Several factors such as gene mutations, gene variations and adverse environmental events concur to the pathogenesis of Autism Spectrum Disorder (ASD). Among the environmental factors involved in the pathogenesis of ASD, prenatal exposure to the antiepileptic drug valproic acid (VPA) has been repeatedly associated with increased risk of neurodevelopmental delay and autistic symptoms in the offspring (1, 2). When given during gestation, depending on the time window of exposure and the administered dose, VPA is able to cause congenital malformations (3, 4) and to induce core autistic symptoms in the offspring, such as impaired communication, reduced sociability and stereotyped behaviors (5, 6). Based on this clinical evidence, prenatal exposure to VPA in rodents has been validated as a preclinical model of ASD mimicking one of the environmental factors involved in the pathogenesis of the disease (7–9). Indeed, rodents exposed to VPA recapitulate core and comorbid features of ASD and are characterized by neurochemical alterations resembling those found in patients with autism, thus confirming the good face and construct validity of this preclinical model (7, 9–11).

Although the mechanisms by which VPA interferes with development causing autistic-like features in the exposed offspring need to be still elucidated, an imbalance between GABAergic and glutamatergic neurotransmission has been reported, which is also implicated in ASD etiology (12, 13). Indeed, defects in establishing and maintaining the balance between excitatory and inhibitory neurotransmission has been considered crucial in the pathogenesis of ASD (14). In particular, this imbalance might be responsible for disrupted neural connectivity and ASD-like behaviors such as impaired social interaction (15–19). Accordingly, male rats prenatally exposed to VPA show an excitatory/inhibitory imbalance, similar to ASD patients (20, 21). Moreover, the GABA excitatory-inhibitory shift that normally occurs during delivery is abolished in VPA-exposed animals, while restoring this shift rescues their aberrant behavioral phenotype (22).

Increased oxidative stress has been identified as a possible mechanism through which environmental factors exert their deleterious effects on the developing brain (23), which may be further exacerbated by their interaction with genetically susceptible alleles, thus increasing the risk of ASD. Not surprisingly, oxidative stress plays a role in the embriotoxicity induced by VPA (24, 25).

N-acetylcysteine (NAC) is a synthetic derivative of the endogenous amino acid L-cysteine that is able to modulate glutamatergic neurotransmission and the excitatory/inhibitory ratio (26, 27). Furthermore, the anti-oxidant properties of NAC are relevant to the pathogenesis of several neuropsychiatric disorders (28–33). Indeed, NAC is an essential substrate for the enzyme γ-glutamyl-cisteine synthetase (GCL) in the synthesis of glutathione (GSH), the main antioxidant in tissues (34). GSH redox imbalance plays an essential role in the onset and progression of several neurodegenerative and neuropsychiatric diseases and it may be used as a biomarker for their diagnosis (35). GSH deficit and GSH redox imbalance have been reported in individuals with autism, supporting a protective role of the GSH system in ASD development (36–41).

The antioxidant properties of NAC and its ability to modulate glutamatergic neurotransmitters have prompted studies focused on the potential clinical application of NAC in ASD (33). However, the studies conducted to date have provided controversial results, with some studies reporting beneficial effects of NAC administration in children with ASD (42), and other studies reporting minimal or no effects (41). In this context, basic research in validated preclinical models of ASD is essential to guide larger clinical trials in order to determine whether core symptoms of ASD may respond to treatment with NAC and to clarify the mechanisms involved in such effects.

Here, we explored the impact of repeated administration of NAC on core autistic-like features (such as social impairments and stereotyped behaviors) in the VPA rat model of ASD. Furthermore, since GSH redox imbalance seems to be specifically observed in some brain regions of autistic patients (37, 43), we measured GSH content in prefrontal cortex (PFC), dorsal striatum (DS), nucleus accumbens (NAc), amygdala (AMY), cerebellum (CER) and hippocampus (HIPP) of VPA-exposed animals, in order to understand if brain region-specific GSH variations may contribute to VPA-induced autistic-like traits. Last, the mRNA expression of genes encoding scaffolding (PSD-95), transcription regulation proteins (MeCP2, CaMKII), the subunits of NMDA (NR1, NR2a, and NR2b) and AMPA (GLUR1 and GLUR2) receptors, as well as the metabotropic glutamate receptors mGLUR1 and mGLUR5, that are all involved in the regulation of the post-synaptic terminal activity, was evaluated in the same brain regions.

