Sex-dependent effects of imperatorin on autistic-like behaviors and BDNF signaling in rats prenatally exposed to valproic acid

Abstract

Autism spectrum disorder (ASD) is defined by cognitive impairments such as learning and memory decline, and social interaction and communication deficits. This study investigated the potential effects of Imperatorin (IMP), a furocoumarin and a phytochemical isolated from the leaves of Ferulago trifida on valproic acid (VPA)-induced behavioral and brain-derived neurotrophic factor (BDNF) alterations in a rat model of ASD. Open field test (locomotion, anxiety), grooming test, marble burying test (OCD-like behavior), novel object recognition memory test, and real-time PCR (BDNF levels) were conducted to collect data. The results showed increased locomotion in ASD-like rats in both sexes with greater effect in females, while IMP restored it in males and decreased it in females. Anxiety was increased in ASD-like rats in both sexes with greater effect in females; IMP had no effect. Grooming was increased in ASD-like rats in both sexes with greater effect in females, while IMP restored it in males and decreased it in females. Marbles buried was increased in ASD-like rats in both sexes, while IMP decreased it in males. Memory was impaired in ASD-like rats in both sexes; IMP had no effect. BDNF was decreased in ASD-like rats in both sexes with greater effect in females, while IMP increased it in both sexes. In conclusion, our findings showed hyperlocomotion, anxiety-like behavior, OCD-like behavior, memory deficit, and BDNF downregulation as ASD-related features, with evidence of sex-specific patterns and partial therapeutic effects of IMP.

Introduction

Autism spectrum disorder (ASD) is defined as a diverse neurodevelopmental disability characterized by deficits in social interaction and communication, often associated with repetitive or stereotyped behaviors, abnormal sensory processing, and restricted interests (Lord et al., 2018). The prevalence of autism is steadily increasing, and in the United States, the latest estimate of the prevalence of ASD is 1 in 36 eight-year-old children, making it a major health problem (Grosvenor et al., 2024). Affected individuals are found across all socioeconomic and racial backgrounds (Shaw et al., 2023). ASD is more common in males, leading to the concept of a “female protective effect.” (Lord et al., 2018, Shaw et al., 2023). Boys are 4.2 times more likely to develop autism than girls (Maenner et al., 2023); however, a recent study has shown an increase in the prevalence of autism in females (Harrop et al., 2024). The etiology of autism is multifactorial, with genetic, epigenetic, and environmental origins implicated in the development of the condition (Lu et al., 2022). A number of environmental characteristics, including maternal age, prenatal and postnatal issues such as hypoxia and birth trauma, exposure to toxic metals, maternal infection, air pollution, endocrine-disrupting chemicals, and valproic acid (VPA) administration, are considered as risk factors for ASD (Modabbernia et al., 2017).

Among animal models of autism, VPA administration is widely applied to initiate ASD-like symptoms by disrupting normal trajectories of fetal brain development. A recent study has shown that prenatal administration of VPA (600 mg/kg, 12.5 gestational day) leads to autistic-like behaviors in offspring (Hosseini et al., 2025). There are other studies using this model (Ay et al., 2024, Tuncak et al., 2025). One of the most studied pathobiological reasons that leads to such neurodevelopmental abnormalities is the alteration of brain-derived neurotrophic factor (BDNF) levels (Almeida et al., 2014, Saghazadeh and Rezaei, 2017). Neurotrophic factors, particularly BDNF are significantly contributed to neurogenesis, neuroplasticity, and synaptogenesis, and are unavoidably linked to the regulation of typical, abnormal, and neuropsychiatric-related behaviors (Abbasi et al., 2024b, Motamedi-Manesh et al., 2025, Samizadeh et al., 2024, Vaseghi et al., 2023). Due to the crucial role of BDNF in modulating neuroplasticity, BDNF dysfunction is implicated in the pathophysiology of ASD (Hosseini et al., 2025). Neuroplasticity changes may arise from disturbances in synaptic function and in activities of neurotransmitter systems (Takumi et al., 2020), while BDNF has a significant role in the modulation of synaptic functions and neurotransmitter systems (Vaseghi et al., 2023). Meta-analyses have uncovered a positive relationship between autism and increased blood BDNF levels (Liu, S.H. et al., 2021; Saghazadeh and Rezaei, 2017). Increases in the total BDNF levels have also been revealed in the fusiform gyrus of individuals with ASD (Garcia et al., 2012). It has also been shown that BDNF-induced increases in neurogenesis and axonal growth may lead to diverse effects on autistic-like behaviors (Ilchibaeva et al., 2023). On the other hand, other reports have shown reduced serum BDNF levels in autistic patients (Francis et al., 2018, Skogstrand et al., 2019). Our recent preclinical study has shown decreased BDNF levels in the medial prefrontal cortex of autistic-like rats exposed to maternal VPA (Hosseini et al., 2025). It has been reported that prenatal VPA administration dramatically reduces hippocampal BDNF level in the offspring, while treatment with nimodipine improves social performance and elevates BDNF level (Kumar et al., 2022). Treatment with rapamycin also improves social behavior by elevating hippocampal BDNF level in rats exposed to prenatal VPA (Zhang et al., 2017). Furthermore, previous study has shown that induction of BDNF overexpression in the hippocampus and the frontal cortex of BTBR mice (a rodent model of ASD) reduces anxiety and the severity of stereotyped behaviors, and improves social interest, respectively (Kaminskaya et al., 2024). Therefore, increased BDNF may be beneficial for the alleviation of ASD-like symptoms in rodent models of ASD.

Imperatorin (IMP) [9-(3-methylbut-2-enyloxy)-7H-furo[3,2-g] chromen-7-one] is a naturally occurring furanocoumarin that possesses various medicinal features and has been found in plants, fruits, and vegetables such as Angelica dahurica, Angelica archangelica (Apiaceae), and celery (Apium graveolens L) (Baek et al., 2000). IMP has pharmacological properties, including anti-inflammatory, neuroprotective, antioxidant, antidepressant, and anticonvulsant effects, and acts as an antagonist of calcium channel and GABA transaminase inhibitor (Cao et al., 2017, Cao et al., 2014, He et al., 2007, Huang et al., 2012, Liao et al., 2023, Luszczki et al., 2007). IMP has also been found to be influential in the function of acetylcholine (Ach), the necessary neurotransmitter for cognitive processes (Fylaktakidou et al., 2004, Kontogiorgis et al., 2006). According to earlier findings, IMP ameliorates memory (Budzynska et al., 2015, Chowdhury et al., 2018) and learning (Zheng et al., 2020) deficits. An experimental model of Alzheimer’s disease has revealed that IMP mitigates the impairment of memory formation and consolidation induced by scopolamine (Budzynska et al., 2015). Furthermore, IMP significantly ameliorates lipopolysaccharide (LPS)-induced memory deficits via upregulating BDNF levels (Chowdhury et al., 2018). IMP has been found to reverse cognitive impairments induced by prenatal stress, and this effect is related to heightened hippocampal BDNF (Zheng et al., 2020); this study has shown that IMP treatment significantly increases BDNF, TrkB, cAMP-response element binding (CREB) mRNA expression levels in the hippocampus of prenatally-stressed female offspring, leading to learning and memory improvements (Zheng et al., 2020). Therefore, IMP may affect the function of BDNF; although evidence is very limited.

