The incidences of alveolar/bronchiolar adenomas and carcinomas in cumene-treated B6C3F1 mice were significantly greater than those of the control animals. We evaluated these lung neoplasms for point mutations in the K-ras and p53 genes that are often mutated in humans. K-ras and p53 mutations were detected by cycle sequencing of PCR-amplified DNA isolated from paraffin-embedded neoplasms. K-ras mutations were detected in 87% of cumene-induced lung neoplasms, and the predominant mutations were exon 1 codon 12 G to T transversions and exon 2 codon 61 A to G transitions. P53 protein expression was detected by immunohistochemistry in 56% of cumene-induced neoplasms, and mutations were detected in 52% of neoplasms. The predominant mutations were exon 5, codon 155 G to A transitions, and codon 133 C to T transitions. No p53 mutations and one of seven (14%) K-ras mutations were detected in spontaneous neoplasms. Cumene-induced lung carcinomas showed loss of heterozygosity (LOH) on chromosome 4 near the p16 gene (13%) and on chromosome 6 near the K-ras gene (12%). No LOH was observed in spontaneous carcinomas or normal lung tissues examined. The pattern of mutations identified in the lung tumors suggests that DNA damage and genomic instability may be contributing factors to the mutation profile and development of lung cancer in mice exposed to cumene.
Introduction
Lung cancer is the most frequently diagnosed cancer in the world and the most common cause of cancer mortality worldwide. This high mortality rate is largely owing to the late stage of diagnosis and the poor response to therapy. The need to develop better diagnostic techniques and therapies is urgent. Mouse models have been used to study carcinogenesis of human lung cancers, and many of the major genetic alterations detected in human lung cancers have also been identified in mouse lung tumors.
Mouse alveolar/bronchiolar adenomas and carcinomas, which are common spontaneous and chemical-induced lung tumors in mice, are similar to human adenocarcinomas in histomorphology and molecular characteristics, including activation of the K-ras gene (Meuwissen and Berns 2005; Nikitin et al. 2004). The patterns of mutations in cancer genes, such as ras and p53, have been found to aid in the understanding of tumorigenesis in rodents and humans (Harris 1993; Le Calvez et al. 2005; Maronpot et al. 1995; Osada and Takahashi 2002; Reynolds 1987). For example, in some neoplasms, the profile of activating mutations in ras genes or inactivating mutations in the p53 gene is specific for particular chemicals and differs from those detected in spontaneous neoplasms (Sills et al. 1999; Sills et al. 2004).
Cumene, or isopropylbenzene, is a constituent of crude oil used primarily for the production of phenol and acetone. It is a good solvent for fats and resins and has been suggested as a replacement for benzene in many industrial applications. The annual production of cumene in the United States is high and increasing, with 4.49 billion pounds produced in 1993 (National Library of Medicine Hazardous Substance Data Bank 2000). Because it is a major commodity chemical, there is a high potential for many workers to be exposed to cumene. The most probable route of human exposure is inhalation of contaminated air (Jackson 1985) from cumene evaporated into the environment during production and processing from petroleum refining, combustion of petroleum products, and use of a variety of products containing cumene.
National Toxicology Program (NTP) two-year, whole-body–inhalation studies revealed that treatment of B6C3F1 mice with cumene significantly increased the incidence of alveolar/bronchiolar adenomas and carcinomas in all groups of exposed males and females. Cumene is not genotoxic, which was demonstrated by several studies involving bacterial and mammalian cells in culture and in in vitro studies involving mice and rats (NTP 1996; National Library of Medicine Hazardous Substance Data Bank 2003; US Environmental Protection Agency 1997). In vitro cell transformation assays using BALB/3T3 mouse embryo cells and unscheduled DNA synthesis assays using rat primary hepatocytes yielded conflicting results regarding a cumene effect that were not reproducible. Cumene was weakly positive with no clear dose response for the induction of micronuclei in rat bone marrow at doses ranging from 78 to 2500 mg/kg intraperitoneally (NTP 1996).
The goal of this study was to evaluate both spontaneous and cumene-induced lung neoplasms for mutations in the K-ras and p53 genes, which are considered important in the pathogenesis of human lung cancer. Another goal was to identify chemical-specific mutations that might serve as biomarkers of occupational exposure with possible relevance to human health.