Although ASD is among the most severe chronic childhood disorders, it lacks effective treatments, and its complex etiology is still a matter of investigation. Multiple factors are thought to be involved in the pathogenesis of the disease (i.e., genetic, environmental factors and their interaction), thus explaining the heterogeneity of symptoms displayed by autistic patients.

In recent years, several studies from different research groups reported reduced GSH levels in the brain and blood of ASD children (36, 38, 53). This depletion of GSH may be critically involved in the pathogenesis of ASD through different mechanisms. First, it may lead to increased oxidative stress toxicity. Second, it may induce excitotoxicity through over-activation of excitatory glutamatergic neurotransmission (54). Third, as GSH is an epigenetic modulator, its deficiency may negatively affect brain development in the presence of triggering environmental factors, thus contributing to ASD. One of the environmental factors known to be involved in the pathogenesis of ASD is maternal exposure to VPA. There is indeed clinical evidence that maternal use of VPA during pregnancy can induce a wide range of abnormalities in the exposed children, ranging from structural malformations to more subtle autistic-like behaviors (1, 5, 8, 55, 56). Based on the robust clinical evidence, prenatal exposure to VPA in rodents has been validated as an environmentally-triggered preclinical model of ASD (8, 9, 57).

Since NAC has antioxidant properties, it is a precursor of glutathione and it modulates glutamatergic neurotransmission, we explored the effects of repeated NAC administration on core autistic-like features, brain GSH metabolism and mRNA expression of genes that regulate synaptic activity in the VPA rat model of ASD.

We found that repeated administration of NAC reverted the social deficit displayed by VPA-exposed rats in the three-chamber test. Since the last NAC administration was given 2.5 h prior the behavioral experiments, we cannot exclude that the animals were still under acute NAC influence when performing the task. Yet, repeated administration of NAC has previously found to be effective: when NAC was administered for 10 days at the same dose used in the present study, it reverted the deficit in social interaction displayed by male VPA-exposed rats (27). However, while NAC was found to ameliorate the repetitive behaviors performed by VPA-exposed rats in the open field test (47), we found that NAC did not counteract the repetitive/stereotyped behaviors displayed by VPA-exposed rats in the hole board test. Thus, the positive effect of NAC on stereotypic behaviors in VPA-exposed animals may be dependent on the behavioral task and the experimental protocol used. To support the preclinical findings, a recent meta-analysis of randomized clinical trials showed that NAC supplementation for 8–12 weeks significantly improved irritability, hyperactivity and social awareness in ASD patients (58), thus supporting the idea that certain behavioral traits observed in autism may benefit of NAC treatment.

The brain is particularly susceptible to oxidative stress, due to high oxygen consumption (~20% of oxygen), low endogenous antioxidant defense, large quantity of iron and copper, and abundance of polyunsaturated fatty acids in neuronal membranes (59).

GSH plays a key role in the maintenance of the brain redox equilibrium and imbalances in the GSH redox system are an important factor in the pathophysiology of ASD (36, 38, 53, 60).

Some studies have previously reported significantly reduced GSH levels in post mortem cerebellum and temporal cortex of children with ASD, along with lower activity of the enzyme responsible for the GSH synthesis, γ-glutamyl cysteine ligase (GCL) (35, 37, 40, 60). In light of this, we measured brain GSH levels in VPA-exposed rats, also exploring the potential region-specificity of these changes. Our findings show that, compared to SAL-exposed rats, GSH concentrations are significantly decreased in the prefrontal cortex, nucleus accumbens, amygdala, cerebellum and hippocampus of VPA-exposed animals. These data support the GSH depletion found in post mortem cerebellum of children with ASD but, at the same time, they highlight a more generalized GSH deficiency in several brain regions, with the only exception for the dorsal striatum, where GSH levels appear increased.

Two principal factors can be responsible for the reduction of GSH levels found in VPA-exposed animals: a decreased synthesis, due to a direct or indirect inhibition of the enzyme GCL, and/or an increase of the GSH consumption by reactions catalyzed by the GSH-dependent antioxidant enzymes (GPx, GR, GST). Exposure to toxic metals, for instance, can block GSH synthesis and rise thiol excretion, thus exacerbating redox dysfunctions (61, 62). Also the gut microbiota, which is dysregulated in ASD, has been found to modulate glutathione metabolism (63), potentially contributing to predispose children to the gastrointestinal dysfunction often occurring in the autistic phenotype (38, 64–66).