Since BDNF alterations are commonly observed in the pathophysiology of ASD, it would be interesting to investigate the possibility that IMP may be associated with ASD-like symptoms through BDNF function. Given that there are no related studies on the association of IMP with the amelioration of ASD-like symptoms, the present study has been designed to evaluate its efficacy in ameliorating ASD-like behavioral changes in rats exposed to prenatal VPA administration.

Material and Method

Induction of ASD-like model and Imperatorin administration

In order to induce ASD-like model in rats, an intraperitoneal injection of VPA (Darou Pakhsh Co., Tehran, Iran) was done at the dose of 600 mg/kg on 12.5 gestational day (Afshari et al., 2024b, Margedari et al., 2024). Imperatorin (IMP) was prepared in the Institute of Medicinal Plants, ACECR, Karaj, Iran. It was injected at the dose of 20 mg/kg on 34–48 PND (15-day treatment, 15 injections). The administration method was done by gavage (Cao et al., 2014, Zhang et al., 2023). IMP was extracted from the leaves of Ferulago trifida (F. trifida, a species of Apiaceae family). IMP extraction and preparation were done by Saeed Tavakoli. The aerial parts of F. trifida were collected from Qazvin Province, near the village of Ovan in Alamut region, Iran. The plant was first identified by Dr. Valiollah Mozaffarian (Research Institute of Forest and Rangelands, Tehran, Iran) and a voucher specimen was deposited in the herbarium of Faculty of Pharmacy, Tehran University of Medical Sciences, Tehran, Iran (No.THE-6562) (Tavakoli et al., 2017). We evaluated behavioral functions and BDNF expression levels on 49 PND to assess the effect of IMP treatment during juvenile period on behavioral functions and BDNF expression levels in the adulthood. In other words, we evaluated the effect of IMP treatment during juvenile period (the neurodevelopmental state, an important period in the pathophysiology of ASD). The first behavioral test was done 24 h after the last injection.

Animals

In the current study, 16 female and 4 male Wistar rats (8–9 weeks old, weighing 180–200 g) were randomly assigned as dams and sires, respectively. Each male rat was housed with four female rats for breeding purposes. The sires served as controls and were not subjected to any experimental intervention. One male and four females were kept together in the same cage for 15 days to allow mating and confirmation of pregnancy, which was determined by the presence of a vaginal plug. Following confirmation of pregnancy, the male rats were removed from the cages.

Pregnant female rats were then divided into two groups: eight dams received saline injection, while the remaining eight dams were used to establish an ASD-like offspring model by receiving a single intraperitoneal injection of Valproic Acid (VPA; 600 mg/kg) on gestational day 12.5. Saline was administered to the control dams on the same gestational day.

From each dam, 3–4 male and/or female offspring were randomly selected for subsequent behavioral evaluations. A total of 56 offspring rats, including 28 males and 28 females (5 weeks old, weighing 150–160 g), were included in the study and allocated into eight experimental groups (n = 7 per group).

The offspring rats were housed in large Plexiglas cages, with seven animals per cage, under standard laboratory conditions consisting of a 12 h light/12 h dark cycle (lights on at 7:00 a.m.) and a constant ambient temperature of 22 ± 2°C. All animals were bred and maintained in the Cognitive Neuroscience Laboratory, Medicinal Plants Research Center, Institute of Medicinal Plants, ACECR, Karaj, Iran. Food and water were provided ad libitum throughout the study. All experimental procedures were conducted exclusively during the light phase, between 9:00 a.m. and 3:00 p.m.

Our experimental protocol was designed in accord with National Institutes of Health Guide for the Care and Use of Laboratory Animals (National Research Council Committee for the Update of the Guide for the and Use of Laboratory, 2011) and ARRIVE guidelines (Drummond et al., 2010). Notably, one day prior to behavioral assessment, all rats were habituated to the testing environments for 5 min each, including the apparatus used for the open field, grooming, marble burying, and novel object recognition tests, with a 60-minute interval between habituation sessions. Likewise, a 60-minute interval was maintained between consecutive behavioral tests.

Preparing IMP

IMP was extracted from the leaves of Ferulago trifida (F. trifida, a species of Apiaceae family). F. trifida grows in Qazvin Province near the village of Ovan Alamut region. This genus consists of seven species in Iran, most of them are endemic to Iran or have spread in regions of Anatolia, Syria, Lebanon, and Iraq (Tavakoli et al., 2017). The extraction process from the leaves of F. trifida was initiated using petroleum ether, methanol, and chloroformprovided by Pars Chemie Company. 1 H (500 MHz) and 13 C (125 MHz) NMR spectra were analyzed on a 500 MHz Bruker spectrometer (Germany), applying CDCl3 as the solvent and TMS as an internal standard (Merck, Germany). Mass analysis was obtained from an Agilent HP-5973 mass spectrometer (USA). Open-column chromatography (CC), was done using silica gel (70–230 mesh), Sephadex LH-20 (Merck, Germany), and analytical grade solvents (Merck, Germany). The full description of preparing IMP in our institute has been introduced in previous studies (Amini et al., 2022, Tavakoli et al., 2018). Also, a schematic feature shows a brief scheme for the fractionation process from leaves, please see the cited study (Yassaa et al., 2024).

Open field test

Locomotor activity and anxiety-like behavior were assessed using the open field test, a validated method for evaluating these behaviors in rodents. The test arena was a clear plexiglass box measuring 50x50x50 cm, divided into 16 equal squares. Locomotor activity was quantified by counting the number of squares the animal traversed. Anxiety-like behavior was measured by the amount of time the animal spent in the central squares (Houshyar et al., 2024, Kamaei et al., 2023). The total duration of the test was 300 s.

Grooming behavior

The rats were recorded for a duration of 600 s, 60 min after completion of the open field test. The amount of time (in seconds) spent in bilateral grooming was recorded, with alternations in the total duration of the test in comparison to the cited study (Melo-Carrillo and Lopez-Avila, 2013). This test was done 1 h following OFT.

