Prefrontal Cortex-Specific Knockdown of Neurexin-1 in Rats Induces Anxiety-Like Behavior, Repetitive Behaviors, and Altered Social Interactions: A Proteomic Study

Neurodevelopmental disorders (NDDs) encompass heterogeneous conditions with early-onset impairments in cognition, communication, adaptive behavior, and motor function, often persisting across the lifespan. Genetic and environmental etiologies are complex and overlapping among disorders such as ASD, ADHD, ID, and schizophrenia. Neurexins, especially NRXN1, are key synaptic organizers implicated in multiple NDDs via mutations, deletions, and splicing dysregulation. Prior mouse studies show that full α-neurexin knockout is lethal, while NRXN1α heterozygosity yields variable behavioral phenotypes, and clinical 2p16.3 (NRXN1) deletions are associated with PFC dysfunction and NDD risk. While NRXN1’s role in synaptogenesis is established, its impact on neurogenesis and neurite development remains controversial. This study investigates how prefrontal NRXN1 knockdown affects juvenile rat behavior relevant to NDDs (anxiety, repetitive behaviors, social interaction) and neurite outgrowth, and explores underlying molecular changes via proteomics.

- Genetic overlap across ASD, ADHD, schizophrenia, and other NDDs includes NRXN1 CNVs and mutations; NRXN1α heterozygous deletions in mice have produced increased anxiety and altered social behavior in some studies, while others report limited sociability effects. - Human data link NRXN1 deletions (often heterozygous, affecting promoter and early exons) to ASD, ID, epilepsy, and broader neuropsychiatric phenotypes, with PFC involvement suggested by 2p16.3 deletions. - Contradictory findings exist regarding neurite morphology in NRXN1-deficient human neurons: some reports show unchanged neurite measures, others show decreased neurite number/length. - The PFC is central to executive functions, social cognition, and emotion regulation; alterations in PFC structure/connectivity are common in NDDs and relate to social and repetitive phenotypes. - Prior work indicates NRXN1 orchestrates transsynaptic networks and affects learning/memory in ADHD models, and reduced α-neurexin-1 may diminish functional brain network efficiency involving thalamic–PFC connectivity.

Animals: Male Sprague-Dawley rats, prepubertal (3 weeks on arrival), housed at 23 ± 1 °C, 50 ± 5% humidity, standard light/dark cycle with ad libitum food/water. Ethics: Nanjing Medical University IACUC approval (IACUC-1711004). Acclimation: 1 week before procedures. In vivo NRXN1 knockdown: At 4 weeks, rats anesthetized with 10% chloral hydrate (0.35 ml/100 g i.p.). Stereotactic injections targeted medial prefrontal cortex (PFC) with AAV9-NRXN1-GFP shRNA targeting exon 1 (titer 1 × 10^12 TU/ml) or control AAV-Nc-GFP. Coordinates: 20° angle; AP +2.6 mm, ML ±0.7 mm from bregma, DV −3.0 mm. Infusion: 10 μL at 1 μL/min; needle left 10 min to prevent reflux. Groups (n=27 total): WT (n=9), Sh-Nc (n=9), Sh-NRXN1 (n=9). Postoperative monitoring for 2 weeks before behavioral testing. Injection sites verified by GFP IHC on 25–30 μm coronal sections. Behavioral assays: Conducted 2 weeks post-injection. Open-Field Test (OFT): 40 × 40 × 45 cm black arena, overhead light; 10-min session recorded by infrared camera; cleaned with 75% ethanol between trials. Outcomes: total distance, average speed, time in center, rearing frequency, self-grooming bouts and time (manual scoring). