Skip to main content
PMC home page
The Journal of Clinical Investigation logoLink to The Journal of Clinical Investigation
. 2003 Mar 15;111(6):887–895. doi: 10.1172/JCI15475

Inactivation of hMLH1 and hMSH2 by promoter methylation in primary non-small cell lung tumors and matched sputum samples

Yi-Ching Wang 1, Yung-Pin Lu 1, Ruo-Chia Tseng 1, Ruo-Kai Lin 1, Jer-Wei Chang 1, Jung-Ta Chen 2, Chuen-Ming Shih 3, Chih-Yi Chen 4
PMCID: PMC153761  PMID: 12639995

Abstract

We performed a genetic and epigenetic study of the hMLH1 and hMSH2 mismatch repair genes in resected primary tumors from 77 non-small cell lung cancer (NSCLC) patients. The molecular alterations examined included the loss of mRNA and protein expression as well as promoter methylation, and the allelic imbalance of the chromosomal regions that harbor the genes. We found that 78% and 26% of patients showed at least one type of molecular alteration within the hMLH1 and hMSH2 genes, respectively. Promoter methylation of the hMLH1 gene was present in 55.8% of tumors, and was significantly associated with the reduction in mRNA and protein expression (P = 0.001). A 72% concordance of aberrant methylation in sputum samples with matched resected tumors was found. In addition, a 93% consistency between the promoter methylation and the mRNA expression of the hMSH2 gene was found in 14 female NSCLC patients. However, no correlation was found between the expression of hMLH1 and hMSH2 proteins and the allelic imbalance of five microsatellite markers closely linked to the genes. Our results suggest that hMLH1 is the major altered mismatch repair gene involved in NSCLC tumorigenesis, and that promoter methylation is the predominant mechanism in hMLH1 and hMSH2 deregulation. In addition, promoter methylation of the hMLH1 gene may be identified in sputum samples to serve as a potential diagnostic marker of NSCLC.

Introduction

Lung cancer is one of the most common malignancies in the world and is the leading cause of cancer deaths in industrial countries, including Taiwan (1). Recently, a marker for genetic instability, microsatellite instability (MSI), has been identified in a class of familial colon carcinomas known as hereditary nonpolyposis colorectal cancer (2, 3) and in a wide variety of other human cancers including lung cancer (4–8). It was reported that MSI was associated with mutations in mismatch repair genes such as hMLH1 and hMSH2 (2, 3) in hereditary nonpolyposis colorectal cancer. Somatic mutations and promoter hypermethylation of mismatch repair genes have also been shown in a proportion of colon, gastric, and endometrial tumors with MSI (5–10). However, the data on mismatch repair gene alterations in lung cancer is scarce.

To examine the etiological association of genetic instability in lung tumorigenesis, we previously investigated the frequency of MSI as well as the association between MSI and expression of hMLH1 mismatch repair protein in 68 patients with non-small cell lung cancer (NSCLC) (11). Forty-one percent of the patients demonstrated instability in multiple tested microsatellite markers and 77% of MSI patients showed no expression of hMLH1 protein. In addition, hMLH1 protein expression was undetectable in 60.9% of NSCLCs. In support of our study, Xinarianos et al. (12) have also reported that 58.6% of NSCLC samples had reduced expression levels of the hMLH1 protein. However, they found no mutations in the promoter or exons of the hMLH1 gene examined.

To further identify the molecular basis for loss of hMLH1 protein expression and to examine the possible involvement of another mismatch repair protein, hMSH2, in NSCLC tumorigenesis, we performed a comprehensive genetic and epigenetic study of protein and mRNA expression as well as promoter methylation and loss of heterozygosity (LOH) of the hMLH1 and hMSH2 genes in 77 resected primary NSCLC samples, performed in parallel to the clinicopathological analyses of the patients. Our data indicate that promoter methylation is the predominant mechanism in immunohistochemical negativity of the hMLH1 and hMSH2 genes and that promoter methylation of the hMLH1 gene can be identified in sputum samples to serve as a potential diagnostic marker of NSCLC.

Methods

Study population, tumor samples, and sputum samples.

The study subjects were 77 patients diagnosed with primary NSCLC admitted to Veterans General Hospital, Taichung, Taiwan, between 1993 and 2000. Forty-two of them had squamous cell carcinomas (SQs), 28 had adenocarcinomas (ADs), four had adenosquamous cell carcinomas, and three had large-cell carcinomas. Histological classification of the tumor types and stages was performed according to the World Health Organization classification method and the Tumor, Node, Metastasis System, respectively. Information on the smoking history of the lung cancer patients was obtained from hospital records. The study was reviewed and approved by the institution’s Surveillance Committee, which allowed us to obtain tissue samples and all pertinent followup information.

Surgically resected tumor samples were immediately snap-frozen and were subsequently stored in liquid nitrogen. For the methylation assay, genomic DNA was prepared using proteinase K digestion and phenol/chloroform extraction, followed by ethanol precipitation. For LOH analysis, samples were microdissected in order to recover tumor tissue from up to five serial 5-μm sections of paraffin-embedded tumor specimens. Following dewaxing in xylene, genomic DNA was extracted according to the standard methods described above.