Materials and Methods
Lung Neoplasms
Male and female B6C3F1 mice were exposed to 0, 125 (females only), 250, 500, or 1000 (males only) ppm cumene (50 animals per group) by whole-body inhalation, six hours per day, five days per week for two years (NTP, CAS Number 98-82-8). Incidences of cumene-induced lung tumors were evaluated using the poly-3 test. Husbandry and experimental procedures were in compliance with the requirements set forth by the Public Health Service’s Guide for the Care and Use of Laboratory Animals. At necropsy, tissues were fixed in 10% neutral-buffered formalin, routinely processed, embedded in paraffin, sectioned to a thickness of 5 μm, and stained with hematoxylin and eosin. Subsequently, five unstained serial sections, 10 μm thick, were prepared from paraffin blocks containing alveolar/bronchiolar adenomas or carcinomas. To isolate adequate amounts of DNA, neoplasms greater than 1 mm in diameter were identified for analysis. Fifty-two cumene-induced alveolar/bronchiolar neoplasms (six adenomas and forty-six carcinomas), seven spontaneously occurring carcinomas, and six normal lung tissues were evaluated for K-ras mutations in exons 1 and 2 (codons 12, 13, and 61) and p53 mutations in exons 5–8.
DNA Isolation, Amplification, and Cycle Sequencing
DNA was isolated and extracted from paraffin-embedded sections containing lung neoplasms and normal lung tissues and amplified by PCR. Details of the use of nested primers for K-ras and p53 genes have been described previously (Lambertini et al. 2005; Sills et al. 1995). Positive DNA controls for K-ras and p53 mutations and controls lacking DNA were run with all sets of reactions. PCR products were purified using a QIAquick Gel Extraction Kit (Qiagen, Valencia, CA, USA). The purified samples were sequenced using a cycle-sequencing kit (US Biochemical, Cleveland, OH, USA), which incorporated [α-33p]-dideoxynucleotide triphosphate (ddNTP) terminators (A, C, G, T) into the sequencing products. Detected mutations were confirmed by repeat analysis, starting from amplification of the original DNA extract.
Immunohistochemistry for p53 Protein
Alveolar/bronchiolar adenomas and carcinomas were examined for p53 protein expression by immunohistochemical analysis, using an avidin-biotin-peroxidase detection system (Vectastain Rabbit Elite Kit, Vector Laboratories, Burlingame, CA, USA). The immunohistochemical staining for expression of mutant p53 protein was performed as previously described (Hsu et al. 1981; Hong et al. 2000; Hong et al. 2003; Maronpot et al. 1993; Sills et al. 2004). A 1:300 dilution of the primary polyclonal rabbit antibody CM-5 (Vector Laboratories, Burlingame, CA, USA), which detects accumulation of the mutant p53 in rodents, was used. Normal rabbit serum (Vector Laboratories, Burlingame, CA, USA) was used as the negative control in place of the primary antibody. Tissue specimens from a p53 transgenic mouse (mutation in codon 135 of p53) served as the positive control.
Loss of Heterozygosity (LOH) Analysis
Microsatellite marker Mts 1 near the p16 tumor suppressor gene on chromosome 4 was used to amplify polymorphic regions between strain C57BL/6 (B) and C3H/He (H) mice (Jackson Laboratories, Bar Harbor, ME, USA) to identify portions of tumors with LOH at the p16 locus. The Mts 1 forward (5′-GA TTT CTA CGG AAA GCC CTG-3′) and reverse (5′-TAT TGT GCA TTT GTG TGT CTG G-3′) primers were located 2395 and 2172 bp upstream of the translational start site, respectively. PCR amplification was performed on forty cumene-induced lung neoplasms (thirty-six carcinomas, four adenomas) and seven spontaneous B6C3F1 carcinomas. Controls lacking DNA were included with all amplifications. The PCR products were resolved on 4% NuSieve (FMC BioProducts, Rockland, ME, USA)–agarose (3:1) to separate B and H alleles.
The inner PCR products from paraffin-embedded plus eight frozen treated neoplasms (seven carcinomas and one adenoma) were labeled by incorporation with [α-33p]-dATP (MP/ICN Biomedicals, Irvine, CA, USA), using single-stranded conformation polymorphism (SSCP) analysis to distinguish between the H and B alleles on a 0.5X MDE (Cambrex Bio Science, Rockland, ME, USA) gel at 3W for fifteen hours at room temperature.
