Increased TSPO alleviates neuropathic pain by preventing pyroptosis via the AMPK-PGC-1α pathway

Abstract

Neuropathic pain poses a significant clinical challenge, largely due to the incomplete understanding of its molecular mechanisms, particularly the role of mitochondrial dysfunction. Bioinformatics analysis revealed that pyroptosis and inflammatory responses induced by spared nerve injury (SNI) in the spinal dorsal horn play a critical role in the initiation and persistence of neuropathic pain. Among the factors involved, TSPO (translocator protein) emerged as a key regulator. Our experimental findings showed that TSPO expression was upregulated during neuropathic pain, accompanied by mitochondrial dysfunction, specifically manifested as impaired mitochondrial biogenesis, disrupted mitochondrial dynamics (including insufficient expression of mitochondrial biogenesis and fusion-related proteins, as well as significantly increased expression of fission-related proteins), and activation of pyroptosis. Pharmacological upregulation of TSPO, but not its downregulation, effectively alleviated SNI-induced pain hypersensitivity, improving mitochondrial function and reducing pyroptosis. Immunofluorescence staining confirmed that TSPO was primarily localized in astrocytes, and its expression mirrored the protective effects on mitochondrial health and pyroptosis prevention. PCR array analysis suggested a strong association between TSPO and the mitochondrial regulation pathway AMPK-PGC-1α. Notably, inhibition of AMPK-PGC-1α abolished TSPO effects on mitochondrial balance and pyroptosis suppression. Furthermore, Mendelian randomization analysis of GWAS data indicated that increased TSPO expression was linked to pain relief. Through drug screening, molecular docking, and behavioral assays, we identified zopiclone as a promising TSPO-targeting drug for pain treatment. In summary, this study enhances our understanding of the molecular interplay between TSPO, mitochondrial health, and neuropathic pain, highlighting TSPO as a potential therapeutic target for pain management.

Graphical Abstract

Background

Neuropathic pain is a chronic condition caused by nerve injury or disease, characterized by hyperexcitability in somatosensory nociceptive circuits that severely affects quality of life [1, 2]. Despite extensive research, effective treatments for this debilitating condition remain limited in clinical practice [3]. A key feature of neuropathic pain is central sensitization, driven by maladaptive synaptic plasticity in the spinal dorsal horn, which, together with widespread neuroinflammation, amplifies pain hypersensitivity and facilitates pain spread [4, 5]. Throughout neuropathic pain progression, activated glial cells release cytokines and other inflammatory mediators, intensifying neuroinflammation and reinforcing central sensitization [4]. This interaction between central sensitization and neuroinflammation not only exacerbates pain but also makes it more challenging to control. Therefore, identifying therapeutic targets that can reduce neuroinflammation and modulate glial cell function may be critical to effectively managing neuropathic pain.

Emerging evidence suggests that pyroptosis, an inflammatory form of programmed cell death, plays a significant role in the development of neuropathic pain [6,7,8,9]. Mitochondria serve as central regulators of pyroptosis: under pathological conditions, damaged mitochondria release reactive oxygen species (ROS), amplifying pyroptosis and promoting inflammation. Furthermore, activated gasdermin D (GSDMD) forms pores in the mitochondrial membrane, leading to the release of pro-inflammatory mitochondrial contents, including cytochrome c and mitochondrial DNA (mtDNA), thereby disrupting mitochondrial integrity and creating a cycle of damage and inflammation [10]. Maintaining mitochondrial health is essential, as these organelles rely on biogenesis, fusion, and fission to meet cellular energy needs and maintain their structure. Disruption of these dynamics impairs mitochondrial function, increases ROS production, and leads to mtDNA damage, further intensifying pyroptosis. Pathways such as AMPK-PGC-1α regulate mitochondrial biogenesis and maintain mitochondrial dynamics, and dysregulation of these pathways exacerbates mitochondrial damage [10].

The 18 kDa translocator protein (TSPO), formerly known as the peripheral benzodiazepine receptor (PBR), is a high-affinity binding site for benzodiazepine drugs located predominantly on the outer mitochondrial membrane. Unlike central benzodiazepine receptors, TSPO exhibits distinct pharmacological and structural characteristics [11]. Discovered in rat kidneys in 1977, TSPO was subsequently identified in the brains of mammals, where it is expressed at low levels, predominantly in microglial cells, astrocytes, endothelial cells, vascular smooth muscle cells, and peripheral macrophages that infiltrate the central nervous system [12,13,14]. TSPO plays a pivotal role in inflammatory responses, particularly neuroinflammation, and serves as an important biomarker for brain inflammation imaging techniques such as positron emission tomography (PET) and single-photon emission computed tomography (SPECT) [15,16,17]. It is also implicated in various inflammatory conditions, including retinal neuroinflammation [18], post-hemorrhagic brain inflammation [19], heart dysfunction [20], and liver inflammation [21, 22]. In addition, TSPO is critical in mitochondrial functions, such as the regulation of oxidative stress, iron homeostasis, and cholesterol transport [22]. It is involved in processes like apoptosis, cell death, inflammation, and immune responses [13, 23, 24]. TSPO deficiency can result in mitochondrial dysfunction, contributing to hypoxia, angiogenesis, and a metabolic shift toward glycolysis in glioblastoma [25]. Furthermore, TSPO activation has been shown to alleviate pain in various neuropathic pain models [26, 27].

Despite these insights, the precise mechanisms by which TSPO regulates inflammation remain incompletely understood, and further research is needed to fully elucidate its role in these processes. In this study, through bioinformatics analysis of spinal dorsal horn tissue following SNI, we identified pyroptosis, especially TSPO closely associated with SNI-induced neuropathic pain. Experimentally, we found that TSPO was upregulated after SNI, along with mitochondrial dysfunction as well as pyroptosis. Pharmacological increase but not decrease of TSPO alleviated neuropathic pain, along with improved mitochondrial health and pyroptosis. By immunofluorescence staining, TSPO was mainly located in astrocytes. Upregulation of TSPO specific in astrocytes mirrored the protective effects. PCR array analysis showed that TSPO supported mitochondrial biogenesis and regulated mitochondrial fusion and fission via the AMPK-PGC-1α pathway. Intriguingly, Mendelian randomization analysis indicated that increase of TSPO was linked to pain relief. Therefore, to facilitate clinical translation, we identified TSPO-targeted drugs, zopiclone, through drug databases, highlighting the potential of repurposing existing drugs for pain treatment.

Materials

Data processing and download of the SNI dataset

To examine the causal mechanisms linked to SNI, we obtained the microarray dataset GSE18803 from the Gene Expression Omnibus (GEO) [28]. We carried out differential expression analysis to identify differentially expressed genes (DEGs) between SNI samples (n = 6) and control samples (n = 6) using the “limma” R package [29], applying significance thresholds of P < 0.05 and |log2(fold change, FC)|> 1. We generated a volcano plot and heatmap using R 4.0.3 software. Enrichment analyses for Gene Ontology (GO) and KEGG pathways were performed utilizing the ‘clusterProfiler’ R package [30]. Additionally, gene set enrichment analysis (GSEA) was executed with GSEA 4.1.0 software, referencing gene sets from the MSigDB database. We identified a total of 1,576 mitochondria-related genes (Mito-RGs) from the MSigDB [31]. Using the “WGCNA” R package, we constructed a co-expression network based on gene expression data from GSE18803, designating SNI and control samples as trait data [32]. Sample clustering was conducted with the hclust function to detect outliers, utilizing the parameter “method = average.” We approximated the optimal soft threshold for establishing a scale-free network and employed a dynamic shear tree algorithm to segment modules. Correlation analysis helped identify SNI-associated modules, and SNI genes (SNIGs) were derived through Module Membership (MM) and Gene Significance (GS) analyses.

