Abstract

5-Hydroxymethylcytosine (5-hmC) may represent a new epigenetic modification of cytosine. While the dynamics of 5-hmC during neurodevelopment have recently been reported, little is known about its genomic distribution and function(s) in neurodegenerative diseases such as Huntington's disease (HD). We here observed a marked reduction of the 5-hmC signal in YAC128 (yeast artificial chromosome transgene with 128 CAG repeats) HD mouse brain tissues when compared with age-matched wild-type (WT) mice, suggesting a deficiency of 5-hmC reconstruction in HD brains during postnatal development. Genome-wide distribution analysis of 5-hmC further confirmed the diminishment of the 5-hmC signal in striatum and cortex in YAC128 HD mice. General genomic features of 5-hmC are highly conserved, not being affected by either disease or brain regions. Intriguingly, we have identified disease-specific (YAC128 versus WT) differentially hydroxymethylated regions (DhMRs), and found that acquisition of DhmRs in gene body is a positive epigenetic regulator for gene expression. Ingenuity pathway analysis (IPA) of genotype-specific DhMR-annotated genes revealed that alternation of a number of canonical pathways involving neuronal development/differentiation (Wnt/β-catenin/Sox pathway, axonal guidance signaling pathway) and neuronal function/survival (glutamate receptor/calcium/CREB, GABA receptor signaling, dopamine-DARPP32 feedback pathway, etc.) could be important for the onset of HD. Our results indicate that loss of the 5-hmC marker is a novel epigenetic feature in HD, and that this aberrant epigenetic regulation may impair the neurogenesis, neuronal function and survival in HD brain. Our study also opens a new avenue for HD treatment; re-establishing the native 5-hmC landscape may have the potential to slow/halt the progression of HD.

INTRODUCTION

DNA methylation constitutes a critical epigenetic modification that plays pivotal roles in chromatin structure remolding, gene silencing, embryonic development, differentiation and maintenance of cellular identity (1–5). In mammals, methylations at the fifth position of cytosine (5 mC) are catalyzed by de novo methyltransferases (DNMT3a, DNMT3b) and subsequently maintained by DNMT1 (6–9). In mammalian central nervous system (CNS), 5 mC plays a significant role in the temporal control of neural differentiation and neurodevelopment (10,11), and alternation in the 5 mC profile of genomic DNA has been recently linked to some neurological disorders such as hereditary sensory and autonomic neuropathy type-1 (12), autosomal-dominant cerebellar ataxia deafness and narcolepsy (ADCA-DN) (13), Rett syndrome-like phenotypes (14), HD (15) and Alzheimer's disease (16).

Recent studies demonstrate that 5 mC can be converted to 5-hydroxymethylcytosine (5-hmC) by the ten-eleven translocation (TET) family proteins through Fe (II) and alpha-ketoglutarate-dependent hydroxylation (17,18). 5-hmC is preferentially enriched in several cell types, including self-renewing and pluripotent stem cells (17,19) and certain neuronal cells (20). Although 5-hmC in mouse embryonic stem (ES) cells exists at an extremely high level, it decreases significantly during ES cell differentiation (17,21), and then reconstructs during the developmental process resulting in an elevated 5-hmC level in terminally differentiated cells, such as Purkinje neurons (20). The fact that 5-hmC is roughly 10-fold more enriched in neurons than in some peripheral tissues or ES cells (20,22–24) suggests that 5-hmC is a stable epigenetic modification which may provide an epigenetic signature specific to a particular neuronal function or dysfunction in the brain. Consistent with this notion, recent studies with genome-wide mapping of 5-hmC in the cerebellum and hippocampus revealed the enrichment of 5-hmC in developmentally activated genes or cell type-active genes (24–26). These studies point to a critical role of TET-mediated 5-hmC modification in CNS developmental processes and hint at the possibility that deviation of 5-hmC levels from these signature patterns may associate with diseases.

Huntington's disease (HD) is an autosomal-dominant inherited, fatal neurodegenerative disorder characterized by chorea, gradual but inexorable cognitive decline and psychiatric disturbances (27,28). HD is caused by the expanded polyglutamine (polyQ) stretch in the amino-terminus of huntingtin protein (HTT), a 350 kDa ubiquitously expressed protein of unknown function (27,29,30). Despite the extensive studies regarding the gain-of-toxic function of polyQ expansion in HTTexp (31–43), the exact mechanism(s) underlying the cause of neurodegeneration in HD remain obscure. Given that 5-hmC may function as a new epigenetic marker in the CNS, we sought to determine whether HTTexp might derange this new epigenetic modification.

A yeast artificial chromosome transgenic mouse model of HD (YAC128, with 128Q-stretch) is a well-characterized HD mouse model that harbors a full-length mutant huntingtin, which can faithfully recapitulate many of the behavioral and neuropathological features of the human condition (38,44–48). Since the striatum is the most vulnerable region relative to other regions such as cortex in HD, in this study, both striatum and cortex regions from WT and HD mice were analyzed, and this pairwise comparisons between each age, genotype and tissue could provide novel insights into the HD-specific links between 5-hmC modification and HD pathogenesis.

By employing a selective chemical labeling method for 5-hmC reported recently (49), we have generated the first genome-wide maps of 5-hmC in striatum and cortex from both WT and YAC128 HD mice brains. Herein, we provide evidence showing that loss of the 5-hmC marker is a new epigenetic feature of HD, and that this aberrant epigenomic regulation of 5-hmC may impair neurogenesis, neuronal survival and functions in HD brain at a very early stage, contributing to the striatal and cortical neuronal loss in advanced HD. Of potentially therapeutic significance, our study also opens a new possible avenue for HD treatment by showing it may be possible to target the cellular/biochemical pathways, and thereby potentially reestablishing the 5-hmC levels and landscape in HD brains.

RESULTS

Marked reduction of 5-hmC in whole genome of YAC128 HD mouse brain

Basing on the fact that 5-hmC is highly enriched in terminally differentiated neurons (20,22–24) and the importance of 5-hmC in developmental processes (24,25,26), we sought to determine whether HTTexp might derange this new epigenetic modification. YAC128 mouse is a HD mouse model that can faithfully recapitulate the pathological progression of the disease in humans. The main characteristic of the disease is neuronal loss in the striatum region of the brain. However, neurons that are absent from the cortex region also occurs in advanced HD patients. Dot-blots of 5-hmC from striatum and cortex DNA samples of 6-week-, 3-month- and 8-month-old mice (n= 5) revealed that the levels of 5-hmC increased substantially with age in both striatum and cortex, with the highest 5-hmC levels peaking at 3 months of age (Fig. 1A and F). Surprisingly, YAC128 HD striatum and cortex samples exhibited markedly lower content of 5-hmC when compared with age-matched wild-type (WT) brain tissues (Fig. 1A and F). To test whether the change of 5-hmC levels is specific for striatum and cortex, DNA samples of hippocampus, cerebellum and peripheral tissues—tails from WT and YAC128 HD mice were also compared. The reduction of 5-hmC levels in YAC128 mouse brains could also be detected in hippocampus but not in cerebellum and peripheral tissues such as tails (Supplementary Material, Fig. S1), indicating that the decrease of 5-hmC levels in YAC128 HD is brain regional specific and neuronal tissue specific. Immunofluorescence (IF) experiments further confirmed a significant reduction of 5-hmC in both striatum and cortex of HD brains (Fig. 1B–E), suggesting that a significant reduction of 5-hmC is a genomic feature in HD brains. Consistent with a recent study of the 5-hmC distribution in cerebellum and hippocampus (24), we also observed the perfect colocalization of the 5-hmC signal with NeuN, a marker of mature neurons, in both striatum and cortex, indicating that 5-hmC build up may occur primarily in neurons in mammalian brains.

Figure 1.

Quantification of the 5-hmC content in WT and YAC128 HD mouse brain tissues. (A) Dot-blot assay of 5-hmC in WT/YAC128 mouse striatum and cortex genome. 50, 20 and 10 ng striatal and cortical DNA samples from 6-week-, 3-month- and 8-month-old WT and YAC128 mice were dot-blotted using a 5-hmC-specific antibody. 5-hmC-containing DNA was used as the positive control, and normal C, 5-mC, 5-carC-containing DNA were used as the negative controls to verify the specificity of 5-hmC antibody. (BE) IF of 5-hmC (red) in striatum and cortex regions from 6-week-old WT (B) and YAC128 (C), 4-month-old WT (D) and YAC128 (E) mouse brains. Brain sections were counterstained with NeuN antibody and Hochest33342 to visualize neurons (green) and nuclei (blue). (F) Quantification of the striatal and cortical 5-hmC content in WT and YAC128 DNA samples extracted at 6 weeks, 3 and 8 months of age. Striatal and cortical 5-hmC signals were normalized to 5-hmC density of 6-week-old YAC128 striatum and cortex, respectively (n = 5, mean ± s.e.; *P < 0.05; **P < 0.01).

Figure 1.

Quantification of the 5-hmC content in WT and YAC128 HD mouse brain tissues. (A) Dot-blot assay of 5-hmC in WT/YAC128 mouse striatum and cortex genome. 50, 20 and 10 ng striatal and cortical DNA samples from 6-week-, 3-month- and 8-month-old WT and YAC128 mice were dot-blotted using a 5-hmC-specific antibody. 5-hmC-containing DNA was used as the positive control, and normal C, 5-mC, 5-carC-containing DNA were used as the negative controls to verify the specificity of 5-hmC antibody. (BE) IF of 5-hmC (red) in striatum and cortex regions from 6-week-old WT (B) and YAC128 (C), 4-month-old WT (D) and YAC128 (E) mouse brains. Brain sections were counterstained with NeuN antibody and Hochest33342 to visualize neurons (green) and nuclei (blue). (F) Quantification of the striatal and cortical 5-hmC content in WT and YAC128 DNA samples extracted at 6 weeks, 3 and 8 months of age. Striatal and cortical 5-hmC signals were normalized to 5-hmC density of 6-week-old YAC128 striatum and cortex, respectively (n = 5, mean ± s.e.; *P < 0.05; **P < 0.01).

To gain a genome-wide view of 5-hmC distribution, we next employed a 5-hmC-specific chemical labeling and enrichment technology coupled with high-throughput deep sequencing of 5-hmC-containing DNA fragments (49). Striatal and cortical DNA samples from the brains of 3-month-old WT/YAC128 mice were prepared and deeply sequenced to determine the effect of HTTexp on the 5-hmC epigenome at the early stage of HD. About 41.3, 40.8 42.6 and 34.1 million reads were obtained from WT striatum, YAC128 striatum, WT cortex and the YAC128 cortex DNA samples, respectively. When the reads from these four samples were mapped to the mouse reference genome (mm9), reads mapped to multiple locations were filtered out, leaving 20.4, 16.6, 22.9 and 18.4 million uniquely mapped reads that were used for subsequent analysis (Table 1). Clusters of read (peaks) were then identified by using MACS software (50). Overall 7.97 × 104, 3.68 × 104, 6.99 × 104 and 4.23 × 104 peaks were detected in the four samples, respectively (Table 1, Supplementary Material, Dataset S1–4).

