Annals of Oncology Advance Access originally published online on September 25, 2007
Annals of Oncology 2008 19(1):109-114; doi:10.1093/annonc/mdm395
lung cancer |
Analysis of the relationship between p53 immunohistochemical expression and risk factors for lung cancer, with special emphasis on residential radon exposure
1 Department of Preventive Medicine and Public Health, University of Santiago de Compostela, Santiago de Compostela
2 Galician Agency for Health Technology Assessment, Galician Department of Health Santiago de Compostela
3 Department of Pathological Anatomy, Santiago de Compostela Clinical University Hospital, Santiago de Compostela Spain
4 Department of Genetics and Complex Diseases, Harvard School of Public Health, Harvard University Boston, MA, USA
5 Preventive Medicine Service, Santiago de Compostela Clinical University Hospital, Santiago de Compostela Spain
* Correspondence to: Prof. A. Ruano-Ravina, Department of Preventive Medicine and Public Health, School of Medicine, C/San Francisco s/n, University of Santiago de Compostela, CP 15782 Santiago de Compostela, Spain. Tel: +34-981-581237; Fax: +34-981-572282; E-mail: mralbert{at}usc.es
 |
Abstract |
|---|
 Top Abstract introductionsubjects and methodsresultsdiscussionfundingAcknowledgementsReferences |
|---|
Background: Indoor radon exposure has been postulated as the second risk factor of lung cancer after tobacco. The objective of this work is to analyze if there exists any effect on p53 immunohistochemical expression mainly due to radon exposure and other risk factors for lung cancer.
Patients and methods: The tumor samples of a case series of 163 lung cancer cases were analyzed to know the p53 staining. The staining was classified into four categories from no staining to intense staining (>60%). This staining was correlated with radon exposure, tobacco consumption, having worked in risk occupations for lung cancer and alcohol consumption.
Results: Only 72 samples could be analyzed for immunohistochemistry and some of these samples were sequenced from exons 4–8. No association was observed for staining intensity and radon exposure and also for tobacco and occupation. A slight association with a more intense staining was observed for high alcohol intake. In the four samples with a staining >60% that could be sequenced from exons 4 to 8, no mutation was observed in the p53 gene.
Conclusion: There is no association between radon exposure and p53 expression, indicating that maybe the effect of radon is not mediated through p53 alterations.
Key words: gene p53, immunohistochemistry, lung neoplasms, radon
|
introduction |
|---|
|
TopAbstractintroductionsubjects and methodsresultsdiscussionfundingAcknowledgementsReferences |
|---|
Lung cancer is one of the leading causes of mortality nowadays [1, 2]. Its main risk factor is tobacco followed by radon exposure. Radon has been established as carcinogenic by the International Agency for Research on Cancer since 1988 [3]. Residential radon exposure is thought to participate in
10%–20% of all lung cancer cases and this is the reason why authorities in various countries have established action levels that should not be surpassed at homes [4, 5]. Radon is an odorless, colorless and insipid gas with a half-life of 3.8 days. Its daughters (Po 218, Pb 214, Bi 214 and Po 214), with a much shorter half-life, emit alpha radiation that can penetrate into bronchial cells when inhaled. The exact mechanism by which alpha radiation leads finally to lung cancer is not exactly known, although there are various hypotheses. Some of them indicate that even one alpha particle is sufficient to alter one cell and its neighbors through mechanisms of cellular communication [6].
One of the key elements of cell regulation is the p53 gene. Risk factors for lung cancer such as tobacco or occupation have shown to have a direct effect on this gene by causing mutations [7–12]. Nevertheless, most of the studies relating radon and p53 gene have been carried out on miners and only two in indoor radon-exposed individuals [12, 13]. Immunohistochemistry (IHC) has revealed itself as a technique that can identify alterations in a protein that could reflect a gene alteration (i.e. a mutation) with a relatively good performance since the protein measured should indicate the genetic integrity or at least show modifications in its expression if the gene is altered.
