Memory Decline and Aberration of Synaptic Proteins in X-Linked Moesin Knockout Male Mice
Article information
Abstract
Objective
This study aims to investigate may moesin deficiency resulted in neurodevelopmental abnormalities caused by negative impact on synaptic signaling ultimately leading to synaptic structure and plasticity.
Methods
Behavioral assessments measured neurodevelopment (surface righting, negative geotaxis, cliff avoidance), anxiety (open field test, elevated plus maze test), and memory (passive avoidance test, Y-maze test) in moesin-knockout mice (KO) compared to wild-type mice (WT). Whole exome sequencing (WES) of brain (KO vs. WT) and analysis of synaptic proteins were performed to determine the disruption of signal pathways downstream of moesin. Risperidone, a therapeutic agent, was utilized to reverse the neurodevelopmental aberrance in moesin KO.
Results
Moesin-KO pups exhibited decrease in the surface righting ability on postnatal day 7 (p<0.05) and increase in time spent in the closed arms (p<0.01), showing increased anxiety-like behavior. WES revealed mutations in pathway aberration in neuron projection, actin filament-based processes, and neuronal migration in KO. Decreased cell viability (p<0.001) and expression of soluble NSF adapter protein 25 (SNAP25) (p<0.001) and postsynaptic density protein 95 (PSD95) (p<0.01) was observed in days in vitro 7 neurons. Downregulation of synaptic proteins, and altered phosphorylation levels of Synapsin I, mammalian uncoordinated 18 (MUNC18), extracellular signal-regulated kinase (ERK), and cAMP response element-binding protein (CREB) was observed in KO cortex and hippocampus. Risperidone reversed the memory impairment in the passive avoidance test and the spontaneous alternation percentage in the Y maze test. Risperidone also restored the reduced expression of PSD95 (p<0.01) and the phosphorylation of Synapsin at Ser605 (p<0.05) and Ser549 (p<0.001) in the cortex of moesin-KO.
Conclusion
Moesin deficiency leads to neurodevelopmental delay and memory decline, which may be caused through altered regulation in synaptic proteins and function.
INTRODUCTION
Moesin, a member of the ezrin, radixin and moeisn (ERM) protein family, plays a key role in cellular structure by linking the cell membrane to the actin cytoskeleton [1,2]. In neurons, actin polymerization, the process of converting G-actin to Factin, is essential for synaptic plasticity, particularly in the structural changes associated with dendritic spines [3,4].
Deficiency of moesin disrupt actin dynamics, leading to instability of the actin cytoskeleton [5-7], which affects synaptic function, particularly maintaining the neuronal structures necessary for proper signaling and plasticity. Actin regulate synaptic vesicle trafficking at nerve terminals, through binding of Synapsin to actin filaments. Therefore, disruption in actin polymerization, can impair the ability of Synapsin protein to bind synaptic vesicles, altering vesicle release and synaptic function. In Drosophila, moesin was essential for neuronal morphogenesis and long-term memory, with knockdown experiments revealing significant alterations in neuronal cytoskeletal structure [6].
Synapsins are critical regulators of synaptic plasticity, directly involved in synapse formation, maturation, and neurotransmitter release [8-11]. Synapsins interact with actin filaments to regulated by various proteins and phosphatases, including mitogen-activated protein kinase (MAPK) and protein kinase A (PKA), influencing their interactions with synaptic vesicles, actin filaments, and other synaptic proteins [11,12]. Synapsin is phosphorylated at distinct sites by different kinases. Cyclin dependent kinase 5 (CDK5), as a MAPK, phosphorylate Ser549 of Synapsin [13], and this phosphorylation modulates the interaction between Synapsin 1-associated synaptic vesicle dynamics, and cytoskeletal organization [14]. Moreover, we have previously reported that clozapine-induced chloride channel 4 and moesin activation modulated CDK5 expression [15].
Extracellular signal-regulated kinase (ERK), a kinase activated downstream of CDK5, also plays a pivotal role in synaptic signaling. ERK phosphorylates Synapsin at specific sites (Ser-9) [16], reducing its affinity for actin and synaptic vesicles, thus enhancing vesicle availability for release [17]. ERK regulate actin dynamics within the postsynaptic density (PSD), contributing to the structural stability of dendritic spines. ERK activation further leads to the phosphorylation and activation of cAMP response element-binding protein (CREB), a transcription factor that promotes the expression of genes necessary for synaptic growth, stability, and plasticity [18,19]. CREB-driven gene expression supports the long-term structural changes in synapses, such as spine growth and increased receptor density, which are crucial for sustained synaptic plasticity [19,20].
The PSD is another key structure that relies on actin filaments to maintain synaptic stability and facilitate synaptic remodeling [3,4]. Recent studies indicate that PSD scaffold proteins, including SAPAP, Shank, and Homer, can drive actin polymerization and bundling independently of other regulatory proteins, underscoring the importance of actin in PSD function [21]. Moesin deficiency may weaken these critical PSD-actin interactions, leading to structural and functional changes in synapses. These findings suggest that the deficiency of moesin disrupt actin polymerization, which in turn affects Synapsin functionality and the integrity of the PSD. We hypothesized that moesin dysfunction may lead to changes in synaptic plasticity and results in neurodevelopmental abnormalities.
Risperidone, a second-generation antipsychotic, exerts its therapeutic effects not only by antagonizing dopamine D2 and serotonin 5-hydroxytryptamine 2A receptors but also by influencing the molecular architecture of synapses [22-24]. Studies have shown that the expression of these PSD-related genes is dose-dependent and that risperidone progressively recruits different cortical and subcortical regions, ultimately reorganizing synaptic networks. This reorganization of synaptic structure, mediated by the modulation of glutamatergic and dopaminergic signaling pathways, may underlie risperidone’s efficacy in improving cognitive deficits and behavioral disorders in neuropsychiatric conditions.
