Abdominal Ultrasound Stimulation Alleviates Negative Symptoms Through Modulation of Serotonin Signaling and Gut Microbiota in the MK-801 Model of Schizophrenia

Abstract

Background and Hypothesis

Abdominal low-intensity pulsed ultrasound (LIPUS) stimulation has potential as a novel therapeutic strategy against neuroinflammation via inhibition of inflammatory responses in the colon. This study aimed to evaluate whether abdominal LIPUS could alleviate MK-801-induced schizophrenia-like negative symptoms through gut–brain communication.

Study Design

Rats administered with MK-801 were treated daily for 5 days with either LIPUS or Lactobacillus plantarum PS128, while another group of MK-801-administered rats received no treatment. Following LIPUS or PS128 treatment, rats underwent behavioral testing, western blot analysis, and histological examination. Changes in the gut bacteria composition were examined through 16S rRNA sequencing analysis.

Study Results

MK-801 administration reduced NMDAR1 and VGAT expression in the medial prefrontal cortex (mPFC) of rats, leading to an imbalance in the excitation/inhibition (E/I) ratio. It also decreased 5-HT1AR and 5-HT2AR density, resulting in reduced concentrations of dopamine and serotonin (5-HT). This induced prepulse inhibition, anhedonia, and social withdrawal behaviors, accompanied by a reduction in gut microbiota diversity. Abdominal LIPUS stimulation effectively lessened the MK-801-induced reduction in gut microbiota diversity, restored NMDAR1, 5-HT1AR, and 5-HT2AR density, enhanced dopaminergic neuron activity, and increased dopamine and 5-HT release in the mPFC, thereby reversing behavioral abnormalities.

Conclusions

These results suggest that abdominal LIPUS alleviates MK-801-induced schizophrenia-like negative symptoms by modulating serotonin signaling and the gut microbiota.

Introduction

Schizophrenia is a major psychiatric disorder with a worldwide incidence of approximately 1%.¹ While psychopharmacological treatments can alleviate the positive symptoms of schizophrenia, their overall efficacy is limited by persistent negative symptoms and cognitive deficits.² The N-methyl-D-aspartate (NMDA) receptor hypofunction hypothesis of psychosis posits that reduced activity of NMDA receptors (NMDARs) on gamma-aminobutyric acid (GABA) interneurons within the medial prefrontal cortex (mPFC) leads to an imbalance between excitatory and inhibitory neurotransmission. Specifically, NMDAR hypofunction diminishes the activity of GABAergic interneurons, resulting in decreased inhibition of pyramidal neurons. This disinhibition leads to increased excitatory glutamate signaling, contributing to neural circuit dysfunctions associated with psychosis.^3,4 Furthermore, pyramidal glutamatergic neurons in the mPFC project to dopaminergic neurons in the ventral tegmental area (VTA), influencing dopamine release and potentially contributing to the development of psychotic symptoms. Proper NMDAR signaling is crucial for facilitating GABAergic transmission in the mPFC, maintaining the necessary excitation/inhibition (E/I) balance for normal cognitive and behavioral functions.^5,6 Numerous studies have reported structural abnormalities in the cortex and impaired function of the mPFC in individuals with schizophrenia.^7,8 Negative symptoms of schizophrenia, such as anhedonia and lack of motivation, have been linked to reduced dopamine (DA) activity in the frontal cortex.⁹

The microbiota–gut–brain axis (MGBA) is a bidirectional communication that links the gastrointestinal tract and brain.¹⁰ A number of reports have highlighted the potential role of the MGB axis in schizophrenia, with the gut microbiota modulating the bidirectional signaling between the central nervous system (CNS) and the enteric nervous system.¹¹ The disruption of the microbiota and microbial metabolites can influence behavioral patterns in mice.¹² The gut microbiota has been associated with psychiatric disorders, particularly in the context of depression in which clinical studies have suggested the presence of dysbiosis.¹³ Furthermore, the gut microbiome has the ability to synthesize functional neurotransmitters, such as serotonin (5-HT), DA, Glu, and GABA, through the digestion and metabolic breakdown of dietary compounds. Certain neurotransmitters produced by the gut microbiota are capable of traversing the blood–brain barrier directly, thereby affecting the functionality of the CNS.¹⁴