The objective of this study is to investigate if factors associated with the appearance of lung cancer, with special emphasis on residential radon exposure, do have an effect on p53 immunohistochemical expression.
|
subjects and methods |
|---|
|
TopAbstractintroductionsubjects and methodsresultsdiscussionfundingAcknowledgementsReferences |
|---|
origin of the case series
The samples analyzed belong to a case–control study carried out from 1992 to 1995 in the Santiago de Compostela Health District Area (Galicia, NW of Spain) which has been published elsewhere [14]. Briefly, this study showed a risk of 2.96 [confidence interval (CI) 95% 1.29–6.79] for those exposed to radon levels >148 Bq/m3 versus nonexposed and an interaction between radon exposure and smoking was observed. This study was included in the European Pooling Study published by Darby et al. [15], which showed a linear effect between radon exposure and lung cancer. The cases were 163 individuals with anatomopathologically diagnosed lung cancer and were collected consecutively at the reference hospital along this period.
information collected and radon measurements
A detailed interview was done, and the case was asked about some aspects of his/her lifestyle and risk factors of lung cancer. Detailed information was collected on tobacco consumption history and also on previous jobs in occupations posing a risk for lung cancer according to a classification published by Ahrens and Merletti [16]. Alcohol consumption took into account beer, wine and spirits and their alcohol content was added in one variable (alcohol consumption in ml/week). A radon detector was placed in each house of the included cases. It was placed under standard conditions, namely, in the room in which study subjects spent most of the time at home (generally the bedroom) away from doors and windows and never <65 cm from the floor. All detectors were in place in the subjects’ homes for a minimum period of 90 days. The detectors were of an alpha-track type and were sent to the United States to be revealed by an Enviromental Protection Agency, United States (EPA) approved laboratory.
immunohistochemistry and DNA sequencing
All the biopsies of the included cases were retrieved from the files of the Department of Pathological Anatomy, Santiago de Compostela Clinical University Hospital, which is the reference hospital in the study area. Immmunohistochemical staining of paraffin-embedded slides was carried out using the Envision method for detection and visualization, which consists of a dextran polymere with all the enzyme molecules joined directly. In this case, peroxidase was used (EnVision-horseradish peroxidase). The slides were deparaffinized and hydrated into water after being cut in 4 µm pieces. The slides were placed in special slides from DakoCytomation (Glostrup, Denmark) (Capillary Gap Microscope Slides, 75 µm, ref: S2024) that were washed with different alcohol solutions. After that, the immunohistochemistry began with an automatic process (Dako TECHMATE TM500 Glostrup Denmark). A mAb specific for p53 was used (Novocastra Laboratories, clon: D07, ref: NCLp53D07) with an optimal dilution of 1/20. The percentage of positively staining tumor cells was scored as follows: negative result (no positive cells); scattered cells, focal, until 10%; moderate positivity (between 11% and 60%); and diffuse positivity (>60%).
The most positive stained samples (>60%) were sequenced in searching for mutations on the exons 4–8 of the p53 gene. Samples were extracted with KIT QUIAamp DNA MICRO (Qiagen, Hilden, Germany) and a nested PCR was carried out.
statistical analysis
Bivariate and multivariate analyses were carried out. An analysis of variance was carried out when analyzing the intensity of staining with radon concentration, tobacco consumption in pack-years, years worked in a risk occupation for lung cancer and alcohol consumption in ml/week. In the case that the variances were not homogeneous, the Kruskal–Wallis test was applied. A further analysis consisted in comparing the samples with the most extensive staining with the remaining. This last comparison was done with the Student's t-test when applicable.
