Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Characterization of Acetylation of Histone H3 at Lysine 9 in the Trigeminal Ganglion of a Rat Trigeminal Neuralgia Model

Characterization of Acetylation of Histone H3 at Lysine 9 in the Trigeminal Ganglion of a Rat... Hindawi Oxidative Medicine and Cellular Longevity Volume 2022, Article ID 1300387, 13 pages https://doi.org/10.1155/2022/1300387 Research Article Characterization of Acetylation of Histone H3 at Lysine 9 in the Trigeminal Ganglion of a Rat Trigeminal Neuralgia Model 1,2,3,4,5 1,2,3,4,5 1,2,3,4,5 1,2,3,4,5 Wenbin Wei , Yuemin Liu , Yating Qiu , Minjie Chen , 1,2,3,4,5 1,2,3,4,5 1,2,3,4,5 Yiwen Wang , Zixiang Han , and Ying Chai Department of Oral Surgery, Shanghai Ninth People’s Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China College of Stomatology, Shanghai Jiao Tong University, Shanghai, China National Center for Stomatology, Shanghai, China National Clinical Research Center for Oral Diseases, Shanghai, China Shanghai Key Laboratory of Stomatology, Shanghai, China Correspondence should be addressed to Yating Qiu; qiuyating@hotmail.com and Minjie Chen; chenmj_9hospital@126.com Received 2 March 2022; Accepted 12 April 2022; Published 4 May 2022 Academic Editor: Ajinkya Sase Copyright © 2022 Wenbin Wei et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Trigeminal neuralgia (TN) is a chronic neuropathic pain disorder characterized by spontaneous and elicited paroxysms of electric- shock-like or stabbing pain in a region of the face. The epigenetic regulation of TN is still obscure. In current study, a rat TN model subject to carbamazepine (CBZ) treatment was established, and transcriptome- and genome-scale profiling of H3K9ac and HDAC3 was performed by RNA-seq and ChIP-seq. We observed that H3K9ac levels in the trigeminal ganglion were lower in the TN rats compared with those in the control, and CBZ treatment led to recovery of H3K9ac levels. Further, we found that HDAC3 was overactivated, which interfered with H3K9 acetylation due to higher phosphorylation in TN compared with that in the control. Finally, the phosphokinase leucine-rich repeat kinase 2 (LRRK2) was demonstrated to contribute to HDAC3 activity via the MAPK signaling pathway. Taken together, we identified a regulatory mechanism in which the phosphate groups transferred from activated ERK and LRRK2 to HDAC3 caused genome-scale deacetylation at H3K9 and resulted in the silencing of a large number of genes in TN. The kinases or important enzymes within this regulatory axis may represent important targets for TN therapy and prevention. TN [2]. In addition, the pathogenesis of TN involves neuro- 1. Introduction degeneration, epilepsy, and infection. A large number of Trigeminal neuralgia (TN) manifests as transient, paroxys- neurophysiological, neuroimaging, and histological studies mal but severe pain that occurs in the trigeminal nerve dis- show that the underlying pathophysiological mechanism in tribution area. TN is a common cause of facial pain in TN involves the focal demyelination of primary trigeminal individuals aged over 50 years old. The incidence of TN in afferents near the entry of the trigeminal root into the pons females is higher than in males. TN is frequently both mis- [3, 4]. Such demyelination results in hyperexcitability of the diagnosed and underdiagnosed, and its incidence is variably focally demyelinated primary afferents due to the insuffi- reported between studies, with a range from 4.3 to 27 new ciency of prompt resting potential re-establishment in axons. cases per 100,000 people per year [1]. The compression of An epigenetic trait is a stably heritable phenotype result- ing from changes in a chromosome without alterations in blood vessels at the trigeminal nerve has been observed in more than 30% of TN patients. The academic community the DNA sequence. Histone H3 is one of the most exten- has reached broad agreement that a neurovascular conflict sively modified of the five primary histone proteins, and with compression is the most common cause of classical acetylation at different lysine residues of histone H3 may 2 Oxidative Medicine and Cellular Longevity 20 minutes after the rats adapted to the test cage. The face play key roles in gene regulation. A previous study has reported larger numbers of acetylated H3K9-, H3K18-, and grooming behavior was analyzed offline by an independent H3K27-positive cells in the TN group than in the sham- observer who was blind to the treatment administered to operation group [2, 5]. Interestingly, the notion that H3K9 the rats. For each observation session of the rats over 10 global acetylation decreases in the Schwann cells in the tri- minutes, the total duration of asymmetrical face grooming geminal ganglia of a nerve injury mouse model induced by episodes as unilateral strokes with the dominant paw was trigeminal inflammatory compression [6], whereas that his- counted [14]. For open field test, the animals were placed tone deacetylase inhibitors (HDACi) can prevent the persis- into an open-field hardboard box for 6 min individually to tent hypersensitivity of neurons from neuropathic pain [7], assess the motor activity, exploratory behavior, and emo- which indicate that cells in different areas of trigeminal root tional state. During the original period, 1 min was allowed entry zone and ganglion contain display a different response for accommodation, and the next 5 min was for recording, to TN. Moreover, various HDAC subclusters have also been including both horizontal and vertical movement and dwell found to exert varying effects on pain vulnerability and time. In addition, the mice were allowed optional movement excitability of neurons [8, 9]. Therefore, the epigenetic regu- in the box during the experiment. lation of the gene expression in response to compressive Rats were anesthetized with 60 mg/kg pentobarbital and stress-induced imbalanced homeostasis in the trigeminal shaved the hair and made a mid-line scalp incision with ganglion and trigeminal root entry zone is still obscure. equal distances anterior and posterior to the center of the In the current study, we constructed a rat model of TN eyes to expose the skull and nasal bone. After dissecting the edge of the orbit free formed by the maxillary, frontal, subject to carbamazepine (CBZ) treatment a widely used antiepileptic drug in clinical practice and performed tran- lacrimal, and zygomatic bones, infraorbital nerve was dis- scriptome- and genome-scale H3K9ac profiling. CBZ as the sected freely from the surrounding connective tissue for first-line pharmacological treatment for trigeminal neuralgia the consequent experiments. The animal study was reviewed [10] has been determined to function as HDACi [11]. Our and approved by the Medical Ethics Committee of Shanghai Jiao Tong University School of Medicine. study attempted to determine the alterative H3K9ac signa- ture and elucidate the role of the HDAC3 in TN [6, 7]. The large number of differentially expressed genes (DEGs) as 2.2. Immunofluorescence (IF) Assay. Frozen trigeminal well as the differential H3K9ac and HDAC3 signals across nerves sections (10 μm) were cut by a cryostat (Thermo the whole genome identified in this study reveals novel con- Fisher Scientific). Unlike the nerve root region with the sur- rounding distribution of nuclei, the trigeminal ganglion con- nections between H3K9ac and gene expression and the underlying mechanism of HDAC3 in H3K9ac modulation, tains robust nuclei in central region. Sections were fixed with emphasizing the importance of epigenetic alterations in reg- 4% solution of paraformaldehyde and washed with ulating transcription during TN. phosphate-buffered saline (PBS) and then permeabilized with 0.1% Triton-X 100 (Millipore) and blocked with 5% horse serum in PBS. Immunostaining of samples was per- 2. Materials and Methods formed using antibodies against H3K9ac (1: 500, CST, #9649) or GFAP (1: 200, CST, #3670) overnight at 4 2.1. Animal Study. Adult female Sprague-Dawley rats (200– C. After 220 g) were randomly divided into two groups: one group washing with PBS for 4 times (5 min per time), the second- received chronic compression of the trigeminal nerve root ary goat antibodies against rabbit (for H3K9ac) or mouse only at left side (n =30), and the other sham-operated group (for GFAP) (1 : 20000, Jackson ImmunoResearch) were used, was considered as the negative control (n =20). The TN and incubation was performed for 2 h at room temperature; model was prepared as previously described [12]. Five days this was followed with further incubation with DAPI for after TN model establishment, the TN group was further 15 min and washing with PBS four times. After drying in divided into two groups: one was treated with 100 mg/kg/d air, mounting medium (Thermo Fisher Scientific) was added CBZ (Sigma) dissolved into corn oil as previous described dropwise to the tissue slices, and they were coated with cover [13] for an additional five days (n =15). All rats were sacri- slips. The images were captured by fluorescence microscopy ficed in the tenth day. DM2000 (Leica Biosystems). Positive staining was statisti- Orofacial mechanical allodynia testing was conducted at cally analyzed using ImageJ. the same side of injured tergeminal nerve as previously described [5]. Rats were lightly restrained by their tail and 2.3. RNA Sequencing (RNA-seq). Trigeminal ganglion tissues allowed to stand on a mesh platform. The von Frey filaments were stored in 1 mL TRIzol (Thermo Fisher Scientific) and (Ugo Basile) were applied to their left vibrissal pad border ground in liquid nitrogen. Next, 100 μL chloroform was and cheek to collect baseline mechanical thresholds. Stimu- added, and the cells were fully mixed and centrifuged at lation always began with the filament that produced the low- the highest speed at 4 C for 10 min. The supernatant was est force (bending forces of 0.008–6 g) and stopped when the transferred into a new tube, isopropanol was added to the threshold was found within the vibrissal pad of the rats. Each same volume, and cells were centrifuged at the highest speed rat was tested five times at intervals of 3–5 s. The face at 4 C for 10 min. The precipitate was washed with 75% cold grooming episodes (ipsilateral, contralateral, or bilateral face ethanol and dissolved in DEPC water. The concentration grooming) of the rats in the three groups were recorded by a and quality of RNA were measured using a Nanodrop video camera. Video recording of face grooming was started 2000 (Thermo Fisher Scientific) and an Agilent Bioanalyzer Oxidative Medicine and Cellular Longevity 3 were harvested and mixed with 1 μg phosphoserine/threo- 2100 (Agilent). A total of 4 μg of RNA from each group was used for library preparation using the NEBNext Ultra Direc- nine (Abcam, ab17464), HDAC3 (CST, #85057), LRRK2 tional RNA Library Prep Kit for Illumina (NEB) following (CST, #13046), or IgG Rabbit IgG antibody and 40μl flurry the manufacturer’s instructions and sequenced on an Illu- IgA beads (Thermo Fisher Scientific) for rotating overnight mina HiSeq platform. at 4 C. Immunoprecipitates were washed with IP buffer The raw data were trimmed for adaptors, and low- (20 mM HEPES (pH 7.9), 350 mM NaCl, 0.1% NP-40, 1 mM quality reads were filtered out using Trimmomatic [15]. DTT, 0.2 mM PMSF, 2 mg/mL leupeptin, and 2 mg/mL apro- The quality of clean reads was checked using FastQC [16]. tinin) and purified with RIPA buffer (50 mM Tris (pH 7.4), Next, clean reads were aligned to the latest rat genome assem- 150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate, 0.1% bly Rnor6.0 using Hisat2 [17]. The transcripts were assembled, SDS) with 1% proteasome inhibitor cocktail and 1% PMSF. and the expression levels were estimated with FPKM values The protein lysate was subjected to western blot