Lateral Habenula 5-HT₂A Receptors Mediate Social Deficits in a Valproic Acid-Induced Rodent Model of Autism Spectrum Disorder Authors: Chi Zhang, Haosen Sang, Beilin Zhang, Zhiruo Zhao, Haoze Shang, Xiangxun Wang, Shuxin Luan, Jinpeng Wang Published: 2026 ABSTRACT AbstractAutism Spectrum Disorder (ASD) is characterized by social deficits, yet underlying neural mechanisms remain poorly understood. Using a valproic acid (VPA)-induced rodent model of ASD, we demonstrate that hyperactivity of lateral habenula (LHb) neurons, driven by upregulated 5-HT₂A receptors, mediates social impairments. VPA-exposed rats showed significant reductions in sociability, increased anxiety-like behaviors, and elevated repetitive actions. These behavioral changes coincided with enhanced LHb neuronal activity, including increased c-Fos expression and spontaneous firing rates. Transcriptomic and molecular analyses revealed upregulation of 5-HT₂A receptors in the LHb, while electrophysiology confirmed that 5-HT₂A activation augmented neuronal excitability through increased sEPSC frequency/amplitude and firing rates. Crucially, causal validation demonstrated that overexpressing 5-HT₂A receptors in wild-type rats induced ASD-like social deficits, whereas receptor knockdown in VPA-exposed rats reversed sociability impairments. Ablation of LHb neurons similarly rescued social behavior, confirming pathological LHb hyperactivity. Our findings identify LHb 5-HT₂A receptor-mediated hyperexcitability as a novel mechanism driving social dysfunction in ASD. Targeted suppression of this pathway may offer therapeutic strategies for core symptoms.Supplementary InformationThe online version contains supplementary material available at 10.1007/s12035-026-05950-5.Keywords: Autism Spectrum Disorder, Lateral habenula, Three-chamber social test, 5-HT2A receptor, Neuron firing Introduction ASD is a highly heterogeneous neurodevelopmental condition, with its core pathological features characterized by impaired social motivation and maladaptive behaviors. Its global prevalence continues to rise, imposing a substantial socioeconomic burden [1]. In China, the number of affected children aged 0–14 years is estimated at approximately 2 million, with a significantly higher incidence in males than in females [2]. Although behavioral interventions can partially alleviate symptoms, pharmacological therapies targeting core social deficits remain extremely limited [3]. This therapeutic dilemma stems from the complexity of ASD pathophysiology. Previous studies have predominantly focused on dysfunction within the “social brain” network, including the prefrontal cortex, amygdala, and superior temporal gyrus, or on the imbalance between glutamatergic and γ-aminobutyric acid (GABA)ergic systems [3, 4], while insufficient attention has been paid to the neural circuits and molecular regulators underlying social motivation. Understanding these mechanisms in model systems may provide insights into the neural basis of social deficits in ASD. The LHb, an evolutionarily conserved dorsal thalamic nucleus, functions as a pivotal hub integrating emotional, cognitive, and behavioral regulation. It receives input from limbic structures such as the prefrontal cortex, amygdala, and nucleus accumbens via the stria medullaris, and projects through the fasciculus retroflexus to monoaminergic nuclei including the ventral tegmental area (VTA) and dorsal raphe nucleus (DRN), thereby forming a bidirectional regulatory network over dopaminergic and serotonergic (5-HT) systems [5]. Clinical neuroimaging studies have reported increased habenular volume in patients with ASD, and machine learning models based on habenular morphology can predict ASD risk with high accuracy [6]. Animal studies further demonstrated that LHb hyperactivation suppresses reward-related pathways, thereby reducing social interaction [7]. Collectively, these findings suggest that LHb dysfunction may contribute to social deficits in ASD and related animal models, although the specific molecular mechanisms and causal relationships remain unresolved. Aberrant 5-HT signaling is a critical pathological hallmark of ASD. Early clinical studies observed significantly elevated peripheral 5-HT levels in children with autism [8], whereas central 5-HT signaling deficits have been directly linked to impaired social motivation [9]. At the molecular level, activation of 5-HT2A receptors enhances neuronal excitability through downstream pathways [10]. Notably, postmortem analyses revealed reduced 5-HT2A receptor binding in the prefrontal cortex of ASD patients [11], while aberrant receptor expression in animal models can induce ASD-like behaviors [12]. This apparent paradox underscores brain region–specific differences in 5-HT2A receptor function. The 5-HT2A receptor is also widely expressed in the LHb; however, whether its activity modulates neuronal excitability within the LHb and thereby contributes to ASD-related social impairments has not yet been clarified. In this study, we employed a prenatal VPA exposure paradigm, which produces rodents exhibiting ASD-like behavioral phenotypes, to investigate the neural mechanisms underlying social deficits. Through behavioral assays and molecular analyses, we demonstrate that excessive expression of 5-HT2A receptors in the LHb drives neuronal hyperexcitability and contributes to ASD-like social impairments in this model. These results not only provide a novel framework for understanding the neuropathology of ASD circuits but also establish an experimental foundation for targeted modulation of LHb activity to alleviate core symptoms. Materials and Methods Animals Sprague Dawley rats (males: 280–320 g; females: 240–260 g) were obtained from Changchun Yisi Experimental Animal Technology Co., Ltd. Rats were housed in individually ventilated cages (IVCs) with two animals per cage under standard conditions (ad libitum access to food and water). All animals were housed in a controlled environment with a temperature of 22 ± 1 °C, relative humidity of 50–60%, and a 12-h light/dark cycle (light on at 7:00 AM). Animals were maintained on a standard rodent chow diet and had free access to water. The housing room was kept quiet and free from external disturbances to minimize stress. Animals were acclimated to the facility for at least 7 days prior to any experimental procedures. Autism Spectrum Disorder Rat Model After one-week acclimatization, Sprague–Dawley rats (males > 280 g; females > 240 g) were paired overnight (1:1). Females with vaginal plugs or sperm-positive smears confirmed at 8:00 AM were designated at embryonic day 0.5 (E0.5) and individually housed. Multiple pregnant dams (n = 6–8 per group) were randomly assigned to receive intraperitoneal injection of either valproic acid (VPA, 600 mg/kg in saline) or equal-volume saline at E12.5. This dosage and timing were selected based on established protocols for inducing ASD-like behavioral phenotypes in rodent models. Offspring were weaned at postnatal day 21, and one male and one female pup were randomly selected from each litter for subsequent behavioral and molecular analyses. This sampling strategy ensured that litter was treated as the experimental unit of analysis, with individual offspring serving as technical replicates within each litter. The VPA-exposed and saline-exposed groups were designated as the VPA model group and control group, respectively, reflecting their status as an environmental risk factor model rather than a direct recapitulation of human ASD. However, upon detailed behavioral analysis (Supplementary Fig. 1), we observed that female offspring did not exhibit the hallmark ASD-like phenotypes—specifically, significant deficits in social interaction and repetitive behaviors—consistent with previous reports in the literature[13–15]. This finding prompted us to focus our subsequent experiments on male offspring, as they more reliably recapitulated the behavioral features of ASD in this model. This decision was made to ensure robust and reproducible behavioral outcomes for our mechanistic investigations. All experimental procedures were designed to minimize litter effects, with equal numbers of pups selected from each litter and balanced across experimental conditions. Statistical analyses incorporated litter as a random effect to account for potential within-litter correlations, following best practices for developmental neurotoxicity studies. Behavior Test Open Field Test (OFT) Rats were placed individually in a clean open-field arena (100 × 100 × 50 cm) for a 30-min session: 25-min habituation followed by 5-min test period. Between trials, the arena was thoroughly cleaned with 75% ethanol to eliminate odor cues and air-dried. Locomotor activity was recorded during the test phase using the Smart 3.0 Tracking System (Panlab), with total distance traveled and time spent in the central zone (50 × 50 cm) quantified by Smart 3.0 software. Three-Chamber Test (TCT) The three-chamber social test apparatus (90 × 60 × 35 cm) was divided into three equally-sized compartments (60 × 30 cm each) by transparent partitions with a central 70 × 70 mm passage. Wire mesh cages (diameter 20 cm) were placed in both side chambers. After a 5-min habituation period in the center chamber, sociability testing commenced: an unfamiliar rat (Stranger 1) was enclosed in one side cage while the opposite cage remained empty; subject-interaction time with Stranger 1 vs. empty cage was recorded for 10 min. For social novelty preference testing, a second unfamiliar rat (Stranger 2) was introduced to the previously empty cage, with interaction times recorded for both strangers over 10 min. The arena was thoroughly cleaned with 75% ethanol between trials to eliminate odor cues. Marble Burying Test (MBT) Individual rats were placed in clean cages (40 × 25 × 20 cm) containing 5-cm deep corncob bedding. Twenty glass marbles (Ø16 mm) were arranged in a 4 × 5 grid on the bedding surface. The number of marbles buried (> 2/3 covered by bedding) was recorded during a 10-min session. Stereotaxic Surgery and Microinjection Glass micropipettes (tip diameter < 50 μm) were backfilled with viral vectors and mounted onto a microinjection pump (Longer Pump). Under isoflurane anesthesia, rats were secured in a stereotaxic frame (RWD). Following scalp incision and skull exposure (disinfected with 10% povidone-iodine), bilateral craniotomies (Ø0.5 mm) were performed above the lateral habenula. Dura mater was punctured using a bent 25G needle. Pipettes were lowered to DV −4.9 mm, held for 10 min, then 300 nL virus was injected per side (100 nL/min flow rate) with 10-min post-injection dwell time. Scalps were sutured and disinfected, followed by subcutaneous 5% dextrose saline (5 mL) and intraperitoneal penicillin (50,000 IU). Rats recovered singly at 26 °C with behavioral tests commencing 3 weeks post-surgery. The AAV vector (serotype: AAV9, titer: 5 × 1012 vg/mL) was obtained from Wuhan Shumimi Brain Science and Technology Co., Ltd. A total volume of 250 nL per side was injected at the following coordinates: AP = 3.15 mm, ML = ± 0.9 mm, DV = −4.9 mm (relative to bregma). Location Verification Following behavioral tests, rats were anesthetized with isoflurane. For cannula placement verification, 100 nL pontamine sky blue was microinjected via cannula (50 nL/min). Brains were rapidly extracted and fixed in 4% PFA overnight, then sectioned coronally (400 μm) at LHb levels using a vibratome. Sections were PBS-hydrated and cannula tracks visualized under stereo microscopy. For viral expression analysis, transcardial perfusion was performed with saline followed by 4% PFA. Brains were post-fixed 24 h, cryoprotected in 10–30% sucrose gradients, embedded in OCT, and cryosectioned (4 μm). Fluorescence microscopy confirmed viral expression in LHb (Bregma −3.14 to −4.16 mm). Immunohistochemistry (IHC) Brains were fixed in 4% PFA for 24 h, paraffin-embedded, and sectioned coronally at LHb levels (3 μm). Sections underwent xylene-ethanol dewaxing/rehydration, antigen retrieval (121 °C, 2 min in citrate buffer, pH 6.2), and endogenous peroxidase blocking (3% H₂O₂, 25 min). After PBS washes, sections were blocked with 3% BSA (30 min), then incubated with primary antibodies (4 °C overnight in humid chamber). Following PBS washes, HRP-conjugated secondary antibodies were applied (50 min, RT). DAB development was monitored microscopically, with hematoxylin counterstaining (3 min) and differentiation (1% acid alcohol). Sections were dehydrated through graded ethanol-xylene series and mounted with neutral resin. LHb-positive cells were quantified using ImageJ. Western Blot (WB) LHb tissues were dissected from coronal sections (Bregma −2.70 to −4.20 mm) of anesthetized rats (1% pentobarbital, 80 mg/kg i.p.), snap-frozen, and homogenized in RIPA buffer containing 1 mM PMSF and protease inhibitors. Lysates were centrifuged (12,000 × g, 20 min, 4 °C), with supernatants quantified by BCA assay. Proteins (30 μg/lane) were separated on 10% SDS-PAGE gels and transferred to PVDF membranes (400 mA, 25 min, 4 °C). After blocking with 5% non-fat milk/TBST (1 h), membranes were incubated overnight at 4 °C with primary antibodies: anti-5-HT₂A receptor (1:1000, Abmart TA3077) and anti-GAPDH (1:1000, Abclonal AC033). Following