Reduced cortico-accumbal excitatory input due to Nav1.2 haploinsufficiency impairs sociability independently of dopamine

5 Scn2a, Nav1.2, autism spectrum disorders, schizophrenia, nucleus accumbens, 6 parvalbumin-expressing fast-spiking inhibitory neurons 7 8 ORCiD 9 Toshimitsu Suzuki https://orcid.org/0000-0003-0585-5692 10 Hiroaki Mizukami https://orcid.org/0000-0001-8954-874X 11 Kenta Kobayashi https://orcid.org/0000-0002-7389-3693 12 Tetsushi Yamagata https://orcid.org/0000-0002-5734-7932 13 Hiroshi Nomura https://orcid.org/0000-0002-6172-4788 14 Hideki Hida https://orcid.org/0000-0002-9679-4508 15 Kazuhiro Yamakawa https://orcid.org/0000-0002-1478-4390 16 17 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2026. ; https://doi.org/10.64898/2026.04.15.718826doi: bioRxiv preprint 3

1 Mutations in SCN2A, which encodes the voltage-gated sodium channel Nav1.2, are 2 associated with a wide spectrum of neurodevelopmental and neuropsychiatric disorders, 3 including epilepsy, autism spectrum disorder (ASD), and schizophrenia. Although 4 dysfunction of SCN2A-dependent neural circuits has been implicated in these disorders, 5 the circuit mechanisms underlying social behavioral abnormalities remain poorly 6 understood. Here, we investigated the neural circuit basis of social behavioral deficits 7 associated with Scn2a dysfunction, focusing on the nucleus accumbens (NAc), a key 8 hub in cortico-limbic circuits that regulates emotional and motivational behaviors. 9 Using conditional genetic and chemogenetic approaches in mice, we examined the roles 10 of dorsal telencephalic excitatory neurons, including those in the cerebral cortex, 11 hippocampus, and amygdala, as well as parvalbumin-positive fast-spiking interneurons 12 (PV⁺ FSIs) in the NAc. Mice with Scn2a haploinsufficiency in dorsal telencephalic 13 excitatory neurons (Scn2afl/+/Emx1-Cre) exhibited reduced sociability in the three-14 chamber social interaction test. Similarly, chemogenetic inhibition of NAc PV⁺ FSIs 15 decreased sociability without affecting locomotor activity or anxiety-like behavior. 16 Scn2afl/+/Emx1-Cre mice also showed a trend toward reduced prepulse inhibition of the 17 acoustic startle response. Notably, dopamine release into the NAc in the Scn2afl/+/Emx1-18 Cre and systemic Scn2a heterozygous knockout (Scn2a+/-) mice was largely comparable 19 to that in control mice. Together, these findings indicate that reduced activity of dorsal 20 telencephalic excitatory neurons or NAc PV⁺ FSIs is sufficient to impair sociability 21 independently of mesolimbic dopamine hypofunction. Our results highlight a potential 22 role of cortico-accumbal circuits in social behavioral deficits associated with SCN2A 23 dysfunction. 24 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2026. ; https://doi.org/10.64898/2026.04.15.718826doi: bioRxiv preprint 4

1 The voltage-gated sodium channel Nav1.2, encoded by SCN2A, plays a critical 2 role in the initiation and propagation of action potentials in excitatory neurons, thereby 3 shaping cortical network excitability and information processing [Catterall, 1992; 4 Catterall et al., 2005]. Nav1.2 is abundantly expressed in the axons of cortical and 5 hippocampal excitatory neurons [Liao et al., 2010; Ogiwara et al., 2018; Yamagata et 6 al., 2023], as well as in dopaminergic neurons in the substantia nigra pars compacta and 7 ventral tegmental area (VTA) [Yang et al., 2019]. In addition, Nav1.2 is present in 8 specific populations of inhibitory neurons, including medium spiny neurons (MSNs) in 9 the striatum [Miyazaki et al., 2014] and GABAergic interneurons in the neocortex [Li et 10 al., 2014; Yamagata et al., 2017]. 11 Pathogenic mutations in SCN2A have been implicated in a broad spectrum of 12 neurodevelopmental and neuropsychiatric disorders, including epilepsy, autism 13 spectrum disorder (ASD), intellectual disability (ID), and schizophrenia [Sugawara et 14 al, 2001; Kamiya et al, 2004; Ogiwara et al, 2009; Buxbaum et al, 2012; Rauch et al, 15 2012; de Ligt et al, 2012; Tavassoli et al, 2014; Fromer et al, 2014; Hoischen et al, 16 2014; Li et al, 2016; Johnson et al, 2016; Carroll et al, 2016; Balakrishna et al, 2020]. 17 Consistent with these clinical findings, rodent models with Scn2a haploinsufficiency 18 exhibit epileptic phenotypes, cognitive impairments, and behavioral abnormalities 19 reminiscent of SCN2A-associated disorders in humans [Ogiwara et al., 2018; Middleton 20 et al., 2018; Tatsukawa et al., 2019; Suzuki et al., 2024]. Our previous studies using 21 genetic models of SCN2A- and STXBP1-related epilepsies demonstrated that reduced 22 excitatory transmission from cortical pyramidal neurons onto striatal parvalbumin-23 expressing (PV+) fast-spiking interneurons (FSIs) is sufficient to trigger epilepsy 24 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2026. ; https://doi.org/10.64898/2026.04.15.718826doi: bioRxiv preprint 5 [Ogiwara et al., 2018; Miyamoto et al., 2019]. Consistently, pharmacological inhibition 1 of cortico-striatal excitatory inputs onto FSIs using Ca²⁺-permeable AMPA receptor 2 antagonist induces generalized seizures in both rodents and non-human primates [Gittis 3 et al., 2011; Aupy et al., 2024]. More recently, we further demonstrated that selective 4 suppression of PV+ FSIs in the anteromedial shell of the NAc is sufficient to elicit 5 convulsive seizures, underscoring the critical role of striatal microcircuits in seizure 6 generation [Suzuki et al., 2026]. These findings highlight the importance of cortico-7 striatal circuits and PV⁺ FSIs in regulating network excitability. Accumulating evidence 8 also suggests that SCN2A dysfunction contributes to behavioral abnormalities relevant 9 to psychiatric disorders. For example, mice with Scn2a deficiency in brain regions 10 implicated in schizophrenia and ASD, such as the medial prefrontal cortex (mPFC) and 11 the VTA, display alterations in prepulse inhibition (PPI) of the acoustic startle response 12 [Suzuki et al., 2024]. PPI is widely recognized as an endophenotype of schizophrenia 13 [Braff et al., 2001; Geyer et al., 2006; Powell et al., 2009; Takahashi et al., 2011; Mena 14 et al., 2016]. In addition, mice lacking Scn2a in the mPFC exhibit increased sociability, 15 decreased locomotor activity, and enhanced anxiety-like behavior, whereas mice lacking 16 Scn2a in the VTA show minimal behavioral abnormalities apart from altered vertical 17 activity. These observations suggest that Scn2a dysfunction in distinct brain regions 18 may differentially contribute to schizophrenia- and ASD-related phenotypes. In contrast, 19 other group reported that deletion of Scn2a in VTA dopaminergic neurons reduces 20 neuronal firing and dopamine release, leading to hyperactivity, impaired sociability, and 21 reduced anxiety-like behavior [Li et al., 2025]. Similar abnormalities are observed in 22 systemic Scn2a heterozygous knockout mice (Scn2a+/-) and are alleviated by acute 23 levodopa administration, indicating dopamine system hypofunction [Li et al., 2025]. 