Methylation-dependent transition rates are dependent on local sequence lengths and genomic regions

Recently, Fryxell and Moon (2005) examined methylation-dependent transition rates (5mC deamination rates), which were calculated by the difference between the CpG transition and GpC transition rates, using 4,437 transition mutations in CpG or GpC dinucleotides. They concluded that 5mC deamination rates were highly dependent on local GC content but not on local sequence lengths over which GC content was calculated or the genomic regions where the mutations occurred. Here, we reexamined these statements by using 292,216 CpG-->TpG/CpA and GpC-->GpT/ApC mutations, an increase of 66 times as much data. Contrary to Fryxell and Moon's conclusions, our analysis indicated that 5mC deamination rates in the human genome were dependent on both the local sequence length and the genomic region. Some explanations for their conclusions were provided.     

Asymmetrical distribution of CpG in an 'average' mammalian gene.

The frequency and distribution of the rare dinucleotide CpG was examined in 15 mammalian genes. CpG is highly methylated at cytosine in mammalian DNA (1,2) and 5-methylcytosine (5mC) is thought to undergo a transition mutation via deamination to produce thymine (3). This would result in the accumulation of TpG and CpA and depletion of CpG during evolution (4). Consistent with this hypothesis, the gene sample of 26,541 dinucleotides contained CpG at 40% the frequency expected by base composition and the CpG transition products, TpG+CpA, were significantly elevated at 124% of expected random frequency. However, because CpG occurs at only 25% of expected random frequency in the genome, the sampled genes were considerably enriched in this dinucleotide. CpGs were asymmetrically distributed in sequences flanking the genes. 5'-flanking sequences were enriched in CpG at 135% of the frequency expected assuming a symmetrical distribution of all the CpGs in the sampled genes (p less than 0.01), while 3'-flanking regions were depleted in CpG at 40% of expected values (p less than 0.0001). This asymmetry may reflect the role of 5-methylcytosine in gene expression. In contrast the frequencies of GpC and GpT+ ApC did not differ significantly from that predicted by base composition and these dinucleotides were not asymmetrically distributed.     

Cytosine deamination plays a primary role in the evolution of mammalian isochores

DNA melting is rate-limiting for cytosine deamination, from which we infer that the rate of cytosine deamination should decline twofold for each 10% increase in GC content. Analysis of human DNA sequence data confirms that this is the case for 5-methylcytosine. Several lines of evidence further confirm that it is also the case for unmethylated cytosine and that cytosine deamination causes the majority of all C-->T and G-->A transitions in mammals. Thus, cytosine deamination and DNA base composition each affect the other, forming a positive feedback loop that facilitates divergent genetic drift to high or low GC content. Because a 10 degrees C increase in temperature in vitro increases the rate of cytosine deamination 5. 7-fold, cytosine deamination must be highly dependent on body temperature, which is consistent with the dramatic differences between the isochores of warm-blooded versus cold-blooded vertebrates. Because this process involves both DNA melting and positive feedback, it would be expected to spread progressively (in evolutionary time) down the length of the chromosome, which is consistent with the large size of isochores in modern mammals.     

Recognition and potential mechanisms for replication and erasure of cytosine hydroxymethylation

Cytosine residues in mammalian DNA occur in at least three forms, cytosine (C), 5-methylcytosine (M; 5mC) and 5-hydroxymethylcytosine (H; 5hmC). During semi-conservative DNA replication, hemi-methylated (M/C) and hemi-hydroxymethylated (H/C) CpG dinucleotides are transiently generated, where only the parental strand is modified and the daughter strand contains native cytosine. Here, we explore the role of DNA methyltransferases (DNMT) and ten eleven translocation (Tet) proteins in perpetuating these states after replication, and the molecular basis of their recognition by methyl-CpG-binding domain (MBD) proteins. Using recombinant proteins and modified double-stranded deoxyoligonucleotides, we show that DNMT1 prefers a hemi-methylated (M/C) substrate (by a factor of >60) over hemi-hydroxymethylated (H/C) and unmodified (C/C) sites, whereas both DNMT3A and DNMT3B have approximately equal activity on all three substrates (C/C, M/C and H/C). Binding of MBD proteins to methylated DNA inhibited Tet1 activity, suggesting that MBD binding may also play a role in regulating the levels of 5hmC. All five MBD proteins generally have reduced binding affinity for 5hmC relative to 5mC in the fully modified context (H/M versus M/M), though their relative abilities to distinguish the two varied considerably. We further show that the deamination product of 5hmC could be excised by thymine DNA glycosylase and MBD4 glycosylases regardless of context.     

