At birth, the mammalian brain contains tens of billions of neurons. Although the number does not increase much as the animal grows, there are many dramatic changes to their size and structure. These changes allow the neurons to communicate with one another, develop into networks, and learn the tasks of the adult brain. One way that these changes occur is by the accumulation of chemical marks on each neuron’s DNA that help dictate which genes switch on, and which turn off.

One of the most common ways that DNA can be marked is through the addition of a chemical group called a methyl group to one of the four DNA bases, cytosine. This process is called methylation. When methylation occurs, cytosine becomes 5-methylcytosine, or 5mC for short.

In 2009, researchers found another modification present in the DNA in the brain: 5-hydroxymethylcytosine, or 5hmC. This modification appears when a group of proteins called the Tet hydroxylases turn 5mC into 5hmC. Converting 5mC to 5hmC normally helps cells remove marks on their DNA before they divide and expand. This is important because the newly generated cells need to be able to accumulate their own methylation marks to perform their roles properly. However, neurons in the brain accumulate 5hmC after birth, when the cells are no longer dividing, indicating that 5hmC may be required for the neurons to mature.

Stoyanova et al. set out to determine whether mouse neurons need 5hmC to get their adult characteristics by tracking the chemical changes that occur in DNA from birth to adulthood. Some of the mice they tested produced 5hmC normally, while others lacked the genes necessary to make the Tet proteins in a specific class of neurons, preventing them from converting 5mC to 5hmC as they differentiate. The results reveal that neurons do not mature properly if 5hmC is not produced continuously following the first week of life. This is because neurons need to have the right genes switched on and off to differentiate correctly, and this only happens when 5hmC accumulates in some genes, while 5hmC and 5mC are removed from others. The data highlight the role of the Tet proteins, which convert 5mC into 5hmC, in preparing the marks for removal and demonstrate that active removal of these marks is essential for neuronal differentiation.

Given the role of 5hmC in the development of neurons, it is possible that problems in this system could contribute to brain disorders. Further studies aimed at understanding how cells control 5hmC levels could lead to new ways to improve brain health. Research has also shown that if dividing cells lose the ability to make 5hmC, they can become cancerous. Future work could explain more about how and why this happens.

Development of the mammalian brain requires generation of hundreds of millions of neuronal progenitors that differentiate into distinct cell types with refined functional properties. Although morphological and physiological maturation of most neurons occurs between mid-gestation and a few months or years after birth, the vast majority of CNS neurons must maintain a stable differentiated state for the life of the organism and remain sufficiently plastic to participate in novel behaviors. Studies of signaling molecules and transcriptional programs have identified many mechanisms that orchestrate critical steps in neurogenesis, cell type diversification, neuronal migration, axonal pathfinding, and differentiation. Although it has been established that epigenetic regulatory mechanisms are critical for proper development of all cell types, our knowledge of the precise roles of these mechanisms in neuronal differentiation, function, and vitality remains rudimentary.

5-Hydroxymethylcytosine (5hmC) is produced from 5-methylcytosine (5mC) by the Ten-eleven translocation dioxygenases (Tet1, Tet2, Tet3) (Iyer et al., 2009; Tahiliani et al., 2009). It is present at approximately 10-fold higher levels in neurons than peripheral cell types (Globisch et al., 2010; Kriaucionis and Heintz, 2009), and its distribution across the genome of adult neurons is cell specific and correlated with active gene expression (Mellén et al., 2017; Mellén et al., 2012; Szulwach et al., 2011). The initial discovery that 5hmC accumulates within active gene bodies, coupled with the discovery that MeCP2 can bind probes containing 5hmC with high affinity, led to the proposal that MeCP2 binding within active genes facilitates their expression (Mellén et al., 2012). Further studies demonstrating that MeCP2 binds 5hmC at high affinity in non-CG dinucleotides but does not bind to 5hmCG overturned this model (Ayata, 2013; Gabel et al., 2015; Mellén et al., 2017) by revealing that accumulation of 5hmCG and the depletion of 5hmCH in gene bodies is correlated with less MeCP2 binding and increased expression (Ayata, 2013; Gabel et al., 2015; Mellén et al., 2017). Imaging of the binding and diffusion of single MeCP2 molecules in living neurons lacking Dnmt3a or Tet1, Tet2, and Tet3 demonstrate that its binding is exquisitely sensitive to the levels of both 5mC and 5hmC (Piccolo et al., 2019). Given these data and the sensitivity of neuronal function to MeCP2 gene dosage (Chahrour et al., 2008; Nan and Bird, 2001), relief of the repressive functions of MeCP2 through Tet-mediated conversion of high-affinity 5mCG-binding sites to low-affinity 5hmCG sites, which we have referred to as functional demethylation (Mellén et al., 2017), provides an important mechanism for modulation of chromatin structure and transcription.

In dividing cells, 5hmC serves as an intermediate in DNA demethylation because maintenance DNA methyltransferases do not recognize hemi-hydroxymethylated cytosines in order to reestablish methylation (Wu and Zhang, 2017). Consequently, 5hmC is lost passively and replaced by C due to replicative dilution. Loss of function studies in mouse embryonic stem cells (ESCs) (Dawlaty et al., 2013) and lymphocyte lineages Lio and Rao, 2019 have demonstrated that the role of Tet-mediated replicative DNA demethylation is to provide full accessibility to regulatory regions necessary for expression of genes required for differentiation. For example, at the activation-induced deaminase (AID) locus in B cells, Tet activity is required for demethylation and activation of enhancer regions that modulate AID expression to enable class switch recombination (Lio et al., 2019). Tet-mediated replication-dependent passive demethylation is thought to be common in many dividing cell types, including progenitor cells in the developing nervous system (MacArthur and Dawlaty, 2021).

Most neurons exit the cell cycle during mid-gestation and remain relatively simple and undifferentiated until birth. Following parturition, they initiate an elaborate program of differentiation that includes dramatic increases in size, morphological complexity, and connectivity. Since these neurons are postmitotic and remain so throughout life, removal of the repressive effects of 5mC by passive demethylation cannot occur. However, in addition to functional demethylation and passive demethylation, a third pathway for DNA demethylation, often referred to as active demethylation, has been proposed based on the finding that 5hmC can be further oxidized by Tet proteins to produce 5-formylcytosine (5fC) and 5-carboxylcytosine (5caC). Removal of 5fC and 5caC by thymine DNA glycosylase (TDG)-dependent base excision repair (BER) provides another mechanism for 5hmC-mediated DNA demethylation (He et al., 2011). Clear evidence that this pathway can operate in cultured cells has been presented, and detailed biochemical studies have delineated the mechanisms operating in TDG-BER DNA demethylation (Weber et al., 2016). Active DNA demethylation is ideally suited for remodeling DNA methylation in postmitotic neurons.

