VGLUT1 and PSD‐95 Expression Remains Stable in the Prefrontal and Cerebellar Cortices of the VPA Autism Rat Model
Megan Reveley, Gavin Ray Owen, Nancy Tumba, Oladiran Ibukunolu Olateju, Busisiwe Constance Maseko
TL;DR
This study found no changes in VGLUT1 and PSD-95 protein levels in rat models of autism, suggesting these proteins may not be involved in the condition's development in these brain regions.
Contribution
The study provides new evidence that VGLUT1 and PSD-95 expression remains stable in a VPA-induced autism rat model in specific brain regions.
Findings
VGLUT1 and PSD-95 protein expression levels were unchanged in the prefrontal cortex of VPA-induced rats.
No significant differences in VGLUT1 and PSD-95 expression were found in the cerebellar hemisphere of VPA-induced rats.
Abstract
Autism spectrum disorder (ASD) has recently been described as a synaptopathy where dysregulation at the level of the synapse is thought to evoke an excitation–inhibition (E/I) imbalance implicated in its pathogenesis. The mechanisms through which alterations in glutamatergic signaling bring about an E/I imbalance remain elusive. Vesicular glutamate transporter 1 (VGLUT1) and postsynaptic density protein‐95 (PSD‐95) are two major regulatory proteins of glutamatergic signaling. This study aimed to determine whether a valproic acid‐induced (VPA) rat model of autism would be associated with changes in the protein expression levels of VGLUT1 and PSD‐95 in the prefrontal cortex (PFC) and cerebellar hemisphere (CH). Sprague‐Dawley rats were obtained from saline control (n = 3) and VPA‐induced (n = 3) groups. Consumption of VPA during pregnancy increases the propensity for the development of…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
FIGURE 1
FIGURE 2
FIGURE 3
FIGURE 4| Prefrontal cortex | Fold change | CV % | |||
|---|---|---|---|---|---|
| A | B | Mean ± SD | |||
| PSD‐95 | 1.09 | 1.39 | 1.24 ± 0.21 | 17 | |
| VGLUT1 | 62 kDa | 1.21 | 1.18 | 1.20 ± 0.02 | 2 |
| 55 kDa | 1.03 | 1.03 | 1.03 | 0 | |
| Cerebellar hemisphere | Fold change | CV % | |||
|---|---|---|---|---|---|
| A | B | Mean ± SD | |||
| PSD‐95 | 0.73 | 1.11 | 0.92 ± 0.27 | 29 | |
| VGLUT1 | 62 kDa | 1.32 | 1.24 | 1.28 ± 0.06 | 4 |
| 55 kDa | 0.65 | 0.45 | 0.55 ± 0.14 | 26 | |
- —Female Academic Leaders Fellowship
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsAutism Spectrum Disorder Research · Neuroscience and Neuropharmacology Research · Genetics and Neurodevelopmental Disorders
Introduction
1
Autism is a pervasive neurodevelopmental disorder defined by three main altered behavioral domains as described by the Diagnostic and Statistical Manual of Mental Disorders, 5th Edition (DSM‐V; American Psychiatric Association 2013). It forms part of autism spectrum disorders (ASD) which acknowledges the highly heterogenic phenotypic expression of autism (Faras et al. 2010; Sauer et al. 2021). The exact etiology of autism remains poorly understood. However, it has been attributed to a combination of risk factors; namely, immunological, environmental insults, epigenetic modifications, and certain susceptibility genes which can cause significant neuroanatomical and neurochemical rearrangements in the early development of the brain (Marotta et al. 2020). Particular attention has been devoted toward the excitatory/inhibitory imbalance (E/I imbalance) hypothesis proposed by Rubenstein and Merzenich (2003), who suggested that an imbalance in excitatory (glutamatergic) and inhibitory (γ‐aminobutyric acid; GABAergic) neurotransmission may be implicated in the pathophysiology of ASD (Hollestein et al. 2023).
The E/I balance refers to the homeostatic regulation of excitatory and inhibitory synaptic inputs either at the level of a single neuron or as an entirety, at the level of neural circuits (Chen et al. 2022). Homeostatic regulation is associated with an ideal brain state that is functionally differentiated allowing for the effective interpretation of salient representations (Yang et al. 2023). Rubenstein and Merzenich (2003) attributed the hyperexcitable state of cortical circuits in an autism model to an increase in the E/I ratio as a result of reduced GABAergic signaling. Inhibition is thought to sharpen the selectivity of excitatory postsynaptic potentials. Therefore, the loss of inhibition is translated to enhanced “noise” impacting information processing and the refinement of neural circuits which underlay sensory perception, memory, learning, behavior, and social engagement (Rubenstein and Merzenich 2003). The imbalance arises during critical periods of neurodevelopment where there is heightened neural plasticity increasing the propensity for aberrant neural circuitry (Colomar et al. 2023). However, contrary to this hypothesis, a decrease in the E/I balance has also been postulated (Gonçalves et al. 2017). This suggests that differences evoked in the E/I balance may be brain region‐specific due to the variable extents of connectivity throughout the brain (Manyukhina et al. 2022). Despite conflicting views with regards to the directionality of the E/I imbalance, it is plausible to consider the implication of an E/I imbalance in the pathogenesis of ASD (Uzunova et al. 2016).