24 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2026. ; https://doi.org/10.64898/2026.04.15.718826doi: bioRxiv preprint 6 Based on these findings, dysfunction of SCN2A-dependent neural circuits, 1 including cortico-striatal and mesolimbic dopamine pathways, has been proposed to 2 contribute to the pathophysiology of ASD and schizophrenia. However, the circuit 3 mechanisms through which Scn2a dysfunction leads to social behavioral abnormalities 4 remain incompletely understood. In the present study, we sought to elucidate the neural 5 circuit mechanisms underlying ASD-related social behavioral deficits associated with 6 SCN2A dysfunction. In particular, we focused on the NAc, a key node in cortico-limbic 7 circuits that integrates excitatory inputs from the cerebral cortex, hippocampus, and 8 amygdala and regulates emotional and motivational behaviors. Using conditional 9 genetic and chemogenetic approaches, we show that Scn2a haploinsufficiency in dorsal 10 telencephalic excitatory neurons and selective inhibition of NAc PV⁺ FSIs reduce 11 sociability. Furthermore, these behavioral abnormalities occur without detectable 12 reductions in dopamine release in the NAc, suggesting that impaired activity of 13 excitatory neurons or FSIs in cortico-striatal circuits may contribute to social behavioral 14 deficits independently of mesolimbic dopamine hypofunction. 15 16 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2026. ; https://doi.org/10.64898/2026.04.15.718826doi: bioRxiv preprint 7

1 Dorsal telencephalic excitatory neuron dysfunction and NAc FSI inhibition impair 2 sociability 3 To selectively delete the Scn2a gene in excitatory neurons of the dorsal telencephalon, 4 including the cerebral cortex, hippocampus, olfactory bulb, and amygdala, we generated 5 conditional knockout mice (Scn2afl/+/Emx1-Cre) using Emx1-Cre driver mice [Iwasato 6 et al., 2000; Iwasato et al., 2004; Ogiwara et al., 2018], which mediate excitatory 7 neuron-specific recombination in these regions (Fig. 1A, B). We then assessed 8 sociability using the three-chamber social interaction test. Control mice showed a 9 significant preference for the cage containing a stranger mouse over the empty cage. In 10 contrast, mutant mice spent less time near the stranger mouse, indicating a reduced 11 preference for the stranger mouse (Fig. 1C). These results indicate that Scn2a 12 haploinsufficiency in dorsal telencephalic excitatory neurons impairs sociability and 13 reduces interest in social stimuli. 14 Next, we examined whether reduced activity of PV⁺ FSIs in the NAc produces a 15 similar behavioral phenotype. The NAc receives projections from subsets of excitatory 16 neurons in the cerebral cortex, hippocampus, and amygdala [Scofield et al., 2016; 17 Marinescu et al., 2024; Xu et al, 2024] (Fig. 1B) and plays a key role in emotional and 18 motivational processing [Castro et al., 2019]. To suppress NAc PV⁺ FSI activity, we 19 injected an adeno-associated virus (AA V) encoding the inhibitory DREADD receptor 20 hM4D(Gi) into the NAc of PV-Cre mice (PV-Cre/NAc-hM4Di). In this line, Cre 21 recombinase is selectively expressed in PV⁺ FSIs, enabling targeted expression of 22 hM4D(Gi). Histological analyses using the same mice confirmed that the AA V 23 injections were accurately targeted to the intended region, as we previously reported 24 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2026. ; https://doi.org/10.64898/2026.04.15.718826doi: bioRxiv preprint 8 [Suzuki et al., 2026]. We then performed the three-chamber test. Before clozapine-N-1 oxide (CNO) administration, PV-Cre/NAc-hM4Di mice displayed normal sociability, 2 similar to control mice (Fig. 1D). After CNO administration, however, sociability was 3 reduced, and mice spent less time near the cage containing the stranger mouse (Fig. 4 1E). Together, these results suggest that both reduced excitatory neuronal function in the 5 dorsal telencephalon and decreased activity of NAc PV⁺ FSIs impair sociability. 6 7 Selective Scn2a heterozygous deficiency in dorsal telencephalic excitatory neurons 8 shows a trend toward reduced PPI 9 We next assessed PPI in Scn2afl/+/Emx1-Cre mice. Startle responses to acoustic stimuli 10 at two intensities (110- and 120-dB) were comparable between mutant and control 11 (Scn2afl/+) mice (Fig. 2A). In contrast, the percentage PPI of the startle response to a 12 120-dB pulse with 74- and 78-dB prepulses tended to be lower in mutant mice, although 13 the differences did not reach statistical significance (Fig. 2B). 14 15 Inactivation of NAc FSIs has no effect on locomotor activity or anxiety‐like 16 behavior 17 We next assessed spontaneous exploratory behavior using the open field test. PV-18 Cre/NAc-hM4Di mice showed no significant differences in total distance traveled (Fig. 19 3A) or time spent in the center zone (Fig. 3B), either before or after CNO 20 administration. These results indicate that chemogenetic inactivation of NAc FSIs does 21 not affect locomotor activity or anxiety-like behavior under these conditions. 22 23 Dopamine release in the NAc is comparable between Scn2a mutant and control 24 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2026. ; https://doi.org/10.64898/2026.04.15.718826doi: bioRxiv preprint 9 mice 1 Finally, we measured dopamine (DA) release in the NAc under novel environmental 2 conditions and following methamphetamine administration. Although our experimental 3 procedures differed from those of a previous study [Li et al., 2025], DA release in 4 Scn2afl/+/Emx1-Cre mice (Fig. 4A, B) and Scn2a+/- mice (Fig. 4C, D) was comparable 5 to that in control mice. These findings suggest that the sociability deficits observed in 6 Scn2afl/+/Emx1-Cre mice and PV-Cre/NAc-hM4Di mice cannot be explained solely by 7 reduced DA release in the NAc. Instead, they raise the possibility that altered activity in 8 other circuits, such as dorsal telencephalic excitatory neurons or NAc FSIs, also 9 contributes to social behavioral deficits. 10 11 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2026. ; https://doi.org/10.64898/2026.04.15.718826doi: bioRxiv preprint 10