Evidence that continued accumulation of 5hmC is required in differentiating neurons and that it can participate in active demethylation in vivo is beginning to emerge. In studies of cerebellar granule cell development in ESC-derived GC cultures, primary GC cultures, and slice preparations, manipulations of Tet activity and 5hmC levels provide strong evidence that 5hmC is required for expression of axon guidance and ion channel genes, and that proper development of the GC dendritic arbor requires 5hmC (Zhu et al., 2016). Demethylation of 5hmC-GFP transfected DNA fragments retrieved after several days in culture display reduced 5hmC levels at several sites as assessed by bisulfite sequencing (BSSeq) in HEK 293 cells and primary hippocampal neurons (Guo et al., 2011). Changes in methylation levels at GGCC sites in the genome of hippocampal dentate gyrus neurons in response to electroconvulsive shock have been documented using the methylation-sensitive cut counting method that employs the methylation or hydroxymethylation-sensitive restriction enzyme HpaII and its methylation-insensitive isoschizomer Msp1 (Guo et al., 2011). Global measurements of 5mC and 5hmC by mass spectroscopy in the hippocampus have shown that both decrease in response to the induction of seizure (Kaas et al., 2013). And in germline Tet1 knockout mice, analysis of the Npas4 and c-Fos promoters has shown that Tet1 is required for their proper regulation in both the cerebral cortex and hippocampus (Rudenko et al., 2013). Despite these observations, our knowledge of the roles of continued 5hmC accumulation in active DNA demethylation and postmitotic differentiation remains rudimentary.

We report here single nucleotide resolution studies of DNA methylation and hydroxymethylation, transcription, chromatin accessibility, and H3K4me3 and H3K27me3 histone marks in postmitotic, differentiating Purkinje cells (PCs). As PCs transition from relatively small, multipolar immature cells to fully elaborated large neurons with complex dendritic arbors and hundreds of thousands of synapses (McKay and Turner, 2005), 5hmC continues to accumulate, and DNA methylation and hydroxymethylation are reconfigured as epigenetic and transcriptional programs progress. Our data confirm previous studies of the relationships between transcription, DNA methylation, DNA hydroxymethylation, and chromatin organization (Lister et al., 2013; Mellén et al., 2017; Tsagaratou et al., 2017; Tsagaratou et al., 2014), and they reveal several novel modes of transcriptional activation and repression. Notably, we identify a class of developmentally induced PC-specific genes that are highly expressed and lose both 5mC and 5hmC in the final stages of PC differentiation.

We report also studies of newly generated Tet1, Tet2, Tet3 PC-specific triple knockout (Pcp2TetTKO) mouse lines in which recombination is activated in the first postnatal week. These data demonstrate that postmitotic transcriptional and epigenetic maturation in PCs, including transcription of many ion channels and active demethylation of late expressed genes, requires continued oxidation of 5mC to 5hmC. Taken together, our data demonstrate that active demethylation occurs in select genes in postmitotic neurons, and that 5hmC plays an essential role in refining the transcriptional and epigenetic status of PCs during the final stages of differentiation.

In mice, PC progenitors complete their final cell cycle between e10.5 and e13.5 within the ventricular zone of the developing cerebellar anlage (Leto et al., 2016). As they exit cell cycle and commit to a PC fate, they express the transcription factors Lhx1/Lhx5 (Chizhikov et al., 2006; Morales and Hatten, 2006). To characterize 5hmC accumulation, transcription and chromatin organization during their postmitotic development, we chose to analyze PCs at the start of their accelerated differentiation (P0), during their rapid phase of morphological development (P7) and as fully mature, differentiated neurons (adult) (Figure 1A–B). We employed fluorescence-activated nuclear sorting (FANS) to purify Itpr1-positive PC nuclei in two biological replicates (Figure 1C–E, Figure S1A) (Xu et al., 2018). Since PCs are extremely rare and difficult to isolate, we optimized low input protocols for genomic profiling. We used ~20,000 nuclei for transcriptional profiling, ~200,000 nuclei for BSSeq and oxidative bisulfite sequencing (OxBSSeq), 25,000 nuclei for the assay for transposase-accessible chromatin sequencing (ATACSeq), and 25,000 nuclei for H3K27me3 and H3K4me3 chromatin immunoprecipitation sequencing (ChIPSeq) (Figure 1E–F). Analysis of biological replicates for each of the assays revealed that the sorted nuclei expressed markers of Purkinje neurons, and did not express genes marking the two most abundant cerebellar cell types, granule cells and glia (Figure S1B). The Pearson correlation coefficients between each replicate at all timepoints and techniques (RNASeq, OxBSSeq, ATACSeq, and ChIPSeq) were high (over 0.9), establishing that our data are reproducible and of very high quality (Figure 1—figure supplement 1C-F). Bisulfite conversion and oxidation rates for the OxBSSeq datasets were within the expected range (Figure 1—figure supplement 1G).

Figure 1 with 1 supplement see all

Download asset Open asset

It is important to note that we chose to focus on OxBSSeq (Booth et al., 2013) as a definitive technology for analysis of DNA methylation because, in contrast to BSSeq alone, it allows us to assess the contributions of both 5mC and 5hmC to neuronal development. Given the distinct functions of these two modifications, analysis of their separate contributions to epigenetic regulation of the genome is more informative than BSSeq alone. For example, in many genes, 5mCG is depleted and 5hmCG accumulates in the gene body as its expression increases between P0 and adult. This is clearly evident for the gene encoding the cyclic GMP-dependent protein kinase (Prkg1), whose expression in PCs has been shown to be essential for long-term depression (Feil et al., 2003). In this case, the OxBSSeq data reveal a transition from 5mCG to 5hmCG that cannot be detected in the BSSeq data as the gene is activated between P0 and adult (Figure 1—figure supplement 1H). These data illustrate that a precise evaluation of DNA methylation status in relation to other programs unfolding during development for this gene and many others is advanced significantly by inclusion of OxBSSeq data.

Transcriptional programs altered during PC differentiation

PC progenitors complete their final divisions in the cerebellar primordium between e11 and e13 (Butts et al., 2014) and remain as a multilayered, simple migrating cell population until birth (P0). In the first postnatal week, they organize into a monolayer as the cerebellum enlarges and begin the transition from a multipolar primitive neuronal morphology to one of the largest neurons in the brain with a characteristic, highly elaborate planar dendritic arbor. As they begin the second postnatal week (P7), PCs undergo an important developmental transition that includes refinement of their climbing fiber input, formation of many thousands of parallel fiber synapses, and myelination of their axons (Leto et al., 2016; McKay and Turner, 2005). The tremendous increase in synaptogenesis and connectivity continues until PCs attain their mature morphology and functions at 3–4 weeks of age. As these programs unfold, there is a global decrease of 5mCG and global increase of 5hmCG (Figure 2—figure supplement 1A).

To identify genes whose transcription is stable and those that are dynamically regulated during PC differentiation, we conducted differential expression analysis between P0 and adult Purkinje neurons, filtering for significance of p < 0.01 and log2 fold change of >2 in either direction (Figure 2A–B, Supplementary file 1). Using these criteria, there were 922 developmentally repressed genes (down-regulated in adult) enriched in categories involved in cell signaling and axon guidance pathways (e.g. Grid1, Pde1c; Figure 1F). We identified 432 genes with increased expression as PCs differentiate encoding proteins known to be important for the mature functions of PCs including calcium ion buffering and transport, regulation of the inositol triphosphate signaling pathway, and RNA splicing. As expected, similar analysis comparing P0 and P7 or P7 and adult data revealed genes overlapping with those identified in the overall P0 and adult comparative data (Figure 2—figure supplement 1B-D), although additional categories of RNA metabolism, synaptic signaling, and ion transport are enriched in the P7 to adult analysis (Figure 2—figure supplement 1C). As anticipated, the P7 profiles we have included in our analysis have been essential in assessing both the developmental course of epigenetic events studied here and the consequences of loss of 5hmC in PCs (see below).

