MAGE A1-A6 RT-PCR and MAGE A3 and p16 methylation analysis in induced sputum from patients with lung cancer and non-malignant lung diseases

The melanoma antigen gene (MAGE) A1-A6 RT-PCR system was developed for the detection of lung cancer cells in the sputum. However, we identified MAGE expression in some patients with non-malignant lung diseases. To understand these patterns of MAGE expression, we performed MAGE A3 methylation-specific PCR (MSP) and p16 MSP. We collected 24 biopsy specimens of lung cancer tissue and performed MAGE A1-A6 RT-PCR, MAGE A3 MSP and p16 MSP. RNA and DNA were simultaneously extracted from induced sputum specimens of 133 patients with lung diseases and 30 random sputum specimens of healthy individuals and the 3 molecular analyses were performed. The patients were diagnosed as 65 cases of lung cancer and 68 of benign lung diseases. Positive rates of MAGE A1-A6 RT-PCR, MAGE A3 MSP and p16 MSP were as follows: in lung cancer tissue, 87.5, 58.3 and 70.8%; in the sputum of lung cancer patients, 50.8, 46.2 and 63.1%; benign lung diseases, 10.3, 30.9 and 39.7%; and healthy individuals, 3.3, 6.7 and 3.3%. Of the 40 MAGE-positive cases, 33 were diagnosed with lung cancer and 7 as having benign lung diseases. From the 7 cases of MAGE-positive benign lung diseases, 6 cases showed methylation abnormalities. The MAGE-positive group revealed significantly higher rates of methylation abnormalities. Of the 40 MAGE-positive cases, 39 cases were found to be lung cancer or benign lung diseases with abnormal methylation. Thus, MAGE expression in the sputum suggests the presence of lung cancer cells or pre-cancerous cells.


Introduction
For the detection of lung cancer cells in the sputum, we developed melanoma associated gene (MAGE) primers that can amplify MAGE A1-A6 simultaneously (1), and evaluated the positive rates in induced sputum of patients with lung cancer (2). Meanwhile, we detected MAGE expression not only in patients with lung cancer but also in some patients with non-malignant lung diseases, including tuberculosis and inflammation. Expression of MAGE by de-methylation of the MAGE promoter region has been demonstrated (3), and its expression was known to be restricted to germ cells and cancer cells (4,5). However, Mecklenburg et al reported that MAGE expression was also detected in patients with inflammatory diseases (6).
As inflammation is a critical component of tumor progression, repeated and chronic inflammation can increase cancer risk (7). Considering the course of lung carcinogenesis, before which molecular biological change has occurred, MAGE expression in the sputum of inflammatory disease can be regarded as a reasonable event. However, applying this result to clinical diagnosis can increase the probability of false positive diagnoses and lead to difficulties in patient care. For the clinical application of MAGE reverse transcription polymerase chain reaction (RT-PCR), it would be necessary that the MAGE result be carefully interpreted along with additional molecular findings and clinical evidence.
This study was designed to evaluate the performance of MAGE RT-PCR, MAGE A3 methylation-specific PCR (MSP), and p16 MSP using induced sputum of patients with lung cancer or non-cancerous inflammatory diseases, and to compare the results of genetic tests with the patients' clinical findings. Eventually, we attempted to elucidate the clinical significance of MAGE expression in the sputum of patients with pulmonary diseases. During the same period, 133 samples of induced sputum specimens collected from patients with pulmonary problems who visited Yeungnam University Hospital were added to sputum RNA extraction solution (iC&G Co., Daegu, Korea), and stored in a -70˚C refrigerator until required for RNA extraction. Induced sputum was obtained after inhaling Berotec solution (Boehringer Ingelheim, Ingelheim, Germany) and 16 ml of 3% hypertonic saline. To determine the clinical diagnosis for the patients, history taking, physical examinations, bronchoscopy, computed tomography (CT) scan, and histopathological biopsies were performed.

Subjects
Random sputa were collected from 30 healthy volunteers, and treated equally. All stored specimens were blindly transferred to the Department of Laboratory Medicine at Daegu Catholic University Medical Center, then MAGE A1-A6 RT-PCR, MAGE A3 MSP, and p16 MSP were carried out. Tissue and sputum procurement procedures were approved by the Institutional Review Board of the Yeungnam University Hospital. Informed consent was obtained from all patients.