Table 1.

Summary of sequencing data

Samples Total reads Uni-map reads Peak numbers Gene numbers 
WT striatum 41321310 20449138 79774 12215 
WT cortex 42684513 22975167 69976 11683 
Yac128 striatum 40863760 16658192 36826 8862 
Yac128 cortex 34137033 18413874 42361 9672 
Samples Total reads Uni-map reads Peak numbers Gene numbers 
WT striatum 41321310 20449138 79774 12215 
WT cortex 42684513 22975167 69976 11683 
Yac128 striatum 40863760 16658192 36826 8862 
Yac128 cortex 34137033 18413874 42361 9672 

A read is a putative 5-hmC-containing DNA fragment that has been sequenced. Uni-map reads are the reads that map to a unique locus in the genome. A peak is a cluster of uni-map reads identified using the MACS software. Genes that contain peaks can also be identified.

Genome-scaled distribution and epigenomic characteristics of 5-hmC in YAC128 brains

The genome-scaled density of 5-hmC reads of each sample was determined using the binned data and visualized in an IGV Browser (IGV 1.4.2, http://www.broadinstitute.org/igv/). As shown in Figure 2A, the global 5-hmC density of WT striatum and cortex tissues was much higher than those in YAC128 counterparts, which was consistent with the dot-blot results, the 5-hmC IF results and the peak counts. The 5-hmC distribution pattern in different chromosomes was unremarkable among four samples and was consistent with previous results from cerebellum and hippocampus (24) (Fig. 2B). Clustering analysis clearly showed distinctive patterns of cortex- and striatum-specific 5-hmC enrichment (Fig. 2C). The same brain regions with different genotypes (WT or YAC128) exhibit similar cluster distribution patterns, though the genotype-specific cluster distribution pattern could still be noted in both WT and YAC128 samples (Fig. 2C).

Figure 2.

Genomic mapping and features of 5-hmC across two brain regions in WT/YAC128 mice. (A) Genome-scale of 5-hmC profiles and enrichment in mouse striatum and cortex. The heatmap represents the read densities which have been equally scaled and then normalized based on the total number of mapped reads per sample. (B) Chromosomal read densities of 5-hmC in cortex and striatum from WT/YAC128 mice brains. The chromosome-wide densities were determined as reads per chromosome divided by the total number of reads in millions. Expected values were determined by dividing 106 by the total genome length and multiplying by the chromosome length. (C) Heatmap of 5-hmC levels of RefSeq genes. Five distinct clusters of genes were identified based on the dynamic changes of their 5-hmC levels in WT/YAC128 striatum and cortex. The heatmap represents the normalized 5-hmC tag density of each gene. (D) Normalized densities of 5-hmC reads (per kilobase per million) in regions with various genomic features (obtained from the University of California Santa Cruz tables, mm9) including promoters (−2 kb to +0.5 kb relative to TSS), exons (5′ UTR, 3′ UTR, coding exons), introns and intergenic regions. (E) Normalized peak densities of DhMR (differential 5-hmC region) in regions with various genomic features, including promoters (−2 kb to +0.5 kb relative to TSS), exons (5′ UTR, 3′ UTR, coding exons), introns and intergenic regions. The dynamic peaks were identified between WT and YAC128 with each brain region (cortex or striatum). (F) Normalized 5-hmC read densities across the transcript unit of all reference genes in striatum and cortex from both WT and YAC128 brains. Each protein-coding gene body was normalized to 0–100%. 5-hmC densities are shown along the transcript unit with both 5′ and 3′ flanking regions. 5-hmC densities were normalized to the total number of aligned reads from each sample (in millions). (G) Normalized densities of 5-hmC reads on repetitive sequences of striatum and cortex samples from WT/YAC128 mice brains. For each type of repeat (annotated by RepeatMasker), relative enrichment of 5-hmC between different samples at different repeat classes was assessed by determining the fraction of total reads aligned to each class of repetitive element.

Figure 2.

Genomic mapping and features of 5-hmC across two brain regions in WT/YAC128 mice. (A) Genome-scale of 5-hmC profiles and enrichment in mouse striatum and cortex. The heatmap represents the read densities which have been equally scaled and then normalized based on the total number of mapped reads per sample. (B) Chromosomal read densities of 5-hmC in cortex and striatum from WT/YAC128 mice brains. The chromosome-wide densities were determined as reads per chromosome divided by the total number of reads in millions. Expected values were determined by dividing 106 by the total genome length and multiplying by the chromosome length. (C) Heatmap of 5-hmC levels of RefSeq genes. Five distinct clusters of genes were identified based on the dynamic changes of their 5-hmC levels in WT/YAC128 striatum and cortex. The heatmap represents the normalized 5-hmC tag density of each gene. (D) Normalized densities of 5-hmC reads (per kilobase per million) in regions with various genomic features (obtained from the University of California Santa Cruz tables, mm9) including promoters (−2 kb to +0.5 kb relative to TSS), exons (5′ UTR, 3′ UTR, coding exons), introns and intergenic regions. (E) Normalized peak densities of DhMR (differential 5-hmC region) in regions with various genomic features, including promoters (−2 kb to +0.5 kb relative to TSS), exons (5′ UTR, 3′ UTR, coding exons), introns and intergenic regions. The dynamic peaks were identified between WT and YAC128 with each brain region (cortex or striatum). (F) Normalized 5-hmC read densities across the transcript unit of all reference genes in striatum and cortex from both WT and YAC128 brains. Each protein-coding gene body was normalized to 0–100%. 5-hmC densities are shown along the transcript unit with both 5′ and 3′ flanking regions. 5-hmC densities were normalized to the total number of aligned reads from each sample (in millions). (G) Normalized densities of 5-hmC reads on repetitive sequences of striatum and cortex samples from WT/YAC128 mice brains. For each type of repeat (annotated by RepeatMasker), relative enrichment of 5-hmC between different samples at different repeat classes was assessed by determining the fraction of total reads aligned to each class of repetitive element.

Our genome-wide mapping results showed that 5-hmC was enriched in gene bodies (including exons, introns, 5′ and 3′ UTRs), and depleted in intergenic regions (Fig. 2D–F). In addition, 5-hmC distributed in repetitive elements similarly among the four samples, with 80% of the repetitive element-mapped reads distributing in LINEs, SINEs and LTRs (Fig. 2G). We also examined the levels of 5-hmC in CpG islands of the genome in comparison with their flanking sequences. 5-hmC was enriched in intragenic CpG islands but depleted in promoter and intergenic CpG islands (Fig. 3A–D). Therefore, the distribution patterns of 5-hmC in different categories of genomic regions are highly conserved and not affected by genotypes (WT and YAC128) and neuronal tissues (striatum and cortex).

Figure 3.

Characterization of 5-hmC around CGIs and identification of disease-related DhMRs. (AD) Normalized 5-hmC read densities across the CGIs belonging to the promoter (Red), intragenic (Green) and intergenic (Blue) regions of each sample. Each CpG island was normalized to 0–100%. 5-hmC tag density was plotted from 500% upstream to 500% downstream of the normalized CGIs. (EH) Disease-related (i.e., genotype-specific) DhMRs. DhMRs identified in bidirectional comparisons of 5-hmC peaks between WT striatum and YAC128 striatum were assigned to genotype-specific (i.e., disease-related) striatum-DhMRs (WT versus YAC128, 698 DhMRs; YAC128 versus WT, 49 DhMRs); Disease-related cortex-DhMRs were identified similarly (WT versus YAC128, 324 DhMRs; YAC128 versus WT, 38 DhMRs).

Figure 3.

Characterization of 5-hmC around CGIs and identification of disease-related DhMRs. (AD) Normalized 5-hmC read densities across the CGIs belonging to the promoter (Red), intragenic (Green) and intergenic (Blue) regions of each sample. Each CpG island was normalized to 0–100%. 5-hmC tag density was plotted from 500% upstream to 500% downstream of the normalized CGIs. (EH) Disease-related (i.e., genotype-specific) DhMRs. DhMRs identified in bidirectional comparisons of 5-hmC peaks between WT striatum and YAC128 striatum were assigned to genotype-specific (i.e., disease-related) striatum-DhMRs (WT versus YAC128, 698 DhMRs; YAC128 versus WT, 49 DhMRs); Disease-related cortex-DhMRs were identified similarly (WT versus YAC128, 324 DhMRs; YAC128 versus WT, 38 DhMRs).

Identification of disease-related differential 5-hmC regions (DhMRs) in YAC128 brain tissues

Recent studies reported that DhMRs varies between different mouse brain development stages, different brain tissues (cerebellum versus hippocampus) (24) and even different neuronal cell types (26), suggesting that genomic 5-hmC enriched loci may serve as a novel epigenetic modification involved in gene regulation. Therefore, we asked whether expression of HTTexp in brains may perturb the DhMR construction. Pairwise comparisons of 5-hmC peaks between WT and YAC128 samples from either striatum or cortex allowed us to identify the disease-related (i.e., genotype-specific) DhMRs at the genome-wide level (Table 2). YAC128 mice at 3 months of age do not show any motor and pathological defects, therefore, identification of HTTexp-gained/lost DhMRs of 3-month-old mice can provide new insights into the links between 5-hmC modification and HD onset. We identified 747 DhMRs in striatum of which 49 were up-regulated and 698 were down-regulated in YAC128 striatum comparing with WT striatum (Fig. 3E and F) (Table 2, Supplementary Material, Dataset S5), localized to 30 and 406 genes, respectively. Meanwhile, 362 DhMRs were identified from cortex samples, 38 were up-regulated and 324 were down-regulated in YAC128 cortex relative to WT cortex (Fig. 3G and H) (Table 2, Supplementary Material, Dataset S5), corresponding to 28 and 171 genes, respectively. Thus, HTTexp-induced perturbance of 5-hmC epigenome potentially influences the expression of 436 striatal genes and 199 cortical genes in HD at a very early stage.

Table 2.