Figure 2 with 1 supplement see all

Download asset Open asset

Loss of 5mC and 5hmC in active genes during PC differentiation

Our data reveal additional, surprising features that have not been documented in postmitotic cells. Thus, in a small subset of highly expressed genes (e.g. Cep76, Itpr1, Mtss1) there is a profound loss of both 5hmCG and 5mCG during the terminal, postmitotic stage of PC differentiation (Figures 1F and 3). To investigate the apparent demethylation over this class of genes, we computationally identified DNA methylation valleys (DMVs) as previously described (Jeong et al., 2014; Mo et al., 2015; Xie et al., 2013). In brief, DMVs were characterized by filtering undermethylated regions (UMRs) for length (>5 kb) and merging any regions within 1 kb (Burger et al., 2013; Figure 3—figure supplement 1A-B, Supplementary file 2). This revealed multiple regions longer than 15 kb, an unusual finding given that the average length of DMVs is ~5 kb (Figure 3—figure supplement 1D-E; Jeong et al., 2014). As expected from previous studies, a small number of these genes are inactive, fully demethylated, inaccessible, and covered by high levels of H3K27me3 repressive marks (Figure 3A, Foxd1, Figure 3—figure supplement 1C). These genes acquire their characteristics early in development, and their epigenetic features are stable.

Figure 3 with 1 supplement see all

Download asset Open asset

The majority of very large DMVs in PCs, however, arise during differentiation through loss of 5mC and 5hmC over active, highly expressed genes as PCs mature (Figure 3, Itpr1, Mtss1). In these genes, the timing of changes in DNA methylation, chromatin accessibility, and dynamic histone marks varies from gene to gene, but the overall progression of events associated with their expression is similar (Figure 3). At birth, these genes are actively transcribed (Figure 3A), 5hmCG is enriched in the gene body (Figure 3B) and 5mCG is depleted (Figure 3C), their promoters are accessible (Figure 3D), and H3K4me3 is present over their promoters (Figure 3E). These epigenetic properties resemble moderately expressed genes in many cell types. As differentiation proceeds and transcription increases, there is a gradual decrease in both 5mCG, 5hmCG, and H3K27me3 histone marks, and ATAC accessibility and H4K4me3 marks spread from the promoter into the gene body. To further support these data, we have included analysis of independent BSSeq experiments that detect the combined levels of 5mC and 5hmC (Figure 3A, BSSeq). These data also reveal that hypo-DMRs (differentially methylated regions) present at P0 expand and fuse as PCs lose both methylation and hydroxymethylation to form the very large DMRs that are characteristics of these genes (Figure 3A, hypo-DMR track). As shown in Figure 3B–E and Figure 3—figure supplement 1I-K, quantitation of these features demonstrates that that loss of DNA methylation and hydroxymethylation occurs over the entire gene body for these genes as H3K4me3 activating histone marks accumulate and ATAC accessibility increases.

Interestingly, many of the active genes associated with broad DMVs are PC-specific and highly expressed (Figure 3—figure supplement 1F). GO analysis indicates that they are involved in the inositol triphosphate/calcium signaling pathways, and a subset is associated with ataxia and autism (Figure 3—figure supplement 1H). They do not accumulate modified cytosines in CpH context (Figure 3—figure supplement 1G). These data provide strong evidence that DNA demethylation can occur in postmitotic neurons, and that it is enhanced in a specific class of genes that are very highly expressed and functionally important. Although our mass spectroscopic analysis of genomic DNA from differentiating PCs (Figure 3—figure supplement 1L) failed to detect 5fC or 5caC, their involvement as transient intermediates in the loss of 5mC and 5hmC cannot be ruled out because the small fraction of the genome covered by this gene class may preclude their detection (Guo et al., 2011; He et al., 2011; Ito et al., 2011).

PC-specific Tet1, Tet2, Tet3 triple knockout mouse lines

Despite convincing evidence that Tet-mediated active DNA demethylation can occur in mouse ESCs through the TDG-BER pathway (He et al., 2011; Weber et al., 2016), evidence that this can occur in vivo has been difficult to obtain because such a proof requires loss or complete inhibition of all three Tet oxidases in a single cell type after cells have exited the cell cycle permanently. To provide this evidence, we generated PC-specific Tet1/Tet2/Tet3 triple knockout (Pcp2TetTKO) mouse lines (Figure 4A). We chose to drive Cre recombinase expression using an engineered Pcp2 BAC employed previously to generate accurate Cre driver lines (Gong et al., 2007) because the onset of expression of the Pcp2 gene and the corresponding BAC vector occurs approximately 1 week after birth as the cerebellum enters its terminal phase of development and maturation. The Pcp2Cre BAC was introduced by pronuclear injection into ova from females carrying floxed alleles of all three Tet genes (Figure 4—figure supplement 1A-B) yielding several founder lines. PCR analysis of the floxed regions of each Tet gene in purified PC genomic DNA, and RNASeq analysis confirmed the deletion of exons from all three TET proteins (Figure 4—figure supplement 1C-D). The lines chosen for analysis displayed no gross motor phenotype as assessed by rotarod performance, survival was normal, and recombination activity was specific to PCs (Figure 4B–D, Figure 4—figure supplement 2A-D). The deletion of the Tet proteins did not affect the expression of Itpr1 significantly, and we confirmed that it could still be used for purification of PCs (Figure 4—figure supplement 1E-F). Evaluation of the quality control metrics of the datasets showed strong correlation coefficients and enrichment for Purkinje markers (Figure 4—figure supplement 1G-H). Cre recombinase activity, and thus recombination in the Tet genes, began at approximately P7 more than 2 weeks after PCs have completed their last division (Figure 4—figure supplement 2A-D). These lines, therefore, allow assessment of the consequences of loss of Tet activity during the rapid phase of somatic and dendritic growth in postmitotic, differentiating Purkinje neurons.

Figure 4 with 2 supplements see all

Download asset Open asset

Since the discoveries that 5hmC is present at high levels in mammalian neuronal genomes (Kriaucionis and Heintz, 2009) and that it is produced from 5mC by the Tet oxidases (Tahiliani et al., 2009), its possible roles as a stable epigenetic mark and as an intermediate in DNA demethylation have been intensively investigated (Wu and Zhang, 2017). The present study adds to a growing body of work demonstrating that 5hmC plays a critical role in development and function of the nervous system. A central finding of this study is that continued Tet activity and 5hmC accumulation during PC differentiation is necessary for acquisition of their refined transcriptional and epigenetic properties. It is noteworthy that these phenotypes are evident despite loss of Tet function in the Pcp2TetTKO late in PC differentiation. It seems likely based on these findings and the gradual accumulation of 5hmC evident in our data that there is a continual requirement for 5hmC as a driver for epigenetic remodeling in most differentiating, postmitotic neurons. Although studies of neural stem cells (Li et al., 2015; Xu et al., 2012), cerebellar granule cells (Zhu et al., 2016), and embryonic brain development have suggested that 5hmC plays a significant role in neural progenitors and developing granule cells, a continual requirement for 5hmC may be particularly important for PCs, pyramidal cells, and other long range projection neurons whose differentiation program includes postmitotic development of sophisticated morphological features and abundant synapses.

