CpG dinucleotides are extensively underrepresented in mammalian genomes. It is widely accepted that genome-wide CpG depletion is predominantly caused by an elevated CpG > TpG mutation rate due to frequent cytosine methylation in the CpG context. Meanwhile the CpG content in genomic regions called CpG islands (CGIs) is noticeably higher. This observation is usually explained by lower CpG > TpG substitution rates within CGIs due to reduced cytosine methylation levels.

Mammalian CpG islands (CGIs) normally escape DNA methylation in all adult tissues and developmental stages. However, in our previous study we unexpectedly identified many methylated CGIs in human peripheral blood leukocytes. Methylated CpG dinucleotides convert to TpG dinucleotides through deaminization of their cytosine bases more frequently than hypomethylated CpG dinucleotides. Therefore, we wondered how methylated CGIs in germline or non-germline cells maintain their CpG-rich sequences. It is known that events such as germline hypomethylation, CpG selection, biased gene conversion (BGC), and frequent CpG fixation can contribute to the maintenance of CpG-rich sequences in methylated CGIs in germline or non-germline cells. However, it has not been investigated which of the processes maintain CpG-rich sequences of methylated CGIs in each genomic position.

Epigenetic modifications and nucleosome positioning play an important role in modulating gene expression. However, how the patterns of epigenetic modifications and nucleosome positioning are established around promoters is not well understood. Here, we have addressed these questions in a series of genome-wide experiments coupled to a novel bioinformatic analysis approach. Our data reveal a clear correlation between CpG density, promoter activity and accumulation of active or repressive histone marks. CGI boundaries define the chromatin promoter regions that will be epigenetically modified. CpG-rich promoters are targeted by histone modifications and histone variants, while CpG-poor promoters are regulated by DNA methylation. CGIs boundaries, but not transcriptional activity, are essential determinants of H2A.Z positioning in vicinity of the promoters, suggesting that the presence of H2A.Z is not related to transcriptional control. Accordingly, H2A.Z depletion has no impact on gene expression of arrested mouse embryonic fibroblasts. Therefore, the underlying DNA sequence, the promoter CpG density and, to a lesser extent, transcriptional activity, are key factors implicated in promoter chromatin architecture.

DNA methylation is being increasingly recognized to play a role in regulation of hepatitis B virus (HBV) gene expression. The aim of this study was to compare the CpG island distribution among different HBV genotypes. We analyzed 176 full-length HBV genomic sequences obtained from the GenBank database, belonging to genotypes A through J, to identify the CpG islands in the HBV genomes. Our results showed that while 79 out of 176 sequences contained three conventional CpG islands (I-III) as previously described, 83 HBV sequences harbored only two of the three known islands. Novel CpG islands were identified in the remaining 14 HBV isolates and named as CpG island IV, V, and VI. Among the eight known HBV genotypes and two putative genotypes, while HBV genomes containing three CpG islands were predominant in genotypes A, B, D, E, and I; genotypes C, F, G, and H tended to contain only two CpG islands (II and III). In conclusion, the CpG islands, which are potential targets for DNA methylation mediated by the host functions, differ among HBV genotypes, and these genotype-specific differences in CpG island distribution could provide new insights into the understanding of epigenetic regulation of HBV gene expression and hepatitis B disease outcome.

Methylation occurs frequently at 5'-cytosine of the CpG dinucleotides in vertebrate genomes; however, this epigenetic feature is rarely observed in CpG islands (CGIs) or CpG clusters in the promoter regions of genes. Aberrant methylation of the promoter-associated CGIs might influence gene expression and cause carcinogenesis. Because of the functional importance, multiple algorithms have been available for identifying CGIs in a genome or a sequence. They can be categorized into the traditional algorithms (e.g., Gardiner-Garden and Frommer (1987), Takai and Jones (2002), and CpGPRoD (2002)) or statistical property based algorithms (CpGcluster (2006) and CG cluster (2007)). We reviewed the features of these algorithms and evaluated their performance on identifying functional CGIs using genome-wide methylation data. Moreover, identification of CGIs is an initial step in many recent studies for predicting methylation status as well as in the design of methylation detection platforms. We reviewed the benchmarks and features used in these studies.

The authors analyzed whether the promoter hypermethylation of cancer-related genes was involved in the tumorigenesis of malignant gliomas.

Somatic embryogenesis (SE) is a more rapid and controllable method for plant propagation than traditional breeding methods. However, it often suffers from limited efficiency. SERK1 promotes SE in several plants, including pineapple (Ananas comosus L.). We investigate the embryonic cell-specific transcriptional regulation of AcSERK1 by methylation analysis of CpG islands in AcSERK1 regulatory sequences. This revealed differences in the methylation status of CpG islands between embryonic callus and non-embryonic callus; the methylation inhibitor 5-azaC increased AcSERK1 expression and also accelerated SE. These findings indicate that the expression of AcSERK1 is regulated epigenetically. This study lays the foundation for further analysis of epigenetic regulatory mechanisms that may enhance the efficiency of SE in pineapple and other plants.

