Abstract
Dysfunctional gene expression in nociceptive pathways plays a critical role in the development and maintenance of neuropathic pain. Super enhancers (SEs), composed of a large cluster of transcriptional enhancers, are emerging as new players in the regulation of gene expression. However, whether SEs participate in nociceptive responses remains unknown. Here, we report a spinal-specific SE (SS-SE) that regulates chronic constriction injury (CCI)-induced neuropathic pain by driving Ntmt1 and Prrx2 transcription in dorsal horn neurons. Peripheral nerve injury significantly enhanced the activity of SS-SE and increased the expression of NTMT1 and PRRX2 in the dorsal horn of male mice in a bromodomain-containing protein 4 (BRD4)-dependent manner. Both intrathecal administration of a pharmacological BRD4 inhibitor JQ1 and CRISPR-Cas9-mediated SE deletion abolished the increased NTMT1 and PRRX2 in CCI mice and attenuated their nociceptive hypersensitivities. Furthermore, knocking down Ntmt1 or Prrx2 with siRNA suppressed the injury-induced elevation of phosphorylated extracellular-signal-regulated kinase (p-ERK) and glial fibrillary acidic protein (GFAP) expression in the dorsal horn and alleviated neuropathic pain behaviors. Mimicking the increase in spinal Ntmt1 or Prrx2 in naive mice increased p-ERK and GFAP expression and led to the genesis of neuropathic pain-like behavior. These results redefine our understanding of the regulation of pain-related genes and demonstrate that BRD4-driven increases in SS-SE activity is responsible for the genesis of neuropathic pain through the governance of NTMT1 and PRRX2 expression in dorsal horn neurons. Our findings highlight the therapeutic potential of BRD4 inhibitors for the treatment of neuropathic pain.
SIGNIFICANCE STATEMENT SEs drive gene expression by recruiting master transcription factors, cofactors, and RNA polymerase, but their role in the development of neuropathic pain remains unknown. Here, we report that the activity of an SS-SE, located upstream of the genes Ntmt1 and Prrx2, was elevated in the dorsal horn of mice with neuropathic pain. SS-SE contributes to the genesis of neuropathic pain by driving expression of Ntmt1 and Prrx2. Both inhibition of SS-SE with a pharmacological BRD4 inhibitor and genetic deletion of SS-SE attenuated pain hypersensitivities. This study suggests an effective and novel therapeutic strategy for neuropathic pain.
Introduction
Neuropathic pain is a leading cause of chronic pain worldwide, with prevalence in the general population estimated at 3 to 17% (Cavalli et al., 2019). Although treatment strategies for neuropathic pain have been improved in recent years, the advanced and effective management approaches are still lacking, largely because the molecular mechanisms underlying the neuropathic pain are not well understood (Zhao et al., 2017; Pan et al., 2021b). Aberrant expression of ion channels, enzymes, and cytokines/chemokines in the dorsal horn of the spinal cord are essential for the genesis of neuropathic pain (Zheng et al., 2023), but how these dysfunctional changes are regulated remains poorly understood.
Super enhancers (SEs) are large clusters of transcriptional enhancers in close genomic proximity that drive expression of key genes defining cell identity (Chen et al., 2004; Pott and Lieb, 2015). The rapid development of genome-wide sequencing technology opened up new avenues in SE research. Whyte and his colleagues redefined SEs, clarified the identification method, and characterized their function in regulating cell identity and disease Whyte et al. (2013). Super enhancer (SE) activity is characterized by strong enrichment for the binding of transcriptional coactivators, such as bromodomain-containing protein 4 (BRD4), mediator 1 (Sabari et al., 2018), monomethylated histone H3 at lysine 4, and acetylated histone H3 at lysine 27 (H3K27ac; Font-Tello et al., 2020). Consequently, the chromatin occupancy of these factors is used as a measure of SE activity. SEs exhibit a unique structure and have stronger transcriptional activation and a stronger regulatory ability for their governed genes than other types of enhancers (Whyte et al., 2013). In particular, SEs can drive the expression of cell-specific or tissue-specific genes and are associated with cellular differentiation, proliferation (Thandapani, 2019), and embryonic development (Agrawal and Rao, 2021). They also participate in pathologic processes in cancer (Liang et al., 2016), diabetes (Sun et al., 2018), systemic lupus erythematosus (Vahedi et al., 2015), and neurodegenerative diseases (Achour et al., 2015). Pharmacological inhibition of BRD4 by JQ1 blocks communication between SEs and oncogene target promoters (Donati et al., 2018); BRD4 inhibition is therefore considered a promising treatment for solid malignancies and blood disorders (Dawson et al., 2011; Lovén et al., 2013; Zhao et al., 2016). In 2013 five SEs associated with pathogenic genes were observed in the brain tissue of patients with Alzheimer's disease (Sun and Zhang, 2016). Subsequent studies have shown that SEs are often activated in specific neuronal cells or tissues (Hnisz et al., 2013). In a mouse model of Huntington's disease, SEs are linked to downregulation of striatal-neuron-specific gene expression (Achour et al., 2015). Isolated neurons from NMDA-treated mouse cerebral cortex had increased neuronal excitability; these neurons also showed increased activity in 295 SEs and decreased activity in 113 SEs, and the SE-driven genes were mostly specific to cerebral cortex neurons. These studies suggest that SEs are abundant in CNS tissue and may be an essential player in neurologic disease.
Here, we report that chronic constriction injury (CCI) of peripheral nerves, a model of neuropathic pain, led to increased SE activity in mouse dorsal horn. We identified a unique native SE that is specifically expressed in mammalian spinal tissues—the spinal-specific super enhancer (SS-SE). Peripheral nerve injury increases SS-SE activity in the ipsilateral dorsal horn. This upregulation is required for the development and maintenance of nerve-injury-induced neuropathic pain via positive governance of Ntmt1 (N-terminal Xaa-Pro-Lys N-methyltransferase 1) and Prrx2 (paired related homeobox 2) gene expression in the dorsal horn. SS-SE may thus be a suitable therapeutic target for neuropathic pain.
Materials and Methods
Animals
Male and female BALB/c mice 7–8 weeks old were used in this study. Mice were supplied by the Experimental Animal Center of Xuzhou Medical University and were raised in a standard facility. They were group housed (four per cage) with ad libitum access to food and water, kept on a 7:00 A.M. to 19:00 P.M. light cycle under conditions of constant humidity and temperature (18–26°C). All procedures used were approved by the Administration Committee of Experimental Animals, Jiangsu Province, China. All of the experimenters were blind to the treatment condition.
Animal surgery
To model neuropathic pain, CCI of the sciatic nerve was conducted as previously reported (Decosterd and Woolf, 2000; Xia et al., 2020). Briefly, mice were anesthetized with 3% sevoflurane. The sciatic nerve was exposed after bluntly separating the fascial plane between the gluteus superficialis and biceps femoris muscles, and the nerve was then loosely ligated with 4-0 silk thread around four sites at intervals of ∼1 mm. Sham animals received identical surgery but without the ligation.
Chromatin immunoprecipitation sequencing
The lumbar (L)3–L5 segments of the ipsilateral/contralateral dorsal spinal cord were separated rapidly and snap frozen in liquid nitrogen for subsequent chromatin immunoprecipitation sequencing (ChIP-seq) analysis (Pan et al., 2021a). ChIP and sequencing were performed with an H3K27ac antibody (Shanghai Kangcheng Biotech). The sequencing data were compared with the mouse genome (MM10) in BowTie software (version 2.1.0), and the data were processed in MACS (Model-based Analysis of ChIP-Seq) version 1.4.2 software. The relevant SE coding and noncoding genes were annotated according to the latest Arraystar LncRNA UCSC and UCSC RefSeq databases.