Drugs and reagents

4',6-Diamidino-2-phenylindole (DAPI) was sourced from Beyotime Biotechnology (China). Ro 5–4864 (4'-Chlorodiazepam; MW: 319.19; purity: ≥ 99.19%) [33], PK 11195 (RP 52028; MW: 352.86; purity: ≥ 99.05%) [34], Compound C (MW: 399.49; purity: ≥ 98.14%) [35] and Zopiclone (N-Desmethyl zopiclone-d₈,MW:374.78; purity: ≥ 99.79%) [36] were obtained from MedChem Express (USA). Lipopolysaccharide (LPS, L4391) was obtained from Sigma. DMEM was obtained from GIBCO/BRL, Life Technologies (Carlsbad, CA, USA). Donkey anti-Rabbit IgG H&L (Alexa Fluor® 488), anti-Mouse IgG H&L (Alexa Fluor® 546), and anti-Goat IgG H&L (Alexa Fluor® 546) were obtained from Invitrogen (United Kingdom; Cat # A-31573, A-31571, and A-21447). Primary antibodies against AMPK, NLRP3, p-AMPK, TFAM, c-Fos, and Cleaved Gasdermin D were sourced from Cell Signaling Technology (USA; Cat# 2532, 15,101, 50,081, 7495, 4384, and 32,112). An anti-TSPO antibody was obtained from Novus Biologicals (USA; Cat# NB100-41398). Antibodies against NeuN, GFAP, and Iba1 were obtained from Abcam (United Kingdom; Cat# ab104224, ab7260, and ab178846). Antibodies against GAPDH and Caspase-1 were obtained from Proteintech Group (USA; Cat# 10,494–1 and 22,915–1). Antibodies against DRP1, PGC-1α, and MFN2 were obtained from Affinity Biosciences (USA; Cat# DF7037, AF5395, and DF8106).

Animals and ethics statement

Adult male C57BL/6 J mice, with an average weight of 20 ± 2.5 g, were obtained from Liaoning Changsheng Biotechnology in Liaoning, China. The mice were maintained on a 12-h light–dark cycle (from 8 a.m. to 8 p.m.) and had continuous access to food and water. All animal studies were conducted in compliance with the US National Institutes of Health (NIH) Guide for the Care and Use of Laboratory Animals and received approval from the Jilin University Administration Committee of Experimental Animals (Approval No. 20230609–01). Testing was performed in a double-blinded manner.

Spared nerve injury (SNI) model

The SNI surgery was conducted according to the method previously outlined [37]. In summary, the animals were induced with isoflurane anesthesia at 2.5% and maintained at 2%. A surgical incision was made in the left lateral thigh to access the sciatic nerve, which was carefully separated from the surrounding muscle to expose its main trunk and three branches. SNI was induced by ligating and cutting segments of the common peroneal and tibial nerves measuring 2–4 mm. In the sham-operated mice, the sciatic nerve and its branches were exposed, but no further intervention was performed.

Von Frey filament assay

The sensitivity of the mechanical paw withdrawal threshold (PWT) in mice was examined using calibrated von Frey filaments (Stoelting, Wood Dale, IL). These filaments were positioned perpendicular to the plantar surface of the hind paw and pressed with adequate force to create a bend. A withdrawal of the paw occurring within 4 s was categorized as a positive response. If the initial filament did not elicit a response, the filament with the next higher force was tested; if a response was observed, the filament with the next lower force was applied instead. Filaments ranging from 0.04 to 1.4 g were utilized for this assessment. The force needed to achieve a 50% probability of a withdrawal response was calculated employing the "up-down" technique. PWT measurements were recorded before the SNI procedure (day -1) and on postoperative days 3, 7, 14, 21, and 28.

Astrocyte-specific TSPO constructs (AS-TSPO)

Recombinant adeno-associated viruses (rAAVs) designed to express enhanced green fluorescent protein (EGFP) and TSPO were utilized for both broad-spectrum and astrocyte-specific overexpression of TSPO. These viruses were produced by Heyuan Biotechnology Co., Ltd (OBIO, Shanghai, China), with the following stock titers: pCAAV-CMV-Tspo-linker-EGFP-3 × FLAG-WPRE at 4.4 × 10¹² and pAAV-GfaABC1D-Tspo-linker-EGFP-3 × FLAG-WPRE also at 4.4 × 10¹².

Virus vector delivery

The AAV vector was administered into the spinal dura mater of anesthetized mice, adhering to a previously established protocol [38]. The target vertebral column was identified through palpation, and the paraspinal muscles on the left side of the T13-L1 intervertebral space were excised, followed by a partial laminectomy to facilitate injection. Using a micro syringe (RWD 79013, Shenzhen, China), 0.6 μL of AAV was injected at a rate of 50 nL/min. A stereotaxic apparatus (RWD 69100, Shenzhen, China) was employed to position the injection needle approximately 500 μm from the midline, with the tip advanced into the spinal dorsal horn at a depth of 250 μm from the dorsal root entry zone surface. The needle was retained in the spinal cord for 10 min prior to being removed. The incision was closed with interrupted sutures, and iodine disinfectant was applied to the area.

Tissue extraction

For the Western blot (WB) experiments, we evaluated the expression changes of TSPO and pyroptosis-related proteins at different time points, including sham, day 3, day 7, day 14, and day 21 post-SNI. Tissue samples for WB analysis were collected at these respective time points. For all other experiments, including immunofluorescence, tissue samples were consistently collected on day 7 post-SNI. All mice were anesthetized and perfused via the left ventricle with 0.9% saline to remove blood, as confirmed by observing the liver turning pale.

For WB experiments: The spinal cord was exposed by removing the lamina, and the right side of the spinal cord was marked for orientation. The entire spinal cord was carefully extracted, with the lumbar enlargement identified at the T13-L1 vertebrae. The spinal cord was halved along the midline, discarding the right half. The left half was dissected on ice to isolate the dorsal horn, comprising approximately one-quarter of the spinal cord tissue, while discarding the ventral horn. Rapid handling was maintained throughout to preserve tissue integrity.

For IF experiments: Following perfusion with 0.9% saline, 20 ml of 4% paraformaldehyde was used for further fixation. After observing whole-body rigidity, the spinal column was extracted, and the tissue was post-fixed in 4% paraformaldehyde overnight. The spinal cord was processed similarly by removing the lamina, marking the right side for orientation, and isolating the lumbar enlargement. The tissue was halved along the midline, retaining the left dorsal horn while discarding the right half and ventral horn. The dorsal horn tissue was dehydrated in 30% sucrose until it sank, then embedded in OCT compound. Frozen sections were prepared using a cryostat for subsequent staining.

Intrathecal injection

Drugs were administered via intrathecal injection. Under anesthesia, a 0.25 mm insulin needle was carefully inserted into the L5-L6 intervertebral space. Proper placement was confirmed by observing a tail flick response. A solution containing the drug dissolved in 10 µl of saline was injected, and the needle was held in place for 15 s to prevent leakage.

