.001) for phospho-p53, and mean counts of 51.4±2.1 (p<0.0001) and 45.7±7.2 (p<0.0001) for caspase-3, respectively. Mice expressing the highest level of the mutant RAS transgene exhibited no statistically significant changes in either p53 phosphorylation or the levels of cleaved caspase-3. Previous IHC results reported by our laboratory (Floyd et al., 2006) indicated increased expression of phospho-p53 in lung lesions of mice when treated with 500 μg/ml of DOX; however, this determination was made using a subjective and qualitative analysis of the IHC staining than the mixed model approach used in this study.

Figure 3.

A) Immunohistochemical detection and quantitation of expression or phosphorylation of RAS downstream effectors. Statistical evaluation of the immunohistochemical determination of phospho-p53, p19, caspase-3, phospho-AKT, survivin, and phospho-JNK in lung tumor tissue from bitransgenic mice treated with 25, 100, or 500 μg/ml of DOX relative to the surrounding benign tissue. A mixed model approach was utilized to determine the signal intensity; results are expressed as the mean cell counts ± the standard error. *, denotes p-values b etween the normal and overall carcinogenic tissue was significant at ≤.05. B) Representative IHC slides showing staining with antibodies against JNK, p53, p38, and Ki-67 in surrounding normal tissue and lesions from mice treated with 25, 100, or 500 μg/ml of DOX; 20-40x magnification, images captured with the Image J software. C) Determination of mRNA expression of p19^ARF and survivin. The levels of expression of p19^ARFand survivin were determined by real-time fluorescent RT-PCR in normal lung tissue and adenomas from Ki-ras^G12C mice treated with 0, 25, 100, or 500 μg/ml of DOX. Each gene value is expressed as the average copy number ± SE and is corrected for the level of expression of GAPDH. Statistical differences in gene expression were determined by comparing samples from treated bitrangenic mice at each of the 3 doses to untreated bitransgenic mice (control samples). **, denotes significant difference in gene expression between treated tissue and control tissue with p value ≤ 0.05. *, denotes marginally significant difference in gene expression between treated tissue and control tissue with p value = 0.08.

IHC analyses comparing surrounding normal lung tissue to proliferative tissue also revealed an increase in survivin expression in lung lesions of mice that expressed higher levels of the mutant RAS transgene (treated with 100 and 500 μg/ml of DOX), represented by mean counts of 54.7±2.3 and 16.3±1.7 , respectively. There was also an increase seen in p19 expression levels in lung lesions, but only at the highest levels of transgene expression (p<0.001).

Additional phospho-specific antibodies were used to examine activation of the AKT and JNK pathways. A significant increase in AKT phosphorylation was only seen in lung lesions of mice that received the 100 μg/ml dose of DOX (p<0.001), as compared to surrounding normal tissue. A significant increase in JNK expression (p<0.0001) was seen in lung lesions of mice that received the 25 μg/ml dose of DOX, represented by a mean count of 13.7±2.7. However, there was a significant decrease in JNK expression seen in lung lesions of mice that expressed higher levels of the mutant RAS transgene, represented by mean counts of −7.3±2 .5 (p=0.004) and −11.0±1.6 (p<.0001) in mice treated with 100 μg/ml and 500 μg/ml of DOX, respectively. IHC analysis was also used to examine expression levels or phosphorylation of additional proteins (Ki-67, ERK, cyclin D1, and p38). In all mice expressing the mutant RAS transgene, there was a significant increase in staining for Ki-67, phospho-ERK, cyclin D1, and phospho-p38 (Fig. 3A). Whereas the levels of phospho-ERK and cyclin D1 were maximally induced at the lowest level of transgene expression, phosphorylation of p38 and expression of Ki-67 exhibited a dose-dependent response. For phospho-p38, the staining intensity at the 25 μg/ml dose was significantly lower (p <0.02) than the 100 or 500 μg/ml dose; for Ki-67, the 500 μg/ml dose was significantly higher (p < 0.0001) than the two lower doses.

To further confirm the results obtained by IHC, we isolated mRNA from lung tumors of CCSP/Ki-ras mice treated with each of the 3 doses of DOX and normal tissue from untreated mice. Real time RT-PCR (Figure 3C) showed that the mRNA expression levels for p19^ARF and survivin were comparable to that of the protein levels determined by IHC analysis. Because of potential contamination from non-tumorous lung tissue, the levels of RNA expression may underestimate the actual changes in RNA levels in the tumors.

Discussion

Previous studies from our laboratory described the development and characterization of a bitransgenic mouse model that allows for the conditional expression of the mutant human Ki-ras^G12C allele in a lung-specific manner from a tetracycline regulated promoter (Floyd et al., 2005). In this study, treatment of bitransgenic mice with 3 different doses of DOX resulted in a statistically significant, dose dependent increase in transgene expression (Figure 2A), as well as a dose dependent increase in tumor multiplicity (Figure 2B).