RNA and DNA extraction and RT-PCR. RNA and DNA extraction from sputum specimens was conducted with an iC&G extraction kit using magnetic beads. RNA from tissues was extracted according to the TRI Corporation's instructions, and its DNA was extracted using TRI remnant. RNA was reverse-transcribed using ImProm-II reverse transcription (RT) reagents (Promega Corp., Madison, WI), and MAGE A1-A6 expression was amplified using iC&G PCR reagents. cDNA integrity was confirmed by GAPD amplifications.
Methylation-specific PCR. Unmethylated cytosine was changed to uracil in the CG nucleotide gene of extracted DNA using Cp genome change reagent (Chemicon, Temecula, CA). Nested PCR was performed on treated genomic DNA for amplification of MAGE A3 and p16 MSP. At first,    Table I. The amplified products of MAGE A3 MSP, and p16 MSP were clearly visualized in Fig. 1 and Fig. 2, respectively.
Statistical analysis. The repeated-measure one factor analysis of the Cochran test was performed to compare the positive rates of MAGE A1-A6 RT-PCR, MAGE A3 MSP, and p16 MSP. The Chi-square test was used for comparison of positive rates among the patients' groups. Statistical analyses were conducted using SPSS 14.0 software (SPSS Inc., Chicago, IL). A P-value of <0.05 was considered to indicate statistical significance.

Results
Subjects. The 133 enrolled patients were diagnosed as 65 lung cancer and 68 benign lung diseases. Patients of benign lung diseases were classified as follows: 25 no active lung disease, 16 pulmonary tuberculosis, 11 pneumonia, 11 inflammatory diseases, 2 bullae, 2 pleural effusion of unknown cause, and 1 right middle lobe syndrome.
Positive rates for MAGE A1-A6 RT-PCR, MAGE A3 MSP, and p16 MSP according to the patient group. Positive rates for MAGE A1-A6 RT-PCR, MAGE A3 MSP, and p16 MSP were as follows. In tissues of patients with lung cancer, 87.5, 58.3, and 70.8%; in induced sputa of patients with lung cancer, 50.8, 46.2, and 63.1%; in induced sputa of patients with benign lung disease, 10.3, 30.9, and 39.7%; in random sputa of healthy people, 3.3, 6.7, and 3.3%. All 3 tests showed statistically significant results using the sputum of lung cancer, benign diseases, and healthy groups (P<0.05) (Table II). In the group of lung cancer, MAGE RT-PCR revealed statistically significant higher positive rates while in benign lung diseases, significant lower positive rates than MAGE A3 and p16 MSP results (P<0.05).
Positive rates for MAGE RT-PCR, MAGE A3 MSP, and p16 MSP in patients with benign lung diseases. MAGE expression rates were high in patients with inflammatory diseases and tuberculosis (18.2 and 18.8%, respectively), while positive rates for MAGE A3 unmethylation and p16 methylation were high in patients with tuberculosis (56.3 and 62.5%) and in those with pneumonia (45.5 and 54.5%). The average positive rate  of MAGE, MAGE A3 unmethylation, and p16 methylation in patients with benign lung diseases were 10.3, 30.9, and 39.7% (Table III), showing a statistically significant difference (Table II).
Positive rates for MAGE A3 MSP and p16 MSP in regard to MAGE expression. In the 40 MAGE positive cases of lung cancer and benign lung diseases, positive rates for MAGE A3 unmethylation and p16 methylation were 67.5 and 75.0%, respectively. These results showed significant differences compared with the 26.9% positive rates of MAGE A3 unmethylation and 40.9% of p16 methylation in the 93 MAGE-negative cases (Fig. 3).

Clinical analysis and methylation abnormality in the MAGE-positive patients.
Of the 40 MAGE-positive cases, 33 were diagnosed as lung cancer and 7 as benign lung diseases. The diagnosis of 7 benign lung diseases was as follows: 3 pulmonary tuberculosis, 2 inflammatory diseases, 1 bullae, and 1 no active lung disease. From the 7 cases of benign lung diseases, 5 showed methylation abnormality in both MAGE A3 and p16. One inflammation showed methylation abnormality in p16 only and 1 bullae showed no methylation abnormality in either gene (Table IV).