Western blotting

Astrocytes and left spinal dorsal horn tissue were lysed using RIPA buffer (C500008; Sangon Biotech, Shanghai, China), and protein concentrations were assessed with a BCA assay (P0010, Beyotime Biotechnology). The samples were subjected to separation via 10% SDS-PAGE and subsequently transferred to PVDF membranes (IPVH00010, Millipore, USA). The membranes were blocked in 5% BSA in TBS-T for 2 h at room temperature and then incubated overnight at 4 °C with primary antibodies, including p-AMPK, AMPK, NLRP3, Cleaved Gasdermin D, pro-Caspase1, Caspase-1, PGC-1α, DRP1, MFN2, TFAM, and GAPDH (all at a dilution of 1:1000). After incubation, the membranes were scanned using a GS800 densitometer (Fusion FX6-XT, Wheelabrator, USA) and analyzed with Multi Gauge software (Fuji, Tokyo, Japan). Band intensities were quantified with FIJI software (NIH ImageJ), normalized to GAPDH, and presented as fold changes relative to the control samples.

Immunofluorescence staining

Tissue samples from the left dorsal horn of the spinal cord were fixed in 4% paraformaldehyde (PFA) overnight. After fixation, samples were dehydrated, embedded, and sectioned at 30 µm for later staining. The sections underwent PBS washing and were blocked for 1 h at room temperature with a solution containing 5% BSA and 1% Triton X-100. They were then incubated overnight at 4 °C with the following primary antibody combinations: TSPO (1:200) with NeuN (1:400), TSPO (1:200) with GFAP (1:400), TSPO (1:200) with Iba1 (1:400), or c-Fos (1:200) with NeuN (1:400). After an additional washing step, sections were exposed to secondary antibodies at 37 °C for 1 h, counterstained with DAPI, washed again, and observed using a Laser Confocal Electron Microscope (Nikon, Japan).

Isolation and culture of astrocytes

Spinal cord tissue was obtained from 24-h-old mice, minced, homogenized, and passed through a 70 µm filter before being transferred to culture flasks. On the third day, cultures were shaken at 200 rpm for 20 min, and the medium was subsequently refreshed. This process was repeated on the sixth day. Astrocyte purity was evaluated using GFAP staining. Cultures were maintained in DMEM supplemented with 10% FBS and 1% (v/v) penicillin/streptomycin at 37 °C in a humidified incubator with 5% CO₂. Cells at passages 6 to 8 were utilized for all experiments. To induce pyroptosis, astrocytes were primed with lipopolysaccharide (LPS, 100 ng/mL) for 24 h, followed by treatment with ATP (5 mM) for 30 min [39]. Ro 5–4864 (100 µg/mL) was added 2 h post-pyroptosis induction. For investigating the AMPK pathway, astrocytes were pretreated with 5 µg/mL Compound C 1 h prior to Ro 5–4864 administration [40, 41].

Mitochondrial assessment in astrocytes

The JC-1 fluorescent probe (Beyotime Biotechnology, China) was used to assess mitochondrial membrane potential (MMP) changes. Astrocytes were incubated with 1 µM JC-1 in culture medium for 30 min at 37 °C. After incubation, cells were washed to eliminate excess dye and then visualized with a laser-scanning confocal microscope (Nikon, Tokyo, Japan). The red and green fluorescence intensities were quantified using ImageJ software (version 4.1; Bethesda, MD, USA), with data presented as the red-to-green fluorescence ratio. To evaluate mitochondrial ROS levels, MitoSox Red dye (Invitrogen, M36008) was applied, while Mito Tracker Red (Invitrogen, M24426) was used to assess mitochondrial function. After treatments, astrocytes were incubated with MitoSox (20 µM) or Mito Tracker (50 µM) for 30 min at 37 °C. Cells were then washed three times with PBS, followed by Hoechst staining for nuclei at 37 °C. Images were acquired using a fluorescence microscope.

Transmission electron microscope

Spinal tissue samples were perfused and preserved in 2.5% glutaraldehyde. After fixation, the sections were exposed to 1% osmium tetroxide for 1 h, then washed and progressively dehydrated through an ascending series of alcohol concentrations. The specimens were embedded in Epon 812 epoxy resin, sectioned into ultrathin slices of 60 nm thickness, stained with uranyl acetate and lead citrate, and examined under a transmission electron microscope (JEOL Ltd., Tokyo, Japan).

Cell viability and LDH detection

Cell viability was evaluated using the Cell Counting Kit-8 (CCK-8, Epizyme, China) according to the manufacturer’s guidelines, while LDH release was quantified with the LDH Cytotoxicity Assay Kit (Beyotime) as instructed in the protocol, with absorbance recorded at 490 nm.

Enzyme-linked immunosorbent assay (ELISA)

Protein samples were extracted, and concentrations were determined as previously outlined. Levels of mature IL-1β and IL-18 in the spinal dorsal horn were measured using commercial ELISA kits (R&D Systems, Minneapolis, MN, USA) following the manufacturer’s instructions.

Oxidative stress and Reactive Oxygen Species (ROS) assessment in astrocytes

The DCFH-DA probe (S0035S, Beyotime Biotechnology) was used to quantify cytosolic reactive oxygen species (ROS) in astrocytes. Following specified treatments, cells were stained based on established protocols, and images were captured with a fluorescence microscope. Additionally, lysates from the spinal dorsal horn and astrocytes were analyzed for total superoxide dismutase (SOD) activity, glutathione (GSH) levels, and malondialdehyde (MDA) concentration. The following kits were utilized: Total Superoxide Dismutase Assay Kit (WST-8, Cat#S0101, Beyotime Biotechnology), Total Glutathione Assay Kit (Cat#S0052, Beyotime Biotechnology), and Lipid Peroxidation MDA Assay Kit (Cat#S0131, Beyotime Biotechnology).

Quantitative real-time PCR for mitochondrial DNA content

Mitochondrial DNA (mtDNA) copy number was measured through real-time quantitative RT-PCR using the mtDNA Copy Number Kit (Cat# MCN3, Detroit R&D). DNA was isolated from spinal dorsal horn tissues and astrocytes with the Qiagen DNeasy Blood and Tissue Kit. The total DNA concentration was determined using a NanoDrop 1000 (Thermo Fisher Scientific). Mitochondrial DNA levels were quantified by normalizing the mitochondrial gene (cytochrome b) against the nuclear gene (GAPDH). Quantitative real-time PCR was employed to assess mtDNA content in both tissue homogenates and cultured cells.

qRT-PCR array

Mouse mitochondrial gene expression profiles were assessed using the Mitochondria PCR Array, following the protocol provided by the manufacturer (Wcgene Biotech, Shanghai, China). Data analysis was conducted using R software (version 4.3.2).

Mendelian randomization and SMR analysis of TSPO and pain phenotypes

We extracted genome-wide association study (GWAS) data for the TSPO gene from the eQTLGen database (https://eqtlgen.org/), applying a significance threshold of P < 5e-8 to identify robust instrumental variables. Next, we retrieved GWAS data on multiple pain phenotypes from the IEU database (https://gwas.mrcieu.ac.uk/), specifically for back pain (ukb-b-9838), facial pain (ukb-b-17107), headache (ukb-b-12181), hip pain (ukb-b-7289), knee pain (ukb-b-16254), and neck or shoulder pain (ukb-b-18596). We used the TwoSampleMR R package to conduct Mendelian Randomization (MR) analysis, assessing the causal relationships between TSPO and the selected pain phenotypes [42]. To further validate the results, Summary-based Mendelian Randomization (SMR) analysis (https://yanglab.westlake.edu.cn/software/SMR/) was conducted [43]. For significant MR results, we performed Bayesian colocalization analysis to further eliminate false positive results.

