Carcinogenesis vol.29 no.9 pp.18451849, 2008 doi:10.

1093/carcin/bgn169 Advance Access publication July 16, 2008

Aberrant promoter hypermethylation in serum DNA from patients with silicosis

Shigeki Umemura, Nobukazu Fujimoto1,, Akio Hiraki2, Kenichi Gemba1, Nagio Takigawa, Keiichi Fujiwara3, Masanori Fujii, Hiroshi Umemura4, Mamoru Satoh4, Masahiro Tabata, Hiroshi Ueoka5, Katsuyuki Kiura, Takumi Kishimoto6 and Mitsune Tanimoto
Department of Hematology, Oncology and Respiratory Medicine, Graduate School of Medicine, Dentistry and Pharmaceutical Sciences, Okayama University, Okayama 7008558, Japan, 1Department of Respiratory Medicine, Okayama Rosai Hospital, 1-10-25 Chikkomidorimachi, Okayama 702-8055, Japan, 2Division of Epidemiology and Prevention, Aichi Cancer Center Research Institute, Nagoya 4648681, Japan, 3Department of Respiratory Medicine, Okayama Medical Center, Okayama 7011192, Japan, 4Department of Molecular Diagnosis, Graduate School of Medicine, Chiba University, Chiba 2608670, Japan, 5Department of Medical Oncology, National Hospital Organization Sanyo Hospital, Ube 7550241, Japan and 6Department of Internal Medicine, Okayama Rosai Hospital, 1-10-25 Chikkomidorimachi, Okayama 702-8055, Japan
To whom correspondence should be addressed. Tel: 81 86 2620131; Fax: 81 86 2623391; Email: nfuji@okayamah.rofuku.go.jp

It is well established that patients with silicosis are at high risk for lung cancer; however, it is difcult to detect lung cancer by chest radiography during follow-up treatment of patients with silicosis because of preexisting diffuse pulmonary shadows. The purpose of this study is to evaluate the usefulness of detection of serum DNA methylation for early detection of lung cancer in silicosis. Serum samples from healthy controls (n 5 20) and silicosis patients with (n 5 11) and without (n 5 67) lung cancer were tested for aberrant hypermethylation at the promoters of the DNA repair gene O6-methylguanine-DNA methyltransferase (MGMT), p16INK4a, ras association domain family 1A (RASSF1A), the apoptosis-related gene death-associated protein kinase (DAPK) and retinoic acid receptor b (RARb) by methylation-specic polymerase chain reaction. Aberrant promoter methylation in at least one of ve tumor suppressor genes was detected more frequently in the serum DNA of silicosis patients with lung cancer than in that of patients without it (P 5 0.006). Furthermore, the odds ratio of having lung cancer was 9.77 (P 5 0.009) for those silicosis patients with methylation of at least one gene. Extended exposure to silica (>30 years) was correlated with an increased methylation frequency (P 5 0.017); however, methylation status did not correlate with age, smoking history or radiographic ndings of silicosis. These results suggest that testing for aberrant promoter methylation of tumor suppressor genes using serum DNA may facilitate early detection of lung cancer in patients with silicosis.