In knowledge of this, autism has recently been considered a synaptopathy where dysfunction at the level of the synapse is thought to evoke an imbalance between glutamatergic (excitatory) and GABAergic (inhibitory) signaling (Uzunova et al. 2016). Synapses demand the tight regulation and structurally organized interaction between proteins at presynaptic terminals that navigate exocytotic release to those that facilitate reception at postsynaptic neurons. In this study, we investigated two proteins, vesicular glutamate transporter 1 (VGLUT1) and postsynaptic density protein‐95 (PSD‐95), involved in glutamatergic signaling. VGLUT1 orchestrates the active reuptake of glutamate into synaptic vesicles (SVs) of presynaptic terminals (Wojcik et al. 2004; Wilson et al. 2005; Herman et al. 2014). Thus, VGLUT1 is an essential protein for maintaining synaptic efficacy through the regulation of glutamate homeostasis where the extent of protein expression has a direct influence on the availability of glutamate in the synaptic cleft following exocytotic release. According to Herzog et al. (2006), VGLUT1‐deficient mice exhibited a significant reduction in glutamate quantal release. Reduced quantal release was correlated to an increased proportion of silent synapses which diminishes the elicitation of postsynaptic‐mediated currents (EPSCs), ultimately impacting the excitability of neurons. It is plausible that altered expression of VGLUT1 could be implicated in the development of an E/I imbalance in the prefrontal cortex (PFC) and cerebellar hemisphere (CH) of the valproic acid (VPA)‐induced rat model of autism.
At the postsynaptic membrane, PSD‐95 functions as a scaffolding protein where it coordinates and assembles neurotransmitter receptors and molecules involved in cell adhesion and signal transduction events. In conjunction with other proteins, it defines the molecular architecture and composition of the highly dynamic site referred to as the PSD (Feyder et al. 2010; Gao and Mack 2021). The most significant implication of PSD‐95 in autism is thought to be its involvement with inotropic receptors N‐methyl‐d‐aspartic acid receptors (NMDA) and α‐amino‐3‐hydroxyl‐5‐methyl‐4‐isox‐azoleproprionic acid receptors (AMPAR) (Coley and Gao 2019). NMDARs and AMPARs play a vital role in synaptic plasticity by influencing the frequency of neurotransmission at glutamatergic synapses (Coley and Gao 2019).
Synaptic plasticity encompasses both changes in the synaptic strength and efficacy of neuronal connections—fundamental to cortical development and cognitive processes such as learning and memory. It is achieved through long‐term potentiation (LTP) and long‐term depression (LTD) events, which involve the activity‐dependent strengthening and weakening of synapses, respectively. This infers the critical role of PSD‐95 in regulating synaptic plasticity due to its association and functional implications with NMDARs and AMPARs. PSD‐95 governs the highly dynamic recruitment, trafficking, and retention of these receptors within the PSD (Coley and Gao 2018). Several studies seem to suggest that reduced protein expression of PSD‐95 results in a decrease in AMPAR‐mediated current and a corresponding increase in NMDAR clustering and enhanced NMDAR‐mediated current, which in turn affects the induction of LTP in the PFC (Keith and El‐Husseini 2008; Béïque et al. 2006). Similarly, another study revealed that the overexpression of PSD‐95 resulted in an increase in AMPAR expression which translated to enhanced dendritic spine growth thereby, contributing to the characteristic hyperconnectivity in local circuitry seen in ASD (El‐Husseini et al. 2000). It is critical to investigate PSD‐95 as changes in the protein expression could a disrupt the E/I balance due to an alteration in synaptic plastic events and dendritic arborization, resulting in atypical connectivity seen across brain regions.
In the present study, a VPA‐induced rat model of autism was used to assess the protein expression levels of VGLUT1 and PSD‐95 in the PFC and CH. VPA is an antiepileptic drug and mood stabilizer used to treat the symptoms of bipolar psychiatric disorders (Favre et al. 2013). It was found that use during pregnancy increased the propensity for the development of autism in children—creating its wide spread application as an environmentally induced model (Phiel et al. 2001). This makes it a valuable model for investigating the possible molecular mechanisms which underlay the pathophysiology of the autistic phenotype and enable screening of therapeutic pursuits (Nicolini and Fahnestock 2018).
The PFC and CH are both pivotal brain regions pertaining to the pathophysiology of autism. The PFC is responsible for executive functions including decision‐making, self‐regulation, goal‐directed behavior, emotional responsivity, and working memory which are typically seen to be impaired in ASD (Leisman et al. 2023). Several studies have suggested significant changes in the connectivity and intrinsic neuronal properties of the PFC in ASD (Courchesne et al. 2011, Beopoulos et al. 2022). A study by Rinaldi et al. (2008) using the VPA‐induced rat model of autism revealed hyperconnectivity and enhanced plasticity rendering these local circuits more sensitive to external stimulation. This incidentally results in the retainment of redundant synaptic connections impacting the signal‐to‐noise ratio and the inability of distal brain regions to effectively orchestrate and modulate these activities—possibly accounting for some of the symptoms seen in ASD (Rinaldi et al. 2007). Another study by Coley and Gao (2019), who used PSD‐95 knockout mice revealed that the mice presented with learning and working memory difficulties, as well as a lack of sociability which they attributed to glutamatergic dysregulation in the medial PFC.