DhMRs identified between samples of pairs

Control Treatment
 
WT striatum WT cortex Yac128 striatum Yac128 cortex 
WT striatum 1182 49 ND 
WT cortex 361 ND 38 
Yac128 striatum 698 ND 149 
Yac128 cortex ND 324 425 
Control Treatment
 
WT striatum WT cortex Yac128 striatum Yac128 cortex 
WT striatum 1182 49 ND 
WT cortex 361 ND 38 
Yac128 striatum 698 ND 149 
Yac128 cortex ND 324 425 

DhMRs of a pair of samples were identified using the MACS software with one sample as the control and the other as the treatment. The DhMRs identified in such a way represent 5-hmC regions whose 5-hmC contents were higher in the treatment sample. Another set of DhMRs was acquired if the control and treatment samples were switched. Eight pairwise comparisons between striatum and cortex in WT and YAC128 were conducted. The comparisons of certain pairs of samples were not defined and were labeled as ND.

Correlation of disease-related DhMRs with gene transcription

Whether 5-hmC modification positively or negatively regulate gene expression remains elusive. We therefore investigated the correlation of 5-hmC-enrichment in genes with the gene mRNA levels. Nine genes were selected based on their 5-hmC contents in different genotypes of different tissues, Creb5, Dnm2 and Gnaq manifested significantly higher 5-hmC peaks in WT striatum than in YAC128. Grm3, Gabbr2, Trpc7 had higher 5-hmC peaks in WT cortex than in YAC128. Ryr3, Grid2 and Capn2 possessed significantly higher 5-hmC peaks in YAC128 cortex than in WT. Our Q-PCR results revealed that, except in Grm3 gene, all other DhMRs-containing genes (such as striatal Creb5, Dnm2, Gnaq and cortical Gabbr2 and Trpc7 in WT brains, and cortical RyR3, Grid2, Capn in YAC128 brains) exhibited a positive correlation with their mRNA levels in the tissues of two different genotypes (Fig. 4A–F; Supplementary Material, Fig. S2). The genomic view of 5-hmC-enriched regions in Creb5, Trpc7 and Ryr3 indicated that DhMRs are located mainly in gene bodies (Fig. 4B, D and F, Supplementary Material, Fig. S2).

Figure 4.

Association of disease-related DhMRs with gene transcription. (A) The Creb5 gene showed a higher transcription level in WT striatum than in YAC128 striatum (n = 4, P < 0.05). (B) Genomic view one of the DhMRs lost in YAC128 striatum comparing with WT striatum located in Creb5 gene. (C) The Trpc7 gene showed a higher transcription level in WT cortex than in YAC128 cortex (n = 4, P < 0.05). (D) Genomic view one of the DhMRs lost in YAC128 cortex comparing with WT cortex located in Trpc7 gene. (E) The Ryr3 gene showed a higher transcription level in YAC128 cortex than in WT cortex (n = 4, P < 0.05). (F) Genomic view one of the DhMRs acquired in YAC128 cortex comparing with WT cortex located in Ryr3 gene. Data are represented as mean ± s.e. (*P < 0.05 comparing with the gene transcription of WT tissue). WTC, WT cortex; YAC128C, YAC128 cortex; WTS, WT striatum; YAC128S, YAC128 striatum.

Figure 4.

Association of disease-related DhMRs with gene transcription. (A) The Creb5 gene showed a higher transcription level in WT striatum than in YAC128 striatum (n = 4, P < 0.05). (B) Genomic view one of the DhMRs lost in YAC128 striatum comparing with WT striatum located in Creb5 gene. (C) The Trpc7 gene showed a higher transcription level in WT cortex than in YAC128 cortex (n = 4, P < 0.05). (D) Genomic view one of the DhMRs lost in YAC128 cortex comparing with WT cortex located in Trpc7 gene. (E) The Ryr3 gene showed a higher transcription level in YAC128 cortex than in WT cortex (n = 4, P < 0.05). (F) Genomic view one of the DhMRs acquired in YAC128 cortex comparing with WT cortex located in Ryr3 gene. Data are represented as mean ± s.e. (*P < 0.05 comparing with the gene transcription of WT tissue). WTC, WT cortex; YAC128C, YAC128 cortex; WTS, WT striatum; YAC128S, YAC128 striatum.

Implications of aberrant 5-hmC epigenome on HD pathogenesis

Gene ontology analysis of the 436 genes with disease-related DhMRs in striatum and 199 genes in cortex allowed us to identify the canonical pathways and biological functions which could be important for HD onset. The ingenuity pathway analysis (IPA) of striatum-DhMRs-containing genes cataloged a number of canonical pathways such as Wnt/β-catenin signaling, axonal guidance signaling, GABA receptor signaling, glutamate receptor signaling, Dopamine-DAPRR32 feedback in cAMP signaling, mTOR signaling and synaptic long-term potentiation (Supplementary Material, Fig. S4A and Fig. 5A and B; Supplementary Material, Dataset S7). A few of these pathways such as glutamate receptor signaling, CREB signaling in neurons also appeared in IPA-cataloged pathways of cortex-DhMRs-genes (Supplementary Material, Fig. S4C; Supplementary Material, Dataset S7), suggesting that striatum and cortex may share some common pathological mechanisms in HD. Intriguingly, striatum-specific IPA-clustered pathways such as Wnt/β-catenin/Sox pathway, axonal guidance signaling, GABA receptor signaling and Dopamine-DAPRR32 feedback in cAMP signaling clearly represent striatum selective pathological mechanisms. In support of the effectiveness of our IPA pathway identification results, aberrant glutamate receptor/calcium signaling/dopamine signaling/neuronal CREB signaling pathways have been previously implicated in HD pathogenesis (35,39,40,42,46,51–59) (Fig. 5A). More interestingly, we observed that dysregulation of the Wnt/β-catenin/Sox pathway and the axonal guidance signaling pathway, which are very important for neurogenesis and neuron survival (60,61), could be a new pathological mechanism(s) for HD onset (Fig. 5B). Genes involved in each specific pathway are listed in Supplementary Material, Dataset S7.

Figure 5.

Pathway analyses of the disease-related DhMRs-located genes. The functions data represent the logarithm of P-values calculated by Fisher's exact test, with a threshold for statistical significance, P < 0.05. Canonical pathways with P < 0.05 were defined as significant. Genes involved in each specific pathway and function are listed in Supplementary Material, Datasets S7 and 8, respectively. (A) Enrichment of genes acquired/lost DhMRs in YAC128 striatum in glutamate receptor signaling, dopamine/cAMP/DARPP32 feedback pathway, CREB signaling in neuron and LTP. Color indicates the strength of fold change of 5-hmC read densities. Red represents 5-hmC peak decrease in YAC128; Green represents 5-hmC peak increase in YAC128. The fold change value of each gene encoding the protein in the signaling is listed below. GNAI2 (Gαi), 4.38; CAMK4, 8.33; PRKAG2 (PKA), 4.27; GNAQ (Gαq), 4.11; CACNA1C (VDCC-α1C, β4), 4.97; PPP2R2C (PP2A, regulatory subunit B), 4.11; KCNJ6, 6.55; CREB5, 4.71; CACNA1A (VDCC-α1A, P/Q), 5.62; GNB1 (Gβ1), 5.05; GRIA1 (AMPAR1), 8.63; GRID2 (IGluRδ2, ionotropic glutamate receptor delta 2), 8.79. (B) Enrichment of genes acquired/lost DhMRs in YAC128 striatum in the wnt/β-catenin/Sox pathway. Color indicates the strength of fold change of 5-hmC read densities. Red represents 5-hmC peak decrease in YAC128; Green represents 5-hmC peak increase in YAC128. The fold change value of each gene encoding the protein in the signaling is listed below. WNT4, 5.37; WNT9B, 4.35; WIF1, 6.7; SFRP1, 4.21; LRP6, 6.16; NLK, 6.94; TLE3 (Groucho), 4.55; GNAQ (Gαq), 4.11; PPP2R2C (PP2A, regulatory subunit B), 4.11; SOX5, 4.93; TCF4, 4.45.

Figure 5.

Pathway analyses of the disease-related DhMRs-located genes. The functions data represent the logarithm of P-values calculated by Fisher's exact test, with a threshold for statistical significance, P < 0.05. Canonical pathways with P < 0.05 were defined as significant. Genes involved in each specific pathway and function are listed in Supplementary Material, Datasets S7 and 8, respectively. (A) Enrichment of genes acquired/lost DhMRs in YAC128 striatum in glutamate receptor signaling, dopamine/cAMP/DARPP32 feedback pathway, CREB signaling in neuron and LTP. Color indicates the strength of fold change of 5-hmC read densities. Red represents 5-hmC peak decrease in YAC128; Green represents 5-hmC peak increase in YAC128. The fold change value of each gene encoding the protein in the signaling is listed below. GNAI2 (Gαi), 4.38; CAMK4, 8.33; PRKAG2 (PKA), 4.27; GNAQ (Gαq), 4.11; CACNA1C (VDCC-α1C, β4), 4.97; PPP2R2C (PP2A, regulatory subunit B), 4.11; KCNJ6, 6.55; CREB5, 4.71; CACNA1A (VDCC-α1A, P/Q), 5.62; GNB1 (Gβ1), 5.05; GRIA1 (AMPAR1), 8.63; GRID2 (IGluRδ2, ionotropic glutamate receptor delta 2), 8.79. (B) Enrichment of genes acquired/lost DhMRs in YAC128 striatum in the wnt/β-catenin/Sox pathway. Color indicates the strength of fold change of 5-hmC read densities. Red represents 5-hmC peak decrease in YAC128; Green represents 5-hmC peak increase in YAC128. The fold change value of each gene encoding the protein in the signaling is listed below. WNT4, 5.37; WNT9B, 4.35; WIF1, 6.7; SFRP1, 4.21; LRP6, 6.16; NLK, 6.94; TLE3 (Groucho), 4.55; GNAQ (Gαq), 4.11; PPP2R2C (PP2A, regulatory subunit B), 4.11; SOX5, 4.93; TCF4, 4.45.

IPA of DhMR-annotated genes cataloged a variety of functions such as tissue development, embryonic development, nervous system development and function, organ development, cell-to-cell signaling and interaction, behavior, cellular assembly and organization, molecular transport, cardiovascular disease and genetic disorder that may be related to striatum pathology (Supplementary Material, Fig. S4B; Supplementary Material, Dataset S8). Similar to IPA-pathway results, some of these functions also appeared in IPA cataloged functions of cortex DhMR genes (Supplementary Material, Fig. S4D; Supplementary Material, Dataset S8), suggesting that some dysfunctions affected by HTTexp were similar in striatum and cortex in HD. A number of striatum-specific IPA-clustered functions such as behavior, molecular transport, cardiovascular disease and genetic disorder may represent striatum selective pathological mechanisms. Genes involved in each specific function are listed in Supplementary Material, Dataset S8.