The dysregulation of various neurotransmitters, such as 5-HT and DA deficiency in the brain, may play a role in the pathogenesis of schizophrenia.¹⁵ Among the known serotonin receptor subtypes, the 5-HT1A and 5-HT2A receptors are considered to be the significant causal factor in the development of the schizophrenia brain.¹⁶ NMDAR antagonists, such as dizocilpine (MK-801), are able to produce both positive and negative symptoms in animal models of schizophrenia.¹⁷ While there are currently no behavioral models that allow for the direct assessment of potential control over negative symptoms, social withdrawal is a hallmark of deficit symptoms.¹⁸ The 5-HT1A and 5-HT2A receptors are crucial modulators of the negative symptom spectrum associated with schizoaffective disorders.¹⁶ In addition, decreased postsynaptic 5-HT1A receptor density is associated with impaired serotonergic signal transduction, potentially leading to reduced DA release in the frontal cortex.¹⁹

Transcranial ultrasound stimulation (TUS) has been demonstrated to enhance cognitive function and mitigate neuropathological changes in MK-801-treated rats.²⁰ Transcranial ultrasound stimulation contributes to the regulation of NMDA receptors and the modulation of c-Fos activity, suggesting its potential as an effective antipsychotic treatment for schizophrenia.²¹ Furthermore, our previous research revealed that applying abdominal LIPUS can mitigate LPS-induced neuroinflammation by suppressing colonic inflammation.^22,23 Abdominal LIPUS stimulation has emerged as a potential therapeutic approach for alleviating colitis-induced behavioral disorders by modulating MGBA signaling.²⁴ Nevertheless, the impact of abdominal low-intensity pulsed ultrasound (LIPUS) stimulation on the MGB axis and negative symptoms of schizophrenia remain to be explored. This study aimed to investigate whether abdominal LIPUS could alleviate negative symptoms (anhedonia and social withdrawal) of schizophrenia by modulating the microbiota and 5-HT in a murine model, potentially laying the groundwork for the use of this technology for treating CNS diseases.

Methods

Abdominal Ultrasound Setup

Abdominal LIPUS was administered using a therapeutic ultrasound generator (ME740, Mettler Electronics, Anaheim, CA, USA) equipped with a 1-MHz plane transducer (ME7413, Mettler Electronics) with a 4.4-cm² effective radiating area (Figure S1A). The ultrasound parameters were configured with 2-ms burst durations at a 20% duty cycle and a repetition frequency of 100 Hz. The spatial average intensity on the transducer’s planar surface was set to 0.5 W/cm² and this value was confirmed using a radiation force balance (Precision Acoustics, Dorset, UK) in degassed water.

Prior to LIPUS treatment, the abdominal fur of the rats was carefully shaved. An ultrasound transmission gel (Pharmaceutical Innovations, Newark, NJ, USA) was applied between the transducer and the abdominal area to enhance the efficiency of ultrasound transmission. The treatment area included the entire abdomen, ranging from the diaphragm to the groin. Each sonication session lasted 5 min, with a 5-min rest interval between the first and second and between the second and third sessions. In total, the daily LIPUS stimulation lasted for 15 min.

Schizophrenia Model Induction and Experimental Protocols

All animal procedures were performed in accordance with the Guidelines for the Care and Use of Laboratory Animals, with the study protocol receiving approval from our Institutional Animal Care and Use Committee. Male Sprague Dawley rats (BioLASCO Taiwan Co., Ltd., Yilan City, Taiwan), weighing 230-280 g, were used in this study. Prior to LIPUS stimulation, each rat was anesthetized with inhaled 2% isoflurane in 2 L/min oxygen; during anesthesia, the rat was in the prone position, and its body temperature was maintained at 37°C using a heating pad. Rats receiving Lactobacillus plantarum PS128 (Bened Biomedical Co., Ltd., Taipei, Taiwan) treatment were orally administered 50 mg of lyophilized bacterial powder, equivalent to 1.5 × 10¹⁰ colony-forming units, suspended in normal saline once per day for 5 consecutive days. The animals were randomly divided into 4 groups—the Sham (sham-operated) group, the MK-801 group, the MK-801 + LIPUS group, and the MK-801 + PS128 group—for subsequent biochemical analysis, histological assessment, and behavioral evaluation. MK-801 was administered to the animals in the MK-801, MK-801 + LIPUS, and MK-801 + PS128 groups, with the last 2 groups also receiving daily LIPUS and PS128 treatment, respectively, for 5 consecutive days, beginning 1 day after MK-801 administration (Figure S1B). Body weight was recorded daily, and spleen weight, along with colon length (measured from the ileocecal junction to the anus), was assessed at the time of sacrifice.