assay. using the StringTie algorithm with default parameters [18]. Differential mRNA and lncRNA expression among the groups 2.6. Western Blot (WB) Assay. The protein lysate was then were evaluated using the R package Ballgown [19], and the sig- analyzed by SDS-PAGE and transferred to PVDF mem- nificance of differences was computed using the Benjamini branes (Bio-Rad Laboratories, Hercules, USA). The mem- and Hochberg (BH) p value adjustment method [20]. Gene brane was blocked with 5% fat-free milk in PBST for annotation was performed using the Ensembl Genome 30 min, followed by incubation overnight at 4 C with final Browser database (http://www.ensembl.org/index.html). The dilutions of primary antibodies against p-HDAC1 (Ser421, R package ClusterProfiler was used to annotate the DEGs Ser423) (1: 1000, Thermo Fisher Scientific, PA5-36911), using Gene Ontology (GO) terms and Kyoto Encyclopedia HDAC1 (1: 2000, Abclonal, A19571), p-HDAC2 (Ser394) of Genes and Genomes (KEGG) pathways [21]. (1: 1000, Thermo Fisher Scientific, PA5-105021), HDAC2 (1: 2000, Abclonal, A19626), p-HDAC3 (Ser424) (1: 1000, 2.4. Chromatin Immunoprecipitation Sequencing (ChIP-Seq). Thermo Fisher Scientific, PA5-99339), HDAC3 (1: 2000, In brief, trigeminal ganglion tissues were incubated with Abclonal, A19537), p-HDAC8 (Ser29) (1: 1000, Thermo 1mL IP buffer (20 mM HEPES (pH 7.9), 350 mM NaCl, Fisher Scientific, PA5-105031), HDAC8 (1: 2000, Abclonal, 0.1% NP-40, 1 mM DTT, 0.2 mM PMSF, 2 mg/mL leupeptin, A5829), p-CK2β (Ser209) (1: 1000, Thermo Fisher Scientific, and 2 mg/mL aprotinin) [22, 23] on ice for 30 min; then, the 44-1090G), CK2 (1: 2000, Abclonal, A2869), p-LRRK2 nuclei were harvested by centrifugation at the highest speed. (Ser935) (1: 1000, Thermo Fisher Scientific, PA5-114599), The supernatant containing nuclei were sonicated to break LRRK2 (1: 2000, Abclonal, A17253), p-SRC (Tyr529) (1: up the genomic DNA into 200–500 bp fragments. Next, 1000, Thermo Fisher Scientific, PA5-97356), SRC (1: 2000, 10% lysates were saved as input, and the remaining were Abclonal, A19119), p-ERK1/2 (Thr202, Tyr204)(1: 1000, incubated with 1 μg IP-grade antibodies of H3K9ac (CST, Abclonal, AP1120), ERK1/2 (1: 2000, Abclonal, A19630), #9649) and HDAC3 (Abclonal, A19537) at 4 C overnight. p-MEK (Ser271, Tyr275) (1: 1000, Abclonal, AP0535), This was followed by incubation for 2 h with protein A beads MEK (1: 2000, Abclonal, A12950), or GAPDH (1: 2000, Pro- at 37 C to pull down the bound DNA fragments. teintech Group, #60004-1). Next, the membrane was washed ′ three times and then incubated with HRP-conjugated sec- For high-throughput sequencing, we added 3 -dA over- ondary antibodies (Proteintech Group). Membranes could hangs to the H3K9ac or HDAC3 enriched or input DNA be stripped using stripping buffer (Abcam, ab270550) at and ligated them to the adapter to build a DNA library. DNA libraries with ligated adapters were isolated based on 52 C for 30 min via gently shaking, washed by PBST, then blocked by 5% fat-free milk for 2 h, and reincubated by anti- the appropriate size for sequencing, using the Illumina bodies at 4 Hiseq2000 platform. The raw sequence reads of input and C overnight. The blotting bands were developed with ECL plus immunoblotting detection reagents (Thermo IP were trimmed based on adaptors, and low-quality reads Fisher Scientific) using UVP Chemstudio Plus System (Ana- were filtered out using Cutadapt (v1.9.1) and Trimmomatic (v0.35). The quality of the clean reads was checked using lytik Jena) and captured using ImageJ. Fastqc. The clean reads were mapped to the rat genome (assembly Rnor6.0) using the Bowtie 2 (v2.2.6) algorithm 2.7. HDAC Activity Assay. HDAC activity/inhibition assay kits for HDAC1 (Abx155637), HDAC2 (Abx258089), HDAC3 [24], and peak calling (p <0:01) was performed using MACS (Abx257885), and HDAC8 (Abx515264) were purchased 2 (v2.1.1) [25]. The differentially bound genes were analyzed based on p values less than 0.05 and annotated using Diff- from AmyJet Scientific Co. Ltd. Trigeminal ganglion tissues were incubated with RIPA buffer (50 mM Tris (pH 7.4), Bind [26]. The relevant peaks on the genomic loci were visu- 150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate, 0.1% alized using the Integrative Genomics Viewer (IGV). GO SDS) with 1% proteasome inhibitor cocktail and 1% PMSF. analysis was used to determine the biological functions of The protein concentration was quantified using the BSA genes associated with the differential peaks [27]. The raw ChIP sequencing data were submitted to the ArrayExpress method. The volume of different samples was adjusted to ensure that the total protein was identical. HDAC activities database and registered under the accession number E- were processed following the manufacturer’sinstructions. MTAB-10792 and 10793. 2.5. Immunoprecipitation (IP). Previous steps were similar to 2.8. Statistical Analysis. Data are presented as the mean ± those followed for the ChIP-seq assay. Trigeminal ganglia standard deviation for three independent experiments. The 4 Oxidative Medicine and Cellular Longevity 5 40 p = 0.042 p = 0.018 p = 0.025 p = 0.009 0 0 Control TN CBZ Control TN CBZ (n = 20) (n = 15) (n = 15) (n = 20) (n = 15) (n = 15) (a) (b) 40 30 p = 0.023 p = 0.012 25 p = 0.038 p = 0.027 20 15 0 0 Control TN CBZ Control TN CBZ (n = 20) (n = 15) (n = 15) (n = 20) (n = 15) (n = 15) (c) (d) 50 𝜇m 50 𝜇m 50 𝜇m GFAP Control TN CBZ (e) Figure 1: Confirmation of the TN rat model. (a)–(d) Behavior examination of orofacial mechanical allodynia by stimulation threshold and head withdrawal frequency and face grooming as well as open field test by central dwell time in the control (n =20), TN rat model (n =15), and TN rat model administered CBZ treatment (n =15). The p values were given by the comparison between control and TN, as well as TN and CBZ. (e) GFAP staining in trigeminal root entry zone by IF (×200 magnification); white arrows indicate distal extension of GFAP- positive astrocytic process in the trigeminal root. grooming behavior was also significantly increased after the differences in values were analyzed using one-way ANOVA. Statistical significance was set at p value less than 0.05. All operation and alleviated at day 10 with CBZ treatment analysis is performed by SPSS 20.0. (F =6:244, p =0:012; F =3:055, p =0:038) (Figure 1(c)). Open field test showed that CBZ could compromise the hyperactivity from TN effect (F =3:628, p =0:023; F = 3. Results 3:314, p =0:027) (Figure 1(d)). Nonetheless, we noticed that 3.1. Decreased H3K9ac Profiling in Trigeminal Ganglia in TN CBZ did not cause behavioral retardation in rats compared to Rats. Thirty adult female SD rats were subjected to mechan- control, indicating that the probable side effects of CBZ on ical compression force of the trigeminal root entry zone by rats’ activity with the current dose did not exist in our sys- retrograde insertion of a plastic filament from the right infe- tem. Moreover, a distal extension of glial fibrillary acidic rior orbital fissure for the TN model. After five days, half of protein (GFAP) staining at the trigeminal root entry zone them were treated with 100 mg/kg/d CBZ for an additional in the TN model at day 5 relative to the control indicated five days. Behavioral examination of orofacial mechanical that astrocytes started to break through the boundary allodynia showed that, compared with the 20 rats in the con- established by Schwann cells (Figure 1(e)). These results trol group, stimulus intensity (F =6:681, p =0:009; F = suggest the successful establishment of a rat TN model for subsequent experiments. 4:412, p =0:018) and mechanical hyperalgesia (F =3:447, p =0:025; F =2:893, p =0:042) were increased after the Next, H3K9ac was investigated by IHC and ChIP-seq. operation and attenuated at day 10 following CBZ treatment We observed fewer H3K9ac-positive nuclei within the tri- (Figures 1(a) and 1(b)). Likewise, the frequency of face- geminal ganglia of the TN model than in the control group Mechanical stimulation Face grooming time (s) threshold (g) Central dwell time (s) Latencies of head withdrawal (s) Oxidative Medicine and Cellular Longevity 5 Control TN CBZ 50 𝜇m 50 𝜇 m 50 𝜇 m DAPI p = 0.041 50 𝜇m 50 𝜇 m 50 𝜇 m H3K9ac p = 0.037 50 𝜇m 50 𝜇 m 50 𝜇 m Merge (a) 0.6 35,000 TN vs. CBZ, p = 0.012 30,000 TN vs. Control, p = 0.005 25,000 0.4 20,000 p = 0.003 15,000 p < 0.001 0.2 10,000 5,000 Control TN CBZ 0.0 TSS 2.0 Kb –2.0 Kb Control TN CBZ (b) (c) 100 Control TN CBZ 1,026 521 Th Control TN CBZ Promoter Intron 5′UTR 3′UTR Exon Distal intergenic regions (d) (e) Figure 2: Overview of H3K9ac enrichment in the TN rat model. (a) Global H3K9ac staining in the trigeminal ganglion of the control, TN rat model, and TN rat model administered CBZ treatment (×200 magnification). The p-values were given by the comparison of H3K9ac positive puncta occupied in all cells between control and TN, as well as TN and CBZ. (b) Overall calling peaks of H3K9ac in the rat genome of control rats, TN rat model, and TN rat model administered CBZ treatment, by ChIP-seq. The p values were given by the comparison of genomic H3K9ac peak counts between control and TN, as well as TN and CBZ. (c) Metagene profiles of genome-wide H3K9ac in the control, TN rat model, and TN rat model administered CBZ treatment. (d) The distribution of differentially H3K9ac- enriched regions in the genomic contexts of the control, TN rat model, and TN rat model administered CBZ treatment. (e) Diagram showing the intersection of genes with differential H3K9ac enrichment in the control, TN rat model, and TN rat model administered CBZ treatment. Genome-wide H3K9ac peaks e genomic distribution of H3K9ac peaks (%) Read count per million mapped reads H3K9ac positive puncta/cell (%) Control TN CBZ 6 Oxidative Medicine and Cellular Longevity Control TN CBZ 2,104 1,672 1,522 (a) 0.6 0.4 r = 0.487 p = 0.002 0.2 0.0 –0.2 Promoter 5′UTR Exon Intron 3′UTR Distal intergenic region (b) Biological process Activation of MAPKK activity Positive regulation of peptidyl-tyrosine phosphorylation Neuroinflammatory response Regulation of cell aging Cellular response to nutrient Axon extension Negative regulation of dendrite development Sensory perception of pain GTP metabolic process Positive regulation of histone modification Negative regulation of protein oligomerization Positive regulation of dephosphorylation Central nervous system neuron axonogenesis Negative regulation of phosphorylation Regulation of cellular response to growth factor stimulus 0 5 10 15 –log P (c) Figure 3: Continued. Pearson correlation coefficient (r) between H3K9ac and transcription Oxidative Medicine and Cellular Longevity 7 GO:0019233 GO:2000171 GO:1900015 GO:0050890 GO:0051930 GO:0031670 GO:0048675 GO:0015914 GO:1904029 GO:0031058 GO:0090342 GO:0032874 GO:0008654 GO:0006925 GO:0046039 GO:0045165 GO:0002534 GO:0021955 GO:0070932 GO:0000186 GO:0150076 GO:0006925: inflammatory cell apoptotic process GO:0150076: neuroinflammatory response GO:0046039: GTP metabolic process GO:0043409 GO:0043409: negative regulation of MAPK cascade GO:0070932: histone H3 deacetylation (d) Figure 3: Overview of the transcriptome in the TN rat model. (a) Diagram showing intersecting DEGs in the control, TN rat model, and TN rat model administered CBZ treatment. (b) Correlation between RNA levels and H3K9ac enrichment at different genomic regions of DEGs, by Pearson correlation analysis. (c) Gene ontology analysis of DEGs with differential H3K9ac enrichment, for biological processes. (d) Regulatory connections between different terms of biological processes by directed acyclic graph. The key intermediate and terminal GO items are labeled by circle and box. p value less than 0.001 and 0.0001 are highlighted by orange and red. (p =0:037) and in rats subjected to CBZ treatment ment were all included among the DEGs. Pearson