IMR Press / JIN / Volume 20 / Issue 3 / DOI: 10.31083/j.jin2003057
Open Access Original Research
TET3-mediated accumulation of DNA hydroxymethylation contributes to the activity-dependent gene expression of Rab3a in post-mitotic neurons
Show Less
1 The First People’s Hospital of Huaihua, Affiliated to University of South China, 418008 Huaihua, Hunan, China
2 Department of Neurosurgery, Zhongnan Hospital of Wuhan University, 430071 Wuhan, Hubei, China
3 Department of Neurosurgery, Third Affiliated Hospital of Naval Medical University, 200438 Shanghai, China

These authors contributed equally.

J. Integr. Neurosci. 2021, 20(3), 529–539;
Submitted: 16 April 2021 | Revised: 28 May 2021 | Accepted: 23 July 2021 | Published: 30 September 2021

Rab3a, a subtype protein in the Rab3 family amongst the small G proteins, is closely associated with the learning and memory formation process. Various neuronal stimuli can induce the expression of Rab3a; however, how DNA modification is involved in regulating its expression is not fully understood. Ten-eleven translocation (TET) proteins can oxidate methylcytosine to hydroxymethylcytosine, which can further activate gene expression. Previous studies reported that TET-mediated regulation of 5hmC induced by learning is involved in neuronal activation. However, whether Tet protein regulates Rab3a is unknown. To understand the role of TET-mediated 5hmC on Rab3a in neuronal activation, we adopted a KCl-induced depolarization protocol in cultured primary cortical neurons to mimic neuronal activity in vitro. After KCl treatment, Rab3a and Tet3 mRNA expression were induced. Moreover, we observed a decrease in the methylation level and an increase of hydroxymethylation level surrounding the CpG island near the transcription start site of Rab3a. Furthermore, recently, Formaldehyde-Assisted Isolation of Regulatory Elements (FAIRE) has proven powerful in identifying open chromatin in the genome of various eukaryotes. Using FAIRE-qPCR, we observed a euchromatin state and the increased occupancy of Tet3, H3K4me3, and H3K27ac at the promoter region of Rab3a after KCl treatment. Finally, by using shRNA to knockdown Tet3 prior KCl treatment, all changes mentioned above vanished. Thus, our findings elucidated that the neuronal activity-induced accumulation of hydroxymethylation, which Tet3 mediates, can introduce an active and permissive chromatin structure at Rab3a promoter and lead to the induction of Rab3a mRNA expression.

DNA hydroxymethylation
Neuronal activity
DNA modification
DNA methylation
1. Introduction

Rab3a protein, a subtype protein in the Rab3 family in the small G proteins of the Ras superfamily, was first identified at chromosome 19 band p13.2 in 1989 [1]. It is highly enriched in the synaptic vesicles of neuronal cells, and it is also the most abundant Rab protein in the brain. Rab3a is composed of 220 amino acids and plays a role in the release of neurotransmitters. It is generally considered one of the most prominently studied activity-dependent molecules in the Rab3 family, and it is involved in exocytosis, plasma membrane repair, nerve fiber growth, and vesicle formation [2-7]. Dysregulation of Rab3a expression may precipitate deterioration in memory, learning impairment, and neurodegenerative diseases [8,9].

Currently, the epigenetic machinery that is involved in regulating the Rab3a expression under neuronal activation remains unknown. Moreover, DNA methylation and demethylation are two vital epigenetic regulatory mechanisms for gene expression [10], especially in neurons. There are two ways of DNA demethylation including passive demethylation (dependent on DNA replication) and active demethylation (independent of DNA replication) [11], and generally only active demethylation happens in post-mitotic neurons. Recent studies have revealed a variety of DNA modifications in post-mitotic neurons, including 5-hydroxymethylcytosine (5hmC), 5-formylcytosine (5fC), and 5-carboxylcytosine (5caC), which are the main DNA modifications that have been found in the active DNA demethylation pathway. Intriguingly, it has been shown that these 5mC oxidative derivatives are not transit products of active DNA demethylation; instead, they can form a stable epigenetic regulating state to promote gene expression [12-14]. Interestingly, there is an CpG island at the promoter region of Rab3a, and it may suggest methylation and demethylation processes would be involved in regulating Rab3a expression in an activity-dependent manner.

The Ten-eleven translocation (TET) protein family, including Tet1, Tet2, and Tet3, plays an essential role in DNA demethylation [15]. It can oxidize 5mC into 5hmC and subsequently into 5fC and 5caC [16], which is considered as the main protein family that initiates DNA demethylation mechanism. TET1 and TET3 contain a CXXC zinc finger domain at their amino-terminus known to bind CpG sequences [17], whereas TET2 partners with IDAX, an independent CXXC-containing protein [18]. Among them, Tet3 shows the highest expression in brain regions [19]. Tet3 is involved in cell survival, angiogenesis, neurogenesis, antioxidant defense, DNA repair, and metabolism. Moreover, previous research shows that Tet3, but not Tet1, exhibits a time-dependent increase in mRNA expression in an activity-dependent manner [20]. In line with this result, Tet3 may be a critical regulator of DNA demethylation during neuronal activation. However, so far, no relevant literature regarding the mechanism of Tet3 regulating Rab3a has been reported.

To date, the majority of studies have focused on the function of Rab3a, but the molecular mechanisms of how the epigenetic machinery regulates its expression under neuronal stimulation have yet to be uncovered. As described above, we hypothesize that Tet 3 may mediate the oxidation of 5mC to 5hmC, 5fC, and 5Cac in the CpG island of the Rab3a promoter region and promote Rab3a expression in an activity-dependent manner. This study can help reveal the interrelationship between Rab3a and Tet3 and providing novel insights into their functional relevance to provide the basis for further research.