DNA methylation has been shown to play an important role in the silencing of tumor suppressor genes in various tumor types. In order to have a system-wide understanding of the methylation changes that occur in tumors, we have developed a differential methylation hybridization (DMH) protocol that can simultaneously assay the methylation status of all known CpG islands (CGIs) using microarray technologies. A large percentage of signals obtained from microarrays can be attributed to various measurable and unmeasurable confounding factors unrelated to the biological question at hand. In order to correct the bias due to noise, we first implemented a quantile regression model, with a quantile level equal to 75%, to identify hypermethylated CGIs in an earlier work. As a proof of concept, we applied this model to methylation microarray data generated from breast cancer cell lines. However, we were unsure whether 75% was the best quantile level for identifying hypermethylated CGIs. In this paper, we attempt to determine which quantile level should be used to identify hypermethylated CGIs and their associated genes.

In contrast to most genes in mammalian genomes, imprinted genes are monoallelically expressed depending on the parental origin of the alleles. Imprinted gene expression is regulated by distinct DNA elements that exhibit allele-specific epigenetic modifications, such as DNA methylation. These so-called differentially methylated regions frequently overlap with CpG islands. Thus, CpG islands of imprinted genes may contain special DNA elements that distinguish them from CpG islands of biallelically expressed genes. Here, we present a detailed study of CpG islands of imprinted genes in mouse and in human. Our study shows that imprinted genes more frequently contain tandem repeat arrays in their CpG islands than randomly selected genes in both species. In addition, mouse imprinted genes more frequently possess intragenic CpG islands that may serve as promoters of allele-specific antisense transcripts. This feature is much less pronounced in human, indicating an interspecies variability in the evolution of imprinting control elements.

There has been much interest in CpG islands (CGIs), clusters of CpG dinucleotides in GC-rich regions, because they are considered gene markers and involved in gene regulation. To date, there has been no genome-wide analysis of CGIs in the fish genome. We first evaluated the performance of three popular CGI identification algorithms in four fish genomes (tetraodon, stickleback, medaka, and zebrafish). Our results suggest that Takai and Jones' (2002) algorithm is most suitable for comparative analysis of CGIs in the fish genome. Then, we performed a systematic analysis of CGIs in the four fish genomes using Takai and Jones' algorithm, compared to other vertebrate genomes. We found that both the number of CGIs and the CGI density vary greatly among these genomes. Remarkably, each fish genome presents a distinct distribution of CGI density with some genomic factors (e.g., chromosome size and chromosome GC content). These findings are helpful for understanding evolution of fish genomes and the features of fish CGIs.

To develop a reliable method for whole genome analysis of DNA methylation.

In higher eukaryotes, up to 70% of genes have high levels of nonmethylated cytosine/guanine base pairs (CpGs) surrounding promoters and gene regulatory units. These features, called CpG islands, were identified over 20 years ago, but there remains little mechanistic evidence to suggest how these enigmatic elements contribute to promoter function, except that they are refractory to epigenetic silencing by DNA methylation. Here we show that CpG islands directly recruit the H3K36-specific lysine demethylase enzyme KDM2A. Nucleation of KDM2A at these elements results in removal of H3K36 methylation, creating CpG island chromatin that is uniquely depleted of this modification. KDM2A utilizes a zinc finger CxxC (ZF-CxxC) domain that preferentially recognizes nonmethylated CpG DNA, and binding is blocked when the CpG DNA is methylated, thus constraining KDM2A to nonmethylated CpG islands. These data expose a straightforward mechanism through which KDM2A delineates a unique architecture that differentiates CpG island chromatin from bulk chromatin.

Long interspersed elements (LINEs), through both self-mobilization and trans-mobilization of short interspersed elements and processed pseudogenes, have made an indelible impact on the structure and function of the human genome. One consequence is the creation of new CpG islands (CGIs). In fact, more than half of all CGIs in the genome are associated with repetitive DNA, three-quarters of which are derived from retrotransposons. However, little is known about the epigenetic impact of newly inserted CGIs. We utilized a transgenic LINE-1 mouse model and tracked DNA methylation dynamics of individual germline insertions during mouse development. The retrotransposed GFP marker sequence, a strong CGI, is hypomethylated in male germ cells but hypermethylated in somatic tissues, regardless of genomic location. The GFP marker is similarly methylated when delivered into the genome via the Sleeping Beauty DNA transposon, suggesting that the observed methylation pattern may be independent of the mode of insertion. Comparative analyses between insertion- and non-insertion-containing alleles further reveal a graded influence of the retrotransposed CGI on flanking CpG sites, a phenomenon that we described as "sloping shores." Computational analyses of human and mouse methylomic data at single-base resolution confirm that sloping shores are universal for hypomethylated CGIs in sperm and somatic tissues. Additionally, the slope of a hypomethylated CGI can be affected by closely positioned CGI neighbors. Finally, by tracing sloping shore dynamics through embryonic and germ cell reprogramming, we found evidence of bookmarking, a mechanism that likely determines which CGIs will be eventually hyper- or hypomethylated.

