Electroacupuncture alleviates central post-stroke pain in rats by modulating miR-21-5p/TGF-β/Smads pathway Authors: Guihua Tian, Yang Wu, Zhihao Shang, Nan An, Huifeng Hao, Ke He, Youxiang Su, Luyu Yang, Meiyue Wang, Xinyi Li Published: 2025 ABSTRACT AbstractBackgroundCentral post-stroke pain (CPSP) poses a persistent challenge with limited responses to current pharmacological interventions. Recent evidence implicates miR-21 and its downstream TGF-β/Smads signaling pathway in pain pathophysiology. This study investigates whether the analgesic effect of electroacupuncture (EA) is related to miR-21/TGF-β/Smads signaling pathway in CPSP.MethodsCPSP was induced in rats by collagenase-IV injection into the right thalamic ventral posterolateral nucleus to produce a focal hemorrhagic stroke lesion. miR-21-5p and Smad7 expressions were manipulated by adenoviral transfection. Pain thresholds were assessed during 15 days after CPSP induction, followed by brain tissue analysis using nissl staining, immunofluorescence, immunohistochemistry, and molecular assays. A dual-luciferase reporter assay was employed to verify the direct binding interaction between miR-21-5p and Smad7.ResultsThe results showed EA significantly alleviated CPSP, and meanwhile, downregulated miR-21-5p, upregulated Smad7, and reduced TGF-β1, p-Smad2, and p-Smad3 expressions, glial hyperactivation and inflammatory factors (IL-1β, IL-6, TNF-α) in CPSP rats. In addition, miR-21-5p inhibition resulted in analgesic effect, while miR-21-5p overexpression or Smad7 inhibition counteracted the effect of EA.ConclusionsThese results indicate that the mechanisms underlying EA-induced analgesia involve downregulating miR-21-5p, consequently inhibiting neuroinflammation by regulating Smad7/TGF-β/Smads signaling pathway, providing a novel insight for probing the molecular mechanism of analgesic effect of EA.Supplementary InformationThe online version contains supplementary material available at 10.1186/s10194-025-02247-1.Keywords: Central post-stroke pain, Electroacupuncture, miR-21-5p, Smad7, Glial activation, Neuroinflammation Introduction Central post-stroke pain (CPSP), a type of chronic neuropathic pain initiated by cerebrovascular trauma, distresses 11% of stroke survivors, wreaking havoc on their physiological and psychological well-being and imposing substantial socio-economic challenges [1]. CPSP most frequently occurs following hemorrhagic stroke lesions affecting the thalamus, characterized by pronounced neuroinflammation in the brain, including neuronal damage and glial activation [2–4]. The primary approach to CPSP management involves pharmaceutical interventions, including anticonvulsants, antispasmodics, antidepressants, and analgesics. However, their effectiveness is often constrained by the extensive side effect profiles observed in patients [5–7]. Therefore, there is a critical need to identify novel mechanisms and develop innovative therapeutic approaches. Electroacupuncture (EA), in which specific points on the body called acupoints are stimulated electrically, has gained widespread acclaim in pain management for its promising efficacy and minimal side effects [7–12]. Clinical studies have demonstrated a significant therapeutic effects of acupuncture treatment by relieving pain of CPSP patients [5, 9, 11]. In our prior study, we found that EA mitigates CPSP caused by thalamic injury within a week, attenuates neuronal damage and curbs abnormal glial activation [4]. However, there are few studies exploring how EA improves CPSP and the important molecular mechanisms involved. Therefore, in-depth research on the molecular mechanisms of EA to alleviate CPSP will provide an important basis for the clinical treatment of CPSP. Studies indicate a close correlation between the expression of microRNA-21(miR-21) and neuropathic pain [13]. miR-21-5p, the 5′-terminal complementary strand of miR-21, is upregulated in dorsal root ganglion neurons following spared nerve injury [14]. Inhibition of miR-21-5p significantly relieves neuropathic pain [15]. Studies confirmed that miR-21-5p facilitating intercellular communication among neurons, glial cells, and macrophages via exosome secretion [16]. Additionally, miR-21-5p modulates inflammatory responses and the expression of genes associated with pain have unveiled in some researches [17], suggesting that miR-21-5p may serve as a potential therapeutic target for neuropathic pain management. Smad7, a key regulator of intracellular signal transduction, functions as a negative modulator of the TGF-β/Smad signaling pathway and may be subject to regulation by miR-21-5p [18, 19]. This pathway is known to orchestrate activation of glial cells and promote neuroinflammation responses that are integral to pain sensation, making Smad7 crucial in pain pathogenesis [20–22]. In our study, we found that EA could modulate miR-21-5p level in brain tissue after thalamic injury, but whether EA inhibits neuroinflammation induced by thalamic injury through modulating miR-21-5p/TGF-β/Smads to ameliorate CPSP is unclear. The role of miR-21-5p in analgesia in EA is significant for further discovering the pathogenesis of CPSP or developing new strategies for treating CPSP. Consequently, this study examined the impact of EA on miR-21-5p, p-Smad2,Smad7, TGF-β1and p-Smad3 expression