2. Materials and methods
2.1 Primary cortical neuron cell culture and KCl treatment

Cortical tissue was isolated from E18 mouse embryos in a sterile atmosphere. To dissociate the tissue, it was finely chopped followed by gentle pipetting to create single cell suspension. To prevent clumping of cells due to DNA from dead cells, tissue was treated with 2 unit/μL of DNase I (Thermo EN0521). Cells finally went through the 40 μm cell strainer (BD Falcon) and were plated onto 6 well plate coated with Poly-L-ornithine (Sigma P2533) at a density of 1 × 106 cells per well. The medium used was Neurobasal media (Gibco) containing B27 supplement (Gibco), 1X Glutamax (Gibco), and 1% Pen/Strep (Sigma). Primary cortical neurons were stimulated by adding of the final concentration of 25 mM KCl as indicated in the experiments and incubated at 37 C for 3 hours. Simultaneously, the same volume of NB media was added to the primary cortical neurons, which served as the KCl-control. The process of embryo isolation followed the protocol approved by the Animal Ethics Committee of the First People’s Hospital of Huaihua.

2.2 Chromatin immunoprecipitation

Chromatin immunoprecipitation (ChIP) was performed following modification of the Invitrogen ChIP kit protocol. Cells were fixed in 1% formaldehyde and cross-linked cell lysates were trimmed with Covaris in 1% SDS lysis buffer to generate chromatin fragments with an average length of 300 bp by using peak power: 75, duty factor: 2, cycle/burst: 200, duration: 900 secs and temperature: between 5 C to 9 C. The chromatin was then immunoprecipitated using the specific antibody for each target. Then, fixed chromatin samples were incubated with Tet3 (Active motif, Cat No: 61395), H3K4me3 (Abcam, Cat No: ab8580) and H3K27ac (Abcam, Cat No: 4760) as well as antibodies overnight at 4 C. Also, an equivalent amount of IP/ChIP grade Rabbit IgG (Cell Signaling Technology, Cat No: 2729) was used for non-specificity control. Protein-DNA-antibody complexes were precipitated with protein G-magnetic beads (Invitrogen) for 1 hr at 4 C, followed by three washes in low-salt buffer, and three washes in high-salt buffer. The precipitated protein-DNA complexes were eluted from the antibody with 1% SDS and 0.1 M NaHCO3, then incubated for 4 hrs at 60 C in 200 mM NaCl to reverse formaldehyde cross-link. Following proteinase K digestion, phenol-chloroform extraction, and ethanol precipitation, samples were subjected to qPCR using primers specific for 200 bp segments corresponding to the target regions.

2.3 Tet3 knockdown experiment

Lentiviral plasmids were generated by inserting either Tet3 shRNA (shRNATet3: GCCTGTTAGGCAGATTGTTCT) or scrambled control (scrambled control: CCTAAGGTTAAGTCGCCCTCG) fragments immediately downstream of the human H1 promoter in a modified FG12 vector (FG12H1, derived from the FG12 vector originally provided by David Baltimore, CalTech). Lentivirus was prepared and maintained as per previously published protocols [20].

2.4 DNA/RNA extraction

Cell pellets were collected by 1000 × g for 5 mins, followed by RNA and DNA extraction. Four hundred μL of homogenate were used for DNA extraction, and 100 μL were used for RNA extraction. DNA extraction was carried out using DNeasy Blood & Tissue Kit (Qiagen) with RNAse A (5 prime), and RNA was extracted using Trizol reagent (Invitrogen). Both extraction protocols were followed according to the manufacturer’s instructions. The concentrations of DNA and RNA were measured by Qubit assay (Invitrogen).

2.5 RT-qPCR

1 ug RNA was employed for cDNA synthesis using the PrimeScript Reverse Transcription Kit (Takara). Quantitative PCR was performed on a RotorGeneQ (Qiagen) cycler with SYBR-Green Master mix (Qiagen) using primers for target genes and for beta-actin as an internal control (Rab3a Forward primer: TCCCAGTCCCCTGGAAAAAC and Reverse primer: GCTCCTCCTTTAGGAACTCGG; Tet1 Forward primer: ACACAGTGGTGCTAATGCAG and Reverse primer: AGCATGAACGGGAGAATCGG; Tet2 Forward primer: AGAGAAGACAATCGAGAAGTCGG and Reverse primer: CCTTCCGTACTCCCAAACTCAT; Tet3 Forward primer: GTGGTCGGACAGTGAACACAA and Reverse primer: GTTGGGCTGGTTGAGGTTCTT; and beta-actin Forward primer: GGCTGTATTCCCCTCCATCG and Reverse primer: CCAGTTGGTAACAATGCCATGT (the primers were designed based on mm10). All transcript levels were normalized to beta-actin mRNA using the ΔΔCT method, and each PCR reaction was run in duplicate for each sample and repeated at least twice.