The mammalian genome is punctuated by CpG islands (CGIs), which differ sharply from the bulk genome by being rich in G + C and the dinucleotide CpG. CGIs often include transcription initiation sites and display 'active' histone marks, notably histone H3 lysine 4 methylation. In embryonic stem cells (ESCs) some CGIs adopt a 'bivalent' chromatin state bearing simultaneous 'active' and 'inactive' chromatin marks. To determine whether CGI chromatin is developmentally programmed at specific genes or is imposed by shared features of CGI DNA, we integrated artificial CGI-like DNA sequences into the ESC genome. We found that bivalency is the default chromatin structure for CpG-rich, G + C-rich DNA. A high CpG density alone is not sufficient for this effect, as A + T-rich sequence settings invariably provoke de novo DNA methylation leading to loss of CGI signature chromatin. We conclude that both CpG-richness and G + C-richness are required for induction of signature chromatin structures at CGIs.

DNA methylation at CpG islands (CGIs) is one of the most intensively studied epigenetic mechanisms. It is fundamental for cellular differentiation and control of transcriptional potential. DNA methylation is involved also in several processes that are central to evolutionary biology, including phenotypic plasticity and evolvability. In this study, we explored the relationship between CpG islands methylation and signatures of selective pressure in Homo Sapiens, using a computational biology approach. By analyzing methylation data of 25 cell lines from the Encyclopedia of DNA Elements (ENCODE) Consortium, we compared the DNA methylation of CpG islands in genomic regions under selective pressure with the methylation of CpG islands in the remaining part of the genome. To define genomic regions under selective pressure, we used three different methods, each oriented to provide distinct information about selective events. Independently of the method and of the cell type used, we found evidences of undermethylation of CGIs in human genomic regions under selective pressure. Additionally, by analyzing SNP frequency in CpG islands, we demonstrated that CpG islands in regions under selective pressure show lower genetic variation. Our findings suggest that the CpG islands in regions under selective pressure seem to be somehow more "protected" from methylation when compared with other regions of the genome.

CpG islands (CGIs) are primarily promoter-associated genomic regions and are mostly unmethylated within highly methylated mammalian genomes. The mechanisms by which CGIs are protected from de novo methylation remain elusive. Here we show that insertion of CpG-free DNA into targeted CGIs induces de novo methylation of the entire CGI in human pluripotent stem cells (PSCs). The methylation status is stably maintained even after CpG-free DNA removal, extensive passaging, and differentiation. By targeting the DNA mismatch repair gene MLH1 CGI, we could generate a PSC model of a cancer-related epimutation. Furthermore, we successfully corrected aberrant imprinting in induced PSCs derived from an Angelman syndrome patient. Our results provide insights into how CpG-free DNA induces de novo CGI methylation and broaden the application of targeted epigenome editing for a better understanding of human development and disease.

The analysis of tumours using biomarkers in blood is transforming cancer diagnosis and therapy. Cancers are characterised by evolving genetic alterations, making it difficult to develop reliable and broadly applicable DNA-based biomarkers for liquid biopsy. In contrast to the variability in gene mutations, the methylation pattern remains generally constant during carcinogenesis. Thus, methylation more than mutation analysis may be exploited to recognise tumour features in the blood of patients. In this work, we investigated the possibility of using global CpG (CpG means a CG motif in the context of methylation. The p represents the phosphate. This is used to distinguish CG sites meant for methylation from other CG motifs or from mentions of CG content) island methylation profiles as a basis for the prediction of cancer state of patients utilising liquid biopsy samples. We retrieved existing GEO methylation datasets on hepatocellular carcinoma (HCC) and cell-free DNA (cfDNA) from HCC patients and healthy donors, as well as healthy whole blood and purified peripheral blood mononuclear cell (PBMC) samples, and used a random forest classifier as a predictor. Additionally, we tested three different feature selection techniques in combination. When using cfDNA samples together with solid tumour samples and healthy blood samples of different origin, we could achieve an average accuracy of 0.98 in a 10-fold cross-validation. In this setting, all the feature selection methods we tested in this work showed promising results. We could also show that it is possible to use solid tumour samples and purified PBMCs as a training set and correctly predict a cfDNA sample as cancerous or healthy. In contrast to the complete set of samples, the feature selections led to varying results of the respective random forests. ANOVA feature selection worked well with this training set, and the selected features allowed the random forest to predict all cfDNA samples correctly. Feature selection based on mutual information could also lead to better than random results, but LASSO feature selection would not lead to a confident prediction. Our results show the relevance of CpG islands as tumour markers in blood.