Sputum samples were obtained from 29 NSCLC patients 3 months before surgery, following standardized procedures. In addition, sputum samples obtained from ten cancer-free individuals served as control during methylation assays. DNA was extracted from 50% of the homogenized pellets of sputum using a tissue kit (QIAGEN GmbH, Hilden, Germany) according to the manufacturer’s specifications.

Analysis of protein expression: immunohistochemistry assay.

Blocks of paraffin-embedded tumors were cut into 5-μm slices and then processed using protocols described previously (11). The tumors were analyzed for hMLH1 and hMSH2 protein expression by the immunohistochemistry assay. Monoclonal antibodies used were G168-728 (1:250; PharMingen, San Diego, California, USA) for the hMLH1 protein and FE11 (1:50; Oncogene Science, Cambridge, Massachusetts, USA) for the hMSH2 protein. The normal staining pattern for hMLH1 and hMSH2 is nuclear. Tumor cells that exhibited an absence of nuclear staining in the presence of non-neoplastic cells and infiltrating lymphocytes with nuclear staining were considered to have an abnormal pattern. Staining results were examined without knowledge of the status of the molecular analyses.

Analysis of mRNA expression: RT-PCR assay.

Total RNA was prepared from matched pairs of primary tumors and nearby normal lung tissues using TRIZOL reagent (Gibco; Invitrogen, Carlsbad, California, USA). SuperScript reverse transcriptase (Gibco; Invitrogen) was used to synthesize cDNA with protocols provided by the manufacturer.

Expression levels of the hMLH1 and hMSH2 genes were detected by a multiplex PCR assay using the β-actin gene as an internal control to indicate the amount of RNA in each reaction. The primer nucleotide sequences were as follows: for the hMLH1 gene, sense, 5′-GTGCTGGCAATCAAGGGACCC-3′, antisense, 5′-CACGGTTGAGGCATTGGGTAG-3′; for the hMSH2 gene, sense, 5′-GTCGGCTTCGTGCGCTTCTTT-3′, antisense, 5′-TCTCTGGCCATCAACTGCGGA-3′; for the β-actin gene, sense, 5′-ACACTGTGCCCATCTACGAGG-3′, antisense, 5′-AGGGGCCGGACTCGTCATACT-3′. PCR was performed for 35 cycles with an annealing temperature of 66°C using cDNA synthesized from 50 ng total RNA.

To quantify the relative levels of mRNA expression in the multiplex RT-PCR assay, the value of the internal standard (β-actin) in each reaction was used as the baseline gene expression of that sample and the relative value was calculated for the hMLH1 and hMSH2 genes for each tumor and matched normal samples. Tumor cells that exhibited a lower expression level than did normal cells were considered to have an abnormal pattern.

HpaII-based multiplex PCR methylation assay for the hMLH1 gene in tumor and sputum samples.

The promoter methylation status of the hMLH1 gene of 77 tumor samples and 29 sputum samples was investigated using HpaII-based multiplex PCR analysis. Genomic DNA (200 ng) was double digested either with 10 U of methylation-sensitive enzyme (HpaII; New England Biolabs Inc., Beverly, Massachusetts, USA) or without enzyme for 10 hours at 37°C. One hundred nanograms of the digested DNA was subjected to multiplex PCR. The primer nucleotide sequences for the hMLH1 promoter region (–670 to –67 bp) containing four HpaII sites used for analyses of resected tumor samples were as described (5). The internal control used for analyses was the IFN-β1 sequence (sense, 5′-ATGAGCTACAACTTGCTTGGA-3′, antisense, 5′-TCAGTTTCGGAGGTAACCTGT-3′), which contains no HpaII site. PCR was performed for 35 cycles at an annealing temperature of 64°C. For sputum samples, a second PCR using 1:100 of the product of the first PCR was performed. Methylation was determined by the ratio of the normalized intensity (hMLH1/internal control) of the samples pretreated with HpaII to that of the undigested samples. DNA of normal lung tissues from 28 NSCLC patients and of sputum samples from ten cancer-free individuals was examined; ratios were less than 0.2 for all samples. Therefore, a ratio greater than 0.2 was defined as aberrant methylation.

Methylation-specific PCR assay for the hMLH1 and hMSH2 genes.

The promoter methylation status of the hMLH1 gene of all tumor samples and of the hMSH2 gene of tumor samples from 14 female lung cancer patients was determined by chemical treatment with sodium bisulfite and subsequent methylation-specific PCR (MSP) analysis as described (13). Primer sequences were as described (14). PCR was performed for 40 cycles with annealing temperatures of 65°C and 62°C for unmethylated and methylated reactions, respectively, using 50 ng bisulfite-modified DNA. All PCRs were performed with positive controls for both unmethylated and methylated alleles and no DNA control.

Treatment of lung cancer cells with 5-aza-2′-deoxycytidine.

The human lung cancer cell lines CL2 and CL1-5-F4 were derived from Taiwanese patients with AD of the lung (kindly provided by P.-C. Yang, Department of Internal Medicine, National Taiwan University Hospital, Taipei, Taiwan). Cells were plated at 105 per 100-mm culture dish on the day before treatment. The cultures were treated for three doubling times with 1 μM 5-aza-2′-deoxycytidine (5-aza-dC). The 5-aza-dC–containing medium was changed after each cell doubling during the treatment. On the day after the third doubling, cells were harvested for analysis of their methylation status using an MSP assay and analysis of mRNA production by an RT-PCR assay. In experiments for protein analysis, the cells were given fresh media without 5-aza-dC for two more days and then fixed and stained for protein expression as judged by immunohistochemistry.