To determine LOH in fifty cumene-induced lung neoplasms (forty-six carcinomas, four adenomas) and seven spontaneous carcinomas at marker D6MC012 near the K-ras gene on chromosome 6 in B6C3F1 mice, SSCP analysis was used. PCR amplification with [α-33p]-labeled and unlabeled dATP was performed as described previously (Sills et al. 2004). The MDE gel was electrophoresed at 3W at room temperature for fifteen hours to separate the H and B alleles. The gels were dried and exposed to x-ray film overnight.
Results
Mutation Analysis
The incidence of alveolar/bronchiolar adenomas and carcinomas in male and female treated groups was significantly greater than those of the controls (Table 1). A higher frequency of K-ras mutations (45/52, 87%) was observed in the cumene-induced lung neoplasms, as compared to spontaneous lung neoplasms from control animals (historical 33/117, 28%; concurrent 1/7, 14%) (Table 2). These mutations were observed in 91% (41/45) of males and 57% (4/7) of females. K-ras mutations were detected in 67% (4/6) of the adenomas (one in codon 13 and three in codon 61) and in 89% (41/46) of the carcinomas. The predominant K-ras mutations were codon 12 GTT (G to T transversions) and codon 61 CGA (A to G transitions). Three codon 12 CGT mutations and one codon 61 CTA mutation were found in the cumene-induced neoplasms, but none were found in spontaneous lung neoplasms (0/124, concurrent controls combined with historical controls) (Table 2; Figures 1 and 2). There was no significant increase in the incidence of mutations at codon 13 in the cumene-induced lung neoplasms.
p53 protein expression was detected by immunohistochemistry in 29/52 (56%) of cumene-induced lung neoplasms (Figure 3), and mutations were identified in 27/52 (52%) of the neoplasms. p53 protein expression and gene mutation were not detected in the seven spontaneous carcinomas or in the six normal lung tissues except one spontaneous carcinoma with positive p53 protein expression (Table 3). The predominant p53 mutations were identified in exon 5 (24/27, 89%) and in exon 7 (3/27, 11%) (Table 3, Figure 4). No mutation was detected at exons 6 and 8 in the samples examined. The p53 mutations were observed in 58% (26/45) of males and 14% (1/7) of females.
Mutations were detected in 50% (3/6) of the adenomas and 52% (24/46) of the carcinomas. There was a dose-dependent increase in incidence of K-ras and p53 mutations; however, a similar mutation spectrum of both mutations was detected in cumene-induced neoplasms regardless of whether the neoplasms were adenomas or carcinomas.
LOH Analysis
LOH on chromosome 6 near the K-ras gene was observed in 6/50 (12%; 4–H and 2–B) treated carcinomas, 0/6 of adenomas, and 0/7 of the spontaneous carcinomas examined (Figure 5). LOH of the C3H/He allele (H) was observed in 5/40 (13%) cumene-induced carcinomas, 1/6 of adenomas, and 0/7 of the spontaneous carcinomas examined at the microsatellite marker on chromosome 4 near the p16 gene (Figure 6).
Discussion
There was a high frequency (87%) of K-ras mutations in cumene-induced alveolar/bronchiolar neoplasms compared to that in spontaneous alveolar/bronchiolar neoplasms from untreated B6C3F1 mice (28% historical database; 14% concurrent controls). The predominant mutations were K-ras codon 12 G to T transversions (GGT to GTT, 21%) and codon 61 A to G transitions (CAA to CGA, 25%), which clearly differed from those identified in untreated control mice (0.008% and 2%, respectively). Point mutations at codon 12 of the K-ras gene are activating mutations, rendering ras insensitive to the down-regulatory action of GTPase activating proteins and thereby locking the protein in the active state and promoting cellular transformation (Chad and Geoff 2000).
The high frequency and pattern of K-ras mutations in mouse lung tumors may directly depend on the nature of the chemical carcinogen or its metabolites. In our study, the development of lung neoplasms in B6C3F1 mice exposed to cumene may involve multiple carcinogenic processes including direct DNA damage and/or indirect DNA damage. Metabolites of cumene may have caused DNA adducts and subsequent point mutations. Side-chain oxidation of cumene (isopropylbenzene) is rapid and extensive and occurs in both hepatic and extrahepatic tissues including the lung (Sato and Nakajima 1987), with the secondary alcohol 2-phenyl-2-propanol being the principal metabolite in rats (Research Triangle Institute 1989; US Environmental Protection Agency 1997) and humans (Lee 1987; Meuwissen and Berns 2005; US Environmental Protection Agency 1997). The C-isopropyl bonds are readily cleaved, and the remaining electrophilic carbon moiety may form DNA adducts and cause subsequent DNA damage. Previous studies showed that the related benzene is carcinogenic and genotoxic (Abemethy et al. 2004; Gut et al. 1996; Snyder and Hedli 1996; Valentine et al. 1996).
