Integrated genomic and epigenomic analyses pinpoint biallelic gene inactivation in tumors

Aberrant methylation of CpG islands and genomic deletion are two predominant mechanisms of gene inactivation in tumorigenesis, but the extent to which they interact is largely unknown. The lack of an integrated approach to study these mechanisms has limited the understanding of tumor genomes and cancer genes. Restriction landmark genomic scanning (RLGS; ref. 1) is useful for global analysis of aberrant methylation of CpG islands, but has not been amenable to alignment with deletion maps because the identity of most RLGS fragments is unknown. Here, we determined the nucleotide sequence and exact chromosomal position of RLGS fragments throughout the genome using the whole chromosome of origin of the fragments and in silico restriction digestion of the human genome sequence. To study the interaction of these gene-inactivation mechanisms in primary brain tumors, we integrated RLGS-based methylation analysis with high-resolution deletion maps from microarray-based comparative genomic hybridization (array CGH; ref. 3). Certain subsets of gene-associated CpG islands were preferentially affected by convergent methylation and deletion, including genes that exhibit tumor-suppressor activity, such as CISH1 (encoding SOCS1; ref. 4), as well as genes such as COE3 that have been missed by traditional non-integrated approaches. Our results show that most aberrant methylation events are focal and independent of deletions, and the rare convergence of these mechanisms can pinpoint biallelic gene inactivation without the use of positional cloning.     

Distribution, silencing potential and evolutionary impact of promoter DNA methylation in the human genome

To gain insight into the function of DNA methylation at cis-regulatory regions and its impact on gene expression, we measured methylation, RNA polymerase occupancy and histone modifications at 16,000 promoters in primary human somatic and germline cells. We find CpG-poor promoters hypermethylated in somatic cells, which does not preclude their activity. This methylation is present in male gametes and results in evolutionary loss of CpG dinucleotides, as measured by divergence between humans and primates. In contrast, strong CpG island promoters are mostly unmethylated, even when inactive. Weak CpG island promoters are distinct, as they are preferential targets for de novo methylation in somatic cells. Notably, most germline-specific genes are methylated in somatic cells, suggesting additional functional selection. These results show that promoter sequence and gene function are major predictors of promoter methylation states. Moreover, we observe that inactive unmethylated CpG island promoters show elevated levels of dimethylation of Lys4 of histone H3, suggesting that this chromatin mark may protect DNA from methylation.     

Epigenome analyses using BAC microarrays identify evolutionary conservation of tissue-specific methylation of SHANK3

CpG islands are present in one-half of all human and mouse genes and typically overlap with promoters or exons. We developed a method for high-resolution analysis of the methylation status of CpG islands genome-wide, using arrays of BAC clones and the methylation-sensitive restriction enzyme NotI. Here we demonstrate the accuracy and specificity of the method. By computationally mapping all NotI sites, methylation events can be defined with single-nucleotide precision throughout the genome. We also demonstrate the unique expandability of the array method using a different methylation-sensitive restriction enzyme, BssHII. We identified and validated new CpG island loci that are methylated in a tissue-specific manner in normal human tissues. The methylation status of the CpG islands is associated with gene expression for several genes, including SHANK3, which encodes a structural protein in neuronal postsynaptic densities. Defects in SHANK3 seem to underlie human 22q13 deletion syndrome. Furthermore, these patterns for SHANK3 are conserved in mice and rats.     