This study hypothesizes that moesin deficiency disrupts actin dynamics within neurons, negatively impacting synaptic structure and function, which may impair synaptic signaling, ultimately leading to behavioral changes associated with neurodevelopmental abnormalities. Furthermore, we investigated whether risperidone, a drug commonly used for behavioral management in neurodevelopmental conditions can restore synaptic function and mitigate the synaptic vesicle trafficking deficits induced by moesin deficiency. This study aims to test these hypotheses, shedding light on the role of moesin in neurodevelopment and synaptic function, while evaluating the mechanism underneath therapeutic potential of risperidone in neurodevelopmental deficits.
METHODS
Generation of moesin-knockout mice
Moesin-knockout mice (KO) was generated by Macrogen (Seoul, South Korea) using the CRISPR/Cas9 system. The moesin-KO were created by targeting the Exon 2 region located at qC3 on the X chromosome (Figure 1A). All experiments were performed using moesin-KO (male hemizygous, female homozygous). Male hemizygous moesin-KO and female heterozygous moesin-KO C57BL/6N mice (F1) were interbred to obtain second-generation mice. Subsequently, the hemizygous male and female homozygous KO from the F2 generation were used. Throughout all the experiments, gene modifications were randomly confirmed by western blot analysis. All animal protocols were approved by the National Center for Mental Health and Dongguk University (NCMH-2108-001-004-01, IACUC-2020-001-1, IACUC-2021-054-1, and IBC-2021-007-02) and were performed in accordance with the National Institute of Health guidelines. The animals were housed in a controlled environment at a constant temperature (22°C±1°C) with a 12 hours light/12 hours dark cycle (lights on at 9:00 A.M.) and given free access to food and water.

Schematic representation of the experimental design and procedures. A: Exon 2 of the moesin gene on chromosome X was deleted to generate moesin-KO. Neuronal cells were obtained from the cortex of WT or moesin-KO on embryonic day 15. B: Neurodevelopmental testing was performed at postnatal day 7 for hemizygous male KO and homozygous female KO. The open field test was used to assess anxiety-related behavior in male hemizygous KO at the age of 2 months, and the elevated plus maze test was used to assess anxiety-related at the age of 5 months. C: Mice were intraperitoneally injected with saline or risperidone for 2 weeks (5 days per week). After a 1-week washout period, behavioral tests for memory function (passive avoidance test and Y-maze test) were performed at 8 weeks (2 months) of age. D: Exome sequencing and sperm analysis were conducted at 3 months of age. KO, knockout mice; WT, wild-type mice.
Schematic representation of the experimental design and procedures according to mice age
Neuronal cells were obtained from the cortices of wild-type mice (WT) or moesin-KO on embryonic day 15 (Figure 1A). Neurodevelopmental testing was performed on postnatal day 7 (PND7) in hemizygous male and homozygous female KO. The open field test was used to assess anxiety-related behavior in male hemizygous KO at 2 months of age of 2 months, and the elevated plus maze test was used at 5 months age of 5 months (Figure 1B). For the risperidone treatment experiment, male hemizygous KO were intraperitoneally injected with saline (WT, KO) or risperidone (KO-risperidone) for 2 weeks (5 days per week). After a 1-week washout period to minimize the effects of chronic stress on learning and memory, behavioral tests for memory function (passive avoidance and Y-maze tests) were performed at 8 weeks of age (2 months) (Figure 1C). Exome sequencing and sperm analysis were conducted at 3 months of age (Figure 1D).
Whole exome sequencing (WES)
For the generation of standard exome capture libraries, we used the Agilent SureSelect Target Enrichment protocol for the Illumina paired-end sequencing library (Version C2, December 2018) together with 1 μg input genomic DNA (gDNA). The SureSelectXT Mouse All Exon probe set was used in all cases. Quantification of DNA and quality were performed using PicoGreen and agarose gel electrophoresis. We used 1 μg of each cell line’s gDNA diluted in EB Buffer. We sheared to a target peak size of 150–200 bp using the Covaris LE220 focused-ultrasonicator (Covaris, Woburn, MA, USA) according to the manufacturer’s recommendations. Load the 8 microTUBE Strip into the tube holder of the ultrasonicator and shear the DNA using the following settings: mode, frequency sweeping; duty cycle, 10%; intensity, 5; cycles per burst, 200; duration, 60 seconds×6 cycles; temperature, 4°C–7°C. The fragmented DNA is repaired, an ‘A’ is ligated to the 3' ends, and Agilent adapters are then ligated to the fragments. Once ligation is assessed, the adapter-ligated product is polymerase chain reaction (PCR) amplified by PCR. For exome capture, 250 ng of DNA library was mixed with hybridization buffers, blocking mixes, RNase block, and 5 μL of SureSelect all exon capture library, according to the standard Agilent SureSelect Target Enrichment protocol. Hybridization to the capture baits was conducted at 65°C using a heated thermal cycler lid option at 105°C for 24 hours on a PCR machine. The captured DNA was washed and amplified. The final purified product was quantified using quantitative PCR (qPCR) according to the qPCR Quantification Protocol Guide (KAPA Library Quantification kits for Illumina Sequencing platforms) and qualified using TapeStation DNA Screentape D1000 (Agilent). Sequencing was performed using the NovaSeq platform (Illumina, San Diego, CA, USA). We identified a list of genes that showed variations compared with the reference genes. This list was then subjected to gene ontology (GO) term analysis using the Metascape software.
Sperm analysis
The left cauda epididymis was placed in 2 mL of saline solution at 37°C and cut into small pieces to facilitate complete sperm release, maintaining the temperature at 37°C for 5 minutes. The sperm sample was then diluted 10-fold, and 10 μL of this sample was placed between a cover glass and a hemocytometer to assess sperm count and morphology.
Behavioral testing for neurodevelopment in infant mice
Neurodevelopmental testing was performed on PND7 in hemizygous male and homozygous female KO (Figure 1B).
Surface righting test
Surface righting was scored as follows: 2 points for righting within 1 second, 1 point for righting after more than 1 but less than 2 seconds, and 0 points for righting after more than 2 seconds [25].
Negative geotaxis test
At PND7, the time taken for the pups to steer to a head-up position after being placed in a head-down position on a 30° incline was recorded. The following scoring scale was used: 0, no response within 60 seconds; 1, response within 60 seconds; and 2, response within 30 seconds [25].