Marble burying test

Obsessive-compulsive disorder (OCD)-like behavior can be assessed in rodents using the marble burying test. In this paradigm, each rat is placed in a Plexiglas cage containing marbles arranged on the surface of the bedding material (typically 10 marbles, arranged in 2 rows of 5). The standard glass marbles used are 15 mm in diameter and weigh 5–6 g. It’s important to note that variations in the number of marbles (ranging from 4 to 25) and placement methods exist across different studies (de Brouwer et al., 2019). Each rat remains in the cage for a 30-minute period without disturbance. The number of marbles buried is quantified as a measure of OCD-like behavior. This test was administered 1 h after the grooming test.

Novel object recognition test (NORT)

The Novel Object Recognition Test (NORT) is a behavioral assay used to evaluate memory function in rodents. The NORT consisted of a habituation session, a familiarization session, and a test session. During the habituation session, rats were introduced to an empty open-field Plexiglas arena (45 cm × 45 cm × 45 cm) for 10 min, with no objects present in the arena. This session was intended solely to allow the animals to become familiar with the testing environment and apparatus and to reduce novelty-related stress before the object-based sessions. A one-hour interval was maintained between the habituation and familiarization sessions. Although longer intervals, including a one-day interval, are commonly used in NORT protocols, we used a one-hour interval in accordance with our previous studies (Gholami-Zanjanbar et al., 2024, Samizadeh et al., 2024). Because the habituation session serves primarily to familiarize the animals with the apparatus, we considered a one-hour interval sufficient for this purpose. In addition, maintaining a shorter interval allowed the behavioral assessment to be conducted closer to the intervention, thereby avoiding an unnecessary prolongation of the time between IMP administration and behavioral testing, during which the effect of the intervention may diminish over time. This approach was therefore used consistently across all experimental groups. Following a one-hour interval, the familiarization session was conducted for 10 min, during which two identical objects were placed in opposite corners of the arena. The test session was conducted 24 h after the familiarization session. During this session, one of the familiar objects was replaced with a novel object, and the rat was allowed to explore the arena freely for 10 min. The familiar and novel objects differed in shape, color, and texture. To eliminate potential confounding olfactory cues, all objects were thoroughly cleaned with a 1% acetic acid solution between trials (Garcia-Partida et al., 2022, Samizadeh et al., 2024). This test was performed 1 h after the marble burying test. The difference in exploration time between the novel (N) and familiar (F) objects was calculated using the discrimination index (DI) as follows: DI = (time N − time F) / (time N + time F) (Romero-Miguel et al., 2023).

Real-time PCR

Tissue collection and preparation

One-hour post-testing, three rats were selected from the seven animals in each experimental group (n = 3 from n = 7 per group) and euthanized using CO₂ gas in a specialized chamber. These three animals were selected for tissue collection and subsequent molecular analysis. These animals were randomly selected from each experimental group and were used as a representative subset for the molecular analysis. The number of animals used for molecular analysis was selected according to previous studies (Abbasi et al., 2024b, Heikal et al., 2026, Payamani et al., 2025). Brains were promptly extracted, and the medial prefrontal cortex (mPFC) was carefully dissected and placed in a liquid nitrogen tank. The entire extraction process was designed to be completed within 2 min to preserve tissue integrity, as per previous methodologies (Talaee et al., 2024a). Thus, while seven animals per group were included in the experimental groups and behavioral assessments, three animals from each group were used for tissue collection and BDNF gene expression analysis (Talaee et al., 2024a).

Sample processing for BDNF analysis

After 24 h, 50 mg of the mPFC samples were suspended in 10 ng of phosphate-buffered saline (PBS) at pH 7.4 and homogenized. The samples were then centrifuged at 2000–3000 rpm for 20 min. The resulting supernatants were carefully collected for the assessment of Brain-Derived Neurotrophic Factor (BDNF) expression levels. All processed samples were stored at −80°C. Detailed protocols for these steps are available in prior studies (Abbasi et al., 2024b, Rahimi-Danesh et al., 2023).

Real-time PCR for BDNF expression

Real-time PCR reactions were conducted using Takara SYBR Premix Ex Taq II (Tli RNaseH Plus) (2X concentration) in a final volume of 20 µl on a StepOnePlus Real-Time PCR System (Applied Biosystems). A volume of 2 µl of synthesized cDNA was used for each reaction. The optimal annealing temperature for the primer pairs was determined to be 64°C. Quantification of the target gene (BDNF) was performed using the standard curve method. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was utilized as the reference gene for data normalization. All samples were run in duplicate, and the average values were used for subsequent analysis. The specificity of the PCR products was confirmed by the presence of a single melting curve peak.

The primers used for real-time PCR are detailed in Table 1. The study focused on assessing BDNF expression in the medial prefrontal cortex due to its established significance in the pathophysiology of autism spectrum disorder (Courchesne and Pierce, 2005, Li et al., 2021, Mediane et al., 2024, Rinaldi et al., 2008, Zhang et al., 2018).

Table 1.

The primers used for real-time PCR.

Gene Name	Primer sequence
r-GAPDH-F	AGGTCGGTGTGAACGGATTTG
r-GAPDH-R	TGTAGACCATGTAGTTGAGGTCA
r-Bdnf-F	GTCCCTTCTACACTTTACCTCT
r-Bdnf-R	TCTTTCACCCTTTCCACTCC

Experimental groups

The present study included eight experimental groups, with seven animals per group (n = 7/group). All seven animals in each group were included in the behavioral assessments. Following completion of the behavioral experiments, three animals from each group were selected for tissue collection and subsequent BDNF gene expression analysis (n = 3/group for molecular analysis). The groups have been summarized below:

Male Saline–Saline: Mother received saline (0.5 mL/rat); male offspring received saline (0.5 mL/rat) during PND 34–48 (15 days).

Female Saline–Saline: Mother received saline (0.5 mL/rat); female offspring received saline (0.5 mL/rat) during PND 34–48 (15 days).

Male ASD-like–Saline: Mother received a single VPA injection (600 mg/kg); male offspring received saline (0.5 mL/rat) during PND 34–48 (15 days).

Female ASD-like–Saline: Mother received a single VPA injection (600 mg/kg); female offspring received saline (0.5 mL/rat) during PND 34–48 (15 days).

Male Saline–IMP (20 mg/kg): Mother received saline (0.5 mL/rat); male offspring received IMP (20 mg/kg) during PND 34–48 (15 days).

Female Saline–IMP (20 mg/kg): Mother received saline (0.5 mL/rat); female offspring received IMP (20 mg/kg) during PND 34–48 (15 days).

Male ASD-like–IMP (20 mg/kg): Mother received a single VPA injection (600 mg/kg); male offspring received IMP (20 mg/kg) during PND 34–48 (15 days).

Female ASD-like–IMP (20 mg/kg): Mother received a single VPA injection (600 mg/kg); female offspring received IMP (20 mg/kg) during PND 34–48 (15 days).

All the experimental timeline and groups, injections, and the tests have been shown in Fig. 1.

Fig. 1.