FAIRE-qPCR was performed as previously described [21]. In details, Primary cortical neurons were fixed with molecular grade formaldehyde (16%, Thermo fisher) at room temperature (22–25 C) to a final concentration of 1% and incubated for 5 min. Glycine was then added to achieve a final concentration of 125 mM for 5 min at room temperature to stop fixation. Two rounds of phosphate buffer saline wash were performed, and the cells were collected using centrifugation at 2000 rpm for 4 mins and stored at –80 C. Fixed cell pellets were then treated with ChIP lysis buffer as described above and samples sonicated using Covaris to generate chromatin fragments with an average length of 300 bp (using peak power: 75, duty factor: 2, cycle/burst: 200, duration: 900 secs and temperature: between 5 C to 9 C). Cellular debris was cleared by spinning at 15,000 rpm for 10 mins at 4 C. DNA was isolated by adding an equal volume of phenol-chloroform (Sigma) and followed by vertex and spinning at 15,000 rpm for 15 mins at 4 C. The aqueous phase was isolated and stored in a fresh 1.5 mL microcentrifuge tube. An additional 500 uL of Tris-EDTA buffer solution was added to the organic phase, vortexed and centrifuged again at 15,000 rpm for 15 mins at 4 C. The aqueous phase was isolated and combined with the first aqueous fraction. Another phenol-chloroform extraction was performed on the pooled aqueous fractions to ensure that all protein had been removed. The DNA was isolated using the previously described DNA extraction procedures in the ChIP protocol. Input DNA isolation was conducted using Genomic DNA isolation kit (QIAGEN, Hilden, Germany). DNA enrichment in the FAIRE samples was determined using a qPCR analysis with a Step One Plus system (Applied Biosystems) and SYBR green fluorescence. The cycle number required to reach the threshold was recorded and analyzed. PCR was performed using the supernatant DNA (FAIRE sample) and the input DNA (control sample). Recovery of FAIRE-extracted DNA in comparison to Input DNA was then determined by qPCR using the ΔΔCT method.

2.7 5mC/5hmC DIP-qPCR

1 μg of genomic DNA was diluted to 130 μL ultrapure water (Invitrogen) and trimmed with an average size of about 300 bp prior to capture. Five mC/5hmC capture was performed using a DIP grade 5mC/5hmC antibody (Active Motif; 5mC antibody Cat No: 61255; 5hmC antibody Cat No: 39791) to capture 5mC/5hmC enriched genomic regions. The procedure was adapted from the manufacturer’s protocol for methyl DNA immunoprecipitation (Active Motif). Five hundred ng of trimmed DNA and 1 μg of 5mC/5hmC antibody was used for each immunoprecipitation reaction and all selected target sequences based on mm10 genome (CpG site proximal Rab3a promoter: chr8: 70,754,926-70,755,263, forward primer: GTCTCGCATTTGGCATTCTT and reverse primer: ATGCACGCACAACGTATCAC; upstream site: chr8: 70,753,542-70,753,842, forward primer: TCATAGCGGCTTGGAAGTCT and reverse primer: TGATGAGTCCCCAGAGAACC; and downstream site: chr8: 70,755,842-70,756,148, forward primer: CTCCTTCACTCCAGCCTTTG and reverse primer: GGCTAGCCTCAGCTTCCTTT) were normalized for input DNA and then for their own controls by using the ΔΔCTmethod. Beta-actin (forward primer: GGCTGTATTCCCCTCCATCG; reverse primer: CCAGTTGGTAACAATGCCA TGT) was used as reference. Each qPCR reaction was run in duplicate for each sample and repeated at least twice. To validate, the antibody specificity and accuracy, IP/ChIP grade Rabbit IgG (Cell Signaling Technology, Cat No: 2729) and Methylated DNA Standard Kit (Active Motif; Cat No: 55008) was used.

2.8 Statistical analysis

The results were expressed as mean ± SD. Statistical analyses were performed using GraphPad Prism version 8.0 (GraphPad Software). Independent sample t-tests, paired sample t-tests, independent samples t-tests with Welch’s correction, and Mann-Whitney U tests were used. P-values of less than 0.05 were considered to indicate a significant difference.

3. Results
3.1 The increased expression of Rab3a and Tet3 and decreased expression of TET1 under KCl-induced depolarization in cultured primary cortical neurons

It has been shown that Tet-meditated 5hmC can lead to induction of synapsis-related gene expression [22]. To extend our understanding on how Tet family proteins and 5hmC regulate Rab3a expression, we first tested the expression of Tet proteins and Rab3a in vitro. Neurons derived from embryonic day 18 (E18) were maintained for 7 days in vitro (DIV). Then, 25 mM final concentration of KCl were added to culture neurons following protocols published previously (Fig. 1A). Post-KCl treatment, compared with the KCl-group, the mRNA expression levels of Rab3a increased 2.5 folds (P < 0.001) (Fig. 1B); there was a significant decrease (P < 0.01) of Tet1 expression (Fig. 1C); no significant alteration was observed in the Tet2 expression (Fig. 1D); whereas Tet3 mRNA expression was 1.7-fold greater (P < 0.001, Fig. 1E). Together, these results suggest that induction of Tet3, but not Tet1 or Tet2, may correlate with Rab3a mRNA expression, and we hypothesized that Tet3 may regulate Rab3a expression through active DNA demethylation.

Fig. 1.

The transcription levels of Rab3a and Tet3 were increased in primary cortical neurons induced by KCl, whereas the opposite results were observed on Tet1 expression and no obvious changes were observed in Tet2. (A) Workflow for the path KCl-induced experiments in primary cortical neurons. (B–E) Real time quantitative PCR of reverse transcription (RT-qPCR) assay to detect mRNA level of Rab3a (B), Tet1 (C), Tet2 (D), and Tet3 (E) in KCl- and KCl+ groups. Values represent mean ± SEM. Independent samples t tests were applied for comparisons between the KCl- and KCl+ groups vs. the corresponding control group. *P < 0.05, **P < 0.01, ***P < 0.001, significance values for comparisons with the corresponding control group.

