Electroacupuncture ameliorates Autism Spectrum Disorder via modulating the gut-brain axis depending on the integrity of vagus nerve

Abstract

Autism spectrum disorder (ASD) is a neurodevelopmental disease characterized by behavioral and neurological abnormalities. Numerous pieces of evidence indicate a strong association between ASD and neuroinflammation mediated by gut microbiota and microglial activation. Previous studies have shown that the therapeutic effects of an acupuncture protocol targeting the bacteria-gut-brain axis in a well-established ASD mouse model induced by prenatal exposure to valproic acid (VPA). We demonstrated that electroacupuncture significantly alleviates behavioral symptoms in VPA model. However, the precise mechanisms remain insufficiently elucidated. In this study, we confirmed that electroacupuncture markedly improved behavioral symptoms in ASD mice. We conducted gut microbiota transplantation from electroacupuncture-treated mice to untreated ASD mice, improving behavioral outcomes in untreated ASD mice. Conversely, by transplanting gut microbiota from ASD mice into electroacupuncture-treated mice, we successfully mitigated the beneficial behavioral effects of acupuncture. We analyzed inflammatory markers in the microglial activation from cerebral cortex and hippocampus tissues, revealing that acupuncture exerts robust anti-neuroinflammatory effects in ASD mice. To further validate the mechanism, we performed vagotomy in ASD mice, which abolished the therapeutic benefits of acupuncture. Our findings establish that the behavioral improvements observed in ASD mice are intricately linked to the diversity and abundance of gut microbiota. Furthermore, regulatory effects of electroacupuncture on ASD behaviors are mediated via bacteria-gut-brain axis, dependent on intact vagus nerve signaling. This study provides compelling evidence for the potential of acupuncture to modulate central neuroinflammation through vagus nerve-mediated gut microbiota regulation, offering novel avenue into its therapeutic application for neurodevelopmental disorders such as ASD.

Introduction

Autism spectrum disorder (ASD) is a neurodevelopmental condition characterized by core symptoms including deficits in social communication, repetitive behaviors, and restricted interests. Despite its profound impact, there are currently no specific pharmacological treatments that effectively address the disorder, necessitating long-term rehabilitation efforts. The prognosis for many children with ASD remains poor, imposing substantial emotional and economic burdens on affected families and society. Recent epidemiological data indicate a marked increase in the global prevalence of ASD. In the United States, approximately 1 in 36 children has been diagnosed with ASD, according to the latest report [1]. Similarly, studies from China estimate the prevalence of ASD among children to range from 0.7% to 1% [2]. This upward trend in diagnosis highlights the growing societal and clinical attention directed toward ASD.

Neuroinflammation is a classical theory to understand the neurodevelopmental abnormalities and disease progression observed in children with ASD. Elevated levels of pro-inflammatory cytokines, such as IFN-γ (interferon-gamma), IL-6 (interleukin-6), IL-17a (interleukin-17a), and TNF-α (tumor necrosis factor-alpha), have been consistently identified in the plasma of individuals with ASD. Notably, these elevations are positively correlated with core symptoms of ASD, particularly social communication deficits [3]. Furthermore, various pro-inflammatory cytokines have been detected within the brain tissue of ASD patients, both in animal models and postmortem human samples, suggesting that neuroinflammation plays a critical role in the onset and progression of ASD [4, 5]. Microglia, as primary immune cells within the central nervous system (CNS), are pivotal in activating immune responses and signaling inflammation in the brain. Both microglia and astrocytes contribute to the pathophysiological alterations associated with ASD development [6]. Acupuncture has been demonstrated to mitigate neuroinflammation in various neurological conditions, including Parkinson’s disease, stroke, cognitive impairments, and depression [7–10]. These anti-inflammatory effects are thought to be mediated, at least in part, through the upregulation of brain-derived neurotrophic factor (BDNF), thereby exerting neuroprotective effects [11, 12].

In recent decades, the gut microbiota has garnered significant attention as a critical regulator of CNS development and function, primarily through its involvement in the gut-brain axis. The vagus nerve (VN) plays a pivotal role in facilitating direct and indirect communication between the gut and the brain, acting as a major conduit for bidirectional signaling between the gut microbiota and the CNS [13, 14]. Peripheral signals originating from the gut are transmitted via afferent VN fibers to the dorsal vagal complex (DVC), a brainstem structure encompassing the nucleus of the solitary tract (NTS), the area postrema, and the dorsal motor nucleus of the vagus (DMV). The DVC is extensively interconnected with various brain regions, including the hypothalamus, thalamus, hippocampus, amygdala, medial prefrontal cortex, and other cortical areas, facilitating the integration of gut-derived information into broader neural networks. Notably, stimulation of the VN has been associated with activation in the hippocampus and medial prefrontal cortex, regions critical for cognitive and emotional processes [15]. These findings emphasize the VN’s central role in mediating gut-brain communication and its potential impact on CNS modulation.

Our clinical observations have demonstrated that acupuncture at gut-regulating acupoints, such as Zusanli (ST36) and Tianshu (ST25), significantly improves core symptoms of ASD, including verbal skills, imitation, and emotional expression in children. In addition to these therapeutic effects, acupuncture has been shown to enhance gut microbiota diversity and increase the abundance of beneficial bacterial populations, thereby contributing to the regulation of physiological functions [16–18]. The modulation of gastrointestinal function and the associated gut microbiota by acupuncture is closely linked to the functionality and structural integrity of the VN. Previous research has highlighted the critical role of the VN in mediating the regulatory effects of acupuncture on gut physiology [19–21]. Notably, our earlier study confirmed that electroacupuncture at ST36 can ameliorate inflammatory responses and promote mucosal healing in a rat model of intestinal ischemia-reperfusion injury via the VN-associated cholinergic anti-inflammatory pathway [22]. These findings underscore the importance of the VN in acupuncture-mediated gut and systemic regulation, providing further insights into its therapeutic mechanisms.

Therefore, based on the observed regulatory effects of acupuncture on gut microbiota and its modulation through the VN, as well as the critical role of the VN in the gut-brain axis of ASD, we hypothesize that electroacupuncture exerts its therapeutic effects by regulating gut microbiota and alleviating central inflammation in ASD via a VN-mediated mechanism. To investigate, we established an ASD mouse model induced by valproic acid (VPA) and administered electroacupuncture treatment to the VPA-acupuncture group. Post-intervention, behavioral tests were conducted to assess the effects of acupuncture. Additionally, gut microbiota from electroacupuncture-treated mice was transferred to untreated ASD mice to evaluate the role of microbiota in mediating the therapeutic effects of acupuncture. Using rescue experimental designs, we analyzed inflammatory markers in both central and peripheral tissues, as well as microglial activation in the cerebral cortex and hippocampus, in ASD mice with and without electroacupuncture treatment. Furthermore, to confirm the role of the VN, we performed truncal vagotomy in a subset of ASD mice to eliminate the effects of acupuncture. The findings from these experiments provide compelling evidence that electroacupuncture can mitigate CNS inflammation via VN-mediated modulation of gut microbiota. These results offer new perspectives on the mechanisms underlying the therapeutic effects of acupuncture in ASD and the pivotal role of the gut-brain axis.