LOH at chromosome regions 3p21.3-22 and 2p16.3-22 in microdissected tumor samples.

Five microsatellite polymorphic markers for the hMLH1 and hMSH2 genes were used for LOH analysis: D3S3718 (3p21.33; 0.9 cM distal to the hMLH1 locus), D3S1768 (3p21.3-22; 2.5 cM distal to the hMLH1 locus), D3S1612 (3p21.3-22; 2.5 cM distal to the hMLH1 locus), D2S123 (2p16.3; 3.6 cM proximal to the hMSH2 locus), and D2S1788 (2p22.3, 11.5 cM distal to the hMSH2 locus). Twenty nanograms of genomic DNA from normal lung tissues or microdissected tumor samples of 77 patients was used for PCR-based LOH analysis. The primer sequence for each microsatellite marker was obtained from the National Center for Biotechnology Information database. The allelic ratio was calculated as (T1/T2):(N1/N2), the ratio of tumor (T) alleles to normal (N) alleles. LOH was defined as an allelic ratio above 2.0 or below 0.5.

Analysis of MSI.

MSI was analyzed in 21 sputum DNA samples using four microsatellite markers: D3S1234 (3p14), D9S162 (9p22-23), D13S170 (13q22-31), and D17S786 (17p13.1). These markers were selected because they were highly sensitive at detecting MSI in our previous study (11). PCR conditions were as described (11). MSI was revealed by the presence of expansions or contractions of repeats in sputum DNA that were absent from normal DNA from lung tissue of the same patient. An MSI-positive result was defined as one demonstrating instability in one or more markers.

Statistical analysis.

The Pearson χ2 test was used to compare the frequency of hMLH1 and hMSH2 alterations between NSCLC patients with different characteristics, including sex and smoking status, and various clinicopathological parameters, such as tumor type and tumor stage, respectively. Comparison of age between patients with and without the alteration was made by the two-sample t test. The Pearson χ2 test was also used for comparative analysis of expression of hMLH1 and hMSH2 proteins with allelic imbalance data.

Results

Protein and mRNA expression of hMLH1.

We investigated hMLH1 protein expression using immunohistochemical analysis (Figure 1, a–c), and mRNA expression using RT-PCR analysis (Figure 2) in tumors from 77 primary NSCLC patients. The association of altered protein and mRNA expression with the clinicopathological parameters of patients was also examined (Table 1). Forty specimens (51.9%) exhibited altered protein expression, and 43 specimens (55.8%) showed an alteration of mRNA expression. A reduction in hMLH1 protein expression was found more frequently in tumors from nonsmoking patients than in those from smoking patients (P = 0.06). In addition, the lack of mRNA expression was significantly more prevalent in AD patients (71.4%, 20 of 28) than in SQ patients (45.2%, 19 of 42) (P = 0.031).

Figure 1.

Figure 1

Open in a new tab

Immunohistochemical staining of hMLH1 protein expression (a–c) and hMSH2 protein expression (d–f) in paraffin-embedded sections of lung tumor specimens. hMLH1 nuclear immunoreactivity was found in a and b. hMSH2 nuclear immunoreactivity was found in d and e. Tumors negative for hMLH1 and hMSH2 are shown in c and f, respectively. Original magnification ×100.

Figure 2.

Figure 2

Open in a new tab

Representative examples of multiplex RT-PCR analysis of the hMLH1 and hMSH2 genes in lung cancer patients. Amplification of the β-actin gene indicated the initial RNA template used for each reaction. Patient numbers are the sample numbers. N and T represent the paired normal and tumor lung cells from the same patient. M, molecular marker.

Table 1.

Genetic alterations of the hMLH1 gene in relation to the clinicopathological parameters of resected NSCLC tumors

graphic file with name JCI0315475.t1.jpg

Open in a new tab

PROMOTER METHYLATION OF HMLH1.

We examined the methylation status of the hMLH1 gene in 77 resected tumors using HpaII-based PCR methylation analysis and an MSP assay. Representative examples of HpaII-based PCR methylation analysis are shown in Figure 3b. Figure 3a shows virtually no methylation in the hMLH1 promoter of 28 normal lung tissues examined. Methylation of the hMLH1 gene was present in 43 of the 77 (55.8%) tumors (Table 1). A 74% concordance of the methylation status for the hMLH1 gene was found between the HpaII-based PCR and the MSP assays (Figure 3d).

Figure 3.

Figure 3

Open in a new tab

Representative examples of promoter methylation analysis of the hMLH1 (a–d) and hMSH2 (e) genes. (a–c) HpaII-based multiplex PCR assay for the hMLH1 promoter. (a) Normal lung tissues, (b) tumor lung samples, and (c) matched sputum samples, indicated by n, t, and s before subject identification number. +, HpaII digest; –, mock digest. Positions of the target hMLH1 gene and control IFN-β1 gene are indicated. Normal lung tissues showed no hMLH1 promoter methylation. MSP assay for the hMLH1 promoter (d) and (e) hMSH2 promoter. Primer sets used for amplification are designated U for unmethylated or M for methylated genes. The data for lymphocytes containing unmethylated hMSH2 are shown for the experimental control. + and – in the rows labeled methyl and mRNA represent the status of promoter methylation and mRNA expression.