1 The present study provides new insight into the circuit mechanisms underlying social 2 behavioral deficits associated with Scn2a dysfunction. Using conditional genetic and 3 chemogenetic approaches, we show that reduced activity of dorsal telencephalic 4 excitatory neurons or NAc PV⁺ FSIs is sufficient to impair sociability. Notably, these 5 behavioral alterations occur in the absence of detectable changes in DA release in the 6 NAc, indicating that dysfunction of cortico-accumbal microcircuits, rather than 7 mesolimbic DA hypofunction alone, can drive social deficits. 8 A key finding is that Scn2a haploinsufficiency in dorsal telencephalic excitatory 9 neurons reduces sociability, in contrast to previous reports of increased social 10 interaction in mice with systemic Scn2a haploinsufficiency or mPFC-specific deletion 11 [Tatsukawa et al., 2019; Suzuki et al., 2024; Li et al., 2025]. This discrepancy 12 underscores the circuit-dependent roles of Scn2a in social behavior. One critical 13 difference lies in the extent and composition of affected circuits. Whereas mPFC-14 specific deletion primarily alters prefrontal outputs, Emx1-Cre–mediated deletion 15 targets a broader population of excitatory neurons across the cortex, hippocampus, and 16 amygdala. These regions provide convergent glutamatergic inputs to the NAc, a central 17 hub for integrating social and motivational signals. Thus, a widespread reduction in 18 excitatory drive may decrease the salience or reward value of social stimuli, leading to 19 reduced sociability. Consistent with this interpretation, direct chemogenetic inhibition of 20 NAc PV⁺ FSIs recapitulated the social deficit phenotype. PV⁺ FSIs provide strong 21 feedforward inhibition onto MSNs and regulating the timing and gain of striatal output. 22 Reduced FSI activity is therefore likely to disrupt the excitation-inhibition balance 23 within the NAc, impairing the encoding of socially relevant information. Together, these 24 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2026. ; https://doi.org/10.64898/2026.04.15.718826doi: bioRxiv preprint 11 findings support model in which Scn2a dysfunction produces bidirectional effects on 1 sociability depending on the affected circuit, with cortico-accumbal hypofunction 2 leading to social deficits, in contrast to hyper-sociability observed following more 3 restricted perturbations. 4 In addition to social behavior, Scn2a haploinsufficiency in dorsal telencephalic 5 excitatory neurons showed a trend toward reduced PPI. This is consistent with our 6 previous finding that mPFC-specific Scn2a deletion impairs PPI [Suzuki et al., 2024], 7 supporting a central role for cortical excitatory circuits in sensorimotor gating. Given 8 that the Emx1-Cre driver includes mPFC neurons, the observed PPI deficit likely 9 reflects impaired prefrontal function. The convergence of PPI deficits across dorsal 10 telencephalic and mPFC-specific manipulations further suggests that distributed cortical 11 network dysfunction, rather than a single locus, underlies impaired sensorimotor gating. 12 In contrast, the enhancement of PPI observed following Scn2a deletion in VTA 13 dopaminergic neurons [Suzuki et al., 2024] highlights the distinct and potentially 14 opposing contributions of cortical and dopaminergic systems. 15 Chemogenetic inhibition of NAc PV⁺ FSIs did not significantly alter locomotor 16 activity or anxiety-like behavior in the open field test. This contrasts with previous 17 reports showing hyperactivity in systemic Scn2a haploinsufficiency [Tatsukawa et al., 18 2019] and reduced locomotion following mPFC-specific deletion [Suzuki et al., 2024]. 19 These differences are consistent with the notion that the NAc is more closely associated 20 with motivational and emotional processing than with general motor control. 21 Accordingly, PV⁺ FSIs in the NAc may preferentially regulate socially and 22 motivationally relevant behaviors. Alternatively, compensatory mechanisms within 23 striatal circuits may preserve locomotor function, or subtle behavioral changes may not 24 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2026. ; https://doi.org/10.64898/2026.04.15.718826doi: bioRxiv preprint 12 be captured by the open field paradigm. 1 Based on these findings, we propose a circuit model (Fig. 5) in which reduced 2 excitatory drive from the dorsal telencephalon leads to decreased activation of PV⁺ 3 FSIs, resulting in dysregulated striatal output. Because FSIs receive dense glutamatergic 4 inputs from cortical and limbic regions, reduced excitatory neuronal activity is likely to 5 indirectly suppress FSI function. This, in turn, may alter activity patterns in MSNs and 6 disrupt downstream signaling within ventral striatal circuits. Such disruption could 7 impair neural processes underlying reward prediction, social salience, and action 8 selection. Notably, our previous work demonstrated that impaired cortico-striatal 9 excitation onto the striatal FSIs contributes to epileptogenesis [Miyamoto et al., 2019], 10 suggesting that PV⁺ FSIs represent a critical convergence point linking network 11 excitability and behavioral regulation. Thus, NAc PV⁺ FSIs may function as an 12 integrative hub through which cortical deficits are translated into behavioral 13 abnormalities, including both social deficits and impaired PPI. 14 Recent work by Li et al. (2025) demonstrated that Scn2a deletion in VTA 15 dopaminergic neurons reduces DA release in the NAc and leads to social deficits, 16 supporting a DA-dependent mechanism. In contrast, we found that DA release in the 17 NAc is largely preserved in both dorsal telencephalon-specific and systemic Scn2a 18 mutant mice despite impaired sociability. These findings indicate that social deficits in 19 Scn2a models cannot be explained solely by mesolimbic DA hypofunction. Instead, 20 they support the existence of at least two parallel pathways: (1) a DA-dependent 21 pathway involving VTA dysfunction, and (2) a DA-independent pathway involving 22 cortico-striatal circuit dysfunction. These mechanisms are not mutually exclusive and 23 may interact. For example, cortical inputs could modulate DA neuron activity indirectly, 24 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2026. ; https://doi.org/10.64898/2026.04.15.718826doi: bioRxiv preprint 13 while DA signaling may influence striatal microcircuit dynamics. In addition, the VTA 1 provides glutamatergic projections to the NAc, which may also contribute to motivated 2 behavior [Qi et al., 2016]. Although not examined here, Scn2a dysfunction in these 3 projections could represent an additional mechanism and warrants further 4 investigations. 5 In summary, our findings demonstrate that reduced activity of dorsal 6 telencephalic excitatory neurons or NAc PV⁺ FSIs is sufficient to impair sociability, 7 independently of changes in NAc dopamine release. These results highlight the 8 importance of cortico-accumbal microcircuits in social behavior and support a 9 framework in which SCN2A-related neuropsychiatric phenotypes arise from circuit-10 specific dysfunction across distributed brain networks. Targeting striatal microcircuits 11 or restoring excitation-inhibition balance may therefore represent promising therapeutic 12 strategies for SCN2A-assiciated disorders. 13 14 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2026. ; https://doi.org/10.64898/2026.04.15.718826doi: bioRxiv preprint 14

