2.1 Animals

Healthy male Sprague Dawley rats weighing 180–220 g were purchased from Beijing Vital River Laboratory Animal Technology Co. Ltd. All experiments adhered to the guidelines of the Committee for Research and Ethical Issues of the International Association for the Study of Pain (IASP) [29]. This research was approved by the Ethics Committee of Beijing University of Chinese Medicine, Ethics Approval No: BUCM-2024011502-1031. Rats were raised in a controlled environment (temperature 18°C–22°C, humidity 40%–70%) with a 12 h light–dark cycle, with water and food available ad libitum. The animals were habituated to the environment for 7 days before all experiments.

2.2 Establishment of spared nerve injury (SNI) model

The SNI model was created using a procedure similar to that described by Decosterd and Woolf [30]. Rats were anesthetized with 2% isoflurane in an oxygen/air mixture, and the left sciatic nerve was exposed. The common peroneal and tibial nerves were tightly ligated 8 mm distal to the sciatic trifurcation, and the nerve was transected distal to the ligation. To prevent regeneration, 2 mm of the distal stump of the cut nerve was removed. Careful attention was given to preserving the integrity of the sural nerve. Sham surgery rats underwent the same surgical procedure as the SNI rats, but the nerves were left intact. To maintain consistent experimental conditions, all surgical operations were performed by the same individual. Following surgery, the rats were placed in their familiar standard environment, where they had previously acclimated, and were provided with unrestricted access to food and water.

2.3 EA Intervention

During the EA stimulation, the rat was restrained but left in an awake state. EA was performed on the first day after SNI. Two stainless steel acupuncture needles (0.25 × 25 mm) were inserted at a depth of 8–10 mm into the left Huantiao (GB30) acupoint and 3–5 mm into the left Yanglingquan (GB34) acupoint. The electrical stimuli were administered using a HANS-200A Acupuncture Point Nerve Stimulator (Nanjing, China), which, with a frequency of 2 Hz and the EA stimulation intensity was increased in gradient by 1-2-3 mA, each stimulation intensity lasting for 10 min. The rats treated once every other day for 42 days. The acupoints in the SNI + sham EA group were the same as those in the SNI + EA group. Only superficial acupuncture (at a depth of 0.5–1 mm) was performed without EA stimulation. Rats in the Sham group and SNI group underwent grasping and fixation similar to those in the SNI + EA group but without EA stimulation.

2.4 Withdrawal Threshold to Mechanical Stimulation

The 50% mechanical paw withdrawal threshold (PWT) was determined using the “up and down” method [31]. In brief, animals were placed on a metal mesh floor covered with an inverted clear plastic cage (18 × 8 × 8 cm) and the plantar surface of the paw was stimulated with a series of ascending force von Frey monofilaments (Stoelting, Wood Dale, IL). Eight von Frey monofilaments with approximately equal logarithmic incremental (0.224) bending forces were chosen (0.40, 0.60, 1.00, 2.00, 4.00, 6.00, 8.00, and 15.00 g). Each trial started with a von Frey force of 2.00 g delivered perpendicularly to the lateral plantar surface of the left hindpaw. An abrupt withdrawal of the foot during stimulation or immediately after the removal of the filament was recorded as a positive response. Whenever there was a positive or negative response, the next weaker or stronger filament was applied, respectively. This procedure was done until six stimuli after the first change in response had been observed. The 50% PWT was calculated using the following formula: 50% PWT (g) =  (10^X_f+^κδ)/10⁴ where X_f is the value of the final von Frey filament used (in log units), κ is a value measured from the pattern of positive/negative responses, and δ = 0.224, which is the average interval (in log units) between the von Frey filaments. The 50% PWT of an allodynic rat is defined as 4.0 g, that is, non-noxious tactile stimulus induce the withdrawal response [32].

2.5 Spontaneous Pain Behavior

To assess spontaneous pain behavior, animals were placed in plexiglass containers and recorded for 60 min. The recorded video was then played back in slow motion to accurately count the number of instances of licking and flinching/shaking/lifting with the ipsilateral hind paw. Movements related to exploration, locomotion, body repositioning, and grooming were not considered. Each instance of licking was assigned two points, while each instance of flinching/shaking/lifting was assigned one point. The total points accumulated during the 1-h recording were used as the spontaneous pain score [33, 34].