3.2 KCl treatment leads to the accumulation of 5hmC at Rab3a promoter

To further test our hypothesis, we examined the methylation state of the Rab3a promoter region. By using methylated DNA immunoprecipitation (MeDIP)- and hydroxymethylated DNA immunoprecipitation (hMeDIP)-qPCR, we accessed the 5mC and 5hmC level surrounding the transcription start site of Rab3a (Fig. 2A). To explore the specific binding of MeDIP and hMeDIP, we add IgG control in both Tet3 and histone modification ChIP as well as DIP experiments. The result showed that MeDIP and hMeDIP can have specific binding in our experiment (P < 0.001, Fig. 2B,C). Intriguingly, we observed reciprocal changes of 5hmC and 5mC at the promoter region of Rab3a, which contains a CpG island. The hMeDIP–qPCR and MeDIP–qPCR results revealed a 0.5-fold decrease in 5mC levels (P < 0.01, Fig. 2D) and 2-fold increase in 5hmC levels (P < 0.001, Fig. 2E) for the KCl treatment. Furthermore, to test whether this demethylation effect is specific to this CpG island containing protomer region, we also checked around 1000 bp downstream and upstream of this aera and found no significant methylation level change (Fig. 2F,G). These results suggest that there is a specific KCl-induced accumulation of 5hmC at the CpG island of the Rab3a promoter in post-mitotic neurons.

Fig. 2.

CpG island demethylation levels increased at the Rab3a promoter for KCl treatment, but not downstream or upstream. (A) Schematic illustration of the Rab3a gene region. The gene is located on chromosome 8 from the 7,075,500 bp to 70,756,300 bp regions. The target sequence was located in the CpG islands region of the Rab3a promoter. TSS: transcriptional start site. (B) Validation of (h)MeDIP-seq data using ChIP-qPCR for adenomatous polyposis coli (APC) no-mods. (C) Validation of IgG 5(h)mC at CpG sites compared with corresponding normal IgG. (D) MeDIP results showing the DNA methylation status of the Rab3a promoter. Five independent repeats were performed. (E–G) hMeDIP results showing the DNA demethylation status of the Rab3a promoter, downstream, and upstream. Values represent mean ± SEM. Independent samples t tests were used for comparisons between the control and KCl treatment group. *P < 0.05, **P < 0.01, ***P < 0.001, significance values for comparisons with the corresponding control group. APC, Adenomatous polyposis coli.

3.3 Accumulation of 5hmC is associated with an active open chromatin state at the promoter region of Rab3a gene

Previous studies have shown 5hmC accumulation is associated with euchromatin structure. By using formaldehyde-assisted isolation of regulatory elements (FAIRE)–qPCR, we examined chromatin accessibility in the Rab3a promoter region. We observed a significant increase of Tet3 occupancy (P < 0.001, Fig. 3A) an open chromatin structure state at the 5hmC accumulation site (P < 0.001, Fig. 3B) under the KCl depolarization condition. It has been shown that the permissive histone modifications are required for the initiation of transcription [23]. Thus, we employed chromatin immunoprecipitation (ChIP)-qPCR to capture the profile of histone modification at the same locus. As expected, we observed increased occupancy of H3K4me3 and H3K27ac at the KCl-treated group (P < 0.01, Fig. 3C,D). These results suggest Tet3-mediated 5hmC accumulation leads to active euchromatin structure at the Rab3a promoter during neuronal depolarization. In addition, ChIP-qPCR for permissive chromatin marks H3K4me3 or H3K27ac was used to substantiate this possibility (Fig. 3E).

Fig. 3.

Tet3-mediated 5hmC accumulation resulted in an active open chromatin state at Rab3a promoter. (A) ChIP-qPCR occupancy of Tet3 in Rab3a proximal region. (B) FAIRE enrichment in Rab3a promoter was quantified by real-time PCR on the basis of the threshold cycle value (CT). (C,D) ChIP-qPCR occupancy profiles of H3K4me3 and H3K27ac in the Rab3a proximal region. (E) The same chromatin was used for control ChIP-qPCR experiments with control IgG. Values for nonspecific binding (as determined by using control IgG) were subtracted. Values represent mean ± SEM. Independent samples t-tests were used for comparisons between the control and KCl treatments group. *P < 0.05, **P < 0.01, ***P < 0.001, significance values for comparisons with the corresponding control group.

3.4 Tet3 knockdown inhibits activity induced accumulation of 5hmC associated permissive chromatin structure at Rab3a promoter and Rab3a expression

To further verify our hypothesis, we designed the shRNA specifically targeted to Tet3, and packaged it into lenti-virus. This lenti-virus Tet3 shRNA could sufficiently carry the shRNA into the primary cortical neuron, as shown in Fig. 4A and this shRNA can significantly reduce the expression of Tet3 mRNA (Fig. 4B). We then sought to examine if Tet3 is necessary for Rab3a expression under KCl-induced conditions. The Tet3 and Rab3a mRNA expression levels after KCl-induced depolarization significantly decreased with Tet3 shRNA included when compared with scramble shRNA (P < 0.001, Fig. 4B; P < 0.001, Fig. 4C). ChIP-qPCR demonstrated that this shRNA can disrupt the Tet3 occupancy at the Rab3a promoter as compared to the scramble shRNA control (P < 0.001, Fig. 4D).

Moreover, we also measured the effect of Tet3 shRNA on 5hmC levels. The result confirmed reduction of Tet3 level can inhibit activity-induced 5hmC enrichment at the Rab3a promoter region (P < 0.001, Fig. 4E) followed by prevention of an open chromatin structure (P < 0.01, Fig. 4F). Furthermore, we also evaluated whether Tet3 knockdown has any effect on the occupancy of H3K4me3 and H3K27ac. ChIP-qPCR results revealed that Tet3 knockdown inhibited previously observed activity-induced occupancy for H3K4me3 (P < 0.01, Fig. 4G) and H3K27ac (P < 0.001, Fig. 4H). Thus, these results support our hypothesis, according to which activity-induced Tet3-mediated 5hmC accumulation leads to an active and permissive chromatin structure at the Rab3a promoter followed by increasing of Rab3A expression.