Recent studies also seem to suggest a possible role of cerebellar dysfunction in ASD. The cerebellum is often thought to be purely involved in unconscious motor control and coordination (Becker and Stoodley 2013). However, there appears to be a growing consensus on its involvement in cognitive functions, particularly that of social cognition which is a major behavioral deficit associated with ASD. Carta et al. (2019) proposed that the cerebellum greatly influences the reward circuitry of social behavior through its direct projection to the ventral tegmental area. Repeated stimulation of the cerebellum has been shown to increase the levels of circulating dopamine in the medial PFC of mice but in the case of ASD cerebello‐evoked dopamine release was not present (Rogers et al. 2011, 2013). Not only has the cerebellum been implicated in regulating dopamine release, changes in the neuroanatomy of the cerebellum are said to be among the most consistent findings in autism with the predominant pathology being the loss of Purkinje cells which constitute the major outflow of the cerebellum (Won et al. 2013; D'Mello et al. 2015). It is conceivable how this could impact the development of extensive and reciprocal connections that make up the cerebro‐cerebellum loops. D'Mello et al. (2015) proposed that possible disruptions in cerebro‐cerebellar loops could severely hinder the functional and structural specialization of cortical regions manifesting as the deficits seen in ASD. Given that the PFC and CH have been implicate in the neuropathology of autism and its associated symptomology, this motivates further investigation into these brain regions.
This study aimed to investigate whether the VPA‐induced rat (Rattus norvegicus) model of autism is associated with changes in the protein expression levels of PSD‐95 and VGLUT1 in the PFC and CH. Densitometric quantification of western blots was used to compare the extent of protein expression of VPA‐induced rats to saline‐treated controls. Findings from this research may offer important insights into the molecular mechanisms underlying the E/I imbalance in ASD, potentially through the modulation of synaptic plasticity and maturation during critical periods of neurodevelopment.
Methods and Materials
2
Experimental Design
2.1
The VPA‐Induced Model of Autism
2.1.1
The present study was issued an ethical clearance (2022/03/07/B) by the Animal Research Ethics Committee (AREC) of the University of the Witwatersrand, Johannesburg, South Africa. Pregnant Sprague‐Dawley (R. norvegicus) rats were allocated into two experimental groups: VPA‐induced and control groups. On gestational Day (GD) 13, one group of rats received a single intraperitoneal injection of 600 mg/kg of sodium valproate (Sigma‐Aldrich, P4543) dissolved in 2 mL of isotonic (0.9%) sterile saline, whereas the second was given an equivalent amount of isotonic saline on the same day (Yang et al. 2016). For better randomness, groups were divided randomly by a third party not involved in the study. Rats had access to water and rat chow ad libitum and were kept on a 12 h light/dark cycle at a constant room temperature (24°C ± 2°C). Births progressed naturally, and on postnatal Day (PD) 55, the rats were euthanized. The brains were immediately extracted and bisected along the midline into two hemispheres. For this study, the left PFC and cerebral hemisphere were used, with three male brains each for the experimental and control groups. Dissections were carried out on ice‐cold phosphate buffered saline (PBS) (0.01 M PBS, pH ∼ 7.4) under a dissecting microscope. The olfactory bulb was first removed, followed by a coronal cut approximately 2 mm anterior to the genu of corpus callosum, corresponding to the region presumed to be the PFC. In addition, the brainstem (including the pons, midbrain, and medulla) was carefully excised to facilitate the removal of the cerebellum. Isolated tissues, weighing approximately 0.025 g, were homogenized to a final concentration of 10% (w/v) in 0.01 M PBS using a G50 homogenizer (Coyote Biosciences, China). The resulting homogenates were immediately snap‐frozen in liquid nitrogen and stored at −80°C for subsequent tissue processing.
Total Protein Extraction From Rat Brain Tissue
2.1.2
For total protein extraction, a probe sonicator (Bandelin, Germany) was used to lyse the cells. Homogenates were sonicated on ice at an amplitude of 20%, alternating cycles of 5 s “On” and 10 s “Off” for a total of 2 min. This process facilitated the release of intracellular content and fragmentation of DNA, yielding uniform brain tissue lysates suitable for subsequent protein analyses.
Standardization of Brain Tissue Lysates Using a Bicinchoninic Acid (BCA) Assay
2.1.3
The BCA assay is a highly sensitive colorimetric method used for quantifying total protein content—necessary for standardization where an optimal protein concentration of brain homogenates can be determined to ensure reliable western blot detection across all proteins of interest. The Pierce BCA Assay Kit (Thermo Fischer Scientific, MA, USA) was used in accordance with manufacture instructions. A bovine serum albumin (BSA) standard curve was generated. Samples together with working solutions were added in duplicates as 10× and 20× dilutions, respectively. The absorbance was read using a Bio‐Rad iMark microplate reader (Microplate Manager Software, Bio‐Rad, USA) at a wavelength of 570 nm and analyzed using Microplate Manager 6, version 6.3.
