Introduction {#s1} ============ Epigenetic and genetic alterations are associated with the development of many human cancers [@pone.0072114-Kulik1]. The methylation status of several tumour suppressor genes have been found to be associated with a significant number of cancers, such as *RASSF1A*, *E-cadherin*, *CDH13*, *p16*, and *SLC5A8* [@pone.0072114-Kristensen1]. *CDH13* is hypermethylated in many kinds of solid cancers, including prostate [@pone.0072114-Fujii1], breast [@pone.0072114-Jin1], colorectal [@pone.0072114-Waki1], lung [@pone.0072114-Fujii2], gastric [@pone.0072114-Kim1], and renal [@pone.0072114-Sasaki1] cancers. *CDH13* encodes a Ca^2+^-independent cell adhesion molecule and interacts with the actin cytoskeleton through its cytoplasmic domain [@pone.0072114-Yamada1], [@pone.0072114-Vleminckx1]. The encoded protein is thought to play a role in tumour suppression by interacting with c-met, regulating its expression [@pone.0072114-Moriyama1]. Loss of the *CDH13* expression in the tumour tissue might result in a higher migratory capacity. *CDH13* hypermethylation is detected in various cancer types, and the expression of *CDH13* is restored by demethylation of this gene [@pone.0072114-Waki1], [@pone.0072114-Chin1]--[@pone.0072114-Toyooka1]. The results of previous studies have suggested that *CDH13* promoter methylation could serve as a specific biomarker for early detection of renal cell carcinoma (RCC) [@pone.0072114-Toyooka1], [@pone.0072114-Fukami1], as well as for the monitoring of the disease status of the patients [@pone.0072114-Yanagawa1]. The methylation status of *CDH13* in lung tissue has been reported in only three previous studies. Yanagawa *et al.* [@pone.0072114-Yanagawa1] reported that *CDH13* hypermethylation was detected in 30.2% (15 of 49) of lung cancer tissues, and this occurred at a higher frequency than in adjacent normal tissues (5.3%, 2 of 38; P = 0.0082). Toyooka *et al.* [@pone.0072114-Toyooka1] found that *CDH13* hypermethylation was significantly associated with a larger tumour size (P = 0.001), lymph node metastasis (P = 0.04), and poor disease-specific survival (P = 0.01). In another study, Yanagawa *et al.* [@pone.0072114-Yanagawa2] reported that the *CDH13* hypermethylation status was significantly associated with the methylation status of the well-established lung cancer biomarkers *p16*, *APC*, *RASSF1A* and *p14*. Taken together, previous findings suggest that the methylation status of *CDH13* is closely associated with lung cancer. However, the role of *CDH13* hypermethylation in the development of lung cancer is not completely understood. We hypothesised that the frequency of *CDH13* hypermethylation is higher in a high risk group of population in lung cancer patients and their smoking partners than in a low risk group of population. To test this hypothesis, we evaluated the frequency of *CDH13* hypermethylation in tissue samples from a total of 105 lung cancer patients and from their smoking partners in a case-control study. Materials and Methods {#s2} ===================== Ethics Statement {#s2a} ---------------- All participants provided written informed consent. The study protocol was approved by the Ethics Review Committee of Chiba University Graduate School of Medicine. Study Subjects {#s2b} -------------- A total of 105 lung cancer patients and their smoking partners (105 pairs) were prospectively enrolled from April 2006 to December 2012 in Japanese Red Cross Nagoya Daiichi Hospital (Nagoya, Japan). Of the 105 patients, 80 were diagnosed with primary lung cancer and 25 were diagnosed with recurrent cancer. The primary lung cancers comprised 34 adenocarcinoma (ADC) and 44 squamous cell carcinoma (SQC) cases. Seventeen and 34 pairs had the same sex and opposite sex, respectively. In a comparison between the same and opposite sexes, we found no significant differences in the age, smoking, and lung cancer patient characteristics (age: P = 0.28; smoking: P = 0.48; pack-years: P = 0.13; sex: P = 0.47; histological type: P = 0.11; pT stage: P = 0.16; pN stage: P = 0.12; histology differentiation: P = 0.07) (Data not shown). All participants were eligible for the study if they were male and ≥20 years of age, and had smoked ≥10 packs of cigarettes per year or for ≥20 years [@pone.0072114-Ito1]. We excluded smokers who smoked ≤10 packs of cigarettes per year or for \<20 years. Patients who met any of the following criteria were excluded: a history of other primary cancers or another second malignant disease; an acute infectious disease within 3 months of the diagnosis; or receipt of chemotherapy or radiotherapy before surgery. Patient information was retrieved from the hospital medical records. The patient characteristics recorded were age, sex, smoking habit, history of chronic obstructive pulmonary disease (COPD) and/or asthma, current pathological stage, and differentiation grade. Methods {#s2c} ------- We obtained tumour tissue samples from the primary and recurrent lung cancer sites from the patients. All samples were taken from the primary site of the lung. Signed informed consent was obtained from the patients, and a trained pathologist with no knowledge of the patient's clinical condition collected all tissue specimens. All tissue samples were cut into multiple 3-µm thick sections. We used the tissue immediately after resection or biopsies for methylation analysis. We used the surgically resected tumour tissues from which we obtained the histological sections to determine the histological types and stages of the tumour [@pone.0072114-Saleh1]. Tumour size was defined as the maximum diameter. DNA extraction and *CDH13* Promoter Methylation Analysis {#s2d} -------------------------------------------------------- Tissue samples were obtained from all 105 patients and their smoking partners. We extracted genomic DNA from the resected specimens using standard methods. DNA methylation was examined using the following primers: 5′-CCTTCAATACCCGCACCCCA-3′ (forward) and 5′-GGAGGTGCCTGTGGTTCTTT-3′ (reverse) with MSP assay. The primers amplified a 116-bp sequence at −152 bp from the transcription start site in exon 1 of the *CDH13* gene (Genbank accession no. NM_005101) [@pone.0072114-Toyooka1]. The methylation status was determined with MSP for an unmethylated reaction (unmethylated primers) and a methylated reaction (methylated primers). The PCR conditions were as follows: 95°C for 2 min; 55 cycles of 95°C for 30 sec, 60°C for 30 sec, and 72°C for 30 sec; 72°C for 7 min; and 4°C until analysis. We also performed a standard MSP method in all cases to ensure that the PCR products did not contain amplification artifacts. PCR products were separated on a 1% agarose gel and visualised under ultraviolet light after staining with ethidium bromide. We evaluated the results of each PCR assay independently by three operators to increase the accuracy and reliability of the results. Results {#s3} ======= *CDH13* Hypermethylation in the Primary Tumour Tissue from Patients with Lung Cancer {#s3a} ------------------------------------------------------------------------------------ The patient characteristics are shown in [Table 1](#pone-0072114-t001){ref-type="table"}. We first measured the frequency of *CDH13* hypermethylation in the tumour tissue samples from the 105 patients and their smoking partners. In a total of 105 pairs, 45 pairs (42.9%) showed *CDH13* hypermethylation. Hypermethylation was detected in 29.6% (31/105) of the primary cancer tissue samples, and in 14.3% (14/101) of the paired samples from the smoking partners. *CDH13* hypermethylation was detected in 23.8% (14 of 59) of the patients with ADC and 20.0% (17 of 85) of the patients with SQC. There were significant differences between ADC and SQC (P = 0.005) in *CDH13* hypermethylation. The majority of the cases with *CDH13* hypermethylation were found in adenocarcinoma (23.3%) or squamous cell carcinoma (25.5%). 10.1371/journal.pone.0072114.t001 ###### Characteristics of study subjects and CDH13