Assessment of Behavior

The prepulse inhibition (PPI) test was employed to evaluate alterations in sensorimotor gating. To examine attentional processing, the PPI of the acoustic startle response was assessed. Rats were placed in startle chambers for a 30-min adaptation period. Following acclimatization to background white noise (70 dB for 30 s), the rats were exposed to a startle pulse at 120 dB for 40 ms and a prepulse, consisting of a pure tone at 73, 76, or 82 dB for 20 ms. Each trial featured an average intertrial interval of 15 s. The PPI response to each of the 3 prepulse intensities was calculated using the following formula: PPI (%) = (mean peak amplitude of pulse-only sessions − mean peak amplitude of prepulse/pulse sessions)/mean peak amplitude of pulse-only sessions × 100.

Sucrose preference was calculated using the following formula: sucrose preference percentage (%) = (sucrose solution consumption [g]/(sucrose solution consumption [g] + water consumption [g])) × 100. All rats underwent adaptive training from day 1 to day 4. During this period, 2 bottles of pure water were provided on days 1 and 2, followed by 2 bottles of 1% sucrose solution on day 3, and 1 bottle of pure water and 1 bottle of 1% sucrose solution on day 4. After a 12-h period of food and water deprivation, each rat was offered 200 mL of pure water and 200 mL of 1% sucrose solution. The amounts of pure water and sucrose solution consumed were recorded after 1 h and again after 12 h.

A social interaction test was conducted in a black plexiglass open-field arena (80 × 80 × 30 cm) under semi-dim lighting conditions. In this test, 2 unfamiliar rats from the same treatment group were allowed to interact with each other for 5 min. The time spent on behaviors such as sniffing, following, and climbing was recorded as an indicator of socialization, while the time spent on avoiding was noted as an indicator of social withdrawal.²⁵ Total interaction was calculated for each rat according to the following formula: Time spent on following + Time spent on sniffing − Time spent on avoiding behaviors.

Western Blot Analysis

Fresh brain tissue from the mPFC was homogenized in T-Per extraction reagent, which was supplemented with Halt Protease Inhibitor Cocktail (Pierce Biotechnology, Inc.). The lysates were centrifuged, and the resulting supernatants were collected. Protein concentrations were determined using a protein assay reagent (Bio-Rad, CA, USA). Samples containing 30 μg of protein were separated on a 15% sodium dodecyl sulfate–polyacrylamide gel and then transferred onto Immun-Blot polyvinylidene fluoride membranes (Bio-Rad, CA, USA). After transfer, the membranes were blocked for at least 1 h with milk (Anchor) and incubated overnight with rabbit antibodies against VGluT1 (Abcam, 1:2000), VGAT (GeneTex, 1:2000), GAD65/GAD67 (GeneTex, 1:2000), and GAPDH (GeneTex, 1:5000) at 4°C. After 4 10-min washes with PBST, the membranes were incubated for 1 h at room temperature with secondary antibodies (goat antimouse IgG HRP, GeneTex, 1:5000 for GAPDH). After washing with PBST buffer, signals were developed using Western Lightning Pro ECL reagent (Bio-Rad, California, USA). The gel images were captured with a biomolecular imager (ImageQuant LAS 4000; GE Healthcare Life Sciences, Pennsylvania, USA) and analyzed with Image J software (National Institute of Health, Bethesda, MD, USA) to estimate the integral optical density of the protein bands.

Histological Assessment

The rats were intracardially perfused with 4% paraformaldehyde in phosphate buffer under anesthesia. Their brains were postfixed in 4% paraformaldehyde overnight and then transferred to phosphate-buffered saline (PBS) containing 30% sucrose for cryoprotection. Coronal sections of the brain, 10 μm thick, were cut using a microtome. After being incubated for 30 min at 37°C in blocking solution containing fetal bovine serum (SKU 04-001-1A, Biological Industries, USA), the sections were immunostained with the following primary antibodies: rabbit anti-NMDAR1 (NR1, 1:200; GeneTex, California, USA); goat antiparvalbumin (1:200; GeneTex, California, USA); rabbit anti-5-HT1A (1:200; Novus Biologicals, Centennial, CO, USA) and 5-HT2A (1:200; Alomone Labs, Ltd.); rabbit antineuronal nuclei (NeuN, 1:500; GeneTex, California, USA); mouse anti-c-Fos (1:200; Abcam, Cambridge, USA); and rabbit antityrosine hydroxylase (TH, 1:500; GeneTex, California, USA). The brain sections were incubated with the primary antibodies overnight, followed by washing and incubation with Alexa Fluor 488- or Alexa Fluor 594-conjugated secondary antibodies (1:500; Abcam, Cambridge, MA, USA) at room temperature. The samples were then incubated with DyLight594-conjugated secondary antibody for 1 h at 37 °C, washed 3 times with PBS, and counterstained with DAPI. For each rat, 3 sections were analyzed under consistent exposure conditions, with at least 3 randomly selected fields from the mPFC captured per section. Double immunofluorescence was visualized and photographed using a fluorescence microscope (Leica DM 6000B, Mannheim, Germany). The number of cells positive for NR1–PV, 5-HT1A–NeuN, and 5-HT2A–NeuN double labeling was quantified in 3 nonoverlapping fields in the mPFC at 400 × (318 × 237 μm²) magnifications. The number of cells positive for c-Fos–TH double labeling was counted in the VTA within a 318 × 237 μm² area across 3 nonoverlapping fields at 400 × magnification.