1 Animals 2 All mice were maintained under a 12 h light/dark cycle with ad libitum access to food 3 and water. Scn2a floxed mice [Ogiwara et al., 2018], in which exon 2 is flanked by loxP 4 sites, and Empty spiracles homolog 1 (Emx1)-Cre knock-in mice [Iwasato et al., 2000; 5 Iwasato et al., 2004] were maintained on a C57BL/6J background. To generate Scn2a 6 conditional knockout mice, homozygous Scn2a floxed (Scn2afl/fl) mice were crossed 7 with Emx1-Cre knock-in mice. Parvalbumin (PV)-Cre transgenic mice [Tanahira et al., 8 2009] were maintained on a C57BL/6J background. Scn2a heterozygous knock-out 9 (Scn2a+/-) mice were described previously [Planells-Cases et al., 2000] and were 10 maintained on a C57BL/6J background. 11 12 AA Vs 13 The plasmid construct for double floxed Gi-coupled hM4D DREADD fused with 14 mCherry under the control of EF1a promoter (pAA V-EF1a-DIO-hM4D(Gi)-mCherry) 15 was a gift from Dr. Bryan Roth (Addgene plasmid 50461). Packaging of the plasmid 16 vector into AA Vs, as well as purification and quantification of the viruses, were 17 performed at the Division of Genetic Therapeutics, Jichi Medical University, and at the 18 section for Viral Vector Development, National Institute of Physiological Sciences. 19 20 Stereotaxic surgery 21 PV-Cre transgenic mice (>8 weeks of age, both sexes) were anesthetized with isoflurane 22 (1.0–2.5%) and placed in a stereotaxic apparatus (Stoelting). Bilateral injections of 23 AA V5-EF1α-DIO-hM4D(Gi)-mCherry (7.5 × 1012 viral genomes/ml; 250 nl per site) or 24 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2026. ; https://doi.org/10.64898/2026.04.15.718826doi: bioRxiv preprint 15 phosphate-buffered saline (PBS; 250 nl per site) into the NAc were performed using a 1 microinjector (Nanoliter 2020 Injector; World Precision Instruments) fitted with a 2 pulled glass capillary at a flow rate of 100 nl/min. Stereotaxic coordinates were 3 determined based on a mouse brain atlas [Paxinos et al., 2001]: NAc (anteroposterior, 4 mediolateral, and dorsoventral, in mm): +1.60, ±1.00, -5.00 and -4.50; and +0.80, 5 ±1.00, -5.20 and -4.50. 6 7 Measurements of dopamine release 8 Extracellular dopamine levels in the NAc of male mice aged 8–15 months were 9 quantified utilizing an in vivo microdialysis system. Under anesthesia induced by a 10 combination of medetomidine (0.3 mg/kg), midazolam (4 mg/kg), and butorphanol (5 11 mg/kg), a guide cannula (AG-6; EICOM) was implanted into the NAc shell. The 12 coordinates for implantation, based on the mouse brain atlas [Paxinos et al., 2001], were 13 as follows (in mm): anteroposterior +1.30, mediolateral +0.70, dorsoventral -4.20. The 14 cannula was affixed to the skull using stainless steel screws and dental acrylic cement. 15 Three to four days post-surgery, a dialysis probe equipped with a 1 mm-length 16 membrane (FX-I-6-01, EICOM) was inserted into the guide cannula. This probe was 17 perfused with artificial cerebrospinal fluid, composed of 147.2 mM NaCl, 4.0 mM KCl, 18 and 2.3 mM CaCl2, at a flow rate of 0.8 μl/min using a microsyringe injector (ESP-64, 19 EICOM). Dialysate samples were collected and injected every 30 minutes using an auto 20 injector (EAS-20S, EICOM). The samples were subsequently separated using an SC-21 50DS column (EICOM) to facilitate dopamine measurement via high-performance 22 liquid chromatography (HPLC) with an electrochemical detection system (HTEC-500, 23 EICOM), where the working electrode was maintained at +700 mV (vs. Ag/AgCl) at a 24 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2026. ; https://doi.org/10.64898/2026.04.15.718826doi: bioRxiv preprint 16 temperature of 25 °C. The mobile phase consisted of 83 % 0.1 M citrate/0.1 M sodium 1 acetate buffer (pH 3.9) and 17 % methanol, containing 140 mg/l sodium 1-2 octanesulfonate and 5 mg/l EDTA-2Na, delivered at a flow rate of 0.23 ml/min. 3 Protocol was controlled and data analysis was performed using Power Chrom software 4 (ADI instruments Japan). 5 The animal was acclimated for over two hours in the experimental arena, which 6 measured 30 × 30 × 35 cm, until the baseline stabilized. Following the two-hour 7 measurement period (four samples), the animals were administered methamphetamine 8 (1.0 mg/kg, i.p.), and samples were obtained for more than two hours. 9 10 Three‐chamber social approach test 11 The three-chamber apparatus consisted of a rectangular clear Plexiglas box (43 × 63 12 cm) divided into three equal-sized chambers (21 × 43 cm) by transparent Plexiglas 13 partitions with small openings that allowed mice to freely move between chambers. A 14 cylindrical wire cage was placed in each corner of the two side chambers to enclose an 15 unfamiliar 11–12-week-old C57BL/6J male mouse (stranger). Test male mice (3–5 16 months of age) were first habituated to the apparatus for 10 min, with an empty wire 17 cage placed in each of the side chambers. For the sociability test, a wire cage containing 18 a stranger mouse was placed in one side chamber, and the subject mouse was initially 19 placed in the center chamber and allowed to freely explore the entire apparatus for a 10 20 min. The location of the stranger mouse (left or right chamber) was alternated between 21 trials. Behavior was recorded using a video camera mounted above the apparatus. The 22 total time spent in proximity to each cage (empty or containing the stranger mouse) was 23 quantified manually in a blinded manner. 