Drug screening and molecular docking

We searched for TSPO-related compounds in three drug databases: DGIdb (https://www.dgidb.org/), DrugBank (https://go.drugbank.com/), and TTD (https://db.idrblab.net/ttd/), selecting compounds present in all three for molecular docking studies. Using PubChem database (http://pubchem.ncbi.nlm.nih.gov/), we retrieved and converted 2D ligand structures into 3D with ChemOffice 20.0, saving them in mol2 format. The TSPO protein sequence (Uniprotkb: P30536) was sourced from UniProt database (https://www.uniprot.org/), and a crystal model was built using AlphaFold (https://alphafold.ebi.ac.uk), saved in PDB format. Molecular Operating Environment (MOE) 2019 was used for energy minimization, target protein preprocessing, and active site identification. Docking simulations were conducted in MOE 2019 with 50 runs per experiment, with binding affinities assessed by docking scores and visualized through PyMOL 2.6.0 and Discovery Studio 2019.

Statistical analysis

Statistical analyses for bioinformatics were conducted using the R language (version 4.3.2). Data analysis and graph construction were carried out with GraphPad Prism 8.3.0. Results are presented as mean values ± standard error of the mean (SEM). Two-tailed Student’s t-tests were utilized for comparisons between two groups, while one-way or two-way ANOVA with Tukey’s multiple comparisons was employed for analyses involving multiple groups. Significance levels are indicated as follows: *p < 0.05, **p < 0.01, ***p < 0.001.

Results

Characterization of SNI-induced changes in the spinal dorsal horn

To characterize the changes of spinal dorsal horn after SNI, we conducted a differential gene expression analysis and identified 91 DEGs (Fig. 1A), Among them, 90 genes upregulated and 1 downregulated (p-value < 0.05, |log2(fold change)|> 1), with a volcano plot illustrating their distribution (Fig. 1B). The DEGs were enriched in inflammation and immune-related biological processes by GO analyses (Supplementary Fig. 1A). KEGG revealed an enrichment of Fc gamma R-mediated phagocytosis, NOD-like receptor signaling, chemokine signaling, and cytosolic DNA-sensing pathways (Fig. 1C). The NOD-like receptor (NLR) signaling pathway, particularly the NLRP3 inflammasome, was notably activated after SNI. GSEA results further supported these findings, highlighting a strong association with NLRP3 inflammasome, cytokine signaling, and pyroptosis in SNI (Fig. 1D-F). These results suggest that pyroptosis and mitochondrial dysfunction play critical roles in neuropathic pain process after SNI.

Fig. 1

Characterization of SNI-induced changes in the spinal dorsal horn. A Heatmap showing differential gene expression in Sham and SNI. B Volcano plot showing the significantly upregulated and downregulated genes between SNI and Sham groups. C Pathways enriched by KEGG showed important biological pathways associated with differentially expressed genes. D-F GSEA results for enrichment of NLRP3 inflammasome pathway, Cytokine signaling in immune systems and Pyroptosis. G Module-trait relationship analysis by WGCNA showed correlations between gene modules and clinical traits

To identify SNI-related genes, we performed Weighted Gene Co-Expression Network Analysis (WGCNA). Sample clustering was well-defined (Supplementary Fig. 1B), and we selected a soft threshold of 7 basing on a scale-free topology model fit (R² > 0.9) and high mean connectivity (Supplementary Fig. 1C, D). After merging highly correlated modules using a clustering height cut-off of 0.25, 24 gene modules were identified for further analysis (Supplementary Fig. 1E). Module-trait relationships showed that the yellow module exhibited a strong negative correlation with the control group (r = -0.98, p = 4e-08) and a positive correlation with SNI (r = 0.98, p = 4e-08) (Fig. 1G; Supplementary Fig. 1F). Furthermore, it was demonstrated that these gene networks were independent of each other (Supplementary Fig. 1G). GO and KEGG analyses of yellow module genes revealed significant enrichment in inflammatory pathways, consistent with differential expression patterns (Supplementary Fig. 1I, J). These findings further highlight the essential role of inflammation in the spinal dorsal horn in response to SNI.

TSPO was correlated to pyroptosis and inflammation in neuropathic pain

Previous studies have shown that pyroptosis and inflammasome activation are strongly linked to mitochondrial dysfunction [44]. Consistently, our GSEA analysis confirmed the existence of mitochondrial impairment after SNI (Fig. 2A-C). Moreover, we performed a Venn analysis of DEGs, SNI-related genes (yellow module), and mitochondrial genes (Fig. 2D), which identified TSPO and UCP2 as the critical genes involved in neuropathic pain (Supplementary Fig. 2A, B). Existing evidence indicates that TSPO plays a role in pain modulation; however, the underlying precise mechanisms remain to be elucidated.

Fig. 2

TSPO was correlated to pyroptosis and inflammation in neuropathic pain. A-C GSEA results for enrichment of mitochondrial complex I assembly, enrichment of the electron transport chain and mitochondrial complex III assembly. D Venn diagram illustrating the intersection between SNI-related genes, mitochondrial genes, and differentially expressed genes (DEGs). E Schematic diagram of the experimental flow. F Von Frey results for the Sham and SNI groups at different time points (days 0, 3, 7, 14, and 21), n = 10 per group. G Western blot showing the expression levels of TSPO, NLRP3, Caspase-1, and GSDMD-N at various time points post-SNI. H–K Quantification of the relative expression of TSPO, NLRP3, Caspase-1, and GSDMD-N at various time points post-SNI, n = 10 per group. L, M ELISA results showing the levels of IL-1β and IL-18, n = 6 per group. Data are represented as mean ± SEM, ***p < 0.001

Therefore, we validated the bioinformatics findings through experimental approaches. Neuropathic pain was induced using the SNI model, and pain thresholds in mice were assessed on days 0 (baseline), 3, 7, 14, and 21 post-surgeries (Fig. 2E). To control potential confounding effects of the surgical procedure, a sham-operated and control group was included in the study (Supplementary Fig. 8). As expected, SNI mice showed increased sensitivity to mechanical stimuli compared to sham controls (Fig. 2F). Western blot analysis revealed that TSPO expression was elevated on day 7 post-SNI and maintained till day 21 or longer. At the same time, pyroptosis markers (NLRP3, GSDMD-N, Caspase-1) were also significantly elevated on day 7 and maintained high levels thereafter (Fig. 2G-K). This suggests that day 7 post-SNI is a critical time point for these processes. ELISA analysis further confirmed that inflammatory cytokines IL-1β and IL-18 were upregulated on day 7 (Fig. 2L, M). These results demonstrate a positive correlation between TSPO expression and pyroptosis.

Pharmacological increase but not decrease of TSPO alleviated neuropathic pain

To investigate the relationship between TSPO and neuropathic pain, we modulated TSPO expression pharmacologically (Fig. 3A). As TSPO was positively correlated with pyroptosis, we firstly intrathecally administered PK 11195, a TSPO inhibitor at 2 μg (in 10 μl saline) daily from the first day post-SNI [27]. Von Frey tests were conducted two hours later after each administration, and the results revealed that inhibition of TSPO amplified the hypersensitivity to mechanical stimuli in SNI mice (Fig. 3B). This result was also supported by immunofluorescence results, which showed a significant increase in the number of c-Fos + neurons after TSPO inhibition Fig. 3C; Supplementary Fig. 2C). PK 11195 administration resulted in elevated levels of inflammatory cytokines IL-1β and IL-18 (Fig. 3D, E), as well as increased pyroptosis-related proteins, including NLRP3, GSDMD-N, and Caspase-1 (Fig. 3F; Supplementary Fig. 2D-G). The above results indicate that the decreased expression of TSPO weakens its inhibitory effect on pyroptosis.