Sodium Dodecyl Sulfate–Polyacrylamide Gel Electrophoresis (SDS–PAGE) and Western Blot
2.1.4
SDS–PAGE separates proteins according to size, before being transferred onto a nitrocellulose membrane during western blot for subsequent immunodetection.
Samples were run in duplicate on reducing SDS–PAGE using Laemmli gels comprised of a 4% (w/v) stacking gel (pH ∼ 6.8) and a 10% (w/v) separating gel (pH ∼ 8.8) in tank buffer (25 mM Tris–HCl, 192 mM glycine, 0.1% (w/v) SDS, pH ∼ 8.3) and transferred to a nitrocellulose membrane through a wet western blot transfer. The transfer was performed at a constant 35 V overnight placed at 4°C in the TE22 Mighty Small Transfer Tank (Hoefer Inc., USA) using blotting buffer (50 mM Tris–HCl, 200 mM glycine, pH ∼ 8.3, containing 0.1% (w/v) SDS and 20% (v/v) methanol).
Membranes were blocked in 3% (w/v) BSA for 1 h at room temperature, followed by a 10‐min Tris‐buffered saline wash (TBS; 0.2 M NaCl and 20 mM Tris–HCl, pH ∼ 7.4). Monoclonal primary antibodies were diluted in 0.5% (w/v) BSA in TBS (TBS‐BSA) as follows: anti‐PSD‐95 (1:1000, mouse ant‐rat, Abcam, ab13552), anti‐VGLUT1 (1:250, mouse anti‐rat, Abcam, ab134283), and anti‐β‐actin (1:1000, mouse anti‐rat, Biolegend, #653802). Incubations were for 1 h at room temperature. Membranes were washed with TBS, except the blots probed with anti‐β‐actin were washed in higher salt containing‐TBS (5 M NaCl, 200 mM Tris–HCl, pH ∼ 7.4) to minimize the extent of nonspecific binding. Membranes were incubated with goat anti‐mouse IgG secondary antibody conjugated to horseradish peroxidase (HRP) (1:5000, Abcam, ab205719) in 0.5% (w/v) BSA (TBS‐BSA) for 1 h at room temperature, followed by further TBS washes (3 × 5 min/wash). The HRP signal was developed using Pierce ECL Western Blotting Substrate (Thermo Fisher Scientific, MA, USA) and chemiluminescent signals were captured using the ChemiDoc MP imaging system (Bio‐Rad, USA).
Data Acquisition and Analysis
2.2
Densitometric Analysis
2.2.1
The Quantity One version 4.6.9 Software Program (Bio‐Rad, USA) was used to perform densitometric analysis of western blots. Protein expression levels of PSD‐95 and VGLUT1 were expressed as a function of trace quantity and normalized to β‐actin (internal loading control, ILC) using the following formula:
An averaged normalized trace quantity for each treatment group was calculated from the three biological replicates. Fold changes in PSD‐95 and VGLUT1 protein expression levels in the PFC and CH of VPA‐induced rats, relative to the controls, were then determined using the following formula:
Statistical Analysis
2.2.2
All continuous data were captured in Microsoft Excel (Microsoft Corporation, 2024) and expressed as mean ± standard deviation (SD). The statistical software R (R core team, Vienna, Austria, https://www.R‐project.org/, 2024) was used to run Shapiro–Wilk tests to determine normality of the data, and multiple unpaired Students t‐tests were performed to ascertain if there were any significant differences in the protein expression levels. A p < 0.05 was statistically significant. A percentage coefficient of variation (CV %) was used to assess the precision of technical replicates. Please refer to the Supporting Information for further information regarding trace quantity analysis and CV %.
Results
3
The BSA standard curve was used to make interpolations of total protein concentrations of brain tissue lysates to ensure standardization for western blots (Figure S2). Densitometric analysis of western blots allowed for ratiometric comparison between control and VPA‐induced rats to evaluate changes in the protein expression levels of PSD‐95 and VGLUT1 in the PFC and CH. Western blots were run in duplicate to assess the reproducibility of the assay and the precision of the results obtained. Representative blots have been used in the results section. Please see Supporting Information for all western blots conducted in this study following optimization as well as t‐test results obtained.
VPA‐Induced Autism Does Not Significantly Alter the Protein Expression Level of PSD‐95 in the PFC
3.1
Western blot analysis of PSD‐95, a scaffolding protein that is abundantly expressed within excitatory synapses, was performed to investigate the protein expression levels in a VPA‐induced rat model of autism compared to saline‐treated controls (Figure 1A and Table S1). Densitometric analysis revealed no statistically significant differences in mean normalized trace quantities between control and VPA‐induced rats in the PFC for both technical replicates A and B (p > 0.05) (Figure 1B). This indicates that a VPA‐induced model of autism does not have a discernible effect on the extent of protein expression of PSD‐95, in the PFC.
