The effect of the chemopreventive agent D,L-alpha-difluoromethylornithine (DFMO) on the incidence of skin squamous cell carcinoma was studied in SENCAR mice treated weekly with topical applications of benzo(a)pyrene (B(a)P) (0.15 mmol, 2 x /week) on the dorsal skin. Animals were randomized to receive either chow or chow supplemented with DFMO (1 g/1 kg) and studied at 10, 15, 20, 25, and 30 weeks of B(a)P treatment. Morphometric analyses at each timepoint evaluated the epidermal thickness (ET) and the number of epidermal nucleated layers (NL). The ET increased from 12-17 microns as early as 10 weeks after B(a)P treatment, reaching 22 microns at 20 weeks, and 27 microns at 25 weeks (130% increase). The NL also increased markedly. A relatively modest increase in ET was observed in animals treated with B(a)P and DFMO (16% at 15 weeks, 53% at 20 weeks, and 85% at 25 weeks) as compared to controls. The relative increase in NL showed a similar pattern. Although extensive epidermal hyperplasia was seen early, clear-cut focal premalignant lesions were not identifiable before week 20 of B(a)P treatment. At 20 weeks, the most frequently noted focal premalignant lesions in carcinogen-treated animals (without DFMO) were moderate dysplasias. At 25 and 30 weeks, a large increase was seen in the incidence of more advanced dysplastic lesions and invasive carcinomas. In the group treated with B(a)P and DFMO, a marked reduction in the number of carcinomas was observed at 25 and 30 weeks. At 25 weeks, DFMO reduced tumor yield from 5.8 to 3.2 carcinomas per mouse. At 30 weeks, the reduction was from 13.1 to 5.7 carcinomas per mouse (57% reduction). Collectively, these data emphasize the strong chemopreventive effect of DFMO against tumors in the mouse skin complete carcinogenesis model, as indicated by the reduction of overall skin tumor incidence and the decreased epidermal hyperplasia in DFMO-treated animals. Morphometrically defined increases in ET and NL can be used as early biomarkers of DFMO chemoprevention in mouse skin tumorigenesis.

We studied the frequency and pattern of p53 mutations in 16 mouse skin primary squamous carcinomas induced by chemical carcinogens and their 19 matched metastases. The molecular changes were analyzed by polymerase chain reaction‐single‐strand conformation polymorphism and subsequent direct sequencing analysis. Eleven mutations of the p53 gene were detected in a total of eight primary tumors, and 10 mutations were detected in nine metastases. Only four pairs had identical mutations in primary and paired metastatic tumors. Eight mutations in six pairs were detected in primary tumors but not in their metastases, and four mutations from three matched pairs were detected in metastases but not in primary tumors. The four pairs that contained the same mutations in both the primary and secondary tumors had lymph‐node metastases, and all mutations occurred in exon 8. Conversely, five of six pairs with p53 mutations only in primary tumors had lung metastases and only one of the mutations occurred in exon 8. None of the mutations found only in metastases were located in exon 8. These data indicate that p53 mutations are prevalent in lymph‐node metastases and infrequent in lung metastases of mouse skin tumors and that primary tumors with exon 8 mutations may be more likely to metastasize to the lymph nodes. © 1995 Wiley‐Liss Inc.