Gene silencing in cancer by histone H3 lysine 27 trimethylation independent of promoter DNA methylation

Epigenetic silencing in cancer cells is mediated by at least two distinct histone modifications, polycomb-based histone H3 lysine 27 trimethylation (H3K27triM) and H3K9 dimethylation. The relationship between DNA hypermethylation and these histone modifications is not completely understood. Using chromatin immunoprecipitation microarrays (ChIP-chip) in prostate cancer cells compared to normal prostate, we found that up to 5% of promoters (16% CpG islands and 84% non-CpG islands) were enriched with H3K27triM. These genes were silenced specifically in prostate cancer, and those CpG islands affected showed low levels of DNA methylation. Downregulation of the EZH2 histone methyltransferase restored expression of the H3K27triM target genes alone or in synergy with histone deacetylase inhibition, without affecting promoter DNA methylation, and with no effect on the expression of genes silenced by DNA hypermethylation. These data establish EZH2-mediated H3K27triM as a mechanism of tumor-suppressor gene silencing in cancer that is potentially independent of promoter DNA methylation.     

CpG island methylator phenotype underlies sporadic microsatellite instability and is tightly associated with BRAF mutation in colorectal cancer

Aberrant DNA methylation of CpG islands has been widely observed in human colorectal tumors and is associated with gene silencing when it occurs in promoter areas. A subset of colorectal tumors has an exceptionally high frequency of methylation of some CpG islands, leading to the suggestion of a distinct trait referred to as 'CpG island methylator phenotype', or 'CIMP'. However, the existence of CIMP has been challenged. To resolve this continuing controversy, we conducted a systematic, stepwise screen of 195 CpG island methylation markers using MethyLight technology, involving 295 primary human colorectal tumors and 16,785 separate quantitative analyses. We found that CIMP-positive (CIMP+) tumors convincingly represent a distinct subset, encompassing almost all cases of tumors with BRAF mutation (odds ratio = 203). Sporadic cases of mismatch repair deficiency occur almost exclusively as a consequence of CIMP-associated methylation of MLH1 . We propose a robust new marker panel to classify CIMP+ tumors.     

Epigenetic remodeling in colorectal cancer results in coordinate gene suppression across an entire chromosome band

We report a new mechanism in carcinogenesis involving coordinate long-range epigenetic gene silencing. Epigenetic silencing in cancer has always been envisaged as a local event silencing discrete genes. However, in this study of silencing in colorectal cancer, we found common repression of the entire 4-Mb band of chromosome 2q.14.2, associated with global methylation of histone H3 Lys9. DNA hypermethylation within the repressed genomic neighborhood was localized to three separate enriched CpG island 'suburbs', with the largest hypermethylated suburb spanning 1 Mb. These data change our understanding of epigenetic gene silencing in cancer cells: namely, epigenetic silencing can span large regions of the chromosome, and both DNA-methylated and neighboring unmethylated genes can be coordinately suppressed by global changes in histone modification. We propose that loss of gene expression can occur through long-range epigenetic silencing, with similar implications as loss of heterozygosity in cancer.     

Polycomb-mediated methylation on Lys27 of histone H3 pre-marks genes for de novo methylation in cancer

Many genes associated with CpG islands undergo de novo methylation in cancer. Studies have suggested that the pattern of this modification may be partially determined by an instructive mechanism that recognizes specifically marked regions of the genome. Using chromatin immunoprecipitation analysis, here we show that genes methylated in cancer cells are specifically packaged with nucleosomes containing histone H3 trimethylated on Lys27. This chromatin mark is established on these unmethylated CpG island genes early in development and then maintained in differentiated cell types by the presence of an EZH2-containing Polycomb complex. In cancer cells, as opposed to normal cells, the presence of this complex brings about the recruitment of DNA methyl transferases, leading to de novo methylation. These results suggest that tumor-specific targeting of de novo methylation is pre-programmed by an established epigenetic system that normally has a role in marking embryonic genes for repression.     

Computational identification of promoters and first exons in the human genome.

The identification of promoters and first exons has been one of the most difficult problems in gene-finding. We present a set of discriminant functions that can recognize structural and compositional features such as CpG islands, promoter regions and first splice-donor sites. We explain the implementation of the discriminant functions into a decision tree that constitutes a new program called FirstEF. By using different models to predict CpG-related and non-CpG-related first exons, we showed by cross-validation that the program could predict 86% of the first exons with 17% false positives. We also demonstrated the prediction accuracy of FirstEF at the genome level by applying it to the finished sequences of human chromosomes 21 and 22 as well as by comparing the predictions with the locations of the experimentally verified first exons. Finally, we present the analysis of the predicted first exons for all of the 24 chromosomes of the human genome.     

