Genome-wide methylation and expression profiling identifies promoter characteristics affecting demethylation-induced gene up-regulation in melanoma

Abstract

Background: Abberant DNA methylation at CpG dinucleotides represents a common mechanism of transcriptional silencing in cancer. Since CpG methylation is a reversible event, tumor supressor genes that have undergone silencing through this mechanism represent promising targets for epigenetically active anti-cancer therapy. The cytosine analog 5-aza-2'-deoxycytidine (decitabine) induces genomic hypomethylation by inhibiting DNA methyltransferase, and is an example of an epigenetic agent that is thought to act by up-regulating silenced genes.

Methods: It is unclear why decitabine causes some silenced loci to re-express, while others remain inactive. By applying data-mining techniques to large-scale datasets, we attempted to elucidate the qualities of promoter regions that define susceptibility to the drug's action. Our experimental data, derived from melanoma cell strains, consist of genome-wide gene expression data before and after treatment with decitabine, as well as genome-wide data on un-treated promoter methylation status, and validation of specific genes by bisulfite sequencing.

Results: We show that the combination of promoter CpG content and methylation level informs the ability of decitabine treatment to up-regulate gene expression. Promoters with high methylation levels and intermediate CpG content appear most susceptible to up-regulation by decitabine, whereas few of those highly methylated promoters with high CpG content are up-regulated. For promoters with low methylation levels, those with high CpG content are more likely to be up-regulated, whereas those with low CpG content are underrepresented among up-regulated genes.

Conclusions: Clinically, elucidating the patterns of action of decitabine could aid in predicting the likelihood of up-regulating epigenetically silenced tumor suppressor genes and others from pathways involved with tumor biology. As a first step toward an eventual translational application, we build a classifier to predict gene up-regulation based on promoter methylation and CpG content, which achieves a performance of 0.77 AUC.

Figure 1

Decitabine effect as related to promoter methylation level and CpG content. Panel A. Genes are stratified by promoter CpG content (increasing from left to right) and methylation level (low methylation in top row, high methylation in bottom row). Plotting post-treatment expression level as a function of basal levels demonstrates the distribution of promoters in each bin (orange dots) in the context of all promoters (blue dots). Panel B. Summary of -log p-value from Wilcoxon signed-rank test assessing the likelihood that basal expression < post-treatment expression. The dotted line corresponds to -log10 of 0.05. The promoter category with the most significant trend toward up-regulated genes has highly methylated promoters with intermediate CpG content. As methylation level decreases, up-regulation is apparent in progressively more CpG dense promoters.

Figure 2

Defining decitabine-response by fold-change expression and delta expression. Panel A. Genes are stratified by promoter CpG content (increasing from left to right) and methylation level (low methylation in top row, high methylation in bottom row). Delta expression is plotted as a function of basal expression level. Upregulated genes are defined as those with >= 2-fold expression increase (left of diagonal line) and delta expression >= 5,000 (above horizontal line). Panel B. Percent of upregulated genes per CpG and methylation bin, with maximal up-regulation observed at high methylation and intermediate CpG content. Stars and circles indicate promoter categories with statistically significant enrichment or depletion in the number of up-regulated genes, respectively.

Figure 3

Promoter methylation and CpG content of decitabine-responsive genes in MCF-7 cell line. Of eighteen decitabine-induced genes from the MCF-7 breast cancer cell line, three are highly methylated, five have intermediate methylation levels, and ten have low levels of methylation. Highly methylated, decitabine-responsive genes have lower promoter CpG content. With increasing levels of methylation, CpG content of responsive genes increases. For extremely CpG-dense promoters, only those with low methylation levels are up-regulated. These observations are in accord with the melanoma data.

Figure 4

Predicting decitabine-responsive genes using logistic regression. Receiver operating curve for predicting decitabine response on the basis of promoter CpG content and methylation level. Models were trained and tested on the YUMAC cell line using ten-fold cross-validation. The plot shows the mean across ten runs, with 2*standard error bars. The area under the curve (AUC) is 77%, compared to 72% and 63%, respectively, for models built on promoter methylation status or CpG content alone.