Cliff avoidance test
Pups were recorded as either withdrawing or turning away from the edge of the platform within 20 seconds of placement, with their chest touching the edge of the platform (10×10×10 cm). Zero points were given if the mouse did not respond within 20 seconds, 1 point if it ran backward, and 2 points if it dodged or turned around [25].
Behavioral testing for anxiety in young and mature adult mice
X-linked gene hemizygosity leads to recessive phenotypes in males, resulting in a higher prevalence of neurodevelopmental disorders in males [26]. We used hemizygous male KO to better understand the molecular basis of the X-linked neurodevelopmental disorder phenotype (Figure 2B).

Moesin-KO show abnormal behaviors. We investigated whether moesin-KO exhibit symptoms of developmental disorders. An open field test to measure anxiety status (A and B), and an elevated plus-maze test was performed on male mice 5 months of age (C). All data were analyzed using the t-test and are presented as the mean±standard error of the mean of six determinations. *p<0.05; **p<0.01. KO, knockout mice; WT, wild-type mice.
Open field test
We measured the time for which the subject remained adjacent to the outer wall of the maze compared to the center zone, which is indicative of anxiety-like behavior. The test that was conducted in a previous study [27].
Elevated plus maze test
The apparatus was a maze with a plus (+) sign elevated above the floor. It consisted of two closed arms positioned opposite each other, two open arms positioned opposite each other, and a central area. We measured the time that the participants spent in the center, open-arm, or closed-arm zones of the maze. Throughout the experiment, a video camera positioned above the maze captured their movements for 5 minutes and was subsequently analyzed using the SMART3.0 program.
Learning and memory testing after drug administration in adolescent mice
Risperidone, a drug approved by the US Food and Drug Administration for the treatment of children with autism from the age of 5 years [28,29] was treated to determine whether the behavioral and molecular aberrance observed during development in moesin-KO [30] could be reversed.
Drug administration
Risperidone (Sigma-Aldrich, St. Louis, MO, USA) was dissolved in 25% hydroxypropyl-β-cyclodextrin (TCI, Sigma-Aldrich) at a concentration of 10 g/0.1 mL. One-month-old mice were divided into three groups: control group (n=13; WT, intraperitoneal injection of saline), KO group (n=10; KO, intraperitoneal injection of saline), and KO-risperidone group (n=10; KO+Ris, intraperitoneal injection of 0.2 mg/kg risperidone). Mice were intraperitoneally injected with saline (WT and KO) or risperidone (KO-risperidone) for 2 weeks (5 days per week). After a 1-week washout period, behavioral tests were performed at 8 weeks (2 months) of age (Figure 1C).
Passive avoidance test
The passive avoidance test was used to measure memory function. The test that was conducted in a previous study [31].
Y-maze test
Y-maze tests were performed on hemizygous male KO at 8 weeks (2 months) after 2 weeks of drug administration and 1-week washout. The Y-maze was used to assess short-term memory. The test that was conducted in previous study [32].
Neuronal cell culture
Primary cortical neurons were plated on a 24-well plate to analyze cell viability. On days in vitro 7 (DIV7), a 3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl 2H-tetrazolium bromide (MTT) solution was added to the cell culture medium. After 1 hour, it was dissolved in DMSO and measured using a spectrometer at a wavelength of 580 nm (Molecular Devices, San Jose, CA, USA).
Western blotting
Prefrontal cortical and hippocampal tissues were isolated from 3-month-old male KO and WT male mice at 3 months of age. Brain tissues were homogenized using a tissue grinder containing 10% w/v buffer, 50 mM Tris-base (pH 7.5), 10 mM sodium pyrophosphate, 1% NP-40, a phosphatase inhibitor cocktail, and a protease inhibitor cocktail. Cortical neurons were lysed in a sample buffer containing 6.25 mM Tris (pH 6.8), 1% sodium dodecyl sulfate, 10% glycerol, and 0.0125% bromophenol. The tissue lysates (30 μg) were electrophoresed and transferred to 0.45 μm nitrocellulose membranes, which were incubated with anti-GFAP (ab7260, 1:1,000, Abcam, Cambridge, MA, USA), anti-MAP2 (ab5392, 1:1,000, Abcam), anti-Sox2 (#4900, CST, 1:1,000), anti-DCX (#91954, CST, 1:1,000), anti-soluble NSF adapter protein 25 (SNAP25) (#5308, CST, 1:1,000), anti-postsynaptic density protein 95 (PSD95) (#3409, CST, 1:1,000), anti-phospho-PSD95 (#45737, CST, 1:1,000), or anti-moesin (#3146, CST, 1:1,000), anti-syntaxin1A (#18572, CST, 1:1,000), anti-synaptophysin (#5467, CST, 1:1,000), anti-CDK5 (#14145, CST, 1:1,000), anti-radixin (#2636, CST, 1:1,000), anti-synapsin 1 (#2312, CST, 1:1,000), anti-phospho-synapsin 1 (Ser9) (#2311, CST, 1:1,000), anti-phospho-synapsin 1 (Ser605) (#88246, CST, 1:1,000), anti-phospho-synapsin 1 (Ser549) (#PA1-4697, 1:1,000; Thermo Fisher Scientific, Waltham, MA, USA), anti-ERK (#4695, CST, 1:1,000), anti-phospho-ERK (#4370, CST, 1:1,000), anti-CREB (#9197, CST, 1:1,000), anti-phospho-CREB (#9190, CST, 1:1,000), anti-mammalian uncoordinated18 (MUNC18) (#13414, CST, 1:1,000), anti-phospho-MUNC18 (ab183484, 1:1,000, Abcam) anti-β-actin (#4967, CST, 1:1,000) for 16 hours at 4°C. After primary antibody incubation, each membrane was incubated with goat anti-rabbit horseradish peroxidase-conjugated antibody (1:5,000; Thermo Fisher Scientific) or goat anti-mouse horseradish peroxidase-conjugated antibody (1:5,000; Thermo Fisher Scientific) for 1 hour at room temperature. Band images were acquired using a Molecular Imager ChemiDoc XRS+ (Bio-Rad, Hercules, CA, USA), and band densities were analyzed by using Image Lab TM software version 2.0.1 (Bio-Rad).