DISCUSSION

In this study, we observed that mutant huntingtin (HTTexp) is associated with a marked reduction of the 5-hmC landscape, indicating that loss of the 5-hmC marker is a new epigenetic feature of HD. Moreover, we found that acquisition of DhMRs in gene bodies is a positive regulator for neuronal gene expression. Thus, by perturbing the normal regulation of 5-hmC epigenome, HTTexp could disturb gene expression and cause neuronal dysfunction long before any signs of neuronal death. We identified a number of pathways involving neuronal development/maturation and neurotransmitter signaling that could be important for the onset of HD. Our study also points to a new avenue for HD treatment, reestablishing 5-hmC levels and landscape may have the potential to slow/halt the progression of HD.

YAC128 mice as models of HD

Selective degeneration of striatal neurons is a pathologic hallmark of HD, and neurons missing in the cortex region have also been shown to occur in advanced HD patients. Several HD mouse models have been created to study the underlying mechanisms for HD (38). YAC128 transgenic mice express a mutant huntingtin gene containing 128 CAG repeats under the human endogenous regulatory elements (44). YAC128 mice exhibit slow progressive motor deficits, cognitive impairment, shortened lifespan and progressive selective degeneration of striatal neurons, thereby recapitulating many key features of the human disease (38,44–48); therefore, the results obtained from YAC128 HD mice are expandable to human disease conditions. YAC128 HD mice do not show apparent motor and neuropathological deficits at 3 months of age (44,46). Since we aimed to explore the potential role of 5-hmC in HD onset, we generated detailed 5-hmC epigenomic maps from mouse brains at 3 months of age which represents the early stage of the disease. Because we were using a mixture of genomic DNAs extracted from five mice from either WT or YAC128 brains, the variations of different individuals in 5-hmC epigenome can be excluded in our study.

Loss of 5-hmC in HD neurons

It has been recently recognized that 5-hmC distribution undergoes an interesting dynamic change in the entire genome during development. The level of 5-hmC in mouse ES cells is quite high. It decreases significantly during ES cell differentiation (17,21), then rebounds during the developmental process resulting in elevated 5-hmC in terminally differentiated cells, particularly in neurons (20,23,24,49,62). Consistent with a recent report that 5-hmC levels increase during postnatal development in cerebellum and hippocampus (24), we observed a substantial increase in 5-hmC in both striatum and cortex regions, with the highest level observed at 3 months of age, followed by reduced levels at 8 months of age, suggesting that continuous remodeling of the 5-hmC epigenomic architecture takes place in striatum and cortex. More importantly, by comparing 5-hmC levels between age-matched DNA samples from WT and HD mouse brain, we found that 5-hmC markedly decreased in both striatum and cortex in YAC128 mice as early as 6 weeks of age, indicating that HTTexp somehow causes deficiencies in the 5-hmC landscape in HD neurons (Fig. 1A and F). A genome-wide mapping of 5-hmC reads density of 3-month-old brain tissues authentically confirmed the significant diminishment of 5-hmC in YAC128 striatum (∼55% reduction in peak number) and cortex (∼40% reduction in peak number) (Fig. 2A, Table 1). Thus, genome-wide loss of the 5-hmC marker is a new epigenetic feature of HD, and the decrease of 5-hmC signal could be potentially regarded as a biomarker for HD.

The mechanism(s) underlying the deficiency of 5-hmC reconstruction during postnatal neuronal development in HD brain need to be resolved. Dysfunction of Tet family members presumably influences 5-hmC levels (17,18), and MeCP2 may also regulate 5-hmC levels through neuronal activity-dependent DNA demethylation (63–65). No Tet and MeCP2 mutations have been reported in HD, suggesting that other mechanism(s) interfering with 5-hmC homeostasis may play a role in down-regulation of 5-hmC in HD. Intriguingly, we observed that the significant down-regulation of Tet2, Tet3 and up-regulation of MeCP2 may account for one of the molecular mechanisms underlying global loss of 5-hmC in YAC128 striatum (Supplementary Material, Fig. S3A). However, expression levels of Tet2, Tet3 and MeCP2 failed to show detectable changes in YAC128 cortex (Supplementary Material, Fig. S3B), suggesting that tissue-specific mechanisms may apply to 5-hmC diminishment in HD cortex. Unlike the striatum-specific reduction of the expression of Tet2 and Tet3 in YAC128 HD mice, Tet1 expression decreases in both striatum and cortex of YAC128 brains (Supplementary Material, Fig. S3), suggesting that reduction of 5-hmC levels in cortex may be caused by lower expression of Tet1 in YAC128 brains. HTTexp has been shown to bind directly to DNA (66), raising the possibility that HTTexp may directly affect the accessibility of epigenetic modifiers to genome. HTTexp also change the neuronal activity (35,40,42,51–54), and normal neuronal activity has been known to cause localized loss of DNA methylation (65,67–69). Other factors such as HTTexp-altered histone modifications (70–75) and DNA-bound transcription factors (15,43,76,77) affecting DNA methylation could also influence 5-hmC levels indirectly. The full mechanistic details of mutant HTT-induced loss of 5-hmC in HD neurons remain to be determined.

Epigenomic features of 5-hmC in HD brains

Large-scale methylome data indicate that cytosine methylation is present throughout mammalian genomes (in intergenic regions, coding regions, mobile elements and certain promoters) and an inverse correlation between cytosine methylation and CpG density: CpG-poor DNA, which comprises most of the genome, shows high levels of cytosine methylation, whereas CpG islands remain mostly unmethylated (78,79). We observed that 5-hmC peaks tended to be depleted in CpG islands in promoter and intergenic regions, but were highly enriched in the intragenic CpG islands, suggesting that promoter and intergenic CpG may have a specific mechanism to protect from 5-hmC modification. Consistent with recent reports of 5-hmC distribution in hippocampus and cerebellum (24) and in ES cells, (80–83), our data revealed that the distribution patterns of 5-hmC in different categories of genomic regions are highly conserved in different brain tissues and not penetrated by HTTexp.

Correlation of DhMR with gene transcription

It is well established that methylated DNA at the gene promoter region blocks the binding of transcriptional factors and thus represses gene expression (84). While 5-hmC has been proposed to serve as either a stable epigenetic modification to DNA that is distinct from 5-mC or an intermediate of 5-mC demethylation, the association between 5-hmC enrichment and the gene transcription activity remains an open question (19, 49, 80, 81, 85–87). Genome-wide expression profiling of Tet1-depleted mouse ES cells revealed that Tet1 may have both repressive and activating functions on its direct target genes (81,88,89), and a reduced 5-hmC level was associated with both up- and down-regulated genes in melanomas compared with nevi (90), suggesting that 5-hmC may play dual functions in transcription regulation in ES or cancer cells. In this study, we examined the expression level of genes which were genotype-specifically acquired DhMRs in WT or in YAC128 brain tissues, and found that eight out of nine genes with DhMRs are positively correlated with gene transcription (Fig. 4A, C and E, Supplementary Material, Fig. S2A–L). We did not find any examples of gene transcription that were negatively associated with 5-hmC enrichment. The genomic IGV view of disease-related DhMRs of those genes indicated that 5hmC-enriched loci occurred mainly in introns of the gene bodies (Fig. 4B, D and F, Supplementary Material, Fig. S2A–L), suggesting that intragenic 5-hmC-enrichment is likely a positive epigenetic regulator of gene expression. Our findings are in agreement with a recent report that 5-hmC is associated with developmentally activated genes and actively transcribed genes in cerebellum and hippocampus of mouse brains (24). Furthermore, in cerebellum neurons, many actively transcribed genes display both significant 5-hmC enrichment within the gene bodies and the depletion of 5 mC (26). Therefore, it is possible that the regulation of 5-hmC and associated gene transcription in neuronal cells are different from that in dividing cells such as ES and cancer cells.

Functional implications of disease-related DhMRs

In this study, by cataloging the genes acquiring genotype-specific DhMRs in striatum and cortex samples, we identified a number of changed pathways and functions that may be important for HD onset. Some pathways such as glutamate receptor signaling, calcium signaling, dopamine-DAPRR32 feedback signaling and neuronal CREB signaling have been previously associated with neurodegeneration in HD (35, 39, 40, 42, 46, 51–54, 56–59, 91). It is well established that CREB-mediated transcription plays a key role in neuronal functions and neuronal survival. Several lines of evidence (15,39,57–59) including the present study (Fig. 5A) observe that CREB functionality is significantly decreased in HD neurons. Moreover, our data here demonstrated that not only CREB itself but also upstream members such as AMPAR, VDCC-α1C, IGluR-δ2 (ionotropic glutamate receptor δ2) and CaMKIV were down-regulated due to decreased 5-hmC during the early stage in HD. Therefore, epigenetic alternations in the 5-hmC marker of CREB pathway members may contribute to the early pathogenic process in HD.

It is somewhat surprising that pathways involving neurogenesis/neuronal differentiation such as Wnt/β-catenin/Sox pathway and axonal guidance signaling are highly enriched in IPA analysis of disease-related striatum-DhMR-marked genes (Supplementary Material, Fig. S4A,B and 5B; Supplementary Material, Dataset S7). The Wnt/β-catenin/Sox pathway regulates both the proliferation of neural progenitor cells and their differentiation to neurons in the subventricular and subgranular zones (SVZs) (92–94). Moreover, the Wnt signaling pathway is an obligate component of neural progenitor cell differentiation into neurons via activation of NeuroD1 (60,61), which is a key factor for neurogenesis, differentiation and neuron survival. Genes associated with neurogenesis and neuronal differentiation (Supplementary Material, Dataset S7) such as Wnt4, Wnt9B, LRP6, NLK, Sox5, SFRP1, Slit1, Slit3, EPHA7 and Kalrn show decreased 5-hmC signals and presumably lower expression levels in HD striatum. Our findings here are consistent with the recent literature, indicating that disturbed neurogenesis is involved in pathogenesis of HD (95,96). Our data also suggest that delayed and/or insufficient endogenous neurogenesis to counteract the striatum neuropathology may occur at a very early stage of HD. Several reagents that potentiated SVZ neurogenesis have been shown to be beneficial in HD mouse models (97–100), manifesting the important role of impaired neurogenesis in HD onset. Therefore, reconstructing the 5-hmC epigenome in HD brains could form a new therapeutic target against HD onset/progression.

MATERIALS AND METHODS

Preparation of genomic DNA

YAC128 HD transgenic mice were obtained from The Jackson Laboratory (Bar Harbor, ME, USA). Generation and breeding of YAC128 transgenic mice (FVBN_NJ background strain) have been described previously (44). Striatum and cortex regions were dissected from 6-week-, 3-month-, and 8-month-old female mouse brains as described previously (53,101). Genomic DNA was prepared using a Wizard Genomic DNA Purification kit (Promega Cat.:A1120) by following the manufacturer's instructions. Equal amounts of genomic DNAs extracted from either striatum or cortex in WT mice (n = 5) were pooled together and named as WT striatum or WT cortex DNA sample, respectively. YAC128 mouse brain striatum and cortex DNA samples were prepared similarly. The four groups of genomic DNA samples were named as WT striatum, YAC128 striatum, WT cortex and YAC128 cortex, respectively. All animal experiments were reviewed and approved by the Institute of Zoology Institutional Animal Care and Use Committee and were conducted according to the committee's guidelines.