TBST washes, HRP-conjugated secondary antibodies (1:10,000, Abclonal) were applied (1 h, RT). ECL detection was performed using a Tanon imaging system, with band intensities quantified by ImageJ. Reverse Transcription-Polymerase Chain Reaction (RT-PCR) Hb tissues were dissected from coronal brain sections of anesthetized rats (1% pentobarbital), snap-frozen, and homogenized in TransZol UP reagent (TransGen). Total RNA was extracted per manufacturer's protocol, including phase separation (12,000 × g, 15 min, 4 °C), isopropanol precipitation, and 75% ethanol washes. RNA quality was verified by A260/A280 ratio (> 1.8). cDNA was synthesized from 1 μg RNA using HiScript II Reverse Transcriptase (Vazyme) with gDNA wiper treatment. HTR2A amplification utilized specific primers (F: 5′-GGTCATCATGGCAGTGTCCC-3′; R: 5′-AAGGCCACCGGTACCCATAC-3′) in 20-μL reactions containing 2 × Taq Master Mix. Cycling parameters: 95 °C 3 min; 35 cycles of 95 °C 30 s, 60 °C 30 s, 72 °C 30 s; final extension 72 °C 5 min. PCR products were electrophoresed on 1.5% agarose/TAE gels (110 V, 30 min) and band intensities quantified using ImageJ. Brain Slice Preparation Following anesthesia induction with isoflurane, mice were rapidly decapitated, and heads were immersed in ice-cold, carbogen-saturated (95% O₂/5% CO₂) slicing solution containing (in mM): 92.0 NMDG, 26.0 NaHCO₃, 25.0 glucose, 20.0 HEPES, 10.0 NaCl, 10.0 MgSO₄, 2.5 KCl, 1.25 NaH₂PO₄, and 0.5 CaCl₂. Brains were dissected, trimmed, and mounted in a vibrating microtome (VT1200S, Leica, Germany). Coronal slices (300 μm thickness) containing the lateral habenula (LHb) were sectioned and collected. Slices were incubated in slicing solution at 32 °C for 10 min, then transferred to carbogen-saturated artificial cerebrospinal fluid (ACSF) containing (in mM): 116.0 NaCl, 26.0 NaHCO₃, 25.0 glucose, 2.5 KCl, 2.0 CaCl₂, 1.25 NaH₂PO₄, and 1.0 MgSO₄. After 30 min recovery at room temperature, electrophysiological recordings were initiated. Electrophysiology Recording pipettes (3–5 MΩ open-tip resistance) were pulled from borosilicate glass capillaries using a Sutter P-97 puller (Sutter Instrument, USA) and filled with internal solution containing (in mM): 115.0 potassium gluconate, 15.0 KCl, 10.0 Na₂-phosphocreatine, 10.0 HEPES, 4.0 Mg-ATP, 0.4 Na-GTP, and 0.1 EGTA (pH 7.3; osmolality 290–300 mOsm/kg). Whole-cell current-clamp recordings were performed using an EPC-10 amplifier (HEKA, Germany) with series resistance < 12 MΩ. Neuronal spontaneous activity was recorded without holding current. The 5-HT2A receptor agonist (Product name: 5-HT2A receptor agonist-1, MedChemExpress, USA; Catalog number: HY-145393; CAS: 2,698,331–34-7) was dissolved in DMSO to a stock concentration of 10 mM and diluted to a final concentration of 30 μM in the extracellular bath solution. The final DMSO concentration in the recording solution was < 0.1%. The agonist was bath-applied during electrophysiological recordings to assess its effects on LHb neuronal activity. Signals were sampled at 20 kHz via PatchMaster software (HEKA, Germany) and analyzed using Igor Pro 6.2 (WaveMetrics, USA). Statistical Analysis Data are presented as mean ± SEM. Normality was assessed by Shapiro–Wilk test. Parametric data used Student's t-test (two groups) or one-way ANOVA (multiple groups) with Tukey/Dunnett post hoc tests; non-parametric data employed Mann–Whitney U test (unpaired) or Wilcoxon test (paired). All tests were two-tailed, with sample sizes and specific tests indicated in figure legends. All statistical analyses were performed using linear mixed models with treatment as fixed effect and litter as random effect to account for the hierarchical structure of the data. Analyses used GraphPad Prism v9.3, with significance denoted as *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. Result Establishment of the ASD Model Through Prenatal VPA Exposure Offspring prenatally exposed to VPA exhibited marked somatic developmental abnormalities and distinct behavioral alterations, thereby establishing an autism spectrum disorder (ASD) model. The experimental timeline is illustrated in Fig. 1A. Rats in the VPA group displayed widespread tail malformations (Fig. 1B), consistent with previously reported[16, 17] phenotypes induced by VPA in rodent models of ASD. In the marble burying test (MBT) (Fig. 1C), the number of buried marbles was significantly increased in the VPA group (Fig. 1D), and both the duration of marble burying and self-grooming behaviors were prolonged compared to controls (Fig. 1E–F), indicating heightened repetitive and stereotyped behaviors characteristic of ASD-like phenotypes. In the open field test (OFT), no significant difference was observed between groups in total locomotor distance (Fig. 1G), indicating that VPA exposure did not impair baseline motor function. These findings support the use of the VPA model for studying social and repetitive behaviors without confounding motor deficits. In the three-chamber test (TCT) (Fig. 1H-I), during the sociability phase (phase I), VPA rats spent significantly less time interacting with the unfamiliar rat 1 (S1) compared with controls (Fig. 1J), and Social Preference Index (SPI)—calculated as (Time in Social Chamber)/(Time in Social Chamber + Time in Empty Chamber)—was also reduced (Fig. 1K). These findings demonstrate impaired overall social motivation in the VPA group. Collectively, these results confirm that the VPA-induced rat model effectively recapitulates the key behavioral features of ASD. Prenatal VPA exposure induces ASD-like phenotypes in rats. (A) Experimental timeline. (B) Tail malformations in VPA-exposed offspring. (C) MBT apparatus. (D) Quantification of the number of marbles buried in the MBT. VPA-exposed rats buried significantly more marbles than control rats (p < 0.01, unpaired t-test). (E) Quantification of the time spent burying marbles. VPA-exposed rats spent significantly more time engaged in marble burying behavior than control rats (p < 0.05, unpaired t-test). (F) Quantification of the time spent in self-grooming behavior. VPA-exposed rats exhibited significantly prolonged self-grooming compared to control rats (p < 0.01, unpaired t-test). (G) Quantification of total locomotor distance traveled in the OFT. No significant difference was observed between VPA-exposed and control rats (n.s., unpaired t-test). (H) Three-chamber test (TCT) apparatus. (I) Representative tracking paths of control (black) and VPA-exposed (red) rats during the sociability phase of the TCT, showing reduced interaction with the stranger rat (S1) in the VPA group. (J)Quantification of time spent in the interaction zone with the