Statistical analysis
For sample size calculation, we referred to previous studies with similar research designs [33-35] and then calculated the sample sizes using the G*Power program. We performed the Shapiro-Wilk test for normality before conducting an independent t-test. All data values were analyzed using an independent t-test and expressed as the mean±standard error of the mean. P<0.05 was set as statistically significant. For western blot analysis, the first three control samples were set to 100%, and the other samples were analyzed. In the risperidone administration experiment, differences between the WT and KO groups, followed by differences between the KO and KO+ risperidone groups, were examined using independent t-tests.
RESULTS
Moesin-KO showed abnormal motor development in infancy and increased anxiety behaviors during adolescence and adulthood
First, we investigated whether moesin deficiency leads to behavioral changes associated with neurodevelopmental abnormalities. Surface righting, negative geotaxis, and cliff avoidance tests were performed on 7-day-old mice to observe whether moesin-deficient mice exhibit symptoms of developmental disorders. Male moesin-KO pups showed a significant decrease in scores in the surface righting test (WT, 1.8±0.2; KO, 0.8±0.4; p<0.05, n=5) but not in the cliff avoidance test on PND7 (WT, 1.6±0.4; KO, 1.4±0.2; n=5) and the negative geotaxis test (WT, 2.0±0.0; KO, 2.0±0.0; n=5) (Table 1). Female moesin-KO pups showed a significant decrease in scores in the surface righting test (WT, 2.0±0.0; KO, 0.8±0.2; p<0.001, n=5) and in the cliff avoidance test on PND7 (WT, 2.0±0.0; KO, 1.4±0.2; p<0.05, n=5) but not in the negative geotaxis test (WT, 2.0±0.0; KO, 1.8±0.2; n=5) (Table 1).

Surface righting, negative geotaxis, and cliff avoidance were performed on 7-day-old male and female mice to observe whether moesin-KO show symptoms of developmental disorders
An open field test was performed to measure the anxiety status in male mice at 2 months (adolescence). KO showed a significant decrease in the center distance traveled (WT, 251.0±35.0; KO, 118.0±22.0; n=7, p<0.01), the time spent in the center (WT, 34.6±10.5; KO, 10.0±2.0; n=7, p<0.05) and increase in the time spent in the periphery (WT, 265.4±10.5; KO, 290.4±2.0; n=7, p<0.05) compared to WT (Figure 2A and B), suggesting increased anxiety-like behavior [36].
We performed the elevated plus maze test [37], the gold standard paradigm for anxiety. We observed that male hemizygous KO continued to show anxiety-like behavior at 5 months. KO showed a significant decrease in time spent in the center zone (WT, 78.7±11.4, n=13; KO, 49.4±4.3, n=20; p<0.01) and open arms zone (WT, 54.2±9.5, n=13; KO, 24.9±8.6, n=20; p<0.05), and a significant increase in time spent in the closed arms zone (WT, 167.2±15.4, n=13; KO, 225.8±11.1, n=20; p< 0.01) compared to WT (Figure 2C), suggesting increased anxiety-like behavior.
Detection of variants from WES data
Next, we performed WES in moesin-KO to determine the disruption of signal pathways caused by moesin deficiency. An in-depth analysis of WES data was conducted to identify chromosomal alterations across the entire genome. The data revealed a distinct distribution pattern of alternative bases across chromosomes (Figure 3A). Chromosome 1 exhibited the highest number of base alterations, indicating a notable genomic variability. Further classification of the alterations revealed that exonic variants constituted the majority, with 215 identified changes, indicating a significant impact on proteincoding regions (Figure 3B). Within the single nucleotide variants, transitions such as G>A, C>T, and A>G were the most prevalent, following that respective order (Figure 3C).

Exome sequencing results in moesin-KO compared to reference genes. A: Number of altered bases per chromosome. This figure shows a bar graph depicting the distribution of altered bases across different chromosomes in the moesin knockout condition. B: Counts of different types of altered bases. This result provides a breakdown of the alterations by type, including substitutions, insertions, and deletions. C: SNV class of altered bases. This graph categorizes the SNVs identified in the exome sequencing into their respective classes based on their genomic impact. KO, knockout mice; SNV, single nucleotide variant.
The Y chromosome exhibits a single gene alteration. Upon comparing the Ddx3y gene on the Y chromosome, which is known for its reduced expression potentially leading to azoospermia, to the reference gene, we identified three base alterations at positions G>A 1277438, C>T 1277449, and G>A 1277454. In addition, a mutation in Dmrtc1a, a transcription factor expressed in the testes and involved in spermatogene-sis, was also identified on the X chromosome.
Because sperms are easily accessible and have a cytoskeletal structure that depends on actin filaments for both motility and head morphology [38], we measured sperm motility and morphological changes to confirm that KO indeed possess actin filament abnormalities caused by moesin deficiency. Analysis of spermatozoa count, and morphology demonstrated a significantly reduced number of spermatozoa and a higher prevalence of morphological abnormalities compared to the WT (Supplementary Figure 1). This is consistent with the hypothesis that moesin deficiency disrupts actin filament organization, contributing to abnormal sperm morphology, particularly in the head and tail structure [39]. Following the alteration of genes in moesin-KO, we conducted subsequent GO term analysis. We identified GO terms related to neuron projection, actin filament-based processes, and brain development, including the regulation of neuron migration (Figure 4). Notably, critical genes, such as Abl2, which is essential for current activity and synaptic localization; Alk, which influences cell growth and brain development; and Reelin, a pivotal gene in the regulation of neuronal migration, dendritic growth, and branching, as well as synaptic formation and plasticity, were significantly decreased. The downregulation of these genes may potentially lead to aberrant neuronal projections and impaired brain development.

GO-term analysis of genes altered in exome sequencing results. A: Top 20 most significantly altered GO-terms. This panel displays the top 20 GO-terms with the highest degree of change, with neuron projection development, actin filament-based process, and regulation of plasma membrane bounded cell projection organization ranking among the highest. B: Network analysis of GO-terms. This panel illustrates the network relationships between the identified GO-terms, showing how these biological processes are interconnected. GO, gene ontology.