that inhaled crystalline silica is a human lung carcinogen (5). Chan et al. (6) reported that 33 (2.3%) of 1490 workers diagnosed with silicosis in Hong Kong died from lung cancer. Lung cancer spreads gradually throughout the body and has a poor prognosis because it is largely unresponsive to chemotherapy. Only those patients diagnosed at an early stage of the disease can be cured by complete resection. However, it is difcult to detect lung cancer by chest radiography or computed tomography during the follow-up treatment of patients with silicosis because of preexisting diffuse pulmonary shadows (7,8). A reliable clinical marker for early detection of lung cancer in patients with silicosis is thus urgently needed. Epigenetic changes such as hypermethylation increasingly appear to play a role in carcinogenesis. DNA methylation is one form of epigenetic variability in mammalian cells. As aberrant hypermethylation in the CpG-rich promoter regions of many tumor suppressor genes interferes with transcription, hypermethylation may contribute to the development and progression of various cancers by abolishing tumor suppressor gene function (911). Aberrant hypermethylation of tumor suppressor genes has been observed in various specimens from lung cancer patients, including serum, plasma, sputum, lavage uid and epithelial brushing (12,13). These ndings raise the possibility that methylation analysis of such materials may be a useful tool forcancer detection. A number of studies have reported increased methylation of several tumor suppressor genes in lung cancer cells, including the DNA repair gene O6-methylguanine-DNA methyltransferase (MGMT) (14,15), p16INK4a (16), ras association domain family 1A (RASSF1A) (17), apoptosis-associated genes such as deathassociated protein kinase (DAPK) (18) and retinoic acid receptor b (RARb) (19); in contrast, methylation of these genes is rarely reported in non-malignant lung tissue (14). We demonstrated previously that screening for promoter methylation in these ve genes using serum (20) and pleural uid (21) DNA can be useful for early detection of lung cancer. Thus, detection of tumor suppressor gene promoter hypermethylation in serum DNA may facilitate early detection of lung cancer in patients with silicosis. Here, we examined whether promoter hypermethylation could identify lung cancer in a clinical setting by testing the promoter methylation status of ve tumor suppressor genes in serum DNA from 78 silicosis patients with and without lung cancer. Materials and methods
Study population The subjects included 78 patients with silicosis at Okayama Rosai Hospital (n 5 76) or Bizen City Yoshinaga Hospital (n 5 2) between 2004 and 2006 and 20 healthy controls. Of the 78 patients with silicosis, 11 were diagnosed with lung cancer. The 20 control subjects had no occupational history of silica exposure and they were matched to the cases by gender, age and smoking status. The characteristics of the study population are summarized in Table I. Patients with silicosis were dened as those who have occupational histories for silica exposure in the industries such as stone quarries, granite production, ceramic and pottery industries and steel production and showed profusion rates from one to four on radiographs based on International Labour Organization classication (22). Histological subtypes of lung cancers were based on World Health Organization classication (23). The clinical stage of disease was assessed using the International Staging System (24). Sample collection and DNA extraction Peripheral blood samples (6 ml) were collected to investigate the methylation status of the serum DNA. The serum (2 ml) was isolated by centrifugation at 3000 r.p.m. for 10 min and stored at 80C until use. Serum DNA was extracted using a QIAamp DNA Blood Midi Kit (Qiagen, Hilden, Germany) according to the manufacturers instructions. We also examined methylation status of the tumor tissues obtained from surgical resection or autopsy. Tumor DNA was extracted from formalin-xed, parafn-embedded lung cancer tissues using QIAamp DNA Mini Kit (Qiagen) according to the manufacturers instructions. The researchers were unaware of each patients diagnosis.

Downloaded from carcin.oxfordjournals.org by guest on September 14, 2010

Introduction Occupational exposure to silica occurs in a large number of industries and circumstances, including mines and stone quarries as well as in the production of granite, ceramics, pottery and steel. In the USA, it is estimated that of more than one million workers occupationally exposed to free crystalline silica dust each year, 59 000 will eventually develop silicosis and $300 will die from the disease (World Health Organization Fact sheet N238, 2000. http://www.who.int/ mediacentre/factsheets/fs238/en/). Growing evidence indicates that the association between silicosis and lung cancer is causal (14). In 1997, the International Agency for Research on Cancer concluded
Abbreviations: CI, condence interval; DAPK, death-associated protein kinase; MSP, methylation specic polymerase chain reaction; MGMT, O6-methylguanineDNA methyltransferase; PCR, polymerase chain reaction; RARb, retinoic acid receptor b; RASSF1A, ras association domain family 1A; SD, standard deviation.

The Author 2008. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org

1845

S.Umemura et al.

Table I. Patient characteristics Healthy controls Age Median (range) Sex Male/female Smoking status S/NS/unknown Exposure period (years) Median (range) 0 30 .30 Unknown Radiographic nding PR 1/2/3/4 Histology Ad/Sq/Sm Stage I/II/III/IV 74 (6085) 17/3 14/6/0 0 20 Silicosis without lung cancer 71 (5186) 63/4 47/15/5 35.5 (147) 22 40 5 11/26/9/21 Silicosis with lung cancer 72 (5682) 10/1 9/2/0 33 (1040) 4 7 0 4/5/0/2 6/4/1 5/0/3/3

a t-test with unequal variance. An unconditional logistic regression model was applied to estimate the odds ratios and 95% condence intervals (CIs) for the occurrence of lung cancer. Those patients without methylation were used as a reference group to estimate the odds ratios for those patients with one or more methylated genes. Crude and multivariate models were examined. The factors considered in the multivariate model included age, gender, smoking status, radiological ndings and the exposure period. A chi-square or Fishers exact test was used to examine the distribution in categorical variables. Statistical signicance was dened as P , 0.05. All statistical analyses were conducted using SPSS version 10 (SPSS, Chicago, IL).