Schematic timeline of the experimental design. Pregnant rats received a single intraperitoneal injection of valproic acid (VPA; 600 mg/kg) or saline on gestational day 12.5. Male and female offspring were assigned to saline- or VPA-exposed groups and received either saline (0.5 mL/rat) or IMP (20 mg/kg) during postnatal days (PND) 34–48 for 15 consecutive days. Behavioral assessments were conducted following the treatment period, with a 1-h interval between successive behavioral tests. At the end of the behavioral assessments, animals were euthanized and the medial prefrontal cortex (mPFC) was collected for BDNF expression analysis. The experimental groups were: Male Saline–Saline, Female Saline–Saline, Male ASD-like–Saline, Female ASD-like–Saline, Male Saline–IMP, Female Saline–IMP, Male ASD-like–IMP, and Female ASD-like–IMP.

Statistical analyses

SPSS software (V.26) was used to analyze data. All the data were analyzed by Kolmogorov-Smirnov test to evaluate normal distribution and all the results were not significant (data distribution was normal). Three-way ANOVA and post hoc Tukey’s were used to compare groups. Data were expressed as mean ± SD with individual data points shown for each animal and P < 0.05 was considered as the level of statistical significance.

Results

Locomotor activity

Three-way ANOVA results showed that the effects of sex (F_1,48 = 12.57, P < 0.001), ASD (F_1,48 = 38.00, P < 0.001), IMP (F_1,48 = 73.25, P < 0.001), sex*IMP (F_1,48 = 15.58, P < 0.001), ASD*IMP (F_1,48 = 45.44, P < 0.001) were significant, but not the effects of sex*ASD (F_1,48 = 0.24, P > 0.05) and sex*ASD*IMP (F_1,48 = 0.01, P > 0.05). Post hoc Tukey test showed that locomotor activity was increased in ASD rats in both sexes (P < 0.001) with greater effect in females (P < 0.01), while IMP restored hyperlocomotion in males (P < 0.001) and decreased it in females (P < 0.01) (Fig. 2).

Fig. 2.

Locomotor activity (squares passed) in the open field test in all experimental groups. Three-way ANOVA was used to assess the significant/non-significant effects of all variables including ASD, IMP, sex*ASD, sex*IMP, ASD*IMP, and sex*ASD*IMP. Post hoc Tukey test was also used to assess the significant difference between each group. (***P < 0.001 compared to saline-control rats; +++P < 0.001 compared to saline-ASD-like rats; ^^P < 0.01 compared to respective male rats; n = 7). Individual data points are shown for each animal. Partial overlap of some points may occur because of the close proximity of individual values.

Anxiety-like behavior

Three-way ANOVA results showed that the effects of sex (F_1,48 = 8.67, P < 0.05), IMP (F_1,48 = 658.20, P < 0.001), sex*IMP (F_1,48 = 5.16, P < 0.05) were significant, but not the effects of ASD (F_1,48 = 1.20, P > 0.05), sex*ASD (F_1,48 = 0.07, P > 0.05), ASD*IMP (F_1,48 = 1.59, P > 0.05), and sex*ASD*IMP (F_1,48 = 0.35, P > 0.05). Post hoc Tukey test showed that time spent in the middle squares was decreased in ASD rats in both sexes (P < 0.001) with greater effect in females (P < 0.01), while IMP could not affect it (Fig. 3).

Fig. 3.

Time spent (anxiety-like behavior) in the open field test in all experimental groups. Three-way ANOVA was used to assess the significant/non-significant effects of all variables including ASD, IMP, sex*ASD, sex*IMP, ASD*IMP, and sex*ASD*IMP. Post hoc Tukey test was also used to assess the significant difference between each group. (***P < 0.001 compared to saline-control rats; ^^P < 0.01 compared to respective male rats; n = 7). Individual data points are shown for each animal. Partial overlap of some points may occur because of the close proximity of individual values.

Grooming

Three-way ANOVA results showed that the effects of sex (F_1,48 = 47.89, P < 0.001), ASD (F_1,48 = 36.21, P < 0.001), IMP (F_1,48 = 366.95, P < 0.001), sex*ASD (F_1,48 = 5.32, P < 0.05), sex*IMP (F_1,48 = 67.63, P < 0.001), ASD*IMP (F_1,48 = 53.60, P < 0.001) were significant, but not the effect of sex*ASD*IMP (F_1,48 = 0.16, P > 0.05). Post hoc Tukey test showed that grooming was increased in ASD rats in both sexes (P < 0.001) with greater effect in females (P < 0.001), while IMP restored it in males (P < 0.001) and decreased it in females (P < 0.001) (Fig. 4).

Fig. 4.

Grooming behavior (OCD-like behavior) in all experimental groups. Three-way ANOVA was used to assess the significant/non-significant effects of all variables including ASD, IMP, sex*ASD, sex*IMP, ASD*IMP, and sex*ASD*IMP. Post hoc Tukey test was also used to assess the significant difference between each group. (***P < 0.001 compared to saline-control rats; +++P < 0.001 compared to saline-ASD-like; ^^^P < 0.001 compared to respective male rats; n = 7). Individual data points are shown for each animal. Partial overlap of some points may occur because of the close proximity of individual values.

Marbles buried

Three-way ANOVA results showed that the effects of sex (F_1,48 = 20.63, P < 0.001), ASD (F_1,48 = 23.80, P < 0.001), IMP (F_1,48 = 500.63, P < 0.001), sex*ASD (F_1,48 = 14.97, P < 0.001), sex*IMP (F_1,48 = 17.69, P < 0.001), ASD*IMP (F_1,48 = 27.20, P < 0.001), and sex*ASD*IMP (F_1,48 = 0.16, P > 0.05) were significant. Post hoc Tukey test showed that marbles buried was increased in ASD rats in both sexes (P < 0.001), while IMP decreased it only in males (P < 0.001) (Fig. 5).

Fig. 5.

Marbles buried in the marble burying test in all experimental groups. Three-way ANOVA was used to assess the significant/non-significant effects of all variables including ASD, IMP, sex*ASD, sex*IMP, ASD*IMP, and sex*ASD*IMP. Post hoc Tukey test was also used to assess the significant difference between each group. (***P < 0.001 compared to saline-control rats; +++P < 0.001 compared to saline-ASD-like; ^^^P < 0.001 compared to respective male rats; n = 7). Individual data points are shown for each animal. Partial overlap of some points may occur because of the close proximity of individual values.

Novel object recognition memory

Three-way ANOVA results showed that only the effects of ASD (F_1,48 = 222.38, P < 0.001) was significant. However, the other effects were not significant: sex (F_1,48 = 0.09, P > 0.05), IMP (F_1,48 = 0.01, P > 0.05), sex*ASD (F_1,48 = 0.25, P > 0.05), sex*IMP (F_1,48 = 0.39, P > 0.05), ASD*IMP (F_1,48 = 0.39, P > 0.05), and sex*ASD*IMP (F_1,48 = 0.25, P > 0.05) were significant. Post hoc Tukey test showed that novel object recognition memory was impaired in ASD rats in both sexes (P < 0.001), while IMP had no effect on it (Fig. 6).