Moesin-KO neurons had decreased cell viability and expression of neuronal cell markers
Primary cortical neurons from moesin-KO pups were harvested, cell viability was checked, and the cells were stained with each antibody to examine the involvement of moesin in neuronal cell viability and morphology. In moesin-deleted neurons, cell viability decreased by 40% (WT, 95.8±11.76; KO, 48.01±8.92; p<0.001, n=3) (not shown). Moreover, moesin-KO significantly reduced the levels of SNAP25 (WT, 99.67±1.7; KO, 39.33±4.9; p<0.001, n=3), a presynaptic protein, and PSD95 (WT, 99.68±6.9; KO, 67.33±2.3; p<0.01, n=3), a postsynaptic protein (Figure 5A) in DIV7 neurons.

Moesin-KO neurons decreased cell viability and the expression of neuronal cell markers. Relative expression levels of GFAP, MAP2, Sox2, DCX, SNAP25, PSD95, and moesin are displayed. At 7 days (A) and 14 days (B) after seeding cortical neurons, molecular changes in neuronal cell markers were examined by immunoblotting. All data were analyzed by t-test and are presented as the mean±SEM of three replicates. The bar is 100 μm. *p<0.05; **p<0.01; ***p<0.001. KO, knockout mice; WT, wild-type mice.
At 14 days after seeding the cortical neurons (DIV 14), moesin-KO neurons had significantly decreased Sox2 (WT, 100.01±14.10; KO, 39.01±11.24; p<0.05, n=3), DCX (WT, 100.14±4.73; KO, 57.67±4.48; p<0.01, n=3), PSD95 (WT, 99.67±25.51; KO, 54.12±14.01; p<0.05, n=3) and GFAP (WT, 100.00±6.03; KO, 13.67±4.91; p<0.001, n=3) levels, while there was no difference in MAP2 (WT, 103.00±3.51; KO, 56.00±25.54; n=3) or SNAP25 (WT, 100.00±0.58; KO, 89.67±6.33; n=3) levels (Figure 5B).
Moesin-KO decreased the expression of synaptic proteins and phosphorylation of signaling proteins in the adult brain
The expression or activation of synaptic proteins in the cortex and hippocampus was used to examine the downstream targets of moesin, which was confirmed by immunoblotting. First, the amount of radixin, an ERM family member, was observed, and there was no difference in its expression in cells that could not express moesin (Figure 6A). The expression level of CDK5 was meaningfully suppressed by moesin deletion in both neuronal tissues (cortex, WT, 99.4±5.8; KO, 28.0±2.5, p<0.001; hippocampus, WT, 94.7±3.9; KO, 32.3±6.5, p< 0.01, n=3) (Figure 6A and B). Moreover, the levels of synaptic proteins, such as Syntaxin 1A, Synaptophysin, MUNC18, and Synapsin I, were examined in the hippocampus and cortex of moesin-KO. Syntaxin 1A expression was reduced in the brain (cortex, WT, 92.3±3.9; KO, 60.0±2.9, p<0.01; hippocampus, WT, 96.3±2.7; KO, 57.0±3.5, p<0.001, n=3), but Synaptophysin, and Synapsin I expression was not (Figure 6A-D). However, the phosphorylation levels of Synapsin I (Ser605, cortex, WT, 103.3±3.8; KO, 43.7±3.8, p<0.001; hippocampus, WT, 95.3±3.3; KO, 55.0±3.5, p<0.01; Ser549, cortex, WT, 95.3± 4.2; KO, 46.7±3.8, p<0.01; hippocampus, WT, 93.3±3.4; KO, 68.0±1.2, p<0.01; and Ser9, cortex, WT, 95.0±4.0; KO, 35.7±5.2, p<0.001; hippocampus, WT, 101.3±1.9; KO, 51.7±4.1, p<0.001) and MUNC18 (cortex, WT, 95.3±4.2; KO, 42.7±4.4; p<0.001; hippocampus, WT, 95.3±4.2; KO, 42.7± 4.4, p<0.001) were significantly decreased in the moesin-KO brain (Figure 6C-F). PSD95 expression was decreased in the cortex (WT, 100±0.6; KO, 32.7±4.6, p<0.001) of adult mice but not in the hippocampus. However, phosphorylation was reduced in the hippocampus but not in the cortex (Figure 6E and F). The activation levels of signaling molecules such as ERK and CREB were assessed using phospho-specific antibodies. ERK and CREB expression levels did not change in the brains of the KO. However, phosphorylation of ERK (cortex, WT, 100± 0.9; KO, 19.0±0.6, p<0.001; hippocampus, WT, 96.7±2.8; KO, 45.7±2.6, p<0.001)and CREB (cortex, WT, 99.3±0.3; KO, 23.3±3.4, p<0.001; hippocampus, WT, 100±0.9; KO, 44.0± 3.2, p<0.001) was significantly decreased in the brain of moesin-KO compared to WT (Figure 6E and F).

Moesin-KO decreased the expression of synaptic proteins and phosphorylation of signaling proteins in the adult brain. A and B: The expression or activation level of synaptic vesicle proteins such as Syntaxin 1A, Synaptophysin, CDK5, and radixin were confirmed by immunoblotting in the cortex and hippocampus to examine the downstream target of moesin. Relative expression levels of each protein are presented in bar graphs. C and D: The expression and phosphorylation level of synapsin I were examined in the cortex and hippocampus of moesin-KO and presented in bar graphs. E and F: The activation levels of signaling molecules such as ERK and CREB were checked with phosphorylationspecific antibodies. The expression and phosphorylation levels of MUNC18 and PSD95 were examined in the cortex, and hippocampus of moesin-KO, and their relative expression is presented in graphs. All data were analyzed by t-test and are presented as the mean±SEM of three replicates. **p<0.01; ***p<0.001. KO, knockout mice; WT, wild-type mice; CDK5, Cyclin dependent kinase 5; ERK, extracellular signal-regulated kinase; CREB, cAMP response element-binding protein; SEM, standard error of the mean.