The Polycomb repressive complex 2 (PRC2) mainly mediates transcriptional repression and has essential roles in various biological processes including the maintenance of cell identity and proper differentiation. Polycomb-like (PCL) proteins, such as PHF1, MTF2 and PHF19, are PRC2-associated factors that form sub-complexes with PRC2 core components, and have been proposed to modulate the enzymatic activity of PRC2 or the recruitment of PRC2 to specific genomic loci. Mammalian PRC2-binding sites are enriched in CG content, which correlates with CpG islands that display a low level of DNA methylation. However, the mechanism of PRC2 recruitment to CpG islands is not fully understood. Here we solve the crystal structures of the N-terminal domains of PHF1 and MTF2 with bound CpG-containing DNAs in the presence of H3K36me3-containing histone peptides. We show that the extended homologous regions of both proteins fold into a winged-helix structure, which specifically binds to the unmethylated CpG motif but in a completely different manner from the canonical winged-helix DNA recognition motif. We also show that the PCL extended homologous domains are required for efficient recruitment of PRC2 to CpG island-containing promoters in mouse embryonic stem cells. Our research provides the first, to our knowledge, direct evidence to demonstrate that PCL proteins are crucial for PRC2 recruitment to CpG islands, and further clarifies the roles of these proteins in transcriptional regulation in vivo.

CpG islands (CGIs) are commonly used as genomic markers to study the patterns and regulatory consequences of DNA methylation. Interestingly, recent studies reveal a substantial diversity among CGIs: long and short CGIs, for example, exhibit contrasting patterns of gene expression complexity and nucleosome occupancy. Evolutionary origins of CGIs are also highly heterogeneous. In order to systematically evaluate potential diversities among CGIs and ultimately to illuminate the link between diversity of CGIs and their epigenetic variation, we analyzed the nucleotide-resolution DNA methylation maps (methylomes) of multiple cellular origins. We discover novel 'clusters' of CGIs according to their patterns of DNA methylation; the stably hypomethylated CGI cluster (cluster I), sperm-hypomethylated CGI cluster (cluster II), and variably methylated CGI cluster (cluster III). These epigenomic CGI clusters are strikingly distinct at multiple biological features including genomic, evolutionary, and functional characteristics. At the genomic level, the stably hypomethylated CGI cluster tends to be longer and harbors many more CpG dinucleotides than those in other clusters. They are also frequently associated with promoters, while CGI clusters II and III mostly reside in intragenic or intergenic regions and exhibit highly tissue-specific DNA methylation. Functional ontology terms and transcriptional profiles co-vary with CGI clusters, indicating that the regulatory functions of CGIs are tightly linked to their heterogeneity. Finally, CGIs associated with distinctive biological processes, such as diseases, aging, and imprinting, occur disproportionately across CGI clusters. These new findings provide an effective means to combine existing knowledge on CGIs into a genomic context while bringing new insights that elucidate the significance of DNA methylation across different biological conditions and demography.

There are differences in the distribution and length of HBV CpG islands and the viral mutations contribute greatly to the development of HBV-related diseases. However, little is known regarding the effects of such difference and mutations in HBV genotypes B and C sequences on the regulation of HBV gene expression and their clinical outcomes. To study the distribution, length and genetic trait of CpG islands in normal and mutant sequences of HBV genotypes B and C, 320 HBV isolates from Chinese patients were retrieved from GenBank. Programs CLUSTALX 1.83 and MethPrimer were employed to perform multiple sequence alignments and to predict CpG islands, respectively. 72.0% genotype B isolates contained three conventional CpG islands, and 76.1% genotype C only contained CpG islands II and III. 14.6% genotype B and 7.5% genotype C contained three novel CpG islands. In genotype B, lengths of conventional CpG islands between normal and mutant isolates exhibited substantial variations, but in genotype C, those were relatively stable. CpG island II could be "truncated" or "split". "Truncated" region mutations were associated with structural and functional abnormalities of HBV genes. Rate of "split" CpG island II in genotype B was much higher than that in genotype C. In the majority of isolates from HCC and HBV-ACLF, genotype C lacked CpG island I and novel islands. Distribution, length and genetic trait of CpG islands in HBV genotypes B and C might affect their methylation status, and further affect regulation of HBV gene expression, leading to different clinical outcomes.