| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Epidemiology and Prevention |
Laboratory of Molecular and Cellular Toxicology, Finnish Institute of Occupational Health, FIN-00250 Helsinki, Finland, [K. H-P., A. K.]; IARC, 69372 Lyon Cedex 08, France [P. B., F. N.]; Institute of Environmental Medicine, Karolinska Institute, S-17177 Stockholm, Sweden [F. N., G. P.]; Institute of Carcinogenesis, 115478 Moscow, Russia [A. M.]; Institute of Public Health, 76256 Bucharest, Romania [V. C.]; Regional Epidemiological Centre, I-00198 Rome, Italy [C. F.]; and National Institute of Health and Medical Research (U521), 94805 Villejuif Cedex, France [S. B.]
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Genetic alterations, in particular, mutations in genes controlling cell growth and proliferation, have potential in discovering crucial biological alterations associated with long-term exposure to carcinogens; consequently, they may aid in unraveling pathways leading to tumor formation. In this context, mutations of the tumor suppressor gene p53, which encodes a multifactorial transcription factor controlling cellular response to DNA damage (20) , represent a potentially useful biomarker in the search for etiology, molecular mechanisms, and, hopefully, prevention of environmental cancers (21, 22, 23) . Mutations of the p53 gene occur in about 50% of human lung tumors (24 , 25) , and the mutations observed in lung cancer appear to have typical features (25 , 26) . The international database on p53mutations records mutation data for about 1000 cases of lung cancer (27) . Surprisingly, however, data on smoking status are available for only about one-fourth of these cases, and very few cases are identifiable as documented never-smokers (28) .
This molecular epidemiology investigation was aimed at addressing the association between ETS exposure and lung cancer in never-smokers using the frequency and type of p53 mutations in lung tumors as a biomarker of tobacco-related carcinogenesis.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
PET blocks or fresh tissue samples from 198 cancer patients with a
histologically confirmed diagnosis of lung cancer were sent for
p53 mutation analyses to the Finnish Institute of
Occupational Health. A histological tissue section was cut from each
PET block, stained with H&E, and examined by a pathologist to confirm
the presence of lung tumor tissue. For most of the cases, the
pathologist gave an evaluation of the approximate area of tumor tissue
in each block, and blocks containing 
50% tumor tissue were selected
for DNA extraction and mutation analysis whenever available. However,
the presence of nontumor tissue was not of major concern because
wild-type sequences do not hamper mutation detection by the screening
procedure used (see below). Finally, 157 cancer cases (contributing 229
PET blocks and 20 fresh tumor specimens) with DNA yields suitable for
analysis were investigated for the presence of p53 mutation.
Primary tumor samples were available for all patients, except for two
patients with metastasis, three patients with positive lymphatic
tissue, and two patients with positive pleural biopsy. One patient
presented with two primary lung tumors, one of large cell cancer
histology and the other of adenocarcinoma histology.
p53 Mutation Analysis.
DNA was obtained from fresh and paraffin-embedded, formalin-fixed tumor
tissue samples by phenol-chloroform extraction as described previously
(29)
. We screened the lung tumor DNAs for p53alterations in exons 4–9 and 11 using DGGE. PCR-DGGE of exons
5–9 of the p53 gene was performed as described previously
(29
, 30) . For exon 4, the PCR primers were
5'-CGCCCGCCGCGCCCCGCGCCCGTCCCGCGCCCCCCGCCCGCGGCCCCTGCACCAGCCCCCTC-3'
and 5'-GCAACTGACCGTGCAAGT-3'. For exon 11, the 5' primer was
5'-CTCCCTGATTATGTCTCC-3', and the 3' primer with GC-clamp was
5'-CGCCCGCCGCGCCCCGCGCCCGTCCCGCCGCCCCCGCCCGTCAGT GGGGAACAAGAAG-3'.
The p53 mutations were identified using Sequenase Version
2.0 (United States Biochemical) and Thermo Sequenase radiolabeled
terminator cycle sequencing (Amersham Life Science, Inc.) with primer
designs as described previously (29
, 30)
.