Alu element mutation spectra: molecular clocks and the effect of DNA methylation

In primate genomes more than 40% of CpG islands are found within repetitive elements. With more than one million copies in the human genome, the Alu family of retrotransposons represents the most successful short interspersed element (SINE) in primates and CpG dinucleotides make up about 20% of Alu sequences. It is generally thought that CpG dinucleotides mutate approximately ten times faster than other dinucleotides due to cytosine methylation and the subsequent deamination and conversion of C-->T. However, the disparity of Alu subfamily age estimations based upon CpG or non-CpG substitution density indicates a more complex relationship between CpG and non-CpG substitutions within the Alu elements. Here we report an analysis of the mutation patterns for 5296 Alu elements comprising 20 subfamilies. Our results indicate a relatively constant CpG versus non-CpG substitution ratio of approximately 6 for the young (AluY) and intermediate (AluS) Alu subfamilies. However, a more complex non-linear relationship between CpG and non-CpG substitutions was observed when old (AluJ) subfamilies were included in the analysis. These patterns may be the result of the slowdown of the neutral mutation rate during primate evolution and/or an increase in the CpG mutation rate as the consequence of increased DNA methylation in response to a burst of retrotransposition activity approximately 35 million years ago.     

Role of base excision repair in maintaining the genetic and epigenetic integrity of CpG sites

Cytosine methylation at CpG dinucleotides is a central component of epigenetic regulation in vertebrates, and the base excision repair (BER) pathway is important for maintaining both the genetic stability and the methylation status of CpG sites. This perspective focuses on two enzymes that are of particular importance for the genetic and epigenetic integrity of CpG sites, methyl binding domain 4 (MBD4) and thymine DNA glycosylase (TDG). We discuss their capacity for countering C to T mutations at CpG sites, by initiating base excision repair of G·T mismatches generated by deamination of 5-methylcytosine (5mC). We also consider their role in active DNA demethylation, including pathways that are initiated by oxidation and/or deamination of 5mC.     

The distribution of 5-methylcytosine in the nuclear genome of plants.

We have determined the 5-methylcytosine (5mC) content in high molecular weight DNA, from two dicot (tobacco and pea) and two monocot (wheat and maize) plant species, fractionated according to base composition. The results show that the proportion of 5mC in the genomic fractions increases linearly with their guanine + cytosine (G + C) content while the proportion of non-methylated cytosine remains almost constant. This can be interpreted as a consequence of a difference in mutation pressure related to spontaneous deamination of 5mC to thymine between the different compartments of plant genomes.     

Cytosine methylation and the fate of CpG dinucleotides in vertebrate genomes

Summary The dinucleotide CpG is a hotspot for mutation in the human genome as a result of (1) the modification of the 5 cytosine by cellular DNA methyltransferases and (2) the consequent high frequency of spontaneous deamination of 5-methyl cytosine (5mC) to thymidine. DNA methylation thus contributes significantly, albeit indirectly, to the incidence of human genetic disease. We have attempted to estimate for the first time the in vivo rate of deamination of 5mC from the measured rate of 5mC deamination in vitro and the known error frequency of the cellular G/T mismatch-repair system. The accuracy and utility of this estimate (m _d ) was then assessed by comparison with clinical data, and an improved estimate of m _d (1.66x10^-16 s^-1) was derived. Comparison of the CpG mutation rates exibited by globin gene and pseudogene sequences from human, chimpanzee and macaque provided further estimates of m _d , all of which were consistent with the first. Use of this value in a mathematical model then permitted the estimation of the length of time required to produce the level of CpG suppression currently found in the bulk DNA of vertebrate genomes. This time span, approximately 450 million years, corresponds closely to the estimated time since the emergence and adaptive radiation of the vertebrates and thus coincides with the probable advent of heavily methylated genomes. An accurate estimate of the 5mC deamination rate is important not only for clinical medicine but also for studies of gene evolution. Our data suggest both that patterns of vertebrate gene methylation may be comparatively stable over relatively long periods of evolutionary time, and that the rate of CpG deamination can, under certain limited conditions, serve as a molecular clock.     