*VPA‐induced autism does not change the protein expression level of PSD‐95 in the PFC. (A) A representative western blot is shown for PSD‐95 at postnatal Day 55 for control and VPA‐induced rats, normalized to β‐actin. (B) Technical replicates A and B presented with no statistically significant difference in PSD‐95 protein expression in the PFC between control and VPA‐induced, when comparing normalized trace quantity values obtained from densitometric analysis. Data presented as mean ± SD. p < 0.05 was considered statistically significant. PFC, prefrontal cortex; PSD‐95, postsynaptic density protein‐95; VPA, valproic acid.
VGLUT1 Isoforms (62 and 55 kDa) Were Identified in the PFC With No Changes in the Extent of Protein Expression in a VPA‐Induced Rat Model of Autism
3.2
When assessing for changes in the protein expression level of VGLUT1 in the PFC between control and VPA‐induced rats, western blots revealed a prominent band corresponding to the anticipated molecular weight of 62 kDa. However, an additional distinct band at approximately 55 kDa was consistently identified for both treatment groups, following multiple rounds of optimization. This observation suggested the presence of two VGLUT1 isoforms, both of which contain a conserved linear epitope that is recognized by the antibody, enabling their simultaneous immunodetection (Figure 2A). Interestingly, both control and VPA‐induced groups appeared to have a greater relative abundance of the 55 kDa VGLUT1 isoform compared to the 62 kDa isoform, inferring its preferred upregulation in the PFC (refer to Table S1). For technical replicates A and B, mean normalized trace quantities revealed no statistically significant difference in the extent of protein expression of VGLUT1 isoforms when comparing VPA‐induced rats to the control (p > 0.05) (Figure 2B). This demonstrates the lack of association between VPA‐induced autism and the protein expression level of VGLUT1.
*VPA‐induced rats maintained a stable protein expression level of VGLUT1 isoforms relative to the control, with a greater abundance of 55 kDa isoform seen for both treatment groups. (A) A representative western blot is shown for VLGUT1 at postnatal Day 55 for control and VPA‐induced rats, normalized to β‐actin. Two VGLUT1 isoforms were detected through western blot analysis, with molecular weights of 62 and 55 kDa. (B) When comparing normalized trace quantity values between control and VPA‐induced rats, no statistically significant difference was observed (p > 0.05). Data presented as mean ± SD.p < 0.05 was considered statistically significant. PFC, prefrontal cortex; PSD‐95, postsynaptic density protein‐95; VGLUT1, vesicular glutamate transporter 1; VPA, valproic acid.
A mean fold change of 1.24 was calculated for the protein expression level of PSD‐95 in the PFC of VPA‐induced rats relative to the control. However, a CV % of 17 was obtained which indicates that there was a degree of variability between technical replicates impacting the precision of the assay (Table 1). For VGLUT1 isoforms, the mean fold changes in protein expression levels were 1.20 and 1.03 for 62 and 55 kDa isoforms, respectively, with technical replicates achieving CV % of ≤2. A low CV % of 2 reflected the consistency of the results suggesting that a high level of precision was attained between technical replicates.
In the CH, the Protein Expression Level of PSD‐95 Does Not Change in VPA‐Induced Rats Relative to the Control
3.3
When examining the protein expression levels of PSD‐95 in the CH, the quantification of western blots (Figure 3A) by densitometry revealed no pronounced changes between control and VPA‐induced rats. Indicated by the lack of statistically significant difference between the treatment groups (p > 0.05) (Figure 3B and refer to Table S2). This result demonstrates that the VPA‐induced rat model of autism is not associated with changes in the degree of protein expression of PSD‐95 in the CH.
*The protein expression level of PSD‐95 in the CH is not significantly altered in a VPA‐induced model of autism. (A) A Representative western blot is shown for PSD‐95 at postnatal Day 55 for control and VPA‐induced rats, normalized to β‐actin. (B) Densitometric analysis revealed that VPA‐induced rats maintain PSD‐95 protein expression levels consistent with that of control rats in the CH, further supported by the lack of statistical significance difference between them. Data presented as mean ± SD. p < 0.05 was considered statistically significant. CH, cerebellar hemisphere; PSD‐95, postsynaptic density protein‐95; VPA, valproic acid.
The VPA‐Induced Rat Model of Autism Was Not Associated With Changes in the Protein Expression Levels of VGLUT1 Isoforms in the CH
3.4
Similarly to the PFC, the same two VGLUT1 isoforms were evident in the CH. Qualitatively, there appeared to be a discernible increase in the protein expression level of the 62 kDa isoform in the CH compared that in the PFC (Figure 4A). This suggests that VGLUT1 isoforms may be differentially expressed in brain regions. In the CH, VPA‐induced autism seemed to have no considerable impact on the expression levels of VGLUT1 isoforms relative to the control (Figure 4B and Table S2). The absence of statistically significant difference in mean normalized trace quantities between VPA‐induced and control rats demonstrated that VPA‐induced autism does not evoke changes in the protein expression level of VGLUT1 isoforms in the CH (p > 0.05).