A second principle that has emerged from our studies is that Tet-mediated active DNA demethylation occurs in postmitotic neurons. Our finding that loss of both 5mC and 5hmC in a specific subset of highly expressed gene bodies and ATAC accessible regulatory sites requires Tet oxidase activity provides critical in vivo evidence supporting studies in vitro (Weber et al., 2016) and in cultured cells (He et al., 2011) that have defined this pathway. For some genes (Gpr63, Grid2) the progression toward loss of modified C is easily appreciated. In these cases, conversion of 5mCG to 5hmCG over the gene body is clearly evident between P0 and P7, and continued Tet activity is required to remove both residual 5mCG and 5hmCG to result in DNA demethylation in the adult. Although we have been unable to detect 5fC and 5caC as transient intermediates in this process (Figure 3—figure supplement 1L), this is not surprising given the very small fraction of the PC genome that undergoes active DNA demethylation and the small amounts of genomic DNA that can be obtained from this very rare cell population. Identification of additional components of the of the DNA demethylation pathway that is occurring in PCs will require further genetic studies disrupting candidate genes in the TDG-BER pathway. Since important non-cell autonomous events are required for PC differentiation (Leto et al., 2016), we believe that additional PC-specific knockouts that are restricted to postmitotic cells will be most informative.

An important question arising from these findings is which cells in the nervous system require Tet-dependent active DNA demethylation to complete their developmental programs. Given these data, it seems likely other large neurons with complex and prolonged postmitotic differentiation programs share this requirement. Furthermore, reports that remodeling of DNA methylation may occur in response to a wide variety of stimuli in adult mice (Kaas et al., 2013; Rudenko et al., 2013) suggest that additional studies of Pcp2TetTKO and other precisely constructed mouse models will help elucidate possible functions of 5hmC in neuronal plasticity.

A third advance reported here is the collection of very high-resolution data to understand details of the relationships between DNA methylation and hydroxymethylation, chromatin accessibility, and chromatin organization in a single complex neuronal cell type as it develops from a primitive neuronal precursor to a fully articulated adult neuron. These data have highlighted several features of epigenetic development that remain to be explored including the definition of two distinct relationships between non-CG DNA methylation/hydroxymethylation and transcriptional repression, and the mechanism for selection of specific genes as substrates for active DNA demethylation. Taken together, our data demonstrate that disruption of continued 5hmC formation in the Pcp2TetTKO postmitotic PCs leads to altered transcription of late expressed genes, including a set of ion channels that may be responsible for refining their adult electrophysiological properties.

Although our data reinforce prior studies of the general relationships between DNA methylation, transcription, and chromatin structure, we note that there are many exceptions that require further investigation. One intriguing example is the very different relationships between genes and de novo methylation in CG and non-CG contexts (Lister et al., 2013). In PCs, as in other neurons, substantial enrichment of 5hmCG relative to its general accumulation genome wide occurs in active transcription units, suggesting a mechanism that targets Tet activity to these regions (Figure 2C). While accumulation of 5mCH and 5hmCH can occur within a repressed transcription unit with very nice discrimination between the gene and its surroundings (e.g. Pitpnc1), there are many regions of 5mCH and 5hmCH accumulation that either partially overlap with a gene, or that are present in intergenic regions (e.g. Myo5b/Scarna17/Lipg locus). In all of these regions, however, 5hmC accumulation directly reflects 5mCH levels suggesting 5hmCH localization reflects the pattern of de novo DNA methylation rather than targeted Tet activity. We hope that continued analysis of these datasets will provide a stimulus for additional studies of the mechanisms determining the relative distributions of these important events.

Pcp2TetTKO mice will be made available with a Material Transfer Agreement (MTA). Sequencing data is available on NCBI Gene Expression Omnibus under accession number GSE166423. Code for data analysis is available at https://github.com/estoyanova/EStoyanova_eLife_2021 copy archived at https://archive.softwareheritage.org/swh:1:rev:4c2fa6c102e4f868fc91e0dcb0b8c7155b8f8712.

1. 1. Stoyanova E
   2. Riad M
   3. Rao A
   4. Heintz N

   (2021) NCBI Gene Expression Omnibus

   ID GSE166423. 5-hydroxymethylcytosine is required for terminal differentiation of Purkinje neurons.

   https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE166423

1. Thesis

   1. Ayata P

   (2013) Decoding 5HMC as an Active Chromatin Mark in the Brain and Its Link to Rett Syndrome

   Student Theses and Dissertations.

   https://digitalcommons.rockefeller.edu/student_theses_and_dissertations/238/

   • Google Scholar
2. 1. Booth MJ
   2. Ost TWB
   3. Beraldi D
   4. Bell NM
   5. Branco MR
   6. Reik W
   7. Balasubramanian S

   (2013) Oxidative bisulfite sequencing of 5-methylcytosine and 5-hydroxymethylcytosine

   Nature Protocols 8:1841–1851.

   https://doi.org/10.1038/nprot.2013.115

   • PubMed
   • Google Scholar
3. 1. Buenrostro JD
   2. Giresi PG
   3. Zaba LC
   4. Chang HY
   5. Greenleaf WJ

   (2013) Transposition of native chromatin for fast and sensitive epigenomic profiling of open chromatin, DNA-binding proteins and nucleosome position

   Nature Methods 10:1213–1218.

   https://doi.org/10.1038/nmeth.2688

   • PubMed
   • Google Scholar
4. 1. Burger L
   2. Gaidatzis D
   3. Schübeler D
   4. Stadler MB

   (2013) Identification of active regulatory regions from DNA methylation data

   Nucleic Acids Research 41:e155.

   https://doi.org/10.1093/nar/gkt599

   • PubMed
   • Google Scholar
5. 1. Butts T
   2. Green MJ
   3. Wingate RJT

   (2014) Development of the cerebellum: simple steps to make a “little brain”

   Development 141:4031–4041.

   https://doi.org/10.1242/dev.106559

   • PubMed
   • Google Scholar
6. 1. Carroll TS
   2. Liang Z
   3. Salama R
   4. Stark R
   5. de Santiago I

   (2014) Impact of artifact removal on ChIP quality metrics in ChIP-seq and ChIP-exo data

   Frontiers in Genetics 5:75.

   https://doi.org/10.3389/fgene.2014.00075

   • PubMed
   • Google Scholar
7. 1. Chahrour M
   2. Jung SY
   3. Shaw C
   4. Zhou X
   5. Wong STC
   6. Qin J
   7. Zoghbi HY

   (2008) MeCP2, a Key Contributor to Neurological Disease, Activates and Represses Transcription

   Science 320:1224–1229.

   https://doi.org/10.1126/science.1153252

   • PubMed
   • Google Scholar
8. 1. Chen X
   2. Shen Y
   3. Draper W
   4. Buenrostro JD
   5. Litzenburger U
   6. Cho SW
   7. Satpathy AT
   8. Carter AC
   9. Ghosh RP
   10. East-Seletsky A
   11. Doudna JA
   12. Greenleaf WJ
   13. Liphardt JT
   14. Chang HY

   (2016) ATAC-see reveals the accessible genome by transposase-mediated imaging and sequencing

   Nature Methods 13:1013–1020.

   https://doi.org/10.1038/nmeth.4031

   • PubMed
   • Google Scholar
9. 1. Chizhikov VV
   2. Lindgren AG
   3. Currle DS
   4. Rose MF
   5. Monuki ES
   6. Millen KJ

   (2006) The roof plate regulates cerebellar cell-type specification and proliferation

   Development 133:2793–2804.

   https://doi.org/10.1242/dev.02441

   • PubMed
   • Google Scholar
10. 1. Cochran JN
    2. Geier EG
    3. Bonham LW
    4. Newberry JS
    5. Amaral MD
    6. Thompson ML
    7. Lasseigne BN
    8. Karydas AM
    9. Roberson ED
    10. Cooper GM
    11. Rabinovici GD
    12. Miller BL
    13. Myers RM
    14. Yokoyama JS
    15. Alzheimer’s Disease Neuroimaging Initiative