Fig. 6.

Discrimination index (DI, novel object recognition memory) in all experimental groups. Three-way ANOVA was used to assess the significant/non-significant effects of all variables including ASD, IMP, sex*ASD, sex*IMP, ASD*IMP, and sex*ASD*IMP. Post hoc Tukey test was also used to assess the significant difference between each group. (***P < 0.001 compared to saline-control rats; n = 7). Individual data points are shown for each animal. Partial overlap of some points may occur because of the close proximity of individual values.

BDNF expression level in the medial prefrontal cortex

Three-way ANOVA results showed that only the effects of sex (F_1,16 = 45.00, P < 0.001), ASD (F_1,16 = 105.80, P < 0.001), IMP (F_1,16 = 1008.20, P < 0.001), sex*IMP (F_1,16 = 33.80, P < 0.001), ASD*IMP (F_1,16 = 88.20, P < 0.001) were significant, but not the effects of sex*ASD (F_1,16 = 1.80, P > 0.05), and sex*ASD*IMP (F_1,16 = 0.20, P > 0.05) were significant. Post hoc Tukey test showed that BDNF levels were decreased in ASD rats in both sexes (P < 0.001) with greater effect in females (P < 0.05), while IMP increased it in both sexes (P < 0.001) (Fig. 7).

Fig. 7.

BDNF expression levels in the medial prefrontal cortex in all experimental groups. Three-way ANOVA was used to assess the significant/non-significant effects of all variables including ASD, IMP, sex*ASD, sex*IMP, ASD*IMP, and sex*ASD*IMP. Post hoc Tukey test was also used to assess the significant difference between each group. (***P < 0.001 compared to saline-control rats; +++P < 0.001 compared to saline-ASD-like; ^^^P < 0.001 and ^P < 0.05 compared to respective male rats; n = 3). Individual data points are shown for each animal. Partial overlap of some points may occur because of the close proximity of individual values.

Discussion

In this study, we investigated the potential therapeutic effects of IMP on ASD-like behavioral phenotypes and BDNF changes in VPA-induced ASD-like rat model, with a focus on sex difference. Our results showed hyperlocomotion as a salient phenotype in ASD-like rats, with a greater increase observed in females. Interestingly, IMP restored locomotor activity in males and significantly reduced the hyperlocomotion observed in females. Anxiety-like behavior, measured by the time spent in the center squares, was significantly increased in ASD rats of both sexes, with a greater effect in females, while IMP had no significant impact on this parameter. Similarly, grooming behavior was more prominent in ASD females and IMP was able to ameliorate this behavior, especially in males. Moreover, the number of marbles buried (OCD-like behavior) showed a significant increase in both sexes, whereas the ameliorative effect of IMP was only found in males. Working memory deficit, assessed by novel object recognition memory test, was observed in both male and female ASD rats, and IMP failed to reverse memory impairment. At the molecular level, BDNF levels were significantly decreased in ASD rats, with a greater reduction in females. Interestingly, IMP partially increased it in both sexes. In conclusion, findings showed hyperlocomotion, anxiety-like behavior, OCD-like behavior, working memory deficit, and BDNF downregulation, as major ASD-related features in this model, with evidence of sex-specific patterns and partial therapeutic effects of IMP.

VPA, as neuroprotective and anticonvulsant agent (Xuan et al., 2012), has been shown to be effective in conditions such as bipolar and mood dysregulation spectrum (Romoli et al., 2019), traumatic brain injury (TBI) (Hummel et al., 2024, Wakam et al., 2021), and glioma (Han and Guan, 2021). It has been reported that the therapeutic uses of VPA include epilepsy treatment across different seizure types, bipolar disorder management, and migraine prophylaxis (Rahman et al., 2025). Furthermore, VPA can be effective as an adjuvant therapy in some disorders such as HIV (human immunodeficiency virus), cancer, and neurodegenerative disorders, owing to its histone deacetylase (HDAC) inhibition effect (Lehrman et al., 2005, Routy et al., 2012). The mechanism of action for VPA involves the increase in inhibitory neurotransmissions and modulating voltage-gated ion channels (Rahman et al., 2025). Inhibition of voltage-gated sodium channels, gamma-aminobutyric acid (GABA) transaminase, enhancement of GABA synthesis, and modulation of calcium channels are among the most important mechanisms of action for VPA (Rahman et al., 2025). However, prenatal VPA administration results in significant developmental abnormalities in offspring, including intellectual disabilities (Daugaard et al., 2020), impaired early cortical development, impaired neuronal differentiation and firing, reduced dendritic complexity (Williams, O.O. et al., 2025; Zang et al., 2022) and ASD-like symptoms (Mabunga et al., 2015). The induction of autistic-like behaviors following VPA administration has been shown to be linked with several underlying neurobiological mechanisms involving epigenetic dysregulation (Liu, H. et al., 2021), disrupting autophagy and synaptic homeostasis (Qin et al., 2016), and pathological alterations in the ultrastructure of microglia and synapses (Gassowska-Dobrowolska et al., 2020). In addition, evidence has shown that VPA results in behavioral and inflammatory changes, changes in glial cell density, abnormal development of the cortical plate, neural tube defects in the cranial region, and increases Purkinje cell death, reactive astrocytes, and microglial activation (Kazlauskas et al., 2016, Kuwagata et al., 2009, Turner et al., 1990).