Materials and methods

ASD mouse modeling

All animal experiments were conducted in accordance with ethical guidelines and were approved by the Nanjing University of Chinese Medicine under the experimental protocol number 012071001410 (application number 202208A033). Adult healthy SFP grade BALB/c mice, aged 9–12 weeks, were used in the study, obtained from the Weitong Lihua Laboratory Animal Technology Co., Ltd. in Beijing, China. Female mice weighing 16–20 g and male mice weighing 20–24 g were housed in standard laboratory conditions with four mice per cage. A total of 30 male and 60 female mice were provided with a standard diet and acclimated to the environment for 7 days prior to the experiments. Environmental conditions were maintained with a 12-h light-dark cycle, a constant temperature of 24 °C, and a relative humidity of 40–50%, with free access to food and water.

To establish the ASD mouse model, VPA was administered to induce prenatal exposure effects. Male and female mice were paired in a 2:1 ratio and co-housed overnight. The presence of a vaginal plug the following morning (observed at 8:00) was designated as embryonic day 0 (E0). On embryonic day 11 (E11), pregnant females received an intraperitoneal injection of VPA at a dose of 600 mg/kg (Med Chem Express, USA, HY-10585, concentration of 100 mg/mL). The day of birth of the male pups was designated as postnatal day 0 (P0), and weaning was performed on P21. Behavioral assessments, including the three-chamber social interaction test and the open field test, were conducted on P28 to confirm successful ASD model induction. Following model verification, ASD mice were subjected to electroacupuncture procedures. During these experiments, the mice were restrained and fixed in position to ensure stability, and electroacupuncture was applied only to the designated treatment groups. All ASD mice were randomly assigned to experimental groups to ensure unbiased results.

Acupuncture treatment

Following the successful establishment of the ASD mouse model, electroacupuncture treatment was administered. Acupoint selection and positioning were conducted according to standardized guidelines detailed in the Chinese national textbook *Experimental Acupuncture* [23]. Acupoint locations in mice were determined proportionally with reference to established rat acupoint maps. Specifically, the “Zusanli” (ST36) point was identified on the posterolateral aspect of the mouse’s knee joint, approximately 3 mm distal to the fibular head.

Mice were secured using a custom-made restraining device (Fig. 1A), and bilateral ST36 points were stimulated using disposable 32-gauge acupuncture needles, each 25 mm in length (Hwato Co, China). Needles were inserted to a depth of approximately 3 mm and connected to an electroacupuncture apparatus via custom-designed bipolar needles. This bipolar needle consisted of two acupuncture needles, each fitted with a polyethylene (PE50) tube approximately 10 mm long and 1 mm in diameter, ensuring insulation over the needle handles. The needles were bundled with plastic tape to expose the entire needle shaft while maintaining a 1 mm gap between them to facilitate electrical conductivity [24].

Fig. 1. Experimental Mice and Equipment.

A Custom-designed bipolar needles used in experiment. B Once sufficient anesthesia depth was achieved with isoflurane, the right vagus nerve trunk of the mouse was severed under a microscope.

Numerous pieces of evidence disperse-dense waveform at a frequency of 2/100 Hz (Hwato Co, SDZ-Ⅲ). The current intensity was adjusted to elicit slight twitching of the needle handle, which was well tolerated by the mice. Each treatment session lasted 20 min and was performed once daily for 6 consecutive days, followed by a 1-day rest interval (with a total of 18 sessions conducted). Control mice were derived from healthy adult BALB/c parents (6–8 weeks old) and co-housed without receiving any treatment interventions.

Fecal microbiota suspension and transplantation

During the fourth week of electroacupuncture intervention in the ASD model group, fecal samples were collected 2–4 h after defecation. The samples were processed by mixing with five volumes of sterile saline, followed by thorough stirring and filtration to produce a fecal suspension. The suspension underwent repeated centrifugation and washing to remove impurities. The resulting sediment, containing purified fecal microbiota, was resuspended in three volumes of sterile saline. The suspension was mixed gently to ensure homogeneity and prepared for use as the fecal microbiota transplantation (FMT) suspension.

A 2 mm diameter silicone enema tube was lubricated with liquid paraffin oil before insertion. Mice were positioned in a prone orientation, and their tails were lifted for proper alignment. The tube was gently inserted into the rectum to a depth of 4–5 cm. Mice were held in an inverted position as 0.5 mL of the fecal microbiota suspension was gradually administered. To ensure retention, the tube remained in place for 10 s before careful removal, and a cotton ball was used to apply gentle pressure to the anus. Mice were kept in an inverted position for 1 min before being returned to their cages.

The FMT procedure was conducted on alternating days, three times per week, over a period of two consecutive weeks. Tissue samples were collected for analysis during the fourth week following a two-week observation period.

Vagus nerve transection

Isoflurane anesthesia was administered, and body temperature was maintained at 37 °C using an infrared heating lamp. The depth of anesthesia was confirmed by the absence of the toe pinch reflex. Mice were then positioned appropriately for a unilateral cervical vagus nerve transection (Fig. 1B). The right vagal nerve trunk was carefully severed, as preliminary experiments indicated that bilateral vagus nerve transection frequently resulted in mortality.

Following transection, the incision was sutured, and mice were monitored until they fully regained consciousness. Once stable, they were returned to their cages for continued housing and observation.

Behavioral test

Three-chamber social interaction test

The experimental apparatus consisted of a rectangular polycarbonate box (40 cm × 60 cm × 22 cm) divided into three interconnected chambers (left, middle, and right). The chambers were separated by walls with movable doors, allowing controlled access. Following each trial, the apparatus was cleaned with a 70% ethanol solution and air-dried for 5 min to eliminate residual odors.

Prior to testing, the subject mouse was placed in the middle chamber to acclimate and freely explore the apparatus for 10 min. During the test phase, one chamber housed a conspecific mouse, while the other contained a novel object. The test mouse was again placed in the middle chamber, and its activity was recorded for 10 min using EthoVision 11.0 software. Behavioral metrics including time spent in each chamber and sniffing duration were analyzed. A preference index was calculated to quantify the subject’s interest in the conspecific mouse relative to the novel object.

Chamber preference index= (time in mouse chamber - time in object chamber) / (time in mouse chamber + time in object chamber)

Sniffing preference index= (time sniffing mouse - time sniffing object) / (time sniffing mouse + time sniffing object)

Open field test

The open field apparatus consisted of a square gray resin box measuring 40 cm × 40 cm × 30 cm. The floor of the box was divided into 16 equal squares, with the central 4 squares designated as the center area.