Enzyme-Linked Immunosorbent Assay

The concentrations of DA and 5-HT in the mPFC were determined using enzyme-linked immunosorbent assay kits (DA [RK00642], Abclonal, Woburn, MA, USA; 5-HT [MBS166089], MY BioSource, San Diego, CA, USA), following the procedures outlined by the manufacturers.

Gut Microbiota Analysis

Total genomic DNA was extracted from frozen fecal samples using the CatchGene Stool DNA kit. The DNA concentration was measured with a fluorometer (Qubit 4.0, Thermo Scientific) and standardized to 1 ng/μL for further processing. The full-length 16S rRNA gene (V1-V9 regions) was amplified using barcoded primers specific to the 16S gene. These primers, designed for multiplexed SMRTbell library preparation and sequencing on the PacBio platform, included a 5′ buffer sequence (GCATC) with a 5′ phosphate modification, a 16-base barcode, and degenerate forward or reverse primer sequences targeting the 16S gene. The amplified sequences were grouped into operational taxonomic units (OTUs) based on a 97% similarity threshold. To investigate the factors influencing alpha diversity, we calculated the observed richness, Faith’s phylogenetic diversity (Faith’s PD), and Margalef’s species richness indices for each sample. Differential abundance analysis was performed using a zero-inflated Gaussian log-normal model implemented in the “fitFeatureModel” function of the Bioconductor metagenomeSeq package to identify significant differences in taxa between groups. In addition, Welch’s t-test was applied using STAMP software.

Statistical Analysis

All data are shown as means ± SD. All data, except for the 16S rRNA sequencing data, were analyzed using one-way ANOVA, followed by the Kruskal–Wallis post hoc test to determine significant differences between groups. The level of statistical significance was set at P < .05.

Results

Effects of Abdominal LIPUS Stimulation on Physiological Parameters in the MK-801-Treated Rats

Four days after MK-801 administration, the MK-801-treated rats in the 3 groups showed a significant loss of body weight compared to the Sham group (all P < .05; Figure S1C). The spleen weight/body weight ratio did not differ between the 4 groups (Figure S1D). In addition, no significant difference was found in colon length between the groups (Figure S1E and F).

Abdominal LIPUS Stimulation Attenuated Prepulse Inhibition Deficit and Negative Symptoms in the MK-801-Treated Rats

The PPI test was performed to assess the sensory-motor gating deficit. At all 3 prepulse intensities, the mean PPI percentage of the MK-801 rats was significantly less than that of the Sham group (all P < .01; Figure 1A). LIPUS stimulation significantly attenuated the decrease in PPI percentage of the MK-801 rats at 73 and 76 dB (both P < .01; Figure 1A). At 82 dB, LIPUS stimulation exhibited a tendency to increase the PPI percentage, but the difference was not statistically significant. However, PS128 had no effect on PPI percentage at all 3 prepulse intensities.

The sucrose preference test and the social interaction test were used to evaluate anhedonia and social withdrawal, which represent negative symptoms of schizophrenia.^26,27 In the MK-801-treated rats, the mean sucrose preference was 22.23 ± 5.57%, while in the Sham group, it was 51.42 ± 8.96%. The MK-801-induced decrease in sucrose preference was statistically significant (P < .001; Figure 1B). LIPUS treatment significantly alleviated this decrease in sucrose preference in the MK-801 + LIPUS group compared to the MK-801 group (P < .001; Figure 1B). No significant change was observed following PS128 administration in the MK-801-treated rats.