    (2020) Non-coding and Loss-of-Function Coding Variants in TET2 are Associated with Multiple Neurodegenerative Diseases

    American Journal of Human Genetics 106:632–645.

    https://doi.org/10.1016/j.ajhg.2020.03.010

    • PubMed
    • Google Scholar
11. 1. Dawlaty MM
    2. Breiling A
    3. Le T
    4. Raddatz G
    5. Barrasa MI
    6. Cheng AW
    7. Gao Q
    8. Powell BE
    9. Li Z
    10. Xu M
    11. Faull KF
    12. Lyko F
    13. Jaenisch R

    (2013) Combined Deficiency of Tet1 and Tet2 Causes Epigenetic Abnormalities but Is Compatible with Postnatal Development

    Developmental Cell 24:310–323.

    https://doi.org/10.1016/j.devcel.2012.12.015

    • PubMed
    • Google Scholar
12. 1. Dobin A
    2. Davis CA
    3. Schlesinger F
    4. Drenkow J
    5. Zaleski C
    6. Jha S
    7. Batut P
    8. Chaisson M
    9. Gingeras TR

    (2013) STAR: ultrafast universal RNA-seq aligner

    Bioinformatics 29:15–21.

    https://doi.org/10.1093/bioinformatics/bts635

    • PubMed
    • Google Scholar
13. 1. Eden E
    2. Navon R
    3. Steinfeld I
    4. Lipson D
    5. Yakhini Z

    (2009) GOrilla: a tool for discovery and visualization of enriched GO terms in ranked gene lists

    BMC Bioinformatics 10:48.

    https://doi.org/10.1186/1471-2105-10-48

    • PubMed
    • Google Scholar
14. 1. ENCODE Project Consortium

    (2012) An integrated encyclopedia of DNA elements in the human genome

    Nature 489:57–74.

    https://doi.org/10.1038/nature11247

    • PubMed
    • Google Scholar
15. 1. Feil R
    2. Hartmann J
    3. Luo C
    4. Wolfsgruber W
    5. Schilling K
    6. Feil S
    7. Barski JJ
    8. Meyer M
    9. Konnerth A
    10. De Zeeuw CI
    11. Hofmann F

    (2003) Impairment of LTD and cerebellar learning by Purkinje cell-specific ablation of cGMP-dependent protein kinase I

    The Journal of Cell Biology 163:295–302.

    https://doi.org/10.1083/jcb.200306148

    • PubMed
    • Google Scholar
16. 1. Gabel HW
    2. Kinde B
    3. Stroud H
    4. Gilbert CS
    5. Harmin DA
    6. Kastan NR
    7. Hemberg M
    8. Ebert DH
    9. Greenberg ME

    (2015) Disruption of DNA-methylation-dependent long gene repression in Rett syndrome

    Nature 522:89–93.

    https://doi.org/10.1038/nature14319

    • PubMed
    • Google Scholar
17. 1. Globisch D
    2. Münzel M
    3. Müller M
    4. Michalakis S
    5. Wagner M
    6. Koch S
    7. Brückl T
    8. Biel M
    9. Carell T

    (2010) Tissue Distribution of 5-Hydroxymethylcytosine and Search for Active Demethylation Intermediates

    PLOS ONE 5:e15367.

    https://doi.org/10.1371/journal.pone.0015367

    • PubMed
    • Google Scholar
18. 1. Gong S
    2. Zheng C
    3. Doughty ML
    4. Losos K
    5. Didkovsky N
    6. Schambra UB
    7. Nowak NJ
    8. Joyner A
    9. Leblanc G
    10. Hatten ME
    11. Heintz N

    (2003) A gene expression atlas of the central nervous system based on bacterial artificial chromosomes

    Nature 425:917–925.

    https://doi.org/10.1038/nature02033

    • PubMed
    • Google Scholar
19. 1. Gong S
    2. Doughty M
    3. Harbaugh CR
    4. Cummins A
    5. Hatten ME
    6. Heintz N
    7. Gerfen CR

    (2007) Targeting Cre recombinase to specific neuron populations with bacterial artificial chromosome constructs

    The Journal of Neuroscience 27:9817–9823.

    https://doi.org/10.1523/JNEUROSCI.2707-07.2007

    • PubMed
    • Google Scholar
20. 1. Guo JU
    2. Su Y
    3. Zhong C
    4. Ming G
    5. Song H

    (2011) Hydroxylation of 5-Methylcytosine by TET1 Promotes Active DNA Demethylation in the Adult Brain

    Cell 145:423–434.

    https://doi.org/10.1016/j.cell.2011.03.022

    • PubMed
    • Google Scholar
21. 1. He YF
    2. Li BZ
    3. Li Z
    4. Liu P
    5. Wang Y
    6. Tang Q
    7. Ding J
    8. Jia Y
    9. Chen Z
    10. Li L
    11. Sun Y
    12. Li X
    13. Dai Q
    14. Song CX
    15. Zhang K
    16. He C
    17. Xu GL

    (2011) Tet-Mediated Formation of 5-Carboxylcytosine and Its Excision by TDG in Mammalian DNA

    Science 333:1303–1307.

    https://doi.org/10.1126/science.1210944

    • PubMed
    • Google Scholar
22. 1. Heinz S
    2. Benner C
    3. Spann N
    4. Bertolino E
    5. Lin YC
    6. Laslo P
    7. Cheng JX
    8. Murre C
    9. Singh H
    10. Glass CK

    (2010) Simple Combinations of Lineage-Determining Transcription Factors Prime cis-Regulatory Elements Required for Macrophage and B Cell Identities

    Molecular Cell 38:576–589.

    https://doi.org/10.1016/j.molcel.2010.05.004

    • PubMed
    • Google Scholar
23. 1. Huang DW
    2. Sherman BT
    3. Lempicki RA

    (2009) Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources

    Nature Protocols 4:44–57.

    https://doi.org/10.1038/nprot.2008.211

    • PubMed
    • Google Scholar
24. 1. Huber W
    2. Carey VJ
    3. Gentleman R
    4. Anders S
    5. Carlson M
    6. Carvalho BS
    7. Bravo HC
    8. Davis S
    9. Gatto L
    10. Girke T
    11. Gottardo R
    12. Hahne F
    13. Hansen KD
    14. Irizarry RA
    15. Lawrence M
    16. Love MI
    17. MacDonald J
    18. Obenchain V
    19. Oleś AK
    20. Pagès H
    21. Reyes A
    22. Shannon P
    23. Smyth GK
    24. Tenenbaum D
    25. Waldron L
    26. Morgan M

    (2015) Orchestrating high-throughput genomic analysis with Bioconductor

    Nature Methods 12:115–121.

    https://doi.org/10.1038/nmeth.3252

    • PubMed
    • Google Scholar
25. 1. Hyun K
    2. Jeon J
    3. Park K
    4. Kim J

    (2017) Writing, erasing and reading histone lysine methylations

    Experimental & Molecular Medicine 49:e324.

    https://doi.org/10.1038/emm.2017.11

    • PubMed
    • Google Scholar
26. 1. Ito S
    2. Shen L
    3. Dai Q
    4. Wu SC
    5. Collins LB
    6. Swenberg JA
    7. He C
    8. Zhang Y

    (2011) Tet Proteins Can Convert 5-Methylcytosine to 5-Formylcytosine and 5-Carboxylcytosine

    Science 333:1300–1303.