During the test, the mouse was placed at the center of the apparatus, and its movements were recorded over a 30-min period. Behavioral data, including the total distance traveled and the time spent in the center area, were analyzed using EthoVision 11.0 software.

Immunofluorescence double-Labeling for microglial activation and quantitative morphological assessment

Following perfusion, brain tissues, including the cortex and hippocampus, were dissected and processed for histological analysis. The brain tissue was immersed in 4% paraformaldehyde for 24 h at 4 °C. Subsequently, it was dehydrated in 30% sucrose for 5 days and frozen in OCT at −20 °C for 2 h. The tissue was then sectioned into 20-μm-thick slices. These slices were blocked with Sea BLOCK blocking buffer (Thermo Fisher Scientific, 37527) containing 0.2% Triton X-100 (Sigma, T8787) for 1.5 h. Following the blocking step, the slices were subjected to overnight incubation with anti-Iba1 antibody (Abcam,ab178846,1:200) at 4 °C, followed by a one-hour incubation with Alexa Fluor 594-conjugated secondary antibodies (Thermo Fisher Scientific, A11005,1:300) at 37 °C. DAPI staining (Absin, abs47047616) was performed for 10 min at 37 °C. The expression of cells was captured using a Thunder Imager model microscope (Leica Microsystems Srl, Buccinasco, Italy) and evaluated using Image-Pro Plus 6.0 software. To detect microglial cells, three brain tissue sections at 200x magnification were taken per mouse and were photographed by pathologists who were blinded to the animal grouping.

A morphological quantitative analysis of the immunofluorescence images was performed to assess the immune activity of microglial cells in mice after different treatments. Using Fiji software, we quantified multiple cells by converting fluorescence microscopy images into representative binary formats [25]. This method allowed us to capture and quantify cellular morphological changes, including descriptors of microglial cell shape and size per unit area. For a more precise analysis, manual counting was carried out to determine the total number of microglial cells in different hippocampal regions. Furthermore, Fiji software was employed to calculate the area of each microglial cell and the corresponding cell body area. Images were analyzed by a blinded investigator.

Enzyme-linked immunosorbent assay (ELISA)

In a 96-well polystyrene microplate, 50 μl of standard or sample was added. Then, 100 μl of HRP-Conjugate reagent was added and incubated at 37 °C for 60 min. Subsequently, the plate was washed five times with washing solution for one minute each. After that, 50 μl of chromogenic substrate and Tetramethylbenzidine were added and incubated in darkness at 37 °C for 15 min. Finally, the OD values of each well were measured at a wavelength of 450 nm.The serum sample, hippocampal tissue sample and cortical sample were added to detect the expressions of IL-6, TNF-α, and IL-17A. The above reagents are sourced from Nanjing Jinyibai Biotechnology Institute.

Gut microbiota detection

Total genome DNA from samples was extracted using CTAB method. 16S rRNA genes of distinct regions (16S V3-V4) were amplified used specific primers 338 F and 806 R with the barcode. PCR products were mixed in equidensity ratios. Then, mixture PCR products were purified with Qiagen Gel Extraction Kit(Qiagen, Germany). Sequencing libraries were generated using TruSeq® DNA PCR-Free Sample Preparation Kit (Illumina, USA) following manufacturer’s recommendations and index codes were added. At last, the library was sequenced on an Illumina NovaSeq platform and 250 bp paired-end reads were generated.

The datasets were analyzed using vsearch. The sequences were clustered into operational taxonomic units (OTUs) at a similarity level of 97%. The Ribosomal Database Project (RDP) Classifier tool was applied to classify all sequences into different taxonomic groups. Clustering analyses and PCA were used based on OTU information from each sample using R.16.

Statistical analysis

Statistical analyses were performed using IBM SPSS Statistics 26 (IBM Corp., Armonk, NY, USA). Continuous data are presented graphically as box plots overlaid with scatter points (representing individual observations) in relevant figures. Box plots depict the median and interquartile range (IQR), while scatter points illustrate raw data distributions. All visualizations were generated using R software (version 4.3.1; R Foundation for Statistical Computing, Vienna, Austria).

Prior to inferential analyses, data assumptions were verified through normality and homogeneity of variance testing. Parametric analyses including independent samples T-tests, one-way ANOVA, and two-way ANOVA were employed as appropriate. For significant omnibus effects, post hoc pairwise comparisons were conducted using Tukey’s HSD method for homogeneous variances, Scheffé's procedure for complex contrasts, and Games-Howell test when heterogeneity of variance was detected. Statistical significance was defined as P-value < 0.05. The images and analyses are from three independent experiments except gut microbiota detection and ELISA in serum samples.

Results

Electroacupuncture at ST36 improves social deficits in ASD mice

The three-chamber social interaction test was used to assess the ASD behavior in mice. The test consists of two phases: social ability and social novelty preference.

In the social ability phase of the three-chamber test, the test mouse was placed in the central chamber, flanked by an empty chamber on one side and a chamber containing an unfamiliar mouse (Stranger 1) on the other (Fig. 2A). The number of entries into each chamber (Fig. 2B) and the duration of time spent in each chamber (Fig. 2C) were recorded to evaluate social preference.

Fig. 2. Electroacupuncture at ST36 Improves Social Deficits in ASD Mice.

A Schematic of the first phase of the social ability test. An unfamiliar mouse (Stranger 1) was placed in one chamber, and an empty chamber was placed on the other side. The test mouse was placed in the middle chamber to observe its behavior. In this test, we compared the number of entrance and the time spent in chambers between Control group (n = 14), ASD Group(n = 30) and ASD + EA Group (n = 10). B Bar graph showing the number of entries into the empty chamber and the chamber containing the unfamiliar mouse. Two-way ANOVA was used, followed by Games-Howell test for post hoc pairwise comparisons. Results demonstrated a significant interaction effect between mouse groups (F = 19.963, P < 0.001). The main effect of empty/stranger chamber was not statistically significant (F = 3.661, P = 0.059). The interaction between empty/stranger chamber and mice groups had a statistically significant effect on the experimental results (F = 31.832, P < 0.001). Further interaction analysis revealed that P > 0.05 under the empty chamber condition, while P < 0.05 under the stranger chamber condition, indicating that significant differences were observed only in the comparison among the three groups in the stranger chamber condition. C Bar graph showing the time spent interacting with Stranger. Two-way ANOVA was used, followed by Games-Howell test for post hoc pairwise comparisons. The result shows an effect of mice groups (F = 54.16, P < 0.001)、empty/stranger chamber (F = 34.73, P < 0.001) and of empty/stranger chamber x mice groups interaction (F = 68.50, P < 0.001). D Schematic of the second phase of the social novelty preference test. A familiar mouse (Stranger 1) was placed in one chamber, and a new unfamiliar mouse (Stranger 2) was placed in the other chamber. The test mouse was placed in the middle chamber. In this test, we compared the number of entrance and the time spent in chambers between Control group (n = 14), ASD Group(n = 30) and ASD + EA Group (n = 10). E Bar graph showing the number of entries into the chambers containing the familiar and unfamiliar mice. Two-way ANOVA was used, followed by Scheffé's procedure for post hoc pairwise comparisons. The result shows an effect of mice groups (F = 21.851, P < 0.001)、stranger/familiar chamber (F = 16.923, P < 0.001) and of stranger/familiar chamber x mice groups interaction (F = 7.708, P < 0.001). F Bar graph showing the time spent interacting with the familiar and unfamiliar mice. The images and analyses shown are from three experiments. Two-way ANOVA was used, followed by Scheffé's procedure for post hoc pairwise comparisons. The result shows an effect of mice groups (F = 37.233, P < 0.001)、stranger/familiar chamber (F = 9.762, P = 0.02) and of stranger/familiar chamber x mice groups interaction (F = 18.142, P < 0.001). The median is shown as a straight line and the box denotes the interquartile range. *P < 0.05, **P < 0.001, ***P < 0.001, NS: no significance.