Fig. 3

Pharmacological increase but not decrease of TSPO alleviated neuropathic pain. A Schematic diagram of the experimental flow. B Behavioral changes following TSPO inhibition by PK 11195 were assessed using von Frey testing at multiple time points (days 0, 3, 7, 14, and 21), n = 10 per group. C Representative immunofluorescence images showing c-Fos and NeuN co-staining after PK 11195 treatment, Scale bar = 50 μm. D, E ELISA results showing the levels of IL-1β and IL-18 after PK 11195 administration, n = 6 per group. F Western blot showing the expression levels of TSPO, NLRP3, Caspase-1 and GSDMD-N after PK 11195 treatment. G Behavioral tests of mice receiving different concentrations of Ro 5–4864, n = 6 per group. H Behavioral changes following TSPO activation by Ro 5–4864 were assessed using von Frey testing at multiple time points (days 0, 3, 7, 14, and 21), n = 10 per group. I Western blot showing the expression levels of TSPO, NLRP3, Caspase-1 and GSDMD-N after Ro 5–4864 treatment. J Representative immunofluorescence images showing c-Fos and NeuN co-staining after Ro 5–4864 treatment, Scale bar = 50 μm. K ELISA results showing the levels of the IL-1β and IL-18 after Ro 5–4864 treatment, n = 6 per group. Data are represented as mean ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001

To further verify the role of TSPO, we upregulated its expression using Ro 5–4864. The results showed that Ro 5–4864 at 1 μg and 3 μg significantly reduced pain sensitivity, while 0.25 μg had no effect. Besides, no differences were detected between the 1 μg and 3 μg doses over the 21-day observation period (Fig. 3G). As a result, 1 μg was selected for further experiments. Von Frey tests show that Ro 5–4864 effectively alleviated mechanical allodynia (Fig. 3H). Administration of Ro 5–4864 resulted in the upregulation of TSPO (Fig. 3I; Supplementary Fig. 2H), which was shown to decrease the expression of pyroptosis-related proteins (NLRP3, Caspase-1, and GSDMD-N) by Western blot analysis (Fig. 3I; Supplementary Fig. 2I-K). Immunofluorescence staining indicated a reduction in c-Fos + neurons (Fig. 3J; Supplementary Fig. 2L), and ELISA results demonstrated lower levels of IL-1β and IL-18 (Fig. 3K, L). To exclude the potential interference of the drug itself on the results, we conducted a separate experiment to rule out its effect (Supplementary Fig. 7). Collectively, these findings suggest that upregulated TSPO alleviates neuropathic pain by modulating pyroptosis.

Genetic increase of TSPO in astrocytes alleviated neuropathic pain by reducing pyroptosis

To determine whether the function of TSPO was cell-type specific, immunofluorescence staining was performed on the spinal dorsal horn using anti-TSPO, NeuN, GFAP, and Iba1 antibodies. Interestingly, the results demonstrated that TSPO was predominantly co-expressed with GFAP + astrocytes, with minimal expression observed in neurons or microglia (Fig. 4A). Therefore, we constructed astrocyte-specific TSPO adeno-associated virus (AS-TSPO) to upregulate the expression of TSPO in astrocytes. AS-TSPO was injected into the left spinal dorsal horn in mice (Fig. 4B), and viral effectiveness was verified two weeks later by observing green fluorescence at the injection site and co-localization with astrocytes (Fig. 4C). Western blot confirmed the upregulation of TSPO by AS-TSPO (Fig. 4D; Supplementary Fig. 3B). After that, we investigated the role of TSPO overexpression in astrocytes during neuropathic pian by injection AS-TSPO two weeks before SNI. Von Frey tests indicated that AS-TSPO alleviated mechanical allodynia (Fig. 4E) and reduced the number of c-Fos + neurons (Fig. 4F; Supplementary Fig. 3C). Additionally, inflammatory cytokines, IL-1β and IL-18 (Fig. 4G, H), as well as pyroptosis-related proteins NLRP3, Caspase-1, and GSDMD-N were all suppressed (Fig. 4I; Supplementary Fig. 3D-F).

Fig. 4

Genetic increase of TSPO in astrocytes alleviated neuropathic pain by reducing pyroptosis. A Representative immunofluorescence images showing TSPO/NeuN, TSPO/GFAP, and TSPO/Iba1 co-staining, Scale bar = 50 μm. B Schematic diagram of the experimental flow. C Representative immunofluorescence images showing AS-TSPO colocalized with GFAP + astrocytes, scale bar = 50 μm. D Western blot shows the expression levels of TSPO in AS-TSPO or AS-TSPO-Ctrl treated conditions. E Behavioral changes following TSPO activation by AS-TSPO were assessed using von Frey tests at multiple time points (days 0, 3, 7, 14, and 21), n = 10 per group. F Representative immunofluorescence images showing c-Fos and NeuN co-staining after AS-TSPO treatment, Scale bar = 50 μm. G, H ELISA results showing the levels of IL-1β and IL-18 after AS-TSPO treatment, n = 6 per group. I Western blot showing the expression levels of TSPO, NLRP3, Caspase-1 and GSDMD-N after AS-TSPO treatment. Data are represented as mean ± SEM, ***p < 0.001

Mitochondial dysfuction and AMPK-PGC-1α activation were observed after SNI

Since TSPO has been associated with pyroptosis in alleviating neuropathic pain, we next aimed to investigate the mechanisms by which TSPO regulates pyroptosis. Transcriptome analysis revealed that SNI induces mitochondrial dysfunction and activates oxidative stress (Figs. 2A-C and 5A). Mitochondrial integrity is crucial in regulating pyroptosis, with adenosine 5'-monophosphate-activated protein kinase (AMPK) and peroxisome proliferator-activated receptor γ coactivator 1α (PGC-1α) identified as key regulatory molecules.

Fig. 5

Mitochondial dysfuction and AMPK-PGC-1α activation were observed after SNI. A GSEA results for enrichment of oxidative stress response. B Schematic diagram of the experimental flow. C Volcano plot showing the significantly upregulated and downregulated genes between SNI and AS-TSPO groups, n = 3 per group. D PCR array heat map illustrating the relative mRNA levels of mitochondrial injury-related genes, including those associated with fusion and fission, between the SNI and AS-TSPO groups, n = 3 per group. E Western blot showing the expression levels of p-AMPK, AMPK, PGC-1α, MFN2 and TFAM in SNI and AS-TSPO groups. F–H Levels of antioxidant stress markers (GSH, SOD) and oxidative stress marker (MDA) in different groups, n = 6 per group. I Representative transmission electron microscopy images of mitochondria in the spinal dorsal horn of indicated groups, scale bar = 2 μm. Data are represented as mean ± SEM, ***p < 0.001

By conducting PCR array analysis (Fig. 5B), the overall expression and the up- and down- regulation distribution of mitochondria-related genes were displayed (Fig. 5C; Supplementary Fig. 3A). A heat map showed the expression levels of mitochondrial injury, fusion, fission, and pathway-related genes in the SNI and AS-TSPO groups on day 7 in the spinal dorsal horn (Fig. 5D). Specifically, compared to AS-TSPO, we found that SNI significantly upregulated the expression of mitochondrial damage-related genes (Bak1, Bid) and fission genes (Fis1, Drp1). In contrast, AS-TSPO treatment notably increased the expression of fusion genes (Mfn2, Mfn1, Opa1). This suggests that AS-TSPO treatment may induce a trend where fission gene expression decreases and fusion gene expression increases. Regarding the AMPK-PGC-1α pathway, AS-TSPO treatment significantly upregulated the expression of genes associated with this pathway.