    https://doi.org/10.1126/science.1210597

    • PubMed
    • Google Scholar
27. 1. Iyer LM
    2. Tahiliani M
    3. Rao A
    4. Aravind L

    (2009) Prediction of novel families of enzymes involved in oxidative and other complex modifications of bases in nucleic acids

    Cell Cycle 8:1698–1710.

    https://doi.org/10.4161/cc.8.11.8580

    • PubMed
    • Google Scholar
28. 1. Jeong M
    2. Sun D
    3. Luo M
    4. Huang Y
    5. Challen GA
    6. Rodriguez B
    7. Zhang X
    8. Chavez L
    9. Wang H
    10. Hannah R
    11. Kim SB
    12. Yang L
    13. Ko M
    14. Chen R
    15. Göttgens B
    16. Lee JS
    17. Gunaratne P
    18. Godley LA
    19. Darlington GJ
    20. Rao A
    21. Li W
    22. Goodell MA

    (2014) Large conserved domains of low DNA methylation maintained by Dnmt3a

    Nature Genetics 46:17–23.

    https://doi.org/10.1038/ng.2836

    • PubMed
    • Google Scholar
29. 1. Kaas GA
    2. Zhong C
    3. Eason DE
    4. Ross DL
    5. Vachhani RV
    6. Ming G
    7. King JR
    8. Song H
    9. Sweatt JD

    (2013) TET1 Controls CNS 5-Methylcytosine Hydroxylation, Active DNA Demethylation

    Gene Transcription, and Memory Formation. Neuron 79:1086–1093.

    https://doi.org/10.1016/j.neuron.2013.08.032

    • Google Scholar
30. 1. Kriaucionis S
    2. Heintz N

    (2009) The Nuclear DNA Base 5-Hydroxymethylcytosine Is Present in Purkinje Neurons and the Brain

    Science 324:929–930.

    https://doi.org/10.1126/science.1169786

    • PubMed
    • Google Scholar
31. 1. Krueger F
    2. Andrews SR

    (2011) Bismark: a flexible aligner and methylation caller for Bisulfite-Seq applications

    Bioinformatics 27:1571–1572.

    https://doi.org/10.1093/bioinformatics/btr167

    • PubMed
    • Google Scholar
32. 1. Langmead B
    2. Salzberg SL

    (2012) Fast gapped-read alignment with Bowtie 2

    Nature Methods 9:357–359.

    https://doi.org/10.1038/nmeth.1923

    • PubMed
    • Google Scholar
33. 1. Leto K
    2. Arancillo M
    3. Becker EBE
    4. Buffo A
    5. Chiang C
    6. Ding B
    7. Dobyns WB
    8. Dusart I
    9. Haldipur P
    10. Hatten ME
    11. Hoshino M
    12. Joyner AL
    13. Kano M
    14. Kilpatrick DL
    15. Koibuchi N
    16. Marino S
    17. Martinez S
    18. Millen KJ
    19. Millner TO
    20. Miyata T
    21. Parmigiani E
    22. Schilling K
    23. Sekerková G
    24. Sillitoe RV
    25. Sotelo C
    26. Uesaka N
    27. Wefers A
    28. Wingate RJT
    29. Hawkes R

    (2016) Consensus Paper: Cerebellar Development

    Cerebellum 15:789–828.

    https://doi.org/10.1007/s12311-015-0724-2

    • PubMed
    • Google Scholar
34. 1. Li H
    2. Handsaker B
    3. Wysoker A
    4. Fennell T
    5. Ruan J
    6. Homer N
    7. Marth G
    8. Abecasis G
    9. Durbin R
    10. 1000 Genome Project Data Processing Subgroup

    (2009) The Sequence Alignment/Map format and SAMtools

    Bioinformatics 25:2078–2079.

    https://doi.org/10.1093/bioinformatics/btp352

    • PubMed
    • Google Scholar
35. 1. Li T
    2. Yang D
    3. Li J
    4. Tang Y
    5. Yang J
    6. Le W

    (2015) Critical role of Tet3 in neural progenitor cell maintenance and terminal differentiation

    Molecular Neurobiology 51:142–154.

    https://doi.org/10.1007/s12035-014-8734-5

    • PubMed
    • Google Scholar
36. 1. Li Y
    2. Zheng H
    3. Wang Q
    4. Zhou C
    5. Wei L
    6. Liu X
    7. Zhang W
    8. Zhang Y
    9. Du Z
    10. Wang X
    11. Xie W

    (2018) Genome-wide analyses reveal a role of Polycomb in promoting hypomethylation of DNA methylation valleys

    Genome Biology 19:18.

    https://doi.org/10.1186/s13059-018-1390-8

    • PubMed
    • Google Scholar
37. 1. Lio CW
    2. Zhang J
    3. González-Avalos E
    4. Hogan PG
    5. Chang X
    6. Rao A

    (2016) Tet2 and Tet3 cooperate with B-lineage transcription factors to regulate DNA modification and chromatin accessibility

    eLife 5:e18290.

    https://doi.org/10.7554/eLife.18290

    • PubMed
    • Google Scholar
38. 1. Lio CWJ
    2. Rao A

    (2019) TET Enzymes and 5hmC in Adaptive and Innate Immune Systems

    Frontiers in Immunology 10:210.

    https://doi.org/10.3389/fimmu.2019.00210

    • PubMed
    • Google Scholar
39. 1. Lio CWJ
    2. Shukla V
    3. Samaniego-Castruita D
    4. González-Avalos E
    5. Chakraborty A
    6. Yue X
    7. Schatz DG
    8. Ay F
    9. Rao A

    (2019) TET enzymes augment activation-induced deaminase (AID) expression via 5-hydroxymethylcytosine modifications at the Aicda superenhancer

    Science Immunology 4:eaau7523.

    https://doi.org/10.1126/sciimmunol.aau7523

    • PubMed
    • Google Scholar
40. 1. Lister R
    2. Mukamel EA
    3. Nery JR
    4. Urich M
    5. Puddifoot CA
    6. Johnson ND
    7. Lucero J
    8. Huang Y
    9. Dwork AJ
    10. Schultz MD
    11. Yu M
    12. Tonti-Filippini J
    13. Heyn H
    14. Hu S
    15. Wu JC
    16. Rao A
    17. Esteller M
    18. He C
    19. Haghighi FG
    20. Sejnowski TJ
    21. Behrens MM
    22. Ecker JR

    (2013) Global Epigenomic Reconfiguration During Mammalian Brain Development

    Science 341:1237905.

    https://doi.org/10.1126/science.1237905

    • PubMed
    • Google Scholar
41. 1. Love MI
    2. Huber W
    3. Anders S

    (2014) Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2

    Genome Biology 15:550.

    https://doi.org/10.1186/s13059-014-0550-8

    • PubMed
    • Google Scholar
42. 1. MacArthur IC
    2. Dawlaty MM

    (2021) TET Enzymes and 5-Hydroxymethylcytosine in Neural Progenitor Cell Biology and Neurodevelopment

    Frontiers in Cell and Developmental Biology 9:645335.

    https://doi.org/10.3389/fcell.2021.645335

    • PubMed
    • Google Scholar
43. 1. Marshall LL
    2. Killinger BA
    3. Ensink E
    4. Li P
    5. Li KX
    6. Cui W
    7. Lubben N
    8. Weiland M
    9. Wang X
    10. Gordevicius J
    11. Coetzee GA
    12. Ma J
    13. Jovinge S
    14. Labrie V