24 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2026. ; https://doi.org/10.64898/2026.04.15.718826doi: bioRxiv preprint 17 1 Startle response and prepulse inhibition (PPI) test 2 The startle response and PPI test was conducted as previously described [Nakao et al., 3 2015; Suzuki et al., 2024]. Male mice (6–8 months of age) were placed in a Plexiglas 4 cylinder with a continuous 70 dB white noise background and allowed to habituate for 5 5 min. An acoustic startle stimulus (40 ms; 110- or 120-dB) was presented either alone or 6 preceded by a prepulse stimulus (20 ms; 74- or 78-dB) during the PPI trials. Each test 7 block consisted of two startle alone trials (110- and 120-dB) and four prepulse–startle 8 trials (74-dB prepulse + 110- or 120-dB startle; 78-dB prepulse + startle 110- or 120-dB 9 startle). Each mouse underwent ten blocks. The inter-trial interval varied randomly with 10 an average of 15 s (range: 10–20 s). Startle responses and PPI were recorded 11 automatically. 12 13 Open field test 14 Male mice (3–4 months of age) were placed in a corner of a square open-field apparatus 15 (45 × 45 × 20 cm) illuminated at 100 lx and allowed to freely explore for 120 min. The 16 total distance traveled (cm) and the time spent in the center area (20 × 20 cm) were 17 automatically recorded using manufacturer’s software (Smart3.0, PanLab). 18 19 Statistical analyses 20 Data were analyzed using Student's t-test for two-group comparisons or one-way or 21 two-way analysis of variance (ANOV A), followed by Tukey’s post hoc test for 22 parametric data (KyPlot; KyensLab Inc.), as appropriate. Data are presented as the mean 23 ± standard error of the mean (s.e.m.) or as box-and-whisker plots. In the box-and-24 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2026. ; https://doi.org/10.64898/2026.04.15.718826doi: bioRxiv preprint 18 whisker plots, the boxes indicate the median (thick line), the 25th and 75th percentiles, 1 and the whiskers represent the minimum and maximum values. Statistical significance 2 was defined as P < 0.05. 3 4 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2026. ; https://doi.org/10.64898/2026.04.15.718826doi: bioRxiv preprint 19 Acknowledgments 1 The authors thank Drs. Tamamaki and Tanahira (Kumamoto University) for generously 2 providing the PV-Cre-TG driver line. We also thank the members of the Department of 3 Neurophysiology and Brain Science, the Endowed Department of Cognitive Function 4 and Pathology, the Department of Neurodevelopmental Disorder Genetics, and the staff 5 of the animal facility at Nagoya City University (NCU) for their support. We are 6 grateful to the Research Equipment Sharing Center at NCU for technical assistance. 7 8 Author Contributions 9 TS and KaY contributed to the study conception and design. Material preparation, data 10 collection, and analysis were performed by TS, ST, YY, HM, KK, WN, KoY , TK, YH, 11 TY , SI, HN, HH, KaY . The first draft of the manuscript was written by TS and K aY . All 12 authors have reviewed and approved the final manuscript. 13 14 Funding 15 This work was supported by grants from NCU; JSPS KAKENHI (Grant Numbers 16 22K07620 to TS, 23K27490 to KaY and 25K10820 to TS); and the Grant-in-Aid for 17 Outstanding Research Group Support Program at NCU (Grant Number 2401101). This 18 study also utilized research equipment shared under the MEXT Project for Promoting 19 Public Utilization of Advanced Research Infrastructure (Program for Supporting 20 Construction of Core Facilities; Grant Number JPMXS0441500024). 21 22 Data Availability 23 The datasets generated during and/or analyzed during the current study are available 24 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2026. ; https://doi.org/10.64898/2026.04.15.718826doi: bioRxiv preprint 20 from the corresponding author on reasonable request. 1 2 Declarations 3 Ethics Approval 4 All animal breeding and experimental procedures were approved by the Institutional 5 Animal Care and Use Committee of NCU (approval No. 19-032, approved 20 Dec 6 2022; approval No. 23-024, approved 21 Apr 2023). All procedures were conducted in 7 accordance with the ARRIVE guidelines and the institutional guidelines and regulations 8 of NCU. 9 10 Consent to Participate 11 Not applicable. 12 13 Consent for Publication 14 Not applicable. 15 16 Competing Interests 17 The authors declare no competing interests. 18 19 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2026. ; https://doi.org/10.64898/2026.04.15.718826doi: bioRxiv preprint 21

1 1. Aupy J, Ribot B, Guehl D, Nguyen TH, Burbaud P (2023) Selective striatal fast-2 spiking interneuron inhibition induces cortical seizure. J Neurosci Res 102 3 (1):e25270. doi:10.1002/jnr.25270 4 2. Balakrishna T, Curtis D (2020) Assessment of Potential Clinical Role for Exome 5 Sequencing in Schizophrenia. Schizophr Bull 46 (2):328-335. 6 doi:10.1093/schbul/sbz057 7 3. Braff DL, Geyer MA, Swerdlow NR (2001) Human studies of prepulse inhibition of 8 startle: normal subjects, patient groups, and pharmacological studies. 9 Psychopharmacology (Berl) 156 (2-3):234-258. doi:10.1007/s002130100810 10 4. Buxbaum JD, Daly MJ, Devlin B, Lehner T, Roeder K, State MW, Autism 11 Sequencing C (2012) The autism sequencing consortium: large-scale, high-12 throughput sequencing in autism spectrum disorders. Neuron 76 (6):1052-1056. 13 doi:10.1016/j.neuron.2012.12.008 14 5. Carroll LS, Woolf R, Ibrahim Y , Williams HJ, Dwyer S, Walters J, Kirov G, 15 O'Donovan MC, Owen MJ (2016) Mutation screening of SCN2A in schizophrenia 16 and identification of a novel loss-of-function mutation. Psychiatr Genet 26 (2):60-17 65. doi:10.1097/YPG.0000000000000110 18 6. Castro DC, Bruchas MR (2019) A Motivational and Neuropeptidergic Hub: 19 Anatomical and Functional Diversity within the Nucleus Accumbens Shell. Neuron 20 102 (3):529-552. doi:10.1016/j.neuron.2019.03.003 21 7. Catterall WA (1992) Cellular and molecular biology of voltage-gated sodium 22 channels. Physiol Rev 72 (4 Suppl):S15-48. 23 doi:10.1152/physrev.1992.72.suppl_4.S15 24 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2026. ; https://doi.org/10.64898/2026.04.15.718826doi: bioRxiv preprint 22 8. Catterall WA, Goldin AL, Waxman SG (2005) International Union of Pharmacology. 