Mouse Oocyte Methylomes at Base Resolution Reveal Genome-Wide Accumulation of Non-CpG Methylation and Role of DNA Methyltransferases

DNA methylation is an epigenetic modification that plays a crucial role in normal mammalian development, retrotransposon silencing, and cellular reprogramming. Although methylation mainly occurs on the cytosine in a CG site, non-CG methylation is prevalent in pluripotent stem cells, brain, and oocytes. We previously identified non-CG methylation in several CG-rich regions in mouse germinal vesicle oocytes (GVOs), but the overall distribution of non-CG methylation and the enzymes responsible for this modification are unknown. Using amplification-free whole-genome bisulfite sequencing, which can be used with minute amounts of DNA, we constructed the base-resolution methylome maps of GVOs, non-growing oocytes (NGOs), and mutant GVOs lacking the DNA methyltransferase Dnmt1, Dnmt3a, Dnmt3b, or Dnmt3L. We found that nearly two-thirds of all methylcytosines occur in a non-CG context in GVOs. The distribution of non-CG methylation closely resembled that of CG methylation throughout the genome and showed clear enrichment in gene bodies. Compared to NGOs, GVOs were over four times more methylated at non-CG sites, indicating that non-CG methylation accumulates during oocyte growth. Lack of Dnmt3a or Dnmt3L resulted in a global reduction in both CG and non-CG methylation, showing that non-CG methylation depends on the Dnmt3a-Dnmt3L complex. Dnmt3b was dispensable. Of note, lack of Dnmt1 resulted in a slight decrease in CG methylation, suggesting that this maintenance enzyme plays a role in non-dividing oocytes. Dnmt1 may act on CG sites that remain hemimethylated in the de novo methylation process. Our results provide a basis for understanding the mechanisms and significance of non-CG methylation in mammalian oocytes.     

RNAi, DRD1, and histone methylation actively target developmentally important non-CG DNA methylation in arabidopsis

Cytosine DNA methylation protects eukaryotic genomes by silencing transposons and harmful DNAs, but also regulates gene expression during normal development. Loss of CG methylation in the Arabidopsis thaliana met1 and ddm1 mutants causes varied and stochastic developmental defects that are often inherited independently of the original met1 or ddm1 mutation. Loss of non-CG methylation in plants with combined mutations in the DRM and CMT3 genes also causes a suite of developmental defects. We show here that the pleiotropic developmental defects of drm1 drm2 cmt3 triple mutant plants are fully recessive, and unlike phenotypes caused by met1 and ddm1, are not inherited independently of the drm and cmt3 mutations. Developmental phenotypes are also reversed when drm1 drm2 cmt3 plants are transformed with DRM2 or CMT3, implying that non-CG DNA methylation is efficiently re-established by sequence-specific signals. We provide evidence that these signals include RNA silencing though the 24-nucleotide short interfering RNA (siRNA) pathway as well as histone H3K9 methylation, both of which converge on the putative chromatin-remodeling protein DRD1. These signals act in at least three partially intersecting pathways that control the locus-specific patterning of non-CG methylation by the DRM2 and CMT3 methyltransferases. Our results suggest that non-CG DNA methylation that is inherited via a network of persistent targeting signals has been co-opted to regulate developmentally important genes.     

Multiplex CRISPR-Cas9 editing of DNA methyltransferases in rice uncovers a class of non-CG methylation specific for GC-rich regions

DNA methylation in the non-CG context is widespread in the plant kingdom and abundant in mammalian tissues such as the brain and pluripotent cells. Non-CG methylation in Arabidopsis thaliana is coordinately regulated by DOMAINS REARRANGED METHYLTRANSFERASE (DRM) and CHROMOMETHYLASE (CMT) proteins but has yet to be systematically studied in major crops due to difficulties in obtaining genetic materials. Here, utilizing the highly efficient multiplex CRISPR-Cas9 genome-editing system, we created single- and multiple-knockout mutants for all the nine DNA methyltransferases in rice (Oryza sativa) and profiled their whole-genome methylation status at single-nucleotide resolution. Surprisingly, the simultaneous loss of DRM2, CHROMOMETHYLASE3 (CMT2), and CMT3 functions, which completely erases all non-CG methylation in Arabidopsis, only partially reduced it in rice. The regions that remained heavily methylated in non-CG contexts in the rice Os-dcc (Osdrm2/cmt2/cmt3a) triple mutant had high GC contents. Furthermore, the residual non-CG methylation in the Os-dcc mutant was eliminated in the Os-ddccc (Osdrm2/drm3/cmt2/cmt3a/cmt3b) quintuple mutant but retained in the Os-ddcc (Osdrm2/drm3/cmt2/cmt3a) quadruple mutant, demonstrating that OsCMT3b maintains non-CG methylation in the absence of other major methyltransferases. Our results showed that OsCMT3b is subfunctionalized to accommodate a distinct cluster of non-CG-methylated sites at highly GC-rich regions in the rice genome.