Comparative Genomic Study Reveals a Transition from TA Richness in Invertebrates to GC Richness in Vertebrates at CpG Flanking Sites： An Indication for Context-Dependent Mutagenicity of Methylated CpG Sites

Vertebrate genomes are characterized with CpG deficiency, particularly for GCpoor regions. The GC content-related CpG deficiency is probably caused by context-dependent deamination of methylated CpG sites. This hypothesis was examined in this study by comparing nucleotide frequencies at CpG flanking positions among invertebrate and vertebrate genomes. The finding is a transition of nucleotide preference of 5＇ T to 5＇ A at the invertebrate-vertebrate boundary, indicating that a large number of CpG sites with 5＇ Ts were depleted because of global DNA methylation developed in vertebrates. At genome level, we investigated CpG observed/expected （obs/exp） values in 500 bp fragments, and found that higher CpG obs/exp value is shown in GC-poor regions of invertebrate genomes （except sea urchin） but in GC-rich sequences of vertebrate genomes. We next compared GC content at CpG flanking positions with genomic average, showing that the GC content is lower than the average in invertebrate genomes, but higher than that in vertebrate genomes. These results indicate that although 5＇ T and 5＇ A are different in inducing deamination of methylated CpG sites, GC content is even more important in affecting the deamination rate. In all the tests, the results of sea urchin are similar to vertebrates perhaps due to its fractional DNA methylation. CpG deficiency is therefore suggested to be mainly a result of high mutation rates of methylated CpG sites in GC-poor regions.     

Hydroxyl-radical-induced oxidation of 5-methylcytosine in isolated and cellular DNA

The methylation and oxidative demethylation of cytosine in CpG dinucleotides plays a critical role in the regulation of genes during cell differentiation, embryogenesis and carcinogenesis. Despite its low abundance, 5-methylcytosine (5mC) is a hotspot for mutations in mammalian cells. Here, we measured five oxidation products of 5mC together with the analogous products of cytosine and thymine in DNA exposed to ionizing radiation in oxygenated aqueous solution. The products can be divided into those that arise from hydroxyl radical (•OH) addition at the 5,6-double bond of 5mC (glycol, hydantoin and imidazolidine products) and those that arise from H-atom abstraction from the methyl group of 5mC including 5-hydroxymethylcytosine (5hmC) and 5-formylcytosine (5fC). Based on the analysis of these products, we show that the total damage at 5mC is about 2-fold greater than that at C in identical sequences. The formation of hydantoin products of 5mC is favored, compared to analogous reactions of thymine and cytosine, which favor the formation of glycol products. The distribution of oxidation products is sequence dependent in specific ODN duplexes. In the case of 5mC, the formation of 5hmC and 5fC represents about half of the total of •OH-induced oxidation products of 5mC. Several products of thymine, cytosine, 5mC, as well as 8-oxo-7,8-dihydroguanine (8oxoG), were also estimated in irradiated cells.     

Genome-scale relationships between cytosine methylation and dinucleotide abundances in animals

In mammalian genomes CpGs occur at one-fifth their expected frequency. This is accepted as resulting from cytosine methylation and deamination of 5-methylcytosine leading to TpG and CpA dinucleotides. The corollary that a CpG deficit should correlate with TpG excess has not hitherto been systematically tested at a genomic level. I analyzed genome sequences (human, chimpanzee, mouse, pufferfish, zebrafish, sea squirt, fruitfly, mosquito, and nematode) to do this and generally to assess the hypothesis that CpG deficit, TpG excess, and other data are accountable in terms of 5-methylcytosine mutation. In all methylated genomes local CpG deficit decreases with higher G + C content. Local TpG surplus, while positively associated with G + C level in mammalian genomes but negatively associated with G + C in nonmammalian methylated genomes, is always explicable in terms of the CpG trend under the methylation model. Covariance of dinucleotide abundances with G + C demonstrates that correlation analyses should control for G + C. Doing this reveals a strong negative correlation between local CpG and TpG abundances in methylated genomes, in accord with the methylation hypothesis. CpG deficit also correlates with CpT excess in mammals, which may reflect enhanced cytosine mutation in the context 5'-YCG-3'. Analyses with repeat-masked sequences show that the results are not attributable to repetitive elements.     

TET-TDG Active DNA Demethylation at CpG and Non-CpG Sites

In mammalian genomes, cytosine methylation occurs predominantly at CG (or CpG) dinucleotide contexts. As part of dynamic epigenetic regulation, 5-methylcytosine (mC) can be erased by active DNA demethylation, whereby ten-eleven translocation (TET) enzymes catalyze the stepwise oxidation of mC to 5-hydroxymethylcytosine (hmC), 5-formylcytosine (fC), and 5-carboxycytosine (caC), thymine DNA glycosylase (TDG) excises fC or caC, and base excision repair yields unmodified cytosine. In certain cell types, mC is also enriched at some non-CG (or CH) dinucleotides, however hmC is not. To provide biochemical context for the distribution of modified cytosines observed in biological systems, we systematically analyzed the activity of human TET2 and TDG for substrates in CG and CH contexts. We find that while TET2 oxidizes mC more efficiently in CG versus CH sites, this context preference can be diminished for hmC oxidation. Remarkably, TDG excision of fC and caC is only modestly dependent on CG context, contrasting its strong context dependence for thymine excision. We show that collaborative TET-TDG oxidation-excision activity is only marginally reduced for CA versus CG contexts. Our findings demonstrate that the TET-TDG-mediated demethylation pathway is not limited to CG sites and suggest a rationale for the depletion of hmCH in genomes rich in mCH.     