The increase in tumor multiplicity observed between the 25 and 100 μg/ml doses (1.6 fold) was much smaller than the large increase in transgene expression (17 fold). As our previous studies demonstrated that the total levels of Ki-ras expression are tightly regulated (Floyd et al., 2005), we evaluated the levels of endogenous Ki-ras expression by RT-PCR. We found that as transgene expression was induced to higher levels, endogenous Ki-ras RNA levels decreased. While it is possible that the observed benign tumor phenotype may be due to the increased expression of the mutant transgene, wild type Ki-ras has been shown to exhibit tumor suppressor activity (Zhang et al., 2001). Thus, it is also possible that lower levels of endogenous wild type gene expression are required before the full proliferative potential of the mutant transgene is manifested. These data cannot distinguish between these possible mechanisms and would require the use of knockout mice or siRNA silencing to definitively determine the role of the endogenous and mutant transgene expression in altering signal transduction pathways. It does appear that increased transgene expression at the two higher doses of DOX (100 and 500 μg/ml) was associated with the transition to adenoma development, as we observed predominantly small foci of hyperplasia at the lowest levels of Ki-ras^G12C expression but increased development of adenoma formation at higher levels of mutant RAS transgene expression. Since the levels of phospho-ERK and cyclin D1 were maximally induced at the lowest levels of transgene expression, but other signaling pathways exhibited dose-dependent effects, the results confirm the importance of defects in multiple signaling pathways in driving tumor progression.

p53, a classic tumor suppressor, becomes activated in response to a myriad of cellular stressors, one of which is deregulated oncogene expression (Agarwal et al., 2001;Horn et al., 2007). IHC demonstrated activation of the p53 pathway at lower levels of RAS transgene expression, whereas maximal induction of mutant RAS expression demonstrated a smaller and more subtle increase in phospho-p53 (Figure 3A). It thus appears that at lower levels of mutant RAS expression, the tumor suppressor activity of p53 is engaged and limits tumor development. Concurrently, the lower levels of transgene expression were also associated with increased caspase-3 activity; both of these signaling molecules play critical roles in stimulating apoptosis (Howe et al., 1992). In contrast, there was no significant increase in phospho-p53 and caspase-3 staining noted in the mice treated with 500 μg/ml of DOX relative to normal lung. Interestingly, IHC and real-time RT-PCR (Figure 3) also demonstrated that there was no significant change in p19^ARF expression at lower levels of RAS transgene expression despite the increased phosphorylation of p53, whereas maximal induction of mutant RAS transgene expression resulted in significant elevation of p19^ARF RNA and protein but no significant activation of the p53 pathway. Thus, lower levels of mutant RAS expression appear to induce apoptosis in the lung lesions via a p53-dependent mechanism but do not appear to trigger up-regulation of p19^ARF. However, at higher levels of mutant RAS expression, activation of p53 and induction of apoptosis were inhibited despite increased expression of p19. Along the same lines, survivin RNA expression was increased at the two higher levels of transgene expression, which likely enhanced cell survival. Thus, the combination of increased survival signals and decreased apoptotic signals likely accounts for the enhanced tumor multiplicity and increased occurrence of adenomas observed at high levels of transgene expression.

In addition to the differential effects of RAS dosage on the survivin and p53 pathways, dose dependent alterations in the levels of phosphorylation or expression of other RAS downstream effector pathways were also noted. We observed increased activation of the AKT pathway, as indicated by increased staining for phospho-AKT, in the lungs of mice treated with 100 μg/ml of DOX. Activation of AKT was not observed in mice treated with 25 or 500 μg/ml of DOX (Figure 3). AKT has been found to contribute to cancer progression by stimulating cell proliferation, promoting angiogenesis, inhibiting apoptosis and enhancing cell survival (Massion et al., 2004;Datta et al., 1997), and may contribute to the mechanism by which moderate levels of transgene expression circumvent the tumor suppressor effects of p53 and progress to adenoma.

The c-Jun N-terminal kinase (JNK) is a member of the stress-activated protein kinase (SAPK) family that rapidly transmit mitogenic, apoptotic and other stress-related signals in mammalian cells (31,32). Different levels of mutant RAS gene dosage resulted in differential effects on JNK phosphorylation. Mice that were induced to express higher levels of transgene expression (mice treated with 100 or 500 μg/ml of DOX) exhibited a significant decrease in JNK phosphorylation in their lung lesions compared with the increased phosphorylation of JNK observed in mice that expressed lower levels of mutant RAS transgene. Sustained JNK activation is suggested to contribute to apoptotic cell death (Guo et al., 1998;Pu et al., 2002), and the lack of sustained JNK activation may contribute to the lack of apoptotic activity in mice that exhibited higher levels of transgene expression.

These results provide in vivo evidence for the differential effects of mutant Ki-ras expression on various parameters of lung tumorigenesis - including alterations in tumor multiplicity and signaling events – that are dependent on the levels of mutant RAS expression. The combination of proliferative and anti-proliferative signals stimulated by different levels of RAS transgene expression differentially influence the activation of downstream effector pathways that are initiated at the level of RAS activation, and thereby influence tumor characteristics. Use of this conditionally expressed RAS transgenic model will aid in our understanding of the effects of RAS expression on tumor phenotype, and will provide important information on tumor progression, clinical staging, and response to chemotherapy.

Abbreviations

AC

adenocarcinomas

AD

adenomas

CCSP

Clara cell secretory protein

DOX

doxycycline

GAPDH

glyceraldehyde-3-phosphate dehydrogenase

erk

extracellular regulated kinase

JNK

Jun kinase

MAPK

mitogen activated protein kinase

MEK

mitogen activated erk kinase

NSCLC

non-small cell lung cancer

RT-PCR

reverse transcription-polymerase chain reaction

rtTA

reverse tetracycline trans-activator

tet

tetracycline