Fig. 4.

Tet3-induced promoter demethylation leads to an active and permissive chromatin structure at the Rab3a promoter accompanied by increments of Rab3A expression. (A) Transfection efficiency of lentivirus in primary cortical neurons. Primary neurons were transfected with plasmids expressing shRNA and green fluorescent protein. Green fluorescent protein was detected as a marker of transfection efficiency by fluorescence microscopy. Control: transfection vector; Tet3 shRNA, shRNA transfection. (B,C) RT–PCR analysis of knockdown of Tet3 in primary cortical neurons transduced with lentivirus expressing Tet3 shRNA. (D) Tet3 occupancy after knockdown of Tet3 in primary cortical neurons. (E) hMeDIP-qPCR analysis for 5hmC enrichment after knockdown of Tet3. (F) FAIRE enrichment in the Rab3a promoter after knockdown of Tet3. (G,H) Occupancy of H3K4me3 and H3K27ac after knockdown of Tet3. Data are presented as the mean ± SD of two independent experiments. Independent samples t tests were used for comparison between the control and KCl treatments group. *P < 0.05, **P < 0.01, ***P < 0.001, significance values for comparisons with the corresponding group.

4. Discussion

This study produced the following findings: (1) The expression of Rab3a and Tet3 increased by KCl-mediated depolarization, while the expression of Tet1 decreased. (2) Tet3 can promote 5hmC accumulation at the Rab3a promoter, which leads to an active open chromatin state at the promoter region of the Rab3a gene. (3) Accumulation of 5hmC and an open chromatin state can further facilitate the expression of Rab3a.

TET protein family consists of three members, namely Tet1, Tet2, and Tet3. They belong to a type of dioxygenase that depends on Fe2+ and α-ketoglutarate [15]. The C-terminal catalytic center is comprised of a double-stranded β-folded domain (DSBH) and a cysteine-rich (Cys-rich) region [24]. The DSBH domain can bind Fe2+ and α-ketoglutarate to oxidize the 5mC substrate, while the Cys-rich domain can help stabilize the interaction between DSBH and 5mC. There is a CXXC domain at the N-terminus of the full-length Tet1 and Tet3 [25]. Although all three TET family proteins have the activity of oxidizing 5mC, their functions in the central nervous system are different. Upon neuronal activation, Tet3-mediated 5hmC is highly involved in regulating learning associated genes under neuronal activity induction [20]. Our experimental results confirmed that levels of Tet3 mRNA were increased by KCl in primary cortical neurons, whereas the opposite results were observed on Tet1 expression and no changes on Tet2 expression. Moreover, we observed a significant increment of 5hmC level and a decrease of 5mC at the Rab3a promoter region post-KCl treatment. Moreover, ChIP–qPCR result showed that the occupancy of Tet3 at the Rab3a promotor is increased by KCl treatment. These results indicate that KCl-induced accumulation of 5hmC at the CpG island of Rab3a promoter is mediated by Tet3, but not by Tet1 or Tet2. These results further suggest that different Tet proteins could regulate different genes under physiological conditions in neurons. Notwithstanding these differences, the evidence is clear that in KCl-induced post-mitotic neurons, Tet3 is the key protein regulating 5hmC level at the Rab3A promoter under an activity-induced manner in post-mitotic neurons.

TET proteins can catalyze the conversion of 5mC to 5hmC and subsequently into 5-fC and 5-caC. 5hmC, 5fC, and 5caC may not only be the oxidation products of 5mC but also have their specific epigenetic functions. There are significant differences in the distribution of 5hmC, 5fC, and 5caC in different tissues. For example, 5hmC is 10- if not 100- fold more prevalent than 5fC/5caC, and it is relatively enriched in neurons [26-29]. Compared with 5hmC, 5fC/5caC was converted into an unmodified cytosine by the initiate base excision repair pathway [30-32]. Thus, rather than the intermediate product of the active demethylation process of 5mC, 5hmC is considered a stable epigenetic marker in mammalian cells. It is widely distributed in the nervous system and participates in memory formation, cognition, differentiation and development of neurons, as well as other multifunctional activities [33]. Thus, in our study, we only detected the accumulation of 5mC and 5hmC at the Rab3a promoter region. Nevertheless, the 5hmC level is higher than 5fC and 5caC, and the 5hmC is likely to be localized at the transcription start site [34,35]. Also, Tet3 could directly bind to the Rab3a promoter and mediate the accumulation of 5hmC. To confirm this, we established Tet3 shRNA and found the reduction of Tet3 level can block the activity-induced 5hmC enrichment at the Rab3a promoter region. Meanwhile, the lack of available antibodies that can detect 5-fC and 5-caC prevented us from testing. Further studies are, therefore, needed to reliably check the 5fC and 5caC level at the Rab3a promoter region.

