A genome-wide screen for promoter methylation in lung cancer identifies novel methylation markers for multiple malignancies

Abstract

Background: Promoter hypermethylation coupled with loss of heterozygosity at the same locus results in loss of gene function in many tumor cells. The "rules" governing which genes are methylated during the pathogenesis of individual cancers, how specific methylation profiles are initially established, or what determines tumor type-specific methylation are unknown. However, DNA methylation markers that are highly specific and sensitive for common tumors would be useful for the early detection of cancer, and those required for the malignant phenotype would identify pathways important as therapeutic targets.

Methods and findings: In an effort to identify new cancer-specific methylation markers, we employed a high-throughput global expression profiling approach in lung cancer cells. We identified 132 genes that have 5' CpG islands, are induced from undetectable levels by 5-aza-2'-deoxycytidine in multiple non-small cell lung cancer cell lines, and are expressed in immortalized human bronchial epithelial cells. As expected, these genes were also expressed in normal lung, but often not in companion primary lung cancers. Methylation analysis of a subset (45/132) of these promoter regions in primary lung cancer (n = 20) and adjacent nonmalignant tissue (n = 20) showed that 31 genes had acquired methylation in the tumors, but did not show methylation in normal lung or peripheral blood cells. We studied the eight most frequently and specifically methylated genes from our lung cancer dataset in breast cancer (n = 37), colon cancer (n = 24), and prostate cancer (n = 24) along with counterpart nonmalignant tissues. We found that seven loci were frequently methylated in both breast and lung cancers, with four showing extensive methylation in all four epithelial tumors.

Conclusions: By using a systematic biological screen we identified multiple genes that are methylated with high penetrance in primary lung, breast, colon, and prostate cancers. The cross-tumor methylation pattern we observed for these novel markers suggests that we have identified a partial promoter hypermethylation signature for these common malignancies. These data suggest that while tumors in different tissues vary substantially with respect to gene expression, there may be commonalities in their promoter methylation profiles that represent targets for early detection screening or therapeutic intervention.

Figure 1. Strategy Used to Identify Methylation Candidates by Gene Expression Microarray

NSCLC and HBEC lines were treated with 5-aza and compared to controls (DMSO). We subtracted transcripts induced at least 4-fold in HBEC (n = 133) from the total number induced at least 4-fold in two of seven NSCLC lines (866 loci of 47,000 total), since methylation of these genes is unlikely to be tumor specific. For practical purposes we removed genes that were not expressed in HBEC (n = 460), were duplicate probes (n = 11), or had poor annotations (n = 66). Finally we excluded genes without identifiable 5′ CpG islands (n = 64). The number of genes subtracted from the total induced ≥4-fold in two of seven NSCLC cell lines (n = 866) is indicated next to each description in parentheses. We used the percentage of transcripts associated with 5′ CpG island as a measure of enrichment for the major steps in the filtering process. 37% of all RefSeq transcripts contain 5′ CpG islands; 55% of the 866 5-aza–induced transcripts had 5′ CpG islands; 73% of the final 196 genes had CpG islands. Statistical analysis of these data appears in Table 5.

Figure 2. 5-Aza Treatment Induces Genes Silenced by Promoter Methylation in HBEC and NSCLC Cell Lines

(A) QPCR for p16 and TKTL1 in HBEC and NSCLC. Solid bars are p16 and cross-hatch bars are TKTL1. Data are normalized, relative mRNA expression levels according to the 2^ΔΔCt method. HBEC2, 3, and 4 had similar profiles and were combined; data are averages and error bars are ± standard deviation. p16 status is indicated below each cell line; +, expressed; HD, homozygous deletion; M, methylated. (B) RT-PCR for p16 in the indicated cell lines. GAPDH is a loading control. (C) Methylation-specific PCR for TKTL1 in the indicated samples shows complete methylation in all samples examined, both methylated (M) and unmethylated (U). SssI in vitro-methylated DNA was used as a positive control for the methylated primers, and 5-aza-treated DNA was a positive control for the unmethylated primer sets (for PCR conditions and primer sequences see Protocol S1).

Figure 3. 5-Aza-Induced Gene Set in NSCLC, SCLC, Breast, and Colon Cancer Cell Lines

Heat map for gene induction across NSCLC and other cancer cell lines as indicated. Data are log₂ changes between mock-treated and 1 μM 5-aza treatment in each cell line. Bright red indicates 4-fold or greater up-regulation; intermediate red, at least 2-fold induction; grey, less than 2-fold induction; black, no data. The data are ordered from top to bottom according to the frequency of 4-fold induction across the NSCLC cell lines. The vertical colored bars parallel to the heat map represent the frequency of 4-fold induction in the NSCLC 5-aza induction experiments. Affymetrix probe IDs, GenBank accessions, UCSC cytoband alignment, and gene symbols are represented in order from top to bottom with the colored bars from the heat map indicating fold induction; purple indicates five of seven, blue indicates four of seven, green indicates three of seven, and yellow indicates two of seven. The figure layout was borrowed from [66].

