High-risk oral leukoplakia is associated with aberrant promoter methylation of multiple genes

Human OSCC cell lines (Ca9–22, HSC-2, HO-1-N-1, HSC-3 and SCC-4) were purchased from the Human Science Research Resources Bank (HSRRB, Osaka, Japan). A total of 24 OL tissues (average age, 64.0 years [range, 38–84 years]; 10 male and 14 female) and a total of 26 OSCC tissues (average age, 64.6 years [range, 42–89 years]; 17 male and 9 female) were obtained from patients who underwent biopsies or operations at the University of Tokyo Hospital between Dec. 2009 and Nov. 2011. The OSCCs were graded according to the Union for International Cancer Control (UICC)’s TNM classification. OL was defined as “a predominantly white lesion of the oral mucosa that can not be characterized as any other definable lesion”[27]. The presence or absence of dysplasia in OLs is determined by the degree of cellular abnormality above the epithelial basement membrane as originally defined by the World Health Organisation (WHO) [28]. Normal oral mucosae were obtained from 16 healthy volunteers. Samples were stored in RNAlater (Applied Biosystems, Foster City, CA, USA) at -80 °C until the extraction of genomic DNA. Genomic DNA was extracted by the phenol-chloroform method. This research was approved by the research ethics committee of Graduate School of Medicine and Faculty of Medicine, The University of Tokyo, approval #2819-(1), and informed consent was obtained from all patients and volunteers. Each patient’s tobacco smoking history was obtained in an interview.

Ca9–22 and HSC-2 cells were seeded at a density of 2 × 10⁵ cells ⁄ 10 cm plate on day 0. For 5-aza-2′-deoxycytidine (5-aza-dC; Sigma, St Louis, MO, USA) treatment, the cells were exposed to medium containing 3-μM 5-aza-dC or control medium for 24 h on days 1 and 3, and then harvested on day 5. The doses of 5-aza-dC were adjusted so that the growth of the treated cells was suppressed to 40–80 % that of nontreated cells.

MeDIP − CGI microarray analysis was performed as previously described [29, 30]. Briefly, 5 μg of genomic DNA was immunoprecipitated with an anti-5-methylcytidine antibody (Diagnode, Liége, Belgium), and the precipitated DNA and input DNA were labeled with Cy5 and Cy3, respectively. A human CGI oligonucleotide microarray (Agilent Technologies, Santa Clara, CA, USA) was hybridized with the labeled probes and scanned with a G2565BA microarray scanner (Agilent Technologies). Scanned data were processed with Feature Extraction 9.1 and ChIP Analytics 1.3 software (Agilent Technologies). The signal of the probe was converted into a “Me value,” which represents the methylation level as a value from 0 (unmethylated) to 1 (methylated) [29]. Differentially methylated regions were detected by a comparison of the Me values of the two samples. When three or more consecutive probes in a locus showed differences in the Me value larger than 0.6, the locus was considered to have different methylation statuses. Promoter regions of three genes (HOXA11, NPY, and UCHL1) reported as frequently methylated in multiple cancers, including OSCCs, were used as a methylated control [31‐33]. Promoter regions of three genes (ACTB, B2M, and GAPDH) known as housekeeping genes were used as unmethylated control.

Expression microarray analysis was performed by a GeneChip Human Genome U133 Plus 2.0 expression microarray (Affymetrix, Santa Clara, CA, USA). From 8 μg of total RNA, first-strand cDNA was synthesized with SuperScript III reverse transcriptase (Invitrogen) and T7-(dT)24 primer (Amersham Biosciences, Little Chalfont, UK). Double-stranded cDNA was then synthesized, and biotin-labeled cRNA was synthesized using a Bio-Array HighYield RNA transcript-labeling kit (Enzo Life Sciences, Farmingdale, NY, USA). Twenty micrograms of labeled cRNA was fragmented and hybridized to the GeneChip oligonucleotide microarray with a GeneChip hybridization control kit. The microarray was stained and scanned according to the Affymetrix protocol. The scanned data were processed using GeneChip operating software 1.4. The signal intensity of each probe was normalized so that the average signal intensity of all the probes on a microarray would be 500. The average signal intensity of all the probes for a gene was used as its transcription level. Genes were classified into those with high (>1000), moderate (250–1000), or low (<250) transcriptions according to their signal intensities [30].

Sodium bisulfite treatment was performed as described previously [29] using 500 ng of DNA digested with BamHI (Toyobo, Tokyo, Japan) and suspended in 20 μl of Tris-EDTA (TE) buffer. For MSP, 1 μl of solution was used for PCR reaction with primers specific to methylated (Additional file 1: Table S1) and with primers that targeted the Alu repeat sequence; the latter were used as a control of the amount of bisulfite-treated DNA [34]. Fully methylated DNA was prepared by methylating genomic DNA using SssI-methylase (New England Biolabs, Beverly, MA, USA). Fully unmethylated DNA was prepared by amplifying genomic DNA with phi29 DNA polymerase (GenomiPhi DNA Amplification kit; GE Healthcare UK, Buckinghamshire, UK).

Associations between methylation status and various clinical parameters were evaluated by Fisher’s exact test (two-sided). SPSS Statistics (IBM Corporation, Somers, NY, USA) software version 21.0 (SPSS Inc., Chicago, IL, USA) was used for analysis. P values < 0.05 were considered to indicate significance.