correla- (p =0:041) (Figure 2(a)). Consistent with this result, tion analysis showed that only the differential H3K9ac genome-wide H3K9ac profiling also indicated weakened enrichment at promoter regions correlated positively with the expression of the corresponding genes (r =0:487, p = global H3K9ac peaks on neuron genomes in the TN group relative to the control (p <0:001) and CBZ treatment groups 0:002) (Figure 3(b)). Notably, GO analysis showed that these (p =0:003) (Figure 2(b)). Furthermore, H3K9ac density sug- genes were tightly linked with functional regulation on gested that a unique subset of target genes exhibited com- axons or dendrites as well as inflammatory cytokine stimu- promised H3K9ac enrichment mainly at promoter regions lus (Figure 3(c)). In particular, protein phosphorylated mod- (defined as −2 kb to +2 kb from the transcriptional start site ification, phosphate metabolism process, and GTPase- [TSS]) in TN relative to the control (p =0:005) and CBZ associated activity occupied the central position in the treatment groups (p =0:012) (Figures 2(c) and 2(d)). directed acyclic graph of functional regulation Numerous H3K9ac-targeted sites identified by examining (Figure 3(d)). Collectively, the data indicated that the abnor- the differences between control or CBZ treatment and TN mal H3K9ac contributed to the changes in the expression of were mapped to 477 genes, with 421 genes showing a decrease a large proportion of genes in TN. and 56 an increase in H3K9ac in TN (fold change ðFCÞ >2 or <0.5, p <0:05) (Figure 2(e)). Taken together, our findings 3.3. Excessive Activation of HDAC3 in Trigeminal Ganglia in suggested that TN causes reduced H3K9ac in the genomes TN. Given that previous studies report seemingly controver- of trigeminal ganglion neurons. sial findings on histone H3 acetylation in neuralgia that the positive staining of H3K9, H3K18, and H3K27 acetylation 3.2. Transcriptome of Trigeminal Ganglion Affected by were higher in postoperative days 7, 14, 21, and 28 in male H3K9ac in TN. To further study the expression of target TN rats [5], whereas the global acetylation of H3K9 genes affected by H3K9ac, we also conducted RNA-seq to decreased at day 21 by trigeminal inflammatory compres- investigate DEGs in TN. Based on the results, 164 genes sion in mouse trigeminal ganglia [6], it is important to study showed higher expression while 1,358 genes had lower the role of HDACs in regulating H3K9ac in TN. We focus expression in TN relative to the control and CBZ treatment on class I HDACs in the current study because of the evi- groups were characterized (FC > 2 or<0.5, p <0:05) dence that class I HDACs (HDAC1, 2, 3, and 8) are widely (Figure 3(a)). The genes with differential H3K9ac enrich- implicated in H3K9ac transition in multiple human diseases 8 Oxidative Medicine and Cellular Longevity phospho-Ser/Thr Input IgG Control TN CBZ Control TN CBZ Control TN CBZ GAPDH p-HDAC1 (Ser421/423) HDAC3 HDAC1 (b) p-HDAC2 (Ser394) HDAC activity assay HDAC2 p = 0.033 ^ p = 0.014 p-HDAC3 (Ser424) HDAC3 p-HDAC8 (Ser29) HDAC1 HDAC2 HDAC3 HDAC8 HDAC8 Control TN GAPDH CBZ (a) (c) Figure 4: HDAC3 activity by phosphorylation modification in TN rat model. (a) Phosphorylation of HDAC1, 2, 3, and 8 in the control, TN rat model, and TN rat model administered CBZ treatment, by WB assay. (b) Phosphorylated HDAC3 profiles in the control, TN rat model, and TN rat model administered CBZ treatment, by phosphoserine and threonine IP assay. (c) The activity of HDAC1, 2, 3, and 8 in the control, TN rat model, and TN rat model administered CBZ treatment, by enzyme-linked immunosorbent assay (ELISA); “ ” and “^” represent comparison between TN and control as well as CBZ and TN with statistical significance. and development [28–30]. The expression of class I HDACs the control and CBZ treatment groups (FC > 2 or<0.5, p < was studied in TN by WB assay. Here, the expression of 0:05) (Figure 5(c)). Hence, we identified three lists of genes these four HDACs showed no significant difference between with differentially occupied H3K9ac (477) and HDAC3 the control, TN, and CBZ treatment groups, and we unex- (637) as well as DEGs (1, 522) by comparison between the pectedly found that the phosphorylation of HDAC3 TN and control group, and obtained an intersection of 288 genes. The transcription of these 288 genes was positively (Ser424) was higher in the TN and CBZ treatment groups than in the control (Figure 4(a)). Furthermore, the phos- correlated with H3K9ac alteration (r =0:506, p =0:0017) phorylated serine/threonine pull-down experiment showed and negatively correlated with HDAC3 (r = −0:475, p = that HDAC3 was remarkably modified via phosphorylation 0:0021) compared between control and TN groups, as well in the TN and CBZ treatment groups relative to that in the as positively correlated with H3K9ac alteration (r =0:448, control (Figure 4(b)). However, CBZ treatment strongly p =0:0026) and negatively correlated with HDAC3 (r = attenuated the activity of HDAC3 relative to that of HDAC1, − 0:626, p =0:0013) compared between TN and CBZ 2, and 8 in the in vitro assay (Figure 4(c)). Unexpectedly, groups (Figure 5(d)) (Table S1). Neurod2, C2cd4c, and CBZ treatment produced a similar phosphorylation signa- Sf3b4, with the top correlation coefficient values, confirmed ture of HDAC3 as in the TN group, despite substantially the enrichment of H3K9ac and HDAC3 as well as their attenuating TN. We assumed that CBZ, being an HDACi, transcription using the IGV browser (Figure 5(e)). In did not exert its effects by blocking phosphorylation. conclusion, we found that the activated HDAC3 contributed Next, we performed ChIP-seq analysis of trigeminal gan- to the reduced H3K9ac modification and gene silencing in TN. glia to investigate the genome-wide occupancy of HDAC3 in TN. Opposite to the findings for enrichment of H3K9ac, 3.4. Phosphorylation of HDAC3 Regulated by LRRK2 via global HDAC3 occupancy was remarkably robust in TN Compressive Stress-Induced MAPK/ERK Signaling Pathway. compared with that in the control and CBZ treatment As previously described, three enzymes, namely, casein groups (Figure 5(a)). Our data indicate that the genomic kinase 2 (CK2) [31], leucine-rich repeat kinase 2 (LRRK2) binding ability of HDAC3 was indeed strengthened, espe- [32], and SRC protooncogene, nonreceptor tyrosine- cially at the promoter regions in TN relative to the con- protein kinase (SRC) [33], contributing to the phosphoryla- trol (p =0:019) and CBZ treatment groups (p =0:037) tion of HDAC3 or HDAC3-H1.3 complex were investigated (Figure 5(b)). Moreover, HDAC3 peaks were mapped to to elucidate the mechanism underlying the phosphorylation 637 genes, in which 564 had increased, whereas 73 had of HDAC3 in TN. Initially, we determined a substantial decreased HDAC3 occupancy in TN compared to that in interaction between HDAC3 and LRRK2 rather than CK2 Control TN CBZ H3K9ac enrichment (log ) Oxidative Medicine and Cellular Longevity 9 30,000 p < 0.001 20,000 p < 0.001 10,000 Control TN CBZ (a) Control TN CBZ 0.6 TN vs. CBZ, p = 0.037 TN vs. Control, p = 0.019 0.4 1,210 755 0.2 0.0 –2.0 Kb TSS 2.0 Kb Control TN CBZ (b) (c) –4 –8 –8 –4 4 2 –2 –4 8 –6 –8 Control TN CBZ (d) Figure 5: Continued. HDAC3 enrichment (log ) Read count per million mapped reads Genome-wide HDAC3 peaks Transcription (log ) 2 10 Oxidative Medicine and Cellular Longevity 200 200 200 Control 0 0 200 200 RNA TN 0 0 200 200 CBZ 0 0 200 200 Control 0 0 200 200 TN H3K9ac 0 0 200 200 CBZ 0 0 0 200 200 Control 0 200 200 200 HDAC3 TN 0 0 CBZ 0 0 Neurod2 C2cd4c Sf3b4 (e) Figure 5: Overview of HDAC3 in TN rat model. (a) Overall calling peaks of HDAC3 in the rat genome of the control, TN rat model, and TN rat model administered CBZ treatment, by ChIP-seq. (b) Metagene profiles of genome-wide HDAC3 in the control, TN rat model, and TN rat model administered CBZ treatment. (c) Diagram showing intersecting genes with differential HDAC3 enrichment in the control, TN rat model, and TN rat model administered CBZ treatment. (d) Correlation analysis of DEGs with differential H3K9ac and HDAC3 enrichment in the control, TN rat model, and TN rat model administered CBZ treatment. (e) Gene-browser views of transcriptome, H3K9ac, and HDAC3 profiles of Neurod2, C2cd4c, and Sf3b4 in the control, TN rat model, and TN rat model administered CBZ treatment. Control TN CBZ p-LRRK2 (Ser935) LRRK2 HDAC3 IP Input IgG Control TN CBZ Control TN CBZ Control TN CBZ p-CK2𝛽 (Ser209) CK2 CK2𝛽 LRRK2 p-SRC (Tyr529) SRC SRC HDAC3 GAPDH (a) (b) Control TN CBZ p-ERK1/2 (Thr202, Tyr204) LRRK2 IP Input IgG ERK1/2 p-MEK (Tyr275) ERK1/2 MEK LRRK2 GAPDH (c) (d) Figure 6: The activities of LRRK2 and MAPK signaling pathways in TN rat model. (a) The interaction between LRRK2 and HDAC3 in TN, by HDAC3 IP assay. (b) Phosphorylation of LRRK2, CK2β, and SRC in the control, TN rat model, and TN rat model administered CBZ treatment, by WB assay. (c) Interaction between LRRK2 and ERK1/2 in TN, by LRRK2 IP assay. (d) Phosphorylation of MEK and ERK1/ 2 in the control, TN rat model, and TN rat model administered CBZ treatment, by WB assay. Oxidative Medicine and Cellular Longevity 11 can suppress the activity of c-Jun to alleviate pain [37]. Fur- thermore, D-β-hydroxybutyric acid is reported to not only relieve mechanical and thermal hyperalgesia in rats but also to improve their motor function by reversing the presence of low acetylation and expression of FOXO3a, catalase, and SOD2 in the damaged area [38]. Substantial attenuating effects of CBZ on TN are observed in this study. The number of DEGs in our RNA-seq data, consistent with that reported in a previous study [39], is far higher than genes with differ- ential H3K9ac occupancy, indicating that in addition to H3K9ac, the acetylation levels of other histones also decline in TN. Besides the well-acknowledged inflammatory response and neuronal self-repair in TN [40], the biological functions associated with MAPK activity and the GTP metabolic pro- cess are highlighted in our sequencing data, prompting us to Compressive stress Phosphate group study the activity of HDACs from phosphorylation modifi- H3K9ac cation. We only observe obvious phosphorylation at Ser29 ERK1/2 LRRK2 of HDAC3 in this system. The IP experiments showed one HDAC3 clear band of HDAC3 within the phosphor-Ser/Thr pull down fraction. However, HDAC3 in general has multiple Figure 7: The graphic overview of this study. The phosphate serine and threonine phosphorylation sites [33]. Therefore, groups transported from activated MAPK/ERK pathway and we suspect that the activation of HDAC3 by phosphorylation LRRK2 to HDAC3 causes the repression of H3K9ac and results at different sites determines individual and specific functions in a large number of gene expression silencing in trigeminal in diseases and other biological events. ganglion induced by compressive stress. Consistent with sequencing data, the findings show that the MAPK/ERK signaling pathway substantially contributes or SRC, by IP study. In addition, the binding ability of to providing phosphate groups to LRRK2 and HDAC3. Here, LRRK2 with HDAC3 strengthened in TN and under CBZ we raise the interesting topic of kinase selection for HDAC3 treatment compared with that in the control (Figure 6(a)). phosphorylation in TN. CK2, LRRK2, and SRC are the phos- Next, the protein expression of these kinases did not change phokinases for other HDACs, and they govern the phos- between the control, TN, or CBZ treatment groups, but phorylation of HDAC3 in different tissues [41]. It is robust phosphorylation of LRRK2 (Ser935) was observed in challenging to elucidate the mechanism underlying such TN and CBZ treatment groups relative to the control behavior under special conditions. Furthermore, CK2 and (Figure 6(b)). To trace back to the phosphate groups from SRC are likely to play a role in other HDACs in the patho- LRRK2, LRRK2 pull-down and IP assays of trigeminal root genesis of TN. entry zone of TN were conducted to detect a variety of Overall, we identified a TN regulatory mechanism in enzymes that catalyze phosphate group transmission. We which the phosphate groups transferred from activated ERK and LRRK2 to HDAC3 caused genome-scale deacetylation noticed that MAPK/ERK signaling pathway was the top affected item in TN (Figure 3(c)) and was reported to partic- at H3K9 and resulted in the silencing of a large number of ipate in pathogenesis neuralgia in previous studies [34, 35]. genes (Figure 7). The kinases or important enzymes within We harvested ERK1/2 from the LRRK2 binding proteins this regulatory axis represent potential targets for TN therapy (Figure 6(c)). Consistent with this, the high expression of and prevention. phosphorylated MEK and ERK1/2 indicated that the MAPK/ERK signaling pathway was indeed activated in TN Data Availability and CBZ treatment groups compared with that in the con- trol (Figure 6(d)). The datasets used and/or analyzed during the current study are available from the corresponding authors upon reason- able request. 