RNA isolation and qRT-PCR analysis
Spinal samples were homogenized in Trizol (catalog #9109, Takara). Isolated RNA was reverse transcribed to cDNA (catalog #R233-01, Vazyme) and assayed by qRT-PCR with SYBR Premix Ex Taq (catalog #RR820A, Takara) on a StepOne system (Applied Biosystems). The expression levels of the target genes were quantified using the 2-(ΔΔCT) method normalized against U6 (an internal control) expression level. Primers for specific genes used in this experiment are listed bellow: Prrx2, Forward (F), 5′-CGTGGCACCAAACGAAAGAAG-3′, Reverse (R), 5′-GTAGTGTGT GCGCTCAAATACA-3′; Ntmt1, F, 5′-ACTGTGGCGCTGGTATTGGA-3′, R, 5′-GTCATAGGAGGCAGGCTCA-3′; Ptges, F, 5′-GGATGCGCTGAAACGTGGA-3′, R, 5′-CAGGAATGAGTACACGAAGCC-3′; Asb6, F, 5′-GGCTTCCGCAGGACTACTTT T-3′, R, 5′-AGAATGGAGAGTGAGCCTTCTG-3′; and U6, F, 5′-GCTTCGGCAGCACATATACTAA-3′; R, 5′-CGAATTTGCGTGTCATCCTT-3′.
Immunofluorescence
Immunofluoresence was performed as described previously (Zhang et al., 2023). Briefly, mice were perfused with 4% PFA and the spinal cords rapidly dissected then postfixed with 4% PFA at 4°C overnight, followed by dehydration in 5, 15, and 30% sucrose solution (w/v) overnight; 15-µm-thick sections were cut on a cryostat (CM1950, Leica) and mounted on slides. After antigen repair with citrate buffer and blocking with 5% BSA, the sections were incubated with primary antibodies at 4°C overnight. The following primary antibodies were used in this study: rabbit NTMT1 antibody (1:100; catalog #A7098, Abclonal), PRRX2 antibody (1:200; catalog #DF9729, Affinity), NeuN antibody (1:200; catalog #ab104224, Abcam), glial fibrillary acidic protein (GFAP) antibody (1:200; catalog #G3893, Sigma-Aldrich), and induction of brown adipocytes 1 (IBA1) antibody (1:200; catalog #ab5076, Abcam). The following secondary antibodies were used in this study: Alexa Fluor 488 donkey anti-mouse IgG, Alexa Fluor 488 donkey anti-goat IgG, and Alexa Fluor 594 donkey anti-rabbit IgG (1:500; Thermo Fisher Scientific). Sections were incubated with the corresponding secondary antibodies at room temperature for 1 h. Images were acquired with a STELLARIS 5 Confocal Microscope (Leica).
Western blotting
For immunoblot analysis, whole-cell protein lysates were prepared from dissected dorsal spinal cord after CCI. Protein lysates (20–50 μg) were separated with 12% SDS-PAGE gel and transferred to PVDF membrane using the Mini-PROTEAN Tetra System (Bio-Rad). The membrane was blocked in 5% nonfat dried milk for 1 h at room temperature with shaking, probed with the following primary antibodies: anti-NTMT1 (1:1000; catalog #A7098, Abclonal), anti-PRRX2 (1:500; catalog #DF9729, Affinity), or control GAPDH antibody (1:5000; catalog #60004-1-Ig, Proteintech) overnight at 4°C, followed by washing and incubation with HRP-labeled goat anti-rabbit IgG (H + L; catalog #A0208, Beyotime) for 1 h at room temperature. Protein blots were visualized via chemiluminescence detection (catalog #P0018FS, Beyotime) with an Alliance-Q9 Advanced system. The immunoblot bands from three independent experiments were evaluated in ImageJ software (version 6.0).
Plasmid construction and lentivirus package
Plasmids were prepared according to previously described methods (Pan et al., 2021a). For Ntmt1/Prrx1 overexpression (via Lenti-Ntmt1 and Lenti-Prrx1), primers Prrx2, F, 5′-AATTCGAATTTAAATCGGATCCATGGACAGCGCGGCCGCCGCCT-3′, R, 5′-GATCCTTGCGGCCGCGGATCCTCAGTTCACTGTGGGCACCTG-3′; and Ntmt1, F, 5′-AATTCGAATTTAAATCGGATCCATGACGAGCGAGGTGATTGAG-3′, R, 5′-GATCCTTGCGGCCGCGGATCCTCATCTCAGGGCAAAGCTGTAGA-3′ were prepared with the BamHI restriction site. Three micrograms of PCDH-CMV-MCS-EF1α-gfp-T2A-Puro vector were digested by BamHI (catalog #R0136L, New England BioLabs), then assembled to the PCDH-expressing vector following the instructions in the ClonExpress II One Step Cloning Kit (Vazyme) manual.
For CRISPR-Cas9 guide RNA (gRNA) vector construction, the gRNA sequences (SE1, gRNA-74, Sense (S), 5′-CGGTGCTTTTTTGAATTCGTGTGGTCCGTGTGCCCTGT-3′; Anti-sense (AS), 5′-CCTAGCT AGCGAATTCACAGGGCACACGGACCA-3′; gRNA-107, S, 5′-CGGTGCTTTTT TGAATTCCAGAGGCTCTCGGAGAGGCC-3′; AS, 5′-CCTAGCTAGCGAATTCG GCCTCTCCGAGAGCCTCTG-3′; SE2, gRNA-134, S, 5′-CGGTGCTTTTTTGAA TTCCAAAACTGAACATGGTGCGC-3′; AS, 5′-CCTAGCTAGCGAATTCGCGC ACCATGTTCAGTTTTG-3′; gRNA-91, S, 5′-CGGTGCTTTTTTGAATTCGACCT ACCTGCTTCAGCCTT-3′; AS, 5′-CCTAGCTAGCGAATTCAAGGCTGAAGCAGGTAGGTC-3′; SE3, gRNA-104, S, 5′-CGGTGCTTTTTTGAATTCTACATCACT TTCCAGTCCTC-3′; AS, 5′-CCTAGCTAGCGAATTCGAGGACTGGAAAGTGATGTA-3′; gRNA-187, S, 5′-CGGTGCTTTTTTGAATTCTGGTCACCAGAGTCCCT GCC-3′, AS, 5′-CCTAGCTAGCGAATTCGGCAGGGACTCTGGTGACCA-3′) were respectively assembled to Lenti-CRISPR-V2 plasmid digested with EcoRI (catalog #B7202S, New England BioLabs) via the ClonExpress II One Step Cloning Kit (Vazyme). All constructs were confirmed by Sanger sequencing.
For lentivirus packaging, the constructed core plasmid (16 µg) and two envelope plasmids, PSPAX2 (12 µg) and PMD2G (4.8 µg), were cotransfected into HEK293T cells in 10 cm dishes with Lipofectamine 3000 (catalog #L3000008, Invitrogen). The supernatant was collected 48 h after transfection and concentrated by a Centricon Plus-70 filter unit (catalog #UFC910096, Millipore). The titers of lentivirus used in the experiment were 108 TU/ml, and knockdown and overexpression efficacy was validated by qRT-PCR in spinal cord in vitro and in vivo.
Cell line culture and transfection
HEK293T cells or HT22 cells were cultured in DMEM (catalog #D5796, MilliporeSigma) containing 10% v/v Gibco fetal bovine serum (catalog #10091148, Thermo Fisher Scientific) at 37°C in a humidified incubator with 5% CO2. siRNAs or plasmids were transfected into the cells with ExFect Transfection reagent (catalog #T101, Vazyme) according to instructions from the manufacturer.
Lentivirus, siRNA, gRNA, and inhibitor delivery
Intrathecal injections were performed as previously described (Pan et al., 2019). Briefly, mice were held firmly by the pelvic girdle, and a 30 gauge needle attached to a 25 µl microsyringe was inserted between the L5 and L6 vertebrae. Proper insertion of the needle into the subarachnoid space was verified by a slight flick of the tail after sudden advancement of the needle. A total of 5 µl of 20 μm siRNA, lentivirus, or 250 μm JQ1 inhibitor was administered daily for 2 consecutive days (siRNA-Prrx2, 367S, 5′-GGAAUCGAACCACAUUCAATT-3′; 367AS, 5′-UUGAAUGUGGUUCGAUUC CTT-3′; 564S, 5′-GCCUCGUUGCUCAAGUCUUTT-3′; 564AS, 5′-AAGACUUGAG CAACGAGGCTT-3′; siRNA-Ntmt1, 347S, 5′-GCGCUGGUAUUGGAAGGAUTT-3′; 347AS, 5′-AUCCUUCCAAUACCAGCGCTT-3′; 548S, 5′-GCAAGAGGGUGAGG AACUATT-3′; 548AS, 5′-UAGUUCCUCACCCUCUUGCTT-3′). Animals receiving intrathecal injections of scrambled siRNA, the empty vector, or DMSO vehicle were used as control groups.