5hmC has been proven to be an important epigenetic modification for euchromatin structure. DNA methylation can change chromatin conformation. DNA methylation is also related to histone modification, which may control the accessibility of transcription factors to promoters, followed by increasing transcription of specific genes. Common histone modifications mainly include H3K4me1, H3K4me3, H3K27me1, H3K27me3, and H3K27ac [36]. Among them, H3K4me3 and H3K27ac are two primary histone-modification markers and represent an active euchromatin state that is associated with gene expression [37,38]. In the present study, we observed the elevated occupancy of H3K4me3 and H3K27ac in the KCl group relative to the control group. Moreover, shRNA-targeted Tet3 was designed to knock down Tet3 expression, which inhibits the occupancy of H3K4me3 and H3K27ac. Thus, KCl depolarization triggers the expression of Tet3, and then Tet3 converts 5mC to 5hmC at Rab3a promoter. H3K4me3 and H3K27ac at the promoter were activated at a later stage, and the structure of chromatin changes from heterochromatin (closed chromatin) to euchromatin (open chromatin). Finally, the transcription factor can directly bind to promoters to upregulate the transcription of Rab3a (Fig. 5).

Fig. 5.

Diagram of the differences between the KCl (+) and KCl (-) group involving the mechanism of Tet3-mediated DNA demethylation leading an active and permissive chromatin structure at the Rab3a promoter.

Rab3a, located on the synaptic vesicles of brain neuron cells, belongs to the small guanosine triphosphate (GTP) binding protein, which is thought to plays a key role in the process of neurotransmitter release and membrane transport [39]. Rab3A, as an active GTP hydrolyzing enzyme, can bind to GTP. When Rab3A binds to GTP, GTP/Rab3A forms an active protein conformation with lowered proteolytic activity, which triggers a series of processes by interacting with downstream proteins to regulate the transport of neuronal synaptic vesicles [40,41]. Moreover, GTP hydrolyzed to guanosine diphosphate (GDP) by Rab3a, leads to structural conformation changes, and the protein loses its activity [42]. At the same time, this combination of Rab3A and GDP or GTP is also regulated by a variety of proteins, including guanine nucleotide exchange proteins (GEPs), GTPase-activating proteins (GAPs), and GDP dissociation inhibitors (GDIs) [43-45]. Furthermore, Rab3a is a requirement for brain-derived neurotrophic factor-induced (BDNF) synaptic plasticity associated with learning and memory [46]. Also, a previous study demonstrated that the homeostatic increase in miniature endplate current amplitude after activity blockade is diminished in the Rab3A deletion mouse [47]. This implicates that Rab3a is closely associated with neuronal activation. Generally, environmental factors engage specific receptors on nerve cells and activate ion fluxes, leading to nerve cell activation. Additionally, Rab3a regulates Ca2+-dependent exocytosis of neurotransmitter in the state of neuronal activation. Therefore, the increment of Rab3a expression can be induced by neuronal activation. However, how the expression of Rab3a is regulated remains inconclusive. In our study, we concluded that Tet3 drives KCl-induced expression of Rab3a by increasing DNA hydroxymethylation, chromatin accessibility and active histone modifications at the gene promoter from the perspective of DNA modification. These data therefore point to a critical role of Tet3-mediated oxidation of 5mC at CpG islands of the Rab3a promoter in regulating Rab3a gene expression under neuronal activation.

Admittedly, there are some limitations. Firstly, since the IgG controls should have been performed in parallel with the antibody-specific ChIP, it would be more appropriate to calculate the ddCt by normalizing first to the IgG pull-down over the same locus and then to the input DNA. IgG validation in our experiment were run by different cell batch and not same experimental run with previous ChIP sample. The changes would also not affect the conclusion. Li et al. [48], also published ChIP-qPCR and IgG validation separately. Secondly, the Rab3a gene may produce several different transcripts, which can be regulated by activation of the corresponding alternative promoter’s usage under neuronal activation. In the current study, we only focused on one promoter in the Rab3a genome. Future studies may require to fully elucidate the role of Tet3-mediated 5hmC in regulating activation of alternative promoters of Rab3a.

5. Conclusions

In summary, in this study we have confirmed that the expression of Rab3a mRNA in primary cortical neurons is activity-dependent and related to DNA demethylation mediated by Tet3. Moreover, we found accumulation of 5hmC was associated with induction of H3K4me3 and H3K27ac occupancy, which suggests an active euchromatin structure at the Rab3a promoter. Lentivirus-mediated Tet3 knockdown blocked the effect of DNA demethylation at the Rab3a promoter and reduced the expression of Rab3a mRNA. Therefore, these findings emphasize the critical role of Tet3 in promoting the active epigenetic state and lead to an induction of Rab3a expression in post-mitotic neurons. However, future work will also be needed to translate our in vitro findings into in vivo experiments with the goal of assessing the comprehensive roles of Tet3 in regulating Rab3a during learning and memory formation processes.


TET, Ten-eleven translocation; 5hmC, 5-hydroxymethylcytosine; 5fC, 5-formylcytosine, 5caC, 5-carboxylcytosine; ChIP, Chromatin immunoprecipitation; E18, embryonic day 18; DIV, days in vitro; MeDIP, methylated DNA immunoprecipitation; hMeDIP, hydroxymethylated DNA immunoprecipitation; FAIRE, Formaldehyde Assisted Isolation of Regulatory Elements; APC, Adenomatous polyposis coli. DSBH, double-stranded β-folded domain; Cys-rich, cysteine-rich; GTP, guanosine triphosphate; GDP, guanosine diphosphate; GEPs, guanine nucleotide exchange proteins, GAPs, GTPase-activating proteins; GDIs, GDP dissociation inhibitors; BDNF, brain-derived neurotrophic factor-induced.

Author contributions

KL, LSW carried out the experiments and wrote the manuscripts. YC wrote and edited manuscript and performed RT-qPCR experiments. KL conceived the study, designed experiments and wrote the manuscript. YLS, SHS performed the acquisition, analysis and interpretation of data.

Ethics approval and consent to participate

Not applicable.