Immuno-dot-blot assay

Genomic DNA from 6-week-, 3-month- and 8-month-old WT and YAC128 mice brain striatum and cortex was denatured in TE buffer for 10 min at 95°C and immediately chilled on ice for 5 min. Dot blot was performed on a Bio-Dot Apparatus (#170-6545,Bio-Rad). 10, 20 and 50 ng of each DNA sample was spotted on the positively charged nylon membrane, respectively, then the membrane was baked for 2 h at 80°C until completely dry, followed by UV254 crosslink for 10 min to fix DNA on the membrane. The membrane was then blocked briefly with 5% non-fat milk for 1.5 h at room temperature. The primary rabbit anti-5-hydroxymethylytosine antibody (1:10000, #39769, Active Motif) was applied to the membrane and incubated at RT for 1 h or overnight at 4°C. After incubation with a peroxidase-conjugated anti-rabbit IgG secondary antibody, the signal was visualized by using ECL (Millipore). The dot-blot densities were analyzed with Image J software. The 5-hmC signals were normalized to 6-week-old YAC128 striatum 5-hmC density. The 5-hmC-containing DNA was used as a positive control, and the normal C, 5-mC, 5-carC-containing DNA were used as the negative controls to verify the specificity of 5-hmC antibody.

Immunofluorescent stainings of 5-hmC

Six-week- and 4-month-old WT and YAC128 mice were anesthetized with pentobarbital sodium (8 mg per kg) and perfused with phosphate-buffered saline, followed by perfusion with 4% paraformaldehyde (W/V). Brains were then dehydrated in 30% sucrose (W/V). Brain sections (35 µm) were prepared with a cryomicrotome (Leica, CM1900). For 5-hmC immunostaining, sections were treated with 1m HCl at 37°C for 30 min followed by blocking with 5% FBS in PBS. For primary antibodies, we used a rabbit antibody against 5-hmC (1:20 000, #39769, Active Motif), mouse antibody against neuronal nuclei (1:500, #MAB377, Millipore). For secondary antibodies, we used donkey against mouse IgG conjugated with Alexa488 (1:500, #A11008, Invitrogen) and donkey against rabbit IgG conjugated with Alexa546 (1:500, #A11031, Invitrogen). Sections were counter-stained with the nuclear dye Hochest33324 (#B2261, Sigma). All images of striatum and cortex were taken using a microscope (Leica DMI6000 B TIRF MC) with the same capture parameters.

5-hmC specific chemical labeling and affinity purification

Purified genomic DNA was sonicated into short fragments by Covaris DNA shearing with microTUBEs according to manufacturer's instructions. Then 5-hmC labeling reactions were performed in 75 μl solution containing 50 mm HEPES buffer (pH7.9), 250 mm MgCl2, 100 µm UDP-6-N3-glucose and 80U β-glucosyltransferase. The reaction was incubated for 2 h at 37°C. After the reaction, DNA substrates were purified and buffer-exchanged in H2O via Bio-Rad-Spin columns according to the manufacturer's instructions. Click chemistry was performed with the addition of 150 µm biotin into the DNA solution and incubated for 2 h at 37°C. The DNA samples were then purified by Invitrogen Dynabeads MyOneTM Streptavidin C1 according to manufacturer's instructions.

Sequencing of 5-hmC-enriched genomic DNA

5-hmC enriched genomic DNA libraries were generated following the Illumina protocol for ‘Preparing Samples for CHIP sequencing of DNA’. Then, 20 ng of 5-hmC enriched DNA was used to initiate the protocol. DNA fragments were gel purified after the adapter ligation step. PCR-amplified DNA libraries were quantified on an Agilent 2100 Bio analyzer using a quantitative PCR. We performed 100 bp single end sequencing on Illumina Hiseq2000 to get a 5-hmC-enriched DNA fragment sequence.

Sequence alignment and peak identification

FASTQ sequence files of WT striatum, YAC128 striatum, WT cortex and YAC128 cortex were aligned to the mouse genome (mm9) using Bowtie 0.12.7. In order to improve the quality of the reads, we removed 5 and 10 bases from the 5′ and 3′ ends of the reads, respectively. A read containing the remaining 85 bases was mapped to the genome with no more than three mismatches in the first 75 bp. If a read was mapped to a unique locus of the genome, it was named as a uni-map read. The analysis of genomic distribution of 5-hmC was conducted using these uni-map reads.

Peak identification was performed using a Poisson-based peak identification algorithm (MACS) (50) with uni-map reads. Parameters were as follows: effective genome size = 2.7 × 109; tag size = 85; band width = 300; q-value cutoff = 0.05; other parameters were set to default values.

DhMRs identification

The MACS software was used again to identify DhMRs (differentially hydroxymethylated regions) between two samples with one as the control and the other as the treatment. These DhMRs represented peaks whose 5-hmC content was significantly higher in the treatment sample than in the control sample. Then another set of DhMRs were identified by switching the control and treatment samples. The parameters were set as the following: effective genome size = 2.7 × 109; tag size = 85; band width = 300; q-value cutoff = 0.05; other parameters were set to default values. Genotype-specific (i.e., disease-related) DhMRs were identified by bidirectional comparisons between WT and YAC128 samples from striatum (or from cortex).

Mapping 5-hmC reads to various genomic regions

Mapping of 5-hmC reads, or peaks, or the DhMRs to different types of genomic regions such as exons and introns was performed by comparing the coordinates of uni-map reads with the coordinates of these genomic regions, which were obtained from the University of California Santa Cruz tables for mm9: RefSeq Whole Gene, 5′ UTR, Exon, Intron, 3′ UTR, coding regions, promoter and Intragenic region based on RefSeq Whole Gene. A read/peak/DhMR was mapped to a given genomic region if they overlapped by ≥1 bp based on their coordinates in the genome. For mapping 5-hmC reads to repetitive elements, reads were aligned to the RepeatMasker track of NCBI37v1/mm9 using Bowtie, using the same parameters as in sequence alignment. Aligned reads were assigned to repeat classes defined by RepeatMasker.

Evaluation of the 5-hmC content in different genomic regions

The content of 5-hmC in a particular type of genomic region such as exons, introns, UTRs or a user-defined window was quantified using the number of reads (uni-map reads, more exactly) as the measurement. As the sequencing depth of each sample varied, the number of reads was first normalized by the total reads of the sample to give a value with the unit being reads per million (RPM). A RPM value was further converted to a density value by dividing the former with the length of a genomic region. For evaluating the 5-hmC contents in genomic regions such as chromosomes, gene bodies, exons, introns, UTRs, promoters, intergenic regions, CpG islands, DhMRs, the numbers of reads in the peaks that mapped to these regions were used. Any read that was not in a peak was not counted. For the repetitive elements, however, all reads were counted without considering whether they were in a peak, and abundance in a repeat class was represented by the percentage of reads mapped to this class in all repeat-mapped reads.

Read coverage and visualization

Genomic views of read coverage were generated using Integrated Genomics Viewer tools and browser (IGV 1.4.2, http://www.broadinstitute.org/igv/, last accessed date on May 14, 2013) with a window size of 300.

Ingenuity pathway analysis

Genes with ≥1 DhMRs were listed and used as input dataset for the IPA software (Ingenuity Systems, www.ingenuity.com; Redwood City, CA, USA). By applying IPA, the gene sets can be functionally annotated and the biologically relevant pathways could be identified.

RNA isolation and mRNA level of DhMR-containing genes

The brain striatum and cortex samples from four WT mice and four YAC128 mice of 3-month-old were collected. Total RNA was extracted from 30 mg of each sample using the Tissue RNA Extraction kit (Tiangen) following the standard manual. The complementary DNA (cDNA) of 1 μg RNA sample was reverse-transcribed using the First Strand cDNA Synthesis kit (Promega). The cDNA was diluted with sterilized ddH2O at the rate of 1:4 for Q-PCR.

The expression of DhMR-enriched genes in the four WT and four YAC128 mice striatum and cortex samples was detected and Q-PCRs were performed in triplicate. The gene expression levels were quantified relative to the expression of mouse GAPDH gene, employing an optimized comparative Ct (ΔΔCt) value method. The primers designed for target genes and internal control GAPDH are shown in Supplementary Material, Table S1.

Statistics

Data are expressed as the mean ± standard error of the mean (s. e. m.), and statistical significance of differences between different groups was assessed using the t-test or ANOVA with P < 0.05.

SUPPLEMENTARY MATERIAL

Supplementary Material is available at HMG online.

FUNDING

This work was supported by grants from the National Basic Research Program of China (2011CB965003, 2012CB944702), NSFC (30970931, 30970588 and 31170730) and by a grant from the Knowledge Innovation Program of Chinese Academy of Sciences (KSCX2-YW-R-148). This work was also supported by One-Hundred-Talent Program of CAS (to Dr C. Guo and Dr Q. Sun).

ACKNOWLEDGMENTS

We apologize to those investigators whose work we could not cite due to the reference limit, and gratefully acknowledge their contributions to HD field. We thank Qiaochu Wang, and Yun Wang for help with maintaining the YAC mouse colony, Ilya Bezprozvanny for facilitating transportation of mice strains and Paula L. Fischhaber for proofreading the manuscript.

Conflict of Interest statement. None declared.