Results Methylation status of ve genes and its signicance in the detection of lung cancer We determined the prevalence of MGMT, p16INK4a, RASSF1A, DAPK and RARb methylation in the serum DNA of 78 silicosis patients with or without lung cancer and in 20 healthy controls by MSP. Among the controls, methylation was detected at a frequency of 5.0% for MGMT, 0.0% for p16INK4a, 0.0% for RASSF1A, 10.0% for DAPK and 5.0% for RARb (Table II). Among the silicosis patients without lung cancer, methylation was detected at a frequency of 14.9% for MGMT, 3.0% for p16INK4a, 3.0% for RASSF1A, 11.9% for DAPK and 9.0% for RARb. In contrast, among the silicosis patients with lung cancer, methylation was detected at a frequency of 36.4% for MGMT, 18.2% for p16INK4a, 9.1% for RASSF1A, 18.2% for DAPK and 0.0% for RARb. The total number of methylations per patient in the ve genes was 0.82 [standard deviation (SD), 0.603; 95% CI, 0.411.22] for those patients with silicosis plus lung cancer, which was higher than that for the patients with silicosis alone (0.42; SD, 0.721; 95% CI, 0.240.59; P 5 0.066). Aberrant promoter methylation in at least one of the ve tumor suppressor genes was more frequent in silicosis patients with lung cancer (72.7%) than in those without it (29.9%; P 5 0.006). For those patients with silicosis and methylation in at least one of the ve genes, the sensitivity, specicity, positive predictive values and negative predictive values for the diagnosis of lung cancer were 72.7, 70.1, 28.6 and 94.0%, respectively. Among ve patients with stage I disease of lung cancer, four patients had methylation in at least one of the ve genes. To quantify these PCR results, methylation value was determined using Scion Image software. Methylation value was higher in silicosis patients with lung cancer than those without in four genes tested (Figure 1). Tumor tissues were obtained from 8 of 11 cases with lung cancer, four from surgically resection and four from autopsy, and sufcient DNA was extracted in six of the eight cases. All the tumor tissues showed methylation in at least one gene, and in four of these six cases, methylation was also detected in serum DNA in at least one gene (supplementary Figure 2 is available at Carcinogenesis Online). To validate our results of MSP, methylation of p16INK4a and RASSF1A in the tumor tissues was evaluated using quantitative real-time PCR. Methylation of p16INK4a was detected in two cases and RASSF1A methylation was detected in ve cases. The p16INK4a methylation in two cases and RASSF1A methylation in three cases, detected by MSP, were conrmed by this quantitative PCR (supplementary Table 1 is available at Carcinogenesis Online). Methylation status and the likelihood of lung cancer Given that the silicosis patients with lung cancer exhibited a higher frequency of aberrant methylation than those without it, we analyzed the methylation status and risk of lung cancer in patients with silicosis. Table III shows the results of a crude logistic regression analysis of the correlation between methylation status and the risk of lung cancer. Silicosis patients with at least one methylated gene were 6.26 (95% CI, 1.5126.07) times more likely to have lung cancer than were silicosis patients with no methylated genes (P 5 0.012). To consider the imbalance in the baseline characteristics, we conducted a similar analysis adjusting for age, gender, smoking status, radiologic ndings and the silica exposure period, which were considered associated with the risk of malignancy. After adjusting for these factors,

Downloaded from carcin.oxfordjournals.org by guest on September 14, 2010

S, smoker; NS, non-smoker; PR, profusion rate; Ad, adenocarcinoma; Sq, squamous cell carcinoma; Sm, small-cell carcinoma.