Risperidone administration during adolescence reversed the behavioral and molecular deficits in moesin-KO
Subsequently, we tested whether the antipsychotic drug risperidone could reverse cognitive impairment in moesin-KO. The assessment of learning and memory is easily confounded by repeated trials [40]. Furthermore, as we were testing young animals within the developmental trajectory, re-testing of the same but older animals would result in better performance in the learning process. Therefore, hemizygous male KO were simultaneously injected with either the drug or saline to compare the effect of the drug on learning and memory deficits. Risperidone was administered according to a predetermined schedule for 2 weeks and validated by behavioral experiments. The results showed that risperidone effectively reversed the memory impairment of 2-months old moesin-KO in the passive avoidance test (increase in the latency time for the mice to enter the dark chamber; WT, 77.4±31.0, n=11; KO, -2.4±8.1, n=10; KO+Ris, 112.3±30.1, n=9, WT vs. KO, p< 0.05, KO vs. KO+Ris, p<0.01) (Figure 7A), and the spontaneous alternation percentage (WT, 67.8±2.4, n=13; KO, 51.6± 3.6, n=10; KO+Ris, 63.4±1.9, n=10, WT vs. KO, p<0.001, KO vs. WT+Ris, p<0.001) (Figure 7B, right panel) in the Y maze test.

Risperidone reversed behavioral and molecular deficits in moesin-KO. Risperidone was administered for 2 weeks and validated by behavioral experiments at 2 months of age. A: Risperidone reversed memory impairment in moesin-KO in the passive avoidance test and (B) total entry number (left panel) and spontaneous alternation percentage (right panel) in the Y maze. C: Risperidone induced the expression of PSD95, the phosphorylation of synapsin at Ser605 and Ser549, and ERK phosphorylation in the cortex of moesin-KO, and the data are displayed in the graph. All data were analyzed by one-way ANOVA and are presented as the mean±SEM. *p<0.05; **p<0.01; ***p<0.001. KO, knockout mice; WT, wild-type mice; ERK, extracellular signal-regulated kinase; SEM, standard error of the mean; Ris, risperidone.
Furthermore, risperidone effectively restored the reduced expression of PSD95 (WT, 106.3±3.2; KO, 43.7±1.9; KO+Ris, 85.3±4.7; WT vs. KO, p<0.001, KO vs. KO+Ris, p<0.01, n=3) in the cortex of moesin-KO. The phosphorylation of PSD95 was decreased in the brain of KO+Ris mice compared to KO (WT, 97.0±3.0; KO, 132.7±16.8; KO+Ris, 79.4±7.8, KO vs. KO+Ris, p<0.05, n=3) (Figure 7C). The phosphorylation of Synapsin at Ser605 (WT, 110.0±5.8; KO, 46.3±6.0; KO+Ris, 83.3±6.7, WT vs. KO, p<0.01, KO vs. KO+Ris, p<0.05, n=3) and Ser549 (WT, 126.7±3.1; KO, 30.3±3.0; KO+Ris, 70.7±7.8, WT vs. KO, p<0.001, KO vs. KO+Ris, p<0.001, n=3) in the cortex was also restored in moesin-KO, and ERK phosphorylation was decreased in untreated moesin-KO (WT, 109.7±5.8; KO, 24.0±6.0; KO+Ris, 43.7±6.0; WT vs. KO, p< 0.001, n=3) (Figure 7C).
DISCUSSION
In this study, we found that moesin deficiency leads to synaptic protein dysfunction and memory decline, possibly due to altered regulation in synaptic proteins such as Synapsin phosphorylation and PSD95 expression. Risperidone treatment showed potential in reversing these deficits, improving both synaptic protein function and memory performance. First, we investigated whether moesin deficiency leads to behavioral changes associated with neurodevelopmental abnormalities. We observed that moesin-KO pups exhibited decrease in the surface righting ability on PND7 and increase in time spent in the closed arms, showing increased anxietylike behavior during adolescence. Moreover, we also observed that moesin-KO exhibit phenotypes resembling delayed brain development and memory impairment. Therefore, we performed WES in moesin-KO brain to determine the disruption of neuronal signal pathways caused by moesin deficiency. An in-depth analysis of WES data was conducted to identify chromosomal alterations across the entire genome. The data revealed a distinct distribution pattern of alternative bases across chromosomes. The alterations observed in key genes regulating neuron projection, actin filament-based processes, and brain development, including the regulation of neuron migration, suggesting the genetic ramifications of moesin on neurodevelopmental pathways. In order to understand how moesin deficiency leads to neurodevelopmental delay and memory decline, we investigated altered regulation in synaptic proteins such as Synapsin phosphorylation and PSD95 expression in moesin-KO. We found that the deletion of moesin led to a decrease in the expression of Syntaxin 1A and PSD95, as well as reduced the phosphorylation of Synapsin I at Ser605, Ser549, and Ser9. Also, we found that risperidone, a drug commonly used in these developmental conditions, reversed the memory impairment in the passive avoidance test and the spontaneous alternation percentage in the Y maze test. Risperidone also restored the reduced expression of PSD95 and the phosphorylation of Synapsin at Ser605 and Ser549 in the cortex of moesin-KO, providing further evidence that moesin deficiency leads to behavioral changes associated with neurodevelopmental abnormalities through altered regulation in synaptic proteins and function.
We identified for the first time that moesin-KO exhibit phenotypes resembling delayed brain development and memory impairment. MSNP1AS matched 94% with the mRNA of the X chromosome moesin transcripts. MSNP1AS near rs4307059 is the functional component showing replicated genetic associations, and MSNP1AS is upregulated in patients with the autism-linked rs4307059 T allele [41]. Moreover, MSNP1AS overexpression changes the length and number of neuronal processes and induces neuronal cell death and abnormal migration [42,43]. In down syndrome, down regulation of moesin has been associated with impaired neurite outgrowth and nerve migration in the fetal brain [44]. In addition, moesin has been implicated in Alzheimer’s disease by promoting cell cycle reactivation and neurodegeneration through it interaction with pathogenic tau, which disrupts cytoskeletal stability and accelerates neuronal damage [45]. The inhibition of moesin/radixin expression suppresses growth cone and axon formation in cultured primary neurons [46]. This study showed that moesin dele-tion reduces neuronal cell viability and causes developmental delay, memory impairment, and anxiety in mice, indicating that moesin is an important molecule in neurodevelopment.