    (2020) Epigenomic analysis of Parkinson’s disease neurons identifies Tet2 loss as neuroprotective

    Nature Neuroscience 23:1203–1214.

    https://doi.org/10.1038/s41593-020-0690-y

    • PubMed
    • Google Scholar
44. 1. Martin M

    (2011) Cutadapt removes adapter sequences from high-throughput sequencing reads

    EMBnet.Journal 17:10.

    https://doi.org/10.14806/ej.17.1.200

    • Google Scholar
45. 1. McKay BE
    2. Turner RW

    (2005) Physiological and morphological development of the rat cerebellar Purkinje cell: Purkinje cell output parallels dendritic development

    The Journal of Physiology 567:829–850.

    https://doi.org/10.1113/jphysiol.2005.089383

    • Google Scholar
46. 1. Mellén M
    2. Ayata P
    3. Dewell S
    4. Kriaucionis S
    5. Heintz N

    (2012) MeCP2 Binds to 5hmC Enriched within Active Genes and Accessible Chromatin in the Nervous System

    Cell 151:1417–1430.

    https://doi.org/10.1016/j.cell.2012.11.022

    • PubMed
    • Google Scholar
47. 1. Mellén M
    2. Ayata P
    3. Heintz N

    (2017) 5-hydroxymethylcytosine accumulation in postmitotic neurons results in functional demethylation of expressed genes

    PNAS 114:E7812–E7821.

    https://doi.org/10.1073/pnas.1708044114

    • PubMed
    • Google Scholar
48. 1. Mo A
    2. Mukamel EA
    3. Davis FP
    4. Luo C
    5. Henry GL
    6. Picard S
    7. Urich MA
    8. Nery JR
    9. Sejnowski TJ
    10. Lister R
    11. Eddy SR
    12. Ecker JR
    13. Nathans J

    (2015) Epigenomic Signatures of Neuronal Diversity in the Mammalian Brain

    Neuron 86:1369–1384.

    https://doi.org/10.1016/j.neuron.2015.05.018

    • PubMed
    • Google Scholar
49. 1. Morales D
    2. Hatten ME

    (2006) Molecular Markers of Neuronal Progenitors in the Embryonic Cerebellar Anlage

    The Journal of Neuroscience 26:12226–12236.

    https://doi.org/10.1523/JNEUROSCI.3493-06.2006

    • PubMed
    • Google Scholar
50. 1. Nan X
    2. Bird A

    (2001) The biological functions of the methyl-CpG-binding protein MeCP2 and its implication in Rett syndrome

    Brain and Development 23:S32–S37.

    https://doi.org/10.1016/S0387-7604(01)00333-3

    • Google Scholar
51. 1. Patro R
    2. Duggal G
    3. Love MI
    4. Irizarry RA
    5. Kingsford C

    (2017) Salmon provides fast and bias-aware quantification of transcript expression

    Nature Methods 14:417–419.

    https://doi.org/10.1038/nmeth.4197

    • PubMed
    • Google Scholar
52. 1. Piccolo FM
    2. Liu Z
    3. Dong P
    4. Hsu CL
    5. Stoyanova EI
    6. Rao A
    7. Tjian R
    8. Heintz N

    (2019) MeCP2 nuclear dynamics in live neurons results from low and high affinity chromatin interactions

    eLife 8:e51449.

    https://doi.org/10.7554/eLife.51449

    • PubMed
    • Google Scholar
53. 1. Ramírez F
    2. Ryan DP
    3. Grüning B
    4. Bhardwaj V
    5. Kilpert F
    6. Richter AS
    7. Heyne S
    8. Dündar F
    9. Manke T

    (2016) deepTools2: a next generation web server for deep-sequencing data analysis

    Nucleic Acids Research 44:W160–W165.

    https://doi.org/10.1093/nar/gkw257

    • PubMed
    • Google Scholar
54. 1. Robinson JT
    2. Thorvaldsdóttir H
    3. Winckler W
    4. Guttman M
    5. Lander ES
    6. Getz G
    7. Mesirov JP

    (2011) Integrative genomics viewer

    Nature Biotechnology 29:24–26.

    https://doi.org/10.1038/nbt.1754

    • PubMed
    • Google Scholar
55. 1. Ross-Innes CS
    2. Stark R
    3. Teschendorff AE
    4. Holmes KA
    5. Ali HR
    6. Dunning MJ
    7. Brown GD
    8. Gojis O
    9. Ellis IO
    10. Green AR
    11. Ali S
    12. Chin SF
    13. Palmieri C
    14. Caldas C
    15. Carroll JS

    (2012) Differential oestrogen receptor binding is associated with clinical outcome in breast cancer

    Nature 481:389–393.

    https://doi.org/10.1038/nature10730

    • PubMed
    • Google Scholar
56. 1. Rudenko A
    2. Dawlaty MM
    3. Seo J
    4. Cheng AW
    5. Meng J
    6. Le T
    7. Faull KF
    8. Jaenisch R
    9. Tsai LH

    (2013) Tet1 Is Critical for Neuronal Activity-Regulated Gene Expression and Memory Extinction

    Neuron 79:1109–1122.

    https://doi.org/10.1016/j.neuron.2013.08.003

    • PubMed
    • Google Scholar
57. 1. Schep AN
    2. Wu B
    3. Buenrostro JD
    4. Greenleaf WJ

    (2017) chromVAR: inferring transcription-factor-associated accessibility from single-cell epigenomic data

    Nature Methods 14:975–978.

    https://doi.org/10.1038/nmeth.4401

    • PubMed
    • Google Scholar
58. 1. Schultz MD
    2. He Y
    3. Whitaker JW
    4. Hariharan M
    5. Mukamel EA
    6. Leung D
    7. Rajagopal N
    8. Nery JR
    9. Urich MA
    10. Chen H
    11. Lin S
    12. Lin Y
    13. Jung I
    14. Schmitt AD
    15. Selvaraj S
    16. Ren B
    17. Sejnowski TJ
    18. Wang W
    19. Ecker JR

    (2015) Human body epigenome maps reveal noncanonical DNA methylation variation

    Nature 523:212–216.

    https://doi.org/10.1038/nature14465

    • PubMed
    • Google Scholar
59. 1. Soneson C
    2. Love MI
    3. Robinson MD

    (2015) Differential analyses for RNA-seq: transcript-level estimates improve gene-level inferences

    F1000Research 4:1521.

    https://doi.org/10.12688/f1000research.7563.2

    • PubMed
    • Google Scholar
60. 1. Song Q
    2. Decato B
    3. Hong EE
    4. Zhou M
    5. Fang F
    6. Qu J
    7. Garvin T
    8. Kessler M
    9. Zhou J
    10. Smith AD

    (2013) A Reference Methylome Database and Analysis Pipeline to Facilitate Integrative and Comparative Epigenomics

    PLOS ONE 8:e81148.

    https://doi.org/10.1371/journal.pone.0081148

    • PubMed
    • Google Scholar
61. 1. Szulwach KE
    2. Li X
    3. Li Y
    4. Song CX
    5. Wu H
    6. Dai Q
    7. Irier H
    8. Upadhyay AK
    9. Gearing M
    10. Levey AI
    11. Vasanthakumar A
    12. Godley LA
    13. Chang Q
    14. Cheng X
    15. He C
    16. Jin P

    (2011) 5-hmC-mediated epigenetic dynamics during postnatal neurodevelopment and aging

    Nature Neuroscience 14:1607–1616.