At the protein level, AS-TSPO treatment upregulated AMPK-PGC-1α pathway proteins, MFN2, and TFAM, while decreased DRP1 expression (Fig. 5E; Supplementary Fig. 3G-K). It also reduced oxidative stress levels, indicated by lower MDA and increased antioxidant markers (GSH, SOD) (Fig. 5F-H), and improved mitochondrial health, as shown by elevated mtDNA copy numbers (Supplementary Fig. 3L). Through transmission electron microscopy results, we found that AS-TSPO alleviated mitochondrial damage caused by SNI (Fig. 5I). Overall, TSPO overexpression activated the AMPK-PGC-1α pathway, supporting mitochondrial biogenesis, maintaining the balance of fusion and fission, reducing oxidative stress levels, and sustaining mitochondrial function, collectively influencing pyroptosis and the release of inflammatory factors.

TSPO protected against mitochondrial dysfunction and pyroptosis by enhancing AMPK-PGC1α pathway

To make sure that TSPO regulates mitochondrial function and mitigate pyroptosis and inflammation via the AMPK-PGC-1α pathway, the AMPK inhibitor Compound C was used to block this pathway. Compound C 6 μg/day in 10 μl saline was intrathecal administrated from the first day after SNI [45] (Fig. 6A), and the results showed that Compound C effectively inhibited AMPK-PGC-1α pathway, reducing the expression of mitochondrial biogenesis protein TFAM and fusion protein MFN2, while increasing mitochondrial fission protein DRP1 Fig. 6B; Supplementary Fig. 4A-F). These changes led to elevated MDA levels, reduced antioxidant factors (GSH and SOD), (Supplementary Fig. 4G-I) and mtDNA copy number (Supplementary Fig. 4 J), thereby exacerbating mitochondrial damage (Fig. 6C). Inhibition of the AMPK-PGC-1α pathway also elevated pyroptosis-related inflammatory cytokines IL-1β and IL-18 (Fig. 6D, E), as well as pyroptosis proteins NLRP3, Caspase-1, and GSDMD-N (Fig. 6F; Supplementary Fig. 4 K-M). These changes ultimately led to behavioral alterations, as the c-Fos + neurons (Fig. 6G; Supplementary Fig. 4N) and the analgesic effect of AS-TSPO was significantly diminished by Compound C (Fig. 6H). These findings indicate that TSPO regulates mitochondrial fusion and fission, influencing cellular pyroptosis and alleviating neuropathic pain via the AMPK-PGC-1α pathway.

Fig. 6

TSPO protected against mitochondrial dysfunction and pyroptosis by enhancing AMPK-PGC1α pathway. A Schematic diagram of the experimental flow. B Western blot showing the expression levels of TSPO, p-AMPK, AMPK, PGC-1α, MFN2 and TFAM after inhibition of AMPK. C Representative transmission electron microscopy images of mitochondria after inhibition of AMPK, scale bar = 2 μm. D, E ELISA results showing the levels of the IL-1β and the IL- after inhibition of AMPK, n = 6 per group. F Western blot showing the expression levels of NLRP3, Caspase-1 and GSDMD-N after inhibition of AMPK. G Representative immunofluorescence images showing c-Fos and NeuN co-staining after inhibition of AMPK. H Behavioral testing of von Frey after inhibition of AMPK at different time points (days 0, 3, 7, 14, and 21), n = 10 per group. Data are represented as mean ± SEM, ***p < 0.001

Mitochondrial dysfunction was observed during pyroptosis induction in primary astrocytes

To validate the astrocyte-specific role of TSPO, primary astrocytes were isolated (Fig. 7A). Astrocytes were treated with LPS (100 ng/mL) for 24 h, followed by a 30-min pulse of 5 mM ATP to induce pyroptosis [39]. The pyroptosis model was validated by elevated IL-1β and IL-18 levels in the supernatant (Fig. 7B, C), an increase in PI-stained pyroptotic astrocytes (Fig. 7D; Supplementary Fig. 5A), and higher LDH levels, indicating membrane rupture and enhanced permeability (Supplementary Fig. 5B). TSPO and pyroptosis-related proteins (NLRP3, Caspase-1 and GSDMD-N) were notably upregulated, consistent with findings from animal models (Fig. 7E; Supplementary Fig. 5C-F). From the perspective of mitochondria, after pyroptosis occurs in astrocytes, there is a significant increase in ROS both within the astrocytes and in the mitochondria (Fig. 7F, G; Supplementary Fig. 5G, H). This is accompanied by a decreased membrane potential (Fig. 7H; Supplementary Fig. 5I), reduced antioxidant markers (GSH, SOD), elevated MDA levels (Supplementary Fig. 5 J-L), and suppressed mtDNA levels (Supplementary Fig. 5 M). Additionally, observable mitochondrial network fragmentation occurs (Fig. 7I, J).

Fig. 7

Mitochondrial dysfunction was observed during pyroptosis induction in primary astrocytes. A Representative images of GFAP immunofluorescence staining to identify astrocytes, scale bar = 50 μm. B, C ELISA showing the levels of IL-1β and IL-18 in different groups, n = 6 per group. D Representative images of PI staining demonstrate the number of pyroptotic cells in the different groups, scale bar = 50 μm. E Western blot showing the expression levels of p-AMPK, PGC1, MFN2, and TFAM after ATP + LPS treatment. F Representative images of ROS staining indicate the level of ROS produced by astrocytes after ATP + LPS treatment, scale bar = 50 μm. G Representative images of MitoSOX staining show the level of mitochondrial ROS levels after ATP + LPS treatment, scale bar = 50 μm. H Representative images of JC-1 staining show a shift from aggregates (red) to monomers (green), indicating mitochondrial depolarization, scale bar = 50 μm. I Representative images of Mito-Tracker staining show mitochondrial morphology, scale bar = 20 μm. J Quantification of mitochondrial aspect ratio, n = 6 per group. K Western blot showing the expression levels of TSPO, NLRP3, Caspase-1, and GSDMD-N after ATP + LPS treatment. Data are represented as mean ± SEM, ***p < 0.001

Meanwhile, TSPO was also significantly increased in the pyroptosis group, consistent with our animal experimental results. In the pyroptosis model, there was an imbalance in oxidative stress, with elevated reactive oxygen species (ROS) and mitochondrial damage, which aligns with the findings from the animal experiments. Additionally, we observed a significant increase in proteins related to mitochondrial biogenesis (TFAM, mitochondrial copy number), fusion (MFN2), and fission (DRP1) compared to the normal group. Based on previous experimental results, we speculate that this change may be associated with insufficient upregulation of AMPK-PGC-1α in the pyroptosis model (Fig. 7K; Supplementary Fig. 5N-R).

TSPO-driven modulation of mitochondrial function and pyroptosis in astrocytes via the AMPK-PGC-1α pathway

Next, we investigated the effects of TSPO and the AMPK-PGC-1α pathway on astrocytes. Pyroptotic astrocytes were treated with Ro 5–4864 at various concentrations (0, 10, 50, 100, and 300 nM) for 2 h. CCK-8 assay showed that 100 nM Ro 5–4864 had the most pronounced effect (Supplementary Fig. 6A), which was selected as the standard condition for subsequent treatments. Supplement of Ro 5–4864 significantly reduced the proliferation of pyroptotic astrocytes, as shown by PI staining, and decreased the levels of IL-1β, IL-18, and LDH in the supernatant (Fig. 8A, B; Supplementary Fig. 6B, C). TSPO was elevated with the activation of the AMPK-PGC-1α pathway, which significantly increased the expression of the mitochondrial fusion protein MFN2 and the biogenesis marker TFAM, while attenuating the increase in the fission protein DRP1 (Fig. 8C; Supplementary Fig. 6D-I).