Figure 4. Complete-Linkage Cluster Analysis of 5-Aza-Induced Methylation Candidates in NSCLC and HBECs

Analysis was performed on a panel of 31 NSCLC cell lines (U133A and B) and seven HBEC lines (U133 Plus 2.0) with an overlapping gene set (117 genes). Data are mean-centered log₂ expression values across the samples. Red indicates above the mean; green, below the mean. The 5-aza induction gene set separates cancer from HBEC lines and in most cases these genes are expressed at high levels in HBECs but not in NSCLC.

Figure 5. Complete-Linkage Cluster Analysis of 5-Aza-Induced Methylation Candidates in NSCLC and Normal Lung Tissues

Panel included 46 primary NSCLC samples and 29 counterpart normal lung tissues. Arrays are median-centered and genes are mean centered and colored as in Figure 4. Blue bar indicates normal lung; purple bar indicates tumor tissue. The 5-aza induction gene set clearly distinguishes cancer from normal. Most genes are expressed at higher levels in normal tissues, although not all.

Figure 6. Summary of Methylation-Specific PCR in Matched Primary NSCLC and Adjacent Nonmalignant Tissue

Data are color-coded and grouped as follows: red fill indicates positive methylated product; blue indicates positive unmethylated product. Data are grouped as follows: group I, no methylation in either HBECs or PBC DNA; group II, methylation in HBECs but not PBCs; group III, methylation in PBCs. Data are ordered from top to bottom according to the frequency of methylation in primary NSCLC. “GENE” indicates gene symbol; blue bars indicate loss of heterozygosity (LOH; allele loss) (a net [median] copy number change over 31 cell lines was calculated by taking the median signal over all cell lines); grey, no data; black, control primer sequences. Data are presented in the same order in the top (methylated) and bottom (unmethylated) panel.

Figure 7. Summary of Methylation-Specific PCR in HBEC and NSCLC Cell Lines

From the left, Locus Link ID (LL ID), gene symbol (GENE), in vitro methylated (SssI) DNA mixed with lymphocyte DNA, normal PBC DNA, HBEC lines, NSCLC cell lines, as indicated. Red fill indicates positive methylated product; blue indicates positive unmethylated product. Data are grouped as follows: group I, no methylation in either HBECs or PBC DNA; group II, methylation in HBECs but not PBCs; group III, methylation in PBCs. Data are ordered by the frequency of methylation in primary lung tumors (Figure 6).

Figure 8. Summary of Sodium Bisulfite Sequencing for Seven Genes in PBCs, HBECs, and NSCLC Cells

Between eight and 20 clones were sequenced for each locus in each cell type. Sequencing primers were designed to flank the MSP priming sites and do not include any CpG sites, with the exception of BNC1, which we were not able to amplify outside of the MSP priming sites for cells that harbored methylation. There was no amplification of the methylated primer set in HBECs or PBCs, and no amplification of the unmethylated primer set in the NSCLC cell lines examined. One some occasions the methylated primer set for BNC1 amplified a 289 bp amplicon from an unrelated locus on Chromosome 1. The sequence corresponds to a CpG island in an intronless gene (GPR25) that was heavily methylated in tumors. The unmethylated primer set did not amplify this sequence. Each box represents a composite of clones for that CpG site. Open boxes indicate 0%–25% methylation; light grey, 26%–50%; dark grey, 51%–75%; black, 76%–100% methylation. Raw data are available in Figures S8–S14. Primers and PCR conditions are available upon request.

Figure 9. MSP for Indicated Genes in Ductal Breast Carcinoma DNA for Samples Obtained from UNC

The basal phenotype is based on gene expression profiles demonstrated previously and is characterized by the absence of estrogen receptor and a poor prognosis. Other samples are characterized as luminal. Visible bands corresponding to the appropriate size were counted as positive. 100 bp ladder is at far left. M, methylated product; U, unmethylated product.

Figure 10. Histogram for Methylation Frequency of Indicated Genes in Prostate, Breast, Lung, and Colon Cancer and Companion Normal Tissue

MSP data for indicated genes in breast (n = 14; red bars), lung (n = 20; black bars), prostate (n = 24; pale yellow bars), and colon (n = 24; grey bars) tumors and benign tissue (see Methods). Only samples with matching benign and tumor tissue are represented in the histogram. Gels were run and scored as above. SOX15 was omitted from this figure for clarity. Data for RASSF1A were obtained from [17,24,45,46]; data for p16 were obtained from [17,43,44,46].