Behavioral analysis
All mice were habituated for 1–2 h every day for 2–3 d before basal behavioral testing. For the mechanical pain test, we measured paw withdrawal frequency (PWF) to two calibrated von Frey filaments (0.07 g and 0.4 g, Stoelting; Pan et al., 2021b). Briefly, mice were allowed to habituate for 30 min. Each von Frey filament was applied to the plantar surface of both hindpaws 10 times, with an interstimulus interval >10 min. Quick withdrawal of the paw was regarded as a positive response. The number of positive responses among the 10 applications was recorded as the percentage withdrawal frequency [(number of paw withdrawals/10 trials) × 100 = % response frequencies].
For the thermal withdrawal test, a Model 336 Analgesia Meter (Series 8, IITC Life Science) was used to measure paw withdrawal latencies (PWLs) to noxious heat stimuli. Briefly, mice were placed in a plastic cage that allowed good light transmission. A focused, radiant noxious light source was applied to the plantar surface of the hindpaw. The light beam was automatically turned off when the paw was lifted quickly, and the duration between stimulus onset and offset was recorded as the PWL. Five trials were performed at 5 min intervals. A 20 s cutoff time was used to avoid tissue damage to the paw.
ChIP assay
ChIP was performed according to instructions from the manufacturer for the ChIP Assay Kit (catalog #P2078, Beyotime). Briefly, freshly dissected L3–L5 spinal cord was collected quickly and incubated in 1% formaldehyde for 7 min at room temperature. The cross-linking reaction was stopped with 125 mm glycine. After washing twice with ice-cold PBS containing protease inhibitor cocktail II, tissue pieces were resuspended with SDS lysis buffer and homogenized into a single-cell suspension by Dounce homogenizer. Then, the samples were sonicated by an ultrasonic instrument (Scientz) for five cycles (10 s on, 30 s off). The chromatin supernatant was obtained by centrifugation.
Samples were diluted 10-fold in ChIP Dilution Buffer, and 10% of the supernatant was stored at −20°C as the input control. The remainder was precleared at 4°C for 30 min with Protein A + G Agarose/Salmon Sperm DNA beads. After centrifugation, the chromatin supernatant was immunoprecipitated at 4°C overnight with IgG, H3K27ac antibody (catalog #ab4729, Abcam), then Protein A + G Agarose/Salmon Sperm DNA beads were added to each IP reaction and incubated at 4°C for 1 h with rotation. Chromatin was washed and eluted from the beads according to the instructions from the manufacturer. After removing the beads, cross-linking was reversed with 5 m NaCl at 65°C for 4 h and 20 μg proteinase K incubation for 1 h at 45°C. The eluted DNA was cleaned according to instructions from the manufacturer before qRT-PCR. The region-specific primers for super enhancers (SE1, F, 5′-ACTCCGAA GGCGTAGCGCT-3′, R, 5′-ATGCTGGAGACGTCCTGGCT-3′; SE2, F, 5′-TCGTTCAGCCTTTGAAGTGC-3′, R, 5′-CTGGGTTCAAGCCTCAGGA-3′; SE3, F, 5′-AGGAAGCCACGTTGGTGGA-3′, R, 5′-AGTGGCTGCTTTCCTTATCTC-3′) or for Prrx2 (F, 5′-GTAGCAAATTCGAGGCTAATCT-3′, R, 5′-AGAGTCTCTGGAGGTGATGG-3′) and Ntmt1 (F, 5′-ATCTGGACGTGAGCTGCCTG-3′, R, 5′-GGCTGGACTCAAGATCGCTC-3′) were used in this study. CHIP enrichment efficiency was calculated according to the Ct value by the following formula: Enrichment percentage = 2% × 2(Ct input sample-Ct IP sample), where Ct is the threshold cycle of PCR.
Reporter vector construction and luciferase reporter assay
To design the fragments of the three signal sites (SE1, SE2, and SE3) corresponding to the super enhancer, the fragments were amplified by specific pairs of SE1, SE2, and SE3 primers (Luci-SE1, F, 5′-GCCTCGAGGAGCTCACGCGTATCATGAAGTCCCTGCGCGT-3′, R, 5′-AGCTTCTGCAGATCTACGCGTCTTAAGGAAACGATGTGCCT-3′; Luci-SE2, F, 5′-GCCTCGAGGAGCTCACGCGTTTCAGTTATTGAGATGCGGA-3′, R, 5′-AGCTTCTGCAGATCTACGCGTATGCCTATATCGCTAGCACT-3′; Luci-SE3, F, 5′-GCCTCGAGGAGCTCACGCGTTACCTACTAAGTGCATCTTG-3′, R, 5′-AGCTTCTGCAGATCTACGCGTGGTCTGAACCGGTAGATCG-3′) in mouse genomic DNA and cloned into a pGL6 plasmid (Beyotime) with digestion by MluI enzymes (catalog #R0198L, New England BioLabs). The SE1/SE2/SE3 fragments were inserted into the digested pGL6 plasmid with a ClonExpress II One Step Cloning Kit (Vazyme); pGL6 was used as the control vector. Five hundred nanograms of reporter plasmids SE1/SE2/SE3/pGL6 were transfected into HEK293T cells in a 24-well plate using the ExFect Transfection Reagent (catalog #T101, Vazyme). Cells were lysated and subjected to luciferase assays using the Dual Luciferase Reporter Gene Assay Kit (catalog #RG027, Beyotime) at 36–48 h after transfection, according to the instructions from the manufacturer.
Experimental design and statistical analysis
The size of the experimental groups was determined based on the literature of the field and our previous experience (Pan et al., 2019; 2021b). Each experiment presented was repeated in multiple animals (three to eight mice per sample, see above, ChIP assay, qRT-PCR analysis, et al.). The relevant animals were randomly allocated to experimental groups in all experiments. To specifically regulate the activity of super enhancer or the level of gene expression in dorsal horn, intrathecal injection was used in the delivery of pharmacological inhibitor, virus, or siRNA as described previously (Pan et al., 2021a; Zhang et al., 2023). Double-blind tests were adopted in all behavioral experiments. Additionally, to exclude the effect of locomotor impairment on the pain behavior data, locomotor function was included in this study. Data are presented as mean ± SEM in bar plots in figures. Direct comparisons were made with two-tailed unpaired Student's t tests, multiple t tests, one-way ANOVAs, or two-way ANOVAs using GraphPad version 8.0 as described previously (Pan et al., 2021a). When an ANOVA showed a significant difference, pairwise comparisons between means were tested by the post hoc Tukey method; p < 0.05 was considered statistically significance in all analyses
Data availability
The data that support the findings of this study are available on reasonable request.
Results
Identification of a spinal-specific super enhancer in dorsal spinal horn
SEs play a prominent role in the modulation of cell fate. They are valuable biomarkers of disease-related genes and, as such, may also form therapeutic targets for disease intervention (Bal et al., 2022). We identified SEs in the dorsal spinal horn of mice with CCI (a model of neuropathic pain) through ChIP-Seq against the H3K27ac modification. The results showed a significant increase in the ChIP-Seq H3K27ac signal in the dorsal horn on day 7 after CCI surgery compared with sham surgery, especially in the gene coding region (Fig. 1a). We identified 418 SE-associated genes in the dorsal horn of sham mice but 569 SE-associated genes in the dorsal horn of CCI mice, indicating that more genes were driven to express in the dorsal horn with neuropathic pain progression (Fig. 1b,c). Next, we performed GO (Gene Ontology) and KEGG (Kyoto Encyclopedia of Genes and Genomes) analysis to investigate the functional enrichment of the identified SE-related targets and pathways (Fig. 1d,e). The SEs were associated with genes enriched in RNA polymerase transcription, signal transduction, and nervous system development. Moreover, peripheral nerve injury activated regulation of the MAPK signaling, axon guidance, and dopaminergic synapse pathways in the dorsal horn (Fig. 1d,e).