4. Discussion A large number of preclinical and clinical studies show that Conflicts of Interest histone acetyltransferases (HATs) and HDACs play an important role in the development of neuralgia [36]. Authors declare no conflict of interests. HDACs have a significant analgesic effect, thus drawing much attention from researchers. HDAC1 levels increase in Authors’ Contributions the spinal cord of the spinal nerve ligation rat model, where HDAC1 interacts with a heterodimer composed of c-Jun, W. W. and L. Y. performed all experiments and analyzed the finally activating the JNK signaling pathway to participate data. W. Y. assisted with the animal experiments. H. Z. in the maintenance of neuralgia. LG325 (HDAC1 inhibitor) assisted with the molecular experiments. C. Y. provided 12 Oxidative Medicine and Cellular Longevity [10] E. Gambeta, J. G. Chichorro, and G. W. Zamponi, “Trigeminal assistance with bioinformatics analysis. Q.Y. drafted and neuralgia: an overview from pathophysiology to pharmacolog- revised the manuscript. C. M. designed and supported the ical treatments,” Molecular Pain, vol. 16, article whole project and submitted the manuscript. Wenbin Wei 1744806920901890, 2020. and Yuemin Liu contributed equally to this work. [11] L. Akbarzadeh, T. Moini Zanjani, and M. Sabetkasaei, “Com- parison of anticancer effects of carbamazepine and valproic Acknowledgments acid,” Iranian Red Crescent Medical Journal, vol. 18, no. 10, article e37230, 2016. This work is supported by the Key Biological and Pharma- [12] D. S. Luo, T. Zhang, C. X. Zuo et al., “An animal model for tri- ceutical Projects of Shanghai Science and Technology Com- geminal neuralgia by compression of the trigeminal nerve mission (Grant No. 16411953300), the Clinical Research root,” Pain Physician, vol. 15, no. 2, pp. 187–196, 2012. MDT Project of Shanghai Ninth People’s Hospital, Shanghai [13] Y. Asai, Y. Sakakibara, M. Nadai, and M. Katoh, “Effect of car- Jiao Tong University (Grant No. 201701018), and the Sci- bamazepine on expression of UDP-glucuronosyltransferase ence and Technology Commission of Shanghai Municipality 1A6 and 1A7 in rat brain,” Drug Metabolism and Pharmacoki- (Grant No. 21Y11903500). netics, vol. 32, no. 6, pp. 286–292, 2017. [14] M. Cha, I. Sallem, H. W. Jang, and I. Y. Jung, “Role of transient receptor potential vanilloid type 1 in the trigeminal ganglion Supplementary Materials and brain stem following dental pulp inflammation,” Interna- tional Endodontic Journal, vol. 53, no. 1, pp. 62–71, 2020. Table S1: the list of DEGs with significant H3K9ac and [15] A. M. Bolger, M. Lohse, and B. Usadel, “Trimmomatic: a flex- HDAC3 enrichments. Candidate genes meet with the follow- ible trimmer for Illumina sequence data,” Bioinformatics, ing conditions, FPKM values significantly differential (FC > 2 vol. 30, no. 15, pp. 2114–2120, 2014. or<0.5, p <0:05) comparison between control and TN groups [16] S. Andrews and A. FastQC, Quality Control Tool for High as well as TN and CBZ groups in RNA-seq and ChIP-seq of Throughput Sequence Data, 2013. H3K9ac and HDAC3. (Supplementary Materials) [17] D. Kim, B. Langmead, and S. L. Salzberg, “HISAT: a fast spliced aligner with low memory requirements,” Nature References Methods, vol. 12, no. 4, pp. 357–360, 2015. [18] M. Pertea, G. M. Pertea, C. M. Antonescu, T. C. Chang, J. T. [1] S. Maarbjerg, G. Di Stefano, L. Bendtsen, and G. Cruccu, “Tri- Mendell, and S. L. Salzberg, “StringTie enables improved geminal neuralgia - diagnosis and treatment,” Cephalalgia: an reconstruction of a transcriptome from RNA-seq reads,” international journal of headache, vol. 37, no. 7, pp. 648–657, Nature Biotechnology, vol. 33, no. 3, pp. 290–295, 2015. [19] A. C. Frazee, G. Pertea, A. E. Jaffe, B. Langmead, S. L. Salzberg, [2] S. Peker, O. Kurtkaya, I. Uzun, and M. N. Pamir, “Microanat- and J. T. Leek, “Ballgown bridges the gap between tran- omy of the central myelin-peripheral myelin transition zone of scriptome assembly and expression analysis,” Nature Biotech- the trigeminal nerve,” Neurosurgery, vol. 59, no. 2, pp. 354– nology, vol. 33, no. 3, pp. 243–246, 2015. 359, 2006, discussion 354-9. [20] Y. Shen, S. Zhou, X. Zhao, H. Li, and J. Sun, “Characterization [3] H. Wood, “Trigeminal neuralgia linked to demyelination in of genome-wide DNA methylation and hydroxymethylation multiple sclerosis,” Nature Reviews. Neurology, vol. 16, no. 6, in mouse arcuate nucleus of hypothalamus during puberty p. 298, 2020. process,” Frontiers in Genetics, vol. 11, article 626536, 2020. [4] S. M. Laakso, O. Hekali, G. Kurdo, J. Martola, T. Sairanen, and [21] G. Yu, L. G. Wang, Y. Han, and Q. Y. He, “clusterProfiler: an R S. Atula, “Trigeminal neuralgia in multiple sclerosis: preva- package for comparing biological themes among gene clus- lence and association with demyelination,” Acta Neurologica ters,” Omics: a journal of integrative biology, vol. 16, no. 5, Scandinavica, vol. 142, no. 2, pp. 139–144, 2020. pp. 284–287, 2012. [5] R. Lin, L. Luo, Y. Gong et al., “Immunohistochemical analysis [22] S. Zhou, Y. Shen, S. Zang, X. Yin, and P. Li, “The epigenetic of histone H3 acetylation in the trigeminal root entry zone in role of HTR1A antagonist in facilitating GnRH expression an animal model of trigeminal neuralgia,” Journal of Neurosur- for pubertal initiation control,” Molecular therapy. Nucleic gery, vol. 131, no. 3, pp. 828–838, 2018. acids, vol. 25, pp. 198–206, 2021. [6] R. J. Danaher, L. Zhang, C. J. Donley et al., “Histone deacetyl- [23] Y. Shen, H. Zhao, L. Zhang et al., “The roles of DNA methyl- ase inhibitors prevent persistent hypersensitivity in an orofa- ation and hydroxymethylation at short interspersed nuclear cial neuropathic pain model,” Molecular Pain, vol. 14, article elements in the hypothalamic arcuate nucleus during puberty,” 1744806918796763, 2018. Molecular therapy. Nucleic acids, vol. 26, pp. 242–252, 2021. [7] F. Denk, W. Huang, B. Sidders et al., “HDAC inhibitors atten- [24] B. Langmead and S. L. Salzberg, “Fast gapped-read alignment uate the development of hypersensitivity in models of neuro- with bowtie 2,” Nature Methods, vol. 9, no. 4, pp. 357–359, pathic pain,” Pain, vol. 154, no. 9, pp. 1668–1679, 2013. [8] C. Zhou, Y. Wu, X. Ding, N. Shi, Y. Cai, and Z. Z. Pan, “SIRT1 decreases emotional pain vulnerability with associated CaM- [25] Y. Zhang, T. Liu, C. A. Meyer et al., “Model-based analysis of ChIP-Seq (MACS),” Genome Biology, vol. 9, no. 9, p. R137, KIIα deacetylation in central amygdala,” The Journal of Neuro- science: The Official Journal of the Society for Neuroscience, 2008. vol. 40, no. 11, pp. 2332–2342, 2020. [26] P. Kolodziej-Wojnar, J. Borkowska, Z. Wicik et al., “Alter- [9] Z. Li, Y. Guo, X. Ren et al., “HDAC2, but not HDAC1, regu- ations in the genomic distribution of 5hmC in in vivo aged lates Kv1.2 expression to mediate neuropathic pain in CCI human skin fibroblasts,” International Journal of Molecular rats,” Neuroscience, vol. 408, pp. 339–348, 2019. Sciences, vol. 22, no. 1, p. 78, 2021. Oxidative Medicine and Cellular Longevity 13 [27] Y. Wang, Q. Sun, J. Liang, H. Li, D. M. Czajkowsky, and Z. Shao, “Q-Nuc: a bioinformatics pipeline for the quantitative analysis of nucleosomal profiles,” Interdisciplinary sciences, computational life sciences, vol. 12, no. 1, pp. 69–81, 2020. [28] Q. Lai, W. Du, J. Wu et al., “H3K9ac and HDAC2 activity are involved in the expression of monocarboxylate transporter 1 in oligodendrocyte,” Frontiers in Molecular Neuroscience, vol. 10, p. 376, 2017. [29] H. Ji, Y. Zhou, X. Zhuang et al., “HDAC3 deficiency promotes liver cancer through a defect in H3K9ac/H3K9me3 transition,” Cancer Research, vol. 79, no. 14, pp. 3676–3688, 2019. [30] J. Vecera, E. Bartova, J. Krejci et al., “HDAC1 and HDAC3 underlie dynamic H3K9 acetylation during embryonic neuro- genesis and in schizophrenia-like animals,” Journal of Cellular Physiology, vol. 233, no. 1, pp. 530–548, 2018. [31] H. Patil, C. Wilks, R. W. Gonzalez, S. Dhanireddy, H. Conrad- Webb, and M. Bergel, “Mitotic activation of a novel histone deacetylase 3-linker histone H1.3 protein complex by protein kinase CK2,” The Journal of Biological Chemistry, vol. 291, no. 7, pp. 3158–3172, 2016. [32] K. A. Han, W. H. Shin, S. Jung et al., “Leucine-rich repeat kinase 2 exacerbates neuronal cytotoxicity through phosphor- ylation of histone deacetylase 3 and histone deacetylation,” Human Molecular Genetics, vol. 26, no. 1, pp. 1–18, 2017. [33] J. Seo, G. Guk, S. H. Park et al., “Tyrosine phosphorylation of HDAC3 by Src kinase mediates proliferation of HER2- positive breast cancer cells,” Journal of Cellular Physiology, vol. 234, no. 5, pp. 6428–6436, 2019. [34] Y. Kim, S. Y. Kwon, H. S. Jung et al., “Amitriptyline inhibits the MAPK/ERK and CREB pathways and proinflammatory cytokines through A3AR activation in rat neuropathic pain models,” Korean Journal of Anesthesiology, vol. 72, no. 1, pp. 60–67, 2019. [35] C. L. Li, R. Yang, Y. Sun, Y. Feng, and Y. B. Song, “N58A exerts analgesic effect on trigeminal neuralgia by regulating the MAPK pathway and tetrodotoxin-resistant sodium channel,” Toxins, vol. 13, no. 5, p. 357, 2021. [36] X. Wang, X. Shen, Y. Xu et al., “The etiological changes of acetylation in peripheral nerve injury-induced neuro- pathic hypersensitivity,” Molecular Pain, vol. 14, article 1744806918798408, 2018. [37] M. D. Sanna and N. Galeotti, “The HDAC1/c-JUN complex is essential in the promotion of nerve injury-induced neuro- pathic pain through JNK signaling,” European Journal of Phar- macology, vol. 825, pp. 99–106, 2018. [38] J. Qian, W. Zhu, M. Lu, B. Ni, and J. Yang, “D-β-hydroxybuty- rate promotes functional recovery and relieves pain hypersen- sitivity in mice with spinal cord injury,” British Journal of Pharmacology, vol. 174, no. 13, pp. 1961–1971, 2017. [39] T. Aczel, J. Kun, E. Szoke et al., “Transcriptional alterations in the trigeminal ganglia, nucleus and peripheral blood mononu- clear cells in a rat orofacial pain model,” Frontiers in Molecular Neuroscience, vol. 11, p. 219, 2018. [40] Y. Yao, B. Chang, and S. Li, “Relationship of inflammation with trigeminal neuralgia,” The Journal of Craniofacial Sur- gery, vol. 31, no. 2, pp. e110–e113, 2020. [41] S. Bahl and E. Seto, “Regulation of histone deacetylase activi- ties and functions by phosphorylation and its physiological relevance,” Cellular and Molecular Life Sciences: CMLS, vol. 78, no. 2, pp. 427–445, 2021. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Oxidative Medicine and Cellular Longevity Hindawi Publishing Corporation

Characterization of Acetylation of Histone H3 at Lysine 9 in the Trigeminal Ganglion of a Rat Trigeminal Neuralgia Model

Loading next page...