The institutional review board approved the protocols and written informed consent was obtained from the subjects. Methylation specic polymerase chain reaction Sample DNA was treated with sodium bisulte using a CpGenome DNA Modication Kit (Intergen, Purchase, NY) as described previously (20,21). The primers used for MGMT, p16INK4a, RASSF1A, DAPK and RARb are described elsewhere (20,21). DNA from SBC-3 (25), a small-cell lung cancer cell line with promoter methylation of all tested genes, was used as a positive control. The polymerase chain reaction (PCR) mixture contained 10 PCR buffer [100 mM TrisHCl (pH 8.3), 500 mM KCl and 15 mM MgCl2], deoxynucleotide triphosphates (each at 2.5 mM), 0.5 lM of each primer, 0.75 U Hotstar Taq DNA polymerase (Qiagen) and 3 ll of bisulte-modied DNA in a nal volume of 30 ll. An initial denaturation step at 95C for 15 min was followed by 50 cycles of denaturation at 9095C for 20 s, annealing at the appropriate temperature for 30 s and extension at 72C for 30 s with a nal extension at 72C for 10 min. After amplication, each product was electrophoresed through a 2% agarose gel, stained with ethidium bromide and visualized under ultraviolet illumination. The presence of a band was dened as a methylation-positive result, even if it was faint. Each blood sample was examined in duplicate. Representative results of our methylation analysis are shown in supplementary Figure 1 (available at Carcinogenesis Online). Quantitative analysis of the PCR products was conducted using Scion Image software (http://www.scioncorp.com). The value of the PCR signal of each sample was calculated as the ratio; PCR signal of each sample to positive control. The nal value of the PCR signal was determined as mean value of the ratio in each group. Quantitative real-time PCR Quantitative real-time PCR was also performed with locus-specic primers and dual-labeled uorogenic probes for lung cancer tissue samples. Methylation of p16INK4a and RASSF1A was examined using b-actin as the internal control for DNA quantication. The DNA sequences of primers, which are different from those used in above-mentioned methylation specic polymerase chain reaction (MSP), and probes for these genes were based on published data (26,27). PCR was set up in a reaction volume of 25 ll containing 1 Taqman universal PCR master mix (Applied Biosystems, Foster City, CA), 500 nM of each primer, 150 nM of probe and 3 ll bisulte-treated DNA samples. After an initial denature step at 50C for 2 min and 95C for 15 min, 50 cycles of 15 s at 94C and 1 min at 6064 C were followed. Amplied data were analyzed using software developed by Applied Biosystems. The methylation ratio was dened as the ratio of the uorescence emission intensity values for methylated p16INK4a and RASSF1A to those of b-actin multiplied by 1000. Statistical analysis For each patient, the methylation status of the ve genes was scored as the total number of methylated genes. The methylation score of those patients with silicosis plus lung cancer was compared with those with silicosis alone using

1846

Promoter hypermethylation in silicosis

Table II. Frequency of methylation in ve genes Gene No. of patients (%) Healthy controls, n 5 20 MGMT p16INK4a RASSF1A DAPK RARb 1 (5.0) 0 (0.0) 0 (0.0) 2 (10.0) 1 (5.0) Silicosis without lung cancer, n 5 67 10 2 2 8 6 (14.9) (3.0) (3.0) (11.9) (9.0) Silicosis with lung cancer, n 5 11 4 (36.4) 2 (18.2) 1 (9.1) 2 (18.2) 0 (0.0)