levels in a rat CPSP model, aiming to unravel whether EA alleviates CPSP through modulating miR-21-5p and its downstream TGF-β/Smads signaling pathway to suppress neuroinflammation. Materials and methods Experimental animals Sprague-Dawley male rats (280–300 g, 10–12 weeks old) were obtained from Beijing Viton Lihua Laboratory Animal Technology Co., Ltd., China, and acclimated for one week in a specific pathogen-free (SPF) laboratory environment at Dongzhimen Hospital, Beijing University of Chinese Medicine (temperature: 22–25 °C, humidity: 60% ± 15%, 12-hour light-dark cycle). The experimental protocols were approved by the hospital’s Animal Care and Use Committee (Ethics Approval No. 22 − 01) and complied with the National Institutes of Health Guidelines for the Care and Use of Laboratory Animals. This study adhered to the ARRIVE 2.0 guidelines. Sample-size planning, randomization, allocation concealment, blinding, and inclusion/exclusion criteria were specified a priori. Establishment of the CPSP rat model The CPSP rat model was established following the methodology of Wasserman et al. (2009) [23], by producing a focal thalamic hemorrhagic stroke lesion through stereotactic microinjection of type IV collagenase (Sigma, USA) into the right ventral posterolateral (VPL) nucleus of the thalamus. Before the procedure, rats were anesthetized via intraperitoneal injection of 1% pentobarbital sodium (40 mg/kg), then shaved, disinfected, and securely positioned on a stereotactic apparatus. Holes were drilled in the skull overlying the OE (olfactory bulb,3.8 mm from the coronal sagittal point, 3.3 mm from the midline, and 6.0 mm deep with bregma as reference point according to a rat brain atlas). We used either 0.125 U/0.5 µl of collagenase (CPSP group) or 0.9% saline (sham-operated group) at a rate of 0.1 µl/min with a 1 µl Hamilton microinjector. The needle was rested for 5 min before carefully withdrawn to avoid backflow, and the rats were then placed in a warm, clean cabinet for recovery. After completion of baseline nociceptive assessments, rats were randomly allocated to groups using a computer-generated sequence prepared by a team member not involved in surgery or outcome testing. Allocation concealment was maintained with opaque, sealed envelopes that were opened by the surgeon immediately before stereotaxic injection. Pre-specified exclusion criteria included VPL mis-targeting (histologically verified), perioperative mortality, irrecoverable technical failure, or inability to generate a reflex paw-withdrawal on a given test day; exclusions were logged a priori and were not based on outcome magnitude. Adenovirus construction and transfection Adenoviruses, including AAV9-rno-miR-21-5p-knock down (1.70E + 13.00 v.g/ml), AAV9-rno-miR-21-5p-overexpression (3.91E + 13.00 v.g/ml), AAV9-Smad7-knock down (3.00E + 13.00 v.g/ ml), and AAV9-Smad7-overexpression (2.03E + 13.00 v.g/ml), were commercially constructed (GKN Chemtech, China). These viruses carried a vector viral backbone protein and interference sequences as follows: AAV9-rno-miR-21-5p-knock down with U6-MCS-CAG-EGFP; 5,-TAGCTTATCAGACTGATGTTGA-3; AAV9-rno-miR-21-5p-expression with CMV bGlobin-EGFP-MCS-WPRE-hGH polyA; AAV9-Smad7-knock down (3E + 13v.g/ml) with CMV-NLS-SaCas9-NLS-3xHA-bGHpA-U6-sgRNA, 5,-AGCAGCTCCAGCTGCCGCTCC-3;AAV9-Smad7–overexpression (2.03E + 13.00v.g/mL) with CMV-betaGlobin-MCS-SV40 PolyA. Transfection of adenoviruses was performed 3 weeks prior to model establishment. The location of the injection sites was right lateral ventricle (coordinates: 1.1 mm from the coronal sagittal point, 1.5 mm from the midline, 4.5 mm deep) and the injection was performed as previously described. Electroacupuncture intervention Acupuncture was applied at ‘Baihui’ (DU20) and ‘Zusanli’ (ST36) following the ‘Animal Acupuncture Point Atlas’ guidelines. Using 0.25 mm × 13 mm needles, intervention was commenced on day 1 post-modeling, delivered every other day for 30 min during 15 days at a frequency of 15 Hz and a current of 1 mA. Von Frey test An electronic von Frey anesthesiometer (IITC Life Science, USA) was used to assess mechanical withdrawal threshold. Animals were acclimated to the testing environment for ≥ 15 min before each session. A rigid polypropylene tip was applied perpendicularly to the mid-plantar surface of the left hind paw contralateral to the lesion, with steadily increasing force until a reflexive withdrawal occurred. A valid withdrawal was defined a priori as a rapid paw lift, flick, or brief licking of the stimulated paw; guarding, postural adjustments, or stepping were not scored to minimize dependence on complex motor coordination. Force at withdrawal (g) was recorded. Measurements were obtained at baseline (day 1 before modeling) and on days 3, 7, 11, and 15 post-modeling. Each session comprised three trials separated by 10-min inter-trial intervals; the mean of the three trials was used for analysis. The testing order across animals was kept consistent across days to reduce between-session variability. Hot plate inspection Thermal nocifensive reflexes were measured using an IITC hot/cold plate pain tester (IITC Life Science, USA). The animal was placed on the plate at a fixed temperature (constant across all sessions; device verified before each session) and the hind-paw