In addition, and besides its well-known role as antioxidant, GSH can affect “redox-independent” processes, including the accumulation of extracellular glutamate, thus triggering excitotoxicity and ultimately resulting in neuronal dysfunction and death (36, 67, 68).

GSH metabolism is strongly inter-connected with glutamatergic neurotransmission: the exposure to glutamate results in a dose-dependent decrease of GSH in astrocytes (69) but, at the same time, it represents an essential precursor of the GSH synthesis (70). Furthermore, GSH elicits a modulatory function on glutamate receptors (such as N-methyl-D-aspartate receptors) and regulates some important transcription factors implicated in the immune response pathway (NF-kB) and in neuro-inflammation. Therefore, the GSH homeostasis is crucial for neuronal health and neurodevelopment contributing to the balanced interaction between glutamate and GABA, both essential for synaptic maturation, refinement of neuronal circuitry and regulation of cognition, emotion and behavior (54, 71–73).

Moving from previous studies showing that VPA-induced behavioral deficits in rodents can be rescued by NAC treatment (27, 47), through the GSH-mediated modulation of synaptic activity, here we analyzed the effect of NAC on GSH brain levels in VPA-exposed rats. Unexpectedly, total and free GSH levels did not show significant variations after NAC supplementation, except for two brain regions, the nucleus accumbens and hippocampus, where total GSH (an index of GSH synthesis) was consistently increased. The reason of this differential, area-specific effectiveness of NAC on GSH levels is not still understood, but it has to be considered that NAC is involved not only in GSH production, but in multiple processes underlying the pathophysiology of several neuropsychiatric disorders (74). Indeed, beside its essential role as reducing agent (by its thiol group) and as co-substrate for the GSH synthesis, NAC has also anti-glutamatergic properties, and NAC-treated mice showed reduced amounts of glutamate and excitatory currents in vivo (74–77). Thus, the ability of NAC to rescue the social deficits displayed by VPA-exposed rat, rather than derive from a direct increase in GSH levels, can be the result of its central role in glutamate signaling, oxidative homeostasis and interaction with inflammatory mediators.

An imbalance of the excitatory synaptic signaling is widely reported in ASD (78, 79), especially regarding the excess of glutamatergic signaling (80) connected to impairments in the synaptic proteins expression at the synaptic terminals (particularly in males) (16, 20, 21, 81). For this reason, we evaluated the mRNA expression of genes regulating the neuronal post-synaptic terminal activity in VPA-exposed animals and SAL-exposed controls. Our data confirm that prenatal VPA exposure induces a general region-specific impairment in the expression of genes related with synaptic regulation, and that NAC treatment is able to partially rescue these defects to normal levels or, at least, to reduce their impact. On this basis, it is possible to hypothesize a role for NAC in re-establishing the physiological glutamatergic signaling, which appears altered in ASD. Taken together our results corroborate previous findings (27, 47) showing that NAC administration improves VPA-induced behavioral deficits in rodents, such as anxiety, social impairments and stereotypic behavior, modulating synaptic activity through its antioxidant effects and its ability to modulate glutamatergic neurotransmission.

Overall, this study highlights a general brain depletion of GSH in the VPA rat model of ASD and, interestingly, suggests a differential response of different brain areas to the antioxidant treatment with NAC. However, what still needs to be determined is if the GSH decrease represents a primary defect in ASD or if the ASD pathological assessment leads to a GSH redox imbalance. Nevertheless, it is certain that the GSH deficiency results in a decreased ability to excrete toxic molecules (metals, xenobiotics) that, instead, an antioxidant supplementation may help to counteract.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The animal study was reviewed and approved by Italian Ministry of Health Autorizzazione n. 70/2021-PR.

SS performed, analyzed, and contributed to the design of the behavioral experiments. SP and PL performed, analyzed, and designed the biochemical experiments. EC and JD'A contributed to the behavioral and the biochemical experiments, respectively. SS, PL, and SP wrote the manuscript. VT and FP supervised the project, designed the experiments, and wrote, revised, and edited the manuscript. All authors contributed to the article and approved the submitted version.

VT was supported by Jerome Lejeune Foundation Research grant #1674 (VT), by PRIN 2017 grant (2017SXEXT5) by MIUR (VT), and by Excellence Departments, MIUR-Italy, Grant/Award Numbers: ARTICOLO 1, COMMI 314-337 LEGGE 232/2016, ARTICOLO 1.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.