A logistic regression was carried out with the former variables plus the histological type. In order to improve the statistical power of the regression, the staining was divided into two categories, <10% of the intensity (no staining) and >10% (positive staining). The predictive variables were introduced as dichotomous in the following way: radon, exposed and nonexposed with the cut point on the median concentration; tobacco, with the cut point on the median of the pack-years; years worked in risk occupations for lung cancer, divided in not having worked or having worked for 1 year or more in one of these occupations, alcohol consumption, with the cut point on the median and finally the histological types with squamous type as the reference category. All the odds ratios were calculated with CI 95% and the calculations were done with SPSS 11.5 statistical package.
|
results |
|---|
|
TopAbstractintroductionsubjects and methodsresultsdiscussionfundingAcknowledgementsReferences |
|---|
sample characteristics
From the original 163 cases of the case–control study, 72 were analyzed for immunohistochemistry. The reasons for not having the whole sample analyzed can be found in Table 1, along with other characteristics of these 72 subjects. The main reason for not having all samples analyzed was the small size of the biopsies that resulted in lack of neoplasic tissue after routine stainings for diagnosis and in several artifacts. There were no statistically significant differences in radon concentration between the analyzed samples and those not analyzed (t-test; P = 0.170).
|
Respecting to the analyzed samples in the present paper, the more prevalent histological type was squamous cell and the mean age was 65 years. Most of the cases were males. Radon concentration was quite high, with
20% of the cases >148 Bq/m3, the action level established by the EPA. Alcohol consumption was also moderately high, with the median consumption in the sample equivalent to two daily glasses of wine approximately.
immunohistochemical results
The immunohistochemical results showed that 42% of the samples had a staining <10% while the remaining had a most intense staining, with a 22% of the samples with an extensive staining (Figure 1A and B). When analyzing separately those samples with a radon concentration >200 Bq/m3, no pattern was observed for the immunohistochemistry (data not shown). In the Table 2 is displayed the relationship between the histological type and p53 staining. For none of the histological types a clear pattern of staining is observed with the exception of small cells, where it seems to exist a more intense staining.
|
|
The results of the bivariant analysis are shown in Table 3. None of the analyzed variables has a statistical association with the staining intensity. For radon concentration there is no clear trend. Although there is no statistical significance, some type of trend can be observed for years worked in a risk occupation for lung cancer and for alcohol consumption in order that a more intense staining is observed for a longer time worked in risk occupations and for a higher alcohol intake. A further analysis was carried out by comparing if those subjects with the most intense staining (>60%) had had different exposures than that of the remaining. The results are shown in Table 4. It can be observed that there are no differences for radon exposure, tobacco or occupation and a statistically significant difference for alcohol consumption, although some trend is observed for tobacco and occupation.
|
|
The results of the logistic regression considering p53 immunohistochemistry as a dichotomic result (<10% and
10%) can be observed in Table 5. None of the five independent variables had a statistical significant association with the intensity of p53 staining.
|
DNA amplification results
Only 4 of the 16 samples with this most extensive staining could be amplified and sequenced with a reliable result. No mutations were found in exons 4 through 8 in the p53 gene in any of these samples. The radon concentrations of these four subjects ranged between 11.1 and 51.8 Bq/m3.
|
discussion |
|---|
|
TopAbstractintroductionsubjects and methodsresultsdiscussionfundingAcknowledgementsReferences |
|---|
The most interesting hypothesis of this paper was to know the possible association between residential radon exposure and alterations on the p53 gene expression by immunohistochemistry. Any type of relationship was not found between radon exposure and p53 staining, even for the most stained samples. Furthermore, no mutations were found in the samples analyzed. Some type of trend has been found for occupation and for alcohol consumption and the extent of staining but with statistical significance only for alcohol consumption, probably due to the small sample included.
To our knowledge, only seven studies (eight with the present one) have analyzed the relationship between radon exposure and alterations on the p53 gene [7, 12, 13, 17–20]. Of these, only two have analyzed residential radon exposure [12, 13].