In this study, the social interaction test measured behaviors such as sniffing, following, climbing, and avoidance. The duration of sniffing, following, and climbing behavior was significantly lower in the MK-801 group compared to the Sham group (all P < .05; Figure 1D–F). LIPUS stimulation significantly reversed the effect of MK-801 on sniffing and following behavior (all P < .001; Figure 1D and E), while PS128 did not change this effect. MK-801 administration also led to a significant increase in avoiding behavior compared to the Sham group (P < .001; Figure 1G), and while LIPUS stimulation reversed this effect, PS128 showed no efficacy in the test. While LIPUS stimulation showed a trend toward increased climbing behavior, the difference was not statistically significant. In addition, rats in the MK-801 group exhibited a significant reduction in total interactions compared to the Sham group. LIPUS stimulation reversed this reduction (P < .001; Figure 1H), whereas PS128 did not demonstrate efficacy.

Abdominal LIPUS Mitigated the MK-801-Induced Reduction in NMDAR1 Expression and Restored the Excitation–Inhibition Balance in MK-801-Treated Rats

Double immunofluorescence labeling of PV and NMDAR1 subunits was used to specifically assess NMDAR1 expression in PV interneurons (Figure 2A). The puncta in individual neurons were quantified. The number of NMDAR1 subunit puncta in PV interneurons in the MK-801-treated group was significantly reduced compared to that in the Sham group (20.40 ± 2.70 vs. 40.20 ± 5.26, P < .001; Figure 2B). However, in the MK-801 + LIPUS group, LIPUS treatment significantly mitigated the MK-801-induced reduction in NMDAR1 subunit puncta in PV interneurons (20.40 ± 2.70 vs. 35.80 ± 6.14, P < .01; Figure 2B), whereas PS128 showed no effect on this reduction. In addition, no significant differences were observed in the number of PV interneurons in the mPFC of rats between the 4 groups (Figure 2C).

GABA is synthesized from glutamate through the action of glutamate decarboxylase (GAD), which has 2 isoforms: GAD65 and GAD67. Both isoforms are expressed in GABAergic neurons. No significant difference in either isoform was observed between the 4 groups (Figure 2D and E). VGAT is localized in the vesicles in inhibitory terminals of both GABAergic and glycinergic neurons. MK-801 treatment significantly reduced VGAT expression in the mPFC (P < .01; Figure 2G) without affecting VGluT1 levels (Figure 2F). The VGluT1/VGAT ratio was notably elevated in MK-801-treated rats (P < .01; Figure 2H). LIPUS stimulation significantly restored VGAT expression and the VGluT1/VGAT ratio of the MK-801 + LIPUS group compared to the MK-801 group (both P < .01; Figure 2G and H).

Abdominal LIPUS Reversed MK-801-Induced Downregulation of 5-HT Receptor Expression

Double immunofluorescent labeling of neurons and the 5-HT1A receptor or the 5-HT2A receptor was performed to specifically assess 5-HT1A receptor or 5-HT2A receptor expression in mPFC neurons (Figure 3A and D). The puncta localized in individual neurons were quantified. The MK-801-treated group exhibited a significant reduction in the number of 5-HT1A receptor puncta (19.00 ± 5.92 vs. 33.80 ± 4.32, P < .001; Figure 3B) or 5-HT2A receptor puncta (22.00 ± 2.74 vs. 38.00 ± 5.70, P < .001; Figure 3E) in neurons compared to the Sham group. In the MK-801 + LIPUS group, LIPUS significantly lessened the MK-801-induced reduction in the number of 5-HT1A (19.00 ± 5.92 vs. 31.20 ± 3.56, P < .01; Figure 3B) or 5-HT2A (22.00 ± 2.74 vs. 34.80 ± 3.83, P < .001; Figure 3E) receptor puncta in neurons compared to the MK-801 group. However, PS128 showed no effect on this reduction. In addition, no significant differences in neuron numbers in the mPFC were observed between the 4 groups (Figure 3C, F).

Abdominal LIPUS Reversed MK-801-Induced Downregulation of c-Fos Expression and Dopamine and Serotonin Levels