 
/lp/hindawi-publishing-corporation/characterization-of-acetylation-of-histone-h3-at-lysine-9-in-the-TFfDNpJj2E

References (44)

Publisher
Hindawi Publishing Corporation
Copyright
Copyright © 2022 Wenbin Wei et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
ISSN
1942-0900
eISSN
1942-0994
DOI
10.1155/2022/1300387
Publisher site
See Article on Publisher Site

Abstract

Hindawi Oxidative Medicine and Cellular Longevity Volume 2022, Article ID 1300387, 13 pages https://doi.org/10.1155/2022/1300387 Research Article Characterization of Acetylation of Histone H3 at Lysine 9 in the Trigeminal Ganglion of a Rat Trigeminal Neuralgia Model 1,2,3,4,5 1,2,3,4,5 1,2,3,4,5 1,2,3,4,5 Wenbin Wei , Yuemin Liu , Yating Qiu , Minjie Chen , 1,2,3,4,5 1,2,3,4,5 1,2,3,4,5 Yiwen Wang , Zixiang Han , and Ying Chai Department of Oral Surgery, Shanghai Ninth People’s Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China College of Stomatology, Shanghai Jiao Tong University, Shanghai, China National Center for Stomatology, Shanghai, China National Clinical Research Center for Oral Diseases, Shanghai, China Shanghai Key Laboratory of Stomatology, Shanghai, China Correspondence should be addressed to Yating Qiu; qiuyating@hotmail.com and Minjie Chen; chenmj_9hospital@126.com Received 2 March 2022; Accepted 12 April 2022; Published 4 May 2022 Academic Editor: Ajinkya Sase Copyright © 2022 Wenbin Wei et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Trigeminal neuralgia (TN) is a chronic neuropathic pain disorder characterized by spontaneous and elicited paroxysms of electric- shock-like or stabbing pain in a region of the face. The epigenetic regulation of TN is still obscure. In current study, a rat TN model subject to carbamazepine (CBZ) treatment was established, and transcriptome- and genome-scale profiling of H3K9ac and HDAC3 was performed by RNA-seq and ChIP-seq. We observed that H3K9ac levels in the trigeminal ganglion were lower in the TN rats compared with those in the control, and CBZ treatment led to recovery of H3K9ac levels. Further, we found that HDAC3 was overactivated, which interfered with H3K9 acetylation due to higher phosphorylation in TN compared with that in the control. Finally, the phosphokinase leucine-rich repeat kinase 2 (LRRK2) was demonstrated to contribute to HDAC3 activity via the MAPK signaling pathway. Taken together, we identified a regulatory mechanism in which the phosphate groups transferred from activated ERK and LRRK2 to HDAC3 caused genome-scale deacetylation at H3K9 and resulted in the silencing of a large number of genes in TN. The kinases or important enzymes within this regulatory axis may represent important targets for TN therapy and prevention. TN [2]. In addition, the pathogenesis of TN involves neuro- 1. Introduction degeneration, epilepsy, and infection. A large number of Trigeminal neuralgia (TN) manifests as transient, paroxys- neurophysiological, neuroimaging, and histological studies mal but severe pain that occurs in the trigeminal nerve dis- show that the underlying pathophysiological mechanism in tribution area. TN is a common cause of facial pain in TN involves the focal demyelination of primary trigeminal individuals aged over 50 years old. The incidence of TN in afferents near the entry of the trigeminal root into the pons females is higher than in males. TN is frequently both mis- [3, 4]. Such demyelination results in hyperexcitability of the diagnosed and underdiagnosed, and its incidence is variably focally demyelinated primary afferents due to the insuffi- reported between studies, with a range from 4.3 to 27 new ciency of prompt resting potential re-establishment in axons. cases per 100,000 people per year [1]. The compression of An epigenetic trait is a stably heritable phenotype result- ing from changes in a chromosome without alterations in blood vessels at the trigeminal nerve has been observed in more than 30% of TN patients. The academic community the DNA sequence. Histone H3 is one of the most exten- has reached broad agreement that a neurovascular conflict sively modified of the five primary histone proteins, and with compression is the most common cause of classical acetylation at different lysine residues of histone H3 may 2 Oxidative Medicine and Cellular Longevity 20 minutes after the rats adapted to the test cage. The face play key roles in gene regulation. A previous study has reported larger numbers of acetylated H3K9-, H3K18-, and grooming behavior was analyzed offline by an independent H3K27-positive cells in the TN group than in the sham- observer who was blind to the treatment administered to operation group [2, 5]. Interestingly, the notion that H3K9 the rats. For each observation session of the rats over 10 global acetylation decreases in the Schwann cells in the tri- minutes, the total duration of asymmetrical face grooming geminal ganglia of a nerve injury mouse model induced by episodes as unilateral strokes with the dominant paw was trigeminal inflammatory compression [6], whereas that his- counted [14]. For open field test, the animals were placed tone deacetylase inhibitors (HDACi) can prevent the persis- into an open-field hardboard box for 6 min individually to tent hypersensitivity of neurons from neuropathic pain [7], assess the motor activity, exploratory behavior, and emo- which indicate that cells in different areas of trigeminal root tional state. During the original period, 1 min was allowed entry zone and ganglion contain display a different response for accommodation, and the next 5 min was for recording, to TN. Moreover, various HDAC subclusters have also been including both horizontal and vertical movement and dwell found to exert varying effects on pain vulnerability and time. In addition, the mice were allowed optional movement excitability of neurons [8, 9]. Therefore, the epigenetic regu- in the box during the experiment. lation of the gene expression in response to compressive Rats were anesthetized with 60 mg/kg pentobarbital and stress-induced imbalanced homeostasis in the trigeminal shaved the hair and made a mid-line scalp incision with ganglion and trigeminal root entry zone is still obscure. equal distances anterior and posterior to the center of the In the current study, we constructed a rat model of TN eyes to expose the skull and nasal bone. After dissecting the edge of the orbit free formed by the maxillary, frontal, subject to carbamazepine (CBZ) treatment a widely used antiepileptic drug in clinical practice and performed tran- lacrimal, and zygomatic bones, infraorbital nerve was dis- scriptome- and genome-scale H3K9ac profiling. CBZ as the sected freely from the surrounding connective tissue for first-line pharmacological treatment for trigeminal neuralgia the consequent experiments. The animal study was reviewed [10] has been determined to function as HDACi [11]. Our and approved by the Medical Ethics Committee of Shanghai Jiao Tong University School of Medicine. study attempted to determine the alterative H3K9ac signa- ture and elucidate the role of the HDAC3 in TN [6, 7]. The large number of differentially expressed genes (DEGs) as 2.2. Immunofluorescence (IF) Assay. Frozen trigeminal well as the differential H3K9ac and HDAC3 signals across nerves sections (10 μm) were cut by a cryostat (Thermo the whole genome identified in this study reveals novel con- Fisher Scientific). Unlike the nerve root region with the sur- rounding distribution of nuclei, the trigeminal ganglion con- nections between H3K9ac and gene expression and the underlying mechanism of HDAC3 in H3K9ac modulation, tains robust nuclei in central region. Sections were fixed with emphasizing the importance of epigenetic alterations in reg- 4% solution of paraformaldehyde and washed with ulating transcription during TN. phosphate-buffered saline (PBS) and then permeabilized with 0.1% Triton-X 100 (Millipore) and blocked with 5% horse serum in PBS. Immunostaining of samples was per- 2. Materials and Methods formed using antibodies against H3K9ac (1: 500, CST, #9649) or GFAP (1: 200, CST, #3670) overnight at 4 2.1. Animal Study. Adult female Sprague-Dawley rats (200– C. After 220 g) were randomly divided into two groups: one group washing with PBS for 4 times (5 min per time), the second- received chronic compression of the trigeminal nerve root ary goat antibodies against rabbit (for H3K9ac) or mouse only at left side (n =30), and the other sham-operated group (for GFAP) (1 : 20000, Jackson ImmunoResearch) were used, was considered as the negative control (n =20). The TN and incubation was performed for 2 h at room temperature; model was prepared as previously described [12]. Five days this was followed with further incubation with DAPI for after TN model establishment, the TN group was further 15 min and washing with PBS four times. After drying in divided into two groups: one was treated with 100 mg/kg/d air, mounting medium (Thermo Fisher Scientific) was added CBZ (Sigma) dissolved into corn oil as previous described dropwise to the tissue slices, and they were coated with cover [13] for an additional five days (n =15). All rats were sacri- slips. The images were captured by fluorescence microscopy ficed in the tenth day. DM2000 (Leica Biosystems). Positive staining was statisti- Orofacial mechanical allodynia testing was conducted at cally analyzed using ImageJ. the same side of injured tergeminal nerve as previously described [5]. Rats were lightly restrained by their tail and 2.3. RNA Sequencing (RNA-seq). Trigeminal ganglion tissues allowed to stand on a mesh platform. The von Frey filaments were stored in 1 mL TRIzol (Thermo Fisher Scientific) and (Ugo Basile) were applied to their left vibrissal pad border ground in liquid nitrogen. Next, 100 μL chloroform was and cheek to collect baseline mechanical thresholds. Stimu- added, and the cells were fully mixed and centrifuged at lation always began with the filament that produced the low- the highest speed at 4 C for 10 min. The supernatant was est force (bending forces of 0.008–6 g) and stopped when the transferred into a new tube, isopropanol was added to the threshold was found within the vibrissal pad of the rats. Each same volume, and cells were centrifuged at the highest speed rat was tested five times at intervals of 3–5 s. The face at 4 C for 10 min. The precipitate was washed with 75% cold grooming episodes (ipsilateral, contralateral, or bilateral face ethanol and dissolved in DEPC water. The concentration grooming) of the rats in the three groups were recorded by a and quality of RNA were measured using a Nanodrop video camera. Video recording of face grooming was started 2000 (Thermo Fisher Scientific) and an Agilent Bioanalyzer Oxidative Medicine and Cellular Longevity 3 were harvested and mixed with 1 μg phosphoserine/threo- 2100 (Agilent). A total of 4 μg of RNA from each group was used for library preparation using the NEBNext Ultra Direc- nine (Abcam, ab17464), HDAC3 (CST, #85057), LRRK2 tional RNA Library Prep Kit for Illumina (NEB) following (CST, #13046), or IgG Rabbit IgG antibody and 40μl flurry the manufacturer’s instructions and sequenced on an Illu- IgA beads (Thermo Fisher Scientific) for rotating overnight mina HiSeq platform. at 4 C. Immunoprecipitates were washed with IP buffer The raw data were trimmed for adaptors, and low- (20 mM HEPES (pH 7.9), 350 mM NaCl, 0.1% NP-40, 1 mM quality reads were filtered out using Trimmomatic [15]. DTT, 0.2 mM PMSF, 2 mg/mL leupeptin, and 2 mg/mL apro- The quality of clean reads was checked using FastQC [16]. tinin) and purified with RIPA buffer (50 mM Tris (pH 7.4), Next, clean reads were aligned to the latest rat genome assem- 150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate, 0.1% bly Rnor6.0 using Hisat2 [17]. The transcripts were assembled, SDS) with 1% proteasome inhibitor cocktail and 