Hyperlocomotion and anxiety-like behaviors are among the most commonly reported behavioral abnormalities in animal models of ASD (Kim et al., 2014, Pascual et al., 2023). In accordance with previous study using prenatal VPA-induced male ASD rodent model, our results showed increased locomotion and anxiety-like behaviors (Hosseini et al., 2025), reflecting alterations in neural circuits involved in motor control. For instance, it has been shown that prenatal VPA exposure leads to enhanced motor activity and stereotypic behaviors in mice, accompanied by reduced expression of synaptic proteins (Nlgn1 and PSD-95) in the striatum, indicating impaired excitatory signaling in motor pathways (Hernandez et al., 2023). Moreover, another study has demonstrated motor hyperactivity and increased anxiety-like behaviors along with decreased dopamine synthesis and activity in the striatum and prefrontal cortex of male rats, suggesting dopaminergic hypoactivity in motor and limbic circuits (Cezar et al., 2025). Altered BDNF signaling has also been noted as a downstream effect of prenatal VPA exposure in a previous review study, suggesting its role in the induction of ASD-like phenotypes. BDNF, the pivotal neurotrophin in regulating neuronal development and synaptic plasticity (Park and Poo, 2013), has been reported to be related to the induction of ASD-like behavioral features in prenatal VPA exposure models. A previous meta-analysis study has reported that BDNF plays a main role in autism initiation/propagation, considered it as a critical biomarker of ASD (Saghazadeh and Rezaei, 2017). As previous findings revealed, reduced BDNF levels are associated with disrupted glycogen synthase kinase-3 beta (GSK-3β) and Akt signaling pathways, critical for neurodevelopment (Hosseini et al., 2025). Similarly, a study by Bath et al. (2017) has found that VPA model results in significant motor deficits in both sexes along with a substantial decrease in BDNF protein level in the cerebellum, indicating that reduced BDNF expression in this region may underlie locomotor abnormalities (Bath and Pimentel, 2017). Our recent study has shown hyperlocomotion and increased anxiety-like behavior, grooming, and sniffing in a male rat model of ASD induced by prenatal VPA administration along with decreased BDNF and increased GSK-3beta levels in the medial prefrontal cortex (Hosseini et al., 2025). Furthermore, it has been shown that prenatal VPA administration leads to autistic-like behavioral features and BDNF downregulation in the prefrontal cortex of male rats (Kumar et al., 2022). Also, BDNF levels are significantly decreased in the hippocampus of autistic male rats induced by VPA (Afshari et al., 2024b). Our findings, consistent with previous reports, showed an increase in locomotion and anxiety-like behavior along with the downregulation of BDNF, suggesting the critical role of BDNF underlying these ASD-like behavioral phenotypes. As we know, BDNF and its tyrosine kinase receptor, TrkB, are expressed in hypothalamic nuclei associated with satiety and locomotor activity (Kernie et al., 2000). Also, decreased serum and plasma BDNF levels have been shown in individuals with anxiety disorders (Suliman et al., 2013). Hippocampal BDNF has been shown to be involved in anxiety-like behaviors in animal models, and numerous types of stressors lead to decreased expression of BDNF (Duman, 2002, Rasmusson et al., 2002). Other studies have also shown the relationship between anxiety-like behaviors and BDNF downregulation in rodents (Hosseini et al., 2025, Rio-Alamos et al., 2025). Although some studies have shown inconsistent results (Abbasi et al., 2024b, Kamaei et al., 2023), suggesting that the model of stress or intervention may affect the relationship between BDNF changes and locomotor activity and anxiety-like behavior. A previous systematic review and meta-regression analysis has also reported that although BDNF levels appear to be reduced in individuals with anxiety disorders, this is not consistent across the various anxiety disorders and may largely be explained by the significantly decreased BDNF level found in OCD, suggesting that results further appear to be mediated by differences in sampling methods (Suliman et al., 2013). Our findings, consistent with previous reports, showed an increase in locomotion and anxiety-like behavior along with the downregulation of BDNF in VPA-exposed offspring, suggesting a potential role of BDNF in these ASD-like behavioral phenotypes.

OCD-like behaviors such as marble burying and excessive grooming have been broadly examined in rodents exposed to prenatal VPA. Several studies have reported increased marble burying following prenatal VPA exposure, indicating heightened OCD-like behavior. For example, a previous study has shown increased self-grooming and buried marbles induced by prenatal VTA in male rats, while a combination of low-frequency repeated transcranial magnetic stimulation and superparamagnetic iron oxide nanoparticles reverse these behaviors (Afshari et al., 2024b). Other study has reported that prenatal administration of VPA in male rats leads to significant social deficits and increased marbles buried (Farbin et al., 2025). Furthermore, it has been demonstrated that prenatal administration of VPA in male mice results in ASD-like behaviors, including social deficits, increased marbles buried, and impaired spatial memory (Liu et al., 2025). BDNF alterations are involved in the induction of OCD-like behaviors in ASD-like rodent models. For example, it has been shown that prenatal administration of VPA in male mice leads to increased repetitive behaviors and marbles buried via attenuating AMPK/CREB/BDNF pathway, while arecoline potentially shows therapeutic role through upregulating AMPK/CREB/BDNF pathway and increasing oligodendrocyte maturation and remyelination (Wen et al., 2025). It has also been shown that the prenatal administration of agmatine can prevent social behavior deficits and behavioral alterations in the marble burying test in VPA-exposed male mice via activating the ERK/CREB/BDNF signaling pathway (Chen et al., 2024). The present research has also shown increased OCD-like behavior in the marble burying test along with decreased BDNF levels, while IMP showed a sex-dependent beneficial effect on OCD-like behavior (only in males).

Memory impairments are among the most reported cognitive deficits in rodent models of ASD. A previous study has found that VPA reduces neurogenesis in the hippocampus of male rats, contributing to impairments in memory function (Aranarochana et al., 2021). VPA-exposed male rats exhibit significant deficits in spatial working memory as well as decreased hippocampal cell proliferation and BDNF expression levels (Umka et al., 2010). Additionally, prenatal VPA administration results in deficits in learning and memory processes potentially via modulating BDNF-Akt signaling pathway in male rats, suggesting the involvement of BDNF pathway in memory impairments induced by VPA (Gao et al., 2016). Other study has suggested BDNF-TrkB signaling stimulation as an underlying mechanism of music therapy which enhances hippocampal cell proliferation and mitigates short-term memory deficits in VPA-induced autistic male rat pups (Lee et al., 2016). AMPK/CREB/BDNF downregulation in the prefrontal cortex is involved in memory deficits in male mice exposed to prenatal VPA administration (Wen et al., 2025). Of note, decreased dendritic spine density in pyramidal neurons of the CA1 hippocampal region is involved in the induction of spatial memory impairment following administration of VPA in male rats (Afshari et al., 2024a). Furthermore, a related study in male rats exposed to prenatal VPA administration has shown that hippocampal BDNF downregulation is involved in the induction of spatial memory deficit (Wu et al., 2017). As mentioned, BDNF is a critical neurotrophic factor involved in the modulation of neuroplasticity, neurogenesis, mood regulation, and learning and memory processes (Ghaffarzadegan et al., 2024a, Mohamadian et al., 2023, Talaee et al., 2024b). Many studies have shown the important role of BDNF downregulation in autistic-related cognitive and behavioral deficits (Hosseini et al., 2025, Ilchibaeva et al., 2023, Zohny et al., 2023). Although evidence has also reported increased BDNF serum levels in individual with ASD, showing the wide discrepancies in the evidence (Saghazadeh and Rezaei, 2017). It has also been shown that BDNF serum levels in individuals with ASD may depend on the severity of ASD, showing higher BDNF levels in atypical autistic subjects (clinically milder phenotype) as compared to controls, but not in typical ASD cases (clinically severe phenotype) (Kasarpalkar et al., 2014). The present research (as shown in most preclinical ASD studies) showed novel object recognition memory impairment along with reduced BDNF levels.