Double immunostaining for c-Fos and TH was performed to evaluate the excitability of VTA neurons (Figure 4A). Immunocytochemical analysis showed that c-Fos expression in TH + dopamine neurons in the VTA was significantly reduced in the MK-801-treated group compared to the Sham group (10.40 ± 3.44 vs. 22.80 ± 4.76, P < .001; Figure 4B). The MK-801-induced decrease in c-Fos puncta in TH-positive dopamine neurons was significantly reversed by LIPUS treatment (10.40 ± 3.44 vs. 18.40 ± 2.88, P < .05; Figure 4B). In contrast, PS128 showed no effect on this reduction. No significant differences were observed in the number of TH-positive dopamine neurons in the VTA between the 4 experimental groups (Figure 4C). Figure 4D and E shows the DA and 5-HT levels in the mPFC. DA and 5-HT levels were significantly reduced in MK-801-treated rats (both P < .01; Figure 4D and E). However, this reduction was significantly lessened after LIPUS treatment (both P < .05). However, PS128 showed no effect on this reduction. In addition, the serum level of 5-HT did not significantly differ between the 4 groups, indicating that MK-801, MK-801 + LIPUS, and MK-801 + PS128 treatments did not alter the 5-HT level in the circulatory system (Figure 4F).

Effects of Abdominal LIPUS on the Intestinal Microbiota Composition in MK-801-Treated Rats

To analyze changes in microbiota composition, the relative abundances of microbial taxa were evaluated at phylum, class, and genus levels (Figure 5). Each group had more than 97 distinct OTUs, with 234 OTUs shared between the 4 groups (Figure 5A). Figure 5B illustrates the microbiota composition at the phylum level. The dominant bacterial phyla in the 4 groups were the Firmicutes and Bacteroidota. MK-801 reduced the relative abundance of Firmicutes and increased that of Bacteroidota, but the differences were not significant. Moreover, the relative abundances of the taxa did not differ between the 4 groups. Clostridia, Bacilli, and Bacteroidia were the dominant classes of bacteria identified in the Sham group (Figure 5C). After MK-801 administration, the relative abundance of the Bacilli decreased, but it tended to recover after LIPUS or PS128 treatment. Figure 5D presents the relative abundances of bacteria at the genus level. The results indicate that the primary differences between the Sham and MK-801 groups were in the relative abundances of Roseburia and Ligilactobacillus. Both genera showed a recovering trend after LIPUS or PS128 treatment.

In a mouse model of social defeat stress, the relative abundance of Mollicutes at the class level decreased, with abundance recovering and depressive behaviors improving after treatment.²⁸ This suggests that Mollicutes may be associated with depressive behavior symptoms. Consistent with the above study, a significant reduction in the relative abundance of Mollicutes was observed in the MK-801 group compared to the Sham group (P < .01; Figure 5E). Mollicutes abundance recovered in the LIPUS and PS128 groups, but this recovery was not statistically significant. Furthermore, numerous studies have suggested that the gut microbiota may influence neuroinflammation in the brain via the gut–brain axis, with particular attention given to the genus Roseburia and the species Roseburia hominis.²⁹ Short-chain fatty acids, important metabolites of Roseburia, reduce neuroinflammation through the production of butyrate.³⁰  Roseburia hominis alleviated neuroinflammation in the digestive system.³¹ The relative abundance of Roseburia was significantly reduced in the MK-801 group compared to the Sham group (0.003 ± 0.002 vs. 0.05 ± 0.04, P < .05; Figure 5F), with significant recovery in relative abundance following LIPUS treatment (0.003 ± 0.002 vs. 0.05 ± 0.04, P < .05). Similarly, the relative abundance of Roseburia hominis was significantly reduced in the MK-801 group compared to the Sham group (0.003 ± 0.002 vs. 0.05 ± 0.04, P < .05; Figure 5G), with notable recovery in relative abundance after LIPUS treatment (0.003 ± 0.002 vs. 0.05 ± 0.04, P < .05). However, no significant recovery was observed in the PS128 treatment group. A previous study showed that the genus Ligilactobacillus and the species Ligilactobacillus murinus play key roles in host gut metabolism and immune function, with their abundances being closely linked to intestinal health.³² In a study on depression induced by Dcf1 gene knockout in mice, a decrease in the relative abundance of Ligilactobacillus was observed.³³ The relative abundance of Ligilactobacillus was significantly reduced in the MK-801 group compared to the Sham group (0.05 ± 0.05 vs. 0.30 ± 0.15, P < .05; Figure 5H). Similarly, the relative abundance of Ligilactobacillus murinus was significantly reduced in the MK-801 group compared to the Sham group (0.05 ± 0.05 vs. 0.30 ± 0.15, P < .05; Figure 5I). Although the relative abundances of both Ligilactobacillus and Ligilactobacillus murinus showed an increase in the LIPUS group and the PS128 group, these changes were not statistically significant.