REFERENCES

1
Li
E.
Bestor
T.H.
Jaenisch
R.
Targeted mutation of the DNA methyltransferase gene results in embryonic lethality
Cell
 , 
1992
, vol. 
69
 (pg. 
915
-
926
)
2
Li
E.
Beard
C.
Jaenisch
R.
Role for DNA methylation in genomic imprinting
Nature
 , 
1993
, vol. 
366
 (pg. 
362
-
365
)
3
Jaenisch
R.
Bird
A.
Epigenetic regulation of gene expression, how the genome integrates intrinsic and environmental signals
Nat. Genet.
 , 
2003
, vol. 
33
 
Suppl
(pg. 
245
-
254
)
4
Reik
W.
Stability and flexibility of epigenetic gene regulation in mammalian development
Nature
 , 
2007
, vol. 
447
 (pg. 
425
-
432
)
5
Zemach
A.
McDaniel
I.E.
Silva
P.
Zilberman
D.
Genome-wide evolutionary analysis of eukaryotic DNA methylation
Science
 , 
2010
, vol. 
328
 (pg. 
916
-
919
)
6
Okano
M.
Bell
D.W.
Haber
D.A.
Li
E.
DNA methyltransferases Dnmt3a and Dnmt3b are essential for de novo methylation and mammalian development
Cell
 , 
1999
, vol. 
99
 (pg. 
247
-
257
)
7
Bestor
T.H.
The DNA methyltransferases of mammals
Hum. Mol. Genet.
 , 
2000
, vol. 
9
 (pg. 
2395
-
2402
)
8
Goll
M.G.
Bestor
T.H.
Eukaryotic cytosine methyltransferases
Annu. Rev. Biochem.
 , 
2005
, vol. 
74
 (pg. 
481
-
514
)
9
Rottach
A.
Leonhardt
H.
Spada
F.
DNA methylation-mediated epigenetic control
J. Cell. Biochem.
 , 
2009
, vol. 
108
 (pg. 
43
-
51
)
10
Wu
H.
Coskun
V.
Tao
J.
Xie
W.
Ge
W.
Yoshikawa
K.
Li
E.
Zhang
Y.
Sun
Y.E.
Dnmt3a-dependent nonpromoter DNA methylation facilitates transcription of neurogenic genes
Science
 , 
2010
, vol. 
329
 (pg. 
444
-
448
)
11
Hutnick
L.K.
Golshani
P.
Namihira
M.
Xue
Z.
Matynia
A.
Yang
X.W.
Silva
A.J.
Schweizer
F.E.
Fan
G.
DNA hypomethylation restricted to the murine forebrain induces cortical degeneration and impairs postnatal neuronal maturation
Hum. Mol. Genet.
 , 
2009
, vol. 
18
 (pg. 
2875
-
2888
)
12
Klein
C.J.
Botuyan
M.V.
Wu
Y.
Ward
C.J.
Nicholson
G.A.
Hammans
S.
Hojo
K.
Yamanishi
H.
Karpf
A.R.
Wallace
D.C.
, et al.  . 
Mutations in DNMT1 cause hereditary sensory neuropathy with dementia and hearing loss
Nat. Genet.
 , 
2011
, vol. 
43
 (pg. 
595
-
600
)
13
Winkelmann
J.
Lin
L.
Schormair
B.
Kornum
B.R.
Faraco
J.
Plazzi
G.
Melberg
A.
Cornelio
F.
Urban
A.E.
Pizza
F.
, et al.  . 
Mutations in DNMT1 cause autosomal dominant cerebellar ataxia, deafness and narcolepsy
Hum. Mol. Genet.
 , 
2012
, vol. 
21
 (pg. 
2205
-
2210
)
14
Psoni
S.
Sofocleous
C.
Traeger-Synodinos
J.
Kitsiou-Tzeli
S.
Kanavakis
E.
Fryssira-Kanioura
H.
MECP2 mutations and clinical correlations in Greek children with Rett syndrome and associated neurodevelopmental disorders
Brain Dev.
 , 
2012
, vol. 
34
 (pg. 
487
-
495
)
15
Ng
C.W.
Yildirim
F.
Yap
Y.S.
Dalin
S.
Matthews
B.J.
Velez
P.J.
Labadorf
A.
Housman
D.E.
Fraenkel
E.
Extensive changes in DNA methylation are associated with expression of mutant huntingtin
Proc. Natl Acad. Sci. USA
 , 
2013
, vol. 
110
 (pg. 
2354
-
2359
)
16
Irier
H.A.
Jin
P.
Dynamics of DNA methylation in aging and Alzheimer's disease
DNA Cell. Biol.
 , 
2012
, vol. 
31
 
Suppl 1
(pg. 
S42
-
48
)
17
Tahiliani
M.
Koh
K.P.
Shen
Y.
Pastor
W.A.
Bandukwala
H.
Brudno
Y.
Agarwal
S.
Iyer
L.M.
Liu
D.R.
Aravind
L.
, et al.  . 
Conversion of 5-methylcytosine to 5-hydroxymethylcytosine in mammalian DNA by MLL partner TET1
Science
 , 
2009
, vol. 
324
 (pg. 
930
-
935
)
18
Ito
S.
D'Alessio
A.C.
Taranova
O.V.
Hong
K.
Sowers
L.C.
Zhang
Y.
Role of Tet proteins in 5mC to 5hmC conversion, ES-cell self-renewal and inner cell mass specification
Nature
 , 
2010
, vol. 
466
 (pg. 
1129
-
1133
)
19
Pastor
W.A.
Pape
U.J.
Huang
Y.
Henderson
H.R.
Lister
R.
Ko
M.
McLoughlin
E.M.
Brudno
Y.
Mahapatra
S.
Kapranov
P.
, et al.  . 
Genome-wide mapping of 5-hydroxymethylcytosine in embryonic stem cells
Nature
 , 
2011
, vol. 
473
 (pg. 
394
-
397
)
20
Kriaucionis
S.
Heintz
N.
The nuclear DNA base 5-hydroxymethylcytosine is present in Purkinje neurons and the brain
Science
 , 
2009
, vol. 
324
 (pg. 
929
-
930
)
21
Szwagierczak
A.
Bultmann
S.
Schmidt
C.S.
Spada
F.
Leonhardt
H.
Sensitive enzymatic quantification of 5-hydroxymethylcytosine in genomic DNA
Nucleic Acids Res.
 , 
2010
, vol. 
38
 pg. 
e181
 
22
Munzel
M.
Globisch
D.
Bruckl
T.
Wagner
M.
Welzmiller
V.
Michalakis
S.
Muller
M.
Biel
M.
Carell
T.
Quantification of the sixth DNA base hydroxymethylcytosine in the brain
Angew. Chem. Int. Ed. Engl.
 , 
2010
, vol. 
49
 (pg. 
5375
-
5377
)
23
Li
W.
Liu
M.
Distribution of 5-hydroxymethylcytosine in different human tissues
J. Nucleic Acids
 , 
2011
, vol. 
2011
 pg. 
870726
 
24
Szulwach
K.E.
Li
X.
Li
Y.
Song
C.X.
Wu
H.
Dai
Q.
Irier
H.
Upadhyay
A.K.
Gearing
M.
Levey
A.I.
, et al.  . 
5-hmC-mediated epigenetic dynamics during postnatal neurodevelopment and aging
Nat. Neurosci.
 , 
2011
, vol. 
14
 (pg. 
1607
-
1616
)
25
Wang
T.
Pan
Q.
Lin
L.
Szulwach
K.E.
Song
C.X.
He
C.
Wu
H.
Warren
S.T.
Jin
P.
Duan
R.
, et al.  . 
Genome-wide DNA hydroxymethylation changes are associated with neurodevelopmental genes in the developing human cerebellum
Hum. Mol. Genet.
 , 
2012
, vol. 
21
 (pg. 
5500
-
5510
)
26
Mellen
M.
Ayata
P.
Dewell
S.
Kriaucionis
S.
Heintz
N.
MeCP2 binds to 5hmC enriched within active genes and accessible chromatin in the nervous system
Cell
 , 
2012
, vol. 
151
 (pg. 
1417
-
1430
)
27
Vonsattel
J.P.
DiFiglia
M.
Huntington disease
J. Neuropathol. Exp. Neurol.
 , 
1998
, vol. 
57
 (pg. 
369
-
384
)
28
MacDonald
M.E.
Gines
S.
Gusella
J.F.
Wheeler
V.C.
Huntington's disease
Neuromolecular Med.
 , 
2003
, vol. 
4
 (pg. 
7
-
20
)
29
A novel gene containing a trinucleotide repeat that is expanded and unstable on Huntington's disease chromosomes
The Huntington's Disease Collaborative Research Group
Cell
 , 
1993
, vol. 
72
 (pg. 
971
-
983
)
30
Young
A.B.
Huntingtin in health and disease
J. Clin. Invest.
 , 
2003
, vol. 
111
 (pg. 
299
-
302
)
31
Tobin
A.J.
Signer
E.R.
Huntington's disease: the challenge for cell biologists
Trends Cell Biol.
 , 
2000
, vol. 
10
 (pg. 
531
-
536
)
32
Ross
C.A.
Polyglutamine pathogenesis: emergence of unifying mechanisms for Huntington's disease and related disorders
Neuron
 , 
2002
, vol. 
35
 (pg. 
819
-
822
)
33
Harjes
P.
Wanker
E.E.
The hunt for huntingtin function: interaction partners tell many different stories
Trends Biochem. Sci.
 , 
2003
, vol. 
28
 (pg. 
425
-
433
)
34
Sugars
K.L.
Rubinsztein
D.C.
Transcriptional abnormalities in Huntington disease
Trends Genet.
 , 
2003
, vol. 
19
 (pg. 
233
-
238
)
35
Bezprozvanny
I.
Hayden
M.R.
Deranged neuronal calcium signaling and Huntington disease
Biochem. Biophys. Res. Commun.
 , 
2004
, vol. 
322
 (pg. 
1310
-
1317
)
36
Li
S.
Li
X.J.
Multiple pathways contribute to the pathogenesis of Huntington disease
Mol. Neurodegener.
 , 
2006
, vol. 
1
 pg. 
19
 