Alternatively indirect damage from oxidative stress may have contributed to the mutations. G to T transversions are commonly detected DNA base changes associated with active oxygen species and are consistent with 8-OH-G adducts produced during oxidative damage to DNA (Janssen et al. 1993; Kino and Sugiyama 2005; Shigenaga and Ames 1991; Tchou et al. 1991). In other studies, exposure of B6C3F1 mice to ozone or vanadium pentoxide is thought to result in the generation of hydroxyl radicals, which induce a G to T transversion at codon 12 of K-ras gene (Devereux et al. 2002; Sills et al. 1995). Interestingly, G to T transversion in K-ras codon 12 is the most common mutation detected in human adenocarcinomas (Rodenhuis et al. 1987). In human lung tumors, K-ras mutations appear to correlate with DNA adducts of benzo(a)pyrene and are associated with smoking (Santillo et al. 2001). It is possible that smoking in combination with cumene exposure in humans may have an additive effect on K-ras mutations.
Although K-ras appears to play a critical role in mouse and human lung adenocarcinoma formation, its function is complex. Zhang et al. (2001) proposed that wild-type K-ras may be a mouse lung tumor suppressor gene. Further LOH analysis in cumene-induced lung neoplasms demonstrated allelic loss on chromosome 6 near the K-ras gene in six carcinomas and on chromosome 4 near p16 tumor suppressor gene in five carcinomas. Allele loss of p16 has been detected in human cancers (de Vos et al. 1995; Kratzke et al. 1996) and mouse lung tumors (Patel et al. 2000).
A high frequency (52%) of p53 mutations was detected in cumene-induced alveolar/bronchiolar neoplasms that were correlated by immunohistochemistry with increased p53 protein expression (56%). The presence of p53 protein expression without p53 gene mutation could be a result of mutations outside exons 5–8 or possibly owing to alterations of other proteins downstream of p53 (Greenblatt et al. 1994). The predominance of cumene-induced alveolar/bronchiolar neoplasms that contain p53 mutations may provide a selective advantage for unregulated growth and the avoidance of apopto-sis (Greenblatt et al. 1994; Harris 1996; Osada and Takahashi 2002; Rodin and Rodin 2005). A study of aflatoxin-B1 (AFB1)-induced mouse lung tumors found a high proportion (> 70%) of tumors with p53 accumulation and mutations (Tam et al. 1999), and lung tumors of mice exposed transpla-centally to AZT also had a high proportion (84%) of p53 mutations (Hong et al. 2007). Other studies such as methylene-chloride–induced mouse lung tumors (Hegi et al. 1993), ozone-induced, and vanadium-pentoxide–induced mouse lung tumors (Devereux et al. 2002; Sills et al. 2004) exhibited no or low frequency of p53 mutations. Unlike the mostly random mutation pattern for the AFB1-induced tumors (Tam et al. 1999), the cumene-induced tumors displayed specific p53 mutations. P53 mutations were observed only in exon 5 (24/27, 89%) and exon 7 (3/27, 11%). The mutations observed in the p53 gene in cumene-exposed mice clearly imply this genetic event is related to chemical exposure, since these mutations were not detected in spontaneous tumors.
In conclusion, the patterns of K-ras and p53 mutations identified in the cumene-induced lung tumors suggest that DNA damage and genomic instability may be the contributing factors to the mutation profile and development of lung cancer in these mice. The molecular alterations identified in the cumene-induced lung neoplasms appear to affect the same pathways as those reported in human lung cancer, suggesting that the response in the mouse may be of relevance to humans. Futher research of the cumene mouse lung tumors by microarray analysis underscores the complexity of lung cancer where, in addition to genetic factors, signal transduction pathways and epigenetic mechanisms are also contributing factors to the car-cinogenesis process (Wakamatsu et al. forthcoming).