1% PMSF. and the expression levels were estimated with FPKM values The protein lysate was subjected to western blot assay. using the StringTie algorithm with default parameters [18]. Differential mRNA and lncRNA expression among the groups 2.6. Western Blot (WB) Assay. The protein lysate was then were evaluated using the R package Ballgown [19], and the sig- analyzed by SDS-PAGE and transferred to PVDF mem- nificance of differences was computed using the Benjamini branes (Bio-Rad Laboratories, Hercules, USA). The mem- and Hochberg (BH) p value adjustment method [20]. Gene brane was blocked with 5% fat-free milk in PBST for annotation was performed using the Ensembl Genome 30 min, followed by incubation overnight at 4 C with final Browser database (http://www.ensembl.org/index.html). The dilutions of primary antibodies against p-HDAC1 (Ser421, R package ClusterProfiler was used to annotate the DEGs Ser423) (1: 1000, Thermo Fisher Scientific, PA5-36911), using Gene Ontology (GO) terms and Kyoto Encyclopedia HDAC1 (1: 2000, Abclonal, A19571), p-HDAC2 (Ser394) of Genes and Genomes (KEGG) pathways [21]. (1: 1000, Thermo Fisher Scientific, PA5-105021), HDAC2 (1: 2000, Abclonal, A19626), p-HDAC3 (Ser424) (1: 1000, 2.4. Chromatin Immunoprecipitation Sequencing (ChIP-Seq). Thermo Fisher Scientific, PA5-99339), HDAC3 (1: 2000, In brief, trigeminal ganglion tissues were incubated with Abclonal, A19537), p-HDAC8 (Ser29) (1: 1000, Thermo 1mL IP buffer (20 mM HEPES (pH 7.9), 350 mM NaCl, Fisher Scientific, PA5-105031), HDAC8 (1: 2000, Abclonal, 0.1% NP-40, 1 mM DTT, 0.2 mM PMSF, 2 mg/mL leupeptin, A5829), p-CK2β (Ser209) (1: 1000, Thermo Fisher Scientific, and 2 mg/mL aprotinin) [22, 23] on ice for 30 min; then, the 44-1090G), CK2 (1: 2000, Abclonal, A2869), p-LRRK2 nuclei were harvested by centrifugation at the highest speed. (Ser935) (1: 1000, Thermo Fisher Scientific, PA5-114599), The supernatant containing nuclei were sonicated to break LRRK2 (1: 2000, Abclonal, A17253), p-SRC (Tyr529) (1: up the genomic DNA into 200–500 bp fragments. Next, 1000, Thermo Fisher Scientific, PA5-97356), SRC (1: 2000, 10% lysates were saved as input, and the remaining were Abclonal, A19119), p-ERK1/2 (Thr202, Tyr204)(1: 1000, incubated with 1 μg IP-grade antibodies of H3K9ac (CST, Abclonal, AP1120), ERK1/2 (1: 2000, Abclonal, A19630), #9649) and HDAC3 (Abclonal, A19537) at 4 C overnight. p-MEK (Ser271, Tyr275) (1: 1000, Abclonal, AP0535), This was followed by incubation for 2 h with protein A beads MEK (1: 2000, Abclonal, A12950), or GAPDH (1: 2000, Pro- at 37 C to pull down the bound DNA fragments. teintech Group, #60004-1). Next, the membrane was washed ′ three times and then incubated with HRP-conjugated sec- For high-throughput sequencing, we added 3 -dA over- ondary antibodies (Proteintech Group). Membranes could hangs to the H3K9ac or HDAC3 enriched or input DNA be stripped using stripping buffer (Abcam, ab270550) at and ligated them to the adapter to build a DNA library. DNA libraries with ligated adapters were isolated based on 52 C for 30 min via gently shaking, washed by PBST, then blocked by 5% fat-free milk for 2 h, and reincubated by anti- the appropriate size for sequencing, using the Illumina bodies at 4 Hiseq2000 platform. The raw sequence reads of input and C overnight. The blotting bands were developed with ECL plus immunoblotting detection reagents (Thermo IP were trimmed based on adaptors, and low-quality reads Fisher Scientific) using UVP Chemstudio Plus System (Ana- were filtered out using Cutadapt (v1.9.1) and Trimmomatic (v0.35). The quality of the clean reads was checked using lytik Jena) and captured using ImageJ. Fastqc. The clean reads were mapped to the rat genome (assembly Rnor6.0) using the Bowtie 2 (v2.2.6) algorithm 2.7. HDAC Activity Assay. HDAC activity/inhibition assay kits for HDAC1 (Abx155637), HDAC2 (Abx258089), HDAC3 [24], and peak calling (p <0:01) was performed using MACS (Abx257885), and HDAC8 (Abx515264) were purchased 2 (v2.1.1) [25]. The differentially bound genes were analyzed based on p values less than 0.05 and annotated using Diff- from AmyJet Scientific Co. Ltd. Trigeminal ganglion tissues were incubated with RIPA buffer (50 mM Tris (pH 7.4), Bind [26]. The relevant peaks on the genomic loci were visu- 150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate, 0.1% alized using the Integrative Genomics Viewer (IGV). GO SDS) with 1% proteasome inhibitor cocktail and 1% PMSF. analysis was used to determine the biological functions of The protein concentration was quantified using the BSA genes associated with the differential peaks [27]. The raw ChIP sequencing data were submitted to the ArrayExpress method. The volume of different samples was adjusted to ensure that the total protein was identical. HDAC activities database and registered under the accession number E- were processed following the manufacturer’sinstructions. MTAB-10792 and 10793. 2.5. Immunoprecipitation (IP). Previous steps were similar to 2.8. Statistical Analysis. Data are presented as the mean ± those followed for the ChIP-seq assay. Trigeminal ganglia standard deviation for three independent experiments. The 4 Oxidative Medicine and Cellular Longevity 5 40 p = 0.042 p = 0.018 p = 0.025 p = 0.009 0 0 Control TN CBZ Control TN CBZ (n = 20) (n = 15) (n = 15) (n = 20) (n = 15) (n = 15) (a) (b) 40 30 p = 0.023 p = 0.012 25 p = 0.038 p = 0.027 20 15 0 0 Control TN CBZ Control TN CBZ (n = 20) (n = 15) (n = 15) (n = 20) (n = 15) (n = 15) (c) (d) 50 𝜇m 50 𝜇m 50 𝜇m GFAP Control TN CBZ (e) Figure 1: Confirmation of the TN rat model. (a)–(d) Behavior examination of orofacial mechanical allodynia by stimulation threshold and head withdrawal frequency and face grooming as well as open field test by central dwell time in the control (n =20), TN rat model (n =15), and TN rat model administered CBZ treatment (n =15). The p values were given by the comparison between control and TN, as well as TN and CBZ. (e) GFAP staining in trigeminal root entry zone by IF (×200 magnification); white arrows indicate distal extension of GFAP- positive astrocytic process in the trigeminal root. grooming behavior was also significantly increased after the differences in values were analyzed using one-way ANOVA. Statistical significance was set at p value less than 0.05. All operation and alleviated at day 10 with CBZ treatment analysis is performed by SPSS 20.0. (F =6:244, p =0:012; F =3:055, p =0:038) (Figure 1(c)). Open field test showed that CBZ could compromise the hyperactivity from TN effect (F =3:628, p =0:023; F = 3. Results 3:314, p =0:027) (Figure 1(d)). Nonetheless, we noticed that 3.1. Decreased H3K9ac Profiling in Trigeminal Ganglia in TN CBZ did not cause behavioral retardation in rats compared to Rats. Thirty adult female SD rats were subjected to mechan- control, indicating that the probable side effects of CBZ on ical compression force of the trigeminal root entry zone by rats’ activity with the current dose did not exist in our sys- retrograde insertion of a plastic filament from the right infe- tem. Moreover, a distal extension of glial fibrillary acidic rior orbital fissure for the TN model. After five days, half of protein (GFAP) staining at the trigeminal root entry zone them were treated with 100 mg/kg/d CBZ for an additional in the TN model at day 5 relative to the control indicated five days. Behavioral examination of orofacial mechanical that astrocytes started to break through the boundary allodynia showed that, compared with the 20 rats in the con- established by Schwann cells (Figure 1(e)). These results trol group, stimulus intensity (F =6:681, p =0:009; F = suggest the successful establishment of a rat TN model for subsequent experiments. 4:412, p =0:018) and mechanical hyperalgesia (F =3:447, p =0:025; F =2:893, p =0:042) were increased after the Next, H3K9ac was investigated by IHC and ChIP-seq. operation and attenuated at day 10 following CBZ treatment We observed fewer H3K9ac-positive nuclei within the tri- (Figures 1(a) and 1(b)). Likewise, the frequency of face- geminal ganglia of the TN model than in the control group Mechanical stimulation Face grooming time (s) threshold (g) Central dwell time (s) Latencies of head withdrawal (s) Oxidative Medicine and Cellular Longevity 5 Control TN CBZ 50 𝜇m 50 𝜇 m 50 𝜇 m DAPI p = 0.041 50 𝜇m 50 𝜇 m 50 𝜇 m H3K9ac p = 0.037 50 𝜇m 50 𝜇 m 50 𝜇 m Merge (a) 0.6 35,000 TN vs. CBZ, p = 0.012 30,000 TN vs. Control, p = 0.005 25,000 0.4 20,000 p = 0.003 15,000 p < 0.001 0.2 10,000 5,000 Control TN CBZ 0.0 TSS 2.0 Kb –2.0 Kb Control TN CBZ (b) (c) 100 Control TN CBZ 1,026 521 Th Control TN CBZ Promoter Intron 5′UTR 3′UTR Exon Distal intergenic regions (d) (e) Figure 2: Overview of H3K9ac enrichment in the TN rat model. (a) Global H3K9ac staining in the trigeminal ganglion of the control, TN rat model, and TN rat model administered CBZ treatment (×200 magnification). The p-values were given by the comparison of H3K9ac positive puncta occupied in all cells between control and TN, as well as TN and CBZ. (b) Overall calling peaks of H3K9ac in the rat genome of control rats, TN rat model, and TN rat model administered CBZ treatment, by ChIP-seq. The p values were given by the comparison of genomic H3K9ac peak counts between control and TN, as well as TN and CBZ. (c) Metagene profiles of genome-wide H3K9ac in the control, TN rat model, and TN rat model administered CBZ treatment. (d) The distribution of differentially H3K9ac- enriched regions in the genomic contexts of the control, TN rat model, and TN rat model administered CBZ treatment. (e) Diagram showing the intersection of genes with differential H3K9ac enrichment in the control, TN rat model, and TN rat model administered CBZ treatment. Genome-wide H3K9ac peaks e genomic distribution of H3K9ac peaks (%) Read count per million mapped reads H3K9ac positive puncta/cell (%) Control TN CBZ 6 Oxidative Medicine and Cellular Longevity Control TN CBZ 2,104 1,672 1,522 (a) 0.6 0.4 r = 0.487 p = 0.002 0.2 0.0 –0.2 Promoter 5′UTR Exon Intron 3′UTR Distal intergenic region (b) Biological process Activation of MAPKK activity Positive regulation of peptidyl-tyrosine phosphorylation Neuroinflammatory response Regulation of cell aging Cellular response to nutrient Axon extension Negative regulation of dendrite development Sensory perception of pain GTP metabolic process Positive regulation of histone modification Negative regulation of protein oligomerization Positive regulation of dephosphorylation Central nervous system neuron axonogenesis Negative regulation of phosphorylation Regulation of cellular response to growth factor stimulus 0 5 10 15 –log P (c) Figure 3: Continued. Pearson correlation coefficient (r) between H3K9ac and transcription Oxidative Medicine and Cellular Longevity 7 GO:0019233 GO:2000171 GO:1900015 GO:0050890 GO:0051930 GO:0031670 GO:0048675 GO:0015914 GO:1904029 GO:0031058 GO:0090342 GO:0032874 GO:0008654 GO:0006925 GO:0046039 GO:0045165 GO:0002534 GO:0021955 GO:0070932 GO:0000186 GO:0150076 GO:0006925: inflammatory cell apoptotic process GO:0150076: neuroinflammatory response GO:0046039: GTP metabolic process GO:0043409 GO:0043409: negative regulation of MAPK cascade GO:0070932: histone H3 deacetylation (d) Figure 3: Overview of the transcriptome in the TN rat model. (a) Diagram showing intersecting DEGs in the control, TN rat model, and TN rat model administered CBZ treatment. (b) Correlation between RNA levels and H3K9ac enrichment at different genomic regions of DEGs, by Pearson correlation analysis. (c) Gene ontology analysis of DEGs with differential H3K9ac enrichment, for biological processes. (d) Regulatory connections between different terms of biological processes by directed acyclic graph. The key intermediate and terminal GO items are labeled by circle and box. p value less than 0.001 and 0.0001 are highlighted