Ntmt1/Prrx2-linked-SE is a spinal-specific super enhancer. a, Super enhancer signal distribution. The signal strength of SE signals from −5000 to about +5000 of all genes (transcription start site denoted as +1) were detected in the dorsal horn of male mice with CCI or sham surgery. Dorsal horns were collected on day 7 after surgery and subjected to anti-H3K27ac ChIP-Seq; n = 2 repeats. Male mice were used in the experiments unless female was mentioned. b, c, SEs were identified in the dorsal horn of sham mice (b) or CCI mice (c) according to the ROSE algorithm. d, e, Gene Ontology (d) and KEGG pathway (e) analysis for activated SE-related genes in the dorsal horn on day 7 after CCI surgery. f, The chromosome location, potential associated genes (Ntmt1, Prrx2, Asb6, and Ptges), and the constituent enhancers (E1–E3) of a super enhancer with significantly increased signal after nerve injury. Input is the noise signal from secondary antibody IgG in ChIP-Seq. g, Identification of spinal cord–specificity of the Ntmt1/Prrx2-SE by comparing enhancer characteristics in the same genomic region (Chr 2, 30645088–30797673) from five nervous system tissues and six peripheral tissues. The constituent enhancer information for the other 11 tissues was downloaded from the SEA database of super enhancers.
When we profiled the SEs, we found one (Chr 2, 30645088-30797673, located upstream of Ntmt1, Prrx2, Asb6, and Ptges genes) with a particularly strong H3K27ac signal (Fig. 1f). Through searching the SEA (Similarity Ensemble Approach) database, an SE database (Wei et al., 2016; Chen et al., 2020); we found that this SE was not found in six peripheral tissues (heart, kidney, limb, lung, stomach, testis), or in five other nervous system tissues (cortex, hindbrain, olfactory bulb, midbrain, whole brain; Fig. 1g). These data indicate that this SE appears to be specific to the spinal cord. We therefore termed it SS-SE.
Given that BRD4 plays an integral role in the formation of SEs and promotes gene expression by recruiting transcription factors and RNA polymerase in various tissues (Hnisz et al., 2013; Yang et al., 2022), we investigated whether peripheral nerve injury induced changes in BRD4 expression. We did not observe any differences in BRD4 expression in the dorsal horn on day 7 after CCI surgery (Fig. 2a). We next used a ChIP-PCR assay to assess the binding activity of BRD4 to the SS-SE clusters; each SS-SE cluster fragment could be amplified from the complex immunoprecipitated with BRD4-antibody in nuclear fractions from sham dorsal horn. This amplification did not occur with normal control serum, suggesting specific binding of BRD4 to SS-SE in the dorsal horn. CCI significantly increased the binding activity in the ipsilateral dorsal horn 1.65-, 1.51-, and 1.55-fold on day 7 after surgery, relative to the activity in sham mice at the same time point (Fig. 2b). Given that RNA polymerase II executes the gene transcription after the formation of super enhancer, we examined the binding activities of RNA polymerase II to the SS-SE clusters by pharmacological inhibition by JQ1, a small-molecule bromodomain and extra-terminal (BET) domain inhibitor, can suppress the activity of SEs by blocking the binding of BRD4 (an SE-specific binding protein) to acetylated H3K27 at SEs (Deng et al., 2020). Similarly, the binding activity of RNA polymerase II to the upstream promoter of Ntmt1, Prrx2 genes were significantly enhanced on day 7 after CCI surgery (Fig. 2c). These results indicate that BRD4 may organize assembly of the SS-SE.
Suppression of Ntmt1/Prrx2-SE activity by a pharmacological BRD4 inhibitor alleviates neuropathic pain sensitivities. a, The expression of BRD4 in dorsal horn on day 7 after CCI surgery was not changed; n = 6 mice/group. b, Binding activity of BRD4 to three SS-SE clusters. ChIP-PCR with anti-BRD4 in the dorsal horn of Sham mice or CCI mice on day 7 after surgery; n = 4 mice/group' *p < 0.05, **p < 0.01 versus Sham. c, Binding activity of RNA polymerase II (RPII) to Ntmt1/Prrx2-SE clusters. ChIP-PCR with anti-RPII in the dorsal horn of Sham mice or CCI mice on day 7 after surgery; n = 3 mice/group; ***p < 0.001 versus Sham+DMSO; #p < 0.05, ##p < 0.01 versus CCI+DMSO. d, Intrathecal injection of JQ1 7 d after CCI reduced the activity of Ntmt1/Prrx2-SE in ipsilateral dorsal horn 24 h after injection. SE activity was identified by H3K27ac-ChIP-qPCR. DMSO was used as the vehicle control; n = 4 mice/group, *p < 0.05, **p < 0.01, ***p < 0.001 versus Sham+DMSO; #p < 0.05, ###p < 0.001, CCI+DMSO. e–j, Time course of the effect of intrathecal injection of JQ1 or vehicle control into CCI or sham male (e–g) and female (h–j) mice on the ipsilateral PWF to 0.07 g and 0.4 g von Frey filaments and on PWLs to heat stimuli; n = 8 mice/group, ***p < 0.001, versus the Sham+DMSO group at the corresponding time points; #p < 0.05, ##p < 0.01, ###p < 0.001 versus CCI+DMSO at the corresponding time points; two-way ANOVA, post hoc Tukey's tests. Red arrows indicate CCI or sham surgery. Blue arrows indicate JQ1 or vehicle control injection.
SS-SE is required for neuropathic pain
We next explored whether SS-SE participated in neuropathic pain. Thus, we first examined the effect of pharmacological inhibition by JQ1 on neuropathic pain. Intrathecal injection of JQ1 blocked the CCI-induced increase in H3K27ac binding to SS-SE in ipsilateral dorsal horn 24 h after CCI injury, as shown by the reduced H3K27ac activity for each enhancer cluster (SE1–SE3), compared with the DMSO control (Fig. 2d). Consistent with previous behavioral studies (Pan et al., 2021a), CCI led to mechanical allodynia (0.07 g and 0.4 g von Frey stimuli; Fig. 2e,f) and heat hyperalgesia (Fig. 2g) on the ipsilateral side in DMSO-injected male mice postsurgery. These nociceptive hypersensitivities were attenuated in the JQ1-injected male mice from 6 to 48 h after injection (Fig. 2e–g). Since sex dimorphism in chronic pain is a well-recognized clinical phenomenon and is increasingly becoming an important factor in pain (Mogil, 2020; Luo et al., 2021), we examined the effect of JQ1 administration on female mouse behavior. Similarly, CCI-induced mechanical and thermal pain was alleviated in the JQ1-injected female mice from 6 to 48 h after injection (Fig. 2h–j). We did not observe locomotor changes in either the JQ1 or DMSO mice (Table 1).
Test of locomotor function after injection of manipulation tools
As JQ1 is a nonspecific inhibitor of SS-SE, we next used the CRISPR gene-editing system (Li et al., 2020) to further confirm the specific regulatory role of SS-SE in neuropathic pain. Two gRNAs were designed for per SE cluster. gRNA-74 and gRNA-107 (+74 to about +93, +107 to about +126, first nucleotide in the SE cluster designated as +1) for SE1, gRNA-91 and gRNA-134 (+91 to about +110, +134 to about +153) for SE2, and gRNA-104 and gRNA-187 (+104 to about +123, +187 to about +206) for SE3-were designed and packaged into lentivirus vector, as previously described (Minton et al., 2018). By intrathecally coinjecting gRNA and Lenti-CRISPR-Cas9 (Cas9), we can measure the effect of specific deletion of dorsal horn SS-SE on pain behavior. Coinjection of Cas9 and gRNA-74 but not gRNA-107 into CCI mice 7 d after surgery reversed the increased binding of H3K27ac to SE1 (Fig. 3a). As expected from this result, coinjection of gRNA-74 and Cas9 ameliorated CCI-induced heat and mechanical hypersensitivity (Fig. 3b–d). Coinjection of Cas9 with gRNA-134 but not gRNA-91 for SE2 (Fig. 3e), and with gRNA-104 but not gRNA-187 for SE3 (Fig. 3i) also produced a significant knockout effect in the binding of H3K27ac to SS-SE in the dorsal horn and resulted in attenuation of nerve-injury-induced heat and mechanical hypersensitivity (Fig. 3f–h,j–l). Impairment of locomotor was not observed in these gRNAs or their scrambled control mice (Table 1). Thus, our data strongly suggest that SS-SE could be a therapeutic target for neuropathic pain induced by peripheral nerve injury.