Figure 5J–L presents the results of α-diversity analysis based on 3 different indices. Figure 5J shows Observed Features, representing the actual number of microbial species observed in the sample, which directly measures species richness without considering evolutionary relationships. Compared to the Sham group, the MK-801 group exhibited a significant decrease in Observed Features (246.8 ± 53.57 vs. 176.5 ± 31.55, P < .05), with significant recovery following LIPUS treatment (176.5 ± 31.55 vs. 228.0 ± 31.41, P < .05). Figure 5K shows Faith’s PD index, which takes into account the phylogenetic relationships between species by measuring the length of the evolutionary tree and provides a more comprehensive assessment of diversity. The MK-801 group showed a significant reduction in Faith’s PD compared to the Sham group (10.39 ± 0.87 vs. 12.84 ± 1.47, P < .01), with significant recovery after LIPUS treatment (10.39 ± 0.87 vs. 11.91 ± 1.01, P < .05). Figure 5L shows Margalef’s Index, which integrates species richness and the number of individuals in the sample to reflect species abundance, with higher values indicating greater species richness. The MK-801 group showed a significant decline in Margalef’s Index compared to the Sham group (18.57 ± 3.38 vs. 25.08 ± 5.089, P < .05), with significant recovery after LIPUS treatment (18.57 ± 3.38 vs. 23.82 ± 3.08, P < .05). However, PS128 did not affect the reduction in all 3 indices.

Discussion

Previous research has shown that the administration of noncompetitive NMDAR antagonists, such as MK-801, can produce schizophrenia-like psychopathological symptoms. In this study, MK-801 was used to establish an animal model of schizophrenia that exhibited negative symptoms such as anhedonia and social withdrawal.³⁴ We investigated the effects of abdominal LIPUS, as a potential noninvasive treatment for MK-801-like negative symptoms, via modulation of serotonin signaling and gut microbiota. We showed that abdominal LIPUS stimulation could significantly improve gut microbiota diversity and reverse the E/I imbalance by increasing inhibitory proteins, thereby enhancing GABAergic transmission. In addition, LIPUS restored serotonin receptor density, further increasing the release of dopamine and serotonin, which leads to improvements in negative symptoms.

VGLUT1 is the predominant isotype, with the highest proportion and range of functions, representing the majority of excitatory glutamatergic terminals in the CNS.³⁵ VGAT is found on the vesicles in inhibitory terminals of both GABAergic and glycinergic neurons. The maintenance of a proper balance between VGLUT1 and VGAT is crucial for normal brain function.³⁶ The disruption of the E/I balance in PFC pyramidal neurons has been linked to various PFC-dependent behaviors, including cognitive functions, social interactions, and anxiety, and their alterations, particularly in psychiatric disorders.³⁷ A series of experiments suggest that the social deficits associated with an elevated E/I balance in mPFC cells may be partially alleviated by restoring balance through enhanced inhibitory signaling.³⁸ In a previous study of a rat model of schizophrenia induced by subchronic MK-801 injection, a long-lasting but not permanent E/I imbalance, characterized by reduced VGAT expression in the brain, was observed.³⁹ Our experimental results similarly showed a decrease in VGAT expression in the brain of the rat model subjected to subchronic intraperitoneal injection of MK-801, leading to an increased E/I ratio (Figure 2F–H). After LIPUS treatment, VGAT expression increased, thereby restoring the E/I balance. Therefore, we suggest that the social behavior deficits observed in this study are due to an elevated E/I ratio caused by reduced VGAT expression, which is consistent with the findings of the previous study. In contrast to subchronic MK-801 administration, acute MK-801 injection led to a significant increase in VGLUT1 expression in the mPFC, while VGAT levels remained unchanged. Consequently, the VGLUT1/VGAT ratio was markedly elevated in MK-801-treated rats. Notably, LIPUS stimulation effectively restored both VGLUT1 expression and the VGLUT1/VGAT ratio in the MK-801 + LIPUS group compared to the MK-801 group.²¹ Besides, the results showed that subchronic intraperitoneal injection of MK-801 reduced the expression of c-Fos on TH in the VTA, while after abdominal injection of LIPUS, the expression of c-Fos was significantly restored (Figure 4B). However, it was found that acute injection of MK-801 caused dopamine neurons to overexpress c-Fos, while prestimulation with LIPUS could inhibit MK-801-induced overactivation of dopamine neurons.²¹