The authors gratefully acknowledge grant support from Zhongnan Hospital Wuhan University Start-up grant to KL, The Huaihua Hospital Young Researcher Grant to KL. LSW has been supported by postgraduate scholarship from Wuhan University. The authors would also like to thank Jianjian Zhang and Jincao Chen for helpful comments and lively discussion.


This study was supported by grants from Zhongnan Hospital of Wuhan University Start-up Grant (PTXM2021006).

Conflict of interest

The authors declare no conflict of interest.

Rousseau-Merck MF, Zahraoui A, Bernheim A, Touchot N, Miglierina R, Tavitian A, et al. Chromosome mapping of the human ras-related rab3a gene to 19p13.2. Genomics. 1989; 5: 694–698.
Takai Y, Sasaki T, Shirataki H, Nakanishi H. Rab3a small GTP-binding protein in Ca(2+)-dependent exocytosis. Genes to Cells. 1996; 1: 615–632.
Raiborg C, Stenmark H. Plasma membrane repairs by small GTPase Rab3a. Journal of Cell Biology. 2016; 213: 613–615.
Ishizaki H, Miyoshi J, Kamiya H, Togawa A, Tanaka M, Sasaki T, et al. Role of rab GDP dissociation inhibitor alpha in regulating plasticity of hippocampal neurotransmission. Proceedings of the National Academy of Sciences of the United States of America. 2000; 97: 11587–11592.
Star EN, Newton AJ, Murthy VN. Real-time imaging of Rab3a and Rab5a reveals differential roles in presynaptic function. Journal of Physiology. 2005; 569: 103–117.
Cheng Y, Wang J, Wang Y, Ding M. Synaptotagmin 1 directs repetitive release by coupling vesicle exocytosis to the Rab3 cycle. ELife. 2015; 4.
Schlüter OM, Basu J, Südhof TC, Rosenmund C. Rab3 superprimes synaptic vesicles for release: implications for short-term synaptic plasticity. Journal of Neuroscience. 2006; 26: 1239–1246.
Yang S, Farias M, Kapfhamer D, Tobias J, Grant G, Abel T, et al. Biochemical, molecular and behavioral phenotypes of Rab3a mutations in the mouse. Genes, Brain, and Behavior. 2007; 6: 77–96.
Huang M, Darvas M, Keene CD, Wang Y. Targeted Quantitative Proteomic Approach for High-Throughput Quantitative Profiling of Small GTPases in Brain Tissues of Alzheimer’s Disease Patients. Analytical Chemistry. 2019; 91: 12307–12314.
Chen Z, Riggs AD. DNA methylation and demethylation in mammals. Journal of Biological Chemistry. 2011; 286: 18347–18353.
Xu L, Chen Y, Nakajima S, Chong J, Wang L, Lan L, et al. A Chemical Probe Targets DNA 5-Formylcytosine Sites and Inhibits TDG Excision, Polymerases Bypass, and Gene Expression. Chemical Science. 2014; 5: 567–574.
Li X, Wei W, Ratnu VS, Bredy TW. On the potential role of active DNA demethylation in establishing epigenetic states associated with neural plasticity and memory. Neurobiology of Learning and Memory. 2013; 105: 125–132.
Bachman M, Uribe-Lewis S, Yang X, Burgess HE, Iurlaro M, Reik W, et al. 5-Formylcytosine can be a stable DNA modification in mammals. Nature Chemical Biology. 2015; 11: 555–557.
Alaghband Y, Bredy TW, Wood MA. The role of active DNA demethylation and Tet enzyme function in memory formation and cocaine action. Neuroscience Letters. 2014; 625: 40–46.
Wu X, Zhang Y. TET-mediated active DNA demethylation: mechanism, function and beyond. Nature Reviews. Genetics. 2017; 18: 517–534.
Pan X, Zheng L. Epigenetics in modulating immune functions of stromal and immune cells in the tumor microenvironment. Cellular & Molecular Immunology. 2020; 17: 940–953.
Bronowicka-Kłys DE, Roszak A, Pawlik P, Sajdak S, Sowińska A, Jagodziński PP. Transcript levels of ten-eleven translocation type 1-3 in cervical cancer and non-cancerous cervical tissues. Oncology Letters. 2017; 13: 3921–3927.
Kohli RM, Zhang Y. TET enzymes, TDG and the dynamics of DNA demethylation. Nature. 2013; 502: 472–479.
Szwagierczak A, Bultmann S, Schmidt CS, Spada F, Leonhardt H. Sensitive enzymatic quantification of 5-hydroxymethylcytosine in genomic DNA. Nucleic Acids Research. 2010; 38: e181.
Li X, Wei W, Zhao Q, Widagdo J, Baker-Andresen D, Flavell CR, et al. Neocortical Tet3-mediated accumulation of 5-hydroxymethylcytosine promotes rapid behavioral adaptation. Proceedings of the National Academy of Sciences of the United States of America. 2014; 111: 7120–7125.
Simon JM, Giresi PG, Davis IJ, Lieb JD. Using formaldehyde-assisted isolation of regulatory elements (FAIRE) to isolate active regulatory DNA. Nature Protocols. 2012; 7: 256–267.
Papale LA, Zhang Q, Li S, Chen K, Keleş S, Alisch RS. Genome-wide disruption of 5-hydroxymethylcytosine in a mouse model of autism. Human Molecular Genetics. 2015; 24: 7121–7131.
Strahl BD, Allis CD. The language of covalent histone modifications. Nature. 2000; 403: 41–45.
Kriukienė E, Liutkevičiūtė Z, Klimašauskas S. 5-Hydroxymethylcytosine—he elusive epigenetic mark in mammalian DNA. Chemical Society Reviews. 2012; 41: 6916–6930.