stranger rat (S1) during the sociability phase. VPA-exposed rats spent significantly less time interacting with S1 compared to control rats (p < 0.05, unpaired t-test). (K) Quantification of the SPI, calculated as: SPI = Time in Social Chamber/(Time in Social Chamber + Time in Empty Chamber). VPA-exposed rats exhibited a significantly lower SPI than control rats (p < 0.05, unpaired t-test). Data points represent individual animals. Data: mean ± SEM Enhanced Neuronal Activity in the LHb of VPA-exposed Rats To examine functional alterations of lateral habenula (LHb) neurons in the ASD rat model induced by prenatal VPA exposure, brain tissues were collected immediately after behavioral testing and subjected to c-Fos immunohistochemistry. The immunohistochemical analysis revealed a significant increase in c-Fos–positive neurons in the LHb of VPA-exposed rats compared with controls (Fig. 2A-B). This result suggests that VPA exposure may induce functional alterations in the LHb by enhancing neuronal activity, which is likely associated with the behavioral abnormalities observed in the VPA-exposed rats. Furthermore, electrophysiological recordings corroborated this conclusion, providing additional evidence for the involvement of LHb hyperactivity in the pathogenesis of ASD. Subsequently, we conducted electrophysiological experiments. Figure 2C shows the lateral habenula neurons used in the experiment. Figure 2D shows the original traces of spontaneous action potentials of in LHb neurons of both control (top) group and VPA-exposed rats group (bottom). The results show significant increase in spontaneous frequency (Fig. 2E) in VPA-exposed rats and increase in resting membrane potential (Fig. 2F) in VPA-exposed rats. There is no significant different between control and VPA-exposed rats group in spontaneous action potential slop (Supplementary Fig. 2A), amplitude (Fig. 2G), half width(Supplementary Fig. 2B), and no different in input resistance(g) of recorded neurons. There is no significant different in action potential threshold which also named Rheobase (Supplementary Fig. 2C). Moreover, electrophysiological data corroborated this finding, providing critical experimental evidence for further exploration of the role of the LHb in ASD pathogenesis. Hyperexcitability of LHb neurons in VPA-exposed rats. (A) Representative immunohistochemical (IHC) images showing c-Fos expression (brown staining) in the LHb of Con and VPA-exposed rats. Scale bar = 200 μm (overview) and 50 μm (insets). (B) Quantification of c-Fos-positive neurons in the LHb. VPA-exposed rats exhibited a significant increase in the number of c-Fos-positive neurons compared to control rats (p < 0.05, unpaired t-test). (C) LHb neuron location. (D) Representative original traces of spontaneous action potentials recorded from LHb neurons in control (top) and VPA-exposed (bottom) rats. Note the increased firing frequency in the VPA group. (E) Quantification of spontaneous firing frequency. VPA-exposed rats showed a significantly higher spontaneous firing frequency than control rats (p < 0.05, unpaired t-test). (F) Quantification of RMP. VPA-exposed rats exhibited a significantly depolarized RMP compared to control rats (p < 0.05, unpaired t-test). (G) Unchanged spike amplitude (n.s.). Data points represent individual animals. Data: mean ± SEM; unpaired t-test. (control: n = 7 from 3 rat; VPA: n = 9 from 3 rat) Viral Induction of LHb Neuronal Apoptosis Reverses Social Deficits in VPA-exposed Rats To verify whether the LHb is involved in social behavior in VPA-exposed rats, an apoptosis-inducing virus (rAAV-hSyn-taCasp3-T2A-TEVp-EGFP-WPRE-hGH-pA) or a control virus (rAAV-hSyn-taCasp3-T2A-EGFP-WPRE-hGH-pA) was bilaterally injected into the LHb. This construct serves as a neuron-specific apoptosis-inducing tool, with its core mechanism relying on activation of the Caspase 3 pathway by TEVp to trigger cell death. Three weeks after injection, NeuN expression in virus-infected regions was assessed through immunofluorescence staining (Fig. 3A). Furthermore, c-Fos immunohistochemistry of the LHb revealed that the number of c-Fos–positive neurons was significantly higher in the control virus group (VPA + EGFP) than in the apoptosis virus group (VPA + Casp3) (Fig. 3B、E), indicating that this viral tool effectively induced neuronal apoptosis within the LHb. LHb neuronal apoptosis rescues social deficits in VPA-exposed rats. (A) NeuN staining confirming neuronal loss after AAV-taCasp3. Scale bar = 200 μm. (B) Representative IHC images showing c-Fos expression (brown staining) in the LHb of VPA + EGFP and VPA + Casp3 rats. Scale bar = 200 μm (overview) and 50 μm (insets). (C) Representative trajectories from the open field test (OFT) for VPA + EGFP (black) and VPA + Casp3 (red) rats. (D) Quantification of total locomotor distance traveled in the OFT. No significant difference was observed between VPA + EGFP and VPA + Casp3 groups (n.s., unpaired t-test). (E) Quantification of c-Fos-positive neurons in the LHb. VPA + Casp3 rats exhibited a significantly lower number of c-Fos-positive neurons compared to VPA + EGFP rats (p < 0.01, unpaired t-test). (F) Representative tracking paths of VPA + EGFP (black) and VPA + Casp3 (red) rats during the sociability phase of the TCT, showing reduced interaction with the stranger rat in VPA + EGFP group. (G) Quantification of time spent in the interaction zone with the stranger rat (S1) during the sociability phase. VPA + Casp3 rats spent significantly more time interacting with S1 compared to VPA + EGFP rats (p < 0.05, unpaired t-test). (H) Quantification of the SPI during the sociability phase. VPA + Casp3 rats exhibited a significantly higher SPI than VPA + EGFP rats (p < 0.05) In terms of behavioral outcomes, OFT results (Fig. 3C-D) showed no significant difference in total locomotor distance between the VPA + EGFP and VPA + Casp3 groups, indicating that LHb apoptosis did not affect baseline motor function. In the TCT (Fig. 3F), VPA + Casp3 rats spent more time interacting with S1 (Fig. 3G) and exhibited a higher social preference index (Fig. 3H) than VPA + EGFP rats. These findings indicate that the LHb contributes to the manifestation of social deficits in VPA-exposed rats and that virus-induced neuronal apoptosis can reverse such impairments. Overexpression of 5-HT2A Receptors in the LHb of VPA-exposed Rats and Presynaptic 5-HT2A Receptor Activation Significantly Enhances Neuronal Firing To determine whether the elevated neuronal activity in the LHb of VPA-exposed rats was attributable to altered gene expression, transcriptomic sequencing was performed on LHb tissue. Differentially