Human tobacco-related cancers exhibit a high frequency of G to T transversions in the mutation hot spot region of the p53 tumor suppressor gene, possibly the result of specific mutagens in tobacco smoke, most notably benzo[a]pyrene (B[a]P). No in vivo animal model of B[a]P-induced tumorigenesis has been used, however, to substantiate these molecular epidemiological data experimentally. Direct DNA sequence analysis of the hot spot region (exons 5-8 inclusive) of murine p53 was performed in 20 skin tumors induced by a complete carcinogenesis protocol with B[a]P. Sequence analyses revealed numerous heterozygous missense mutations in carcinomas, specifically in exons 7 and 8 of the p53 gene, and targeting exclusively guanine residues. Moreover, 70% (5/7) of the mutations characterized were G to T transversions. In contrast, direct DNA sequence analysis of 36 skin tumors induced by 7,12-dimethylbenz[a]anthracene (DMBA) in either a complete carcinogenesis protocol or in a two-stage carcinogenesis protocol revealed a 30% frequency of heterozygous p53 mutations, with the majority of mutations found in carcinomas, but only a single G to T transversion (1/8). Thus, while mutation frequencies are similar, the pattern and type of p53 mutations in B[a]P-induced skin tumors differs significantly from the mutation spectra in DMBA-induced squamous neoplasias. These in vivo findings in B[a]P-induced tumors lend support to in vitro and molecular epidemiological evidence, suggesting that the p53 tumor suppressor gene may be a selective target of metabolically activated B[a]P species etiologically associated with human tobacco-related cancers.

Several chemicals that are found in cigarette smoke or diesel oil engine exhausts, such as benzo[a]pyrene (B[a]P) and 1,6-dinitropyrene (DNP) are carcinogenic in experimental animal models. In the present study, we have exposed in vivo the xenotransplanted immortalized human bronchial epithelial cell line BEAS-2B to the ultimate carcinogen of B[a]P, benzo[a]pyrene diolepoxide (BPDE), to DNP or to the benzo[e]pyrene, a less active compound that has tumor-promoting abilities in mouse skin carcinogenesis bioassays. All three compounds were administered using slow-release beeswax pellets. After a 6 month exposure, BPDE produced two tumors in seven transplants, four tumors were seen in 10 transplants treated with DNP and one tumor was observed in five tracheal grafts exposed to B[a]P. All the neoplasms were well-differentiated invasive adenocarcinomas. Tracheal transplants exposed to beeswax without carcinogen did not show any evidence of neoplastic growth, and their luminal surfaces were lined by a single or double layer of cuboidal cells. All lines derived from the adenocarcinomas showed increased in vitro resistance to serum-induced terminal differentiation, gelatinolytic activity, s.c. tumorigenicity and invasive growth in an in vivo assay. When these cell lines were compared with previously described tumor cell lines derived from xenotransplants exposed to cigarette smoke condensate, it became clear that the latter exhibited a more aggressive invasive behavior. Nevertheless treatment with the three chemicals gave rise to tumor cell lines that exhibited a similar invasive behavior in vivo, and were able to penetrate early into the wall of the tracheal transplants in which they were seeded. These data indicate that this system based on xenotransplanted bronchial epithelial cells is a very relevant model to identify human carcinogens and to study mechanisms of bronchogenic cancer pathogenesis.

p53 alterations were studied in a group of 22 primary squamous cell carcinomas (SCC) of the head and neck and in 10 cell lines derived from SCC. Positive immunohistochemical detection of p53 was accomplished in 10 of 22 primary tumors and in 7 of 10 SCC cell lines. Loss of heterozygosity of chromosome 17p, were the p53 gene is localized, was seen in five of seven SCC lines studied. DNA sequencing of the p53 gene of these five cell lines that had lost one allele showed p53 mutations in the remaining allele. In addition, from six primary SCC that exhibited loss of heterozygosity of chromosome 17p, three showed missense mutations of the p53 gene. The mutations of primary tumors and SCC cell lines were scattered in the midregion of the gene, affecting codons 151, 155, 174, 194, 220, 248, and 273. Five of these mutations modified guanine residues, a phenomenon that has been associated with the effect of carcinogens contained in tobacco smoke. Collectively these data show that approximately 50% of primary tumors and cell lines derived from SCC of the head and neck showed abnormalities of the p53 gene. In addition, it is of interest to note that the most invasive cell lines, as determined in an in vivo assay using xenotransplantation of tumor cells into denuded rat tracheal grafts, exhibited the most intense staining. Similarly, of five very advanced primary tumors, four showed intense p53 immunostain. These observations support the evidence that alterations in this tumor suppressor gene could be related to late events in tumor progression.