For DGGE runs, both positive and negative controls were included. The strict criterion used for a positive finding in DGGE or in sequence analysis was that the same band pattern or sequence alteration must be detected in at least two independent PCR amplification products from the original tumor DNA. We had three mutation-positive cases for which a clear, positive DGGE result (mutation homoduplex band, wild-type homoduplex band, and the corresponding two heteroduplex bands) was detected repeatedly in several independent PCRs, but sequencing yielded a readable sequence alteration only once; these DNAs remained in the final results as DGGE-positive DNAs. The difficulties met in sequencing PET samples were not entirely unexpected because cross-linking and fragmentation of DNA caused by tissue fixation are known to cause problems. DGGE screening of p53 mutations, in contrast, has been proven to be a very powerful technique, with a greater sensitivity than many other screening techniques (31 , 32) or direct sequencing (33) , as also experienced in our laboratory (29 , 34) . Furthermore, the sensitivity of DGGE and its versions is not diminished by the presence of wild-type sequences; on the contrary, the presence of both types of sequences results in formation of mutation-wild-type heteroduplexes, which increase the sensitivity (31 , 35) . Reliability and reproducibility of the present mutation analysis strategy were also shown by independent detection of the same mutated sequence or DGGE band pattern from more than one tissue block from the same tumor. All exon 6 (codon 213) polymorphisms of the p53 gene were distinguished from somatic mutations by DGGE (36) and sequencing, and exon 4 (codon 72) polymorphism was excluded by primer design.
Statistical Analysis.
Tobacco consumption was expressed as cigarette equivalents/day (1 cigar
or cigarillo = 2 cigarette equivalents; 1 pipe = 4 cigarette equivalents). Ex-smokers were defined as people who
stopped smoking at least 1 year prior to diagnosis. The average daily
consumption was calculated by dividing the cumulative lifetime tobacco
consumption by the overall duration of smoking. Cumulative consumption
was expressed as pack-years (1 pack = 20 cigarette
equivalents; 1 pack-year = 1 pack/day for 1 year).
Exposure to ETS from the spouse and at the workplace was quantified as
weighted duration (number of hours/day x years of
exposure).
ORs of p53 mutation and the corresponding 95% CIs were
calculated by unconditional logistic regression using the GLIM
statistical package (37)
. All risk estimates were adjusted
for age (
60 years, 61–70 years, and >70 years), gender, center
(Northern Europe, Western Europe, and Eastern Europe plus Brazil), and
histology (adenocarcinoma versus other types of cancer).
Quantitative variables related to tobacco consumption and ETS exposure
were dichotomized at the median in the total population so that
sufficient numbers of individuals were included in each subgroup.
Numbers in the analysis of the type of mutation were too sparse to
permit a stringent statistical analysis.
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. The study included 91 (58%) cases who were never-smokers and 66
(42%) cases who were smokers (42 current smokers and 24 ex-smokers;
mean number of pack-years, 34.5; SD = 22.9). The
subjects were mostly women (111 cases; 71%). Adenocarcinoma was the
most common histological type (89 cases; 57%), followed by squamous
cell carcinoma (35 cases; 22%).
|
. Among smokers, 17 mutation-positive cases (26%) were
detected (Table 2)
. Among smoking cases from the different study areas,
p53 mutations were seen in 13 of 42 cases (31%) from
Northern Europe, 4 of 14 cases (29%) from Western Europe, and 0 of 10
cases (0%) from Eastern Europe and Brazil. Among never-smokers,
p53 mutations were seen in 5 of 57 cases (8.8%) from
Northern Europe, 1 of 9 cases (11%) from Western Europe, and 3 of 25
cases (12%) from Eastern Europe and Brazil, respectively. No
significant difference was observed between these study areas
(P = 0.11 for smokers; P = 0.87 for never-smokers). The multivariate analysis showed that
smokers had a 3-fold (OR, 2.9; 95% CI, 1.2–7.2) elevated risk of
p53 mutation as compared with all never-smokers combined
(i.e., irrespective of ETS exposure; Table 2
). The risk was increased for both former smokers (OR, 3.4; 95% CI,
1.0–11) and current smokers (OR, 2.6; 95% CI, 1.0–7.3). When
compared with never-smokers without ETS exposure, the risk of mutation
in smokers was 4.4-fold (95% CI, 1.2–16). The risks associated with a
higher number of pack-years or a longer duration of smoking were
slightly higher than those in the corresponding groups with lower
exposure; however, the differences were not significant (Table 2)
. The PET samples from three centers (Bucharest, 4 never-smokers and
10 smokers; Poznan, 4 never-smokers and 0 smokers; and Bremen, 1
never-smoker and 0 smokers) were of a somewhat lower quality,
possibly resulting in nondifferential underestimation of mutations.