Sequence context analysis of 8.2 million single nucleotide polymorphisms in the human genome

We analyzed n-mers (n=3-8) in the local environment of 8,249,446 human SNPs and compared their distribution with that in the genome reference sequences. The results revealed that the short sequences, which contained at least one CpG dinucleotide, occurred more frequently in the local SNP sequences than in the genome sequences. To exclude the hypermutability effect of the methylated CpG dinucleotides on the sequence context of SNPs, we examined the distribution patterns for each of the six categories of substitution. We observed the similar pattern (i.e., CpG-containing n-mers vs. non-CpG-containing n-mers) in SNP categories A/G, C/T and C/G but the opposite pattern in category A/T. We next identified 34,928 putative CpG islands in the human genome and located 133,591 SNPs within these islands. In the CpG islands, CpG SNPs were 3.92-fold less prevalent relative to the presence of CpG dinucleotides. Conversely, in the human genome, the frequency of CpG dinucleotides at the polymorphic sites was 6.09 times that in the genome reference sequences. These results support the previous views of mutational suppression at the CpG sites in the CpG islands and hypermutability of the methylated CpG dinucleotides that are prevalent in the non-CpG island sequences in the human genome. Our study represents a comprehensive investigation of the sequence context of SNPs in the human genome and in human CpG islands.     

Sequence context analysis in the mouse genome: single nucleotide polymorphisms and CpG island sequences

A genome-wide view of sequence mutability in mice is still limited, although biologists usually assume the same scenario for mice as for humans. In this study, we examined the sequence context in the local environment of 482,528 mouse single nucleotide polymorphisms (SNPs). We found that CpG-containing short sequences, in general, had more representation in the local sequences of SNPs compared to the genome sequences. The extent of this overrepresentation was stronger in mice than in humans, which is inconsistent with previous observations of the weaker neighboring-nucleotide biases on mouse SNPs. To exclude the CpG effect, we compared the distribution patterns of short sequences among the six categories of SNPs. The results revealed an even stronger pattern in the CpG-containing group for C/G substitution compared to for A/G or C/T substitutions. We next performed the first genome-wide sequence context analysis of SNPs in the mouse CpG islands. SNPs occurring at CpG sites were 3.14-fold less prevalent than expected, suggesting the suppression of methylation-dependent deamination in the CpG islands. The extent of this suppression was less in mice than in humans. Finally, compared with humans, the observations of a greater deficit of CpG dinucleotides, a stronger overrepresentation of CpG-containing n-mers surrounding the polymorphic sites, and a higher SNP/genome ratio of CpG dinucleotides in the mouse genome support the "loss of CpG islands" model in the mouse lineage.     

Detection of heavy methylation in human repetitive DNA subsets by a monoclonal antibody against 5-methylcytosine

A monoclonal antibody (IgM) against 5-methylcytosine (mC) was isolated and characterized. It showed a high specificity for mC with a cross-reactivity of less than 1% with cytosine and 0.1% with thymidine. An improved immunohybridization method, originally developed with polyclonal antibodies (Sano et al. (1980) Proc. Natl. Acad. Sci. USA 77, 3581), was applied to detect mC in immobilized DNA using the new monoclonal preparation. Human genomic DNA was cleaved with the restriction enzyme EcoRI and successively fractionated by malachite-green affinity chromatography and agarose gel electrophoresis. The fractionated DNA was transferred to nitrocellulose paper and treated with the anti-mC monoclonal antibody. Heavy methylation was observed in EcoRI-ladders of repetitive sequences of 1360, 1750, 2200 and 3400 bp, while 340, 660 and 2700 bp fragments were less methylated. The results show that methylation occurs in limited subsets of satellite II and III repetitive DNAs that contain high amounts of methylatable CpG dinucleotides, or CpG clusters.     