CRISPR/Cas9-mediated genetic deletion of Ntmt1/Prrx2-SE attenuates nerve-injury-induced nociceptive hypersensitivity. a, Agarose gel electrophoresis showing the activity of the Ntmt1/Prrx2-SE cluster in dorsal horn on day 5 after intrathecal injection of gRNA-74 (74) or gRNA-107 (107) and Lenti-CRISPR/Cas9 for SE1 in mice 7 d after CCI or sham surgery; n = 3–4 mice/group, *p < 0.05, **p < 0.01 versus Sham+Scr; ##p < 0.01 versus CCI+Scr. Top left, Diagram illustrates CRISPR/Cas9-mediated SEs knockout strategy by deleting the two regions flanking the enhancer cluster of Ntmt1/Prrx2-SE. gRNA, guide RNA. The number following gRNA indicates the gRNA location in each SE cluster (the first nucleotide of each cluster is designated as +1). b–d, Time course of the effect of intrathecal injection of gRNA-74 and Lenti-CRISPR/Cas9 in CCI or sham mice on mechanical allodynia (b, c) and heat hyperalgesia (d). Red arrows indicate CCI or sham surgery. Blue arrows indicate gRNAs and CRISPR-Cas9 injection; n = 8 mice/group, ***p < 0.001 versus Sham+Scr; #p < 0.05, ##p < 0.01, ###p < 0.001 versus CCI+Scr. e, The activity of the Ntmt1/Prrx2-SE2 cluster in the dorsal horn on day 5 after intrathecal injection of sgRNA-134 (134) or gRNA-91 (91) and CRISPR/Cas9 for SE2 in CCI or sham mice; n = 3–4 mice/group, *p < 0.05 versus Sham+Scr, #p < 0.05 versus CCI+Scr. f–h, Time course of the effect of intrathecal injection of sgRNA-134 and CRISPR/Cas9 in CCI or sham mice on mechanical sensitivity (f, g) and heat hyperalgesia (h); n = 8 mice/group, **p < 0.01, ***p < 0.001 versus Sham+Scr mice; ##p < 0.01, ###p < 0.001 versus CCI+Scr mice. i, Ntmt1/Prrx2-SE3 activity on day 5 after intrathecal injection of gRNA-104 (104) or gRNA-187 (187) and CRISPR/Cas9 for SE3 in the dorsal horn 7 d after CCI or sham surgery; n = 3–4 mice/group, *p < 0.05 versus Sham+Scr, #p < 0.05 versus CCI+Scr. j–l, The effect of intrathecal injection of sgRNA-104 and CRISPR/Cas9 in CCI or sham mice on mechanical sensitivity (j, k) and heat hyperalgesia (l) stimuli; n = 8 mice/group, ***p < 0.001 versus Sham+Scr; ##p < 0.01, ###p < 0.001 versus CCI+Scr.
SS-SE-associated Ntmt1 and Prrx2 are upregulated in dorsal horn neurons following peripheral nerve injury
SEs are emerging as essential players in gene transcription. Thus, we next set about to examine whether SS-SE drives expression of its four downstream genes—two overlapping genes Ntmt1 and Asb6 and two distal genes Prrx2 and Ptges. First, we measured the mRNA expression level of these genes in the ipsilateral L3/4 dorsal horn after nerve injury. Ntmt1 and Prrx2 mRNA were significantly increased in a time-course-dependent manner (Ntmt1, 188% on day 7 and 221% on day 14; Prrx2, 133% on day 7 and 174% on day 14; Fig. 4a,b) in dorsal horn after CCI surgery, compared with contralateral side; however, no increases were observed for Asb6 or Ptges (Fig. 4c,d). Of note, none of the four genes showed altered expression in ipsilateral DRG after CCI surgery (Fig. 4e). Levels of both NTMT1 and PRRX2 protein expression in the ipsilateral L3/4 dorsal horn were significantly enhanced on days 7 and 14 after CCI but not on days 0–3 (Fig. 4f,g). Consistent with the results for mRNA in the DRG, CCI did not change the expression level of NTMT1 or PRRX2 protein in the DRG on day 7 after surgery (Fig. 4h).
Nerve injury upregulates SE-related Ntmt1 and Prrx2 in the dorsal horn. a–d, Levels of four potential SE-associated genes—Ntmt1 (a), Prrx2 (b), Asb6 (c), and Ptges (d)—in the ipsilateral dorsal horn up to 14 days after CCI of the unilateral sciatic nerve; n = 6 mice/time point/group, *p < 0.05, **p < 0.01, ***p < 0.001 versus the corresponding contralateral side; two-way ANOVA followed by post hoc Tukey's tests. e, mRNA expression levels of Ntmt1, Prrx2, Ptges, and Asb6 mRNA in ipsilateral DRG on day 7 after CCI or Sham surgery of the unilateral sciatic nerve; n = 4 mice/group; CCI group versus Sham group; two-tailed unpaired Student's t test. f, g, Expression levels of NTMT1 (f) and PRRX2 (g) protein in the ipsilateral dorsal horn up to 14 days after CCI surgery of the unilateral sciatic nerve; n = 4 mice/time point/group, *p < 0.05, **p < 0.01, ***p < 0.001 versus the corresponding contralateral side; two-way ANOVA followed by post hoc Tukey's tests. h, Expression levels of NTMT1 (top) and PRRX2 (bottom) in the ipsilateral DRG on day 7 after CCI or Sham surgery; n = 4 mice/group; CCI group versus Sham group; two-tailed unpaired Student's t test. i–l, Combined immunofluorescence staining for NTMT1 (i, j) or PRRX2 (k, l, red) and NeuN (a neuronal marker, green), GFAP (an astrocyte marker, green), and IBA1 (a microglial marker, green) and their overlay analysis in ipsilateral spinal cord of naive mice; n = 3 repeats. Scale bar, 50 μm.
On the basis of these results, we focused on Prrx2 and Ntmt1 in the subsequent experiments. We used cell-type-specific immunofluorescence staining to identify the distribution of PRRX2 and NTMT1 in the spinal cord. We observed that ∼82.9% of NTMT1-positive cells were colabeled with NeuN (a neuron-specific marker). A few NTMT1-positive cells (7.3%) were colabeled with GFAP (an astrocyte-specific marker), and 4.9% were colabeled with IBA1 (a microglia-specific marker) (Fig. 4i,j). Similarly, the colocalization of PRRX2 in spinal neurons and glial cells was 91.2% with NeuN, 3% with GFAP, and 3.6% with Iba1 (Fig. 4k,l), indicating that Prrx2 and Ntmt1 are predominantly expressed in neurons in the dorsal horn.
SS-SE drives Ntmt1 and Prrx2 upregulation after nerve injury
We hypothesized that SS-SE drives the transcription of spinal Ntmt1 and Prrx2 after peripheral nerve injury. To test this, we conducted in vitro and in vivo tests. We constructed the luciferase reporters pGL6-SE1, pGL6-SE2, and pGL6-SE3 inserted by the SS-SE cluster in the pGL6 promoter (Fig. 5a); pGL6 was used as the control. Forty-eight hours after transfection, pGL6-SE1 robustly increased the activity of the reporter, pGL6-SE3 slightly elevated the activity of the reporter, and pGL6-SE2 did not alter the activity of the reporter (Fig. 5a). These in vitro data indicate that SS-SE1 and SS-SE3 may play an important role in the promotion of gene expression. In the in vivo test, we found that compared with the DMSO control intrathecal injection of JQ1 7 d after CCI (postinjection) abolished the increase in Ntmt1 and Prrx2 mRNA (Fig. 5b,d) and protein (Fig. 5c,e) levels in the ipsilateral L3/4 dorsal spinal horn 24 h after injection. Intrathecal coinjection of Cas9 and gRNA-74 (Fig. 5f,g), gRNA-134 (Fig. 5h,i), or gRNA-104 (Fig. 5j,k) reversed the CCI-induced increase in NTMT1 and PRRX2 protein in the dorsal horn 7 d after injection in CCI mice. Collectively, these results suggest that SS-SE drives the expression of Ntmt1 and Prrx2 in the dorsal horn after nerve injury.