(sample 4 in supplementary Figure 2 is available at Carcinogenesis Online). Case 2 involved an 82-year-old man who had received segmentectomy for squamous cell carcinoma of the lung. Although no methylation was observed in his serum while he was disease free, MGMT methylation was detected in his serum at recurrence. These results indicate that the methylated DNA in the serum of these patients was released from the cancer cells and that methylated DNA in the serum of patients with silicosis may reect undetected cancerous lesions. Discussion It is well established that patients with silicosis are at increased risk for lung cancer (14,28). In the current study, we examined the promoter methylation status of ve tumor suppressor genes using serum DNA from 78 silicosis patients with or without lung cancer. Promoter hypermethylation in at least one of the ve genes was signicantly increased in silicosis patients with lung cancer than in those without it. In addition, among 11 patients with silicosis that developed lung cancer, eight (72.7%) showed methylation in at least one of the ve genes tested. In a previous study of the same ve genes, almost half of the patients (49.5%) with lung cancer had methylation in their serum (20). These results indicate that lung cancer may be associated with an increased frequency of aberrant methylation if it is accompanied by silicosis, although the present study should be interpreted carefully due to the small sample size. We also examined the methylation status of tumor tissues of these patients with lung cancer and demonstrated that all the tumor tissues examined showed methylation in at least one gene. These results and the altered methylation prole in two described cases indicate that the methylated DNA in the serum was released from the cancer cells, though an unexpected result was also shown that MGMT methylation was detected in serum from two patients even though tumors did not show the methylation. We have no clear evidences but speculate two possible explanations of the discrepancy: (i) the tumor DNAs were extracted from formalin-xed, parafnembedded samples and not from fresh-frozen tissues. So DNA pieces of MGMT sequences might be fragmented and (ii) the methylated DNA in their serum might come from malignant or premalignant lesions in organs other than the lung. These should be claried in the future studies. The increased incidence of DNA methylation in silicosis patients with lung cancer compared with those without it motivated us to evaluate serum DNA methylation as a marker for lung cancer. Our analysis showed that patients with serum DNA methylation were $10 times more likely to have lung cancer. These ndings support the use of DNA methylation status as a marker for lung cancer detection in silicosis patients. Early detection of lung cancer in patients with silicosis is a serious clinical problem (14,8,28); thus, a reliable marker for rapid and accurate diagnosis is urgently needed. In the current study, methylation in serum DNA was detected even in patients with early stage of lung cancer. These results suggest that aberrant promoter methylation of tumor suppressor genes in serum DNA may be a valuable marker for the early detection of lung cancer during the follow-up treatment of patients with silicosis. Although further evaluation is essential, these results also suggest that serum DNA methylation may be worth examining routinely, particularly before invasive procedures. Before using serum DNA methylation as a marker for lung cancer during the follow-up of patients with silicosis, several issues must be considered. In the present study, we found that the sensitivity, specicity and positive predictive value of methylation in one or more genes for the diagnosis of lung cancer were 72.7, 70.1 and 28.6%, respectively. These results may not be satisfactory for clinical purposes. One strategy for improving the specicity and sensitivity of testing for lung cancer during follow-up treatment for silicosis is to use a larger number of genes or to apply a quantitative methylation assay (27,29,30). The sensitivity of the Taqman method is reported to be 10-fold higher than conventional qualitative MSP (31). In this study, quantitative PCR was performed for limited samples of lung cancer

Downloaded from carcin.oxfordjournals.org by guest on September 14, 2010

Fig. 1. Methylation value of ve genes. Methylation value was determined as mean value of the ratio: methylated band/positive control quantied with Scion Image.

silicosis patients with methylation in at least one gene were still 9.77 (1.7853.74; P 5 0.009) times more likely to have lung cancer (Table III). These results suggest that methylation of these ve tumor suppressor genes is associated with lung cancer in patients with silicosis. Methylation status and silicosis The total number of methylations per patient in the ve genes was 0.47 (SD, 0.716; 95% CI, 0.310.64) in patients with silicosis, which tended to be higher than in the healthy controls (0.2; SD, 0.523; 95% CI, 0.000.44; P 5 0.061). We next analyzed methylation status and the risk of silicosis. Table IV shows the results of a crude logistic regression analysis of the correlation between methylation status and the risk of silicosis. Patients with at least one methylated gene were 3.17 (95% CI, 0.85511.78) times more likely to have silicosis than were patients with no methylation (P 5 0.084). We then tested for associations between the methylation status of the ve genes and various clinical variables. As shown in Table V, no statistically signicant association between methylation status and clinical variables such as age, gender, smoking status or radiologic ndings was observed; however, we found a statistically signicant association between methylation status and the silica exposure period (P 5 0.017). These results suggest that methylation of these ve tumor suppressor genes is associated with silica exposure and might be associated with silicosis. Among the patients with long exposure to silica (.30 years), 85.7% of the patients with lung cancer have at least one gene hypermethylation compared with 37.5% of those without lung cancer (P 5 0.035) (Table VI). The altered serum methylation status of two silicosis patients with lung cancer We describe two cases that methylation status was examined twice during follow-up treatment for silicosis (supplementary Figure 3 is available at Carcinogenesis Online). Case 1 involved a 69-year-old woman who developed adenocarcinoma of the lung. DAPK methylation was detected in her serum at the time of partial resection of the right upper lobe; however, it disappeared 15 months later. The DAPK methylation was also detected in resected tumor tissue in the case

1847

S.Umemura et al.