expressed genes (DEGs) were identified with a fold change (FC) > 1.4 and p < 0.05, revealing 525 DEGs in total, including 249 upregulated and 276 downregulated genes (Fig. 4A). Gene Ontology (GO) functional enrichment analysis of the upregulated genes revealed significant enrichment in biological processes related to synaptic transmission, neurotransmitter release, and receptor signaling, including "presynaptic modulation of chemical synaptic transmission," "neuropeptide signaling pathway," and "G protein-coupled serotonin receptor activity" (Fig. 4B). In contrast, downregulated genes were primarily enriched in pathways associated with extracellular matrix organization, apoptosis, and ion channel function, including "positive regulation of bone mineralization," "negative regulation of ossification," and "voltage-gated monoatomic ion channel activity" (Fig. 4C). Notably, our RNA-seq analysis revealed that HTR2A is one of the most significantly upregulated genes in the LHb of VPA-treated mice. Furthermore, GO enrichment analysis confirmed that genes involved in G protein-coupled receptor signaling and synaptic transmission were significantly enriched in the VPA group, strongly implicating HTR2A in the pathophysiology of VPA-induced social deficits. RT-PCR analysis confirmed increased HTR2A gene expression in the VPA group relative to controls (Fig. 4D-E), consistent with RNA-seq findings. Western blotting further demonstrated elevated 5-HT2A receptor protein levels in VPA rats (Fig. 4F-G), in agreement with the RT-PCR results. 5-HT₂A receptor upregulation drives LHb hyperexcitability in VPA-exposed rats. (A) Volcano plot of DEGs in the LHb of VPA-exposed rats versus controls (VPA vs. Con; |log₂FC|> 1.4, p < 0.05). Red dots represent significantly upregulated genes, blue dots represent significantly downregulated genes, and gray dots represent non-significant genes. (B-C) GO enrichment analysis of upregulated (B) and downregulated (C) DEGs. Top enriched biological processes are shown, highlighting pathways related to synaptic transmission, receptor signaling, and ion channel function. (D) Representative gel electrophoresis image showing amplified HTR2A and GAPDH (loading control) bands. (E) Quantification of HTR2A mRNA levels normalized to GAPDH. VPA-exposed rats exhibited significantly higher HTR2A mRNA expression than controls (**p < 0.001, unpaired t-test). (F) Representative Western blot image showing 5-HT₂A receptor and GAPDH (loading control) bands. (G) Quantification of 5-HT₂A protein levels normalized to GAPDH. VPA-exposed rats exhibited significantly higher 5-HT₂A protein levels than controls (p < 0.05, unpaired t-test). (H) Experimental protocol for electrophysiological recordings. A 5-HT₂A receptor agonist (30 nM) was bath-applied for 170 s (indicated by the bar) to LHb neurons recorded in perforated voltage-clamp mode at a holding potential of −70 mV. (I) Quantification of sEPSC frequency. (J) Quantification of sEPSC amplitude. Application of the agonist significantly increased both sEPSC frequency and amplitude in LHb neurons (p < 0.05, unpaired t-test for both). (K) Quantification of spontaneous firing rate. (L) Quantification of RMP. Application of the agonist significantly increased the spontaneous firing rate (p < 0.05, unpaired t-test), while the change in RMP did not reach statistical significance (p > 0,05, unpaired t-test). (M) Representative traces of sEPSCs before and during agonist application. (N) Quantification of sEPSC frequency under TTX. The 5-HT₂A receptor agonist significantly increased sEPSC frequency even in the presence of TTX (p < 0.05, unpaired t-test). (O) Quantification of spontaneous firing rate under TTX. The agonist also significantly increased the firing rate of LHb neurons under TTX (p < 0.05, unpaired t-test), confirming a direct postsynaptic mechanism of action. Data points represent individual animals. Data: mean ± SEM. (n = 6 from 6 rat) To further confirm the cellular localization of HTR2A in the LHb, immunofluorescence co-staining for HTR2A and the neuronal marker NeuN was performed. As shown in Supplemental Fig. 3, HTR2A was predominantly expressed in neurons within the LHb, indicating that the upregulation of 5-HT2A receptors occurs primarily in neuronal populations. Importantly, the number of HTR2A-positive neurons was significantly increased in the VPA group compared to the control group, consistent with the elevated mRNA and protein levels observed in RT-PCR and Western blot analyses. We hypothesized that the upregulation of 5-HT2A receptors was responsible for altered firing activity in LHb neurons. To test this, a 5-HT2A receptor agonist was perfused, and changes in action potentials of LHb neurons were recorded (Fig. 4H).Cells were recorded in perforated voltage clamp mode at a holding potential of −70 mV. Serotonin 2 A Receptor Agonist was applied for 170 s (indicated by the bar).Original traces of currents induced by applications of 30 nM Serotonin 2 A Receptor Agonist extracellular. 30 nM Serotonin 2 A Receptor Agonist application significantly increased spontaneous EPSCs frequency (Fig. 4I) and amplitude (Fig. 4J). Also, Increase in membrane excitability of LHb neurons caused by Serotonin 2 A Receptor Agonist. Spontaneous action potentials were recorded in perforated current clamp mode. Serotonin 2 A Receptor Agonist was applied for 170 s (indicated by the bar).Original traces of membrane potential and spontaneous action potentials changes induced by applications of 30 nM Serotonin 2 A Receptor Agonist extracellular. Application of the 30 nM 5-HT₂A receptor agonist significantly increased spontaneous action potential frequency (Fig. 4K), with a trend toward depolarization of resting membrane potential (RMP) that did not reach statistical significance (Fig. 4L). These results suggest that VPA exposure increases neuronal activity in the LHb, primarily through elevated firing rates, which may underlie the functional changes associated with ASD-like behaviors. We also conducted additional experiments using tetrodotoxin (TTX) to block action potential-dependent neurotransmitter release. In the presence of TTX (1 μM), which effectively abolished most presynaptic excitatory events, we observed that the 5-HT₂A receptor agonist (30 nM) still significantly increased the frequency of spontaneous excitatory postsynaptic currents (sEPSCs) and the firing rate of LHb neurons (Fig. 4M–O). This persistent postsynaptic excitation, despite the blockade of presynaptic activity, indicates that the primary mechanism of 5-HT₂A receptor-mediated neuronal excitation in the LHb is postsynaptic. Upregulation of 5-HT2A Receptors in