*Increased expression of 66 kDa isoform in the CH compared to the PFC, with no substantial difference in the degree of protein expression of VGLUT1 isoforms between control and VPA‐induced rats. (A) A Representative western blot is shown for VGLUT1 at postnatal Day 55 for control and VPA‐induced rats, normalized to β‐actin. A greater relative abundance of 62 kDa isoform of VGLUT1 was evident in the CH across both control and VPA‐induced rats compared to that in the PFC. (B) Densitometric quantification of western blots showed no statistical difference in the protein expression levels of VGLUT1 isomers in the CH of VPA‐induced rats relative to the saline‐controls (p > 0.05). This indicated that extent of protein expression of these regulatory proteins of glutamate quantal release remained largely unchanged. Thus, confirming the lack of association between protein expression levels of VGLUT1 isoforms and VPA‐induced autism in the CH. Data presented as mean ± SD. p < 0.05 was considered statistically significant. CH, cerebellar hemisphere; VGLUT1, vesicular glutamate transporter 1; VPA, valproic acid.
A mean relative fold change of 0.92 was calculated for PSD‐95 protein expression in the CH of VPA‐induced rats. A high level of variability was observed between technical replicates revealed by a CV % of 29—reflecting poor reproducibility of the assay. VGLUT1 isoforms had mean relative fold changes of 1.28 and 0.55 for 62 and 55 kDa, respectively (Table 2). For the 62 kDa VGLUT1 isoform a far greater level of precision was attained between replicates supported by a CV % of 4 compared to the 55 kDa VGLUT1 isoform where a CV % of 26 indicated a substantial degree of disparity between the results obtained for technical OLites A and B.
Discussion
4
To explore potential synaptic alterations, we assessed the expression levels of PSD‐95 and VGLUT1 in the PFC and CHs of rats prenatally exposed to VPA. Surprisingly, densitometric analysis of western blots revealed no significant changes in the expression of either synaptic marker in VPA‐exposed rats compared to saline‐treated controls—suggesting that, despite prenatal VPA exposure, baseline levels of these key synaptic proteins remain unaltered in these brain regions.
This was contrary to what was anticipated for PSD‐95 protein expression. A study by Coley and Gao (2019) utilized PSD‐95 knockout mice to reveal that a deficiency in PSD‐95 drastically altered the subunit composition of ionotropic glutamatergic receptors, particularly that of NMDA receptors. The change in subunit composition had a direct impact on the dynamics of glutamatergic synapses in the medial PFC and resulted in concurrent shifts of the E/I balance (Coley and Gao 2019). This was further supported by Béïque et al. (2006), who demonstrated a decrease in AMPAR‐mediated current as a reduction in the AMPA/NMDA ratio in the hippocampi of PSD‐95 knockout mice. A reduction in the ratio was thought to arise from a greater proportion of silent synapses in PSD‐95 deficient mice, where there would be an attenuation of glutamatergic neurotransmission (Béïque et al. 2006). These results are consistent with other findings who took an alternative approach where they investigated the overexpression of PSD‐95. As one would predict, this led to the opposite outcome where AMPA‐mediated current was enhanced due to an increase in the recruitment and function of AMPA receptors—an action thought to be mediated by PSD‐95 (El‐Husseini et al. 2000).
It is undeniable that changes in the extent of PSD‐95 protein expression can have a considerable impact on the functional plasticity of neural circuits. PSD‐95 is a protein that is intricately involved in activity‐dependent changes of synaptic strength through the regulation of neurotransmission receptors (AMPA and NMDA) within the postsynaptic density (Coley and Gao 2018). Studies that reported changes in the E/I balance all had one thing in common in that they actively induced changes in the protein expression level of PSD‐95 through the use of transgenic or knockout mice models of autism (Ergaz et al. 2016). These models were beneficial as they allowed for further investigation into how changes in PSD‐95 protein expression affect synaptic maturation and function, and whether these changes were associated with the autistic phenotype.
This makes one reflect on the choice of model used in the current study. The VPA‐induced model of autism continues to provide valuable insight into possible neurobiological mechanisms which underlay autism—owing to its widespread application (Nicolini and Fahnestock 2018). However, it does have limitations in that it does not capture the strong genetic component associated with autism etiology. It is well known that the VPA‐induced model better represents idiopathic autism thought to arise from a combination of environmental and epigenetic insults (Nicolini et al. 2015). In this study, the VPA‐induced model of autism was not associated with changes in the degree of protein expression of PSD‐95 in the PFC or CH which could suggest that PSD‐95 is not a molecular perpetuator in this form of ASD. But this result does not negate its role in other forms of ASD such as syndromic autism, which refers to incidents of ASD that are a part of a much bigger underlying medical condition, such as Fragile X Syndrome or Rett Syndrome (Rojas 2014). In the case of Fragile X Syndrome, this is one such incidence where syndromic autism has been shown to alter the protein expression of PSD‐95 (Westmark 2013; McCary and Roberts 2012). Therefore, it may prove to be beneficial to investigate the expression of PSD‐95 in combination with autism‐associated susceptibility genes. This finding highlights the need to identify prospective biomarkers that would allow for further delineation between forms of ASD. Biomarker delineation would address the diagnostic challenges that arise from the use of a broad “umbrella” term that relies solely on behavioral characteristics that are not unique to its manifestation—adding further complexity to the disorder. A further limitation is the absence of female brain analysis, which could have offered some more insight on the degree of these two proteins’ expression in female autistic brains.