Miners’ and residential radon studies have not shown a clear pattern of p53 mutations or immunohistochemistry. While some authors postulate that a radon-specific mutation can occur in codon 249 of the exon 7 [17], others do not support this affirmation [7, 12]. The results of the different studies are discrepant and the present study do not show an association between radon and p53. It should be highlighted that the study area has been classified as a high emission area by previous studies [14, 21] since 20% of dwellings exceed the EPA's action level and even though an association has not been observed. Furthermore, although only four samples with the highest staining could be adequately sequenced, none of them showed a mutation in exons 4–8, although neither of these cases had a radon concentration at home >60 Bq/m3 (data not shown).
The mechanisms of radon carcinogenesis are not clear, although the p53 gene, as a key factor in the regulation of the cell cycle possibly plays a role. There could be other factors affecting the p53 expression without needing a mutation or there could be alterations in other genes implicated in carcinogenesis such as tumor suppressor genes. The inactivation of p53 can be due to mutations caused by exposure to environmental mutagens, but it can also be occasionally inactivated through endogenous processes and, in some rare cases, mutations can be hereditary. Some authors indicate that a stabilization of the p53 protein can occur after an oxidative stress, which conducts to lipid peroxidation and could cause this mutation, although this is not clear [22]. The distribution of mutations along the p53 gene is not randomized. They are localized mainly in mutational hot spots placed in a highly preserved DNA sequence.
A G
T transversion on the second position of codon 249 (AGGATG, arginine to methionine) is the most prevalent nonsense mutation in lung cancer. This mutation rarely occurs in other cancers [23] but is frequent in cancers of smokers and nonsmokers and could be due to DNA-polycyclic aromatic hydrocarbon adduct formation.
p53 protein in the cell is tightly regulated, with negligible antibody staining noted in normal tissues. However, inhibition of degradation, mutation or enhanced production of p53 protein (or some combination of these) could all result in antibody recognition of the protein and thus abnormal staining [24]. The p53 can be functionally inactive and not have mutations due to alterations in the mechanisms conducting to its activation, i.e. protein MDM2.
There is a dose dependence of p53 mutation frequency according to the number of cigarettes smoked [7]. This has been found here for the IHC, although the result is not statistically significant. In some cases, the p53 gene shows a mutational spectrum specific for certain carcinogens and sometimes the expression of the p53 protein is inducted as a response to a harm to DNA and is implicated in DNA repair and in apoptosis.
In the study of Yngveson et al. [12], 41% of the samples showed reactivity and the concordance between IHC and sequenciation was 62% and seemed not affected by tobacco or radon exposure. Sensitivity and specificity of IHC were estimated in 57% and 64%, respectively. Other studies have shown a dose–response relationship between radon concentration and p53 prevalence of mutations.
The fact that a unique alpha particle goes through a mammalian cell is sufficient to induce biological consequences. Alpha radiation has an ionization density that coincides with the diameter of the DNA double helix and then has a high probability of causing a break of this double helix [6]. Zhou et al. [25] have pointed that the risks due to alpha particle exposure cannot be extrapolated linearly. The risks are higher, in proportion, for low exposures than for high exposures, although these results belong to an experimental study.
With respect to the relationship among p53 alterations and other exposures, such as tobacco, alcohol or occupation, the association with tobacco use is clear [26]. To our knowledge, this association has not been analyzed for alcohol consumption or occupation and p53 alterations in lung cancer. An overexpression of p53 has been observed for alcohol consumption in oesophageal cancer [27] while alcohol consumption has not been associated with p53 mutations in oral cancer [28].
This study has several limitations. The most important is the low number of samples analyzed for direct p53 gene sequencing. We have reliable information about p53 immunohistochemical expression, but this is an indirect or approximative way to know about p53 gene status. Another limitation is related with the sample size. Although there were initially 163 subjects, these were reduced to 72 due to problems mainly with sample conservation or availability. Those paraffin-embedded samples were usually small sized, and after routine morphological and immunohistochemical studies for diagnosis, no remaining tumoral tissue was available. In addition, the smaller samples exhibited frequently artifacts that precluded immunohistochemical interpretation. However, nearly half of the sample could be finally analyzed. In any case, there were no significant differences in radon concentration between the analyzed and not analyzed samples. In this study, all subjects were interviewed immediately after the diagnosis and the biological samples were those used for the cancer diagnosis, so the progression of the disease should not have influenced the IHC results.