the LHb of Wild-type Rats Mimics ASD-like Social Deficits To investigate whether overexpression of the 5-HT2A receptor induces social deficits by altering LHb neuronal activity, an AAV vector overexpressing HTR2A (rAAV-CaMKⅡα-HTR2A-EGFP-WPREs-pA) or a control virus (rAAV-CaMKⅡα-EGFP-WPREs-pA) was bilaterally injected into the LHb of wild-type (WT) rats. Expression of the viral construct was confirmed in the LHb by frozen sectioning (Fig. 5A) and Western blotting, which validated correct overexpression of the 5-HT2A receptor (Fig. 5B-C). EGFP expression was detectable by 21 days post-injection, allowing for behavioral and electrophysiological assessments at this time point.We performed immunohistochemistry for c-Fos and analyzed its colocalization with EGFP. As shown in Fig. 5D, the number of c-Fos-positive neurons within the EGFP-positive population was significantly higher in the HTR2A-overexpressing group compared to the EGFP control group, indicating that 5-HT₂A receptor overexpression enhances neuronal activity specifically in the targeted neurons(Fig. 5G). LHb 5-HT₂A overexpression induces ASD-like behaviors in WT rats. (A) Schematic illustration of the experimental approach. A recombinant adeno-associated virus (rAAV-CaMKⅡα-HTR2A-EGFP-WPREs-pA) was bilaterally injected into the LHb of WT rats to overexpress the HTR2A fused with EGFP. The viral construct was driven by the CaMKⅡα promoter to target excitatory neurons. (B) Representative Western blot image showing 5-HT₂A receptor and GAPDH (loading control) protein bands in the LHb of WT rats injected with either a control virus (WT + EGFP) or the HTR2A-overexpressing virus (WT + HTR2A). (C) Quantification of 5-HT₂A protein levels normalized to GAPDH. WT rats overexpressing HTR2A exhibited significantly higher 5-HT₂A protein levels than WT + EGFP controls (p < 0.05, unpaired t-test). (D) Representative immunofluorescence images of the LHb in WT + EGFP and WT + HTR2A rats. Co-staining for DAPI (blue, nuclei), EGFP (green, viral expression), and c-Fos (red, neuronal activation marker). The merged image (right panel) shows increased co-localization of EGFP and c-Fos (yellow) in the WT + HTR2A group, indicating enhanced activity specifically in the HTR2A-overexpressing neurons. Scale bar = 100 μm (overview) and 20 μm (insets). (E) Representative tracking paths of VPA + EGFP (black) and VPA + Casp3 (red) rats during the sociability phase of the TCT, showing reduced interaction with the stranger rat in WT + HTR2A group. (F) Quantification of total locomotor distance traveled in the OFT. No significant difference was observed between WT + EGFP and WT + HTR2A groups (n.s., unpaired t-test). (G) Quantification of c-Fos-positive cells among EGFP-expressing neurons in the LHb. WT + HTR2A rats exhibited a significantly higher number of c-Fos⁺/EGFP⁺ cells compared to WT + EGFP rats (p < 0.05, unpaired t-test). (H) Representative tracking paths from the TCT for WT + EGFP (black) and WT + HTR2A (red) rats during the sociability phase. The red trajectories in the WT + HTR2A group show markedly reduced interaction with the stranger rat (S1) compared to the black trajectories in the control group. (I) Quantification of time spent in the interaction zone with the stranger rat (S1) during the sociability phase. WT + HTR2A rats spent significantly less time interacting with S1 compared to WT + EGFP controls (*p < 0.01, unpaired t-test). (J) Quantification of the SPI during the sociability phase. No significant difference was observed between the two groups (n.s., unpaired t-test). Data: mean ± SEM Behavioral analyses further supported these findings. In the OFT, no difference in total locomotor distance was observed between the HTR2A and EGFP groups (Fig. 5E-F), indicating that viral manipulation did not affect general motor function.. In the first phase of the TCT (Fig. 5H), HTR2A rats spent significantly less time interacting with S1 compared with controls (Fig. 5I). However, no significant difference was observed in the social preference index (Fig. 5J). These results suggest that overexpression of 5-HT₂A receptors in the LHb induces social deficits, likely through enhanced neuronal activity in this region. Downregulation of 5-HT2A Receptors in the LHb of VPA-exposed Rats Reverses Social Deficits To further determine whether altered 5-HT2A receptor expression is sufficient to account for social deficits, an shRNA AAV targeting HTR2A (rAAV-hSyn-mCherry-5’miR-30a-shRNA(HTR2A)−3’miR-30a-WPRE) was bilaterally injected into the LHb of VPA-exposed rats (VPA + shRNA group). Frozen sectioning (Fig. 6A) and Western blotting (Fig. 6B-C) confirmed correct expression of the viral construct and effective suppression of HTR2A in the LHb. Immunohistochemistry revealed a significant reduction in c-Fos expression within the EGFP-positive (shRNA-expressing) neurons in the LHb of VPA + shRNA rats compared to VPA-only rats (Fig. 6D), indicating that HTR2A knockdown attenuates the elevated neuronal activity induced by VPA exposure. 5-HT₂A knockdown in LHb rescues ASD-like phenotypes. A) Schematic illustration of the experimental approach. A recombinant adeno-associated virus (rAAV-hSyn-mCherry-5’miR-30a-shRNA(HTR2A)−3’miR-30a-WPRE) was bilaterally injected into the LHb of VPA-exposed rats to knock down the HTR2A expression. The viral construct was driven by the hSyn promoter to target neurons, with mCherry serving as a fluorescent reporter. (B) Representative Western blot image showing 5-HT₂A receptor and GAPDH (loading control) protein bands in the LHb of VPA-exposed rats injected with either a control virus (VPA + EGFP) or the shRNA virus (VPA + shRNA). (C) Quantification of 5-HT₂A protein levels normalized to GAPDH. VPA + shRNA rats exhibited significantly lower 5-HT₂A protein levels compared to VPA + EGFP controls (p < 0.05, unpaired t-test). (D) Representative immunofluorescence images of the LHb in VPA + EGFP and VPA + shRNA rats. Co-staining for DAPI (blue, nuclei), EGFP/mCherry (green/red, viral expression), and c-Fos (red/green, neuronal activation marker). The merged image (right panel) shows a marked reduction in c-Fos expression (green) within the virally transduced (EGFP/mCherry-positive) neurons in the VPA + shRNA group compared to the VPA + EGFP group, indicating attenuated neuronal hyperactivity following HTR2A knockdown. Scale bar = 100 μm (overview) and 20 μm (insets). (E) Representative tracking trajectories from the open field test (OFT) for VPA + EGFP (black) and VPA + shRNA (red) rats. (F) Quantification of total locomotor distance traveled in the OFT. No significant difference was observed between VPA + EGFP and VPA + shRNA