Aberrant promoter methylation of the hMLH1 gene was found in 55% of stage I and II patients, suggesting that it is an early and very frequent event in NSCLC. We then examined whether aberrant methylation might also be found in the sputum of patients with NSCLC and whether aberrant methylation can track the cancer cell even in patients for which conventional sputum cytology is not sensitive enough. Cytologically negative sputum samples were available from 29 NSCLC patients and were analyzed for hMLH1 methylation status (Figure 3c). HpaII-based PCR analysis showed aberrant methylation in 11 of the 29 (37.9%) sputum samples. In addition, a 72% (21 of 29) concordance of sputum samples with matched resected tumors was found (Table 2).

Table 2.

Comparison of hMLH1 promoter methylation and MSI in matched sputum and tumor DNA from NSCLC patients

graphic file with name JCI0315475.t2.jpg

Open in a new tab

To further compare the sensitivity and specificity of promoter methylation and MSI in detecting NSCLC in sputum samples, 21 sputum samples were also examined for MSI. Sensitivity was 60% for methylation analysis, but only 22% for MSI analysis (Table 2).

Protein and mRNA expression of hMSH2.

We investigated expression of hMSH2 protein (Figure 1, d–f) and mRNA (Figure 2) in tumors from 77 primary NSCLC patients. The association of altered protein and mRNA expression with the clinicopathological parameters of patients was also examined (Table 3). Fourteen specimens (18.2%) and thirteen specimens (16.9%) exhibited an alteration of protein and mRNA expression, respectively. Negative hMSH2 protein expression was significantly associated with younger patients (P = 0.021), female patients (42.9%, P = 0.005), and nonsmokers (40.0%, P = 0.003) (Table 3). In addition, the number of mRNA-negative samples was significantly higher in AD patients (28.6%, 8 of 28) than in SQ patients (9.5%, 4 of 42) (P = 0.038).

Table 3.

Genetic alterations of the hMSH2 gene in relation to the clinicopathological parameters of resected NSCLC tumors

graphic file with name JCI0315475.t3.jpg

Open in a new tab

PROMOTER METHYLATION OF HMSH2.

We examined the methylation status of the hMSH2 gene in resected tumors from 14 female lung cancer patients using the MSP assay because reduction in hMSH2 expression appeared to be more prevalent in female patients (Table 3). Representative examples of the MSP analysis are shown in Figure 3e. Methylation of the hMSH2 gene was present in four (28.6%) tumors from female lung cancer patients. These four tumors all lacked hMSH2 mRNA expression. In addition, a 93% (13 of 14) concordance in mRNA expression and promoter methylation patterns was found.

Correlation of promoter methylation with negative mRNA and protein expression, and reactivation of hMLH1 and hMSH2 by 5-aza-dC treatment.

The data for protein expression, mRNA expression, and promoter methylation were then cross-examined to investigate the correlation among these three parameters for the hMLH1 gene (Figure 4a). The results indicated that negative protein expression was significantly associated with negative mRNA expression and promoter methylation of the hMLH1 gene (P = 0.001). The concordant group was found in 68.9–71.5% of the patients analyzed. In addition, negative protein expression was significantly associated with negative hMSH2 mRNA expression (Figure 4b; P = 0.002). However, there was no correlation between altered expression of hMSH2 and hMLH1 (data not shown).

Figure 4.

Figure 4

Open in a new tab

Concordance analysis between protein expression, mRNA expression, and promoter methylation of hMLH1 (a) and hMSH2 (b). The percentage of cases is indicated on the y axis, whereas the type of comparison is shown on the x axis. + indicates positive protein expression (protein), positive mRNA expression (RNA), and hypermethylation (methyl) of the promoter. Numbers above bars indicate percentages of the total concordant group (gray sections) and nonconcordant group (white sections). P = 0.001 for association between protein expression, mRNA expression, and promoter methylation in the hMLH1 gene; P = 0.002 for association between protein expression and mRNA expression in the hMSH2 gene.

To determine whether hMLH1 and hMSH2 promoter methylation could be further linked to the loss of gene expression, two lung cancer cells that showed negative expression and promoter hypermethylation of the hMLH1 or hMSH2 gene were treated with the demethylating agent 5-aza-dC. As shown in Figure 5, treatment with 5-aza-dC successfully restored mRNA and protein expression and demethylated the promoter region in each of the cells that lacked hMLH1 or hMSH2 expression and that harbored a methylated respective promoter.

Figure 5.

Figure 5

Open in a new tab

Effects of 5-aza-dC treatment on CL2 and CL1-5-F4 cells. (a) Demethylation analysis by MSP assay of hMLH1 and hMSH2 in lung cancer cell lines after 5-aza-dC treatment. The presence of the unmethylated product (U) after 5-aza-dC treatment indicates demethylation of the promoter in these cell lines. (b) RT-PCR analysis of mRNA expression of the hMLH1 and hMSH2 genes in cell lines. (c) Immunohistochemistry analysis of hMLH1 (original magnification ×100) and hMSH2 (original magnification ×200) protein expression in cell lines.

No correlation of allelic imbalance at 3p21.3-22 and 2p16.3-22 with hMLH1 and hMSH2 protein expression.