When these samples were excluded, the risk of p53mutation was 4.1 (95% CI, 1.5–12) for ever smokers, 3.7 (95%
CI, 1.1–12) for smokers who had smoked for less than 30 pack-years,
and 4.8 (95% CI, 1.4–17) for those who had smoked for 30 pack-years
or more, as compared with never-smokers.
|
. The frequency of p53 mutations in these ever-exposed
never-smoker groups was 13%, 8%, and 10% respectively. The risk of
p53 mutation was doubled among never-smoking subjects ever
exposed to spousal ETS as compared with never-smokers who were never
exposed to ETS (OR, 2.0; 95% CI, 0.5–8.7; Table 3
). The risk
associated with ETS exposure at the workplace was 0.9 (95% CI,
0.2–5.7), and the risk associated with spousal or workplace ETS
exposure was 1.5 (95% CI, 0.2–8.8; Table 3
). The risk estimates for
ETS exposure were statistically unstable. In further analyses by
duration of ETS exposure, no clear trend was observed. Exposure to ETS
in childhood was not related to p53 mutations (data not
shown). The risks for ETS-exposed never-smokers were practically
unchanged after exclusion of the three centers that had lower quality
samples.
|
. In total, successful sequencing of 23 mutations demonstrated 20
missense (87%), 2 nonsense (9%), and 1 silent (4%) p53mutation. The most common type of base substitution was G:C to A:T
transition, which was the predominant type seen in never-smokers (four
of seven mutations). G:C to T:A transversion was the second most common
type of mutation, with the majority of these seen in smokers (Table 5)
.
|
|
. With regard to tumor histology, 22 of the 26 mutations detected had occurred in the two main histological types, adenocarcinoma or squamous cell carcinoma, and only 4 mutations had occurred in the other cell types. The frequency of mutation-positive adenocarcinomas was 23% (8 of 35 cases) in smokers, 10% (3 of 30 cases) in never-smokers exposed to spousal ETS, and 8% (2 of 24 cases) in never-smokers without spousal ETS exposure. In squamous cell carcinomas, frequencies were 38% (8 of 21 cases), 13% (1 of 8 cases), and 0% (0 of 6 cases) in the corresponding categories.
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Literature data on p53 mutations in lung tumors from lifelong nonsmokers are sparse. In recent works (38, 39, 40, 41) , the mutation frequencies observed in nonsmokers (20–28%) have generally been lower than those seen in smokers but still higher than those reported here. The observed differences in mutation frequencies might be explained by the more stringent definition of never-smoker used in our study in comparison with some of the previous investigations. Additionally, it may be possible that the lower quality DNAs have affected the number of mutations detected. Exclusion of those cases gave a mutation frequency of 11% (9 of 82 cases) for never-smokers and 30% (17 of 56 cases) for smokers. Influence from this source is unlikely because the frequencies come close to the respective values from the total study population. Our prevalences are in good agreement with findings from a recent study from Sweden (42) . An overall mutation frequency of 13.8% (9 of 65 cases) was observed in never-smokers, and a frequency of 7.9% (3 of 38 cases) was observed in never-smokers with low level domestic radon exposure. For adenocarcinoma, the predominant cell type in never-smokers, mutation frequency was 13.2% in never-smokers and was as low as 14.6% in smokers (42) .
The prevalence of p53 mutations observed in smokers was 26%. Typically, mutation prevalence close to 50% has been observed in association with long-term heavy smoking, with lower prevalences reported in some studies, especially for lighter smokers (25 , 26) . In our study, half of the smokers had a lifetime cigarette smoke exposure of <30 pack-years, and the moderate cigarette consumption may have influenced the mutation frequency. This is supported by the observation that smokers with higher pack-years or longer duration of smoking had slightly higher risks of mutation. Another point may be the high proportion of women (61%) and adenocarcinomas (52%), which was due to the effort to make the case series of smokers comparable with that of never-smokers. Mutations have been reported to be less frequent in adenocarcinoma than in squamous cell carcinoma or small cell carcinoma of the lung in many studies (26 , 42) , including our own previous work (34) and the present study.