by orange and red. (p =0:037) and in rats subjected to CBZ treatment ment were all included among the DEGs. Pearson correla- (p =0:041) (Figure 2(a)). Consistent with this result, tion analysis showed that only the differential H3K9ac genome-wide H3K9ac profiling also indicated weakened enrichment at promoter regions correlated positively with the expression of the corresponding genes (r =0:487, p = global H3K9ac peaks on neuron genomes in the TN group relative to the control (p <0:001) and CBZ treatment groups 0:002) (Figure 3(b)). Notably, GO analysis showed that these (p =0:003) (Figure 2(b)). Furthermore, H3K9ac density sug- genes were tightly linked with functional regulation on gested that a unique subset of target genes exhibited com- axons or dendrites as well as inflammatory cytokine stimu- promised H3K9ac enrichment mainly at promoter regions lus (Figure 3(c)). In particular, protein phosphorylated mod- (defined as −2 kb to +2 kb from the transcriptional start site ification, phosphate metabolism process, and GTPase- [TSS]) in TN relative to the control (p =0:005) and CBZ associated activity occupied the central position in the treatment groups (p =0:012) (Figures 2(c) and 2(d)). directed acyclic graph of functional regulation Numerous H3K9ac-targeted sites identified by examining (Figure 3(d)). Collectively, the data indicated that the abnor- the differences between control or CBZ treatment and TN mal H3K9ac contributed to the changes in the expression of were mapped to 477 genes, with 421 genes showing a decrease a large proportion of genes in TN. and 56 an increase in H3K9ac in TN (fold change ðFCÞ >2 or <0.5, p <0:05) (Figure 2(e)). Taken together, our findings 3.3. Excessive Activation of HDAC3 in Trigeminal Ganglia in suggested that TN causes reduced H3K9ac in the genomes TN. Given that previous studies report seemingly controver- of trigeminal ganglion neurons. sial findings on histone H3 acetylation in neuralgia that the positive staining of H3K9, H3K18, and H3K27 acetylation 3.2. Transcriptome of Trigeminal Ganglion Affected by were higher in postoperative days 7, 14, 21, and 28 in male H3K9ac in TN. To further study the expression of target TN rats [5], whereas the global acetylation of H3K9 genes affected by H3K9ac, we also conducted RNA-seq to decreased at day 21 by trigeminal inflammatory compres- investigate DEGs in TN. Based on the results, 164 genes sion in mouse trigeminal ganglia [6], it is important to study showed higher expression while 1,358 genes had lower the role of HDACs in regulating H3K9ac in TN. We focus expression in TN relative to the control and CBZ treatment on class I HDACs in the current study because of the evi- groups were characterized (FC > 2 or<0.5, p <0:05) dence that class I HDACs (HDAC1, 2, 3, and 8) are widely (Figure 3(a)). The genes with differential H3K9ac enrich- implicated in H3K9ac transition in multiple human diseases 8 Oxidative Medicine and Cellular Longevity phospho-Ser/Thr Input IgG Control TN CBZ Control TN CBZ Control TN CBZ GAPDH p-HDAC1 (Ser421/423) HDAC3 HDAC1 (b) p-HDAC2 (Ser394) HDAC activity assay HDAC2 p = 0.033 ^ p = 0.014 p-HDAC3 (Ser424) HDAC3 p-HDAC8 (Ser29) HDAC1 HDAC2 HDAC3 HDAC8 HDAC8 Control TN GAPDH CBZ (a) (c) Figure 4: HDAC3 activity by phosphorylation modification in TN rat model. (a) Phosphorylation of HDAC1, 2, 3, and 8 in the control, TN rat model, and TN rat model administered CBZ treatment, by WB assay. (b) Phosphorylated HDAC3 profiles in the control, TN rat model, and TN rat model administered CBZ treatment, by phosphoserine and threonine IP assay. (c) The activity of HDAC1, 2, 3, and 8 in the control, TN rat model, and TN rat model administered CBZ treatment, by enzyme-linked immunosorbent assay (ELISA); “ ” and “^” represent comparison between TN and control as well as CBZ and TN with statistical significance. and development [28–30]. The expression of class I HDACs the control and CBZ treatment groups (FC > 2 or<0.5, p < was studied in TN by WB assay. Here, the expression of 0:05) (Figure 5(c)). Hence, we identified three lists of genes these four HDACs showed no significant difference between with differentially occupied H3K9ac (477) and HDAC3 the control, TN, and CBZ treatment groups, and we unex- (637) as well as DEGs (1, 522) by comparison between the pectedly found that the phosphorylation of HDAC3 TN and control group, and obtained an intersection of 288 genes. The transcription of these 288 genes was positively (Ser424) was higher in the TN and CBZ treatment groups than in the control (Figure 4(a)). Furthermore, the phos- correlated with H3K9ac alteration (r =0:506, p =0:0017) phorylated serine/threonine pull-down experiment showed and negatively correlated with HDAC3 (r = −0:475, p = that HDAC3 was remarkably modified via phosphorylation 0:0021) compared between control and TN groups, as well in the TN and CBZ treatment groups relative to that in the as positively correlated with H3K9ac alteration (r =0:448, control (Figure 4(b)). However, CBZ treatment strongly p =0:0026) and negatively correlated with HDAC3 (r = attenuated the activity of HDAC3 relative to that of HDAC1, − 0:626, p =0:0013) compared between TN and CBZ 2, and 8 in the in vitro assay (Figure 4(c)). Unexpectedly, groups (Figure 5(d)) (Table S1). Neurod2, C2cd4c, and CBZ treatment produced a similar phosphorylation signa- Sf3b4, with the top correlation coefficient values, confirmed ture of HDAC3 as in the TN group, despite substantially the enrichment of H3K9ac and HDAC3 as well as their attenuating TN. We assumed that CBZ, being an HDACi, transcription using the IGV browser (Figure 5(e)). In did not exert its effects by blocking phosphorylation. conclusion, we found that the activated HDAC3 contributed Next, we performed ChIP-seq analysis of trigeminal gan- to the reduced H3K9ac modification and gene silencing in TN. glia to investigate the genome-wide occupancy of HDAC3 in TN. Opposite to the findings for enrichment of H3K9ac, 3.4. Phosphorylation of HDAC3 Regulated by LRRK2 via global HDAC3 occupancy was remarkably robust in TN Compressive Stress-Induced MAPK/ERK Signaling Pathway. compared with that in the control and CBZ treatment As previously described, three enzymes, namely, casein groups (Figure 5(a)). Our data indicate that the genomic kinase 2 (CK2) [31], leucine-rich repeat kinase 2 (LRRK2) binding ability of HDAC3 was indeed strengthened, espe- [32], and SRC protooncogene, nonreceptor tyrosine- cially at the promoter regions in TN relative to the con- protein kinase (SRC) [33], contributing to the phosphoryla- trol (p =0:019) and CBZ treatment groups (p =0:037) tion of HDAC3 or HDAC3-H1.3 complex were investigated (Figure 5(b)). Moreover, HDAC3 peaks were mapped to to elucidate the mechanism underlying the phosphorylation 637 genes, in which 564 had increased, whereas 73 had of HDAC3 in TN. Initially, we determined a substantial decreased HDAC3 occupancy in TN compared to that in interaction between HDAC3 and LRRK2 rather than CK2 Control TN CBZ H3K9ac enrichment (log ) Oxidative Medicine and Cellular Longevity 9 30,000 p < 0.001 20,000 p < 0.001 10,000 Control TN CBZ (a) Control TN CBZ 0.6 TN vs. CBZ, p = 0.037 TN vs. Control, p = 0.019 0.4 1,210 755 0.2 0.0 –2.0 Kb TSS 2.0 Kb Control TN CBZ (b) (c) –4 –8 –8 –4 4 2 –2 –4 8 –6 –8 Control TN CBZ (d) Figure 5: Continued. HDAC3 enrichment (log ) Read count per million mapped reads Genome-wide HDAC3 peaks Transcription (log ) 2 10 Oxidative Medicine and Cellular Longevity 200 200 200 Control 0 0 200 200 RNA TN 0 0 200 200 CBZ 0 0 200 200 Control 0 0 200 200 TN H3K9ac 0 0 200 200 CBZ 0 0 0 200 200 Control 0 200 200 200 HDAC3 TN 0 0 CBZ 0 0 Neurod2 C2cd4c Sf3b4 (e) Figure 5: Overview of HDAC3 in TN rat model. (a) Overall calling peaks of HDAC3 in the rat genome of the control, TN rat model, and TN rat model administered CBZ treatment, by ChIP-seq. (b) Metagene profiles of genome-wide HDAC3 in the control, TN rat model, and TN rat model administered CBZ treatment. (c) Diagram showing intersecting genes with differential HDAC3 enrichment in the control, TN rat model, and TN rat model administered CBZ treatment. (d) Correlation analysis of DEGs with differential H3K9ac and HDAC3 enrichment in the control, TN rat model, and TN rat model administered CBZ treatment. (e) Gene-browser views of transcriptome, H3K9ac, and HDAC3 profiles of Neurod2, C2cd4c, and Sf3b4 in the control, TN rat model, and TN rat model administered CBZ treatment. Control TN CBZ p-LRRK2 (Ser935) LRRK2 HDAC3 IP Input IgG Control TN CBZ Control TN CBZ Control TN CBZ p-CK2𝛽 (Ser209) CK2 CK2𝛽 LRRK2 p-SRC (Tyr529) SRC SRC HDAC3 GAPDH (a) (b) Control TN CBZ p-ERK1/2 (Thr202, Tyr204) LRRK2 IP Input IgG ERK1/2 p-MEK (Tyr275) ERK1/2 MEK LRRK2 GAPDH (c) (d) Figure 6: The activities of LRRK2 and MAPK signaling pathways in TN rat model. (a) The interaction between LRRK2 and HDAC3 in TN, by HDAC3 IP assay. (b) Phosphorylation of LRRK2, CK2β, and SRC in the control, TN rat model, and TN rat model administered CBZ treatment, by WB assay. (c) Interaction between LRRK2 and ERK1/2 in TN, by LRRK2 IP assay. (d) Phosphorylation of MEK and ERK1/ 2 in the control, TN rat model, and TN rat model administered CBZ treatment, by WB assay. Oxidative Medicine and Cellular Longevity 11 can suppress the activity of c-Jun to alleviate pain [37]. Fur- thermore, D-β-hydroxybutyric acid is reported to not only relieve mechanical and thermal hyperalgesia in rats but also to improve their motor function by reversing the presence of low acetylation and expression of FOXO3a, catalase, and SOD2 in the damaged area [38]. Substantial attenuating effects of CBZ on TN are observed in this study. The number of DEGs in our RNA-seq data, consistent with that reported in a previous study [39], is far higher than genes with differ- ential H3K9ac occupancy, indicating that in addition to H3K9ac, the acetylation levels of other histones also decline in TN. Besides the well-acknowledged inflammatory response and neuronal self-repair in TN [40], the biological functions associated with MAPK activity and the GTP metabolic pro- cess are highlighted in our sequencing data, prompting us to Compressive stress Phosphate group study the activity of HDACs from phosphorylation modifi- H3K9ac cation. We only observe obvious phosphorylation at Ser29 ERK1/2 LRRK2 of HDAC3 in this system. The IP experiments showed one HDAC3 clear band of HDAC3 within the phosphor-Ser/Thr pull down fraction. However, HDAC3 in general has multiple Figure 7: The graphic overview of this study. The phosphate serine and threonine phosphorylation sites [33]. Therefore, groups transported from activated MAPK/ERK pathway and we suspect that the activation of HDAC3 by phosphorylation LRRK2 to HDAC3 causes the repression of H3K9ac and results at different sites determines individual and specific functions in a large number of gene expression silencing in trigeminal in diseases and other biological events. ganglion induced by compressive stress. Consistent with sequencing data, the findings show that the MAPK/ERK signaling pathway substantially contributes or SRC, by IP study. In addition, the binding ability of to providing phosphate groups to LRRK2 and HDAC3. Here, LRRK2 with HDAC3 strengthened in TN and under CBZ we raise the interesting topic of kinase selection for HDAC3 treatment compared with that in the control (Figure 6(a)). phosphorylation in TN. CK2, LRRK2, and SRC are the phos- Next, the protein expression of these kinases did not change phokinases for other HDACs, and they govern the phos- between the control, TN, or CBZ treatment groups, but phorylation of HDAC3 in different tissues [41]. It is robust phosphorylation of LRRK2 (Ser935) was observed in challenging to elucidate the mechanism underlying such TN and CBZ treatment groups relative to the control behavior under special conditions. Furthermore, CK2 and (Figure 6(b)). To trace back to the phosphate groups from SRC are likely to play a role in other HDACs in the patho- LRRK2, LRRK2 pull-down and IP assays of trigeminal root genesis of TN. entry zone of TN were conducted to detect a variety of Overall, we identified a TN regulatory mechanism in enzymes that catalyze phosphate group transmission. We which the phosphate groups transferred from activated ERK and LRRK2 to HDAC3 caused genome-scale deacetylation noticed that MAPK/ERK signaling pathway was the top affected item in TN (Figure 3(c)) and was reported to partic- at H3K9 and resulted in the silencing of a large number of ipate in pathogenesis neuralgia in previous studies [34, 35]. genes (Figure 7). The kinases or important enzymes within We harvested ERK1/2 from the LRRK2 binding proteins this regulatory axis represent potential targets for TN therapy (Figure 6(c)). Consistent with this, the high expression of and prevention. phosphorylated MEK and ERK1/2 indicated that the MAPK/ERK signaling pathway was indeed activated in TN Data Availability and CBZ treatment groups compared with that in the con- trol (Figure 6(d)). The datasets used and/or analyzed during the current study are available from the corresponding authors upon reason- able request. 