37
Hodges
A.
Strand
A.D.
Aragaki
A.K.
Kuhn
A.
Sengstag
T.
Hughes
G.
Elliston
L.A.
Hartog
C.
Goldstein
D.R.
Thu
D.
, et al.  . 
Regional and cellular gene expression changes in human Huntington's disease brain
Hum. Mol. Genet.
 , 
2006
, vol. 
15
 (pg. 
965
-
977
)
38
Kuhn
A.
Goldstein
D.R.
Hodges
A.
Strand
A.D.
Sengstag
T.
Kooperberg
C.
Becanovic
K.
Pouladi
M.A.
Sathasivam
K.
Cha
J.H.
, et al.  . 
Mutant huntingtin's effects on striatal gene expression in mice recapitulate changes observed in human Huntington's disease brain and do not differ with mutant huntingtin length or wild-type huntingtin dosage
Hum. Mol. Genet.
 , 
2007
, vol. 
16
 (pg. 
1845
-
1861
)
39
Cui
L.
Jeong
H.
Borovecki
F.
Parkhurst
C.N.
Tanese
N.
Krainc
D.
Transcriptional repression of PGC-1alpha by mutant huntingtin leads to mitochondrial dysfunction and neurodegeneration
Cell
 , 
2006
, vol. 
127
 (pg. 
59
-
69
)
40
Fan
M.M.
Raymond
L.A.
N-Methyl-d-aspartate (NMDA) receptor function and excitotoxicity in Huntington's disease
Prog. Neurobiol.
 , 
2007
, vol. 
81
 (pg. 
272
-
293
)
41
Trushina
E.
McMurray
C.T.
Oxidative stress and mitochondrial dysfunction in neurodegenerative diseases
Neuroscience
 , 
2007
, vol. 
145
 (pg. 
1233
-
1248
)
42
Bezprozvanny
I.
Calcium signaling and neurodegenerative diseases
Trends Mol. Med.
 , 
2009
, vol. 
15
 (pg. 
89
-
100
)
43
Seredenina
T.
Luthi-Carter
R.
What have we learned from gene expression profiles in Huntington's disease?
Neurobiol. Dis.
 , 
2012
, vol. 
45
 (pg. 
83
-
98
)
44
Slow
E.J.
van Raamsdonk
J.
Rogers
D.
Coleman
S.H.
Graham
R.K.
Deng
Y.
Oh
R.
Bissada
N.
Hossain
S.M.
Yang
Y.Z.
, et al.  . 
Selective striatal neuronal loss in a YAC128 mouse model of Huntington disease
Hum. Mol. Genet.
 , 
2003
, vol. 
12
 (pg. 
1555
-
1567
)
45
Van Raamsdonk
J.M.
Murphy
Z.
Slow
E.J.
Leavitt
B.R.
Hayden
M.R.
Selective degeneration and nuclear localization of mutant huntingtin in the YAC128 mouse model of Huntington disease
Hum. Mol. Genet.
 , 
2005
, vol. 
14
 (pg. 
3823
-
3835
)
46
Tang
T.S.
Chen
X.
Liu
J.
Bezprozvanny
I.
Dopaminergic signaling and striatal neurodegeneration in Huntington's disease
J. Neurosci.
 , 
2007
, vol. 
27
 (pg. 
7899
-
7910
)
47
Fernandes
H.B.
Baimbridge
K.G.
Church
J.
Hayden
M.R.
Raymond
L.A.
Mitochondrial sensitivity and altered calcium handling underlie enhanced NMDA-induced apoptosis in YAC128 model of Huntington's disease
J. Neurosci.
 , 
2007
, vol. 
27
 (pg. 
13614
-
13623
)
48
Mazarei
G.
Neal
S.J.
Becanovic
K.
Luthi-Carter
R.
Simpson
E.M.
Leavitt
B.R.
Expression analysis of novel striatal-enriched genes in Huntington disease
Hum. Mol. Genet.
 , 
2010
, vol. 
19
 (pg. 
609
-
622
)
49
Song
C.X.
Szulwach
K.E.
Fu
Y.
Dai
Q.
Yi
C.
Li
X.
Li
Y.
Chen
C.H.
Zhang
W.
Jian
X.
, et al.  . 
Selective chemical labeling reveals the genome-wide distribution of 5-hydroxymethylcytosine
Nat. Biotechnol.
 , 
2011
, vol. 
29
 (pg. 
68
-
72
)
50
Zhang
Y.
Liu
T.
Meyer
C.A.
Eeckhoute
J.
Johnson
D.S.
Bernstein
B.E.
Nusbaum
C.
Myers
R.M.
Brown
M.
Li
W.
, et al.  . 
Model-based analysis of ChIP-Seq (MACS)
Genome. Biol.
 , 
2008
, vol. 
9
 pg. 
R137
 
51
Zeron
M.M.
Hansson
O.
Chen
N.
Wellington
C.L.
Leavitt
B.R.
Brundin
P.
Hayden
M.R.
Raymond
L.A.
Increased sensitivity to N-methyl-d-aspartate receptor-mediated excitotoxicity in a mouse model of Huntington's disease
Neuron
 , 
2002
, vol. 
33
 (pg. 
849
-
860
)
52
Tang
T.S.
Tu
H.
Chan
E.Y.
Maximov
A.
Wang
Z.
Wellington
C.L.
Hayden
M.R.
Bezprozvanny
I.
Huntingtin and huntingtin-associated protein 1 influence neuronal calcium signaling mediated by inositol-(1,4,5) triphosphate receptor type 1
Neuron
 , 
2003
, vol. 
39
 (pg. 
227
-
239
)
53
Tang
T.S.
Slow
E.
Lupu
V.
Stavrovskaya
I.G.
Sugimori
M.
Llinas
R.
Kristal
B.S.
Hayden
M.R.
Bezprozvanny
I.
Disturbed Ca2+ signaling and apoptosis of medium spiny neurons in Huntington's disease
Proc. Natl Acad. Sci. USA
 , 
2005
, vol. 
102
 (pg. 
2602
-
2607
)
54
Fan
M.M.
Fernandes
H.B.
Zhang
L.Y.
Hayden
M.R.
Raymond
L.A.
Altered NMDA receptor trafficking in a yeast artificial chromosome transgenic mouse model of Huntington's disease
J. Neurosci.
 , 
2007
, vol. 
27
 (pg. 
3768
-
3779
)
55
Jordt
S.E.
Ehrlich
B.E.
TRP channels in disease
Subcell. Biochem.
 , 
2007
, vol. 
45
 (pg. 
253
-
271
)
56
Mittal
S.K.
Eddy
C.
The role of dopamine and glutamate modulation in Huntington disease
Behav. Neurol.
 , 
2013
, vol. 
26
 (pg. 
255
-
263
)
57
Sugars
K.L.
Brown
R.
Cook
L.J.
Swartz
J.
Rubinsztein
D.C.
Decreased cAMP response element-mediated transcription: an early event in exon 1 and full-length cell models of Huntington's disease that contributes to polyglutamine pathogenesis
J. Biol. Chem.
 , 
2004
, vol. 
279
 (pg. 
4988
-
4999
)
58
Okamoto
S.
Pouladi
M.A.
Talantova
M.
Yao
D.
Xia
P.
Ehrnhoefer
D.E.
Zaidi
R.
Clemente
A.
Kaul
M.
Graham
R.K.
, et al.  . 
Balance between synaptic versus extrasynaptic NMDA receptor activity influences inclusions and neurotoxicity of mutant huntingtin
Nat. Med.
 , 
2009
, vol. 
15
 (pg. 
1407
-
1413
)
59
Choi
Y.S.
Lee
B.
Cho
H.Y.
Reyes
I.B.
Pu
X.A.
Saido
T.C.
Hoyt
K.R.
Obrietan
K.
CREB is a key regulator of striatal vulnerability in chemical and genetic models of Huntington's disease
Neurobiol. Dis.
 , 
2009
, vol. 
36
 (pg. 
259
-
268
)
60
Kuwabara
T.
Hsieh
J.
Muotri
A.
Yeo
G.
Warashina
M.
Lie
D.C.
Moore
L.
Nakashima
K.
Asashima
M.
Gage
F.H.
Wnt-mediated activation of NeuroD1 and retro-elements during adult neurogenesis
Nat. Neurosci.
 , 
2009
, vol. 
12
 (pg. 
1097
-
1105
)
61
Inestrosa
N.C.
Arenas
E.
Emerging roles of Wnts in the adult nervous system
Nat Rev. Neurosci.
 , 
2010
, vol. 
11
 (pg. 
77
-
86
)
62
Haffner
M.C.
Chaux
A.
Meeker
A.K.
Esopi
D.M.
Gerber
J.
Pellakuru
L.G.
Toubaji
A.
Argani
P.
Iacobuzio-Donahue
C.
Nelson
W.G.
, et al.  . 
Global 5-hydroxymethylcytosine content is significantly reduced in tissue stem/progenitor cell compartments and in human cancers
Oncotarget
 , 
2011
, vol. 
2
 (pg. 
627
-
637
)
63
Zhou
Z.
Hong
E.J.
Cohen
S.
Zhao
W.N.
Ho
H.Y.
Schmidt
L.
Chen
W.G.
Lin
Y.
Savner
E.
Griffith
E.C.
, et al.  . 
Brain-specific phosphorylation of MeCP2 regulates activity-dependent Bdnf transcription, dendritic growth, and spine maturation
Neuron
 , 
2006
, vol. 
52
 (pg. 
255
-
269
)
64
Nelson
E.D.
Kavalali
E.T.
Monteggia
L.M.
Activity-dependent suppression of miniature neurotransmission through the regulation of DNA methylation
J. Neurosci.
 , 
2008
, vol. 
28
 (pg. 
395
-
406
)
65
Ma
D.K.
Jang
M.H.
Guo
J.U.
Kitabatake
Y.
Chang
M.L.
Pow-Anpongkul
N.
Flavell
R.A.
Lu
B.
Ming
G.L.
Song
H.
Neuronal activity-induced Gadd45b promotes epigenetic DNA demethylation and adult neurogenesis
Science
 , 
2009
, vol. 
323
 (pg. 
1074
-
1077
)
66
Benn
C.L.
Sun
T.
Sadri-Vakili
G.
McFarland
K.N.
DiRocco
D.P.
Yohrling
G.J.
Clark
T.W.
Bouzou
B.
Cha
J.H.
Huntingtin modulates transcription, occupies gene promoters in vivo, and binds directly to DNA in a polyglutamine-dependent manner
J. Neurosci.
 , 
2008
, vol. 
28
 (pg. 
10720
-
10733
)
67
Chen
W.G.
Chang
Q.
Lin
Y.
Meissner
A.
West
A.E.
Griffith
E.C.
Jaenisch
R.
Greenberg
M.E.
Derepression of BDNF transcription involves calcium-dependent phosphorylation of MeCP2
Science
 , 
2003
, vol. 
302
 (pg. 
885
-
889
)
68
Martinowich
K.
Hattori
D.
Wu
H.
Fouse
S.
He
F.
Hu
Y.
Fan
G.
Sun
Y.E.
DNA methylation-related chromatin remodeling in activity-dependent BDNF gene regulation
Science
 , 
2003
, vol. 
302
 (pg. 
890
-
893
)
69
Guo
J.U.
Ma
D.K.
Mo
H.
Ball
M.P.
Jang
M.H.
Bonaguidi
M.A.
Balazer
J.A.
Eaves
H.L.
Xie
B.
Ford
E.
, et al.  . 
Neuronal activity modifies the DNA methylation landscape in the adult brain
Nat. Neurosci.
 , 
2011
, vol. 
14
 (pg. 
1345
-
1351
)
70
Faber
P.W.
Barnes
G.T.
Srinidhi
J.
Chen
J.
Gusella
J.F.
MacDonald
M.E.
Huntingtin interacts with a family of WW domain proteins
Hum. Mol. Genet.
 , 
1998
, vol. 
7
 (pg. 
1463
-
1474
)
71
Sadri-Vakili
G.
Bouzou
B.
Benn
C.L.
Kim
M.O.
Chawla
P.
Overland
R.P.
Glajch
K.E.
Xia
E.
Qiu
Z.
Hersch
S.M.
, et al.  . 
Histones associated with downregulated genes are hypo-acetylated in Huntington's disease models
Hum. Mol. Genet.
 , 
2007
, vol. 
16
 (pg. 
1293
-
1306
)
72
Kazantsev
A.G.
Thompson
L.M.
Therapeutic application of histone deacetylase inhibitors for central nervous system disorders
Nat. Rev. Drug Discov.
 , 
2008
, vol. 
7
 (pg. 
854
-
868
)
73
McFarland
K.N.
Das
S.
Sun
T.T.
Leyfer
D.
Xia
E.
Sangrey
G.R.
Kuhn
A.
Luthi-Carter
R.
Clark
T.W.
Sadri-Vakili
G.
, et al.  . 
Genome-wide histone acetylation is altered in a transgenic mouse model of Huntington's disease
PLoS One
 , 
2012
, vol. 
7
 pg. 
e41423
 