Discussion
Mortality of lung cancer is still high; therefore, early detection of lung cancer is a major issue. The advent of low-dose spiral chest CT, positron emission tomography, and autofluorescence bronchoscopy (AFB) have opened a new perspective for early detection (8), and specific molecular markers have also been developed (9). Although AFB can detect pre-invasive lesions and lung cancers in the central airway, the specificity of AFB is low (10). A diagnostic test with high sensitivity and high specificity has not yet been developed. MAGE is a highly specific tumor marker, and MAGE A3 expression was observed in 35% of lung cancer patients through a multi-center study (11). Based on these findings, MAGE-A3 vaccination has been developed as a promising treatment modality for lung cancer (12,13). Atanackovic et al (12) reported that 14 of 18 lung cancer patients with stage I or stage II disease had no evidence of disease up to 3 years after vaccination with MAGE A3-protein. In addition to MAGE A3, a DNA methylation-based biomarker is considered a rapid and efficient early detection marker for lung cancer (9,14), and a previous study reported that DNA methylation marker targeting 4 genes in sputum showed 94% sensitivity and 90% specificity (14). Anglim et al demonstrated that aberrant promoter methylation of p16 was observed 3 years before diagnosis of squamous cell carcinoma in smokers (9).
In the present study, MAGE gene-positive rates were 87.5% in lung cancer tissues, 50.8% in induced sputum specimens of patients with lung cancer, 10.3% in induced sputum specimens of patients with benign lung diseases, and 3.3% in random sputum specimens of healthy people. The finding of MAGE expression in benign lung diseases is consistent with those of an earlier study, which reported that MAGE A1 or A2 expression is detected in severe bronchitis and severe actinomycosis with concomitant tissue regeneration (6). MAGE expression was considered an early event in lung carcinogenesis (3) and was detected in precancerous lesions. Therefore, MAGE can be used not only as an early detection marker for lung cancer but also as a prevention marker for lung carcinogenesis. In this study, MAGE positive benign lung diseases were mainly comprised of pulmonary tuberculosis and inflammatory diseases and severity of inflammation was not evaluated. Kim et al (15) reported MAGE expression in tissue samples obtained using a percutaneous needle aspiration biopsy of tuberculosis patients. We performed MAGE A1-A6 nested PCR using DNA of Mycobacterium tuberculosis; however, the MAGE gene was not amplified. Therefore, MAGE expression  of tuberculosis patients may be not the result of a false-positive detection by Mycobacterium tuberculosis but the result of the inflammatory process. Compared with the non-tuberculosis group, tuberculosis patients showed coexistence of lung cancer and high incidence rates of lung cancer (16,17). However, determination of MAGE expression in the sputum based simply on clinical conditions cannot reflect molecular biological change at the cellular level. To reflect molecular biological change, additional molecular biological tests associated with lung carcinogenesis are necessary. In order to conduct a proper mutation analysis, a number of tumor cells and a large number of genetic loci should be investigated. Aberrant methylation of the p16 promoter is an important mechanism of lung carcinogenesis (14) and unmethylation of the MAGE A3 promoter is directly associated with MAGE expression. Therefore, we performed MAGE A3 and p16 MSP. The MSP is a promising method for detection of lung cancer because it can detect a small number of cancer cells in sputum that contains a large number of normal cells (9).
In the present study, unmethylation rates for MAGE A3 MSP were 46.2% in sputa of patients with lung cancers, 30.9% in patients with benign lung diseases, and 6.7% in healthy people. Though unmethylation rates for MAGE A3 MSP have not yet been reported in the literature review, these findings are similar to the results of Olaussen et al (18), which reported that positive rates for MAGE A1 MSP were 50% in cytologically negative sputum from lung cancers patients, 45% in sputum showing inflammatory change from smokers, and 6% in cytologically negative sputum from smokers.