SS-SE drives Ntmt1 and Prrx2 upregulation after nerve injury. a, Luciferase activity of a reporter inserted by the individual Ntmt1/Prrx2-SE clusters (pGL6-SE1, 2, and 3; SE1, 2, and 3, respectively) at 48 h after transfection of pGL6-SE1, 2, or 3 or the pGL6 control and PBS negative control (Control) in HEK-293T cells; n = 3 repeats/treatment, ***p < 0.001 versus the pGL6 control group; one-way ANOVA followed by post hoc Tukey's tests. b–e, The effect of suppressing Ntmt1/Prrx2-SE activity by intrathecal injection of the BRD1 inhibitor JQ1 or vehicle control (DMSO) on Ntmt1 (b, c) and Prrx2 (d, e) mRNA and protein expression in the dorsal horn 7 d after CCI or Sham surgery. Data were obtained 6 h after the intrathecal injection; n = 5 mice/group, *p < 0.05, ***p < 0.001 versus Sham+DMSO; #p < 0.05, ###p < 0.001 versus CCI+DMSO. f, g, Effect of Ntmt1/Prrx2-SE1 cluster knockout by CRISPR-Cas9 on NTMT1 (f) and PRRX2 (g) protein levels in the dorsal horn. SE1-sgRNA and Lenti-CRISPR/Cas9 were coinjected 7 d after CCI or Sham surgery, and data were obtained 5 d after coinjection; n = 5 mice/group, *p < 0.05, **p < 0.01 versus Sham+Scr; #p < 0.05 versus CCI+Scr; two-way ANOVA with repeated measures followed by post hoc Tukey's tests. h, i, NTMT1 (h) and PRRX2 (i) protein levels in the dorsal horn 5 d after coinjection of SE2-sgRNA and Lenti-CRISPR/Cas9 in 7 d post-CCI or after sham surgery mice; n = 4 mice/group, **p < 0.01 versus Sham+Scr; #p < 0.05, ###p < 0.001 versus CCI+Scr; two-way ANOVA with repeated measures followed by post hoc Tukey's tests. j, k, Expression levels of NTMT1 (j) and PRRX2 (k) protein in the dorsal horn after deletion of the Ntmt1/Prrx2-SE2 cluster. Data were obtained 5 d after coinjection of SE3-sgRNA and Lenti-CRISPR/Cas9 in CCI or sham mice; n = 4 mice/group, *p < 0.05 versus Sham+Scr; #p < 0.05 versus CCI+Scr; two-way ANOVA with repeated measures followed by post hoc Tukey's tests.
Downregulation of Ntmt1 and Prrx2 attenuates CCI-induced neuropathic pain
We next asked whether Ntmt1 and Prrx2 participate in neuropathic pain. First, we designed two siRNAs targeting Ntmt1 or Prrx2 mRNA. Ntmt1-siRNA-548 (transcription start site designated as +1) significantly decreased the expression of Ntmt1 mRNA in vitro by 45.2%, but Ntmt1-siRNA-347 had no effect on Ntmt1 expression in HT-22 cells (Fig. 6a). Similarly, Prrx2-siRNA-367 but not Prrx2-siRNA-564 reduced the level of Prrx2 mRNA expression by 45.2% in HT-22 cells (Fig. 6g). Therefore, we used Ntmt1-siRNA-548 and Prrx2-siRNA-367 in the in vivo experiment. Intrathecal injection of Ntmt1-siRNA-548 after CCI abolished the CCI-induced increase in NTMT1 in the ipsilateral L3/4 dorsal horn (Fig. 6b,c). Similarly, the increase in PRRX2 after nerve injury was abolished by intrathecal injection of Prrx2-siRNA-367 (Fig. 6h,i). Furthermore, downregulation of Ntmt1 or Prrx2 via injection of Ntmt1-siRNA (Fig. 6d–f) or Prrx2-siRNA (Fig. 6j–l) ameliorated the CCI-induced mechanical and thermal hypersensitivities from 1 d after injection, and these antinociceptive responses lasted at least 2 d. Injection of scrambled siRNA as a control did not have a similar analgesic effect during the observation period (Fig. 6e–f,j–l). Locomotor impairments were not observed in either sham or CCI mice after injection of Ntmt1-siRNA or Prrx2-siRNA or their controls (Table 1). Thus, our data suggest that spinal Ntmt1 and Prrx2 play an essential role in the initiation and maintenance of neuropathic pain.
Blocking Ntmt1 and Prrx2 inhibits the genesis of neuropathic pain. a, The Ntmt1 expression level on 48 h after Ntmt1 siRNA-347 and siRNA-548 (transcription start site designated as +1) transfection in HT-22 cell; **p < 0.01 versus Scr. b, c, Expression levels of dorsal horn Ntmt1 mRNA (b) and protein (c) 3 d after intrathecal injection of Ntmt1-siRNA-548 (siR) or scrambled siRNA (Scr) in CCI mice; n = 5 mice/group, *p < 0.05 versus Sham+Scr group; ##p < 0.01, ###p < 0.001 versus CCI+Scr group; one-way ANOVA with repeated measures followed by post hoc Tukey's tests. d–f, Effect of knockdown with Ntmt1-siRNA on nerve-injury-induced mechanical (d, e) and thermal (f) hypersensitivities; n = 8 mice/group, *p < 0.05, **p < 0.01, ***p < 0.001 versus Sham+Scr group; ##p < 0.01, ###p < 0.001 versus the CCI+Scr group at the corresponding time points; two-way ANOVA with repeated measures followed by post hoc Tukey's tests. g, The Prrx2 expression level on 48 h after Prrx2 siRNA-564 and siRNA-367 transfection in HT-22 cell; **p < 0.01 versus Scr. h, i, Expression levels of dorsal horn Prrx2 mRNA (h) and protein (i) 3 d after intrathecal injection of Prrx2-siRNA-564 (siR) or scrambled siRNA (Scr) in CCI mice; n = 5 mice/group, **p < 0.01, ***p < 0.001 versus Sham+Scr group; #p < 0.05, ##p < 0.01 versus CCI+Scr group; one-way ANOVA with repeated measures followed by post hoc Tukey's tests. j–l, Effect of Prrx2 knockdown with siRNA on CCI-induced mechanical (j, k) and thermal (l) hypersensitivities; n = 6–8 mice/group, ***p < 0.001 versus Sham+Scr group; ##p < 0.01, ###p < 0.001 versus CCI+Scr group at the corresponding time points; two-way ANOVA with repeated measures followed by post hoc Tukey's tests.
Mimicking upregulation of Ntmt1 and Prrx2 induces neuropathic-pain-like behaviors
Next, we examined whether upregulation of spinal Ntmt1 and Prrx2 was sufficient for the induction of nociceptive hypersensitivity. We constructed full-length Ntmt1 and Prrx2 lentivirus expression vectors. Intrathecal injection of Lenti-Ntmt1 increased the expression of Ntmt1 mRNA and protein on day 5 after injection, compared with the control Lenti-Gfp (Fig. 7a,b). Lenti-Ntmt1-induced upregulation of NTMT1 also resulted in a dramatic enhancement in the ipsilateral paw withdrawal frequencies to the 0.07 g and 0.4 g stimuli and a marked reduction in the ipsilateral paw withdrawal latency to thermal stimuli from day 3 after injection to day 14 (the farthest observed) in male (Fig. 7c–e) and female (Fig. 7f–h) naive mice. Neither Lenti-Ntmt1 nor the Lenti-Gfp control had an effect on locomotor behaviors (Table 1). We also observed an increase in Prrx2 mRNA and protein in the dorsal horn day 5 after injection of Lenti-Prrx2 (Fig. 7i,j). As expected, Lenti-Prrx2-induced PRRX2 elevation also led to the genesis of neuropathic-pain-like behavior, evidenced by the increase in ipsilateral paw withdrawal frequencies to 0.07 g and 0.4 g stimuli and by the significant decrease in the ipsilateral paw withdrawal latency to heat stimuli, starting from day 3 after injection and maintained for least 11 d during our observation period in male (Fig. 7k–m) and female (Fig. 7n–p) naive mice. PRRX2 upregulation by Lenti-Prrx2 did not impair locomotor function (Table 1). These observations suggest that in the absence of nerve injury, NTMT1 and PRRX2 upregulation can produce nociceptive behavior.