Table III. Methylation status and risk of lung cancer in patients with pulmonary silicosis Silicosis without lung cancer No. of methylations 0 !1 Age (,70 to !70) Sex (female to male) Smoking status (never to smoker) Radiologic nding (0 versus 1, 2 versus 3, 4) Exposure period to silica ( 30 versus .30)
a

Silicosis with lung cancer

Model 1 Odds ratio 95% CI P value

a

Model 2 Odds ratio 1.00 9.77 1.45 0.86 1.17 4.89 0.38 95% CI P valuea

47 20

3 8

1 6.26

1.50526.071

0.012

1.77653.74 0.3236.482 0.05613.329 0.1608.465 0.78730.346 0.0702.099

0.009 0.630 0.914 0.88 0.089 0.269

Logistic regression model.

Table IV. Methylation status and the risk of silicosis No. of methylations 0 !1
a

Healthy controls 17 3

Silicosis patients 50 28

Odds ratio 1.00 3.17

95% CI

P valuea

Table VI. Comparison of methylation frequency in patients with long silica expose (.30 years) Gene No. of patients (%) Without lung cancer, n 5 40 MGMT p16INK4a RASSF1A DAPK RARb At least one gene 8 2 1 6 4 15 (20.0) (5.0) (2.5) (15.0) (10.0) (37.5) With lung cancer, n 5 7 2 2 0 2 0 6 (28.6) (28.6) (0.0) (28.6) (0.0) (85.7) P valuea

Downloaded from carcin.oxfordjournals.org by guest on September 14, 2010

0.85511.78

0.084

Logistic regression model.

Table V. Univariate analysis of clinical variables and methylation status Variable n Methylation status Negative Age ,70 !70 Gender Male Female Smoking status Non-smoker Smoker Radiologic ndings (PR) 0 1, 2 3, 4 Silica exposure period (years) 0 30 .30 39 59 90 8 23 70 20 46 32 20 26 47 25 42 61 6 16 48 17 31 19 17 21 26 Positive 14 17 29 2 7 22 3 15 13 3 5 21
b

0.035

P value

Chi-square test.

0.51 0.999 0.999 0.151

0.017

PR, profusion rate. a Chi-square test. b Patients with at least one gene hypermethylation.

tissues and we found inconsistency between results by non-quantitative MSP and those by quantitative real-time PCR, such as RASSF1A methylation in sample 2. These ndings might indicate the higher sensitivity of the Taqman method compared with non-quantitative one. Another strategy is to search for the best combination of genes to use for methylation analysis with or without other diagnostic tests, including low-dose spiral computed tomography (3234). Additional studies are warranted. It is of note that a statistically signicant association was detected between methylation status and the silica exposure period, although no direct association between methylation status and silicosis was observed. A previous study using an animal model demonstrated that expression of E-cadherin was signicantly reduced by silica-induced

chronic inammation because of promoter hypermethylation (35). Aberrant promoter methylation has also been reported in precancerous lesions, such as in the dysplasias of patients with lung cancer (36). Our nding that long-term silica exposure is correlated with an increased frequency of methylation suggests that silicosis may be a socalled precancerous lesion (1,37). And methylation status did not correlate with age, smoking history or radiographic ndings of silicosis. There are numerous reports that show age-related methylation in a subset of genes (38,39) or association between smoking and methylation in the lung (4042). We have no clear explanation why methylation status did not correlate with age or smoking history in this study, but it might be due to our non-quantitative method for methylation analysis. We found that the total number of methylations per patient in ve genes tended to be higher in patients with silicosis than in healthy controls. These results support the notion that silicosis patients with methylation in some tumor suppressor genes may already be in an early stage of lung cancer development or that some proportion of these patients already has cancerous lesions. Thus, it will be of great interest to investigate whether the methylation-positive patients with silicosis develop lung cancer in the near future. In conclusion, we showed that aberrant tumor suppressor gene promoter methylation was more frequent in the serum DNA of silicosis patients with lung cancer than in those without it. Our results suggest that testing for aberrant promoter methylation of these genes in serum DNA may aid in the early detection of lung cancer in patients with silicosis. Additional studies are warranted to conrm the value of such analyses and to determine the most informative combination of genes. Funding This research is a part of the research and development and the dissemination projects related to he 13 elds of occupational injuries and illnesses of the Japan Labour Health and Welfare Organization.