Polynuclear aromatic hydrocarbons and the tobacco-specific nitrosamines NNK and N'-nitrosonornicotine are the major known lung carcinogens present in mainstream tobacco smoke and ETS. G:C to T:A transversions in p53 are considered hallmarks of smoking-related lung carcinomas (24 , 25) and are assumed to arise as a direct consequence of benzo(a)pyrene diol epoxide-DNA adducts (43) . In keeping with this, five of six G:C to T:A transversions observed in this study occurred in smokers. The only such transversion in never-smokers occurred in a case without past exposure to ETS. In ETS, NNK and other tobacco-specific nitrosamines form an important class of lung carcinogens, with concentrations 1–2 orders of magnitude higher in undiluted ETS as compared with those in mainstream smoke (4) . Despite the fact that nonsmokers are exposed to ETS, which is strongly diluted, measurable amounts of NNK and its metabolites have been detected in ETS-exposed nonsmokers (19) . Animal data have demonstrated that NNK is a strong lung-specific carcinogen (44) that induces predominantly G:C to A:T transitions in treated animals and other experimental systems (19 , 45 , 46) . In our data, 57% of the mutations identified in lifelong nonsmokers were G:C to A:T transitions. The prevalence in smokers was 25%. The present data on the predominance of G:C to A:T transitions in never-smokers is supported by observations from other studies and the international p53 database (28 , 38, 39, 40, 41) .
Among both smokers and never-smokers, half of the G:C to A:T substitutions observed had occurred at CpG dinucleotide sites. The major mutational hot spots in human cancers occur at CpG sequences in the p53 gene. It is generally presumed that the majority of G:C to A:T mutations at these sites result from the endogenous deamination of methylated cytosine residues, proposed as molecular markers of endogenous mutagenesis processes (25 , 47) . Alternatively, mutational hot spots at methylated CpG sequences in the p53 gene may be a consequence of preferential carcinogen binding at these sites (48 , 49) .
Of the typical hot spot codons of lung cancer, three were mutated in never-smokers (codons 248, 249, and 273), and two were mutated in smokers (codon 249 and 282). The most frequently mutated site, with two mutations in smokers and one in never-smokers, was codon 176, one of the mutation hot spots of human cancer other than lung cancer (27) . Codon 176 mutations have been found by a Chinese study (50) to be overrepresented in esophageal cancer, a cancer type associated with tobacco and alcohol use, but the mutational specificity has been contradicted (51) .
Our study was of a multicentric nature, which was required to obtain a sufficiently large series of never-smoking cases. Such an approach may introduce confounding by variables associated with study center, including genetic susceptibility factors and environmental factors such as diet. We approached this problem by adjusting for study center in the multivariate analyses, although it should be recognized that such an adjustment might have been imperfect in the analysis including both smoking and never-smoking cases because of the uneven distribution of smokers and never-smokers across study centers. As another way to tackle the problem, we carried out analyses restricted to the two study areas (Sweden and France) that contributed the majority of the subjects (63 never-smoking and 53 smoking cases). The mutation frequencies were 9.5% for never-smokers and 30% for smokers, not much different from those in the total population. The restricted analysis yielded a 4-fold (95% CI, 1.3–13) risk of p53 mutation for smokers as compared with all never-smokers combined and indicated an OR of 2.2 (95% CI, 0.3–19) for mutation in nonexposed never-smokers in comparison to the ETS-exposed ones. In all, our primary aim to obtain a representative series of never-smoker cases and select smokers to match the never-smokers on gender and age may have resulted in underestimation of the prevalence of p53 mutation in smokers. Therefore, the results may not be generalizable to a population series of lung cancer cases from the study areas.
Use of biomarkers may obviously be one of the few ways available to strengthen the association between lung cancer risk and exposure to ETS observed in epidemiological studies (12 , 52) . Our study indicated a 3–4-fold increased risk of p53 mutation in smokers, and it suggests that mechanisms of lung carcinogenesis in ETS-exposed never-smokers include mutations in the p53 gene, similar to that in smokers. In addition, differences other than those related to exposure levels are likely to exist in carcinogenic exposures and/or mechanisms between smokers and never-smokers, as suggested by the difference observed in mutation patterns. In conclusion, our work suggests that p53 mutation could serve as a sensitive marker of tobacco-related carcinogenesis, but further confirmation is needed due to statistically unstable results. Clearly, additional investigations of the role of p53 mutation as a biomarker are indicated.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 Supported by the European Commission, DG-XII,
Grant EV5V-CT94–0555; the Ministry of Environment, Finland, Drno
17/742/94; and the Association for Cancer Research (Villejuif,
France). 