Growing evidence suggests that the trillions of microbes residing in our gut play a significant role in mental health and contribute to the development of neuropsychiatric disorders.^40,41 Compared to the Sham group, in the MK-801 group, the number of unique bacterial species was reduced, indicating that MK-801 decreases the abundance of unique bacterial species in the rat gut (Figure 5A). MK-801 decreases the abundance of unique bacterial species in the rat gut primarily by disrupting the gut microbiota composition through its effects on the CNS and gastrointestinal physiology. As an NMDA receptor antagonist, MK-801 induces schizophrenia-like symptoms and affects multiple biological pathways that contribute to gut dysbiosis.⁴² After receiving abdominal LIPUS treatment, an increase in the number of unique species was observed, suggesting that abdominal LIPUS treatment effectively countered the reduction in unique gut microbiota caused by MK-801. In contrast, the PS128 treatment failed to improve this condition. The Proteobacteria and Firmicutes have been shown to significantly increase and decrease, respectively, in individuals with schizophrenia.⁴³ Consistent with previous studies, our results for the MK-801-treated rats showed an increase in the Proteobacteria and a decrease in the Firmicutes compared to the Sham group (Figure 5B). All indices of α-diversity showed significant reductions in the MK-801 group (all P < .05; Figure 5J–L), indicating that MK-801 had a negative impact on gut microbiota diversity. However, after LIPUS treatment, these indices significantly increased, returning to levels similar to those of the Sham group. This suggests that LIPUS treatment effectively restored the gut microbiota diversity reduced by MK-801, demonstrating its potential role in overriding changes in the gut microbiome. However, PS128 did not affect these changes. Lactobacillus plantarum PS128 was chosen for its well-documented psychobiotic properties, particularly its ability to modulate the gut–brain axis by influencing dopamine and serotonin levels, which are crucial for mood regulation and neurological function. Studies have demonstrated its potential benefits in conditions such as major depressive disorder and Parkinson’s disease.^44,45 The failure of the 5-day PS128 treatment to counteract the reduced unique gut microbiota caused by MK-801 may be attributed to the extensive dysbiosis induced by this NMDA receptor antagonist. MK-801 significantly disrupts microbial diversity, and a single probiotic strain administered over a short duration may be insufficient for restoration. More prolonged interventions, multistrain probiotics, prebiotics, or dietary modifications may be required to effectively rebalance gut microbiota.⁴⁶

The pathological mechanisms of schizophrenia are not yet fully understood. This study focused only on the interaction between the serotonin and dopamine systems, while the actual pathophysiology may involve other neurotransmission pathways. Therefore, the results of this study do not fully represent the complex pathological processes of schizophrenia. In addition, a limitation of this study is that it only compared the short-term effects of abdominal LIPUS stimulation with oral PS128 treatment, showing that LIPUS was more effective in improving behavioral abnormalities, regulating brain neural activity, and promoting neurotransmitter release than PS128. However, previous studies have shown that PS128 exhibits significant efficacy in influencing neurotransmitters and improving behavior over longer treatment courses (several weeks).^47,48 This study may not fully demonstrate the potential effects of PS128 due to the brevity of the latter’s application (5 days). Previous research has demonstrated that LIPUS stimulation of the abdominal region can mitigate DSS-induced acute colitis by activating the splenic nerve, thereby triggering the cholinergic anti-inflammatory pathway, with the vagus nerve likely playing a crucial role in this process.⁴⁹ Therefore, further research is needed to explore the underlying mechanisms of abdominal LIPUS in the treatment of schizophrenia. To date, there are few literatures exploring abdominal LIPUS stimulation, so the possible side effects remain to be explored. Another limitation of this study is the challenge in precisely identifying the underlying mechanisms of LIPUS treatment, as the ultrasound exposure was applied to the entire abdominal region of the mice. To address this, our team is currently developing an image-guided focused LIPUS approach to selectively target specific organs or regions for future investigations.

Abdominal LIPUS stimulation can significantly improve gut microbiota diversity and reverse the E/I imbalance by increasing inhibitory proteins, thereby enhancing GABAergic transmission. Furthermore, abdominal LIPUS restores serotonin receptor density, which in turn increases the release of dopamine and serotonin, alleviating negative symptoms. In summary, this study not only demonstrates the potential of LIPUS for improving schizophrenia by modulating neural activity, enhancing neurotransmitter release, and correcting gut microbiota dysregulation but also highlights its clinical value as a noninvasive physical therapy.

Funding

This study was supported by grants from the National Science and Technology Council of Taiwan (no. NSTC 113-2218-E-A49-031- and NSTC 111-2314-B-A49-045-MY3), the Far Eastern Memorial Hospital National Yang Ming Chiao Tung University Joint Research Program (no. 113DN22 and 114DN20).

Conflicts of Interest

None declared.

References

Author notes