groups (n.s., unpaired t-test), indicating that HTR2A knockdown does not affect baseline motor function. (G) Quantification of c-Fos-positive cells among EGFP/mCherry-expressing neurons in the LHb. VPA + shRNA rats exhibited a significantly lower number of c-Fos⁺/EGFP⁺ cells compared to VPA + EGFP rats (*p < 0.01, unpaired t-test). (H) Representative tracking paths from the (TCT for VPA + EGFP (black) and VPA + shRNA (red) rats during the sociability phase. The red trajectories in the VPA + shRNA group show markedly increased interaction with the stranger rat (S1) compared to the black trajectories in the control group. (I) Quantification of time spent in the interaction zone with the stranger rat (S1) during the sociability phase. VPA + shRNA rats spent significantly more time interacting with S1 compared to VPA + EGFP controls (p < 0.05, unpaired t-test). (J) Quantification of the SPI during the sociability phase. VPA + shRNA rats exhibited a significantly higher SPI than VPA + EGFP controls (**p < 0.001, unpaired t-test). Data: mean ± SEM Behaviorally, OFT results showed that no significant difference in total locomotor distance was observed between the VPA + shRNA and VPA groups (Fig. 6E-F), indicating that the two groups exhibited comparable motor function. In the TCT (Fig. 6H), the VPA + shRNA group showed significantly longer interaction time with S1 during the first phase and a higher social preference index compared to VPA rats (Fig. 6I-J). These results demonstrate that downregulation of 5-HT₂A receptors in the LHb rescues social deficits induced by VPA exposure. The reduced neuronal activity in the LHb of the VPA + shRNA group suggests that HTR2A knockdown normalizes hyperactivity in this region, thereby rescuing social deficits induced by VPA exposure. Discussion ASD, as a neurodevelopmental condition, imposes a substantial socioeconomic burden due to its core features of social deficits and repetitive stereotyped behaviors. Although gene–environment interactions are recognized as the fundamental basis of pathogenesis, current interventions provide only limited improvement in core symptoms. Traditional theories have primarily emphasized abnormalities within the “social brain” network or excitation/inhibition imbalance, while the neural regulation of social motivation—an essential component—has often been overlooked [3]. Therefore, it is imperative to elucidate the neuropathological mechanisms underlying social motivation deficits in ASD through targeted experimental studies, thereby providing potential therapeutic targets and strategies for precise treatment of ASD-related social impairments. LHb, with its unique circuit architecture, integrates inputs from the prefrontal cortex and amygdala and regulates monoaminergic outputs [18]. It encodes negative emotional states, mediates reward avoidance, and plays integrative roles in social interaction, pain perception, and sleep–wake regulation. Under physiological conditions, social interaction is perceived as a positive rewarding stimulus, whereas hyperactivation of the LHb may transmit “negative” signals to the dopaminergic and serotonergic systems, thereby diminishing the motivational and hedonic value of social engagement. Animal studies have shown that LHb hyperactivity suppresses dopaminergic neurons in the VTA and serotonergic neurons in the DRN, leading to anhedonia and sustained negative affect, whereas suppression of LHb activity significantly alleviates such symptoms [19, 20]. These findings suggest that LHb hyperexcitability may directly contribute to ASD-like social deficits. Clinical neuroimaging further supports this association, as magnetic resonance imaging (MRI) studies in ASD patients have revealed increased habenular volume. This provides the first evidence at the human level implicating the habenula in ASD pathology, while machine learning models based on bilateral habenular volume can predict ASD risk with high accuracy [21]. Together, these findings highlight functional abnormalities of the habenula as a potential neuropathological basis for social motivation deficits in ASD, positioning it as a candidate neural target for intervention in core symptoms. The VPA-induced ASD rat model is a widely used experimental paradigm, simulating prenatal environmental exposure that disrupts neurodevelopment in offspring and consistently reproduces the core behavioral features of autism [22]. In the present study, both the three-chamber social test and marble burying test demonstrated decreased sociability and enhanced stereotyped behaviors in VPA-exposed rats, confirming the validity of this model for mimicking human ASD core symptoms. Immunohistochemical and electrophysiological analyses revealed significantly elevated c-Fos expression and increased firing frequency of LHb neurons in VPA-exposed rats compared with controls. Collectively, these results suggest that hyperexcitability of the LHb may contribute to the formation of social impairments in VPA-exposed rats. Subsequently, to confirm the role of the LHb in ASD-related social deficits, adeno-associated viral vectors were used to induce LHb neuronal apoptosis. The three-chamber test revealed that apoptosis of LHb neurons effectively alleviated social deficits in VPA-exposed rats, with significant improvements observed in both sociability (phase I) and social preference (phase II) compared with controls. Notably, no significant differences were detected in total locomotor distance during the open field test, indicating that selective ablation of LHb neurons specifically ameliorated autism-like social deficits without producing nonspecific effects on motor function. Transcriptomic analysis of the LHb showed HTR2A was among the most significantly upregulated genes, and its enrichment in receptor signaling pathways strongly implicates it in VPA-induced social deficits. The regulation of reward-related outcomes by the LHb depends on its interaction with the serotonergic system [20]. In ASD patients, social motivation deficits are manifested as diminished social reward experiences, while responses to other types of rewards remain relatively preserved, consistent with this mechanism. The results of transcriptomic analysisresult were validated by RT-PCR and Western blotting. Both ASD patients and animal models exhibit abnormalities in the serotonergic system [23], and functional mutations, haplotypes, and upregulation of HTR2A (in animal models) are strongly associated with autism risk. Furthermore, its polymorphisms have been linked to social phobia and withdrawal behaviors [24], suggesting that this receptor may contribute to social avoidance