withdrawal latency was recorded using a stopwatch integrated with the instrument. A response was defined as a rapid hind-paw lift, flick, or licking. To prevent tissue damage, a 30-s cutoff was enforced. Each session included three trials with ≥ 5-min inter-trial intervals, and the mean latency was used for analysis. Nissl staining Following 15 days of electroacupuncture treatment, nissl staining was utilized to observe neuronal damage. After perfusion with saline and paraformaldehyde, the brains were collected, sectioned, and stained. Tissue sections were sequentially dewaxed to 80% ethanol and stained with Nissl dye (Solarbio Life Science, China) at 56˚C for 1.5 h. Following staining, sections were washed, differentiated in 95% alcohol, dehydrated, and sealed. Nissl body alterations were examined under an optical microscope to assess the average number of Nissl-positive cells. Immunofluorescence Rat brain tissue sections, previously embedded in paraffin, underwent incubation with GFAP (1:250, Abcam, UK), IBA1 (1:250, CST, USA), and Smad7 (1:250, Proteintech, China) antibodies. Following this, nuclei were stained with DAPI (1 µg/ml) from Sigma-Aldrich. Image capture and analysis were conducted with a laser scanning confocal microscope (DM6 B, Leica Microsystems) and Image Lab software (Bio-Rad, USA) respectively. Enzyme-linked immunosorbent assay After-blood collection, the supernatant was gathered following centrifugation at 3000 g for 15 min. Serum levels of TNF-α, IL-1β, and IL-6 were quantified using ELISA kits (Dogesce, China) following the manufacturer’s instructions. Fluorescence in situ hybridization Fifteen days after modeling, the rat brains were excised post-anesthesia and cardiac perfusion, then dehydrated and embedded at -40 °C. The process involved digesting frozen brain tissue sections at room temperature with a 3% pepsin solution, followed by several stages of fixation, pre-hybridization, and hybridization using digoxigenin-labeled oligonucleotide probes. Quantitative real-time PCR (PCR) The right thalamic ventral posterolateral nucleus (VPL) region and surrounding brain tissue were isolated for real-time RT-qPCR analysis. RNA was extracted, its quality and quantity were determined, and a cDNA was synthesized for RT-qPCR on a CFX-96 system. The GAPDH gene served as a control, and the experiment was conducted three times. The miR-21-5p forward primer used was (5′ to 3′) TAGCTTATCAGACTGATGTGGGGG; The Smad7 forward primer used was (5′ to 3′) GAGTCTCGGAGGAAGAGGCT, and reverse primer was CTGCTCGCATAAGCTGCTGG; The GAPDH forward primer used was (5′ to 3′) GGACGAGGAAACACTCTCCTG, and reverse primer was TGAGCCTCCTCCAATTCAACC. Immunohistochemistry Following 15 days of electroacupuncture treatment, anesthetized rats underwent cardiac perfusion, and the brains were extracted. Detection of IBA1-associated protein expression levels was carried out via immunohistochemistry, following a series of steps involving retrieval, dehydration, paraffin embedding, and staining. Western blot Proteins from the right thalamic VPL region and surrounding brain tissue were extracted using RIPA buffer, their concentration determined, and then separated by SDS-PAGE. After electrophoresis, proteins were transferred onto PVDF membranes, blocked, and incubated with primary and secondary antibodies. The membranes were then washed, treated with enhanced chemiluminescence (ECL), and visualized using a ChemiDoc™ MP Imaging System. Cell culture and transfection The CTX-TNA rat astrocyte cell line (Fenghui Bio, China) was cultured in DMEM supplemented with 10% fetal bovine serum. Transfection of the miR-21-5p mimic and the miR-21-5p knock down was performed using riboFECTTM CP Reagent (RiboBio, China) when cell concentration reached 60%-70% in a six-well plate. Initial preparation involved diluting either the mimic or inhibitor and gently blending the mixture. After adding the riboFECT™ CP Reagent, the blend was incubated briefly at room temperature. This mix was then introduced to complete medium void of double antibodies and gently stirred. Finally, the plate was incubated at 37 °C in a CO₂ incubator for 48 h. Dual-Luciferase Reporter gene assay The pmirGLO vector (Promega) was used to construct wild-type and mutant reporter plasmids. The Smad7 WT plasmid contained the predicted miR-21-5p binding site within the 3’UTR sequence, while the MUT plasmid had specific nucleotide changes disrupting this site. Both plasmids were co-transfected with either the miR-21-5p mimic or mimic NC into 293T cells using the Hieff Trans™ Liposomal Transfection Reagent. The transfection mixture per well in a 96-well plate contained 0.45 µL of Hieff Trans™, 0.15 µg of the reporter plasmid, and 2 µL of miRNA (10 µM) in a total volume of 20 µL of Opti-MEM. Following a 6-hour incubation at 37 °C in a 5% CO₂ incubator, the medium was replaced with complete growth medium. For luciferase activity measurement, cells were lysed 48 h post-transfection, and the lysates were collected. Firefly and Renilla luciferase activities were quantified using the Dual Luciferase Reporter Assay System and a GloMax Multi Detection System. To account for variations in transfection efficiency and cell viability, firefly luciferase activity was normalized to Renilla luciferase activity. The relative luciferase activity was