Wild-type C57BL/6J male mice (8 weeks old) were provided by the Animal Experiments Center of Naval Medical University (NMU) (Shanghai, China). All mice were housed in standard cages in a condition of 23°C and 55% humidity under a 12 h-12 h light-dark cycle. All animal experiments were approved by the Scientific Investigation Committee of NMU, and the guidelines of the Ethics Committee of the International Pain Research Association were followed.

Inflammatory pain was induced by CFA (F5881, Sigma-Aldrich), as previously described (18). Briefly, mice were anesthetized with 1% Pentobarbital sodium (50 mg/kg, injected intraperitoneally); 20 µL of CFA or saline was intraplantarly injected into the left hind paw from heel to sole. The mice were divided into the following groups: sham (saline), control (CFA), and JQ1(CFA+JQ1) (MCE, Shanghai, China). Mice in the CFA+JQ1 group mice were intrathecally injected with 100 mM (5µL) of JQ1 1 h before CFA injection, as previously described (8). The sham and control groups were intrathecally injected with 5µL of saline.

HT22, a mouse hippocampal neuron cell line, was purchased from American Type Culture Collection. HT22 cells were cultured in DMEM (L110KJ, Basal Media, China) containing 10% fetal bovine serum (FBS, 10099141, Gibco), 100 U/mL of penicillin, and 100 µg/mL of streptomycin (15070063, Gibco) and grown in a condition of 5% CO2 and 95% humidity at 37°C. HT22 cells were treated with different doses of JQ1 for 24 h, and a drug toxicity test using a Cell Counting Kit 8 (CCK8) was used to find the optimal dose. Then, after pre-treatment with 100 nM or 500 nM of JQ1 for 1 h, HT22 cells were sequentially stimulated with LPS (1 µg/mL) for 24 h and ATP (5 mM) for 30 min to mimic cell pyroptosis.

To detect the effect of JQ1 treatment on the viability of HT22 cells, a CCK-8 kit (MCE, Shanghai, China) was used. Briefly, 1×10³ HT22 cells were seeded in each well of a 96-well plate and treated with different concentration of JQ1 (10, 50, 100, 500, and 1000 nM) for 24 h. 100 μL of complete medium mixed with 10 μL of CCK-8 reagent was added to each well for 1 h. Finally, the absorbance was detected using the infinite M200 PRO (TECAN) at 450 nm.

The L4-L5 spinal cord tissue of modeled mice or treated HT22 cells was homogenized in RIPA Lysis Buffer (Epizyme, Shanghai, China) containing a protease and phosphate inhibitor (Epizyme). The homogenates were centrifuged at 12,000 × g for 10 min. Then, the supernatant liquid was collected. BCA protein assay kit (Epizyme) was used for testing protein concentrations. Equal amounts of proteins were separated on a 10% SDS PAGE and were wet-electro-transferred onto a 0.2 μm polyvinylidene fluoride (PVDF) membranes (Millipore, USA). The membrane was blocked with Protein Free Rapid Blocking Buffer (Epizyme) for 1 h at room temperature on a rocker and then was incubated with specific primary antibody overnight at 4°C. The primary antibodies used were anti-GAPDH (1:1000 dilution, 5174), anti-NLRP3 (1:1000 dilution, D4D8T), NF-κB-p65 (1:1000 dilution, 93H1), anti-GSDMD-F (1:1000 dilution, E9S1X), anti-GSDMD-N (1:1000 dilution, E9S1X), and anti-mature IL-1β (1:1000 dilution, 63124), all purchased from cell signaling technology; anti-BRD4 (1:1000 dilution, servicebio, GB111415); anti-caspase-1-p20 (1:1000 dilution, Santa cruz, sc-398715). The following day, the membrane was washed with TBST (Servicebio, Wuhan, China) three times for 10 min each time, and then incubated with anti-rabbit secondary antibody (1:10,000 dilution, proteintech, SA00001-2) or anti-mouse secondary antibody (1:10,000 dilution, proteintech, SA00001-1) at room temperature for 2 h. Proteins were measured by enhanced chemiluminescence (ECL) reagent (Simuwu, China), imaged with a gel imaging system (Tanon, China), and quantified using ImageJ software (NIH, Bethesda, MD, USA).

Left paw tissues were fixed in buffered 4% paraformaldehyde for 24 h. Then, paw tissues were washed with running water for 3-5 h. Tissues were then routinely dehydrated by alcohol, transparentized, waxed, and embedded with paraffin. Paraffin blocks were cut into thin sections with a thickness of 5 µm. The sections were deparaffinized, hydrated, and stained with hematoxylin and eosin (H&E) solution according to a previously described protocol (7). Images were acquired using light microscopy (Olympus, Tokyo, Japan).