Our study, centered on moesin-KO, revealed substantial genetic alterations on chromosome 1, a known hotspot for health issues due to copy number variations. The contribution of chromosome 1 to diverse diseases is well-documented, particularly in the context of neurodevelopmental and Mendelian disorders [47]. The key findings included changes in Abl2, which is essential for neuronal development and central nerve system (CNS) functioning. These changes may be linked to phosphorylation alterations and molecular weight variations in the ABL protein, which plays a role in neural cell development signaling [48,49]. Our observations of Abl2 complement the existing knowledge about its involvement in neural growth and development, adding new dimensions to our understanding of its functions in the CNS.
Furthermore, the decreased expression of ALK, as evidenced by qPCR analysis, underscores its critical role in neuronal differentiation and survival in the CNS. This finding aligns with existing research that associates ALK with neuronal circuit assembly and behavioral changes in animal models [50-54]. Additionally, our study draws attention to reelin, which has shown a shift in expression during the post-developmental stages, predominantly in GABAergic interneurons. This finding aligns with existing literature on the influence of reelin on synaptic arrangement and dendritic growth [55]. The alterations observed in key genes, such as Abl2, ALK, and Reelin, within our moesin-KO, provides insights into the genetic ramifications of moesin and highlights its significant impact on neurodevelopmental pathways.
Spermatozoa count, and morphology demonstrated a significantly reduced number of spermatozoa and a higher prevalence of morphological abnormalities in KO compared to the WT confirmed that KO possess actin filament abnormalities caused by moesin deficiency. In spermatogenesis, actin dynamic is crucial for proper sperm morphology, including the head and tail structures in the mammalian [56]. Recent studies indicate that the MAPK cascade plays a key regulatory role in spermatogenesis and various male reproductive processes, including sperm maturation, activation, capacitation, and the acrosome reaction [57]. Disruption in actin cytoskeleton dynamics due to moesin deficiency could impact MAPK signaling, further affecting these critical reproductive processes. Dysfunction of moesin may disrupt the actin cytoskeleton and impair the spermatogenesis via MAPK signaling.
Moreover, the downregulation of DEAD-box helicase 3 Ylinked (Ddx3y), a gene located on the Y chromosome, plays an essential role in the proliferation and differentiation of spermatogonia [58], observed in exome sequencing could suggest that moesin deficiency alters upstream regulatory mechanisms affecting Ddx3y expression. Moesin is known to play critical roles in cell signaling and cytoskeletal organization, and its absence may exacerbate the effects of the Ddx3y mutation. However, the exact molecular mechanisms by which moesin and Ddx3y interact remain to be elucidated. Future studies should investigate the relationship between moesin and Ddx3y expression.
In this study, the deletion of moesin led to a decrease in the expression of Syntaxin 1A and PSD95, as well as reduced the phosphorylation of Synapsin I at Ser605, Ser549, and Ser9. Mice deficient in Synapsin I exhibit epileptic seizures from 2 to 3 months of age [59], mild cognitive deficits, including impairments in emotional and spatial memory [60,61], and key symptoms of autism that affect social communication, social behavior, and repetitive behavior [62]. Among the phosphorylation sites of Synapsin I, PKA, Ca2+/calmodulin-dependent protein kinase, ERK, and CDK 1/5 reduce the binding of Synapsin I to synaptic vesicles and/or F-actin, thereby increasing synaptic vesicle availability [63]. Conversely, Src phosphorylating Synapsin I favors renewed synaptic vesicle reassembly by enhancing the binding of Synapsin I to synaptic vesicles and Factin [64]. In this study, we observed a decrease in phosphorylation of Synapsin I at Ser605, Ser549, and Ser9, as well as reduced activities of ERKs and PKA/CREB, which are upstream kinases of Synapsin I, and the expression level of CDK5 was reduced. These results indicate that Synapsin I dysfunction induced by moesin knockout may contribute to neurodevelopmental delays, memory impairment, and anxiety.
In this study, decreased expression of SNAP25 and PSD95 was observed in DIV7 neurons of KO. Moreover, downregulation of synaptic proteins, Syntaxin 1A and CDK5, and altered phosphorylation levels of Synapsin I, MUNC18, ERK, and CREB was observed in KO cortex and hippocampus. PSD95 plays an important role in neuronal development, such as in regulating glutamate transmission, synaptic plasticity, and the morphogenesis of dendritic spines [65-67] and PSD95 dysfunction is related with neuropsychiatric disorders, including schizophrenia, autism, and intellectual disorders [68-70]. PSD95 KO show prefrontal cortex-associated behavioral phenotypes that reflect sociability, cognition, and working memory [71], suggesting that moesin deletion-induced PSD95 expression in the cortex is involved in memory impairment. The release of neurotransmitters by synaptic vesicles involves MUNC18, the soluble N-ethylmaleimide-sensitive factor adhesion protein receptor (SNARE), synaptotagmin-1, N-ethylmaleimide-sensitive factor, SNAP, and Munc13 [72]. SNARE, mainly composed of SNAP25 and Syntaxin-1, regulates membrane fusion and binds to the target membrane of vesicular proteins (T-SNARE) for synaptic vesicle fusion and secretion [73]. MUNC18 plays an important role in regulating brain-derived neurotrophic factors, participating in synapse development, and affecting cognitive function [74,75]. This study also showed MUNC18 dephosphorylation in the cortex and hippocampus of moesin-KO, suggesting that MUNC18 dephosphorylation in the brains of moesin-KO may be involved in delayed neurodevelopmental behavior.