Microdissected genomic DNA from tumors and matched normal lung tissue was examined for the incidence of allelic imbalance at chromosome regions 3p21.3-22 and 2p16.3-22, where hMLH1 and hMSH2 reside, using five microsatellite markers. The frequencies of LOH were 15.8–37.5% and 9.3–36.4% for the regions at 3p21.3-22 and 2p16.3-22, respectively. The patients with LOH at these regions were mostly SQ patients (data not shown). Statistical analysis showed close correlation between the LOH of markers nearby, suggesting that LOH at these markers represents the frequency of allelic imbalance in the region where they are located (data not shown). However, comparative analysis of allelic imbalance data of these five markers with expression of hMLH1 and hMSH2 protein showed no correlation (Table 4; P > 0.05).

Table 4.

Expression levels of hMLH1 and hMSH2 in NSCLC patients in relation to allelic imbalance at chromosomes 3p21.3-22 and 2p16.3-22

graphic file with name JCI0315475.t4.jpg

Open in a new tab

Discussion

Aberrant methylation, which can result in the transcriptional silencing of the target gene, was frequently found in the hMLH1 and hMSH2 alterations in NSCLC tumors. A highly significant correlation was found between hMLH1 gene methylation and negative protein and mRNA expression (P = 0.001), and a 93% concordance of hMSH2 methylation and mRNA expression patterns was shown in tumors from female lung cancer patients, suggesting that promoter methylation is the predominant mechanism by which these two genes are silenced in NSCLC. This conclusion is further strengthened by the re-expression of mRNA and protein, together with demethylation at the promoter region in both the hMLH1 and hMSH2 genes, after 5-aza-dC treatment of two lung cancer cell lines. In addition, the frequent occurrence of promoter methylation of the hMLH1 gene in early-stage tumors and sputum samples indicates its potential use as a diagnostic marker in NSCLC.

In this study, both HpaII and MSP assays were used to examine the methylation status of the hMLH1 gene. HpaII-based PCR analysis examines methylation status at the region from –670 to –67 bp relative to the transcription start site of the hMLH1 gene. This region contains four HpaII sites at –567, –527, –374, and –341 bp. MSP analysis examines the methylation status at region –721 to –598 bp. The primer design can detect a total of seven CpG sites at –715, –709, –695, –693, –621, –619, and –609 bp. The methylation frequencies detected by the HpaII-based PCR and MSP assays were 55.8% and 37.7%, respectively. A 74% concordance of the methylation status for the hMLH1 gene between the two assays was observed. There seem to be at least three possible explanations for the 26% discordance between the two assays. First, the lower methylation frequency detected by the MSP assay compared with the HpaII-based assay may partly result from the fact that the MSP method was designed to analyze more CpG sites than was the HpaII method. Second, the regions examined in the two assays were different, and the methylation frequency of the various CpG sites may have been different. Finally, there may have been several distinct subpopulations present in the discordant tumors, subpopulations that had a mixture of cells with methylated and nonmethylated DNA or that each had methylation at different sites. It has been shown that the presence of promoter hypermethylation for a given gene can be heterogeneous at different CpG sites and/or with various tumor samples (15, 16). Several potential transcription factor binding sites, such as the Sp1, CCAAT-box, C/EBP, and c-Myb binding sites, are located in the region from –552 to –132 bp (5, 17, 18). In addition, the methylation status of the promoter region examined by the HpaII assay is better correlated to the transcriptional activity of the hMLH1 gene than is the MSP assay. therefore, we used the HpaII result for all analyses, including correlation with clinicopathological parameters and gene expression. To ensure the reliability of the HpaII methylation assay, efforts were made to avoid incomplete digestion and false amplification using strict quality-control procedures. We consider technique artifacts to be an unlikely explanation for the discordance between the two assays. It will be important to examine the methylation status of all CpG sites within a much larger region to more precisely define the site critical for gene silencing in the future.

It has been reported that promoter methylation can be a very early event in NSCLC tumorigenesis (19, 20). This is consistent with our results in that hMLH1 methylation was detected in 55% of stage I and stage II tumors (Table 1) and in 37.9% (11 of 29) of the cytologically negative sputum samples. In addition, a 72% (21 of 29) concordance of hMLH1 methylation status was detected in resected tumors and matched sputum samples (Figure 3). Sensitivity and specificity for methylation analysis in sputum DNA were 60% and 86%, respectively (Table 2). To our knowledge, this is the first report of the potential use of promoter methylation of the hMLH1 gene in sputum samples for the diagnosis of NSCLC, and is also the first report on the comparison of detection sensitivity between promoter methylation and MSI in sputum. The data from the hMLH1 methylation analysis were consistent with the data reported previously for the detection of aberrant p16INK4a methylation (100% sensitivity) and O6-methylguanine DNA methyltransferase gene methylation (67%) in sputum DNA (21). Only two studies to date have reported the presence of MSI in the sputum of two SCLC patients (22) and in sputum from three of five lung cancer patients (23). Our data showed better sensitivity by promoter methylation than by MSI in detecting cancer in sputum DNA. Ahrendt et al. (24) have also reported p16 methylation to be more sensitive than MSI in detecting cancer in bronchoalveolar lavage samples. It was estimated that the detection limit can be as sensitive as one methylation allele in 1,000–50,000 unmethylated alleles using the two-stage PCR approach for hMLH1 promoter methylation as we did (17, 21). However, the detection limit is only 1 in 200–500 for MSI analysis (22). Studies of DNA methylation in lung cancer and matched sputum samples to date strongly suggest that the analysis of DNA methylation patterns could become a powerful tool for accurate and early lung cancer diagnosis, with reasonable specificity and sensitivity. Our findings suggest that hypermethylation changes in multiple genes in sputum DNA might detect a very high-risk status or the actual presence of cancer and may be used as a sensitive and reliable molecular diagnostic method. This should be confirmed in larger patient subsets.