Examination of knockout mutants reveals that rice methyltransferases have subfunctionalized to accommodate a distinct cluster of non-CG-methylated sites at highly GC-rich regions in the rice genome.     

Locus-specific control of DNA methylation by the Arabidopsis SUVH5 histone methyltransferase

In Arabidopsis thaliana, heterochromatin formation is guided by double-stranded RNA (dsRNA), which triggers methylation of histone H3 at Lys-9 (H3 mK9) and CG plus non-CG methylation on identical DNA sequences. At heterochromatin targets including transposons and centromere repeats, H3 mK9 mediated by the Su(var)3-9 homologue 4 (SUVH4)/KYP histone methyltransferase (MTase) is required for the maintenance of non-CG methylation by the CMT3 DNA MTase. Here, we show that although SUVH4 is the major H3 K9 MTase, the SUVH5 protein also has histone MTase activity in vitro and contributes to the maintenance of H3 mK9 and CMT3-mediated non-CG methylation in vivo. Strikingly, the relative contributions of SUVH4, SUVH5, and a third related histone MTase, SUVH6, to non-CG methylation are locus-specific. For example, SUVH4 and SUVH5 together control transposon sequences with only a minor contribution from SUVH6, whereas SUVH4 and SUVH6 together control a transcribed inverted repeat source of dsRNA with only a minor contribution from SUVH5. This locus-specific variation suggests different mechanisms for recruiting or activating SUVH enzymes at different heterochromatic sequences. The suvh4 suvh5 suvh6 triple mutant loses both monomethyl and dimethyl H3 K9 at target loci. The suvh4 suvh5 suvh6 mutant also displays a loss of non-CG methylation similar to a cmt3 mutant, indicating that SUVH4, SUVH5, and SUVH6 together control CMT3 activity.     

Divergent DNA methylation contributes to duplicated gene evolution and chilling response in tea plants

The tea plant (Camellia sinensis) is a thermophilic cash crop and contains a highly duplicated and repeat-rich genome. It is still unclear how DNA methylation regulates the evolution of duplicated genes and chilling stress in tea plants. We therefore generated a single-base-resolution DNA methylation map of tea plants under chilling stress. We found that, compared with other plants, the tea plant genome is highly methylated in all three sequence contexts, including CG, CHG and CHH (where H = A, T, or C), which is further proven to be correlated with its repeat content and genome size. We show that DNA methylation in the gene body negatively regulates the gene expression of tea plants, whereas non-CG methylation in the flanking region enables a positive regulation of gene expression. We demonstrate that transposable element-mediated methylation dynamics significantly drives the expression divergence of duplicated genes in tea plants. The DNA methylation and expression divergence of duplicated genes in the tea plant increases with evolutionary age and selective pressure. Moreover, we detect thousands of differentially methylated genes, some of which are functionally associated with chilling stress. We also experimentally reveal that DNA methyltransferase genes of tea plants are significantly downregulated, whereas demethylase genes are upregulated at the initial stage of chilling stress, which is in line with the significant loss of DNA methylation of three well-known cold-responsive genes at their promoter and gene body regions. Overall, our findings underscore the importance of DNA methylation regulation and offer new insights into duplicated gene evolution and chilling tolerance in tea plants.     

MSI4/FVE is required for accumulation of 24-nt siRNAs and DNA methylation at a subset of target regions of RNA-directed DNA methylation

DNA methylation is an important epigenetic mark. In plants, de novo DNA methylation occurs mainly through the RNA-directed DNA methylation (RdDM) pathway. Researchers have previously inferred that a flowering regulator, MULTICOPY SUPPRESSOR OF IRA1 4 (MSI4)/FVE, is involved in non-CG methylation at several RdDM targets, suggesting a role of FVE in RdDM. However, whether and how FVE affects RdDM genome-wide is not known. Here, we report that FVE is required for DNA methylation at thousands of RdDM target regions. In addition, dysfunction of FVE significantly reduces 24-nucleotide siRNA accumulation that is dependent on factors downstream in the RdDM pathway. By using chromatin immunoprecipitation and sequencing (ChIP-seq), we show that FVE directly binds to FVE-dependent 24-nucleotide siRNA cluster regions. Our results also indicate that FVE may function in RdDM by physically interacting with RDM15, a downstream factor in the RdDM pathway. Our study has therefore revealed that FVE, by associating with RDM15, directly regulates DNA methylation and siRNA accumulation at a subset of RdDM targets.     