Fig. 8

TSPO-driven modulation of mitochondrial function and pyroptosis in astrocytes via the AMPK-PGC-1α pathway. A Representative images of PI staining showed the number of pyroptosis cells in different groups, scale bar = 50 μm. B ELISA showing the levels of IL-1β and IL-18 in different groups, n = 6 per group. C Western blot showing the expression levels of TSPO, p-AMPK, PGC1, MFN2, and TFAM in different groups. D Representative images of ROS staining indicates the level of ROS produced by cells in different groups, scale bar = 50 μm. E Representative images of MitoSOX staining shows the level of mitochondrial ROS levels in different groups, scale bar = 50 μm. F Representative images of JC-1 staining shows a shift from aggregates (red) to monomers (green), indicating mitochondrial depolarization, scale bar = 50 μm. G Quantitative analysis of the relative intensity of JC-1 in different groups, n = 6 per group. H Quantification of mitochondrial aspect ratio, n = 6 per group. I Representative images of Mito-Tracker staining show mitochondrial morphology, scale bar = 20 μm. J Representative images of TSPO and Mito-tracker co-staining, scale bar = 20 μm. K Western blot showing the expression levels of NLRP3, Caspase-1, and GSDMD-N in different groups. Data are represented as mean ± SEM, **p < 0.01, ***p < 0.001

Ro 5–4864 improved cellular antioxidant capacity, evidenced by increased GSH and SOD levels, reduced MDA, and restored mtDNA levels (Supplementary Fig. 6 J-M). It also lowered intracellular and mitochondrial ROS (Fig. 8D, E; Supplementary Fig. 6N, O), restored mitochondrial membrane potential (Fig. 8F, G), and reduced mitochondrial network disruption and fragmentation, as confirmed by TSPO and Mito Tracker co-staining (Fig. 8H-J). Pyroptosis-related proteins (NLRP3, Caspase-1, GSDMD-N) were ultimately downregulated (Fig. 8K; Supplementary Fig. 6P-R). However, treatment with the AMPK inhibitor Compound C significantly reduced AMPK and PGC-1α expression, disrupted mitochondrial fusion-fission balance, and impaired mitochondrial function, thereby abolishing the protective effect of TSPO on pyroptosis-related proteins. These findings confirm that TSPO enhances mitochondrial function and regulates pyroptosis in astrocytes through the AMPK-PGC-1α pathway.

Association of TSPO gene with multiple pain phenotypes and validation of potential targeted drugs

To gain the value of this study in clinic, we conducted Mendelian randomization (MR) analysis using GWAS data to explore the association between the TSPO gene and various types of pain, including headache, knee pain, neck and shoulder pain, and back pain (Fig. 9A). The MR analysis indicated that TSPO was associated with neck, shoulder, and back pain. Unexpectedly, we found that the direction of the β-value was consistent and negatively correlated across all pain phenotypes (Fig. 9B), which suggested a negative correlation between TSPO and pain. Increased expression of TSPO was linked to pain relief, further supporting the effectiveness of our strategy to upregulate TSPO for pain alleviation, as demonstrated from a human whole-genome perspective.

Fig. 9

Association of TSPO gene with multiple pain phenotypes and validation of potential targeted drugs. A Schematic diagram showing the localization of TSPO and its association with different pain types, including neck and shoulder pain, back pain, knee pain, and foot pain. B MR results depicting the association of TSPO with various pain phenotypes. β-values and 95% confidence intervals (CI) for each phenotype are shown, with significant associations highlighted in red. C Manhattan plot from the SMR analysis of TSPO expression with neck and shoulder pain, showing the association signals across the genome. D Venn diagram showing the intersection of TSPO-targeted drugs from three different drug databases. Chlormezanone and Zopiclone were identified as common drugs across all databases. E, F Molecular docking of Chlormezanone and Zopiclone with TSPO, showing their respective binding energies. G Behavioral changes at different concentrations of Zopiclone were assessed using the von Frey test at multiple time points (days 0, 3, 7, 14, and 21), n = 10 per group. Data are represented as mean ± SEM, *p < 0.05, **p < 0.01, ***p < 0.001

Next, we further confirmed this causal relationship by performing SMR analysis on TSPO, focusing on various types of pain (Supplementary Fig. 8). Since both MR and SMR tests showed positive results for neck or shoulder pain, we performed Bayesian colocalization analysis to exclude potential pleiotropic effects. The results, by integrating GWAS and eQTL data, revealed that TSPO expression is a causal factor for the neck or shoulder pain phenotype across all analyses (Fig. 9C and Table 1)

Table 1 SMR and HEIDI results for TSPO gene across different pain locations

To uncover clinical drugs associated with TSPO and assess their potential in pain treatment, we performed a comprehensive search in three major drug target databases to identify TSPO-targeting compounds. This search highlighted two drugs—chlormezanone and zopiclone—that were consistently present across all three databases (Fig. 9D). Subsequent molecular docking studies confirmed that both compounds bind within TSPO with binding affinities of -5.6267 kcal/mol for chlormezanone and -6.9176 kcal/mol for zopiclone, signifying favorable binding activity (Fig. 9E, F). The superior binding affinity of zopiclone warranted its selection for further experimental investigation. We evaluated the therapeutic potential of zopiclone through intrathecal administration in SNI model. Initial testing showed that a dose of 0.25 µg was ineffective for pain relief, whereas a 5 µg dosage induced side effects, including drowsiness, complicating behavioral assessments. After dose optimization, a 2 µg dose was identified as the most effective dosage, achieving approximately 50% pain relief by day 21, though with limited efficacy during the early treatment phase (Fig. 9G).

Discussions

This study elucidates the critical roles of mitochondrial dysfunction, pyroptosis, and inflammation in neuropathic pain following SNI in spinal dorsal horn. Our findings highlight TSPO, a mitochondrial gene, as a key regulator of pyroptosis. Upregulation of TSPO reduces the expression of pyroptosis-related proteins (NLRP3, Caspase-1, GSDMD-N) and decreases the release of inflammatory factors (IL-1β, IL-18). Mechanistic studies demonstrate that TSPO regulates mitochondrial biogenesis and the dynamic balance of fusion and fission through the AMPK-PGC1α pathway, enhancing antioxidant capacity, reducing both cellular and mitochondrial ROS levels, and maintaining mitochondrial homeostasis. To further enhance the clinical relevance of this research, we utilized human whole-genome sequencing data to validate TSPO upregulation as a therapeutic strategy for pain relief, identifying clinically relevant drugs targeting TSPO and providing new avenues for repurposing existing medications.

Pyroptosis is a programmed cell death mechanism reliant on inflammasomes, characterized by cell membrane rupture and the release of pro-inflammatory cytokines such as IL-1β and IL-18, which are significant in inflammatory diseases [46]. In neuropathic pain, inflammation plays a key role in inducing hypersensitivity and persistent pain. Numerous studies have shown that inflammatory factors, including TNF-α, IL-6, and IL-1β, are closely related to the onset and maintenance of pain [47,48,49,50,51]. These cytokines enhance pain signal transmission by activating nociceptive neurons or altering synaptic plasticity in the central nervous system. However, despite extensive research on the connection between inflammation and pain, the role of pyroptosis in pain has not been sufficiently explored. Only a few studies have preliminarily indicated that pyroptosis may exacerbate pain by activating inflammatory pathways in the nervous system [52]. Our study highlights the pivotal role of pyroptosis in the induction of pain. We observed elevated expression of pyroptosis-related proteins (NLRP3, Caspase-1, GSDMD-N) and inflammatory factors (IL-1β, IL-18) in the spinal dorsal horn following SNI, which correlates with increased pain sensitivity. In contrast, the upregulation of TSPO mitigates pain behaviors in mice, as evidenced by reduced levels of pyroptosis-related proteins and inflammatory factors.