Notably, our findings showed sex differences in response to VPA administration, as well as in the ameliorating effect of IMP on ASD-like symptoms. Notably, several behavioral and molecular outcomes exhibited sex-dependent differences in magnitude and responsiveness to IMP treatment. These findings are consistent with previous reports suggesting sex-specific vulnerability to prenatal VPA exposure. It has been shown that female VPA-exposed rats only exhibit increased repetitive/stereotypic-like activity, while male VPA-exposed rats exhibit increased repetitive/stereotypic-like activity along with lower sensitivity to pain, higher anxiety, reduced social interaction, and elevated basal levels of corticosterone (Schneider et al., 2008). More hyperlocomotion in female VPA-exposed rats and higher anxiety-like behavior in male VPA-exposed rats have been shown, probably due to sex-related physiological differences (Mamali et al., 2025). A previous study has shown increased anxiety-like behaviors in adolescent VPA-exposed females but not males, while there are no significant sex differences in changes in locomotion, grooming, and recognition memory (Thornton et al., 2021). Male and female rats show different autistic-like traits, with both showing hyperactivity, while only males show reduced sociability and increased anxiety (Mamali et al., 2025). Vitamin D supplementation alleviates ASD-like symptoms in rats, improves gastrointestinal function, and increases serotonin levels in male but not female rats (Wang et al., 2025). Continuous training exercise increases long-term potentiation (LTP) in VPA-exposed male offspring, while has no effect on LTP in VPA-exposed female offspring (Mohammadkhani et al., 2024). Also, VPA exposure leads to elevated repetitive behavior (self-grooming) in early adolescence and impairments in social behavior in adulthood in both sexes of rats, while recognition memory deficit is only observed in males (Souza et al., 2024). Female ASD-like rats show better performance in the discrimination tasks and acoustic startle reflex than males, while male ASD-like rats have been reported to be better than females in the three-chamber social interaction test (Sabzalizadeh et al., 2022). It has been declared that females show heightened sensitivity to VPA-induced stress (Sheibani Tezerji et al., 2025). The potential underlying mechanisms of sex differences in rodent models of ASD have not well been investigated. Exploring sex differences in different brain functions is necessary because nervous system in males and females responds differently to abnormal physiological conditions (Sickmann et al., 2014). Sex-specific structural and functional changes in the hippocampus and the cortex following VPA administration may also be involved in sex differences (Williams, O.O.F. et al., 2025). Also, behavioral differences and other sex differences in ASD-like models may be associated with the function of genes that are differentially affected by VPA in males and females (Weinstein-Fudim et al., 2019). Sex differences in gene expression involved in glutamatergic pathways (such as Unc13a and Cacna1a) have been shown to be important in sex differences in ASD pathology and learning and memory deficits (Weinstein-Fudim et al., 2019), while the present study did not evaluate it.

The effectiveness of IMP, an organic ingredient extracted from several medicinal plants such as Angelica dahurica and Cnidium monnieri, has been largely explored. IMP has neuroprotective and anti-oxidant properties, alleviating cognitive impairments and oxidative stress in animal models (Chowdhury et al., 2018, Huang et al., 2021). Some earlier results have reported anti-anxiety and memory improvement effects in mice exposed to nicotine (Budzynska et al., 2013) and ameliorating effects of IMP on memory in mice exposed to scopolamine (Budzynska et al., 2015). IMP can ameliorate learning and memory impairments through BDNF/TrkB signaling in prenatally-stressed female offspring rats (Zheng et al., 2020). IMP can significantly improve spatial learning and memory deficits in 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP)-exposed mice, as a Parkinson’s model (Liu et al., 2024). Also, IMP improves the impairment of memory acquisition and consolidation induced by scopolamine in mice (Budzynska et al., 2015). IMP can prevent nicotine-induced anxiety- and memory-related responses in mice (Budzynska et al., 2013). Another study has shown that IMP dose-dependently relieves anxiety-like behaviors in the elevated plus maze in mice (Budzynska et al., 2012). However, the present research showed that IMP could not restore memory deficit and anxiety-like behavior induced by prenatal VPA administration. The mechanisms underlying therapeutic effects of IMP have not been properly understood. It has been found that IMP can significantly decrease oxidative stress and inflammation, and induce antioxidant properties (Liao et al., 2023, Yang et al., 2024), leading to cognitive improvements. IMP highly suppresses neuroinflammation and neural oxidative stress in the midbrains of rats exposed to MPTP (Parkinson’s disease model), leading to behavioral improvements (Liu et al., 2024). Other study has shown the beneficial effects of IMP on LPS-induced cognitive deficits in mice via decreasing pro-inflammatory factors (Chowdhury et al., 2018). Furthermore, anti-oxidant properties of IMP lead to cognitive improvements in mice exposed to scopolamine (Budzynska et al., 2015). Therefore, reduced neuroinflammation and oxidative stress, and induction of anti-oxidant effects may be involved in the therapeutic effects of IMP. However, the present research did not assess inflammatory and oxidative stress factors, but evaluated BDNF levels, as the most important neurotrophic factor. Our findings showed that IMP partially increased BDNF levels in both sexes. As we know, reduced BDNF levels in the hippocampus and the prefrontal cortex often leads to disrupted cognitive and behavioral functions (Abbasi et al., 2024b, Manooki et al., 2025). It has been shown that hippocampal BDNF downregulation is related to memory deficits (Ghaffarzadegan et al., 2024b, Tajabadi Farahani et al., 2024), behavioral dysfunctions such as hyperactivity, compulsive-like behaviors, and anxiety-like behaviors (Abbasi et al., 2024b, Chen et al., 2025), and also, neuropsychiatric disorders such as depression, OCD, and schizophrenia (Autry and Monteggia, 2012, Motamedi-Manesh et al., 2025). Furthermore, decreased BDNF levels in the hippocampus, serum, and prefrontal cortex are related to cognitive and behavioral deficits in the offspring exposed to prenatal stress or other stressors, while BDNF upregulation underlies the improvement of cognitive and behavioral deficits (Abkhezr et al., 2024, Vera-Juarez et al., 2025, Zhang et al., 2024). IMP increases BDNF levels in hippocampus and cortex and improves cognitive deficits under inflammatory conditions, such as LPS-induced neuroinflammation (Chowdhury et al., 2018). IMP ameliorates learning and memory impairments induced by prenatal stress via modulating the BDNF/TrkB signaling pathway, as well as regulating the activity of the hypothalamic–pituitary-adrenal (HPA) axis (Wang et al., 2013). IMP can markedly increase BDNF, TrkB, and CREB expression levels in the hippocampus of female offspring rats exposed to prenatal stress, leading to the improvement of cognitive and behavioral deficits (Zheng et al., 2020). Therefore, increased BDNF levels may be one of the underlying mechanisms of IMP involved in its’ beneficial effects on ASD-like behavioral features in the rats’ offspring exposed to prenatal VPA. However, it should be noted that evidence regarding the underlying mechanisms of IMP is very limited and more detailed studies are needed.