2.6 Stereotaxic Injection

Cre-dependent adeno-associated viruses (AAVs) were used to manipulate neural activity in the NTS^Glu-RVLM circuit in this study. AAV injections and cranial window implantations were performed as previously described [35, 36]. The rats were anesthetized with 4% isoflurane, maintained at 2% isoflurane, and positioned in a stereotaxic apparatus (RWD, Shenzhen, China). The skull was exposed via a small incision, and two small holes were drilled (0.50 mm drill bit) into the skull to introduce a microinjection glass pipette into the NTS (AP, −12.75 mm; ML, −1 mm; DV, −6.5 mm) in a volume of 200 nL for each side at 1 nL/s and into the RVLM (AP, −12.25 mm; ML, −2 mm; DV, −7.8 mm) in a volume of 200 nL at 1 nL/s. The syringe was slowly retracted after an additional 10 min diffusion of the virus. The skin was sutured, and the animals were allowed to recover in prewarmed cages before returning to the home cage. For NTS^Glu-RVM circuit tracing, a viral approach was performed to label the NTS glutamatergic neurons projecting to the RVLM: we injected rAAV-CaMKIIα-mCherry into the NTS.

In order to detect the role of the circuit in EA analgesia, we used chemogenetic manipulation. Like other reports, it indicated that ipsilateral EA stimulation could alleviate knee osteoarthritis (KOA) pain behavior by regulation of ipsilateral sympathetic [37], in Figure 1, the results also showed that ipsilateral EA stimulation could alleviate pain behavior and ipsilateral sympathetic activity. Besides, as an important area for regulation of sympathetic outflow, there were reports that indicated that RVLM had predominantly ipsilateral innervation of sympathetic nerves [38, 39], so we injected rAAV-DIO-hM3Dq/hM4Di-mCherry or rAAV-DIO-mcherry into the left NTS and Retro-rAAV-CaMKIIα-Cre virus into the left RVLM, to specifically label the NTS^Glu-RVLM neural circuit, and intraperitoneal injection of CNO (1 mg/kg) half an hour before EA stimulation to activate or inhibit the neural activities. All viruses were purchased from Brain Case (Wuhan, China).

2.7 Extracellular Single-Unit Recordings

Rats were deeply anesthetized with 2% isoflurane in an oxygen/air mixture and placed in a stereotaxic apparatus (RWD, Shenzhen, China). According to the above method, a craniotomy was performed to expose the brain above the hypothalamic NTS and RVLM. Then, an aseptic surgical procedure was performed to implant a single electrode (resistance: 1 M) (KD-SC, KedouBC), with a 100 μm inner electrode diameter and a 50 mm interelectrode distance, into a target cortical area. Slowly advance the electrode within the target brain area to look for neurons with stable discharge. The frequency of the neuronal discharge was analyzed with special software (SpikeHistogram, AD Instruments). Impair the electrode (1 mA, 10 s) to mark the site of the last recorded neuron [40].

2.8 Immunohistochemistry

Forty-two days after the SNI operation, rats were anesthetized with 2% isoflurane and subjected to intracardiac perfusion with 0.9% saline followed by 4% paraformaldehyde (PFA). The L4–L6 DRGs and the brain were postfixed in 4% PFA for 6–8 h. The tissues were subsequently removed and dehydrated in 20% and 30% sucrose successively until they completely sunk to the bottom of the container. Sections were cut at 20 μm for DRGs and brain on a cryostat (CM1950, Leica). The sections were blocked in 5% normal donkey serum and 0.2% Tween 20 in phosphate buffer (PB) for 1 h at 37°C After blocking, the DRG sections were incubated overnight with the primary antibody, which includes mouse antineurfilament-200 (anti-NF200) (1:400, 2836S, Cell Signaling Technology), rabbit anticalcitonin gene-related peptide (CGRP) (1:400, #14959, Cell Signaling Technology), FITC-conjugated isolectin B4 (IB4) (1:400, L2140, Sigma-Aldrich) to label sensory neurons in the DRGs, and sheep antityrosine hydroxylase (anti-TH) (1:1000, PA1-4679, Invitrogen) to visualize sympathetic fibers in the DRGs. The brain sections were incubated overnight with the primary antibody rabbit anti-c-Fos (1:200, 2250S, Cell Signaling Technology) to label activated neurons in the NTS and RVLM, mouse anticalcium-CaM-dependent protein kinase II (CaMK II) (1:200, 66843-1-Ig, Proteintech), and mouse anti-tyrosine hydroxylase (TH) (1:100, 66843-1-Ig, Proteintech) to label glutamatergic neurons in the NTS and presympathetic neurons in the RVLM, respectively. After overnight incubation with primary antibodies at 4°C, the slides were washed three times in PB and incubated with appropriate Alexa Fluor-conjugated secondary antibodies (1:1000, Jackson, USA) for 1 h at room temperature. Then sections were counterstained with the nuclear marker DAPI (100 ng/mL, Sigma-Aldrich) and captured with a confocal imaging system (FV1200, Olympus, Tokyo, Japan).