The results demonstrated that ASD model mice entered the empty chamber more frequently compared to control mice, reflecting potential deficits in social behavior. However, mice in the electroacupuncture treatment group (ASD + EA group) exhibited increased interaction with Stranger 1, resembling the behavior of Control group. These findings suggest that electroacupuncture intervention may ameliorate social deficits in ASD model mice.

In the social novelty preference phase, a second unfamiliar mouse (Stranger 2) was introduced into the previously empty chamber, while the Stranger 1 remained in its original chamber and had become familiar to the test mouse (Fig. 2D). The number of entries into each chamber (Fig. 2E) and the time spent in each chamber (Fig. 2F) were recorded to assess preference for social novelty.

Normal mice displayed a clear preference for interacting with the Stranger 2 over the Stranger 1, indicating a natural inclination toward social novelty. In contrast, ASD model mice exhibited no significant difference in interaction between the strangers, suggesting impaired social novelty recognition. Notably, ASD model mice treated with electroacupuncture demonstrated a preference for the unfamiliar mouse, resembling the behavior of control mice. These results indicate that electroacupuncture may restore social novelty preference in ASD model mice.

These findings indicate that ASD model mice display significant social deficits. Moreover, electroacupuncture can effectively ameliorate these social impairments, as a therapeutic strategy for addressing social behavioral deficits in ASD.

Electroacupuncture at ST36 alleviates anxiety-like and depression-like behaviors in ASD mice

The open field test was employed to assess anxiety-like behaviors in mice by analyzing spontaneous activity within the central zone. Parameters measured included the total time spent and the total distance traveled in the central zone, with movement trajectories recorded for further analysis (Fig. 3A). Compared to the Control group, ASD model mice exhibited significantly reduced time spent and distance traveled in the central zone (Fig. 3B, C, both P < 0.05), indicative of heightened anxiety-like behaviors.

Fig. 3. Electroacupuncture at ST36 Alleviates Anxiety-like and Depression-like Behaviors in ASD Mice.

A The open field test setup and typical movement trajectories of mice from different groups. B, C Time spent by mice in the central area of the open field and distance traveled by mice in the central area of the open field between Control group (n = 12), ASD Group(n = 13) and ASD + EA Group (n = 8). One-way ANOVA was used, followed by Tukey’s HSD method for post hoc pairwise comparisons. The median is shown as a straight line and the box denotes the interquartile range. *P < 0.05, **P < 0.001, ***P < 0.001, NS: no significance.

Following electroacupuncture intervention, ASD mice demonstrated a significant increase in the distance traveled within the central zone (Fig. 3C, P < 0.01) and a trend toward increased time spent in this area (Fig. 3B, P > 0.05). These findings suggest that ASD model mice display anxiety-like and depression-like behaviors, which can be alleviated by electroacupuncture at ST36.

Electroacupuncture at ST36 can improve abnormal central and peripheral inflammatory states in ASD mice

Compared to the Control group, the ASD group exhibited significantly elevated TNF-α levels in hippocampal tissue (Fig. 4C, P < 0.05), alongside a non-significant upward trend in IL-6 and IL-17A levels (Fig. 4A, B, both P > 0.05). Following electroacupuncture intervention, TNF-α levels in the hippocampal tissue of ASD model mice were significantly reduced (Fig. 4C, P < 0.05), accompanied by a downward trend in IL-6 and IL-17A levels, though these changes did not reach statistical significance (Fig. 4A, B, both P > 0.05). These findings indicate the presence of an inflammatory response in the hippocampal tissue of ASD mice, which can be mitigated by electroacupuncture at ST36, highlighting its potential to alleviate hippocampal inflammation in this model.

Fig. 4. Electroacupuncture at ST36 improves abnormal CNS and peripheral inflammatory states in ASD mice.

A–C IL-6, IL-17, and TNF-α levels in hippocampal tissue samples of mice from different groups (n = 4). D–F IL-6, IL-17, and TNF-α levels in cerebral cortex tissue samples of mice from different groups (n = 4).. G–I IL-6, IL-17, and TNF-α levels in serum samples of mice from different groups (n = 4). One-way ANOVA was used, followed by Tukey’s HSD method (figure A-H) and Games-Howell test (figure I) for post hoc pairwise comparisons. The median is shown as a straight line and the box denotes the interquartile range. *P < 0.05, **P < 0.001, ***P < 0.001, NS: no significance.

Compared to the Control group, the ASD group displayed significantly elevated levels of IL-17A and TNF-α in cortical tissue (Fig. 4D, F, both P < 0.01), along with a non-significant upward trend in IL-6 levels (Fig. 4E, P > 0.05). Following electroacupuncture intervention, levels of IL-17A and TNF-α in cortical tissue were significantly reduced (Fig. 4D, F, both P < 0.01), while IL-6 levels showed a downward trend that did not reach statistical significance (Fig. 4E, P > 0.05). These findings suggest that inflammatory responses are present in the cortical tissue of ASD mice and that electroacupuncture at ST36 may attenuate cortical inflammation in this model.

The ASD group exhibited significantly elevated serum levels of IL-17A and TNF-α compared to the Control group (Fig. 4H, I, both P < 0.05), while IL-6 levels showed an upward trend that did not reach statistical significance (Fig. 4G, P > 0.05). Electroacupuncture intervention resulted in a significant reduction in serum TNF-α levels (Fig. 4I, P < 0.05), accompanied by downward trends in IL-6 and IL-17A levels (Fig. 4G, H, both P > 0.05). These findings indicate the presence of peripheral inflammatory responses in ASD mice, which may be ameliorated by electroacupuncture at ST36, suggesting its potential to regulate peripheral inflammation in this model.