Despite the surprising results attained in this study, which revealed no association between VPA‐induced autism and PSD‐95 protein expression levels, the role of PSD‐95 as a potential molecular mechanism through which an E/I imbalance could be evoked should not be disregarded. Autism is a disorder of connectivity and is thought to originate at the site of the synapse (Guang et al. 2018). Thus, it was valuable to assess the possible variation in PSD‐95 protein expression as it is an abundant and an integral scaffolding molecule within the PSD (Gao and Mack 2021). It is plausible that the expression level of other high‐risk autism proteins such as src homology‐3 domain (SH3), multiple Ankyrin repeat domains 3 (SHANK3), neuroligin, and neurexin which associate with PSD‐95 change, rather than PSD‐95 itself (Jiang et al. 2022; Nisar et al. 2022).
Autism is a disorder that arises over critical periods of neurodevelopment where there is heightened synaptic plasticity (Colomar et al. 2023). This makes it difficult to distinguish primary autism deficits from compensatory mechanisms that mediate homeostatic adjustments. The protein expression level of PSD‐95 in control and VPA‐induced rats was assessed at a single point in time. It could prove insightful to conduct a time‐dependent investigation that would allow for spatiotemporal evaluation of PSD‐95 protein expression as it may fluctuate in response to other underlying molecular events.
Mean fold changes in PSD‐95 expression were 1.24 in the PFC and 0.92 in CH, with coefficients of variation of 17% and 27%, respectively. The percentage coefficients of variation, used to assess the reproducibility of the assay, suggest that there was a degree of discrepancy between technical replicates. The same was evident between technical replicates used to assess the protein expression level of VGLUT1. This is understandable as no two gels are the same and western blot assays are susceptible to a high degree of variability as it is a technique that involves a series of independent steps (Mahmood and Yang 2012). In knowledge of inter‐replicate variability, analysis within each gel showed reproducible results between the treatment groups indicating that internal consistency was maintained. Each technical replicate demonstrated that VPA‐induced autism is not associated with changes in the extent of protein expression of either synaptic proteins in the PFC or CH, maintaining relatively stable protein expression to that of the saline‐treated controls.
To date no studies have pursued isoform‐level characterization of VGLUT1. Most have focused on comparisons among VGLUT variants 1–3, derived from a family of related genes, each having unique expression profile allowing for functional and developmental specialization within brain regions (Berry et al. 2012). A salient finding in this study was the detection of two protein isoforms of VGLUT1 in the PFC and CH of VPA‐induced and control rats. Isoforms are functionally similar proteins that arise from alternative splicing of a single gene (Schlüter et al. 2009). Both brain regions had co‐expression of VGLUT1 isoforms but through qualitative assessment, one could evidently see a greater degree of 62 kDa isoform expression in the CH of control and VPA‐induced rats compared to that in the PFC, where there appeared to be preferential expression of 55 kDa isoform. The regional‐differential expression of VGLUT1 isoforms in the PFC and CH may be largely attributed to epigenetic regulation (Pozo et al. 2021). Wojcik et al. (2004) reported that the presynaptic regulation of glutamate quantal release is not dependent upon the VGLUT1 isoform but rather the level of expression of VGLUT1, which was of interest in this study.
As no significant changes in the protein expression levels of VGLUT1 isoforms were detected in the PFC or CH, this diminishes its implication in the development of an E/I imbalance in a VPA‐induced rat model of autism. Lenart et al. (2020) revealed changes in VGLUT1‐encoding genes in a VPA‐induced model of autism but these changes often cannot be effectively translated to proteins in terms of the extent of expression and functionality. Another study showed that VPA‐induced mice had higher basal levels of extracellular glutamate in the prelimbic PFC indicative of hyper‐glutamatergic function (Kurahashi et al. 2024). However, they too investigated VGLUT1 and confirmed no difference in the level of protein expression (Kurahashi et al. 2024). This suggests that VPA‐induced autism is associated with dysregulation in glutamate signaling, but not as a result of changes in the protein expression of VGLUT1. This shifts the view toward inotropic and metabotropic glutamate receptors, their availability, and the removal of excess glutamate from the synaptic cleft by excitatory amino acid transporters (EAAT)—all of which are key regulators of glutamate homeostasis (Montanari et al. 2022; Nisar et al. 2022; O'Donovan et al. 2017).