Głowacki S, Błasiak J. Role of 5-hydroxymethylcytosine and TET proteins in epigenetic regulation of gene expression. Postepy Biochemii. 2013; 59: 64–69.
Tahiliani M, Koh KP, Shen Y, Pastor WA, Bandukwala H, Brudno Y, et al. Conversion of 5-methylcytosine to 5-hydroxymethylcytosine in mammalian DNA by MLL partner TET1. Science. 2009; 324: 930–935.
Globisch D, Münzel M, Müller M, Michalakis S, Wagner M, Koch S, et al. Tissue distribution of 5-hydroxymethylcytosine and search for active demethylation intermediates. PLoS ONE. 2010; 5: e15367.
Song C, Szulwach KE, Fu Y, Dai Q, Yi C, Li X, et al. Selective chemical labeling reveals the genome-wide distribution of 5-hydroxymethylcytosine. Nature Biotechnology. 2011; 29: 68–72.
Mellén 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; 151: 1417–1430.
Feng S, Jacobsen SE, Reik W. Epigenetic reprogramming in plant and animal development. Science. 2010; 330: 622–627.
Pfaffeneder T, Hackner B, Truss M, Münzel M, Müller M, Deiml CA, et al. The discovery of 5-formylcytosine in embryonic stem cell DNA. Angewandte Chemie. 2011; 50: 7008–7012.
He Y, Li B, Li Z, Liu P, Wang Y, Tang Q, et al. Tet-mediated formation of 5-carboxylcytosine and its excision by TDG in mammalian DNA. Science. 2011; 333: 1303–1307.
Chen Y, Damayanti NP, Irudayaraj J, Dunn K, Zhou FC. Diversity of two forms of DNA methylation in the brain. Frontiers in Genetics. 2014; 5: 46.
Robertson J, Robertson AB, Klungland A. The presence of 5-hydroxymethylcytosine at the gene promoter and not in the gene body negatively regulates gene expression. Biochemical and Biophysical Research Communications. 2011; 411: 40–43.
Xu Y, Wu F, Tan L, Kong L, Xiong L, Deng J, et al. Genome-wide regulation of 5hmC, 5mC, and gene expression by Tet1 hydroxylase in mouse embryonic stem cells. Molecular Cell. 2011; 42: 451–464.
Barski A, Cuddapah S, Cui K, Roh T, Schones DE, Wang Z, et al. High-resolution profiling of histone methylations in the human genome. Cell. 2007; 129: 823–837.
Villar D, Berthelot C, Aldridge S, Rayner TF, Lukk M, Pignatelli M, et al. Enhancer evolution across 20 mammalian species. Cell. 2015; 160: 554–566.
Portela A, Esteller M. Epigenetic modifications and human disease. Nature Biotechnology. 2010; 28: 1057–1068.
Coleman WL, Bill CA, Bykhovskaia M. Rab3a deletion reduces vesicle docking and transmitter release at the mouse diaphragm synapse. Neuroscience. 2007; 148: 1–6.
Fukui K, Sasaki T, Imazumi K, Matsuura Y, Nakanishi H, Takai Y. Isolation and characterization of a GTPase activating protein specific for the Rab3 subfamily of small G proteins. Journal of Biological Chemistry. 1997; 272: 4655–4658.
Nagano F, Sasaki T, Fukui K, Asakura T, Imazumi K, Takai Y. Molecular cloning and characterization of the noncatalytic subunit of the Rab3 subfamily-specific GTPase-activating protein. Journal of Biological Chemistry. 1998; 273: 24781–24785.
Ullrich O, Stenmark H, Alexandrov K, Huber LA, Kaibuchi K, Sasaki T, et al. Rab GDP dissociation inhibitor as a general regulator for the membrane association of rab proteins. Journal of Biological Chemistry. 1993; 268: 18143–18150.
Matsui Y, Kikuchi A, Araki S, Hata Y, Kondo J, Teranishi Y, et al. Molecular cloning and characterization of a novel type of regulatory protein (GDI) for smg p25a, a ras p21-like GTP-binding protein. Molecular and Cellular Biology. 1990; 10: 4116–4122.
Brondyk WH, McKiernan CJ, Burstein ES, Macara IG. Mutants of Rab3a analogous to oncogenic Ras mutants. Sensitivity to Rab3a-GTPase activating protein and Rab3a-guanine nucleotide releasing factor. Journal of Biological Chemistry. 1993; 268: 9410–9415.
Park J, Kim J, Lee J, Kim J, Seo J, Kim A. GTP binds to Rab3a in a complex with Ca2+/calmodulin. Biochemical Journal. 2002; 362: 651–657.
Thakker-Varia S, Alder J, Crozier RA, Plummer MR, Black IB. Rab3a is required for brain-derived neurotrophic factor-induced synaptic plasticity: transcriptional analysis at the population and single-cell levels. Journal of Neuroscience. 2001; 21: 6782–6790.
Wang X, Wang Q, Yang S, Bucan M, Rich MM, Engisch KL. Impaired activity-dependent plasticity of quantal amplitude at the neuromuscular junction of Rab3a deletion and Rab3a earlybird mutant mice. Journal of Neuroscience. 2011; 31: 3580–3588.
Li X, Marshall PR, Leighton LJ, Zajaczkowski EL, Wang Z, Madugalle SU, et al. The DNA Repair-Associated Protein Gadd45γ Regulates the Temporal Coding of Immediate Early Gene Expression within the Prelimbic Prefrontal Cortex and Is Required for the Consolidation of Associative Fear Memory. Journal of Neuroscience. 2019; 39: 970–983.
Publisher’s Note: IMR Press stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Back to top