The observation of association of altered hMLH1 and hMSH2 mRNA/protein expression with nonsmoking AD patients is intriguing. Registry data indicate that only a few female lung cancer patients smoke cigarettes in Taiwan (25). In addition, most cases of lung cancer in nonsmoking women are AD, which shows less correlation with smoking (1, 25). It is possible that geographical and/or ethnic factors account for frequent hMLH1 and hMSH2 alterations in Taiwanese NSCLC patients. In addition, environmental factors other than cigarette smoke and/or endogenous factors may be involved in the observed association of hMLH1 and hMSH2 alterations with nonsmoking AD. These possibilities may be attributable to the discrepancy between the present study and a USA study that found no promoter methylation of the hMLH1 gene in 20 NSCLC tumors (26). In addition, the varying results may be due in part to differences in the type and/or stage of tumors analyzed in the contrasting studies. For example, a study analyzing the protein expression levels of hMLH1 and hMSH2 in 33 bronchioloalveolar carcinomas showed no detectable alteration in these proteins (27). Bronchioloalveolar carcinoma is a subtype of lung AD with distinct clinical and pathologic features (28).

We observed that the allelic imbalance at chromosome regions 3p21.3-22 and 2p16.3-22, which harbor the hMLH1 and hMSH2 genes, respectively, was not associated with altered expression of hMLH1 and hMSH2 protein (Table 4). Our data on hMSH2 is in agreement with previous studies that showed low LOH frequency (11–15%) at 2p16.3-22 (29, 30) and no association of allelic imbalance of chromosome 2p with negative hMSH2 protein expression (12). However, Xinarianos et al. (12) found that reduced expression of hMLH1 protein correlated with allelic imbalance at chromosome 3p21 based on the study of a marker (D3S1289) located 17.1 cM from the hMLH1 locus. The investigators concluded that allelic loss seems to be a major genetic event involved in hMLH1 silencing. However, hMLH1 promoter methylation was not performed in their study. The data to date, including our study, suggest that chromosome 3p, especially 3p12, 3p14.2, and 3p21.3, represents highly unstable regions that undergo frequent allele loss associated with lung cancer of SQ and with smoking exposure (12, 29, 31, 32). However, no homozygous deletions or rearrangements in the hMLH1 gene have been found to be associated with lung cancer so far (19, 29, 33). A gene dosage effect on the stoichiometry and the activity of the mismatch repair complex has been proposed to explain the observed high frequency of LOH at the chromosome region harboring the hMLH1 gene (29). However, the 3p21.3 region contains the most common known fragile site in the genome (31), and the frequency of discontinuous LOH reported in various studies makes the possibility of mitotic recombination highly likely (31). Whether the frequent hemizygous 3p21.3 allelic deletions reflect only the fragility of this region with a predisposition to recombination events induced after smoking exposure or whether they also indicate the presence of an underlying tumor suppressor gene (or genes), perhaps resulting in tumorigenesis because of haploinsufficiency, merits further clarification.

Although mutational analysis was not performed on our samples, several studies have shown that no mutation was identified in the promoter region or hot-spot exons of the hMLH1 (19, 29) and hMSH2 (29) genes in NSCLC samples. This suggests that mutations are unlikely to be a major cause of hMLH1/hMSH2 inactivation in lung tumorigenesis. Our data strongly suggest that promoter methylation is the predominant mechanism involved in hMLH1 and hMSH2 alteration in NSCLC. More than 20 cases clearly showed retention of both alleles as well as aberrant methylation, indicating biallelic inactivation of hMLH1 expression by promoter methylation. It is possible that these tumors belong to the so-called CpG island methylator phenotype group and have many genes whose expression is downregulated because of aberrant promoter methylation. Our published data on p16INK4a (34) and preliminary data on FHIT gene analyses showed that patients with aberrant promoter methylation in hMLH1 and/or hMSH2 often had hypermethylation at the promoter of multiple tumor suppressor genes. Although the mechanism of aberrant hypermethylation is unknown, it may be caused by increases in de novo methylation activity (35–37) or by a defect of the protection mechanism against de novo methylation (38). More study of candidate genes and genomic screening techniques that permit simultaneous analysis of promoter methylation of many genes should lead to a more complete knowledge of the epigenetic events occurring in tumors and their functional consequences.

Acknowledgments

This work was supported in part by grants NSC 90-2318-B-003-001 and NSC 91-3112-P-003-001 from the National Science Council, The Executive Yuan, Taipei, Taiwan, Republic of China.

Footnotes

Conflict of interest: The authors have declared that no conflict of interest exists.

Nonstandard abbreviations used: microsatellite instability (MSI); non-small cell lung cancer (NSCLC); loss of heterozygosity (LOH); squamous cell carcinoma (SQ); adenocarcinoma (AD); methylation-specific PCR (MSP); 5-aza-2′-deoxycytidine (5-aza-dC).