1 XLVII. Nomenclature and structure-function relationships of voltage-gated sodium 2 channels. Pharmacol Rev 57 (4):397-409. doi:10.1124/pr.57.4.4 3 9. de Ligt J, Willemsen MH, van Bon BW, Kleefstra T, Yntema HG, Kroes T, Vulto-4 van Silfhout AT, Koolen DA, de Vries P, Gilissen C, del Rosario M, Hoischen A, 5 Scheffer H, de Vries BB, Brunner HG, Veltman JA, Vissers LE (2012) Diagnostic 6 exome sequencing in persons with severe intellectual disability. N Engl J Med 367 7 (20):1921-1929. doi:10.1056/NEJMoa1206524 8 10. Fromer M, Pocklington AJ, Kavanagh DH, Williams HJ, Dwyer S, Gormley P, 9 Georgieva L, Rees E, Palta P, Ruderfer DM, Carrera N, Humphreys I, Johnson JS, 10 Roussos P, Barker DD, Banks E, Milanova V , Grant SG, Hannon E, Rose SA, 11 Chambert K, Mahajan M, Scolnick EM, Moran JL, Kirov G, Palotie A, McCarroll 12 SA, Holmans P, Sklar P, Owen MJ, Purcell SM, O'Donovan MC (2014) De novo 13 mutations in schizophrenia implicate synaptic networks. Nature 506 (7487):179-14 184. doi:10.1038/nature12929 15 11. Geyer MA (2006) The family of sensorimotor gating disorders: comorbidities or 16 diagnostic overlaps? Neurotox Res 10 (3-4):211-220. doi:10.1007/BF03033358 17 12. Gittis AH, Leventhal DK, Fensterheim BA, Pettibone JR, Berke JD, Kreitzer AC 18 (2011) Selective inhibition of striatal fast-spiking interneurons causes dyskinesias. J 19 Neurosci 31 (44):15727-15731. doi:10.1523/JNEUROSCI.3875-11.2011 20 13. Hoischen A, Krumm N, Eichler EE (2014) Prioritization of neurodevelopmental 21 disease genes by discovery of new mutations. Nat Neurosci 17 (6):764-772. 22 doi:10.1038/nn.3703 23 14. Iwasato T, Datwani A, Wolf AM, Nishiyama H, Taguchi Y , Tonegawa S, Knopfel T, 24 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2026. ; https://doi.org/10.64898/2026.04.15.718826doi: bioRxiv preprint 23 Erzurumlu RS, Itohara S (2000) Cortex-restricted disruption of NMDAR1 impairs 1 neuronal patterns in the barrel cortex. Nature 406 (6797):726-731. 2 doi:10.1038/35021059 3 15. Iwasato T, Nomura R, Ando R, Ikeda T, Tanaka M, Itohara S (2004) Dorsal 4 telencephalon-specific expression of Cre recombinase in PAC transgenic mice. 5 Genesis 38 (3):130-138. doi:10.1002/gene.20009 6 16. Johnson MR, Shkura K, Langley SR, Delahaye-Duriez A, Srivastava P, Hill WD, 7 Rackham OJ, Davies G, Harris SE, Moreno-Moral A, Rotival M, Speed D, 8 Petrovski S, Katz A, Hayward C, Porteous DJ, Smith BH, Padmanabhan S, Hocking 9 LJ, Starr JM, Liewald DC, Visconti A, Falchi M, Bottolo L, Rossetti T, Danis B, 10 Mazzuferi M, Foerch P, Grote A, Helmstaedter C, Becker AJ, Kaminski RM, Deary 11 IJ, Petretto E (2016) Systems genetics identifies a convergent gene network for 12 cognition and neurodevelopmental disease. Nat Neurosci 19 (2):223-232. 13 doi:10.1038/nn.4205 14 17. Kamiya K, Kaneda M, Sugawara T, Mazaki E, Okamura N, Montal M, Makita N, 15 Tanaka M, Fukushima K, Fujiwara T, Inoue Y , Yamakawa K (2004) A nonsense 16 mutation of the sodium channel gene SCN2A in a patient with intractable epilepsy 17 and mental decline. J Neurosci 24 (11):2690-2698. doi:10.1523/JNEUROSCI.3089-18 03.2004 19 18. Li J, Cai T, Jiang Y , Chen H, He X, Chen C, Li X, Shao Q, Ran X, Li Z, Xia K, Liu 20 C, Sun ZS, Wu J (2016) Genes with de novo mutations are shared by four 21 neuropsychiatric disorders discovered from NPdenovo database. Mol Psychiatry 21 22 (2):290-297. doi:10.1038/mp.2015.40 23 19. Li L, Huang Q, Hu J, Jin M, Zhuo Y , Ke W, He Q, Xiao Y , Zhang X, Wang W, 24 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2026. ; https://doi.org/10.64898/2026.04.15.718826doi: bioRxiv preprint 24 Cheng TL, Tai Y , Guo F, Yu J, Yulong L, He J, Li B, Shu Y (2025) Selective loss of 1 Scn2a in ventral tegmental area dopaminergic neurons leads to dopamine system 2 hypofunction and autistic-like behaviors. Neuron 113 (18):2997-3014 e2998. 3 doi:10.1016/j.neuron.2025.06.003 4 20. Li T, Tian C, Scalmani P, Frassoni C, Mantegazza M, Wang Y , Yang M, Wu S, Shu 5 Y (2014) Action potential initiation in neocortical inhibitory interneurons. PLoS 6 Biol 12 (9):e1001944. doi:10.1371/journal.pbio.1001944 7 21. Liao Y , Deprez L, Maljevic S, Pitsch J, Claes L, Hristova D, Jordanova A, Ala-8 Mello S, Bellan-Koch A, Blazevic D, Schubert S, Thomas EA, Petrou S, Becker AJ, 9 De Jonghe P, Lerche H (2010) Molecular correlates of age-dependent seizures in an 10 inherited neonatal-infantile epilepsy. Brain 133 (Pt 5):1403-1414. 11 doi:10.1093/brain/awq057 12 22. Marinescu AM, Labouesse MA (2024) The nucleus accumbens shell: a neural hub at 13 the interface of homeostatic and hedonic feeding. Front Neurosci 18:1437210. 14 doi:10.3389/fnins.2024.1437210 15 23. Mena A, Ruiz-Salas JC, Puentes A, Dorado I, Ruiz-Veguilla M, De la Casa LG 16 (2016) Reduced Prepulse Inhibition as a Biomarker of Schizophrenia. Front Behav 17 Neurosci 10:202. doi:10.3389/fnbeh.2016.00202 18 24. Middleton SJ, Kneller EM, Chen S, Ogiwara I, Montal M, Yamakawa K, McHugh 19 TJ (2018) Altered hippocampal replay is associated with memory impairment in 20 mice heterozygous for the Scn2a gene. Nat Neurosci 21 (7):996-1003. 21 doi:10.1038/s41593-018-0163-8 22 25. Miyamoto H, Tatsukawa T, Shimohata A, Yamagata T, Suzuki T, Amano K, Mazaki 23 E, Raveau M, Ogiwara I, Oba-Asaka A, Hensch TK, Itohara S, Sakimura K, 24 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2026. ; https://doi.org/10.64898/2026.04.15.718826doi: bioRxiv preprint 25 Kobayashi K, Kobayashi K, Yamakawa K (2019) Impaired cortico-striatal 1 excitatory transmission triggers epilepsy. Nat Commun 10 (1):1917. 2 doi:10.1038/s41467-019-09954-9 3 26. Miyazaki H, Oyama F, Inoue R, Aosaki T, Abe T, Kiyonari H, Kino Y , Kurosawa M, 4 Shimizu J, Ogiwara I, Yamakawa K, Koshimizu Y , Fujiyama F, Kaneko T, Shimizu 5 H, Nagatomo K, Yamada K, Shimogori T, Hattori N, Miura M, Nukina N (2014) 6 Singular localization of sodium channel beta4 subunit in unmyelinated fibres and its 7 role in the striatum. Nat Commun 5:5525. doi:10.1038/ncomms6525 8 27. Nakao A, Miki T, Shoji H, Nishi M, Takeshima H, Miyakawa T, Mori Y (2015) 9 Comprehensive behavioral analysis of voltage-gated calcium channel beta-10 anchoring and -regulatory protein knockout mice. Front Behav Neurosci 9:141. 11 doi:10.3389/fnbeh.2015.00141 12 28. Ogiwara I, Ito K, Sawaishi Y , Osaka H, Mazaki E, Inoue I, Montal M, Hashikawa T, 13 Shike T, Fujiwara T, Inoue Y , Kaneda M, Yamakawa K (2009) De novo mutations of 14 voltage-gated sodium channel alphaII gene SCN2A in intractable epilepsies. 15 Neurology 73 (13):1046-1053. doi:10.1212/WNL.0b013e3181b9cebc 16 29. Ogiwara I, Miyamoto H, Tatsukawa T, Yamagata T, Nakayama T, Atapour N, Miura 17 E, Mazaki E, Ernst SJ, Cao D, Ohtani H, Itohara S, Yanagawa Y , Montal M, Yuzaki 18 M, Inoue Y , Hensch TK, Noebels JL, Yamakawa K (2018) Nav1.2 haplodeficiency 19 in excitatory neurons causes absence-like seizures in mice. Commun Biol 1:96. 20 doi:10.1038/s42003-018-0099-2 21 30. Paxinos G, Franklin KBJ (2001) The mouse brain in stereotaxic coordinates. 