Mimicking the upregulation of NTMT1 and PRRX2 induces neuropathic pain-like behaviors. a, b, Ntmt1 mRNA (a) and protein (b) levels on day 5 after intrathecal injection of Lenti-Ntmt1 (Ntmt1) or Lenti-Gfp (Gfp) in naive mice; n = 5 mice/group, *p < 0.05 versus the Gfp group; two-tailed unpaired Student's t test. c–h, Effect of Ntmt1 upregulation by Lenti-Ntmt1 on sensitivity to mechanical and thermal stimuli in male (c–e) and female (f–h) naive mice; n = 6-8 mice/group, *p < 0.05, **p < 0.01, and ***p < 0.001 versus the Gfp group; two-way ANOVA with repeated measures followed by post hoc Tukey's tests. i, j, Prrx2 mRNA (i) and protein (j) levels on day 5 after intrathecal injection of Lenti-Prrx2 (Prrx2) or Lenti-Gfp (Gfp) in naive mice; n = 5 mice/group, **p < 0.01 and ***p < 0.001 versus the Gfp group; two-tailed unpaired Student's t test. k–p, Effect of Prrx2 upregulation on sensitivity to mechanical and thermal stimuli in male (k–m) and female (n–p) naive mice; n = 6–8 mice/group, *p < 0.05, **p < 0.01, and ***p < 0.001 versus the Gfp group; two-way ANOVA with repeated measures followed by post hoc Tukey's tests.
NTMT1 and PRRX2 participate in the spinal hyperactivation response
Nerve injury can cause hyperactivity in neurons of the spinal dorsal horn, resulting in the genesis of nociceptive responses. Thus, we next explored whether Ntmt1 and Prrx2 are involved in this hyperactivity by measuring the expression of p-ERK1/2 (a marker for hyperactivation of spinal cells) and GFAP (a marker for astrocytic hyperactivation). Silencing Ntmt1 or Prrx2 by intrathecal injection of Ntmt1-siRNA (Fig. 8a) or Prrx2-siRNA (Fig. 8b), but not the scrambled controls, abolished the CCI-induced increases in p-ERK1/2 and GFAP in the ipsilateral L3/4 dorsal horn on day 2 after injection in CCI mice. Overexpression of NTMT1 or PRRX2 in naive mice by intrathecal injection of Lenti-Ntmt1 (Fig. 8c) or Lenti-Prrx2 (Fig. 8d) increased the levels of p-ERK1/2, GFAP, and IBA1 (a marker for microglial hyperactivation) in the L3/4 dorsal horn on day 5 after injection, compared with the Lenti-Gfp control. These data suggest that Ntmt1 and Prrx2 participate in the nerve-injury-induced hyperactivity in neurons and astrocytes of the spinal dorsal horn. Additionally, the upregulation of Ntmt1 and Prrx2 is sufficient to induce glial activation in the spinal cord. Together, our results suggest that SS-SE regulates nerve-injury-induced nociception via Ntmt1 and Prrx2 (Fig. 9).
Ntmt1 and Prrx2 contribute to neuropathic pain by enhancing cellular hyperactivation in the dorsal horn. a, b, Protein level of p-ERK1/2, a marker of spinal cells hyperactivation) and GFAP (a marker for astrocytic hyperactivation) in the ipsilateral dorsal horn on day 2 after intrathecal injection of Ntmt1-siRNA (siR; a), Prrx2-siRNA (siR; b), or scrambled siRNA (Scr) in CCI mice; *p <0.05, **p < 0.01, ***p < 0.001 versus the corresponding groups; two-way ANOVA, post hoc Tukey's test; n = 5 mice/group. c, d, Protein levels of p-ERK1/2, GFAP, and IBA1 (a marker for microglial hyperactivation) in the dorsal horn on day 5 after intrathecal injection of Lenti-Ntmt1 (Ntmt1; c), Lenti-Prrx2 (d), or Lenti-Gfp in naive mice, *p <0.05, **p < 0.01, ***p < 0.001 versus the corresponding groups; two-way ANOVA, post hoc Tukey's test, n = 4 mice/group.
Activation of the spinal-specific super enhancer contributes to the genesis of neuropathic pain by driving Ntmt1 and Prrx2 expression in dorsal horn neurons. Nerve injury induces the formation of a spinal-specific SE via the binding of H3K27ac and BRD4 to the chromatin region located upstream of the Ntmt1 and Prrx2 genes. This activated SE further recruits transcription factors and RNA polymerase II to promote the transcription of Ntmt1 and Prrx2, resulting in hyperactivation of dorsal horn neurons in the spinal cord. Disruption of this SE by the BRD4 inhibitor JQ1 alleviates nerve-induced pain hypersensitivity by downregulating the NTMT1 and PRRX2 levels in the dorsal horn.
Discussion
Here, we identified the SS-SE using the epigenetic modification H3K27ac and a ChIP assay. We demonstrated that SS-SE is activated by peripheral nerve injury and that the two closest genes, Ntmt1 and Prrx2, can be controlled by SS-SE. Pharmacological inhibition or CRISPR-Cas9-mediated genetic deletion of SS-SE significantly alleviated the nerve-injury-induced hypersensitivity. Mechanistically, SS-SE participates in neuropathic pain through regulation of Ntmt1 and Prrx2, which in turn are involved in hyperactivation of spinal neurons.
SEs mediate high-order epigenetic regulation of gene expression (Suzuki et al., 2017). SEs are a special type of cis-acting regulatory element that combines multiple enhancer-like elements occupied by unique highly active transcriptional regulators such as BRD4, p300 histone acetyl transferase, RNA pol II, and active histone markers (H3K27ac; Suzuki et al., 2017). Given that H3K27ac is often used as a marker of SEs, we measured the occupation level of H3K27ac in chromatin from spinal genome. We found that peripheral nerve injury increases SE activity, and this increase mainly occurs in the gene coding domain sequences. SE-driven genes with differential expression are generally associated with the neuronal excitability. Through ChIP-Seq with H3K27ac, we identified SS-SE, which is located in Chr 2, 30645088–30797673, and is characterized by three enhancer clusters. This feature is markedly distinct from the peak profiling occupied by H3K27ac in six non-nervous-system tissues and five nervous system tissues. Therefore, SE driving of Ntmt1 and Prrx2 appears specific to spinal tissues. We did not analyze global SEs in the dorsal horn; thus, several questions remain to be addressed, that is, How many spinal-specific SEs are there in total? Do other specific SEs have a similar regulatory function in neuropathic pain? How are the other observed nerve-injury-specific SEs formed, and are they involved in nociceptive response?
Mounting evidence indicates that SEs play important roles in regulating cell identity, fate, and stem cell pluripotency and in diseases such as tumorigenesis, diabetes (Sun et al., 2018), and neurodegenerative diseases (Achour et al., 2015). For instance, an SE in the first intron of LMO1 affects susceptibility to neuroblastoma by regulating LMO1 expression (Oldridge et al., 2015). A CD47-associated SE is linked to pro-inflammatory signaling in MCF-7 cells (a breast cancer cell line; Betancur et al., 2017), and disruption of the KLF6 SE can inhibit growth of liver cancer cells (Ri et al., 2020). Currently, pharmaceutical companies are actively exploring small-molecule drugs that can specifically target noncoding regulatory DNA segments like SEs. For example, Syros Pharmaceuticals has developed a candidate drug, SY-5609, that inhibits cyclin-dependent kinase 7, an important enhancer element that controls the expression of various transcription factors.