References

  • 1.Department of Health, The Executive Yuan, Republic of China. 2000. General Health Statistics, 1999. In Health and vital statistics, Republic of China. Republic of China Press. Taipei, Taiwan, Republic of China. 89–112.
  • 2.Fishel R, et al. The human mutator gene homolog MSH2 and its association with hereditary nonpolyposis colon cancer. Cell. 1993;75:1027–1038. doi: 10.1016/0092-8674(93)90546-3. [DOI] [PubMed] [Google Scholar]
  • 3.Bronner CE, et al. Mutation in the DNA mismatch repair gene homologue hMLH1 is associated with hereditary non-polyposis colon cancer. Nature. 1994;368:258–261. doi: 10.1038/368258a0. [DOI] [PubMed] [Google Scholar]
  • 4.Arzimanoglou II, Gilbert F, Barber HR. Microsatellite instability in human solid tumors. Cancer. 1998;82:1808–1820. doi: 10.1002/(sici)1097-0142(19980515)82:10<1808::aid-cncr2>3.0.co;2-j. [DOI] [PubMed] [Google Scholar]
  • 5.Kane MF, et al. Methylation of the hMLH1 promoter correlates with lack of expression of hMLH1 in sporadic colon tumors and mismatch repair-defective human tumor cell lines. Cancer Res. 1997;57:808–811. [PubMed] [Google Scholar]
  • 6.Cunningham JM, et al. Hypermethylation of the hMLH1 promoter in colon cancer with microsatellite instability. Cancer Res. 1998;58:3455–3460. [PubMed] [Google Scholar]
  • 7.Leung SY, et al. hMLH1 promoter methylation and lack of hMLH1 expression in sporadic gastric carcinomas with high-frequency microsatellite instability. Cancer Res. 1999;59:159–164. [PubMed] [Google Scholar]
  • 8.Esteller M, et al. hMLH1 promoter hypermethylation is an early event in human endometrial tumorigenesis. Am. J. Pathol. 1999;155:1767–1772. doi: 10.1016/S0002-9440(10)65492-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Borresen AL, et al. Somatic mutations in the hMSH2 gene in microsatellite unstable colorectal carcinomas. Hum. Mol. Genet. 1995;4:2065–2072. doi: 10.1093/hmg/4.11.2065. [DOI] [PubMed] [Google Scholar]
  • 10.Katabuchi H, et al. Mutations in DNA mismatch repair genes are not responsible for microsatellite instability in most sporadic endometrial carcinomas. Cancer Res. 1995;55:5556–5560. [PubMed] [Google Scholar]
  • 11.Chang JW, et al. Correlation of genetic instability with mismatch repair protein expression and p53 mutations in non-small cell lung cancer. Clin. Cancer Res. 2000;6:1639–1646. [PubMed] [Google Scholar]
  • 12.Xinarianos G, et al. hMLH1 and hMSH2 expression correlates with allelic imbalance on chromosome 3p in non-small cell lung carcinomas. Cancer Res. 2000;60:4216–4221. [PubMed] [Google Scholar]
  • 13.Herman JG, Graff JR, Myohanen S, Nelkin BD, Baylin SB. Methylation-specific PCR: a novel PCR assay for methylation status of CpG islands. Proc. Natl. Acad. Sci. U. S. A. 1996;93:9821–9826. doi: 10.1073/pnas.93.18.9821. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Herman JG, et al. Incidence and functional consequences of hMLH1 promoter hypermethylation in colorectal carcinoma. Proc. Natl. Acad. Sci. U. S. A. 1998;95:6870–6875. doi: 10.1073/pnas.95.12.6870. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Nuovo GJ, Plaia TW, Belinsky SA, Baylin SB, Herman JG. In situ detection of the hypermethylation-induced inactivation of the p16 gene as an early event in oncogenesis. Proc. Natl. Acad. Sci. U. S. A. 1999;96:12754–12759. doi: 10.1073/pnas.96.22.12754. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Maekawa M, et al. DNA methylation analysis using bisulfite treatment and PCR-single-strand conformation polymorphism in colorectal cancer showing microsatellite instability. Biochem. Biophys. Res. Commun. 1999;262:671–676. doi: 10.1006/bbrc.1999.1230. [DOI] [PubMed] [Google Scholar]
  • 17.Deng G, Chen A, Hong J, Chae HS, Kim YS. Methylation of CpG in a small region of the hMLH1 promoter invariably correlates with the absence of gene expression. Cancer Res. 1999;59:2029–2033. [PubMed] [Google Scholar]
  • 18.Ito E, et al. A core promoter and a frequent single-nucleotide polymorphism of the mismatch repair gene hMLH1. Biochem. Biophys. Res. Commun. 1999;256:488–494. doi: 10.1006/bbrc.1999.0368. [DOI] [PubMed] [Google Scholar]
  • 19.Belinsky SA, et al. Aberrant methylation of p16(INK4a) is an early event in lung cancer and a potential biomarker for early diagnosis. Proc. Natl. Acad. Sci. U. S. A. 1998;95:11891–11896. doi: 10.1073/pnas.95.20.11891. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Zochbauer-Muller S, et al. Aberrant promoter methylation of multiple genes in non-small cell lung cancers. Cancer Res. 2001;61:249–255. [PubMed] [Google Scholar]