2.9 Sympathetic Nerve Activity Recording

To assess sympathetic nerve activity in SNI rats, L4–L5 sympathetic nerve trunks were dissociated and recorded using a platinum wire electrode (50 μm). Rats were anesthetized with 10% urethane (1.2 mg/kg, ip) and fixed on the operating table in a supine position. An incision was made along the midline of the abdomen to expose the abdominal aorta. Then the operation was switched to a microscope, and a glass minute needle was used to trace the abdominal aorta longitudinally. The retroperitoneum was separated and mobilized to both sides to expose the left lumbar sympathetic nerve. A plastic wrap was used to isolate the surrounding tissue under the nerve, and paraffin oil was dropped to keep the nerve moist. Fiber tweezers were used to peel off the mucosa and epineurium about 2 mm long on the nerve surface. The electrode was inserted into the sympathetic nerve at about 30°. The signal was amplified by a preamplifier (AM-1800; A–M Systems, Sequim, WA, USA) and collected through the Powerlab data acquisition system using LabChart 8.0 software (ADInstruments, NSW, Australia). Spike frequency was analyzed using spike histograms of LabChart 8.0 (gain, 1000 Hz; low-cut filter, 100 Hz; high-cut filter, 1000 Hz) [41].

2.10 Statistical Analysis

Statistical analyses were performed with IBM SPSS Statistics 21.0 and GraphPad Prism 8.0 for Windows (GraphPad Software). Shapiro–Wilk tests were used to assess normality in the distribution (Gaussian distribution) for each group. For normally distributed data, One-way analysis of variance (ANOVA) with the Tukey method (equal variance) or Tukey–Kramer method (equal variance, unequal number of cases between groups) or Dunnett's T3 method (unequal variance) was used. Additionally, Kruskal–Wallis was employed to compare non-normally distributed data. All data are reported as mean ± SEM, and differences with p < 0.05 were considered statistically significant.

3.1 EA Improved Pain Behaviors and Inhibited Sympathetic Nerve Activity in SNI Rats

To investigate the impact of EA stimulation on pain behaviors in rats suffering from SNI, rats in the SNI + EA group received 2 Hz EA treatment every other day for 42 days post-SNI surgery (Figure 1A). The positioning of acupoints is illustrated in Figure 1B. PWT and spontaneous pain scores were assessed. Results showed a significant decrease in PWT (p < 0.05; Figure 1C) and a notable increase in spontaneous pain scores (p < 0.05; Figure 1D) on the first day post-SNI, which persisted throughout the modeling period, indicating the successful establishment of SNI-induced neuropathic pain. EA treatment effectively raised PWT and decreased spontaneous pain scores (p < 0.05) from day 3 to 42 after surgery (Figure 1C,D), demonstrating that 2 Hz EA stimulation can alleviate mechanical and spontaneous pain induced by SNI. At the same time, using electrophysiological recording techniques to analyze the L4–L5 sympathetic nerves excitability (Figure 1E), we found that the sympathetic nerve excitability was enhanced in SNI rats, and EA treatment could reverse the increase induced by the SNI operation (Figure 1F,G). Sympathetic sprouting also occurred during neuropathic pain conditions in many researches [42, 43]. In our studies, we found the same results as previously described. In SNI rats, obvious sympathetic-sensory coupling was observed up to 6 weeks after the SNI operation (Figure S1A), 42 days after the SNI operation, the expression of TH was notably increased, and EA treatment significantly decreased the expression of TH (Figure 1H,I). Also, as described in other reports TH⁺ nerve fibers predominantly encircled NF200⁺ neurons, rather than CGRP⁺ and IB4⁺ neurons (Figure S1B). The results suggested that the sympathetic nerve might be involved in EA analgesia.