    https://doi.org/10.1038/nn.2959

    • PubMed
    • Google Scholar
62. 1. Tahiliani M
    2. Koh KP
    3. Shen Y
    4. Pastor WA
    5. Bandukwala H
    6. Brudno Y
    7. Agarwal S
    8. Iyer LM
    9. Liu DR
    10. Aravind L
    11. Rao A

    (2009) Conversion of 5-Methylcytosine to 5-Hydroxymethylcytosine in Mammalian DNA by MLL Partner TET1

    Science 324:930–935.

    https://doi.org/10.1126/science.1170116

    • PubMed
    • Google Scholar
63. 1. Tsagaratou A
    2. Äijö T
    3. Lio CWJ
    4. Yue X
    5. Huang Y
    6. Jacobsen SE
    7. Lähdesmäki H
    8. Rao A

    (2014) Dissecting the dynamic changes of 5-hydroxymethylcytosine in T-cell development and differentiation

    PNAS 111:E3315.

    https://doi.org/10.1073/pnas.1412327111

    • PubMed
    • Google Scholar
64. 1. Tsagaratou A
    2. González-Avalos E
    3. Rautio S
    4. Scott-Browne JP
    5. Togher S
    6. Pastor WA
    7. Rothenberg EV
    8. Chavez L
    9. Lähdesmäki H
    10. Rao A

    (2017) TET proteins regulate the lineage specification and TCR-mediated expansion of iNKT cells

    Nature Immunology 18:45–53.

    https://doi.org/10.1038/ni.3630

    • PubMed
    • Google Scholar
65. 1. Weber AR
    2. Krawczyk C
    3. Robertson AB
    4. Kuśnierczyk A
    5. Vågbø CB
    6. Schuermann D
    7. Klungland A
    8. Schär P

    (2016) Biochemical reconstitution of TET1-TDG-BER-dependent active DNA demethylation reveals a highly coordinated mechanism

    Nature Communications 7:10806.

    https://doi.org/10.1038/ncomms10806

    • PubMed
    • Google Scholar
66. 1. Wickham H
    2. Averick M
    3. Bryan J
    4. Chang W
    5. McGowan LD
    6. François R
    7. Grolemund G
    8. Hayes A
    9. Henry L
    10. Hester J
    11. Kuhn M
    12. Pedersen TL
    13. Miller E
    14. Bache SM
    15. Müller K
    16. Ooms J
    17. Robinson D
    18. Seidel DP
    19. Spinu V
    20. Takahashi K
    21. Vaughan D
    22. Wilke C
    23. Woo K
    24. Yutani H

    (2019) Welcome to the Tidyverse

    Journal of Open Source Software 4:1686.

    https://doi.org/10.21105/joss.01686

    • Google Scholar
67. 1. Wu X
    2. Zhang Y

    (2017) TET-mediated active DNA demethylation: mechanism, function and beyond

    Nature Reviews. Genetics 18:517–534.

    https://doi.org/10.1038/nrg.2017.33

    • PubMed
    • Google Scholar
68. 1. Xie W
    2. Schultz MD
    3. Lister R
    4. Hou Z
    5. Rajagopal N
    6. Ray P
    7. Whitaker JW
    8. Tian S
    9. Hawkins RD
    10. Leung D
    11. Yang H
    12. Wang T
    13. Lee AY
    14. Swanson SA
    15. Zhang J
    16. Zhu Y
    17. Kim A
    18. Nery JR
    19. Urich MA
    20. Kuan S
    21. Yen C
    22. Klugman S
    23. Yu P
    24. Suknuntha K
    25. Propson NE
    26. Chen H
    27. Edsall LE
    28. Wagner U
    29. Li Y
    30. Ye Z
    31. Kulkarni A
    32. Xuan Z
    33. Chung W-Y
    34. Chi NC
    35. Antosiewicz-Bourget JE
    36. Slukvin I
    37. Stewart R
    38. Zhang MQ
    39. Wang W
    40. Thomson JA
    41. Ecker JR
    42. Ren B

    (2013) Epigenomic Analysis of Multilineage Differentiation of Human Embryonic Stem Cells

    Cell 153:1134–1148.

    https://doi.org/10.1016/j.cell.2013.04.022

    • PubMed
    • Google Scholar
69. 1. Xu Y
    2. Xu C
    3. Kato A
    4. Tempel W
    5. Abreu JG
    6. Bian C
    7. Hu Y
    8. Hu D
    9. Zhao B
    10. Cerovina T
    11. Diao J
    12. Wu F
    13. He HH
    14. Cui Q
    15. Clark E
    16. Ma C
    17. Barbara A
    18. Veenstra GJC
    19. Xu G
    20. Kaiser UB
    21. Liu XS
    22. Sugrue SP
    23. He X
    24. Min J
    25. Kato Y
    26. Shi YG

    (2012) Tet3 CXXC Domain and Dioxygenase Activity Cooperatively Regulate Key Genes for Xenopus Eye and Neural Development

    Cell 151:1200–1213.

    https://doi.org/10.1016/j.cell.2012.11.014

    • PubMed
    • Google Scholar
70. 1. Xu X
    2. Stoyanova EI
    3. Lemiesz AE
    4. Xing J
    5. Mash DC
    6. Heintz N

    (2018) Species and cell-type properties of classically defined human and rodent neurons and glia

    eLife 7:e37551.

    https://doi.org/10.7554/eLife.37551

    • PubMed
    • Google Scholar
71. 1. Zhang Y
    2. Liu T
    3. Meyer CA
    4. Eeckhoute J
    5. Johnson DS
    6. Bernstein BE
    7. Nusbaum C
    8. Myers RM
    9. Brown M
    10. Li W
    11. Liu XS

    (2008) Model-based analysis of ChIP-Seq (MACS)

    Genome Biology 9:R137.

    https://doi.org/10.1186/gb-2008-9-9-r137

    • PubMed
    • Google Scholar
72. 1. Zhu X
    2. Girardo D
    3. Govek EE
    4. John K
    5. Mellén M
    6. Tamayo P
    7. Mesirov JP
    8. Hatten ME

    (2016) Role of Tet1/3 Genes and Chromatin Remodeling Genes in Cerebellar Circuit Formation

    Neuron 89:100–112.

    https://doi.org/10.1016/j.neuron.2015.11.030

    • PubMed
    • Google Scholar

Author details

1. Elitsa Stoyanova

   Laboratory of Molecular Biology, Howard Hughes Medical Institute, The Rockefeller University, New York, United States

   Contribution

   Conceptualization, Formal analysis, Investigation, Visualization, Writing – original draft, Writing – review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0001-6400-6119

2. Michael Riad

   Laboratory of Molecular Biology, Howard Hughes Medical Institute, The Rockefeller University, New York, United States

   Contribution

   Formal analysis, Investigation, Writing – review and editing

   Competing interests

   No competing interests declared

3. Anjana Rao

   1. Sanford Consortium for Regenerative Medicine, La Jolla, United States
   2. La Jolla Institute for Allergy and Immunology, La Jolla, United States
   3. Department of Pharmacology, University of California San Diego, La Jolla, United States

   Contribution

   Resources, Writing – review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-1870-1775

4. Nathaniel Heintz

   Laboratory of Molecular Biology, Howard Hughes Medical Institute, The Rockefeller University, New York, United States

   Contribution

   Conceptualization, Funding acquisition, Resources, Supervision, Writing – original draft, Writing – review and editing

   For correspondence

   heintz@rockefeller.edu

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-8874-8704

• 44

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.