calculated as the ratio of firefly luciferase to Renilla luciferase activity. Statistical analysis GraphPad Prism 8.0 software was used for all statistical analyses. Results are expressed as mean ± standard deviation (SD). For comparisons among three or more groups, data were analyzed using one-way ANOVA followed by Tukey’s multiple comparisons test. For behavioral data (PWMT and PWTL) measured at multiple time points, two-way repeated measures ANOVA was performed to evaluate the effects of treatment, time, and their interaction, followed by Tukey’s multiple comparisons test for post hoc analysis. A P-value of < 0.05 was considered statistically significant. Results Electroacupuncture mitigates hyperalgesia in CPSP model CPSP rat model was constructed by injecting collagenase into the right VPL region, as illustrated in Fig. 1A. Brain tissue sections confirmed a focal hemorrhagic stroke lesion involving the right VPL region 24 h after injection (Fig. 1B). Mechanical and thermal pain threshold tests demonstrated that the paw withdrawal threshold (PWT) to mechanical stimulation and paw withdrawal thermal latency (PWTL) in CPSP model rats began to decline on day 3 post-modeling, reached their lowest point on day 7, and remained reduced until day 15. To evaluate the analgesic effects of electroacupuncture in CPSP rats, stimulation was applied to the “Baihui” (DU20) and left “Zusanli” (ST36) acupoints, beginning on the first day post-surgery and continuing every other day. Electroacupuncture significantly increased PWMT and PWTL in CPSP rats during the treatment period (Fig. 1C). Electroacupuncture significantly alleviates persistent mechanical and thermal hypersensitivity in the rat model of CPSP, mitigates neuronal damage in CPSP rats, and inhibits glial cell activation and neuroinflammation. A. Workflow of CPSP rat model and EA treatment. B. Site of injection and focal hemorrhagic stroke lesion in the VPL region of the right thalamus 24 h post-injection. C. Pain threshold evaluation (n = 5 or 6). D-E. Nissl stain (purple, neuronal stain) and total neuron numbers estimated on nissl-stained sections (n = 4). Scale bars represent 100 μm as indicated. F. Immunofluorescence labeling for astrocyte marker GFAP (top row, green), microglial marker IBA1 (bottom row, green), DAPI (blue) represents cell nuclei, Scale bars represent 100 μm as indicated. G-H. WB analysis of GFAP and IBA1 (n = 4). I. Cytokine expression was determined by ELISA (n = 6). *** P < 0.001 versus Sham: **** P < 0.0001 versus Sham; #### P < 0.0001 versus CPSP. Statistical analysis was performed using one-way or two-way ANOVA as appropriate Electroacupuncture alleviates neuronal damage, inhibits glial activation, and reduces inflammation in CPSP rats Further Nissl staining revealed a reduction in the number of stained cells and more severe structural damage to neurons in the CPSP group. In contrast, electroacupuncture (EA) treatment mitigated these effects, indicating that EA significantly alleviated neuronal damage in the affected brain regions(Fig. 1D and E). Immunofluorescence and WB results showed significant difference in GFAP and IBA1 protein levels between CPSP and EA treatment groups, astrocyte and microglia activation in CPSP rats. (Fig. 1F-H). ELISA results demonstrated that EA intervention significantly reduced serum levels of the inflammatory cytokines TNF-α, IL-1β, and IL-6 in CPSP rats(Fig. 1I). These findings suggest electroacupuncture can alleviate CPSP-associated pain, potentially through mitigating neuronal damage, inhibiting astrocyte hyperactivation, and reducing inflammation. miR-21-5p is a key target for electroacupuncture to alleviate CPSP Fluorescence in situ hybridisation (FISH) assays found miR-21-5p was localized in neurons (NeuN+) and a significant upregulation of expression was observed in the thalamic region of rats in the CPSP group. This upregulation, however, was diminished in rats receiving electroacupuncture intervention (Fig. 2A-C). This finding suggests that miR-21-5p could be a key target of electroacupuncture in the treatment of CPSP. Electroacupuncture significantly reduced miR-21-5p expression in rats with CPSP model and miR-21-5p co-localized with NeuN. A-C. miR-21-5p in situ hybridization (a and b, n = 3) and RT-PCR (c, n = 4). Scale bars represent 100 μm as indicated To verify the role of miR-21-5p in alleviating CPSP pain by electroacupuncture, we created viral vectors that could effectively knock down or overexpress miR-21-5p (sFig. 1 A). This vector significantly reduced or increased miR-21-5p expression in brain tissues (sFig. 1 B-G). Following the injection of the miR-21-5p knock down viral vector into the right lateral ventricle of rats, we established the CPSP model (Fig. 3A). The results indicated that suppressing miR-21-5p expression significantly elevated mechanical and thermal pain thresholds in CPSP rats. (Fig. 3B), concurrently reducing neuronal damage (Fig. 3C-D), glial activation (Fig. 3E-H), and inflammatory mediators in CPSP rats (Fig. 3I). miR-21-5p knock down significantly alleviates persistent mechanical and thermal hypersensitivity in the rat model of CPSP, mitigates neuronal damage in CPSP rats, and inhibits glial activation and neuroinflammation. A. Workflow of miR-21-5p KD/OE AAV injection. B. Pain threshold evaluation (n = 10). C-D. Nissl stain (purple, neuronal stain) and total neuron numbers counted on nissl-stained sections (n = 3). Scale bars represent 100 μm as indicated. E. Immunohistochemical staining of IBA1 (microglial marker), scale bars represent 100 μm as indicated. F. Immunofluorescence labeling for astrocyte marker GFAP (green), DAPI (blue) represents cell nuclei, scale bars represent 100 μm as indicated. G-H. WB analysis of GFAP and IBA1 (n = 3). I. Cytokine expression was determined by ELISA (n = 6). ** P < 0.01 versus Sham; **** P < 0.0001 versus Sham; && P < 0.01 versus CPSP; &&& P < 0.001 versus CPSP; &&&& P < 0.0001versus CPSP. Statistical analysis was performed using one-way or two-way ANOVA as appropriate On the other hand, the upregulation of miR-21-5p was achieved by injecting an overexpression (OE) virus containing miR-21-5p into the right lateral ventricle, followed by subsequent modeling and electroacupuncture intervention. The virus injection method and workflow are described. When electroacupuncture was applied to CPSP rats overexpressing miR-21-5p, we observed that miR-21-5p upregulation weakened the therapeutic effects of electroacupuncture, as evidenced by reduced mechanical and thermal pain thresholds in the EA-treated group(Fig. 4A), attenuated the repair of neuronal damage (Fig. 4B-C) and faded the ability to inhibit the activation of glial cells (Fig. 4D-G) and inflammatory mediators (Fig. 4H). These results suggest that overexpression of miR-21-5p plays a detrimental role in CPSP by promoting neuronal injury and neuroinflammation. Hence, miR-21-5p is a crucial target for electroacupuncture to alleviate CPSP. Overexpression of miR-21-5p reverses the therapeutic effects of electroacupuncture. A. Pain threshold evaluation (n = 10). B-C. Nissl stain (purple, neuronal stain) and total neuron numbers estimated on nissl-stained sections (n = 3) scale bars represent 100 μm as indicated. D. IBA1 (microglial marker) immunohistochemical staining, scale bars represent 100 μm as indicated. E. Immunofluorescence labeling for astrocyte marker GFAP (green), DAPI (blue) represents cell nuclei, scale bars represent 100 μm as indicated. F-G. WB analysis of GFAP and IBA1 (n = 3). H. Cytokine expression was determined by ELISA (n = 6). ** P < 0.01 versus Sham; ***P < 0.001 versus Sham; **** P < 0.0001 versus Sham; # P < 0.05 versus CPSP + EA; ### P < 0.001 versus CPSP + EA; #### P < 0.0001 versus CPSP + EA. Statistical analysis was performed using one-way or two-way ANOVA as appropriate 3.4 miR-21-5p targets Smad7 for repression in CPSP, involving in electroacupuncture analgesia MicroRNAs mediate post-transcriptional gene silencing by binding to the 3′ untranslated region of target mRNAs, thereby regulating cellular physiological functions and subsequently modulating pain perception [23]. Predictive analyses suggested that miR-21-5p could target Smad7 at the 3’ UTR site (Fig. 5A). The Dual-Luciferase Reporter assay results demonstrated that the miR-21-5p mimic significantly suppressed the luciferase activity of the wild-type vector, whereas it had no significant effect on the mutant vector. This finding indicates that miR-21-5p specifically targets the sequence region of Smad7, confirming a direct interaction between miR-21-5p and Smad7(Fig. 5B). Our study further demonstrated that miR-21-5p overexpression significantly downregulated Smad7 expression in both the rat brain and the CTX-TNA rat astrocyte cell line, whereas miR-21-5p inhibition resulted in Smad7 upregulation in both models. These findings confirm the direct regulatory role of miR-21-5p in modulating Smad7 protein expression (Fig. 5C-E). miR-21-5p targets and suppresses the expression of Smad7. A. Diagram of the binding sites between miR-21-5p and the 3’ UTR of Smad7. B. Double luciferase reporter assays demonstrate that miR-21-5p targets the 3’ UTR of Smad7. C. WB expression of Smad7 in the VPL (ventral posterolateral) in different groups (n = 4). D. WB expression of Smad7 in CTX-TNA rat astrocyte cell line with inhibited miR-21-5p (n = 4). E. WB expression of Smad7 in CTX-TNA rat astrocyte cell line overexpressing miR-21-5p (n = 4) To determine whether Smad7 plays a role in CPSP, we manipulated Smad7 levels through knock down and overexpression (Fig. 6A-G). Overexpression of Smad7 after CPSP resulted in a significant increase in pain threshold (Fig. 6I), suppression of abnormal glial cell activation and neuroinflammation (Fig. 6J-M), Conversely, no changes were induced specifically by knock down of Smad7 after CPSP (Fig. 6H, J-M). These results indicate the involvement of Smad7 in CPSP pathogenesis. Smad7 is a critical target in the CPSP. A. Schematic diagram of the viral vector construction and injection of Smad7 knock down and overexpression (KD represents Smad7 knock down; OE represents Smad7 overexpression). B-D. The effectiveness of Smad7 KD was confirmed by immunofluorescence (b and c, n = 3) and RT-PCR (d, n = 4). E-G. The effectiveness of Smad7 OE was confirmed by immunofluorescence (E and F, n = 3) and RT-PCR (g, n = 4). H-I. Pain threshold evaluation (n = 6 or 8 or 11). J. Immunofluorescence labeling for astrocyte marker GFAP (green), microglial marker IBA1 (green), DAPI (blue) for cell nuclei, all scale bars represent 100 μm as indicated. K-L. WB analysis of GFAP and IBA1 (n = 3). M. Cytokine expression was determined by ELISA (n = 8). **** P < 0.0001 versus Sham; & P < 0.05 versus CPSP; #### P < 0.0001 versus CPSP. ns, P > 0.05 not significant. Statistical analysis was performed using one-way or two-way ANOVA as appropriate Furthermore, when knockdown of Smad7 was combined with electroacupuncture, the analgesic efficacy of electroacupuncture was significantly compromised as evidenced by decreased pain threshold (Fig. 7A), glial cell activation (Fig. 7B-D), and heightened neuroinflammation (Fig. 7E) in the CPSP group. Smad7 is the target through which electroacupuncture exerts its analgesic effects. A. Pain threshold evaluation (n = 6 or 10). B. Immunofluorescence labeling for Smad7 (green), astrocyte marker GFAP (green), microglial marker IBA1 (green), DAPI (blue) represents cell nuclei, scale bars represent 100 μm as indicated. C-D. WB analysis of GFAP, IBA1 and Smad7 (n = 3). E. Cytokine expression was determined by ELISA (n = 6 or 8). *** P < 0.001 versus Sham; **** P < 0.0001 versus Sham; & P < 0.05 versus CPSP + EA; && P < 0.01 versus CPSP + EA; &&&& P < 0.0001 versus CPSP + EA. Statistical analysis was performed using one-way or two-way ANOVA as appropriate Electroacupuncture relieves CPSP through inhibiting miR-21-5p to regulate the TGF-β/Smads signaling pathway Previous studies have emphasized the role of Smad7 in suppressing inflammatory responses and mitigating pain by negatively regulating the TGF-β/Smad signaling pathway [24, 25]. It is highly likely that electroacupuncture alleviates CPSP by modulating miR-21-5p and Smad7, thereby influencing the TGF-β/Smad signaling pathway. Immunoblotting results demonstrated that both electroacupuncture and miR-21-5p inhibition significantly downregulated the expression of TGF-β1, p-Smad2, and p-Smad3 compared to the CPSP group. However, the combination of electroacupuncture and miR-21-5p overexpression resulted in an upregulation of these proteins compared to the electroacupuncture group, suggesting that miR-21-5p plays a crucial role in mediating the effects of electroacupuncture on this pathway(Fig. 8A-B). These findings reinforce the idea that electroacupuncture may exert its analgesic effect by downregulating miR-21-5p, subsequently modulating the TGF-β/Smads signaling pathway (Fig. 8C). Electroacupuncture exerts analgesic effect by influencing miR-21-5p/Smad7, thereby further regulating the TGF-β/Smads pathway in rats with CPSP. A-B. WB expression of TGF-β1, p-Smad2 and p-Smad3 in the VPL (ventral posterolateral) region of rat brain tissue in different groups (n = 3). C. Diagram of the mechanism of electroacupuncture exerts analgesic effect on CPSP Discussion The treatment of CPSP is a clinical challenge, largely limited to the effectiveness to current treatment strategies [26]. Numerous studies have underscored the significant therapeutic potential of acupuncture in treating CPSP, gaining both domestic and international recognition as a promising alternative therapy [27]. Despite this, the precise mechanism underpinning its therapeutic effect remains elusive. Our study discovered that EA can reduce the expression of miR-21-5p, enhance the expression of Smad7, and modulate the TGF-β/Smads signaling pathway, which results in the amelioration of neuronal damage, inhibition of glial cell activation, and reduction of neuroinflammation, consequently leading to a substantial alleviation of CPSP. Pathogenesis of CPSP is mostly related with neuroinflammation, disinhibition, and central sensitization [28]. Studies have demonstrated glial activation in the somatosensory cortex and injured thalamus of CPSP model rats. Likewise, in post-stroke phase, astrocytes and microglia were activated and amassed at the lesion site within hours [29, 30]. Furthermore, activated glial cells facilitated the sustained release of pro-inflammatory mediators, including tumor necrosis factor-alpha (TNF-α), interleukin-6 (IL-6), and interleukin-1 beta (IL-1β). Activated glial cells and inflammatory mediators can further enhance the activation and sensitization of central nervous system neural fibers, leading to pain [31]. In this study, we detected abnormal glial cells activation by measuring GFAP and IBA1 expression in the VPL region of rats. Inflammatory responses were observed by quantifying TNF-α, IL-6, and IL-β levels in rat serum. We further confirmed abnormal glial cell activation and the subsequent release of pro-inflammatory mediators in CPSP. These lines of evidence suggest that activated microglia and astrocytes together with a persistent inflammatory environment are involved in the maintenance of CPSP, supporting previous studies [29, 32]. Moreover, EA effectively alleviated mechanical and thermal hypersensitivity, inhibited aberrant glial cell activation, and reduced neuroinflammation in CPSP rats. These findings strongly suggest that the analgesic efficacy of EA is complex linked to the suppression of glial cell activation and the mitigation of neuroinflammatory processes. Treatment for CPSP ranges from pharmacological therapies to nonpharmacological therapies with each having its own limitations. Studies on the therapeutic mechanism of EA for CPSP are very limited. Numerous studies have shown that electroacupuncture (EA) significantly alleviates inflammatory, neuropathic, cancer-related, and