Electroacupuncture at ST36 can regulate the overactivation of microglial cells in ASD mice

A quantitative analysis was conducted on the number of microglial cells and cell area within a defined region. Compared to the Control group, the ASD group showed a significant increase in the number of Iba1⁺ microglial cells in the hippocampal tissue (Fig. 5B), along with a marked enlargement in both the avg-territory area and avg-some cytoplasmic area (Fig. 5B). Following electroacupuncture intervention, the number of Iba1⁺ microglial cells in the hippocampal tissue of ASD model mice significantly decreased, accompanied by a reduction in the avg-territory area (Fig. 5B), while the avg-some cytoplasmic area showed a declining trend.

Fig. 5. Electroacupuncture at ST36 regulates the overactivation of microglial cells in ASD mice.

A Comparison of fluorescence intensity of surface markers on microglial cells in the hippocampus of mice from Control group, ASD group and EA group (n = 6). B Activated Iba1+ microglial cells and avg-territory area in the hippocampus of mice from Control group, ASD group and EA group (n = 6). One-way ANOVA was used, followed by Tukey’s HSD method (figure B-a, figure B-b) and Games-Howell test (figure B-c) for post hoc pairwise comparisons. Scale bars, 100μm. The median is shown as a straight line and the box denotes the interquartile range. *P < 0.05, **P < 0.001, ***P < 0.001, NS: no significance.

These findings indicate the presence of microglial overactivation in the hippocampus of ASD mice, which can be mitigated by electroacupuncture at ST36.

Differences in the composition of gut microbiota between ASD and electroacupuncture mice

To investigate the effects of electroacupuncture at ST36 on the gut microbiota of ASD mice, alpha diversity analysis was performed. The results demonstrated that ASD mice had lower Chao1 and Shannon indices compared to normal controls, indicating reduced microbial diversity. Following electroacupuncture treatment at ST36, these indices increased, approaching levels observed in normal mice (Fig. 6A, B). Principal component analysis (PCA) further revealed distinct clustering of microbial communities between ASD and normal mice, highlighting structural differences in the microbiota. Following electroacupuncture intervention, the microbial community structure in ASD mice shifted further from the ASD group, suggesting that the treatment significantly influenced gut microbiota composition (Fig. 6C).

Fig. 6. Analysis of Gut Microbiota Differences after acupuncture treatment.

A, B Chao1 index and Shannon index; Group Control: Normal group mice (n = 7); group ASD: ASD model group mice (n = 6); group EA: Electroacupuncture at ST36 group (n = 6). One-way ANOVA was used, followed by Tukey’s HSD method for post hoc pairwise comparisons. The median is shown as a straight line and the box denotes the interquartile range. *P < 0.05, **P < 0.001, ***P < 0.001, NS: no significance. C PCA analysis based on raw OTU data and Euclidean distance with 95% confidence ellipses. D Bar chart of species composition at the phylum level; E Bar chart of species composition at the genus level; F Evolutionary cladogram from Linear Discriminant Analysis.

To examine the species composition across groups, the RDP Classifier algorithm was applied to compare OTU representative sequences and annotate taxonomic information at the phylum and genus levels. At the phylum level, the microbiota in all groups was predominantly composed of p__Bacteroidetes, p__Firmicutes, and p__Proteobacteria(Fig. 6D). At the genus level, the microbiota in each group was mainly composed of g__Muribaculum, g__Alistipes, g__Ligilactobacillus, and g__Duncaniella (Fig. 6E).

To further analyze inter-group species differences, statistical analysis was conducted using the LDA Effect Size (LEfSe) method. Using ANOVA test with p < 0.05, Wilcoxon rank-sum test with p < 0.05, and an LDA score > 3 as differential selection criteria, 44 species with significant inter-group differences were identified. In the healthy group, the abundance of p__Bacteroidetes, c__Bacteroidia, o__Bacteroidales, f__Muribaculaceae, g__Muribaculum, s__Muribaculum_sp, g__Duncaniella, and s__Duncaniella_dubosii was significantly higher than in the ASD and EA groups. In the ASD group, the abundance of p__Candidatus_Saccharibacteria, p__Firmicutes, p__Bacteroidetes, p__Proteobacteria, c__Bacteroidia, o__Bacteroidales, f__Rikenellaceae, g__Alistipes, s__Alistipes_sp., c__Deltaproteobacteria, o__Desulfovibrionales, f__Desulfovibrionaceae, g__Desulfovibrio, and s__Desulfovibrio_sp. was significantly higher than in the Control and EA groups. In the EA group, the abundance of p__Bacteroidetes, c__Bacteroidia, o__Bacteroidales, f__Bacteroidaceae, g__Bacteroides, s__Bacteroides_rodentium, p__Verrucomicrobia, c__Verrucomicrobiae, o__Verrucomicrobiales, f__Akkermansiaceae, g__Akkermansia, s__Akkermansia_muciniphila, p__Firmicutes, c__Clostridia, o__Clostridiales, f__Ruminococcaceae, g__Pseudoflavonifractor, s__Pseudoflavonifractor_capillosus, p__Bacteroidetes, c__Bacteroidia, o__Bacteroidales, f__Prevotellaceae, g__Prevotellamassilia, and s__Prevotellamassilia timonensis was significantly higher than in the Control and ASD groups (Fig. 6F).

Regulatory effects of fecal microbiota transplantation from the electroacupuncture group on ASD mice

Fecal microbiota from the electroacupuncture group (EA group) was transplanted via enema into ASD mice (ASD + FMT(EA) group, FMT: fecal microbiota transplantation). The differences in behavioral outcomes, central and peripheral inflammatory responses, as well as hippocampal microglial activation, were assessed between ASD mice and ASD + FMT(EA) mice.

Behavioral assessments revealed significant differences in social interaction and anxiety-like behaviors between ASD mice and ASD + FMT(EA) mice. In the first phase of the three-chamber social test, ASD mice exhibited a preference for the empty chamber over social interaction (Fig. 7A, P < 0.05), whereas ASD + FMT(EA) mice displayed no significant preference, showing increased interaction with Stranger 1 (Fig. 7A, P > 0.05). In the second phase, a similar pattern was observed, with ASD + FMT(EA) mice spending significantly more time interacting with Stranger 2 compared to ASD mice (Fig. 7B, P < 0.05). Additionally, the number of interactions in this phase demonstrated an increasing trend. These findings suggest that fecal microbiota transplantation from the electroacupuncture-treated group ameliorates social deficits in ASD mice.

Fig. 7. Regulatory Effects of Fecal Microbiota Transplantation from the Electroacupuncture Group on ASD Mice.