1848

Promoter hypermethylation in silicosis

Acknowledgements
We wish to thank Drs Kenji Ogino, Toshiyuki Kozuki, Eiki Ichihara, Katsuyuki Hotta and Shinichi Toyooka for their support and comments on our analysis. Conict of Interest Statement: None declared.

References
1. Pelucchi,C. et al. (2006) Occupational silica exposure and lung cancer risk: a review of epidemiological studies 19962005. Ann. Oncol., 17, 10391050. 2. Smith,A.H. et al. (1995) Meta-analysis of studies of lung cancer among silicotics. Epidemiology, 6, 617624. 3. Tsuda,T. et al. (1997) A meta-analysis on the relationship between pneumoconiosis and lung cancer. J. Occup. Health, 39, 285294. 4. Steenland,K. et al. (1997) Silica, asbestos, man-made mineral bers, and cancer. Cancer Causes Control, 8, 491503. 5. International Agency for Research on Cancer. (1997) Evaluation of carcinogenic risks to humans. Silica, some silicates, coal dust, and para-aramid brils. IARC Press, 68, 41242. 6. Chan,C.K. et al. (2000) Lung cancer mortality among a cohort of men in a silicotic register. J. Occup. Environ. Med., 42, 6975. 7. Okazaki,H. et al. (2004) Improved detection of lung cancer arising in diffuse lung diseases on chest radiographs using temporal subtraction. Acad. Radiol., 11, 498505. 8. Chong,S. et al. (2006) Pneumoconiosis: comparison of imaging and pathologic ndings. Radiographics, 26, 5977. 9. Baylin,S.B. et al. (1998) Alterations in DNA methylation: a fundamental aspect of neoplasia. Adv. Cancer Res., 72, 141196. 10. Baylin,S.B. et al. (2001) Aberrant patterns of DNA methylation, chromatin formation and gene expression in cancer. Hum. Mol. Genet., 10, 687692. 11. Merlo,A. et al. (1995) 5# CpG island methylation is associated with transcriptional silencing of the tumor suppressor p16/CDKN2/MTS1 in human cancers. Nat. Med., 1, 686692. 12. Kim,H. et al. (2004) Tumor-specic methylation in bronchial lavage for the early detection of non-small-cell lung cancer. J. Clin. Oncol., 22, 23632370. 13. Topaloglu,O. et al. (2004) Detection of promoter hypermethylation of multiple genes in the tumor and bronchoalveolar lavage of patients with lung cancer. Clin. Cancer Res., 10, 22842288. 14. Esteller,M. et al. (2001) A gene hypermethylation prole of human cancer. Cancer Res., 61, 32253229. 15. Toyooka,S. et al. (2001) DNA methylation proles of lung tumors. Mol. Cancer Ther., 1, 6167. 16. Kersting,M. et al. (2000) Differential frequencies of p16 (INK4a) promoter hypermethylation, p53 mutation, and K-ras mutation in exfoliative material mark the development of lung cancer in symptomatic chronic smokers. J. Clin. Oncol., 18, 32213229. 17. Burbee,D.G. et al. (2001) Epigenetic inactivation of RASSF1A in lung and breast cancers and malignant phenotype suppression. J. Natl Cancer Inst., 93, 691699. 18. Tang,X. et al. (2000) Hypermethylation of the death-associated protein (DAP) kinase promoter and aggressiveness in stage I non-small-cell lung cancer. J. Natl Cancer Inst., 92, 15111516. 19. Virmani,A.K. et al. (2000) Promoter methylation and silencing of the retinoic acid receptor-b gene in lung carcinomas. J. Natl Cancer Inst., 92, 13031307. 20. Fujiwara,K. et al. (2005) Identication of epigenetic aberrant promoter methylation in serum DNA is useful for early detection of lung cancer. Clin. Cancer Res., 11, 12191225.