3.2 EA Inhibited NTS and RVLM Neuronal Activities

Previous researches reported that NTS and RVLM regulated the sympathetic nerve activity [22]. In order to assess the role of NTS and RVLM on sympathetic in EA analgesia, extracellular discharges in NTS and RVLM were analyzed. Results indicated a notable increase in discharge frequency in the NTS (p < 0.05; Figure 2A,B) and RVLM (p < 0.05; Figure 2C,D) of SNI rats. EA treatment led to a significant reduction in extracellular discharge frequency in these two brain regions (p < 0.05; Figure 2A–D).

The neurons activation of NTS (Figure 2E) and RVLM (Figure 2F) was also assessed by using immunofluorescence, which yielded the same results (Figure 2G~J). EA significantly reduced the expression of c-Fos in the NTS (p < 0.05; Figure 2G,I) and RVLM (p < 0.05; Figure 2H,J). The above results showed that the activity of NTS and RVLM was the same as sympathetic nerve in EA analgesia, suggesting that NTS and RVLM played an important role in sympathetic regulation and EA analgesia.

3.3 Glutamatergic Neurons in the NTS and Presympathetic Neurons in the RVLM Were Activated

Neurons in the NTS and RVLM could classify into different types. After phenylephrine-induced hypertension, 72% of Fos-ir neurons were glutamatergic and 26% were GABAergic in rat NTS [44]. To identify the type of activated neurons in NTS and RVLM, immunofluorescence staining was selected. We found that in NTS, the c-Fos⁺ neurons mainly coexpressed the glutamatergic neurons (Figure 3A,C), and in RVLM, the c-Fos⁺ neurons mainly coexpressed presympathetic neurons (Figure 3B,D). The results further suggested that glutamatergic neurons in NTS and presympathetic neurons in RVLM might be involved in the regulation of sympathetic activity, as well as previous reports.

3.4 The NTS^Glu-RVLM Circuit Underlies the Effect of EA on Sympathetic Nerve Activity and Spontaneous Pain

To study the role of the NTS^Glu-RVLM circuit on the EA regulation, we first determined whether glutamatergic neurons in the NTS project to the RVLM by injecting anterograde tracer virus rAAV-CaMKIIα-mCherry into the NTS in healthy male SD rats (Figure 4A). Successful injection was confirmed by localized expression of mCherry after 3 weeks (Figure 4B). The results indicated that CaMKII-positive nerve fibers were observed in the RVLM (Figure 4C). These findings suggest a direct projection of NTS glutamatergic neurons to the RVLM. Subsequently, a chemogenetics virus was injected into the NTS and RVLM to manipulate the NTS^Glu-RVLM circuit (Figure 4D). The schematic diagram of the experimental process is shown in Figure 4E. The results showed that chemogenetic inhibition of the NTS glutamatergic neurons produced the same effect as EA, both of which could reduce extracellular discharge frequency in the RVLM (Figure 4F,G) and spontaneous pain scores (Figure 4H). In contrast, chemogenetic activation of the NTS glutamatergic neurons negated the effect of EA (Figure 4F–H). However, inhibition of the NTS^Glu-RVLM circuit had no effect on mechanical pain, and activation of the NTS^Glu-RVLM circuit also could not negate the effect of EA on mechanical pain (Figure 4I). Based on these findings, it is speculated that EA exerts an inhibitory effect on the NTS^Glu-RVLM neural circuit to modulate spontaneous pain but not mechanical pain.

In order to test whether the sympathetic nerve was involved in this discrepancy mediated by the NTS^Glu-RVLM circuit, sympathetic nerve activity and sprouting were analyzed under chemogenetic modulation and EA treatment. We found that chemogenetic inhibition of the NTS^Glu-RVLM circuit significantly inhibited the excitability (Figure 5A,B) and sprouting (Figure 5C,D) of sympathetic nerves, just like the effect of EA. Chemogenetic activation of the NTS^Glu-RVLM circuit could reverse the effect of EA on sympathetic nerves (Figure 5A–D). The results showed that the NTS^Glu-RVLM circuit contributed to the regulation of sympathetic nerves in EA treatment and might be the reason that NTS^Glu-RVLM circuit modulation only affected spontaneous but not mechanical pain.