A The number of entries into the stranger mouse’s chamber and the social interaction time with the stranger mouse during the first phase of the three-chamber social test for ASD mice (n = 30) and fecal microbiota transplantation (ASD + FMT(EA)) mice (n = 12). Two-way ANOVA was used, followed by Games-Howell test (figure A-a) and Scheffé's procedure (figure A-b) for post hoc pairwise comparisons. B The number of entries into the stranger chamber and the social interaction time during the second phase of the three-chamber social test for ASD mice (n = 30) and ASD + FMT(EA) mice (n = 12). Two-way ANOVA was used, followed by Games-Howell test for post hoc pairwise comparisons. C, D Typical activity trajectories and differences in activity duration in the open-field test for ASD group (n = 13) and ASD + FMT(EA) group (n = 12). Independent samples T-tests was used. E Levels of inflammatory factors in the hippocampus, cerebral cortex, and serum of ASD group (n = 4) and ASD + FMT(EA) group (n = 4). Independent samples T-tests was used. F Fluorescence intensity of hippocampal microglial surface markers in ASD group (n = 6) and ASD + FMT(EA) group (n = 6) activated Iba1⁺ microglial cells, with comparison of avg-territory area in the hippocampus of mice from different groups. Independent samples T-tests was used. Scale bars, 100μm. The median is shown as a straight line and the box denotes the interquartile range. *P < 0.05, **P < 0.001, ***P < 0.001, NS: no significance.

In the open-field test, ASD + FMT(EA) mice exhibited significantly greater time spent and total distance traveled in the central area compared to ASD mice (Fig. 7C, D, both P < 0.05), indicating that fecal microbiota transplantation from the electroacupuncture-treated group effectively reduces anxiety-like and depressive-like behaviors in ASD mice.

Inflammatory status comparison between the two groups showed that inflammatory cytokine levels in the hippocampus, cerebral cortex, and serum were reduced in the ASD + FMT(EA) group compared to ASD mice (Fig. 7E). Specifically, significant differences were observed in TNF-α levels in the hippocampus and serum, as well as IL-6 levels in the cerebral cortex (Fig. 7E, P < 0.05), suggesting that fecal microbiota transplantation from the electroacupuncture group can alleviate inflammation in ASD mice.

A morphological quantitative assessment of the immunofluorescence images was conducted. Compared to the ASD group, the ASD + FMT(EA) group exhibited a significant decrease in the number of Iba1⁺ microglial cells per unit area in the hippocampal tissue (Fig. 7F, P < 0.01), along with a notable reduction in the avg-territory area (Fig. 7F, P < 0.01), while no significant difference was observed in the avg-some cytoplasmic area (Fig. 7F, P > 0.05).

In summary, fecal microbiota transplantation derived from the electroacupuncture-treated group demonstrated the ability to improve social deficits and alleviate anxiety-like and depressive-like behaviors in ASD mice. Additionally, it modulated both central and peripheral inflammatory responses and exhibited a trend toward attenuating hippocampal microglial activation.

Effects of fecal microbiota transplantation from ASD mice on the efficacy of electroacupuncture

Fecal microbiota from ASD mice was transplanted via enema into electroacupuncture-treated mice (EA + FMT(ASD) group). The behavioral differences, central and peripheral inflammatory profiles, and hippocampal microglial activity were systematically compared between EA group and EA + FMT(ASD) group.

Behavioral test results revealed that, during the first phase of the three-chamber social test, EA mice displayed a significant preference for interacting with Stranger 1 (Fig. 8A, P < 0.01). In contrast, EA + FMT(ASD) mice exhibited no significant interaction preference, with both interaction time and duration significantly reduced compared to EA mice (Fig. 8A, both P < 0.01). A similar trend was observed in the second phase of the social test, where EA mice showed significantly longer interaction times and higher interaction counts with Stranger 2 (Fig. 8B, P < 0.01). These findings suggest that fecal microbiota transplantation from ASD mice diminishes the regulatory effect of electroacupuncture on improving social deficits in ASD mice.

Fig. 8. Effects of Fecal Microbiota Transplantation from ASD Mice on the Efficacy of Electroacupuncture.

A The number of times ASD group (n = 10) and EA + FMT(ASD) group (n = 12) entered the stranger mouse chamber and the social interaction time with the stranger mouse in the first phase of the three-chamber social test. Two-way ANOVA was used, followed by Games-Howell test (figure A-a) and Scheffé's procedure (figure A-b) for post hoc pairwise comparisons. B The number of times ASD group (n = 10) and EA + FMT(ASD) group (n = 12) entered the stranger mouse chamber and the social interaction time with the stranger mouse in the second phase of the three-chamber social test. Two-way ANOVA was used, followed by Scheffé's procedure for post hoc pairwise comparisons. C, D Typical activity tracks in the open-field test for ASD group (n = 8) and EA + FMT(ASD) group (n = 12), including time and distance spent in the central area. Independent samples T-tests was used. E Levels of inflammatory factors in the hippocampus, cerebral cortex, and serum of ASD group and EA + FMT(ASD) group (n = 4). Independent samples T-tests was used. F Fluorescence intensity of hippocampal microglial surface markers in ASD group and EA + FMT(ASD) group (n = 6). Independent samples T-tests was used. Scale bars, 100μm. The median is shown as a straight line and the box denotes the interquartile range. *P < 0.05, **P < 0.001, ***P < 0.001, NS: no significance.

In the open-field test, EA + FMT(ASD) mice demonstrated a significantly reduced total distance traveled in the central area compared to EA mice, while activity time showed a decreasing trend (Fig. 8C, D, total distance P < 0.05, activity time P > 0.05). This indicates that fecal microbiota transplantation from ASD mice attenuates the anxiolytic and antidepressant-like effects of electroacupuncture in ASD mice.

Compared to EA group, EA + FMT(ASD) mice exhibited an upward trend in inflammatory factor levels across the hippocampus, cerebral cortex, and serum (Fig. 8E). Notably, significant increases in TNF-α levels were observed in the hippocampus in EA + FMT(ASD) mice compared to the EA group (Fig. 8E, P < 0.05). These findings suggest that fecal microbiota transplantation from ASD mice attenuates the anti-inflammatory effects of electroacupuncture on the CNS in ASD mice.

A morphological quantitative assessment of the immunofluorescence images was performed. Compared to the EA group, mice that underwent ASD gut microbiota transplantation showed a significant increase in the number of Iba1⁺ microglial cells per unit area in the hippocampal tissue (Fig. 8F, P < 0.01). Additionally, both the avg-territory area and avg-some cytoplasmic area were significantly reduced (Fig. 8F, P < 0.05). These results suggest that fecal microbiota transplantation from ASD mice may attenuate the modulatory effects of electroacupuncture on microglial overactivation in the hippocampus of ASD mice.

In summary, transplantation of fecal microbiota from ASD mice into EA mice diminished the regulatory effects of electroacupuncture on improving social behavior, alleviating anxiety-like and depression-like behaviors, and reducing central and peripheral inflammation in ASD mice. These findings indicate that the gut microbiota may play a critical role in the mechanisms underlying the therapeutic effects of electroacupuncture in ASD mice.

The role of VN integrity in electroacupuncture’s regulation of the gut-brain axis to alleviate central inflammation in ASD

Following truncal vagotomy (TV) in ASD mice, electroacupuncture intervention was administered (EA + TV group). Subsequent analyses compared behavioral outcomes, central and peripheral inflammatory profiles, and hippocampal microglial activity among ASD mice, EA mice, and EA + TV mice.

