Dear Sir,Tobacco smoking is the most important risk factor for lung cancer worldwide. Cancers arise following an accumulation of genetic and epigenetic alterations, including point mutations of genes such as p53 1,2 and Ras, chromosomal deletions 3 and methylation of multiple genes; 4,5 and many of these changes occur in tobacco-related lung cancers. In addition, tobacco smoking is associated with a dose-response increase in p53 mutations. 3 We previously demonstrated that the aberrant methylation status of p16, RASSF1A, APC, RAR and CDH13 genes in lung cancer is related to smoke exposure and histologic type but not to gender. 6 Here, we report the results of further examination for methylation status in 383 Japanese non-small cell lung cancer (NSCLC) cases. The characteristics of our cases are shown in Table I. These samples have extensive data for dosage of tobacco exposure, so we examined the relationship between smoking dosage and methylation status of the 5 genes related to tobacco exposure. Methylation-specific PCR was used for determining the methylation status of samples. 7 Methylation-specific PCR, with sensitivity determined as previously described, could detect one methylated allele in the presence of 10 3 -10 4 unmethylated alleles. 6 DNA from lymphocytes of a healthy volunteer artificially methylated by treatment with Sss1 (New England BioLabs, Beverly, MA) was used as a positive control for methylated alleles. We also used cell lines as a positive control, which we reported previously. 5 PCR products were visualized on 2% agarose gels stained with ethidium bromide.Based on smoking status, we divided the cases into 3 groups: never-smokers (Ͻ100 cigarettes/lifetime), smokers with exposure of Ͻ30 pack-years and smokers with exposure Ն30 packyears. The smoking effects on methylation were examined by the Cochran-Armitage trend test, and the results are shown in Figure 1. In all cases, methylation of p16 and RASSF1A and the methylation index (MI), which was defined as a fraction representing the number of genes methylated/the number of genes tested, 4 showed a dose-related response. Because we previously noted that histologic differences affect the profile of aberrant methylation, 6 we also examined the effects of the degree of smoke exposure on both adenocarcinoma and squamous cell carcinoma, the major subtypes of NSCLC. As most lung cancers in never-smokers are adenocarcinomas, our study was limited by the relatively small number of squamous cell carcinomas (n ϭ 8) in never-smokers. However, for both histologic types, methylation of p16 and MI showed a doserelated response. For adenocarcinoma, RASSF1A and APC methylation also demonstrated a dose-dependent response. While a similar tendency for squamous cell carcinomas was noted, the effect was not statistically significant. For CDH13 and RAR, there was no relationship to extent of smoke exposure.Our present findings confirm and extend our previous findings that p16 and APC methylation are more frequent in eversmokers than in never-smokers. 6 While me...