The SRA methyl-cytosine-binding domain links DNA and histone methylation

Epigenetic gene silencing suppresses transposon activity and is critical for normal development . Two common epigenetic gene-silencing marks are DNA methylation and histone H3 lysine 9 dimethylation (H3K9me2). In Arabidopsis thaliana, H3K9me2, catalyzed by the methyltransferase KRYPTONITE (KYP/SUVH4), is required for maintenance of DNA methylation outside of the standard CG sequence context. Additionally, loss of DNA methylation in the met1 mutant correlates with a loss of H3K9me2. Here we show that KYP-dependent H3K9me2 is found at non-CG methylation sites in addition to those rich in CG methylation. Furthermore, we show that the SRA domain of KYP binds directly to methylated DNA, and SRA domains with missense mutations found in loss-of-function kyp mutants have reduced binding to methylated DNA in vitro. These data suggest that DNA methylation is required for the recruitment or activity of KYP and suggest a self-reinforcing loop between histone and DNA methylation. Lastly, we found that SRA domains from two Arabidopsis SRA-RING proteins also bind methylated DNA and that the SRA domains from KYP and SRA-RING proteins prefer methylcytosines in different sequence contexts. Hence, unlike the methyl-binding domain (MBD), which binds only methylated-CpG sequences, the SRA domain is a versatile new methyl-DNA-binding motif.     

DNA methylation in mouse A-repeats in DNA methyltransferase-knockout ES cells and in normal cells determined by bisulfite genomic sequencing

Mouse ES cells with a null mutation of the known DNA methyltransferase retain some residual DNA methylation and can methylate foreign sequences de novo. We have used bisulfite genomic sequencing to examine the sequence specificity and distributions of methylation of a hypermethylated CG island sequence, mouse A-repeats. There were 13 CG dinucleotides in the region examined, 12 of which were methylated to variable extents in all DNAs. We found that: (1) there is considerable residual DNA methylation in ES cells lacking the known DNA methyltransferase (29% of normal methylation in the complete knockout ES DNA); (2) this other activity methylates at exactly the same CG sites as the major methyltransferase; and (3) differences in the distribution of methylated sites between A-repeats in these DNAs are consistent with this other activity methylating in a random de novo fashion. Also, the lack of any methylation in non-CG sites argues that, in other studies where non-CG methylation sites have been found by bisulfite sequencing, detection of such sites of non-CG methylation is not an inherent artifact in this methodology.     

Distribution of CWG and CCWGG in the Human Genome

Expression of the bacterial CG methyltransferase M?HhaI in mammalian cells appears to generate significant biological effects, while biological effects of the expression of the non-CG methyltransferase M?EcoRII in human cells have not been detected. The association of cytosine methylation with the CG site in mammals is also associated with clustering of CG sites near 5´ control regions (CG-islands) of human genes. Moreover spontaneous deamination of 5-methylcytosine at these sites is thought to lead to the well known deficiency of CG sites in genomes where endogenous CG methyltransferases are expressed. Since these associations are generally taken to imply a biological function for the CG dinucleotide that is associated with its selective methylation by endogenous DNA methylation systems, we have asked whether or not CWG or CCWGG sites are clustered in regions flanking human genes and whether or not an overall deficiency of CWG or CCWGG occurs in the human genome. Using build 36.1, of the human genome, we inspected the regions flanking the 28,501 well known gene loci in the human genome. Our analysis confirmed the expected clustering of CG sites near the 5´ region of known genes and open reading frames. In contrast to the CG site, neither the CWG site nor the CCWGG site recognized by the bacterial methyltransferase M?EcoRII were clustered in any particular region near known genes and open reading frames. Moreover, neither the CCWGG nor the CWG site was depleted in the human genome, again in sharp contrast to the known genomic deficiency of CpG sites. Our findings suggest that in contrast to CG site recognition, human cytosine methyltransferases recognize CWG and CCWGG only at very low frequency if at all.