Behavioral assessments revealed that, during the first phase of the three-chamber social test, EA + TV mice exhibited a social preference comparable to that of EA mice (Fig. 9A, P > 0.05). However, both the frequency and duration of interactions in the EA + TV group showed a decreasing trend relative to the EA group. In the second phase, none of the three groups demonstrated a significant social preference (Fig. 9B, P > 0.05 for all intra-group comparisons). Additionally, EA + TV mice displayed reduced interaction frequency and duration compared to EA mice. These findings indicate that electroacupuncture retains partial modulatory effects on social deficits in ASD mice, even following vagus nerve transection.

Fig. 9. Comparison of the Behavioral Performance, Inflammatory Markers, and Microglial in Electroacupuncture after Truncal Vagotomy.

A Comparison of the first phase of the three-chamber social test among ASD group (n = 30), EA group (n = 10), and EA + TV group (n = 12). Two-way ANOVA was used, followed by Games-Howell test (figure A-a) and Scheffé's procedure (figure A-b) for post hoc pairwise comparisons. B Comparison of the second phase of the three-chamber social test among ASD group (n = 30), EA group (n = 10), and EA + TV group (n = 12). Two-way ANOVA was used, followed by Games-Howell test (figure B-a) and Scheffé's procedure (figure B-b) for post hoc pairwise comparisons. C, D Comparison of the activity trajectories, time spent in the central area, and total distance traveled in the open-field test among ASD group (n = 13), EA group (n = 8), and EA + TV group (n = 6). One-way ANOVA was used, followed by Tukey’s HSD method for post hoc pairwise comparisons. E Comparison of inflammatory factor levels in the hippocampus, cerebral cortex, and serum among groups (n = 4). One-way ANOVA was used, followed by Games-Howell test (figure E a-b) and Tukey’s HSD method (figure E c-i) for post hoc pairwise comparisons. F Comparison of the number of activated microglia in the hippocampal brain tissue among groups (n = 6). One-way ANOVA was used, followed by Tukey’s HSD method for post hoc pairwise comparisons. Scale bars, 100μm. The median is shown as a straight line and the box denotes the interquartile range. *P < 0.05, **P < 0.001, ***P < 0.001, NS: no significance.

In the open-field test, the EA group exhibited increase in both central zone duration and total distance traveled compared to the ASD group, with the distance difference reaching statistical significance (Fig. 9C, D, P < 0.05). In contrast, the EA + TV group demonstrated a reversal of the therapeutic effects induced by electroacupuncture. Similarly, a statistically significant difference in total distance traveled was also observed between groups (Fig. 9C, D, P < 0.05). These findings suggest that electroacupuncture exerts modulatory effects on anxiety-like and depressive-like behaviors in ASD mice, partially mediated by vagal nerve integrity.

Compared to ASD mice, EA + TV mice exhibited a downward trend in inflammatory factor levels within the hippocampus, cerebral cortex, and serum (Fig. 9E). Notably, TNF-α levels in the hippocampus and IL-6 levels in the cerebral cortex and serum were significantly reduced in the EA + TV group (Fig. 9E, P < 0.05). However, when compared to EA-treated mice, EA + TV mice showed an upward trend in inflammatory factor levels across the hippocampus, cerebral cortex and serum(Fig. 9E, P > 0.05), with a significant increase in hippocampal TNF-α levels observed in the EA + TV group (Fig. 9E, P < 0.05). These findings suggest that electroacupuncture intervention following vagus nerve transection exerts only partial regulatory effects on central and peripheral inflammation in ASD mice.

In the morphological quantitative assessment, compared to the ASD group, the number of Iba1⁺ microglial cells and the avg-territory area in the EA + TV group mice were significantly reduced (Fig. 9F, P < 0.01). However, when compared to the EA group, there were no significant differences in the number of Iba1⁺ microglial cells per unit area or avg-some cytoplasmic area in the hippocampal tissue of the EA + TV mice (Fig. 9F, P > 0.05). Notably, the avg-territory area was significantly increased (Fig. 9F, P < 0.01). These findings suggest that electroacupuncture intervention after truncal vagotomy exerts only partial regulatory effects on the overactivation of hippocampal microglia in ASD mice.

In conclusion, electroacupuncture intervention in vagus nerve-transected ASD mice demonstrates a diminished regulatory effect on improving social behaviors, alleviating anxiety-like and depression-like behaviors, and reducing central and peripheral inflammation. These findings suggest that the VN may partially mediate the regulatory mechanisms underlying the effects of electroacupuncture in ASD mice.

Discussion

The precise etiology of most ASD cases remains largely undefined. However, extensive clinical and preclinical studies have demonstrated that individuals with ASD often suffer from gastrointestinal dysfunctions, such as selective eating, constipation, and abdominal pain, which are often correlated with the severity of ASD symptoms [26, 27]. Experimental studies have further confirmed that VPA-induced ASD mouse models exhibit significant disruptions in gut microbiota composition [28]. Notably, the gut microbiota of ASD patients exhibits significantly increased diversity [29], but lacks certain beneficial bacterial taxa, such as Prevotella. Dysbiosis in the gut microbiota also alters the regulation of metabolites, including corticosterone, indoles, and lipopolysaccharides (LPS), which subsequently impact both neurobiological function and gastrointestinal health [30]. Moreover, treatment with Lactobacillus reuteri has been shown to reverse social deficits in VPA-induced ASD mouse models, an effect that appears to be mediated by VN signaling.

Previous research has revealed that microglia in the brains of children with ASD are abnormally activated [31, 32], with a concomitant increase in the expression of TNF and IL-17 [33, 34]. This immune activation-driven neuroinflammation in the central nervous system may disrupt the normal function of microglia, particularly in the process of synaptic pruning [35]. In their activated state, primary microglia release a range of pro-inflammatory cytokines, including IL-1α, IL-1β, IL-6, and TNF-α, as part of their inflammatory response [36–38]. Moreover, compared to resting microglial cells, the cell body area of activated microglia can show a general increase [39, 40]. Targeting and inhibiting microglia-mediated neuroinflammation thus holds promise as a potential therapeutic strategy for ASD [41, 42].

Growing evidence from international research has underscored a strong correlation between gut microbiota, the VN, and the onset, progression, and severity of ASD-related behaviors [43]. Acupuncture has been shown to regulate neuroinflammation by modulating gut microbiota and influencing multiple neural pathways [30]. In our study, transplantation of gut microbiota from electroacupuncture-treated ASD mice into untreated ASD mice effectively inhibited the activation of microglia and alleviated behavioral impairments and neuroinflammation. These findings not only confirm that gut microbiota composition plays a critical role in the pathogenesis of ASD but also highlight the pivotal role of gut microbiota in mediating the therapeutic effects of acupuncture for ASD.