2 To whom requests for reprints should be
addressed, at Laboratory of Molecular and Cellular Toxicology, Finnish
Institute of Occupational Health, Topeliuksenkatu 41 aA, FIN-00250
Helsinki, Finland. Phone: 358-9-47472212; Fax: 358-9-47472208; E-mail: Kirsti.Husgafvel-Pursiainen{at}occuphealth.fi
3 The abbreviations used are: ETS, environmental
tobacco smoke; PET, paraffin-embedded tissue; DGGE, denaturing gradient
gel electrophoresis; OR, odds ratio; CI, confidence interval; NNK,
4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone.
Received 10/25/99. Accepted 4/ 3/00.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  W. Zhou, R. S. Heist, G. Liu, K. Asomaning, D. P. Miller, D. S. Neuberg, J. C. Wain, T. J. Lynch, and D. C. Christiani Second Hand Smoke Exposure and Survival in Early-Stage Non-Small-Cell Lung Cancer Patients Clin. Cancer Res., December 1, 2006; 12(23): 7187 - 7193. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Le Calvez, A. Mukeria, J. D. Hunt, O. Kelm, R. J. Hung, P. Taniere, P. Brennan, P. Boffetta, D. G. Zaridze, and P. Hainaut TP53 and KRAS Mutation Load and Types in Lung Cancers in Relation to Tobacco Smoke: Distinct Patterns in Never, Former, and Current Smokers Cancer Res., June 15, 2005; 65(12): 5076 - 5083. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Li, E. M. Sturgis, L.-E. Wang, R. M. Chamberlain, C. I. Amos, M. R. Spitz, A. K. El-Naggar, W. K. Hong, and Q. Wei Association of a p73 exon 2 G4C14-to-A4T14 polymorphism with risk of squamous cell carcinoma of the head and neck Carcinogenesis, October 1, 2004; 25(10): 1911 - 1916. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. A. Snyder, E. R. Bertone, R. M. Jakowski, M. S. Dooner, J. Jennings-Ritchie, and A. S. Moore p53 Expression and Environmental Tobacco Smoke Exposure in Feline Oral Squamous Cell Carcinoma Vet. Pathol., May 1, 2004; 41(3): 209 - 214. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W.-M. Gao, M. Romkes, R. D. Day, J. M. Siegfried, J. D. Luketich, H. H. Mady, M. F. Melhem, and P. Keohavong Association of the DNA repair gene XPD Asp312Asn polymorphism with p53 gene mutations in tobacco-related non-small cell lung cancer Carcinogenesis, October 1, 2003; 24(10): 1671 - 1676. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Leffondre, M. Abrahamowicz, J. Siemiatycki, and B. Rachet Modeling Smoking History: A Comparison of Different Approaches Am. J. Epidemiol., November 1, 2002; 156(9): 813 - 823. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K Husgafvel-Pursiainen Molecular biomarkers in studies on environmental cancer J. Epidemiol. Community Health, October 1, 2002; 56(10): 730 - 731. [Full Text] [PDF]
|
|
|
|
 |
|
 M. Miyaki, T. Iijima, R. Ishii, Y. Kita, M. Koike, T. Kuroki, and T. Mori Increased Frequency of p53 Mutation in Sporadic Colorectal Cancer from Cigarette Smokers Jpn. J. Clin. Oncol., June 1, 2002; 32(6): 196 - 201. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. M. DeMarini, S. Landi, D. Tian, N. M. Hanley, X. Li, F. Hu, B. C. Roop, M. J. Mass, P. Keohavong, W. Gao, et al. Lung Tumor KRAS and TP53 Mutations in Nonsmokers Reflect Exposure to PAH-Rich Coal Combustion Emissions Cancer Res., September 1, 2001; 61(18): 6679 - 6681. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. H. Vahakangas, W. P. Bennett, K. Castren, J. A. Welsh, M. A. Khan, B. Blomeke, M. C. R. Alavanja, and C. C. Harris p53 and K-ras Mutations in Lung Cancers from Former and Never-Smoking Women Cancer Res., June 1, 2001; 61(11): 4350 - 4356. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. E. Anderson, S. G. Carmella, M. Ye, R. L. Bliss, C. Le, L. Murphy, and S. S. Hecht Metabolites of a Tobacco-Specific Lung Carcinogen in Nonsmoking Women Exposed to Environmental Tobacco Smoke J Natl Cancer Inst, March 7, 2001; 93(5): 378 - 381. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Hainaut and G. P. Pfeifer Patterns of p53 G{->}T transversions in lung cancers reflect the primary mutagenic signature of DNA-damage by tobacco smoke Carcinogenesis, March 1, 2001; 22(3): 367 - 374. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