CDK5 is predominantly expressed in the brain and plays a pivotal role in synaptogenesis, synaptic transmission, and plasticity [76,77]. CDK5 plays a major role in synaptic transmission by phosphorylating several substrates at both the pre- and postsynaptic levels, acting cooperatively with the phosphatase calcineurin and presynaptic voltage-gated Ca2+ channels [78,79]. In addition to Synapsin I, MUNC18 is a presynaptic substrate of CDK5 [80,81]. In this study, moesin-KO showed decreased CDK5 expression, suggesting that reduced CDK5 activity decreases Synapsin I expression and that MUNC18 phosphorylation results in impaired synaptic transmission. Previously, we reported that the chloride channel 4/moesin pathway is involved in NGF treatment-induced PC12 cell and cortical neuron differentiation [82,83] and that chloride channel 4 is upstream of CDK5 [83]. These results suggest that moesin deletion leads to aberrations in synaptic proteins by decreasing CDK5 expression.
In our study, moesin-deleted primary cortical neurons showed lower expression of the neural stem cell marker SRY-box transcription factor (SOX) and the immature neuronal marker doublecortin X (DCX) than WT neurons. Patients with heterozygous loss-of-function mutations in SOX2 show typical defects of the central nervous system, including hippocampal defects (including the dentate gyrus, DG), intellectual disability, and epilepsy [84-86]. The effect of DCX is associated with its interaction with the actin cytoskeleton, affecting not only the development of axons but also the formation of dendritic trees [87]. These results suggest that moesin-KO-induced decrease in SOX2 and DCX expression may induce neurodevelopmental defects.
Furthermore, we observed that risperidone, a drug approved by the US Food and Drug Administration for the treatment of children with autism from the age of 5 years [28,29], recovered impaired memory function and reversed the reduction in the expression of PSD95 and Synapsin phosphorylation at Ser549, Ser609, and ERK in the cortex of moesin-KO. These data are consistent with previous reports showing that antipsychotics promote neurite outgrowth [88] and neural plasticity [89] through molecules involved in PSD and ERK signaling. Risperidone’s modulation of synaptic plasticity may be mediated by its effects on PSD proteins, which govern the structural integrity of synapses. Risperidone regulates PSD proteins such as Arc, Zif-268, and ΔFosB, which are immediate early genes (IEGs) crucial for synaptic plasticity [24]. These IEGs are rapidly transcribed in response to synaptic activity and play a key role in regulating the structural and functional changes at synapses that underlie cognitive and behavioral functions [90]. Arc facilitates AMPA receptor trafficking, affecting synaptic strength [91]. while Zif-268, ΔFosB supports long-term potentiation (LTP), which is essential for memory consolidation [92]. Risperidone inhibits phencyclidine-induced Arc overexpression in the prefrontal cortex and nucleus accumbens, regions commonly associated with psychotomimetic drug-induced synaptic dysfunctions, such as cognitive deficits and behavioral disturbances [93]. Homer1a, another PSD molecule regulated by risperidone, modulates calcium homeostasis and dendritic morphology by disrupting Homer1b/c multimers and altering glutamate signaling [94]. In this study, risperidone restored the reduced expression of PSD95 and the phosphorylation of Synapsin at Ser605 and Ser549 in the cortex of moesin-KO, providing further evidence that moesin deficiency leads to behavioral changes associated with neurodevelopmental abnormalities through altered regulation in synaptic proteins and function.
This study had some limitations. The first is the lack of direct evidence of aberrations in the two primary forms of synaptic plasticity, LTP and long-term depression. However, we have shown that several target proteins of moesin that regulate synaptic functions, such as CDK5, Synapsin, and PSD95, are dysregulated in moesin-KO. Second, because only hemizygous male KO were used for memory tests, the conclusions drawn from the risperidone treatment can only be generalized to males and not to all sexes. Finally, we did not perform in vivo staining for developmental markers or direct measurements of neurotransmission in the postnatal brains of moesin-KO. However, the results showed that moesin-KO induced decreased SOX2 and DCX in primary cortical cultures derived from embryos (15 days) when most neurons in the cortical and subcortical areas were being generated and had migrated [95].
Supplementary Materials
The Supplement is available with this article at https://doi.org/10.30773/pi.2024.0186.
Sperm morphology and motility analysis. A: Comparison of testis weight in WT and KO. B: Sperm count (number×106/mL) in WT and KO. C: Representative image of sperm morphology in WT and KO. Arrows indicate the abnormal-shaped sperm. All data were analyzed by t-test and are presented as the mean±SEM of three replicates. *p<0.05; **p<0.01. KO, knockout mice; WT, wild-type mice.
Notes
Availability of Data and Material
All data and materials will be sent upon request.
Conflicts of Interest
The authors have no potential conflicts of interest to disclose.
Author Contributions
Conceptualization: all authors. Data curation: Hua Cai, Seong Mi Lee, Yura Choi, Bomlee Lee, Soo Jung Im, Hyung Jun Choi, Jin Hee Kim. Fomal analysis: Hua Cai, Seong Mi Lee, Yura Choi, Bomlee Lee, Soo Jung Im, Hyung Jun Choi, Jin Hee Kim. Funding acquisition: Yeni Kim, Boo Ahn Shin, Songhee Jeon. Investigation: Hua Cai, Seong Mi Lee, Yeni Kim, Boo Ahn Shin, Songhee Jeon. Methodology: Yeni Kim, Boo Ahn Shin. Project administration: Yeni Kim, Boo Ahn Shin, Songhee Jeon. Resources: Yeni Kim. Supervision: Yeni Kim, Boo Ahn Shin, Songhee Jeon. Validation: Yura Choi, Yeni Kim. Visualization: Yura Choi, Yeni Kim. Writing—original draft: Songhee Jeon. Writing—review & editing: Yura Choi, Yeni Kim.
Funding Statement
The research grants NRF-2020R1A2C1006910, NRF-2017R1D1A3B03031334, and NRF-2022R1A2C3010655 from the National Research Foundation of Korea, a grant (CRI16019-1) from the Chonnam National University Hospital Biomedical Research Institute, intramural research grants (2018-04, 2019-11, R2020-C, R2021-C) from the National Center for Mental Health, and the Dongguk University Research Program of 2022 supported this research.
Acknowledgements
None