It is noteworthy that IMP showed differential efficacy across the tested behavioral paradigms, demonstrating significant improvements in some areas while remaining ineffective in others, such as anxiety-like behaviors and certain memory tasks. This variability likely stems from the complex, multifaceted nature of VPA-induced neurodevelopmental disruption, which may not be fully compensated by the current treatment regimen. Alternatively, the temporal window of administration or the dosage of IMP may have been insufficient to modulate the specific neural circuits underlying these cognitive and emotional deficits. Future research should investigate a broader dose-range and investigate the potential synergy of IMP with other pharmacological agents to address these refractory symptoms. Furthermore, longitudinal assessments are essential to determine whether the observed behavioral benefits are sustained throughout development or are merely transient effects.

An important finding of the present study is that the behavioral effects of IMP were not uniform across all evaluated domains. While IMP significantly improved hyperlocomotion and repetitive/compulsive-like behaviors, it failed to ameliorate anxiety-like behavior and recognition memory deficits in VPA-exposed rats. Interestingly, these behavioral outcomes were accompanied by a significant restoration of BDNF expression in the mPFC, suggesting that the beneficial effects of IMP may be partially mediated through the enhancement of neuroplasticity-related pathways in this brain region. Given the critical role of the mPFC in executive function, behavioral regulation, social cognition, and the modulation of repetitive behaviors (Ko, 2017), the observed increase in BDNF expression may contribute to the normalization of specific ASD-like behavioral abnormalities induced by prenatal VPA exposure. However, the lack of improvement in all behavioral domains despite increased mPFC BDNF expression highlights the complexity of the neurobiological mechanisms underlying ASD-like phenotypes. This finding suggests that restoration of BDNF signaling within a single brain region may not be sufficient to reverse all behavioral impairments associated with prenatal VPA exposure. For example, memory deficits are strongly associated with hippocampal-dependent synaptic plasticity and neurogenesis (Bekinschtein et al., 2014, Vaseghi et al., 2021), and several studies have reported reduced hippocampal BDNF expression in VPA-exposed animals (Afshari et al., 2024b, Hosseini et al., 2025). Therefore, the absence of memory improvement following IMP treatment may indicate that molecular alterations in memory-related neural circuits persist despite the restoration of mPFC BDNF levels. Similarly, anxiety-like behavior is regulated by a distributed network involving the prefrontal cortex, hippocampus, amygdala, and their reciprocal connections (Kenwood et al., 2022). Consequently, normalization of BDNF expression in the mPFC alone may be insufficient to fully correct anxiety-related behavioral disturbances. Another noteworthy observation is that IMP appeared to exert greater effects on behaviors associated with behavioral control and repetitive responding than on cognitive performance. This dissociation may indicate that distinct neural substrates underlie repetitive behaviors, anxiety-like behavior, and recognition memory deficits in the VPA model. It is therefore possible that IMP preferentially influences neural pathways involved in mPFC-dependent behavioral regulation, while having a more limited impact on hippocampal- or amygdala-dependent functions. Such region-specific effects could explain why improvements in locomotor activity and repetitive behaviors occurred in parallel with increased mPFC BDNF expression, whereas anxiety-like behavior and memory deficits remained largely unaffected.

Limitations of the study

Several limitations should be considered when interpreting our findings. First, BDNF expression was assessed in a randomly selected subset of animals from each experimental group (n = 3/group). Although this approach is commonly used in exploratory molecular studies, the limited sample size may restrict statistical power and the generalizability of the molecular findings. Second, we utilized a single dose of Imperatorin, which may not capture the full dose-response relationship or the optimal therapeutic window for treating ASD-like behaviors. Third, the study focused on specific behavioral domains and molecular markers; therefore, it does not account for the broader neuro-circuitry changes or other neurotransmitter systems involved in the pathogenesis of VPA-induced ASD. Fourth, BDNF expression was assessed only in the medial prefrontal cortex. Although the mPFC is strongly implicated in ASD pathophysiology, the hippocampus is also critically involved in learning, memory, and neuroplasticity. Therefore, evaluating hippocampal BDNF expression could have provided additional mechanistic insight into the observed memory deficits and the lack of IMP efficacy in the novel object recognition task. Finally, as an animal study, these findings require translational validation in human clinical trials to establish therapeutic efficacy and safety.

Conclusion

In conclusion, the present study is the first to examine the effects of IMP on ASD-like behaviors induced by prenatal VPA administration. Our results demonstrate that IMP exerts sex-dependent (and also task-dependent) effects. Specifically, IMP significantly reduced excessive grooming in both sexes and marble burying behavior only in males. In contrast, IMP showed no significant effect on anxiety-like behavior or memory performance in both sexes, suggesting that its influence may be both sex- and task-dependent. Notably, IMP increased BDNF levels in both sexes, with a more pronounced effect in males, suggesting a potential mechanism underlying the beneficial effects of IMP. We suggest that further research be conducted to determine whether IMP could serve as an effective treatment for ASD-like symptoms, with particular attention to sex- and task-dependent effects and the biological pathways involved.

CRediT authorship contribution statement

Marjan Mohamadian: Writing – original draft, Methodology, Formal analysis. Fatemeh Barzi: Investigation, Data curation. Marzieh Jalalian-Javadpour: Validation, Methodology, Data curation. Hamed Fallah: Investigation, Data curation. Mahdi Khaledian: Writing – review & editing, Validation. Saeed Tavakoli: Resources. Salar Vaseghi: Writing – review & editing, Supervision, Project administration, Methodology.

Ethics declaration

This study was conducted in accordance with the ARRIVE (Animal Research: Reporting of In Vivo Experiments) guidelines. Ethics committee approval was not required. Cognitive Neuroscience Lab of the Institute of Medicinal Plants has an approved certificate for the conduction of animal studies.

Ethical Approval

The Institutional Animal Care and Use Committee (IACUC) ethics approval for conducting animal studies in the Cognitive Neuroscience Lab, Medicinal Plants Research Center, Institute of Medicinal Plants, ACECR, Karaj, Iran, is under the letter with the code of IMP (S/1403/44319).

Consent to participate

N/A

Consent to publish

N/A

Clinical trial number

Not applicable

Authors’ contribution

F.B., H.F., and M.J. conducted the experiments. M.J. and M.M. writing and analyzing data, drafting. M.M. and M.K. participated in revisions, designing, and validations. S.T. drug preparation. S.V. designed the study, supervised the research process, and revised the manuscript. All the authors have approved the final version.

Funding

There is no providing financial support to this project.

Declaration of Competing Interest

None.

Data availability

Data will be made available upon a reasonable request.