2nd 22 edn. Academic Press, San Diego ; London 23 31. Planells-Cases R, Caprini M, Zhang J, Rockenstein EM, Rivera RR, Murre C, 24 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2026. ; https://doi.org/10.64898/2026.04.15.718826doi: bioRxiv preprint 26 Masliah E, Montal M (2000) Neuronal death and perinatal lethality in voltage-gated 1 sodium channel alpha(II)-deficient mice. Biophys J 78 (6):2878-2891. 2 doi:10.1016/S0006-3495(00)76829-9 3 32. Powell SB, Zhou X, Geyer MA (2009) Prepulse inhibition and genetic mouse 4 models of schizophrenia. Behav Brain Res 204 (2):282-294. 5 doi:10.1016/j.bbr.2009.04.021 6 33. Qi J, Zhang S, Wang HL, Barker DJ, Miranda-Barrientos J, Morales M (2016) VTA 7 glutamatergic inputs to nucleus accumbens drive aversion by acting on GABAergic 8 interneurons. Nat Neurosci 19 (5):725-733. doi:10.1038/nn.4281 9 34. Rauch A, Wieczorek D, Graf E, Wieland T, Endele S, Schwarzmayr T, Albrecht B, 10 Bartholdi D, Beygo J, Di Donato N, Dufke A, Cremer K, Hempel M, Horn D, Hoyer 11 J, Joset P, Ropke A, Moog U, Riess A, Thiel CT, Tzschach A, Wiesener A, 12 Wohlleber E, Zweier C, Ekici AB, Zink AM, Rump A, Meisinger C, Grallert H, 13 Sticht H, Schenck A, Engels H, Rappold G, Schrock E, Wieacker P, Riess O, 14 Meitinger T, Reis A, Strom TM (2012) Range of genetic mutations associated with 15 severe non-syndromic sporadic intellectual disability: an exome sequencing study. 16 Lancet 380 (9854):1674-1682. doi:10.1016/S0140-6736(12)61480-9 17 35. Scofield MD, Heinsbroek JA, Gipson CD, Kupchik YM, Spencer S, Smith AC, 18 Roberts-Wolfe D, Kalivas PW (2016) The Nucleus Accumbens: Mechanisms of 19 Addiction across Drug Classes Reflect the Importance of Glutamate Homeostasis. 20 Pharmacol Rev 68 (3):816-871. doi:10.1124/pr.116.012484 21 36. Sugawara T, Tsurubuchi Y , Agarwala KL, Ito M, Fukuma G, Mazaki-Miyazaki E, 22 Nagafuji H, Noda M, Imoto K, Wada K, Mitsudome A, Kaneko S, Montal M, 23 Nagata K, Hirose S, Yamakawa K (2001) A missense mutation of the Na+ channel 24 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2026. ; https://doi.org/10.64898/2026.04.15.718826doi: bioRxiv preprint 27 alpha II subunit gene Na(v)1.2 in a patient with febrile and afebrile seizures causes 1 channel dysfunction. Proc Natl Acad Sci U S A 98 (11):6384-6389. 2 doi:10.1073/pnas.111065098 3 37. Suzuki T, Hattori S, Mizukami H, Nakajima R, Hibi Y , Kato S, Matsuzaki M, Ikebe 4 R, Miyakawa T, Yamakawa K (2024) Inversed Effects of Nav1.2 Deficiency at 5 Medial Prefrontal Cortex and Ventral Tegmental Area for Prepulse Inhibition in 6 Acoustic Startle Response. Mol Neurobiol 61 (2):622-634. doi:10.1007/s12035-023-7 03610-6 8 38. Suzuki T, Kondo T, Yamagata T, Hibi Y , Mizukami H, Kobayashi K, Yamakawa K 9 (2026) Reduced activity of nucleus accumbens parvalbumin-expressing fast-spiking 10 inhibitory neurons causes convulsive seizures. bioRxiv:2026.2003.2009.710428. 11 doi:10.64898/2026.03.09.710428 12 39. Takahashi H, Hashimoto R, Iwase M, Ishii R, Kamio Y , Takeda M (2011) Prepulse 13 inhibition of startle response: recent advances in human studies of psychiatric 14 disease. Clin Psychopharmacol Neurosci 9 (3):102-110. 15 doi:10.9758/cpn.2011.9.3.102 16 40. Tanahira C, Higo S, Watanabe K, Tomioka R, Ebihara S, Kaneko T, Tamamaki N 17 (2009) Parvalbumin neurons in the forebrain as revealed by parvalbumin-Cre 18 transgenic mice. Neurosci Res 63 (3):213-223. doi:10.1016/j.neures.2008.12.007 19 41. Tatsukawa T, Raveau M, Ogiwara I, Hattori S, Miyamoto H, Mazaki E, Itohara S, 20 Miyakawa T, Montal M, Yamakawa K (2019) Scn2a haploinsufficient mice display 21 a spectrum of phenotypes affecting anxiety, sociability, memory flexibility and 22 ampakine CX516 rescues their hyperactivity. Mol Autism 10:15. 23 doi:10.1186/s13229-019-0265-5 24 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2026. ; https://doi.org/10.64898/2026.04.15.718826doi: bioRxiv preprint 28 42. Tavassoli T, Kolevzon A, Wang AT, Curchack-Lichtin J, Halpern D, Schwartz L, 1 Soffes S, Bush L, Grodberg D, Cai G, Buxbaum JD (2014) De novo SCN2A splice 2 site mutation in a boy with Autism spectrum disorder. BMC Med Genet 15:35. 3 doi:10.1186/1471-2350-15-35 4 43. Xu Y , Lin Y , Yu M, Zhou K (2024) The nucleus accumbens in reward and aversion 5 processing: insights and implications. Front Behav Neurosci 18:1420028. 6 doi:10.3389/fnbeh.2024.1420028 7 44. Yamagata T, Ogiwara I, Mazaki E, Yanagawa Y , Yamakawa K (2017) Nav1.2 is 8 expressed in caudal ganglionic eminence-derived disinhibitory interneurons: 9 Mutually exclusive distributions of Nav1.1 and Nav1.2. Biochem Biophys Res 10 Commun 491 (4):1070-1076. doi:10.1016/j.bbrc.2017.08.013 11 45. Yamagata T, Ogiwara I, Tatsukawa T, Suzuki T, Otsuka Y , Imaeda N, Mazaki E, 12 Inoue I, Tokonami N, Hibi Y , Itohara S, Yamakawa K (2023) Scn1a-GFP transgenic 13 mouse revealed Nav1.1 expression in neocortical pyramidal tract projection 14 neurons. Elife 12. doi:10.7554/eLife.87495 15 46. Yang J, Xiao Y , Li L, He Q, Li M, Shu Y (2019) Biophysical Properties of Somatic 16 and Axonal V oltage-Gated Sodium Channels in Midbrain Dopaminergic Neurons. 17 Front Cell Neurosci 13:317. doi:10.3389/fncel.2019.00317 18 19 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2026. ; https://doi.org/10.64898/2026.04.15.718826doi: bioRxiv preprint 29 Figure Legends 1 Figure 1. Reduced cortico-accumbal excitatory input and suppression of FSIs in 2 the NAc impair sociability. (A) Schematic of the experimental design. (B) Schematic 3 illustration of neural circuits projecting to the NAc. (C) During the sociability phase of 4 the three-chamber social approach test, Scn2afl/+/Emx1-Cre mice spent less time around 5 the cage containing a stranger mouse, whereas control mice showed a clear preference 6 for the cage containing the stranger mouse over the empty cage. (D, E) In the sociability 7 phase of the three-chamber test, PV-Cre/NAc-hM4Di mice exhibited normal sociability 8 before CNO administration (Pre-CNO; D), but showed reduced sociability after CNO 9 administration (Post-CNO; E). Black dots represent individual mice. The number of 10 mice in each group is indicated in parentheses. 3ch test, three-chamber test; AA V, 11 adeno-associated virus; AMG, amygdala; CNO, clozapine-N-oxide; CTX, neocortex; 12 FSI, fast-spiking interneuron; HIPP, hippocampus; MSN, medium spiny neuron; NAc, 13 nucleus accumbens; OF test, open field test; PBS, phosphate-buffered saline; PPI test, 14 prepulse inhibition test; PV+, parvalbumin-positive. 