4. Discussion A large number of preclinical and clinical studies show that Conflicts of Interest histone acetyltransferases (HATs) and HDACs play an important role in the development of neuralgia [36]. Authors declare no conflict of interests. HDACs have a significant analgesic effect, thus drawing much attention from researchers. HDAC1 levels increase in Authors’ Contributions the spinal cord of the spinal nerve ligation rat model, where HDAC1 interacts with a heterodimer composed of c-Jun, W. W. and L. Y. performed all experiments and analyzed the finally activating the JNK signaling pathway to participate data. W. Y. assisted with the animal experiments. H. Z. in the maintenance of neuralgia. LG325 (HDAC1 inhibitor) assisted with the molecular experiments. C. Y. provided 12 Oxidative Medicine and Cellular Longevity [10] E. Gambeta, J. G. Chichorro, and G. W. Zamponi, “Trigeminal assistance with bioinformatics analysis. Q.Y. drafted and neuralgia: an overview from pathophysiology to pharmacolog- revised the manuscript. C. M. designed and supported the ical treatments,” Molecular Pain, vol. 16, article whole project and submitted the manuscript. Wenbin Wei 1744806920901890, 2020. and Yuemin Liu contributed equally to this work. [11] L. Akbarzadeh, T. Moini Zanjani, and M. Sabetkasaei, “Com- parison of anticancer effects of carbamazepine and valproic Acknowledgments acid,” Iranian Red Crescent Medical Journal, vol. 18, no. 10, article e37230, 2016. This work is supported by the Key Biological and Pharma- [12] D. S. Luo, T. Zhang, C. X. Zuo et al., “An animal model for tri- ceutical Projects of Shanghai Science and Technology Com- geminal neuralgia by compression of the trigeminal nerve mission (Grant No. 16411953300), the Clinical Research root,” Pain Physician, vol. 15, no. 2, pp. 187–196, 2012. MDT Project of Shanghai Ninth People’s Hospital, Shanghai [13] Y. Asai, Y. Sakakibara, M. Nadai, and M. Katoh, “Effect of car- Jiao Tong University (Grant No. 201701018), and the Sci- bamazepine on expression of UDP-glucuronosyltransferase ence and Technology Commission of Shanghai Municipality 1A6 and 1A7 in rat brain,” Drug Metabolism and Pharmacoki- (Grant No. 21Y11903500). netics, vol. 32, no. 6, pp. 286–292, 2017. [14] M. Cha, I. Sallem, H. W. Jang, and I. Y. Jung, “Role of transient receptor potential vanilloid type 1 in the trigeminal ganglion Supplementary Materials and brain stem following dental pulp inflammation,” Interna- tional Endodontic Journal, vol. 53, no. 1, pp. 62–71, 2020. Table S1: the list of DEGs with significant H3K9ac and [15] A. M. Bolger, M. Lohse, and B. Usadel, “Trimmomatic: a flex- HDAC3 enrichments. Candidate genes meet with the follow- ible trimmer for Illumina sequence data,” Bioinformatics, ing conditions, FPKM values significantly differential (FC > 2 vol. 30, no. 15, pp. 2114–2120, 2014. or<0.5, p <0:05) comparison between control and TN groups [16] S. Andrews and A. FastQC, Quality Control Tool for High as well as TN and CBZ groups in RNA-seq and ChIP-seq of Throughput Sequence Data, 2013. H3K9ac and HDAC3. (Supplementary Materials) [17] D. Kim, B. Langmead, and S. L. Salzberg, “HISAT: a fast spliced aligner with low memory requirements,” Nature References Methods, vol. 12, no. 4, pp. 357–360, 2015. [18] M. Pertea, G. M. Pertea, C. M. Antonescu, T. C. Chang, J. T. [1] S. Maarbjerg, G. Di Stefano, L. Bendtsen, and G. Cruccu, “Tri- Mendell, and S. L. Salzberg, “StringTie enables improved geminal neuralgia - diagnosis and treatment,” Cephalalgia: an reconstruction of a transcriptome from RNA-seq reads,” international journal of headache, vol. 37, no. 7, pp. 648–657, Nature Biotechnology, vol. 33, no. 3, pp. 290–295, 2015. [19] A. C. Frazee, G. Pertea, A. E. Jaffe, B. Langmead, S. L. Salzberg, [2] S. Peker, O. Kurtkaya, I. Uzun, and M. N. Pamir, “Microanat- and J. T. Leek, “Ballgown bridges the gap between tran- omy of the central myelin-peripheral myelin transition zone of scriptome assembly and expression analysis,” Nature Biotech- the trigeminal nerve,” Neurosurgery, vol. 59, no. 2, pp. 354– nology, vol. 33, no. 3, pp. 243–246, 2015. 359, 2006, discussion 354-9. [20] Y. Shen, S. Zhou, X. Zhao, H. Li, and J. Sun, “Characterization [3] H. Wood, “Trigeminal neuralgia linked to demyelination in of genome-wide DNA methylation and hydroxymethylation multiple sclerosis,” Nature Reviews. Neurology, vol. 16, no. 6, in mouse arcuate nucleus of hypothalamus during puberty p. 298, 2020. process,” Frontiers in Genetics, vol. 11, article 626536, 2020. [4] S. M. Laakso, O. Hekali, G. Kurdo, J. Martola, T. Sairanen, and [21] G. Yu, L. G. Wang, Y. Han, and Q. Y. He, “clusterProfiler: an R S. Atula, “Trigeminal neuralgia in multiple sclerosis: preva- package for comparing biological themes among gene clus- lence and association with demyelination,” Acta Neurologica ters,” Omics: a journal of integrative biology, vol. 16, no. 5, Scandinavica, vol. 142, no. 2, pp. 139–144, 2020. pp. 284–287, 2012. [5] R. Lin, L. Luo, Y. Gong et al., “Immunohistochemical analysis [22] S. Zhou, Y. Shen, S. Zang, X. Yin, and P. Li, “The epigenetic of histone H3 acetylation in the trigeminal root entry zone in role of HTR1A antagonist in facilitating GnRH expression an animal model of trigeminal neuralgia,” Journal of Neurosur- for pubertal initiation control,” Molecular therapy. Nucleic gery, vol. 131, no. 3, pp. 828–838, 2018. acids, vol. 25, pp. 198–206, 2021. [6] R. J. Danaher, L. Zhang, C. J. Donley et al., “Histone deacetyl- [23] Y. Shen, H. Zhao, L. Zhang et al., “The roles of DNA methyl- ase inhibitors prevent persistent hypersensitivity in an orofa- ation and hydroxymethylation at short interspersed nuclear cial neuropathic pain model,” Molecular Pain, vol. 14, article elements in the hypothalamic arcuate nucleus during puberty,” 1744806918796763, 2018. Molecular therapy. Nucleic acids, vol. 26, pp. 242–252, 2021. [7] F. Denk, W. Huang, B. Sidders et al., “HDAC inhibitors atten- [24] B. Langmead and S. L. Salzberg, “Fast gapped-read alignment uate the development of hypersensitivity in models of neuro- with bowtie 2,” Nature Methods, vol. 9, no. 4, pp. 357–359, pathic pain,” Pain, vol. 154, no. 9, pp. 1668–1679, 2013. [8] C. Zhou, Y. Wu, X. Ding, N. Shi, Y. Cai, and Z. Z. Pan, “SIRT1 decreases emotional pain vulnerability with associated CaM- [25] Y. Zhang, T. Liu, C. A. Meyer et al., “Model-based analysis of ChIP-Seq (MACS),” Genome Biology, vol. 9, no. 9, p. R137, KIIα deacetylation in central amygdala,” The Journal of Neuro- science: The Official Journal of the Society for Neuroscience, 2008. vol. 40, no. 11, pp. 2332–2342, 2020. [26] P. Kolodziej-Wojnar, J. Borkowska, Z. Wicik et al., “Alter- [9] Z. Li, Y. Guo, X. Ren et al., “HDAC2, but not HDAC1, regu- ations in the genomic distribution of 5hmC in in vivo aged lates Kv1.2 expression to mediate neuropathic pain in CCI human skin fibroblasts,” International Journal of Molecular rats,” Neuroscience, vol. 408, pp. 339–348, 2019. Sciences, vol. 22, no. 1, p. 78, 2021. Oxidative Medicine and Cellular Longevity 13 [27] Y. Wang, Q. Sun, J. Liang, H. Li, D. M. Czajkowsky, and Z. Shao, “Q-Nuc: a bioinformatics pipeline for the quantitative analysis of nucleosomal profiles,” Interdisciplinary sciences, computational life sciences, vol. 12, no. 1, pp. 69–81, 2020. [28] Q. Lai, W. Du, J. Wu et al., “H3K9ac and HDAC2 activity are involved in the expression of monocarboxylate transporter 1 in oligodendrocyte,” Frontiers in Molecular Neuroscience, vol. 10, p. 376, 2017. [29] H. Ji, Y. Zhou, X. Zhuang et al., “HDAC3 deficiency promotes liver cancer through a defect in H3K9ac/H3K9me3 transition,” Cancer Research, vol. 79, no. 14, pp. 3676–3688, 2019. [30] J. Vecera, E. Bartova, J. Krejci et al., “HDAC1 and HDAC3 underlie dynamic H3K9 acetylation during embryonic neuro- genesis and in schizophrenia-like animals,” Journal of Cellular Physiology, vol. 233, no. 1, pp. 530–548, 2018. [31] H. Patil, C. Wilks, R. W. Gonzalez, S. Dhanireddy, H. Conrad- Webb, and M. Bergel, “Mitotic activation of a novel histone deacetylase 3-linker histone H1.3 protein complex by protein kinase CK2,” The Journal of Biological Chemistry, vol. 291, no. 7, pp. 3158–3172, 2016. [32] K. A. Han, W. H. Shin, S. Jung et al., “Leucine-rich repeat kinase 2 exacerbates neuronal cytotoxicity through phosphor- ylation of histone deacetylase 3 and histone deacetylation,” Human Molecular Genetics, vol. 26, no. 1, pp. 1–18, 2017. [33] J. Seo, G. Guk, S. H. Park et al., “Tyrosine phosphorylation of HDAC3 by Src kinase mediates proliferation of HER2- positive breast cancer cells,” Journal of Cellular Physiology, vol. 234, no. 5, pp. 6428–6436, 2019. [34] Y. Kim, S. Y. Kwon, H. S. Jung et al., “Amitriptyline inhibits the MAPK/ERK and CREB pathways and proinflammatory cytokines through A3AR activation in rat neuropathic pain models,” Korean Journal of Anesthesiology, vol. 72, no. 1, pp. 60–67, 2019. [35] C. L. Li, R. Yang, Y. Sun, Y. Feng, and Y. B. Song, “N58A exerts analgesic effect on trigeminal neuralgia by regulating the MAPK pathway and tetrodotoxin-resistant sodium channel,” Toxins, vol. 13, no. 5, p. 357, 2021. [36] X. Wang, X. Shen, Y. Xu et al., “The etiological changes of acetylation in peripheral nerve injury-induced neuro- pathic hypersensitivity,” Molecular Pain, vol. 14, article 1744806918798408, 2018. [37] M. D. Sanna and N. Galeotti, “The HDAC1/c-JUN complex is essential in the promotion of nerve injury-induced neuro- pathic pain through JNK signaling,” European Journal of Phar- macology, vol. 825, pp. 99–106, 2018. [38] J. Qian, W. Zhu, M. Lu, B. Ni, and J. Yang, “D-β-hydroxybuty- rate promotes functional recovery and relieves pain hypersen- sitivity in mice with spinal cord injury,” British Journal of Pharmacology, vol. 174, no. 13, pp. 1961–1971, 2017. [39] T. Aczel, J. Kun, E. Szoke et al., “Transcriptional alterations in the trigeminal ganglia, nucleus and peripheral blood mononu- clear cells in a rat orofacial pain model,” Frontiers in Molecular Neuroscience, vol. 11, p. 219, 2018. [40] Y. Yao, B. Chang, and S. Li, “Relationship of inflammation with trigeminal neuralgia,” The Journal of Craniofacial Sur- gery, vol. 31, no. 2, pp. e110–e113, 2020. [41] S. Bahl and E. Seto, “Regulation of histone deacetylase activi- ties and functions by phosphorylation and its physiological relevance,” Cellular and Molecular Life Sciences: CMLS, vol. 78, no. 2, pp. 427–445, 2021.

Journal

Oxidative Medicine and Cellular LongevityHindawi Publishing Corporation

Published: May 4, 2022

There are no references for this article.