This study has also some advantages. The main one is its own objective, which has novelty mainly due to two reasons, the scarcity of published literature and the existing interest in knowing the carcinogenic mechanisms of radon residential exposure. Other advantage is that we were able to adjust the results for many potential confounders on the effect of gene p53 such as tobacco or occupation. This fact, together with the low sample has contributed to the wide CIs obtained in the logistic regression. Another advantage is that the detectors used were long-term radon detectors, which provide a reliable measure for radon exposure along with the fact that most of the subjects included had lived for >20 years in the same dwelling.
As a conclusion, no association has been found between residential radon exposure and p53 expression through p53 immunohistochemistry in this case series. Furthermore, no mutations were found in the four samples with the most intense staining. A slight association with a more intense staining has been found for occupation and alcohol consumption. These results show that radon exposure might not exert its carcinogenic effect through p53 gene, although more studies searching for mutations in this gene are needed. Again, more studies are required in order to understand the carcinogenic mechanisms of residential radon exposure and a good way would be to analyze biological samples of the subjects included in other published residential radon case–control studies.
|
funding |
|---|
|
TopAbstractintroductionsubjects and methodsresultsdiscussionfundingAcknowledgementsReferences |
|---|
Galician Health Service (Spain) (PGIDIT02SAN91802); CIBER Epidemiología y Salud pública (CIBERESP), Spain (to A. R.-R. and J. M. B.-D.).
|
Acknowledgements |
|---|
|
TopAbstractintroductionsubjects and methodsresultsdiscussionfundingAcknowledgementsReferences |
|---|
This work has been done while the first author was doing a research stay at the Harvard School of Public Health. The authors declare no conflict of interests.
Received for publication March 17, 2007. Revision received July 10, 2007. Accepted for publication July 10, 2007.
|
References |
|---|
|
TopAbstractintroductionsubjects and methodsresultsdiscussionfundingAcknowledgementsReferences |
|---|
1. Parkin DM, Bray F, Ferlay J, Pisani P. Global cancer statistics, 2002. CA Cancer J Clin (2005) 55:74–108.
2. Jemal A, Siegel R, Ward E, et al. Cancer statistics, 2006. CA Cancer J Clin (2006) 56:106–130.
3. Radon. IARC Monogr Eval Carcinog Risks Hum (1988) 43:173–259.[Medline]
4. EPA. EPA Assessment of Risk from Radon in Homes. EPA 402-R-03–003, United States Environmental Protection Agency. United States Government (2003).
5. Euratom. 90/143/Euratom: Commission Recommendation of 21 February 1990 on the Protection of the Public Against Indoor Exposure to Radon (1990) http://eur-lex.europa.eu/LexUriServ/LexUriserv.do?uri=CELEX:31990H0143:EN:HTML (09 August 2007, date last accessed).
6. Hall EJ, Hei TK. Genomic instability and bystander effects induced by high LET radiation. Oncogene (2002) 22:7034–7042.[Web of Science]
7. Hainaut P, Pfeifer GP. Patterns of p53 GT transversions in lung cancers reflect the primary mutagenic signature of DNA-damage by tobacco smoke. Carcinogenesis (2001) 22(3):367–374.
8. Hainaut P, Hernandez T, Robinson A, et al. IARC database of p53 gene mutations in human tumors and cell lines: updated compilation, revised formats and new visualization tools. Nucleic Acids Res (1998) 26(1):205–213.