Research indicates that mitochondrial dysfunction is a primary trigger of pyroptosis. In lumbar disc degeneration, oxidative stress in nucleus pulposus cells results in mitochondrial damage, which causes the release of mitochondrial DNA into the cytoplasm. This event subsequently activates the cGAS-STING-NLRP3 axis, leading to the induction of pyroptosis. Therefore, maintaining mitochondrial homeostasis is crucial for regulating pyroptosis [53]. Mitochondrial biogenesis, fusion, and fission are essential for this homeostasis [54,55,56]. Although literature confirms increased expression of mitochondrial fission proteins in the spinal dorsal horn after SNI [57], focusing solely on mitochondrial fission without considering the overall dynamic balance (biogenesis, fusion, and fission) is insufficient to address fundamental issues. External stimuli, such as spinal nerve injury (SNI), cause mitochondrial damage that triggers mitochondrial fission, while mitochondrial biogenesis and fusion respond to maintain mitochondrial health. However, these processes are critical and should not be overlooked, as they are essential for preserving mitochondrial function. In our study, the results indicate an excessive increase in the mitochondrial fission protein DRP1, coupled with an insufficient elevation of proteins related to mitochondrial biogenesis (TFAM) and fusion (MFN2) following SNI. This imbalance contributed to heightened oxidative stress, a decline in mitochondrial membrane potential, elevated reactive oxygen species (ROS) levels, and exacerbated mitochondrial damage.

These changes are significantly regulated by the AMPK-PGC-1α pathway. As a cellular energy sensor, AMPK responds to metabolic and oxidative stress by activating PGC-1α, which upregulates the expression of mitochondrial-related genes, promoting mitochondrial DNA replication, transcription, and biogenesis, thereby improving mitochondrial function under stress conditions [58,59,60]. However, in the SNI group, although AMPK and PGC-1α were upregulated, they were insufficient to restore the balance of mitochondrial fission and fusion. Specifically, the high expression of mitochondrial fission proteins persisted, while the expression of fusion and biogenesis proteins remained inadequately elevated, indicating insufficient activation of the AMPK-PGC-1α pathway following injury. When we upregulated TSPO, we observed a reduction in the expression of mitochondrial fission proteins (DRP1) and an increase in the expression of mitochondrial biogenesis (TFAM) and fusion proteins (MFN2). This restoration of the balance between mitochondrial fusion and fission helped normalize mitochondrial function, reduce oxidative stress, lower ROS production, restore mitochondrial membrane potential, and ultimately improve overall mitochondrial health. These changes resulted in reduced pyroptosis and significantly alleviated pain behaviors in mice.

The downstream regulation of mitochondrial dynamics by the AMPK-PGC-1α signaling pathway is an important and intriguing area of focus. Studies have shown that the AMPK-PGC-1α axis regulates the astrocytic glutathione system to enhance mitochondrial function, thereby protecting it from oxidative and metabolic damage [61]. AMPK also modulates mitochondrial dynamics by activating autophagic flux in neurodegenerative diseases [62]. Furthermore, research suggests that the AMPK-PGC-1α-SIRT3 axis plays a critical role in regulating mitochondrial homeostasis and redox balance [63]. Currently, research primarily highlights the involvement of the glutathione system, autophagic flux, and SIRT3 in mitochondrial dynamics regulation. The regulation of mitochondria by the AMPK-PGC-1α axis remains underexplored and presents an exciting scientific question worthy of further investigation.

Additionally, through human Mendelian randomization analysis, we confirmed the significant association between TSPO expression and pain phenotypes, highlighting its clinical potential in treating neuropathic pain. Using drug databases, we identified Chlormezanone and Zopiclone as promising compounds for TSPO regulation. Chlormezanone, a central muscle relaxant effective for anxiety and muscle tension, is often combined with other analgesics to enhance pain relief in neuropathic conditions, suggesting a potential synergistic effect [64,65,66].

Zopiclone, a commonly used hypnotic for the treatment of insomnia, belongs to the class of non-benzodiazepine sedative-hypnotics [67]. Studies have demonstrated that zopiclone exerts its sedative effects primarily by binding to GABA-A receptors and potentially through interactions with TSPO, modulating cellular responses to enhance its sedative and sleep-promoting effects, thereby improving sleep structure and quality [68, 69]. Clinically, zopiclone is widely used to treat various types of insomnia, including difficulty in falling asleep, maintaining sleep, and early awakening. It is particularly suitable for adults, elderly patients, and individuals with hepatic impairment, offering fewer side effects compared to traditional benzodiazepines. Clinical studies have also revealed that zopiclone can provide sustained relief from chronic neuropathic pain and improve sleep quality in affected patients [70], with enhanced sleep quality potentially contributing to reduced pain perception [71]. Furthermore, evidence suggests that zopiclone, when used in combination with other analgesics, can enhance analgesic efficacy, supporting its role in multimodal pain management strategies [71, 72].

Our molecular docking analysis demonstrated that zopiclone exhibits superior binding energy compared to chlormezanone, corroborating its analgesic effects observed in mice. Behavioral experiments revealed that high doses of zopiclone induce sedation, impacting behavioral assessments, while lower doses showed minimal effects. An intrathecal dose of 2 µg was identified as the most effective, achieving approximately 50% pain relief on postoperative day 21 in the spared nerve injury (SNI) model. However, its limited efficacy during the early treatment phase may be attributed to rapid metabolism, with low doses transiently affecting TSPO expression. Prolonged TSPO expression, necessitating a cumulative effect, appears critical for sustained analgesia. While zopiclone represents a potential TSPO-targeted therapeutic for pain management, its development pathway remains challenging. Our findings provide new insights into its application for neuropathic pain and lay a foundation for future clinical trials and TSPO-focused therapies, although significant efforts are still required for clinical translation.

Despite the significant progress achieved in this study, several limitations remain that need to be addressed in future research. First, our results primarily focus on demonstrating how TSPO regulates mitochondrial function through the AMPK-PGC-1 pathway, influencing pyroptosis and contributing to neuropathic pain. However, multiple factors drive pyroptosis, and our study only investigates the mitochondrial aspect. For instance, Levo-tetrahydropalmatine has been reported to alleviate neuropathic pain by regulating pyroptosis through the Clec7a-MAPK/NF-κB pathway [9]. Additionally, autophagy regulation has been shown to reduce NLRP3 inflammasome-dependent pyroptosis, thereby mitigating neuropathic pain [6, 8]. Other factors, such as potassium efflux, calcium signaling, and lysosomal leakage, have also been shown to influence the regulation of pyroptosis by mitochondria [73]. This limitation underscores the need for future studies to explore alternative pathways and mechanisms contributing to pyroptosis in neuropathic pain. Second, although we identified potential TSPO-targeted drugs through database screening and molecular docking, along with preliminary behavioral assessments, further in-depth research is needed to comprehensively evaluate their efficacy and safety. Future studies should include detailed analyses of drug metabolism, structural assessments, and rigorous preclinical testing to establish their therapeutic potential. Additionally, the potential side effects of these drugs must be investigated to facilitate the translation of these findings into clinical applications. Conducting more extensive research into the clinical utility of TSPO-targeted therapies for neuropathic pain is critical for advancing this field. Addressing these limitations in future studies will contribute to a deeper understanding of pyroptosis mechanisms and enhance therapeutic strategies for neuropathic pain.