Methylation of CpG sites of two X-linked genes coincides with X-inactivation in the female mouse embryo but not in the germ line.

To further our understanding of initiation and imprinting of X-chromosome inactivation, we have examined methylation of specific CpG sites of X-linked Pgk-1 and G6pd genes throughout female mouse development. Methylation occurs around the time of inactivation and earlier for Pgk-1, which is closer to the X-inactivation centre. In female primordial germ cells, the inactive X chromosome escapes methylation; this may underly the reversibility of inactivation at meiosis. Similarly, the genes are unmethylated on the inactive X chromosome in sperm; hence, the imprint specifying preferential X-inactivation in extra-embryonic tissues must reside elsewhere.     

Increased methylation variation in epigenetic domains across cancer types

Tumor heterogeneity is a major barrier to effective cancer diagnosis and treatment. We recently identified cancer-specific differentially DNA-methylated regions (cDMRs) in colon cancer, which also distinguish normal tissue types from each other, suggesting that these cDMRs might be generalized across cancer types. Here we show stochastic methylation variation of the same cDMRs, distinguishing cancer from normal tissue, in colon, lung, breast, thyroid and Wilms' tumors, with intermediate variation in adenomas. Whole-genome bisulfite sequencing shows these variable cDMRs are related to loss of sharply delimited methylation boundaries at CpG islands. Furthermore, we find hypomethylation of discrete blocks encompassing half the genome, with extreme gene expression variability. Genes associated with the cDMRs and large blocks are involved in mitosis and matrix remodeling, respectively. We suggest a model for cancer involving loss of epigenetic stability of well-defined genomic domains that underlies increased methylation variability in cancer that may contribute to tumor heterogeneity.     

Identification of genetic elements that autonomously determine DNA methylation states

Cytosine methylation is a repressive, epigenetically propagated DNA modification. Although patterns of DNA methylation seem tightly regulated in mammals, it is unclear how these are specified and to what extent this process entails genetic or epigenetic regulation. To dissect the role of the underlying DNA sequence, we sequentially inserted over 50 different DNA elements into the same genomic locus in mouse stem cells. Promoter sequences of approximately 1,000 bp autonomously recapitulated correct DNA methylation in pluripotent cells. Moreover, they supported proper de novo methylation during differentiation. Truncation analysis revealed that this regulatory potential is contained within small methylation-determining regions (MDRs). MDRs can mediate both hypomethylation and de novo methylation in cis, and their activity depends on developmental state, motifs for DNA-binding factors and a critical CpG density. These results demonstrate that proximal sequence elements are both necessary and sufficient for regulating DNA methylation and reveal basic constraints of this regulation.     

Global hypomethylation of the genome in XX embryonic stem cells

Embryonic stem (ES) cells are important tools in the study of gene function and may also become important in cell therapy applications. Establishment of stable XX ES cell lines from mouse blastocysts is relatively problematic owing to frequent loss of one of the two X chromosomes. Here we show that DNA methylation is globally reduced in XX ES cell lines and that this is attributable to the presence of two active X chromosomes. Hypomethylation affects both repetitive and unique sequences, the latter including differentially methylated regions that regulate expression of parentally imprinted genes. Methylation of differentially methylated regions can be restored coincident with elimination of an X chromosome in early-passage parthenogenetic ES cells, suggesting that selection against loss of methylation may provide the basis for X-chromosome instability. Finally, we show that hypomethylation is associated with reduced levels of the de novo DNA methyltransferases Dnmt3a and Dnmt3b and that ectopic expression of these factors restores global methylation levels.