  • 21.Palmisano WA, et al. Predicting lung cancer by detecting aberrant promoter methylation in sputum. Cancer Res. 2000;60:5954–5958. [PubMed] [Google Scholar]
  • 22.Mao L, et al. Microsatellite alterations as clonal markers for the detection of human cancer. Proc. Natl. Acad. Sci. U. S. A. 1994;91:9871–9875. doi: 10.1073/pnas.91.21.9871. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Miozzo M, et al. Microsatellite alterations in bronchial and sputum specimens of lung cancer patients. Cancer Res. 1996;56:2285–2288. [PubMed] [Google Scholar]
  • 24.Ahrendt SA, et al. Molecular detection of tumor cells in bronchoalveolar lavage fluid from patients with early stage lung cancer. J. Natl. Cancer Inst. 1999;91:332–339. doi: 10.1093/jnci/91.4.332. [DOI] [PubMed] [Google Scholar]
  • 25.Ger LP, Hsu WL, Chen KT, Chen CJ. Risk factors of lung cancer by histological category in Taiwan. Anticancer Res. 1993;13:1491–1500. [PubMed] [Google Scholar]
  • 26.Esteller M, Corn PG, Baylin SB, Herman JG. A gene hypermethylation profile of human cancer. Cancer Res. 2001;61:3225–3229. [PubMed] [Google Scholar]
  • 27.Aubry MC, et al. DNA mismatch repair genes hMLH1, hMSH2, and hMSH6 are not inactivated in bronchioloalveolar carcinomas of the lung. Cancer. 2001;92:2898–2901. doi: 10.1002/1097-0142(20011201)92:11<2898::aid-cncr10104>3.0.co;2-q. [DOI] [PubMed] [Google Scholar]
  • 28.Liu YY, Chen YM, Huang MH, Perng RP. Prognosis and recurrent patterns in bronchioloalveolar carcinoma. Chest. 2000;118:940–947. doi: 10.1378/chest.118.4.940. [DOI] [PubMed] [Google Scholar]
  • 29.Benachenhou N, Guiral S, Gorska-Flipot I, Labuda D, Sinnett D. High resolution deletion mapping reveals frequent allelic losses at the DNA mismatch repair loci hMLH1 and hMSH3 in non-small cell lung cancer. Int. J. Cancer. 1998;77:173–180. doi: 10.1002/(sici)1097-0215(19980717)77:2<173::aid-ijc1>3.0.co;2-n. [DOI] [PubMed] [Google Scholar]
  • 30.Zienolddiny S, Ryberg D, Arab MO, Skaug V, Haugen A. Loss of heterozygosity is related to p53 mutations and smoking in lung cancer. Br. J. Cancer. 2001;84:226–231. doi: 10.1054/bjoc.2000.1528. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Wistuba II, et al. High resolution chromosome 3p allelotyping of human lung cancer and preneoplastic/preinvasive bronchial epithelium reveals multiple, discontinuous sites of 3p allele loss and three regions of frequent breakpoints. Cancer Res. 2000;60:1949–1960. [PubMed] [Google Scholar]
  • 32.Hirao T, et al. Tobacco smoke-induced DNA damage and an early age of smoking initiation induce chromosome loss at 3p21 in lung cancer. Cancer Res. 2001;61:612–615. [PubMed] [Google Scholar]
  • 33.Roche J, et al. Distinct 3p21.3 deletions in lung cancer and identification of a new human semaphorin. Oncogene. 1996;12:1289–1297. [PubMed] [Google Scholar]
  • 34.Chen JT, Chen YC, Wang YC, Chen CY, Wang YC. Alterations of the p16INK4A gene in resected non-small cell lung tumors and exfoliated cells within sputum. Int. J. Cancer. 2002;98:724–731. doi: 10.1002/ijc.10262. [DOI] [PubMed] [Google Scholar]
  • 35.Belinsky SA, Nikula KJ, Baylin SB, Issa JP. Increased cytosine DNA-methyltransferase activity is target-cell-specific and an early event in lung cancer. Proc. Natl. Acad. Sci. U. S. A. 1996;93:4045–4050. doi: 10.1073/pnas.93.9.4045. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Kanai Y, Ushijima S, Kondo Y, Nakanishi Y, Hirohashi S. DNA methyltransferase expression and DNA methylation of CPG islands and peri-centromeric satellite regions in human colorectal and stomach cancers. Int. J. Cancer. 2001;91:205–212. doi: 10.1002/1097-0215(200002)9999:9999<::aid-ijc1040>3.0.co;2-2. [DOI] [PubMed] [Google Scholar]
  • 37.Mizuno S, et al. Expression of DNA methyltransferases DNMT1, 3A, and 3B in normal hematopoiesis and in acute and chronic myelogenous leukemia. Blood. 2001;97:1172–1179. doi: 10.1182/blood.v97.5.1172. [DOI] [PubMed] [Google Scholar]
  • 38.Graff JR, Herman JG, Myohanen S, Baylin SB, Vertino PM. Mapping patterns of CpG island methylation in normal and neoplastic cells implicates both upstream and downstream regions in de novo methylation. J. Biol. Chem. 1997;272:22322–22329. doi: 10.1074/jbc.272.35.22322. [DOI] [PubMed] [Google Scholar]

Articles from Journal of Clinical Investigation are provided here courtesy of American Society for Clinical Investigation

RESOURCES