The total RNA of the tissues or cells was extracted using TRIzol (R401-01, Vezyme). The RNA was reverse-transcribed to cDNA with a reverse transcription kit (11141EZ60, YESEN) according to the manufacturer’s instructions. An SYBR Green kit (11202ES03, YESEN) was used for polymerase chain reaction (PCR) quantification on the QuantStudio 5 (Applied Biosystems). The cycle threshold (Ct) values of target proteins were collected and normalized to that of β-actin, and the fold change of gene expression was calculated with the 2^-ΔΔCT method. The primers sequence used are as follows: β-actin Forward: 5′-GGCTGTATTCCCCTCCATCG-3′, β-actin Reverse: 5′-CCAGTTGGTAACAATGCCATGT-3′; BRD4 Forward: 5′-GTGAGAAGCTAGGCCGTGTAG-3′, BRD4 Reverse: 5′-AGGCAGGACCTGTTTCAGAGT-3′.

After deep anesthesia with an overdose of pentobarbital sodium, mice were intracardially perfused with PBS until the effluent was clear and bloodless; then, they were injected with 4% ice-cold paraformaldehyde (PFA) until their limbs became stiff. The L4-L5 spinal cord tissues were collected and fixed in PFA for 5 h, and then dehydrated using a 25% sucrose solution until the tissues sunk to the bottom of the tube. The tissues were embedded in O.T.C tissue freezing medium (SAKURA, USA) and frozen to −20°C in a cryostat (Leica, Germany). Each sample was sectioned into 20-μm-thick slices, as previously described (19). A single or double labeling method was used for immunofluorescence. Single-labeling involved incubation of slices from target sections with the BRD4 antibody (Abcam, ab128874) at room temperature overnight, and then with the Cy3-conjugated anti-rabbit IgG (Servicebio, GB21403) for 1 h. Slices were washed with TBST three times (10 min each time) after every incubation. Observation and photographic documentation were conducted using a fluorescence microscope (EclipseE600, Nikon, Japan). To explore the cell types that expressed BRD4 and pyroptosis-related molecules, the sections were incubated with primary antibodies against Iba1 (Servicebio, GB11105), GFAP (Servicebio, GB11096) or NeuN (Servicebio, GB11138) together with caspase-1-p20 (Santa Cruz, sc-398715) or GSDMD (Santa Cruz, sc-393656), respectively. Then, the sections were incubated with secondary antibodies at room temperature for 1 h, washed again. Slices were imaged with a fluorescence microscope.

BRD4 small-interfering RNA (siRNA, Ribo, Shanghai) and a negative control (si-NC) were used to transfect HT22 cells at a confluency of 70-90% using Lipofectamine 3000 (Life Technologies) according to the manufacturer’s instruction. After 48 h, HT22 cells were sequentially stimulated with LPS (1 μg/mL) for 24 h and ATP (5 mM) for 30 min. The gene sequences of the negative control and the BRD4 siRNA used were GCTCAAGACACTATGGAAA and GGTACCAAACACAACTCAA, respectively.

Data were expressed as the mean ± standard error of the mean (SEM), and results were analyzed using GraphPad Prism 7 software. A two-tailed Student’s t-test was used for statistical analysis of two groups, and one-way analysis of variance was used for multiple comparisons with Bonferroni post hoc analysis. Differences were considered statistically significant if the P value was <0.05.

To analyze the role of BRD4 in inflammatory pain, the CFA paw injection model was used to simulate chronic inflammatory pain caused by peripheral tissue injury. The expression of BRD4 in the spinal dorsal horn was detected before and after CFA injection. The expression level of BRD4 protein in the spinal cord was significantly increased and peaked on the third day after CFA injection ( Figures 1A, B ), whereas the expression level was not significantly changed in the sham group. Consistent with western blot results, the mean fluorescence intensity of BRD4 in immunofluorescence staining was significantly enhanced by CFA injection ( Figure 1C ). The mRNA level of BRD4 was measured by qualitative PCR and was remarkably up-regulated, reaching a peak on the seventh day after CFA injection ( Figure S1 ). The cellular localization of BRD4 in the spinal cord was also explored by immunofluorescence. BRD4 was mainly colocalized with NeuN-positive neurons in the spinal cord ( Figure 1D ), rather than with Iba1 (microglia maker) or GFAP (astroglia marker) positive cells. In summary, these results showed that the BRD4 in the spinal dorsal horn was mainly expressed in neurons and was elevated after inflammatory pain.

To evaluate the effects of JQ1 on CFA-induced inflammatory pain, the electronic von Frey test was performed before and after CFA, saline, or JQ1 injection. At 1 h before CFA injection, 100 μM (5 μL) of JQ1 was intrathecally injected, as previously described (8) ( Figure 2A ). CFA was injected subcutaneously into the plantar area of the left paw of mice. This inflammatory pain model could induce a persistent mechanical and thermal allodynia in the ipsilateral paw from 6 h to 14 days after CFA injection (20). Our experimental data were consistent with previous reports. Compared with the hind paw of sham mice, the paw withdrawal threshold in the control mice was decreased from day 1 to day 7 and reached the minimum value on day 7 ( Figures 2B, C ). The mechanical and thermal allodynia induced by CFA was predominantly reversed through JQ1 pre-treatment, suggesting that BRD4 played a vital role in inflammatory pain. CFA induced a severe inflammatory reaction and tissue swelling of the injected paw on the third day after injection (21). Similarly, the left paw swelling had an obvious difference with that of the sham group mice, manifesting as increased accumulation of infiltrated cells in H&E staining. The paw swelling was gradually resolved in the CFA+JQ1 group with that of the control group, manifesting as less inflammatory cell infiltration on day 3 ( Figure 2D ). In summary, JQ1 pre-treatment attenuated CFA-induced inflammatory pain and paw inflammation.