BRD4 is a double bromodomain protein in the BET family and is expressed ubiquitously at high levels (Dey et al., 2019). BRD4 not only binds to acetylated histones to occupy regions in genome but also acts as a prominent component of SEs. Some of these studies used small-molecule inhibitors, such as JQ1, to disrupt BRD4-bound SEs, resulting in oncogene downregulation (Pelish et al., 2015; Bhagwat et al., 2016), suppression of inflammation, and amelioration of LPS-induced sepsis, experimental autoimmune encephalopathy–based neuroinflammation, and diabetes (Bandukwala et al., 2012; Anand et al., 2013; Duan et al., 2017; Santos-Silva et al., 2017). BET inhibitors therefore offer new therapeutic possibilities for a range of diseases. Here, we demonstrated that BRD4 binds to SS-SE and showed that intrathecal injection of the BRD4 inhibitor JQ1 abolished the increase in SS-SE-associated Ntmt1 and Prrx2 expression and attenuated nerve-injury-induced male and female pain hypersensitivity. Contrarily, upregulation of SS-SE-related genes Ntmt1 or Prrx2 caused genesis of nociception-like behavior in male and female naive mice, suggesting there is no gender difference in the role of SS-SE in neuropathic pain.
Importantly, JQ1 administration did not cause any locomotor impairments. These data suggest that BRD4 inhibitors may be useful therapeutic drugs for neuropathic pain. Our findings are consistent with previous reports in which JQ1 effectively ameliorated neuropathic pain after spared nerve injury (Palomés-Borrajo et al., 2021) and spinal cord injury (Sánchez-Ventura et al., 2019); JQ1 also reduced inflammatory pain (Hua et al., 2022) and protected against vincristine-induced peripheral neuropathy (Zhang and Xu, 2020). Functionally, these effects of JQ1 were because of decreased neuronal excitability or decreased microglia/macrophage reactivity. Our work is the first to identify the super-enhancer mechanism underlying the effects of JQ1 on nociception. We believe that our results provide a novel explanation of the analgesic efficacy of BRD4 inhibitors and expands our understanding of the molecular mechanisms. Future studies will be needed to explore whether JQ1 can be used in the clinic.
To confirm the analgesic effects of manipulating SS-SE, we genetically knocked out the SS-SE cluster using the CRISPR-Cas9 system. CRISPR-Cas9 successfully deleted the SS-SE enhancer cluster, reversed the CCI-induced increases in Ntmt1 and Prrx2 expression, and alleviated CCI-induced neuropathic pain. Thus, CRISPR-Cas9 appears to be an effective strategy for deleting SE-associated key genes. A previous study also reports that CRISPR-Cas9-mediated EphA2-SE deletion significantly reduced EphA2 expression, suppressing cell proliferation and invasion (Cui et al., 2021). We used CRISPR-Cas9 to delete each enhancer cluster in SS-SE, whereas Cui et al. (2021) deleted the entire EphA2-SE; however, both strategies knocked down SE genes. From our strategy, we can conclude that knockout of a single SE cluster is sufficient for functional manipulation of the entire SE; each SE cluster plays an important role in the activity of the SE as a whole, and the absence of an SE constituent likely impairs the regulatory function of the SE.
Ntmt1 (also known as Mettl11A or Nrmt1) encodes an S-adenosyl-l-methionine–dependent methyltransferase that is highly conserved and expressed in all tissues (Petkowski et al., 2013). Although the first crystal structure of NTMT1 was determined in 2005, it was not functionally characterized until 2010 (Tooley et al., 2010). N-terminal (Nα) methylation by NTMT1 is an important posttranslational modification that regulates protein–DNA interactions. NTMT1 can methylate an N-terminal X-Y-Lys/Arg consensus sequence (X = Leu, Ile, Trp, Asp, or Glu; Y, uncharged polar or nonpolar amino acids; Petkowski et al., 2012) and almost 300 proteins, including those involved in the structure of chromatin [CENPA (centromere protein A), CENPB (centromere protein B), HP1γ, SET] and DNA repair (Rb, DDB2, PARP3, BAP1), are predicted target proteins of NTMT1 (Conner et al., 2022). Ntmt1 is closely associated with development and several cancers. Knockdown of Ntmt1 results in hypersensitivity of breast cancer cell lines to double-stranded DNA breaks and increased proliferation of the estrogen-receptor-positive breast cancer cells MCF-7 and LCC916 (Bonsignore et al., 2015b). Mice with Ntmt1 deficiency exhibit early postnatal derepression of ZHX2 targets in fetal liver development, leading to liver degeneration (Conner et al., 2022). Furthermore, Ntmt1-knock-out mice have phenotypically defective DNA repair and exhibit premature aging (Bonsignore et al., 2015a). Here, we provide, to our knowledge, the first evidence that Ntmt1 can be driven by SS-SE and participates in nerve-injury-induced neuropathic pain.
Paired related homeobox transcription factor 2 (Prrx2) is another SS-SE-driven key gene. Similar to Ntmt1, Prrx2 was expressed mainly in spinal neurons; their distribution characteristics are consistent with the previous findings using single-cell sequencing (Rosenberg et al., 2018; Milich et al., 2021; Li et al., 2022; Matson et al., 2022). CCI-induced increase in PRRX2 was abolished by both the BRD4 inhibitor JQ1 and CRISPR-Cas9-mediated knockout of SS-SE. Knockdown of Prrx2 markedly reduced nociceptive hypersensitivity in nerve-injured mice, and upregulation of Prrx2 led to neuropathic-pain-like behavior. We speculate that Ntmt1 and Prrx2 are the critical genes driven by SS-SE and that the role of SS-SE in neuropathic pain is mediated by its effects on Ntmt1 and Prrx2. Prrx2 is a member of a subfamily of homeobox genes (Norris et al., 2000). As a transcription factor, the predominant mechanism by which Prrx2 affects disease is through binding to the promoters of target genes to alter their expression (Higuchi et al., 2013, 2014). Prrx2 is important for the development of mesenchymal tissues and the organogenesis of many tissues, as well as tumor progression and myocardial fibrosis (Bai et al., 2020; Liu et al., 2023). Interestingly, a previous report described several master regulators (Twist1, Twist2, Prrx1, and Prrx2) in the fibroblast gene regulatory network that are significantly downregulated in lung fibroblasts but not in skin fibroblasts. Knocking down Prrx2, Hey2, and p53 and upregulating Ascl1 enabled the efficient conversion of human skin fibroblasts to MAP2+ neurons. Thus, Prrx2 is likely a useful tool in cellular differentiation (Li et al., 2019). Here, we found that Prrx2 is a key regulator in the development of neuropathic pain, however, the detailed mechanisms by which Ntmt1 or Prrx2 regulate pain remain to be studied in the future.
In conclusion, this work shows that SS-SE is essential for the initiation and maintenance of neuropathic pain. Our results also highlight the pathways downstream of SS-SE. In particular, both pharmacological inhibition of SS-SE and CRISPR-Cas9-mediated SS-SE knockout significantly alleviated neuropathic pain behaviors via regulation of Ntmt1 and Prrx2 in dorsal horn neurons. Thus, SS-SE serves as a permissive epigenetic signal for the regulation of nerve-injury-induced genes expression and may be a promising target for the therapeutic management of neuropathic pain.
Footnotes
This work was supported by National Natural Science Foundation of China (NNSFC) Grants 82371243, 82171234, and 81971041 (Z.P.); NNSFC Grant 82201391 (Q.W.); NNSFC Grant 32200818 (L.Y.); NNSFC Grant 82171233 (H.-J.W.); Natural Science Foundation of Jiangsu Province Grant BK20201460; Natural Science Foundation of Jiangsu Education Department Key Project 22KJA320008; Jiang Su-Specially Appointed Professor Project; Jiangsu Provincial Association of Science and Technology Youth Science and Technology Talent Recruitment Project; and Jiangsu Provincial Doctors of Entrepreneurship and Innovation Program.
The authors declare no competing financial interests.
- Correspondence should be addressed to Zhiqiang Pan at zhiqiangp2002{at}aliyun.com
