Global and gene-specific promoter methylation changes are related to anti-B[a]PDE-DNA adduct levels and influence micronuclei levels in polycyclic aromatic hydrocarbon-exposed individuals Sofia Pavanello 1 * , Valentina Bollati 2 , Angela Cecilia Pesatori 2 , Lucyna Kapka 3 , Claudia Bolognesi 4 , Pier Alberto Bertazzi 2 and Andrea Baccarelli 2 1 Occupational Health Section, Department of Environmental Medicine and Public Health, University of Padova, Padova, Italy 2 Center of Molecular and Genetic Epidemiology, Department of Environmental and Occupational Health, University of Milan & IRCCS Maggiore Policlinico Hospital, Mangiagalli and Regina Elena Foundation, Milan, Italy 3 Institute of Agricultural Medicine, Lublin, Poland 4 Environmental Carcinogenesis Unit, National Cancer Research Institute, Genoa, Italy We investigated the effect of chronic exposure to polycyclic aro- matic hydrocarbons (PAHs) on DNA methylation states (percent- age of methylated cytosines (%mC)) in Polish male nonsmoking coke-oven workers and matched controls. Methylation states of gene-specific promoters (p53, p16, HIC1 and IL-6) and of Alu and LINE-1 repetitive elements, as surrogate measures of global meth- ylation, were quantified by pyrosequencing in peripheral blood lymphocytes (PBLs). DNA methylation was evaluated in relation to PAH exposure, assessed by urinary 1-pyrenol and anti-ben- zo[a]pyrene diolepoxide (anti-B[a]PDE)-DNA adduct levels, a crit- ical genetic damage from B[a]P. We also evaluated whether PAH- induced DNA methylation states were in turn associated with micronuclei in PBLs, an indicator of chromosomal instability. ' 2009 UICC Key words: DNA methylation; polycyclic aromatic hydrocarbons; lung carcinogens; anti-B[a]PDE-DNA adduct; micronuclei; chronic exposure We found that Alu and LINE-1 methylation levels and those of the inflammatory cytokine IL-6, to a lesser extent, were higher in polycyclic aromatic hydrocarbon (PAH)-exposed coke-oven workers than controls (p < 0.001 and p 5 0.094). Conversely, methylation of p53 and HIC1 tumor suppressors was lower in workers compared with controls (p < 0.001 and p < 0.05). Global and IL-6 hypermethylation and p53 hypomethylation were signifi- cantly correlated, not only to PAH-exposure (urinary 1-pyrenol) but also to the anti-B[a]PDE-DNA adduct levels (p < 0.01). Over- all, linear multivariate regression analysis showed that the only significant determinant of increasing micronuclei (MN) (p < 0.01) was p53 hypomethylation and not the other LINE-1, Alu, p53, HIC1 and IL-6 methylation states. Gene-specific promoters (mainly p53) and global (Alu and LINE-1 repeats) DNA methylation changes in circulating periph- eral blood lymphocytes (PBLs), related to anti-B[a]PDE-DNA adduct and MN, suggest that these events could be determined to identify subjects at high cancer risk. Understanding how environmental factors are involved in can- cer etiology and development is one of the main goals of biomedi- cal research. 1 Extensive research has been conducted to determine how known or potential environmental carcinogens modify the DNA sequence, as a component of malignant transformation. However, an investigative approach exclusively based on genetic damage, as well as on inheritance of genetic variants, has been revealed to be insufficient to account for multistage carcinogenic processes. 2 It has been proposed that DNA methylation, one of the best known epigenetic mechanisms, shares critical roles with DNA mutations in the theoretical continuum for exposure to can- cer. 3,4 One potential mechanism for environmental factors is through hypermethylation or hypomethylation on somatic cells, leading to activation or silencing of key genes in critical pathways of cancer. 5,6 From some years, the disruption of DNA methylation status by exposure to genotoxic agents has been an issue in the lit- erature, but the clear demonstration that such epigenetic altera- tions were caused by one or more specific agents in people has been demanding (for a review see Ref. 7 and references therein). Benzo[a]pyrene (B[a]P), the most well-studied PAH carcinogen (especially of the lung), has a well-established genotoxic mechanism, via metabolic activation to diol epoxide. 8 The stereo- selective binding of anti-benzo[a]pyrene diolepoxide (anti- B[a]PDE), which is efficiently produced at the guanine residue in CpG sequence, resulting in mutations, including mutational ‘‘hotspots’’ in the pulmonary p53 gene, is critical in B[a]P lung tu- morigenesis. 9 P53 is the best characterized B[a]P mutagenic target and, together with tumor suppressor gene p16, is also frequently epigenetically altered in smoking (PAH)-associated lung cancer (for a review see Ref. 10). P53 synergizes in tumor suppression with HIC1 (hypermethylated in cancer), 11 which epigenetically regulates the p53-dependent apoptotic DNA-damage responses. 12 Moreover, p53 can be epigenetically modulated by the inflamma- tory cytokine IL-6, via DNA methyltransferase (DNMT) activa- tion. 13 Global DNA hypermethylation following in vitro chronic exposure to B[a]P was recently revealed in mouse embryonic fibroblasts and was associated to mutation as a consequence of genetic instability. 14,15 PBLs, although not considered target cells for PAH(B[a]P)- induced lung tumorigenesis, can be easily obtained from human subjects and their genetic (PAH(B[a]P)-DNA adducts) and epige- netic alterations (methylation) are correlated with levels in target tissues (e.g., lung). 16–18 Moreover, high levels of adduct and micronuclei (as markers of chromosome instability 19 ) and p53 hypomethylation in PBLs are predictive of lung cancer risk. 20–22 Finally, increased anti-B[a]PDE-DNA and MN levels have been related to occupational (cokeries) and environmental (smoking, Abbreviations: anti-B[a]PDE ()-r-7,t-8-dihydroxy-t-9,10-oxy-7,8,9, 10-tetrahydrobenzo[a]pyrene; B[a]P, benzo[a]pyrene; B[a]P-tetrol-I-1, r- 7,c-10,t-8,t-9-tetrahydroxy-7,8,9,10-tetrahydro-benzo[a]pyrene; BEI, bio- logical exposure index; CBMN, cytokinesis-blocked micronucleus; DNMT, DNA methyltransferase; HIC1, hypermethylated in cancer; HPLC, high-performance liquid chromatography; IL, interleukin; LINE-1, long interspersed nuclear element-1; MN, micronuclei; PAHs, polycyclic aromatic hydrocarbons; PBLs, peripheral blood lymphocytes. Grant sponsor: CARIPLO Foundation; Grant number: 2007-5469; Grant sponsor: Italian Ministry of University and Research; Grant number: PRIN2007-2S2HT8; Grant sponsor: Italian Association for Research on Cancer (AIRC) 2008. *Correspondence to: Occupational Health Section, Department of Environmental Medicine and Public Health, University of Padova, Via Giustiniani 2, 35128 Padova, Italy. Fax: 1390498212542. E-mail: sofia.pavanello@unipd.it Received 8 January 2009; Accepted after revision 31 March 2009 DOI 10.1002/ijc.24492 Published online 7 April 2009 in Wiley InterScience (www.interscience. wiley.com). Int. J. Cancer: 125, 1692–1697 (2009) ' 2009 UICC Publication of the International Union Against Cancer