by modulating serotonergic signaling. We hypothesized that increased 5-HT2A receptor signaling leads to LHb neuronal hyperexcitability, thereby mediating reduced social motivation in VPA-exposed rats (Fig. 7). Proposed mechanism: LHb 5-HT₂A receptors drive social deficits in VPA-exposed rats Whole-cell patch-clamp recordings demonstrated that 5-HT2A receptor activation significantly enhances neuronal firing in the LHb, with this effect being predominantly mediated through postsynaptic mechanisms. This electrophysiological evidence strongly supports our hypothesis that 5-HT2A receptor-driven modulation of LHb neuronal activity plays a causal role in social motivation deficits observed in ASD. The specificity of this mechanism is further corroborated by previous studies showing that pharmacological blockade of 5-HT2A receptors reduces both excitatory synaptic transmission and firing rates in LHb neurons [25], which aligns with our understanding that downregulation of 5-HT2A signaling would dampen neuronal hyperexcitability. These converging findings establish a clear mechanistic link between 5-HT2A receptor function in the LHb and the neural circuitry underlying social behavior, suggesting that targeted modulation of this receptor system could represent a promising therapeutic approach for addressing social motivation deficits in ASD. To further validate this mechanism, adeno-associated viral vectors were employed to overexpress 5-HT2A receptors specifically in the LHb of WT rats. The overexpression group exhibited increased anxiety-like behavior and significantly reduced initial sociability, while the social preference index remained unchanged. This indicates that aberrant upregulation of LHb 5-HT2A receptors is sufficient to induce ASD-like social withdrawal behaviors, although it may not directly impair social discrimination ability, implying that the mechanism is more closely related to motivational regulation rather than cognitive deficits. Critically, in VPA-exposed rats, knockdown of LHb 5-HT2A receptors completely normalized aberrant social behaviors, including contact time and preference index, without inducing locomotor abnormalities. This demonstrates that VPA-induced social deficits are directly dependent on pathological overexpression of 5-HT2A receptors in LHb neurons. We propose that 5-HT2A receptors in LHb neurons regulate excitability to dynamically maintain social motivation and sustain normal social behaviors. Pathological upregulation of these receptors disrupts this balance, constituting a core mechanism of ASD-related social deficits, while targeted inhibition of the receptor can specifically reverse these impairments. It is important to acknowledge the limitations of this study. First, we did not stratify electrophysiological or behavioral analyses by HTR2A expression levels or neuronal identity. This limits our ability to determine whether the observed effects are driven by a specific subpopulation of HTR2A-enriched neurons or reflect a global shift in LHb excitability—a critical consideration for future mechanistic dissection. Second, due to the constraints of our experimental design, we were limited in the number of litters available for analysis. While we have incorporated litter as a random effect in our statistical models, future studies with larger litter numbers would provide more robust estimates of between-litter variability and strengthen the generalizability of our findings. These considerations are important when interpreting the translational potential of our results. Conclusion This study is the first to demonstrate that 5-HT2A receptors in the LHb enhance neuronal excitability through a postsynaptic mechanism, thereby driving the core pathological process underlying social deficits in VPA-Induced Rodent Model. By using a VPA-exposed rat model showing ASD-like phenotypes, we confirmed across multiple experimental dimensions that neuronal activity in the LHb was markedly increased, while virus-induced LHb neuronal apoptosis specifically reversed social impairments. At the molecular level, transcriptomic and biochemical analyses validated abnormal upregulation of 5-HT2A receptors in the LHb, and whole-cell patch-clamp recordings further revealed that receptor agonists enhanced neuronal firing through presynaptic regulation. Crucially, causal verification showed that overexpression of 5-HT2A receptors in the LHb of WT rats induced ASD-like social withdrawal, whereas receptor knockdown in VPA-exposed rats fully restored social function. These findings transcend the limitations of traditional “social brain” and excitation/inhibition imbalance hypotheses, establishing a novel mechanistic pathway—“LHb → 5-HT2A → neuronal hyperexcitability → social deficits”—and providing a precise interventional strategy targeting LHb activity to alleviate core symptoms of ASD. Supplementary Information Below is the link to the electronic supplementary material. (PDF 316 KB) Author Contribution Chi Zhang: Conceptualization, Investigation, Funding acquisition, Writing – original draft. Haosen Sang: Methodology, Investigation, Data curation, Validation. Beilin Zhang: Investigation (behavioral assays), Visualization, Data curation. Zhiruo Zhao: Resources, Project administration, Animal model establishment. Haoze Shang: Investigation (electrophysiology), Software, Formal analysis. Xiangxun Wang: Investigation (molecular biology), Methodology, Validation. Shuxin Luan: Supervision, Writing, Project administration– review & editing. Jinpeng Wang: Conceptualization, Supervision, Project administration, Writing, Funding acquisition, – review & editing. All authors approved the final manuscript. Funding This work was supported by the Jilin Provincial Natural Science Foundation (YDZJ202301ZYTS094) and the Youth Development Fund of The First Hospital of Jilin University (JDYY16202513). Data Availability The data that support the findings of this study are available from the corresponding author upon reasonable request. Declarations Ethics All experimental procedures were approved by the Animal Ethics Committee of Jilin University College of Basic Medical Sciences (Approval ID: 2025397). Clinical Trial Number Not applicable. Conflict of Interest The authors declare no competing interests. Footnotes Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Contributor Information Shuxin Luan, Email: luansx@jlu.edu.cn. Jinpeng Wang, Email: jinpeng@jlu.edu.cn. References Associated Data Supplementary Materials (PDF 316 KB) Data Availability Statement The data that support the findings of this study are available from the corresponding author upon reasonable request.