visceral pain in corresponding animal models [33]. Additionally, genome-wide expression profiling using RNA sequencing in CPSP has identified potential mechanisms, indicating that chemical synaptic transmission, sensory perception of pain, and neuroactive ligand-receptor interactions are the predominant functional pathways involved [34]. However, the specific target of mechanism of analgesic action of EA on CPSP has not yet been identified. Further revealing a deeper understanding of targeting mechanism thalamic damage neural circuits and molecular mechanisms may be the focus of future research on CPSP, which might help for the precise treatment of CPSP. miR-21 stands out as the most extensively researched miRNA. Recent study has found that, in some neurological and inflammatory diseases, overexpression of miR-21 may lead to enhanced inflammatory responses. Study has indicated that elevated miR-21 expression could lead to the release of increased inflammatory mediators by activating glial cells [35]. Within the miR-21 family, which includes miR-21-5p and miR-21-3p, the miR-21-5p specifically originates from the 5’ end of pre-miR-21, which is linked with the development of various types of pain [22]. Numerous studies have demonstrated that miR-21-5p typically exerts an analgesic effect by binding to and regulating the 3′ untranslated region of target genes [36], have emerged as playing crucial roles in the initiation and progression of pain. For instance, miR-21-5p suppresses neuropathic pain development by directly targeting C-C motif ligand 1 (CCL1) and tissue inhibitor of metalloproteinase-3 [37]. Another study identified the involvement of miR-21-5p in the development and maintenance of neuropathic pain through immune-related targets, including C-C motif ligand 5 and tyrosine 3-monooxygenase/tryptophan 5-monooxygenase activation protein epsilon. Additionally, silencing miR-21-5p in sensory neurons has been shown to reverse neuropathic allodynia by activating the TGF-β-related pathway in macrophages. Our study is the first to identify miR-21-5p as a novel target in CPSP and to explore its role in the analgesic effects of EA through both overexpression and knockdown approaches. First, our results confirmed that inhibiting miR-21-5p expression significantly increased mechanical and thermal pain thresholds in CPSP rats, while concurrently reducing neuronal damage and neuroinflammation. Conversely, miR-21-5p overexpression produced the opposite effects. Furthermore, EA significantly suppressed miR-21-5p expression, whereas its upregulation markedly attenuated the analgesic effects of EA on CPSP. These findings support our hypothesis that the ability of EA to alleviate CPSP is closely associated with the inhibition of miR-21-5p. Meanwhile, several studies have suggested a negative regulatory relationship between miR-21-5p and Smad7. In lung cancer cells, miR-21-5p has been shown to promote cell proliferation, migration, and invasion by directly targeting Smad7. Additionally, findings indicate that exosomal miR-21 derived from multiple myeloma cells can induce renal epithelial-mesenchymal transition by modulating the TGF-β/Smad7 signaling pathway [38]. Our study revealed that miR-21-5p, by regulating Smad7, further suppresses the activation of the TGF-β/Smad signaling pathway. Smad7 belongs to the inhibitory Smads (I-Smads) subgroup, which is part of the Smad protein family. This family is classified into three groups: receptor-activated Smads (R-Smads), such as Smad2 and Smad3; common Smads (Co-Smads); and inhibitory Smads (I-Smads), including Smad7 [39]. A strong association between Smad7 and pain has been identified. For instance, halofuginone has been shown to alleviate pain induced by intervertebral disc degeneration by upregulating inhibitory Smad7, thereby suppressing collagen I production and inhibiting the activation of the transforming growth factor-β (TGF-β) signaling pathway [40]. Smad7 acts on the TGF-β1 receptor, regulating the activity of the TGF-β/Smad pathway by blocking the phosphorylation and nuclear import of Smad2 and Smad3, while also inducing the dephosphorylation of Smad2 and Smad3, to inhibit inflammatory responses and alleviate aberrant neuronal activation, thus relieving pain [41, 42]. Moreover, our study revealed a significant reduction in the expression of Smad7, a key target of miR-21-5p, following the occurrence of CPSP. First, our results confirmed that miR-21-5p can indeed negatively regulate the expression of Smad7 through in vivo and in vitro experiments. Then, both the overexpression of Smad7 and the suppression of Smad7 by electroacupuncture intervention confirmed the significant benefits of upregulating Smad7 in CPSP. Besides, TGF-β1, p-Smad2, and p-Smad3 expression levels were significantly increased in the CPSP rat model, and EA inhibited miR-21-5p and upregulated Smad7 expression, which may relieve pain by regulating the TGF-β/Smads signaling pathway. Our study is the first to identify miR-21-5p as a key target in the onset and progression of CPSP. Furthermore, we established a negative regulatory relationship between miR-21-5p and Smad7. Specifically, miR-21-5p directly targets and suppresses Smad7, thereby modulating the TGF-β/Smad signaling pathway. Electroacupuncture, by inhibiting miR-21-5p