Based on the results of our initial experiment, we conducted a reverse study in which gut microbiota from untreated ASD mice was transplanted into ASD mice that had undergone electroacupuncture treatment. This intervention attenuated the regulatory effects of electroacupuncture, including improvements in social behavior, reductions in anxiety and depression, and decreases in CNS and systemic inflammation. These findings provide further evidence that gut microbiota plays a critical role in the mechanisms by which electroacupuncture modulates neuroinflammation in ASD mice.

The gut-brain axis is a well-established hypothesis proposing that gut microbiota can regulate gastrointestinal physiology, immunity, cognition, and behavior through a dynamic, bidirectional communication between the gut and brain [44, 45]. However, a universally accepted mechanism that elucidates how gut bacteria impact the nervous system remains elusive. Current research indicates that this communication is chiefly mediated through neural pathways, the neuroendocrine system, neuroimmune interactions, and, importantly, the gut microbiota itself [30].

Disruptions in the gut microbiota can induce intestinal inflammation, undermining the integrity of the intestinal mucosa and resulting in gut barrier dysfunction. This dysfunction facilitates chronic systemic inflammation, which in turn activates the immune system, increases endothelial permeability, and enables immune cells to migrate through the bloodstream. These cascading effects ultimately disrupt the blood-brain barrier, setting the stage for neuroinflammation [46, 47].

However, this framework does not fully account for the critical role of the VN in the gut-brain axis. Emerging evidence suggests that the VN can transport gut-derived signals directly to the CNS, bypassing the bloodstream [48–50]. This direct neural communication offers an additional mechanism by which gut microbiota can exert influence over brain function.

The integrity of the VN appears to be a crucial factor in how the gut microbiota regulates brain function. The VN has been identified as a key modulator of cognitive and memory-related processes, playing an essential role in the structural remodeling of dendritic spines in the hippocampus. Notably, VN stimulation has been shown to enhance learning and memory behaviors, as well as to promote hippocampal neurogenesis and synaptic plasticity [51, 52]. Conversely, damage to the VN has been associated with alterations in cognitive behavior and neurochemical adaptations [53].

Through the VN pathway, the gut-brain axis influences hippocampal-dependent learning and memory, dendritic spine morphology, and synaptic plasticity, with imbalances in these processes being implicated in neurological dysfunction [54]. In our study we sever the right vagus nerve to observe the effect of vagus nerve integrity on acupuncture modulation of the gut-brain axis. The control experiment revealed that acupuncture at ST36 was significantly less effective in treating ASD following vagus nerve transection. These findings underscore the essential role of VN integrity in the ability of acupuncture to modulate the nervous system.

Acupuncture has gained widespread use in the treatment of neuropsychiatric disorders such as insomnia and depression, owing to its safety, efficacy, and ease of application. Several systematic reviews have highlighted the promising potential of acupuncture in this regard [55, 56]. Acupuncture has been shown to regulate CNS and gastrointestinal functions, as well as modulate the composition and abundance of gut microbiota, thereby promoting systemic functional balance [16, 57, 58]. Acupuncture intervention has been shown to regulate gut microbial alpha/beta diversity and enhance cognitive function in VPA-induced ASD mice [59].

Drawing upon Chinese ancient medical texts, we have systematically compiled and classified the meridians and acupoints that have the potential to simultaneously regulate both the “gut” and “ spirit (Shen)” [60]. Our previous study has investigated the therapeutic effects of acupuncture at gastrointestinal-related points, such as Zusanli (ST36), Sanyinjiao (SP6), Hegu (LI4), and Quchi (LI11), in individuals with ASD. The results revealed a significant correlation between improvements in behavioral symptoms and the alleviation of gastrointestinal dysfunction in children with ASD [61]. Among the many acupuncture points, Zusanli (ST36), located on the Stomach Meridian of Foot-Yangming, stands out as a classic and widely employed point for treating both gastrointestinal disorders and mental health conditions[62, 63]. The underlying mechanisms may involve the improvement of the intestinal barrier, the blood-brain barrier and other factors [64–66]. However, despite these promising results, the precise mechanisms underlying the therapeutic effects of acupuncture remain unclear. Our study, building upon these previous investigations on ST36, further explores the crucial role of the integrity of the VN in acupuncture’s modulation of the gut-brain axis, providing a new perspective for research in this area.

This study demonstrates that acupuncture at ST36 mitigates behavioral abnormalities in ASD mouse models. Moreover, our findings reveal that electroacupuncture at ST36 not only reduces neuroinflammation in the hippocampus and cerebral cortex but also lowers serum levels of inflammatory cytokines. Our experiment demonstrated that acupuncture influences the VN, initiating changes in the gut microbiota along the gut-brain axis. These alterations then regulate microglial activation and exert anti-neuroinflammatory effects in ASD. Based on these findings, we hypothesize that acupuncture modulates VN activity, thereby impacting microglial activity in the cerebral cortex and hippocampus. This modulation may influence the development and pruning of dendritic spines in neurons, ultimately improving memory, emotional regulation, and other cognitive functions in ASD mouse models.

To explore these mechanisms, our future research will focus on comparing the changes in neuronal dendritic morphology in the cerebral cortex and hippocampus of ASD mice before and after electroacupuncture. This investigation will provide deeper insights into the therapeutic effects of acupuncture on nervous system function and its potential role in the treatment of neurodevelopmental disorders.

Conclusions

Our study reveals that electroacupuncture could attenuate CNS and peripheral inflammatory states and overactive of microglial cells, improving social deficits in ASD mice. This effect can be prevented by intestinal microbiota transplantation from untreated ASD mice. The vagotomy in ASD mice abolished the therapeutic benefits of acupuncture, showing the significance of the integrity of vagus nerve in electroacupuncture therapeutic effects.

These findings provide compelling evidence for the potential of electroacupuncture to modulate central neuroinflammation through vagus nerve-mediated gut microbiota regulation, offering novel avenue into its therapeutic application for neurodevelopmental disorders such as ASD.

Author contributions

Dong Chen, Xinyi Yang, and Daiyan Jiao contributed equally as co-first authors and were responsible for the animal behavioral experiments and manuscript preparation. Mengqian Yuan, Bin Xu, and Yancai Li served as co-corresponding authors, overseeing experimental design and quality supervision. Mengqian Yuan was responsible for data analysis and figure preparation. Xiaoyan Chen carried out animal modeling. Wenhui Xiao performed immunofluorescence experiments. Jingjing Zheng and Yingxin Li were in charge of serological testing and gut microbiota analysis. Chao Bao provided guidance on acupuncture techniques.

Fundings

This research is partially supported by National Natural Science Foundation of China, 82004456.

Data availability

Data analyzed in this study were derived from a subset of laboratory mouse experiments and are not publicly available. However, the data can be made available by the corresponding author upon reasonable request.

Competing interests

The authors declare no competing interests.