21. Katayama,H. et al. (2007) Aberrant promoter methylation in pleural uid DNA for diagnosis of malignant pleural effusion. Int. J. Cancer, 120, 21912195. 22. International Labour Ofce. (1980) Guidelines for the Use of ILO International Classication of Radiograph of Pneumoconioses. Occupational Safety and Health, Series No. 22. ILO, Geneva, Switzerland. 23. World Health Organization. (1999) Histological Typing of Lung and Pleural Tumors, 3rd edn. Geneva, Switzerland, WHO. 24. Mountain,C.F. (1997) Revisions in the international system for staging lung cancer. Chest, 111, 17101717. 25. Miyamoto,H. (1986) Establishment and characterization of an adriamycinresistant subline of human small cell lung cancer cells. Acta Med. Okayama, 40, 6573. 26. Hoque,M.O. et al. (2006) Quantitation of promoter methylation of multiple genes in urine DNA and bladder cancer detection. J. Natl Cancer Inst., 96, 9961004. 27. Harden,S.V. et al. (2003) Gene promoter hypermethylation in tumors and lymph nodes of stage I lung cancer patients. Clin. Cancer Res., 9, 1370 1375. 28. Zeka,A. et al. (2006) Lung cancer and occupation in nonsmokers: a multicenter case-control study in Europe. Epidemiology, 17, 615623. 29. An,Q. et al. (2002) Detection of p16 hypermethylation in circulating plasma DNA of non-small cell lung cancer patients. Cancer Lett., 188, 109114. 30. Brabender,J. et al. (2003) Quantitative O6-methylguanine DNA methyltransferase methylation analysis in curatively resected non-small cell lung cancer: associations with clinical outcome. Clin. Cancer Res., 9, 223227. 31. Eads,C.A. et al. (2000) MethyLight: a high-throughput assay to measure DNA methylation. Nucleic Acids Res., 28, E32. 32. Henschke,C.I. et al. (1999) Early Lung Cancer Action Project: overall design and ndings from baseline screening. Lancet, 354, 99105. 33. Swensen,S.J. et al. (2002) Screening for lung cancer with low-dose spiral computed tomography. Am. J. Respir. Crit. Care Med., 165, 508513. 34. McWilliams,A. et al. (2002) Innovative molecular and imaging approaches for the detection of lung cancer and its precursor lesions. Oncogene, 21, 69496959. 35. Blanco,D. et al. (2004) Altered expression of adhesion molecules and epithelial-mesenchymal transition in silica-induced rat lung carcinogenesis. Lab. Invest., 84, 9991012. 36. Brabender,J. et al. (2003) Quantitative O6-methylguanine DNA methyltransferase methylation analysis in curatively resected non-small cell lung cancer: associations with clinical outcome. Clin. Cancer Res., 9, 223227. 37. Taeger,D. et al. (2006) Role of exposure to radon and silicosis on the cell type of lung carcinoma in German uranium miners. Cancer, 106, 881889. 38. Holliday,R. (1985) The signicance of DNA methylation in cellular aging. Basic Life Sci., 35, 269283. 39. Issa,J.P. (2000) CpG-island methylation in aging and cancer. Curr. Top. Microbiol. Immunol., 249, 101118. 40. Kim,D.H. et al. (2001) p16(INK4a) and histology-specic methylation of CpG islands by exposure to tobacco smoke in non-small cell lung cancer. Cancer Res., 61, 34193424. 41. Toyooka,S. et al. (2004) Dose effect of smoking on aberrant methylation in non-small cell lung cancers. Int. J. Cancer, 110, 462464. 42. Toyooka,S. et al. (2003) Smoke exposure, histologic type and geographyrelated differences in the methylation proles of non-small cell lung cancer. Int. J. Cancer, 103, 153160. Received February 6, 2008; revised July 9, 2008; accepted July 10, 2008

Downloaded from carcin.oxfordjournals.org by guest on September 14, 2010

1849