Positive rates for p16 MSP were 63.1% in sputa of lung cancer patients, 39.7% in patients with benign lung diseases, and 3.3% in healthy people. Olaussen et al (18) also reported that positive rates for p16 MSP were 27% in cytologically negative sputum from lung cancers patients, 64% in sputum showing cancerous cytology from smokers, 27% in sputum showing inflammatory change from smokers, and 47% in sputum showing normal cytology from smokers. Among studies performed in the Korean population, one study reported that p16 methylation was detected in 67% of tumor samples (19). However, another study reported 22% positive rates in tumor samples, and 1% in the corresponding non-malignant lung tissues (20). The methylation rate of p16 in lung cancer tissues was about 80% (21,22). Liu et al (22) reported methylation rates of 74.7% in the sputum from lung cancer patients, and 51.4% from people exposed to coal smoke. Using matched specimens from lung cancer patients, Hsu et al (23) reported 37% methylation rates in lung cancer tissues, 33% in sputa, 13% in normal lung tissues, and 14% in sputa from the control group. Although reported rates for p16 methylation were variable, p16 MSP was regarded as a useful molecular marker for use in early detection and prediction of lung cancer.
Therefore, analysis of molecular abnormality in the same specimen using MAGE A1-A6 RT-PCR, MAGE A3 MSP, and p16 MSP simultaneously may be very useful for detection and prediction of lung cancer at the molecular level. This analysis can also be applied to non-cancerous groups to understand the clinical significance of MAGE expression. In this and other studies, positive rates of MSP remain high in the non-cancerous group; therefore, abnormality of MSP was just utilized as a supplemental modality to explain MAGE expression in the sputum, not as a cancer detection tool.
The MAGE-positive group in sputum showed a statistically significant higher MAGE A3 unmethylation and p16 MSP methylation rate than the MAGE negative group. From the 7 cases of benign lung diseases with MAGE expression, 5 showed methylation abnormality in both MAGE A3 and p16, 1 case showed methylation abnormality in p16 only, and 1 case showed methylation abnormality in neither MAGE A3 nor p16. Although the MAGE A3 gene could not be expressed without unmethylation of MAGE A3, another gene could be amplified because common primers that can amplify MAGE A1-A6 together had been used. Moreover, MAGE A3 MSP is not a quantitative MSP of promoter loci, but reflects methylation status of primer binding sites. Therefore, the results of MSP may not be consistent with the results of RT-PCR.
The question of how to interpret MAGE expression in patients with benign lung diseases is an important issue. MAGE proteins form complexes with KAP1, suppress p53-dependent apoptosis, and contribute to cancer development (24). We analyzed MAGE expression acting as a tumor enhancer, and also detected the methylation status of MAGE A3 and p16 using the same specimens. Of 40 MAGE positive cases in the sputa, 39 cases turned out to be lung cancers or benign lung diseases accompanying methylation abnormality. Therefore, MAGE expression in the clinical specimen may suggest the presence of cells in the process of molecular carcinogenesis, even if cancer cells are not visible.
Thus, the clinical significance of MAGE expression in the sputum would be i) the presence of lung cancer cells or ii) pre-cancerous cells. In another study (25), we evaluated MAGE expression in the peritoneal washes of gastric carcinoma patients, and the patients were followed up for 5 years in regard to MAGE expression. As a result, recurrence rates of MAGE-positive cases (45.5%) were much higher than those of MAGE negative cases (9.6%). Although gastric carcinogenesis may not be identical with lung carcinogenesis, MAGE expression in clinical specimens can be considered a strong indication of tumor recurrence. However, not all precancerous cells will develop into cancer, and MAGE expression was regarded as a reversible change. Thus, MAGE expression in specimens of non-malignant patients should be interpreted very carefully, and for proper clinical application, other clinical information, such as 5-year follow-up results would be studied.
In 2001, Jang et al (3) predicted that the MAGE gene would be utilized as a tool for lung cancer prevention. Since then, functions of the MAGE gene as a tumor promoter have been disclosed, and tumor therapeutic agents targeting the MAGE gene have been developed and will soon be applied to lung cancer treatment. The recurrence rate of MAGE-positive cases in peritoneal washes of gastric carcinoma patients was 45.5%. In the present study, the MAGE A3 unmethylation or p16 methylation abnormality was demonstrated in MAGEpositive specimens from non-cancerous patients. Based on these findings, MAGE expression in the sputum may indicate the presence of lung cancer cells or pre-cancerous cells. Therefore, a MAGE positive case in the non-cancer group should be closely followed up. In conclusion, MAGE could be utilized as a cancer prediction tool as well as a cancer detection tool. Further studies including molecular markers, histological examination, and clinical studies targeting non-cancerous patients with MAGE expression are inevitable.