Conclusion
5
The E/I hypothesis proposed by Rubenstein and Merzenich (2003) is considered to be the main pathophysiological mechanism of ASD. However, little is known as to how this imbalance arises, its directionality, and whether it differs across brain regions. We aimed to make an advancement into the pursuit of understanding possible molecular mechanisms that elicit such an imbalance as a consequence of synaptic dysregulation. Densitometric quantification of western blots was used to ratiometrically compare saline‐treated controls and VPA‐induced rats to evaluate changes in the protein expression levels of PSD‐95 and VGLUT1 in the PFC and CH. The VPA‐induced rat model of autism was not associated with changes in the extent of protein expression of PSD‐95 and VGLUT1 in the PFC and CH. This reduces the implication of PSD‐95 and VGLUT1 in evoking an E/I imbalance thought to underlay VPA‐induced autism. The high heterogeneity of autism speaks to the complex biological substrates upon which an E/I imbalance could be elicited. There is no disputing the critical role of glutamate in brain development and the refinement of neural circuitry emphasizing the need for a sustained effort to determine where in glutamatergic neurotransmission is dysregulation being evoked in ASD.
Author Contributions
Conceptualization: Busisiwe Constance Maseko. Methodology: Megan Reveley, Gavin Owen, Nancy Tumba, Oladiran Ibukunolu Olateju, and Busisiwe Constance Maseko. Formal analysis: Megan Reveley, Gavin Owen, Nancy Tumba, Oladiran Ibukunolu Olateju, and Busisiwe Constance Maseko. Investigation: Megan Reveley, Gavin Owen, Nancy Tumba, Oladiran Ibukunolu Olateju, and Busisiwe Constance Maseko. Resources: Busisiwe Constance Maseko. Data curation: Megan Reveley, Gavin Owen, Nancy Tumba, Oladiran Ibukunolu Olateju, and Busisiwe Constance Maseko. Writing – original draft preparation: Megan Reveley. Writing – review and editing: Gavin Owen, Nancy Tumba, Oladiran Ibukunolu Olateju, and Busisiwe Constance Maseko. Visualization: Megan Reveley, Gavin Owen, Nancy Tumba, Oladiran Ibukunolu Olateju, and Busisiwe Constance Maseko. Supervision: Busisiwe Constance Maseko, Gavin Owen, Nancy Tumba, and Oladiran Ibukunolu Olateju. Project administration: Busisiwe Constance Maseko. Funding acquisition: Busisiwe Constance Maseko.
Conflicts of Interest
The authors declare no conflicts of interest.
Supporting information
Supporting Information: dneu70018‐sup‐0001‐SuppMat.pdf
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1American Psychiatric Association . 2013. Diagnostic and Statistical Manual of Mental Disorders. 5th ed. American Psychiatric Publishing.
- 2Becker, E. B. E. , and C. J. Stoodley . 2013. “Autism Spectrum Disorder and the Cerebellum.” In International Review of Neurobiology, Neurobiology of Autism, edited by G. Konopka , 1–34. Academic Press. 10.1016/B 978-0-12-418700-9.00001-0.24290381 · doi ↗ · pubmed ↗
- 3Béïque, J.‐C. , D.‐T. Lin , M.‐G. Kang , H. Aizawa , K. Takamiya , and R. L. Huganir . 2006. “Synapse‐Specific Regulation of AMPA Receptor Function by PSD‐95.” Proceedings of the National Academy of Sciences of the United States of America 103: 19535–19540. 10.1073/pnas.0608492103.17148601 PMC 1748260 · doi ↗ · pubmed ↗
- 4Beopoulos, A. , M. Géa , A. Fasano , and F. Iris . 2022. “Autism Spectrum Disorders Pathogenesis: Toward a Comprehensive Model Based on Neuroanatomic and Neurodevelopment Considerations.” Frontiers in Neuroscience 16: 988735. 10.3389/fnins.2022.988735.36408388 PMC 9671112 · doi ↗ · pubmed ↗
- 5Berry, C. T. , M. P. Sceniak , L. Zhou , and S. L. Sabo . 2012. “Developmental Up‐Regulation of Vesicular Glutamate Transporter‐1 Promotes Neocortical Presynaptic Terminal Development.” P Lo S ONE 7: e 50911. 10.1371/journal.pone.0050911.23226425 PMC 3511412 · doi ↗ · pubmed ↗
- 6Carta, I. , C. H. Chen , A. L. Schott , S. Dorizan , and K. Khodakhah . 2019. “Cerebellar Modulation of the Reward Circuitry and Social Behavior.” Science 363: eaav 0581. 10.1126/science.aav 0581.30655412 PMC 6711161 · doi ↗ · pubmed ↗
- 7Chen, L. , X. Li , M. Tjia , and S. Thapliyal . 2022. “Homeostatic Plasticity and Excitation‐Inhibition Balance: The Good, the Bad and the Ugly.” Current Opinion in Neurobiology 75: 102553. 10.1016/j.conb.2022.102553.35594578 PMC 9477500 · doi ↗ · pubmed ↗
- 8Coley, A. A. , and W.‐J. Gao . 2019. “PSD‐95 Deficiency Disrupts PFC‐Associated Function and Behavior During Neurodevelopment.” Scientific Reports 9: 9486. 10.1038/s 41598-019-45971-w.31263190 PMC 6602948 · doi ↗ · pubmed ↗