The NLRP3 inflammasome promotes the beginning of cell pyroptosis. One study has shown that CFA-induced inflammatory pain increased the level of the NLRP3 inflammasome in the spinal dorsal horn (15). No study has measured the pyroptosis-associated proteins during inflammatory pain. The levels of NLRP3 inflammasome-related proteins (NLRP3, caspase-1-p20) were increased from day 1 to day 7 and reached a peak on day 3 ( Figures 3A, B ). GSDMD-F, GSDMD-N, and mature IL-1β which performs the function of pyroptosis-associated cells were detected by western blot analysis on days 1, 3, and 7 after CFA injection. Notably, compared with the sham group, the control group exhibited expressions of GSDMD-F, GSDMD-N, and mature IL-1β that were remarkably upregulated on days 1, 3, and 7 and that peaked on day 3 after CFA injection ( Figures 3A, C ). The NF-κB signaling pathway usually activates transcription of the NLRP3 inflammasome. Therefore, p65 phosphorylation was analyzed by western blot, which showed that it was up-regulated during inflammatory pain ( Figures 3A, D ). To uncover the main pyroptotic cell type, double labeling immunofluorescence staining was used. Results showed that caspase-1-p20 and GSDMD were mainly expressed in neurons ( Figure 3E ). Our data suggest that NLRP3 inflammasome-mediated pyroptosis played a vital role in CFA-induced inflammatory pain by activating NF-κB.

We showed that CFA-induced inflammatory pain was inhibited by the BRD4 inhibitor, JQ1 ( Figure 2B ). The mechanism by which BRD4 inhibition mediated the effect was investigated in more detail. Results showed that the up-regulation of pyroptosis-associated proteins, namely NLRP3, caspase-1-p20, GSDMD-F, GSDMD-N, and mature IL-1β, induced by CFA injection were reduced by JQ1 pre-treatment ( Figures 4A – C ). Meanwhile, activation of the NF-κB signaling pathway was significantly reduced by JQ1 pre-treatment through inhibition of p65 phosphorylation ( Figures 4A, D ). Immunofluorescence staining also was used to explore the effects of JQ1 pre-treatment on neuronal pyroptosis. JQ1 pre-treatment alleviated the Immunofluorescence intensity of NLRP3, caspase-1-p20, and GSDMD ( Figure 4E ). These data indicated that JQ1 pre-treatment had a protective effect by inhibiting neuronal pyroptosis.

A newly discovered type of programmed cell death, cell pyroptosis can be induced by LPS and ATP stimulation (22). In this study, HT22 cells were sequentially stimulated with LPS (1 µg/mL) for 12 h or 24 h and 5 mM of ATP for 30 min. The relative levels of pyroptosis-related proteins, namely NLRP3, GSDMD-F, GSDMD-N and caspase-1-p20 were gradually elevated over-time, and reached peaks at 24 h after stimulation ( Figure 5A ). CCK-8 was used to study the effects of JQ1 on the viability of HT22 cells. Results showed that the viability of HT22 cells was inhibited by 1000 nM of JQ1 ( Figure 5B ); therefore 100 nM and 500 nM of JQ1 were chose for subsequent experiments. The up regulation of NLRP3, GSDMD-F, and GSDMD-N proteins induced by LPS and ATP exposure was reversed by pre-treatment of JQ1 ( Figures 5C – E ). In summary, data suggested that JQ1, an inhibitor of BRD4, could work against the pyroptosis of HT22 cells in vitro.

To verify the function of BRD4 in inflammatory pain, we tested the role of pyroptosis in HT22 cells with BRD4 knockdown. First, the effect of BRD4 knockdown was tested by western blot, results showed that the expression of BRD4 in the si-BRD4 group was significantly decreased compared with that of the si-NC group ( Figures 6A, B ). Further, relevant alterations of the phosphorylation p65 induced by LPS stimulation were abolished by BRD4 knockdown, thus inhibiting activation of the NF-κB signaling pathway ( Figures 6A, E ). Notably, NLRP3 inflammasome activation and the cleavage of caspase-1 and GSDMD, which were induced by LPS and ATP stimulation, were markedly alleviated in HT22 cells with BRD4 knockdown ( Figures 6A, C, D ). These results demonstrated that the protective effects of BRD4 knockdown could be attributed to the inhibition of pyroptosis in HT22 cells.