and consequently upregulating Smad7 expression, further modulates the TGF-β/Smads signaling pathway, thereby exerting its analgesic effects. This provides significant insights into our understanding of the mechanisms underlying acupuncture-induced analgesia. However, our study has some limitations. Further research is needed to explore in depth electroacupuncture regulates the development of CPSP through various physiological and molecular pathways. Several limitations in the present study warrant consideration. First, we employed a collagenase-induced thalamic hemorrhagic stroke lesion model in the VPL to minimize motor confounds; however, future validation in ischemic paradigms is requisite to establish generalizability across diverse stroke etiologies. Second, while FISH assays localized miR-21-5p predominantly to neurons, we cannot exclude potential glial contributions to the observed phenotypes, necessitating future cell-specific resolution via single-cell approaches. Third, although our findings suggest that EA-mediated suppression of neuroinflammation likely acts as an upstream mechanism to mitigate downstream GABAergic disinhibition and central sensitization, we did not directly quantify neuronal excitability or synaptic transmission. Finally, owing to the high technical complexity and costs associated with stereotactic modeling across numerous experimental cohorts, the sample size per individual group was relatively modest, despite the substantial total number of animals utilized in the study. Notwithstanding these limitations, the present study provides experimental evidence implicating the miR-21-5p/Smad7/TGF-β axis in the analgesic effects of EA, thereby enriching our understanding of the molecular mechanisms underlying CPSP management. Conclusion In conclusion, our study elucidates electroacupuncture reduces central post-stroke pain by modulating miR-21-5p/Smad7 expression, providing a theoretical basis for acupuncture’s analgesic mechanism. Following brain injury, miR-21-5p expression is upregulated in rat cerebral tissue, leading to the suppression of Smad7 and the activation of the TGF-β/Smad signaling pathway. This cascade leads to rapid neuroglial cell activation, triggering neuroinflammation, heightened allodynia. Conversely, electroacupuncture reduces miR-21-5p levels while enhancing Smad7 expression, effectively inhibiting the TGF-β/Smad signaling pathway. This modulation alleviates neuronal damage, suppresses neuroglial activation, and reduces neuroinflammation, culminating in a significant analgesic effect. These findings provide new insights into the underlying analgesic mechanisms of acupuncture. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1: sFig 1: The expression of miR-21-5p and the construction and validation of the miR-21-5p knock down (KD) and overexpression (OE) virus. A. Schematic diagram of the viral vector construction and injection of miR-21-5p knock out and overexpression (KO represents miR-21-5p knock out; OE represents miR-21-5p overexpression). B-D. The expression of miR-21-5p knock down was confirmed by immunofluorescence image (B and D, n = 3) and RT-PCR (D, n = 4). E-G. The expression of miR-21-5p OE was confirmed by immunofluorescence image (E and F, n = 3) and RT-PCR (G, n = 4). Scale bars represent 100 μm as indicated Acknowledgements We thank Pro. Shihong Zhang (Zhejiang University) for revising the paper. Author contributions Guihua Tian and Huifeng Hao are responsible for the conception and design, Yang Wu, Zhihao Shang, Ke He and Xinyi Li were responsible for writing the initial form of the manuscript and its editing. Yang Wu, Xinyi Li, Meiyue Wang and Nan An contributed to the experiments. Meiyue Wang, Yang Wu, Ke He, Su, Luyu Yang, Xinyi Li and Guihua Tian performed the data analysis. All authors contributed to data interpretation and the preparation of figures and tables. Additionally, all authors participated in drafting the manuscript, reviewed its content, and provided final approval for the submitted version. Funding This work was supported by the National Natural Science Foundation of China (81973623, 82574753). Data availability The datasets used and analyzed in this study are available from the corresponding author upon reasonable request. Declarations Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Guihua Tian, Yang Wu and Zhihao Shang contributed equally to this work. Contributor Information Guihua Tian, Email: rosetgh@163.com. Xinyi Li, Email: lxinyilxy@163.com. References Associated Data Supplementary Materials Supplementary Material 1: sFig 1: The expression of miR-21-5p and the construction and validation of the miR-21-5p knock down (KD) and overexpression (OE) virus. A. Schematic diagram of the viral vector construction and injection of miR-21-5p knock out and overexpression (KO represents miR-21-5p knock out; OE represents miR-21-5p overexpression). B-D. The expression of miR-21-5p knock down was confirmed by immunofluorescence image (B and D, n = 3) and RT-PCR (D, n = 4). E-G. The expression of miR-21-5p OE was confirmed by immunofluorescence image (E and F, n = 3) and RT-PCR (G, n = 4). Scale bars represent 100 μm as indicated Data Availability Statement The datasets used and analyzed in this study are available from the corresponding author upon reasonable request.