74
Ryu
H.
Lee
J.
Hagerty
S.W.
Soh
B.Y.
McAlpin
S.E.
Cormier
K.A.
Smith
K.M.
Ferrante
R.J.
ESET/SETDB1 gene expression and histone H3 (K9) trimethylation in Huntington's disease
Proc. Natl Acad. Sci. USA
 , 
2006
, vol. 
103
 (pg. 
19176
-
19181
)
75
Cedar
H.
Bergman
Y.
Linking DNA methylation and histone modification, patterns and paradigms
Nat. Rev. Genet.
 , 
2009
, vol. 
10
 (pg. 
295
-
304
)
76
Hogart
A.
Lichtenberg
J.
Ajay
S.S.
Anderson
S.
Margulies
E.H.
Bodine
D.M.
Genome-wide DNA methylation profiles in hematopoietic stem and progenitor cells reveal overrepresentation of ETS transcription factor binding sites
Genome Res.
 , 
2012
, vol. 
22
 (pg. 
1407
-
1418
)
77
Stadler
M.B.
Murr
R.
Burger
L.
Ivanek
R.
Lienert
F.
Scholer
A.
van Nimwegen
E.
Wirbelauer
C.
Oakeley
E.J.
Gaidatzis
D.
, et al.  . 
DNA-binding factors shape the mouse methylome at distal regulatory regions
Nature
 , 
2011
, vol. 
480
 (pg. 
490
-
495
)
78
Lister
R.
Pelizzola
M.
Dowen
R.H.
Hawkins
R.D.
Hon
G.
Tonti-Filippini
J.
Nery
J.R.
Lee
L.
Ye
Z.
Ngo
Q.M.
, et al.  . 
Human DNA methylomes at base resolution show widespread epigenomic differences
Nature
 , 
2009
, vol. 
462
 (pg. 
315
-
322
)
79
Meissner
A.
Mikkelsen
T.S.
Gu
H.
Wernig
M.
Hanna
J.
Sivachenko
A.
Zhang
X.
Bernstein
B.E.
Nusbaum
C.
Jaffe
D.B.
, et al.  . 
Genome-scale DNA methylation maps of pluripotent and differentiated cells
Nature
 , 
2008
, vol. 
454
 (pg. 
766
-
770
)
80
Ficz
G.
Branco
M.R.
Seisenberger
S.
Santos
F.
Krueger
F.
Hore
T.A.
Marques
C.J.
Andrews
S.
Reik
W.
Dynamic regulation of 5-hydroxymethylcytosine in mouse ES cells and during differentiation
Nature
 , 
2011
, vol. 
473
 (pg. 
398
-
402
)
81
Williams
K.
Christensen
J.
Pedersen
M.T.
Johansen
J.V.
Cloos
P.A.
Rappsilber
J.
Helin
K.
TET1 and hydroxymethylcytosine in transcription and DNA methylation fidelity
Nature
 , 
2011
, vol. 
473
 (pg. 
343
-
348
)
82
Cimmino
L.
Abdel-Wahab
O.
Levine
R.L.
Aifantis
I.
TET family proteins and their role in stem cell differentiation and transformation
Cell Stem Cell
 , 
2011
, vol. 
9
 (pg. 
193
-
204
)
83
Wu
H.
D'Alessio
A.C.
Ito
S.
Wang
Z.
Cui
K.
Zhao
K.
Sun
Y.E.
Zhang
Y.
Genome-wide analysis of 5-hydroxymethylcytosine distribution reveals its dual function in transcriptional regulation in mouse embryonic stem cells
Genes Dev.
 , 
2011
, vol. 
25
 (pg. 
679
-
684
)
84
Bird
A.
DNA methylation patterns and epigenetic memory
Genes Dev.
 , 
2002
, vol. 
16
 (pg. 
6
-
21
)
85
Szulwach
K.E.
Li
X.
Li
Y.
Song
C.X.
Han
J.W.
Kim
S.
Namburi
S.
Hermetz
K.
Kim
J.J.
Rudd
M.K.
, et al.  . 
Integrating 5-hydroxymethylcytosine into the epigenomic landscape of human embryonic stem cells
PLoS Genet.
 , 
2011
, vol. 
7
 pg. 
e1002154
 
86
Jin
S.G.
Wu
X.
Li
A.X.
Pfeifer
G.P.
Genomic mapping of 5-hydroxymethylcytosine in the human brain
Nucleic Acids Res.
 , 
2011
, vol. 
39
 (pg. 
5015
-
5024
)
87
Wu
H.
D'Alessio
A.C.
Ito
S.
Xia
K.
Wang
Z.
Cui
K.
Zhao
K.
Sun
Y.E.
Zhang
Y.
Dual functions of Tet1 in transcriptional regulation in mouse embryonic stem cells
Nature
 , 
2011
, vol. 
473
 (pg. 
389
-
393
)
88
Wu
H.
Zhang
Y.
Mechanisms and functions of Tet protein-mediated 5-methylcytosine oxidation
Genes Dev.
 , 
2011
, vol. 
25
 (pg. 
2436
-
2452
)
89
Xu
Y.
Wu
F.
Tan
L.
Kong
L.
Xiong
L.
Deng
J.
Barbera
A.J.
Zheng
L.
Zhang
H.
Huang
S.
, et al.  . 
Genome-wide regulation of 5hmC, 5mC, and gene expression by Tet1 hydroxylase in mouse embryonic stem cells
Mol. Cell
 , 
2011
, vol. 
42
 (pg. 
451
-
464
)
90
Lian
C.G.
Xu
Y.
Ceol
C.
Wu
F.
Larson
A.
Dresser
K.
Xu
W.
Tan
L.
Hu
Y.
Zhan
Q.
, et al.  . 
Loss of 5-hydroxymethylcytosine is an epigenetic hallmark of melanoma
Cell
 , 
2012
, vol. 
150
 (pg. 
1135
-
1146
)
91
Wang
J.Q.
Chen
Q.
Wang
X.
Wang
Q.C.
Wang
Y.
Cheng
H.P.
Guo
C.
Sun
Q.
Tang
T.S.
Dysregulation of mitochondrial calcium signaling and superoxide flashes cause mitochondrial genomic DNA damage in Huntington's disease
J. Biol. Chem.
 , 
2013
, vol. 
288
 (pg. 
3070
-
3084
)
92
Lie
D.C.
Colamarino
S.A.
Song
H.J.
Desire
L.
Mira
H.
Consiglio
A.
Lein
E.S.
Jessberger
S.
Lansford
H.
Dearie
A.R.
, et al.  . 
Wnt signalling regulates adult hippocampal neurogenesis
Nature
 , 
2005
, vol. 
437
 (pg. 
1370
-
1375
)
93
Adachi
K.
Mirzadeh
Z.
Sakaguchi
M.
Yamashita
T.
Nikolcheva
T.
Gotoh
Y.
Peltz
G.
Gong
L.
Kawase
T.
Alvarez-Buylla
A.
, et al.  . 
Beta-catenin signaling promotes proliferation of progenitor cells in the adult mouse subventricular zone
Stem Cells
 , 
2007
, vol. 
25
 (pg. 
2827
-
2836
)
94
Qu
Q.
Sun
G.
Li
W.
Yang
S.
Ye
P.
Zhao
C.
Yu
R.T.
Gage
F.H.
Evans
R.M.
Shi
Y.
Orphan nuclear receptor TLX activates Wnt/beta-catenin signalling to stimulate neural stem cell proliferation and self-renewal
Nat. Cell. Biol.
 , 
2010
, vol. 
12
 (pg. 
31
-
40
sup pp 31–39
95
Simpson
J.M.
Gil-Mohapel
J.
Pouladi
M.A.
Ghilan
M.
Xie
Y.
Hayden
M.R.
Christie
B.R.
Altered adult hippocampal neurogenesis in the YAC128 transgenic mouse model of Huntington disease
Neurobiol. Dis.
 , 
2011
, vol. 
41
 (pg. 
249
-
260
)
96
Gil-Mohapel
J.
Simpson
J.M.
Ghilan
M.
Christie
B.R.
Neurogenesis in Huntington's disease: can studying adult neurogenesis lead to the development of new therapeutic strategies?
Brain Res.
 , 
2011
, vol. 
1406
 (pg. 
84
-
105
)
97
Grote
H.E.
Bull
N.D.
Howard
M.L.
van Dellen
A.
Blakemore
C.
Bartlett
P.F.
Hannan
A.J.
Cognitive disorders and neurogenesis deficits in Huntington's disease mice are rescued by fluoxetine
Eur. J. Neurosci.
 , 
2005
, vol. 
22
 (pg. 
2081
-
2088
)
98
Cho
S.R.
Benraiss
A.
Chmielnicki
E.
Samdani
A.
Economides
A.
Goldman
S.A.
Induction of neostriatal neurogenesis slows disease progression in a transgenic murine model of Huntington disease
J. Clin. Invest.
 , 
2007
, vol. 
117
 (pg. 
2889
-
2902
)
99
Duan
W.
Peng
Q.
Masuda
N.
Ford
E.
Tryggestad
E.
Ladenheim
B.
Zhao
M.
Cadet
J.L.
Wong
J.
Ross
C.A.
Sertraline slows disease progression and increases neurogenesis in N171–82Q mouse model of Huntington's disease
Neurobiol. Dis.
 , 
2008
, vol. 
30
 (pg. 
312
-
322
)
100
Peng
Q.
Masuda
N.
Jiang
M.
Li
Q.
Zhao
M.
Ross
C.A.
Duan
W.
The antidepressant sertraline improves the phenotype, promotes neurogenesis and increases BDNF levels in the R6/2 Huntington's disease mouse model
Exp. Neurol.
 , 
2008
, vol. 
210
 (pg. 
154
-
163
)
101
Tang
T.S.
Tu
H.
Wang
Z.
Bezprozvanny
I.
Modulation of type 1 inositol (1,4,5)-trisphosphate receptor function by protein kinase a and protein phosphatase 1α
J. Neurosci.
 , 
2003
, vol. 
23
 (pg. 
403
-
415
)

Author notes

Equal contribution.