9. Vähäkangas K. Molecular epidemiology of human cancer risk. Methods Mol Med (2003) 74(1):43–59.[Medline]
10. Vähäkangas K. TP53 mutations in workers exposed to occupational carcinogens. Hum Mutat (2003) 21:240–251.[CrossRef][Web of Science][Medline]
11. Vahakangas KH, Samet JM, Metcalf RA, et al. Mutations of p53 and ras genes in radon-associated lung cancer from uranium miners. Lancet (1992) 339:576–580.[CrossRef][Web of Science][Medline]
12. Yngveson A, Williams C, Hjerpe A, et al. p53 mutations in lung cancer associated with residential radon exposure. Cancer Epidemiol Biomarkers Prev (1999) 8:433–438.
13. Lo YM, Darby S, Noakes L, et al. Screening for codon 249 p53 mutation in lung cancer associated with domestic radon exposure. Lancet (1995) 345:60.[Web of Science][Medline]
14. Barros-Dios JM, Barreiro Carracedo MA, Ruano-Ravina A, Figueiras A. Exposure to residential radon and lung cancer risk in Spain: a population-based case-control study. Am J Epidemiol (2002) 156(6):548–555.
15. Darby S, Hill D, Auvinen A, et al. Radon in homes and risk of lung cancer: collaborative analysis of individual data from 13 European case-control studies. Br Med J (2005) 330:223–226.
16. Ahrens W, Merletti F. A standard tool for the analysis of occupational lung cancer in epidemiologic studies. Int J Occup Environ Health (1998) 4(4):236–240.[Medline]
17. Taylor JA, Watson MA, Deveraux TR, et al. P53 mutation hotspot in radon-associated lung cancer. Lancet (1994) 343:86–87.[CrossRef][Web of Science][Medline]
18. Hollstein M, Bartsch H, Wesch H, et al. p53 gene mutation analysis in tumors of patients exposed to 
-particles. Carcinogenesis (1997) 18(3):511–516.
19. Yang Q, Wesch H, Mueller K-M, et al. Analysis of radon-associated squamous cell carcinomas of the lung for a p53 gene hotspot mutation. Br J Cancer (2000) 82(4):763–765.[CrossRef][Web of Science][Medline]
20. McDonald J, Taylor J, Watson M, et al. P53 and K-ras in radon-associated lung adenocarcinoma. Cancer Epidemiol Biomarkers Prev (1995) 4(7):791–793.[Abstract]
21. Barros-Dios JM, Ruano-Ravina A, Gastelu-Iturri J, Figueiras A. Factors underlying residential radon concentration. Results from Galicia (Spain). Environ Res (2007) 103(2):185–190.[Medline]
22. Robles AI, Linke SP, Harris CC. The p53 network in lung carcinogenesis. Oncogene (2002) 21:6898–6907.[CrossRef][Web of Science][Medline]
23. Hernandez-Boussard TM, Hainaut P. A specific spectrum of p53 mutations in lung cancer from smokers: review of mutations compiled in the IARC p53 database. Environ Health Perspect (1998) 106(7):385–391.[Web of Science][Medline]
24. Kelsey KT, Hirao T, Schned A, et al. A population based study of inmunohistochemical detection of p53 alteration in bladder cancer. Br J Cancer (2004) 90:1572–1576.[CrossRef][Web of Science][Medline]
25. Zhou H, Suzuki M, Randers-Pehrson G, et al. Radiation risk to low fluences of alpha particles may be greater than we thought. Proc Natl Acad Sci USA (2001) 98(25):14410–14415.
26. Hagiwara N, Mechanic LE, Trivers GE, et al. Quantitative detection of p53 mutations in plasma DNA from tobacco smokers. Cancer Res (2006) 66(16):8309–8317.
27. Lee JM, Shun CT, Wu MT, et al. The associations of p53 overexpression with p53 codon 72 genetic polymorphisms in esophageal cancer. Mutat Res (2006) 594(1–2):181–188.[Web of Science][Medline]
28. Chaves AC, Cherubini K, Herter N, et al. Characterization of p53 gene mutations in a Brazilian population with oral squamous cell carcinomas. Int J Oncol (2004) 24(2):295–303.[Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||