15 16 Figure 2. Selective Scn2a deficiency in dorsal telencephalic excitatory neurons 17 shows a trend toward reduced PPI. (A) Acoustic startle responses to two sound 18 stimulus intensities (110- and 120-dB) did not differ significantly between 19 Scn2afl/+/Emx1-Cre mice and control Scn2afl/+ mice. (B) In the PPI test, the percentage 20 PPI of the startle response to a 120-dB pulse preceded by 74- or 78-dB prepulses 21 showed a trend toward reduction in Scn2afl/+/Emx1-Cre mice compared with control 22 mice. Black dots represent individual mice, and the number of mice is indicated in 23 parentheses. 24 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2026. ; https://doi.org/10.64898/2026.04.15.718826doi: bioRxiv preprint 30 Figure 3. Inactivation of NAc FSIs does not affect exploratory or anxiety-like 1 behavior in the open field test. (A) Total distance traveled in the open field test did not 2 differ between PV-Cre/hM4Di mice and control mice injected with PBS into the NAc, 3 either before (−) and after (+) CNO administration. (B) Time spent in the center zone of 4 the open field area was comparable between PV-Cre/hM4Di mice and control mice 5 injected with PBS into the NAc, both before (−) and after (+) CNO administration. 6 Black dots represent individual mice, and the number of mice in each group is indicated 7 in parentheses. 8 9 Figure 4. Dopamine release in the NAc is comparable between Scn2a mutant and 10 control mice. Dopamine release in the NAc of Scn2afl/+/Emx1-Cre mice (A, B) and 11 Scn2a+/- mice (C, D) was measured under a novel environmental condition (A, C) and 12 following methamphetamine administration (B, D). Data are presented as mean ± s.e.m. 13 (A, C) or as box-and-whisker plots (B, D). Black dots represent individual mice (B, D), 14 and the number of mice in each group is indicated in parentheses (A–D). DA, 15 dopamine. 16 17 Figure 5. Cortico-accumbal circuit models for sociability and PPI in Scn2a mutant 18 mice. Schematic models of neural circuit underlying sociability and prepulse inhibition 19 (PPI) in Scn2a mutant mice. These diagrams are modified and simplified from previous 20 published models [Swerdlow et al., 2001; Miyamoto et al., 2019; Cano et al., 2021; 21 Suzuki et al., 2024; Suzuki et al., 2026], and incorporate interpretations based on the 22 present findings. Hypothetical neural circuit models illustrating the mechanisms 23 underlying reduced PPI in Scn2afl/+/Emx1-Cre mice (A) and PV-Cre/hM4Di mice (B). 24 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2026. ; https://doi.org/10.64898/2026.04.15.718826doi: bioRxiv preprint 31 Upward and downward arrows indicate increased and decreased neural transmission, 1 respectively. CN, cochlear nuclei; CTX, neocortex; D1R, dopamine D1 receptor; D2R, 2 dopamine D2 receptor; DA-N, dopaminergic neuron (gray circle) and transmission 3 (gray arrows); FSI, fast-spiking interneuron; GABA, GABAergic neurons (black 4 circles) and transmission (black lines); Glu, glutamatergic neurons (white circles) and 5 transmission (black arrows); Gly, glycinergic neurons (striped circle) and transmission 6 (gray lines); iCS, intratelencephalic cortico-striatal neurons; IN, interneurons; MDT, 7 mediodorsal thalamus; MSN, medium spiny neuron; NAc, nucleus accumbens; Nav1.1, 8 voltage-gated sodium channel α1 subunit; Nav1.2, voltage-gated sodium channel α2 9 subunit; PnC, caudal pontine reticular nucleus; PPI, prepulse inhibition; PPTg, 10 pedunculopontine tegmental nucleus; SMN, spinal motor nerve; VP, ventral pallidum; 11 VTA, ventral tegmental area. 12 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2026. ; https://doi.org/10.64898/2026.04.15.718826doi: bioRxiv preprint Figure 1 A Birth 320 3ch test 4 5 6 7 8 9 101 PPI test Scn2a fl/+/Emx1 -Cre months of age Dopamine measurement Birth 320 4 5 6 7 8 9 101 Scn2a +/– months of age Dopamine measurement 11 12 13 14 15 16 Birth AAV Injection months of age 320 OF test 3ch test 4 5 6 7 8 9 101 PPI test PV-Cre/NAc -hM4D(Gi) Empty Stranger 0 50 100 150 200 250 Pre-CNO Time around cage (s) 0 40 80 120 160 Time around cage (s) Empty Stranger NAc-hM4D(Gi) (n=12) Empty Stranger Empty Stranger NAc-PBS (n=14) Post-CNO NAc-hM4D(Gi) (n=12) NAc-PBS (n=14) P=3.72E-06 P=1.78E-05 P=3.65E-03 D E Three-chamber test 0 50 100 150 200 250 300 350 Empty Stranger Empty Stranger Scn2afl/+/Emx1-Cre (n=18) Scn2afl/+ (n=16) P=1.38E-03 Time around cage (s) C Three-chamber testB CTX, HIPP, AMG, etc. Nav1.2+ Emx1+ NAc FSI MSN PV+ Glu Glu GABA .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2026. ; https://doi.org/10.64898/2026.04.15.718826doi: bioRxiv preprint Figure 2 0 0.02 0.04 0.06 0.08 0.10 0.12 Startle amplitude -150 -100 -50 0 50 100 110dB 74dB Pre-puls inhibition (%) Scn2a fl/+ (n=16) Scn2a fl/+/Emx1 -Cre (n=18) 78dB 120dB 74dB 78dBPre-pulse Startle Prepulse inhibition of startle response 110dB 120dB A B .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2026. ; https://doi.org/10.64898/2026.04.15.718826doi: bioRxiv preprint Figure 3 0 20,000 40,000 60,000 80,000 100,000 120,000 140,000 160,000 0 500 1,000 1,500 2,000 2,500 3,000 3,500 Total distance (cm) Time in center (s) A B – + – + NAc-hM4D(Gi) (n=12) NAc-PBS (n=14) NAc-hM4D(Gi) (n=12) NAc-PBS (n=14) CNO – + – +CNO Open field test .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2026. ; https://doi.org/10.64898/2026.04.15.718826doi: bioRxiv preprint Methamphetamine administration 0 50 100 150 200 0-30 30-60 60-90 90-120 DA (% basal) Time (min) Novel environment Scn2a(fl/+)/Emx1-Cre (n=6) Scn2a(fl/+) (n=6) 0 50 100 150 0-30 30-60 60-90 90-120 DA (% basal) Time (min) Novel environment Scn2a(+/-) (n=8) WT (n=8) 0 5,000 10,000 15,000 20,000 25,000 DA (% basal) Scn2a +/- (n=8) Scn2a +/+ (n=8) 0 1,000 2,000 3,000 4,000DA (% basal) Methamphetamine administration A B C D Scn2a +/– (n=8) Scn2a +/+ (n=8) Scn2a fl/+/Emx1 -Cre (n=6) Scn2a fl/+ (n=6) Figure 4 Scn2afl/+/Emx1-Cre (n=5) Scn2a fl/+ (n=6) Dopamine measurement Dopamine measurement .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2026. ; https://doi.org/10.64898/2026.04.15.718826doi: bioRxiv preprint Figure 5 Glu GABA increase decreaseDopamine NAc VTANav1.1+ PV+ FSI Nav1.2+ D1R+ MSN GABA IN Nav1.2+ DA-N VP MDT Nav1.2+ D2R+ MSN PV+ ?? ? CTX NAc VTANav1.1+ PV+ FSI Nav1.2+ D1R+ MSN GABA IN Nav1.2+ DA-N VP MDT Nav1.2+ D2R+ MSN Nav1.2+ iCS PV+ ?? ? CTXNav1.2+ iCS Scn2afl/+/Emx1-Cre PV-Cre/NAc-hM4D(Gi)A B PPTg CN PnC Gly SMNstartle response PPI ? ? ? ? ? ? PPTg CN PnC SMNstartle response PPI? ? ? ? ? ?? ? ? ? ? ? ? ? .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2026. ; https://doi.org/10.64898/2026.04.15.718826doi: bioRxiv preprint