The distribution of the dinucleotide CpG and cytosine methylation in the vitellogenin gene family

Summary Sequence data from regions of five vertebrate vitellogenin genes were used to examine the frequency, distribution, and mutability of the dinucleotide CpG, the preferred modification site for eukaryotic DNA methyltransferases. The observed level of the CpG dinucleotide in all five genes was markedly lower than that expected from the known mononucleotide frequencies. CpG suppression was greater in introns than in exons. CpG-containing codons were found to be avoided in the vitellogenin genes, but not completely despite the redundancy of the genetic code. Frequency and distribution patterns of this dinucleotide varied dramatically among these otherwise closely related genes. Dense clusters of CpG dinucleotides tended to appear in regions of either functional or structural interest (e.g., in the transposon-like Vi-element ofXenopus) and these clusters contained 5-methylcytosine (5 mC). 5 mC is known to undergo deamination to form thymidine, but the extent to which this transition occurs in the heavily methylated genomes of vertebrates and its contribution to CpG suppression are still unclear. Sequence comparison of the methylated vitellogenin gene regions identified CT and GA substitutions that were found to occur at relatively high frequencies. The predicted products of CpG deamination, TpG and CpA, were elevated. These findings are consistent with the view that CpG distribution and methylation are interdependent and that deamination of 5 mC plays an important role in promoting evolutionary change at the nucleotide sequence level.     

Activation-induced cytidine deaminase deaminates 5-methylcytosine in DNA and is expressed in pluripotent tissues: implications for epigenetic reprogramming

DNA deaminases of the Aid/Apobec family convert cytosine into uracil and play key roles in acquired and innate immunity. The epigenetic modification by methylation of cytosine in CpG dinucleotides is also mutagenic, but this is thought to occur by spontaneous deamination. Here we show that Aid and Apobec1 are 5-methylcytosine deaminases resulting in a thymine base opposite a guanine. Their action can thus lead to C --> T transition mutations in methylated DNA, or in conjunction with repair of the T:G mismatch, to demethylation. The Aid and Apobec1 genes are located in a cluster of pluripotency genes including Nanog and Stella and are co-expressed with these genes in oocytes, embryonic germ cells, and embryonic stem cells. These results suggest that Aid and perhaps some of its family members may have roles in epigenetic reprogramming and cell plasticity. Transition in CpG dinucleotides is the most frequent mutation in human genetic diseases, and sequence context analysis of CpG transitions in the APC tumor suppressor gene suggests that DNA deaminases may play a significant role in tumor etiology.     

Active mammalian replication origins are associated with a high-density cluster of mCpG dinucleotides.

ori-beta is a well-characterized origin of bidirectional replication (OBR) located approximately 17 kb downstream of the dihydrofolate reductase gene in hamster cell chromosomes. The approximately 2-kb region of ori-beta that exhibits greatest replication initiation activity also contains 12 potential methylation sites in the form of CpG dinucleotides. To ascertain whether DNA methylation might play a role at mammalian replication origins, the methylation status of these sites was examined with bisulfite to chemically distinguish cytosine (C) from 5-methylcytosine (mC). All of the CpGs were methylated, and nine of them were located within 356 bp flanking the minimal OBR, creating a high-density cluster of mCpGs that was approximately 10 times greater than average for human DNA. However, the previously reported densely methylated island in which all cytosines were methylated regardless of their dinucleotide composition was not detected and appeared to be an experimental artifact. A second OBR, located at the 5' end of the RPS14 gene, exhibited a strikingly similar methylation pattern, and the organization of CpG dinucleotides at other mammalian origins revealed the potential for high-density CpG methylation. Moreover, analysis of bromodeoxyuridine-labeled nascent DNA confirmed that active replication origins were methylated. These results suggest that a high-density cluster of mCpG dinucleotides may play a role in either the establishment or the regulation of mammalian replication origins.     

Frequency of abnormal human haemoglobins caused by C----T transitions in CpG dinucleotides

A large part of human genetic disease apparently arises from deamination of cytosine residues in methylated CpG dinucleotides. Their mutation rate is known to be high when C is present as 5-methyl-cytosine, but is believed to be normal when it is unmethylated. The beta-globin gene contains five, the gamma-globin gene two, and each of the alpha-globin genes contains 35 CpG dinucleotides. The CpG dinucleotides in the beta and gamma-globin genes are methylated, while those in the alpha-globin genes are under-methylated. One would therefore have expected the CpG dinucleotides to be a frequent source of mutations in the beta and gamma-globin genes, but not in the alpha-globin genes. In fact, the evidence points to CpG dinucleotides being a frequent source of mutations in both the alpha and beta-globin genes. This suggests either that the mutation rates of both methylated and unmethylated CpG dinucleotides are abnormally high, which conflicts with published evidence, or that there is a finite chance of some of these in the alpha-globin genes of certain individuals being methylated and therefore subject to mutation.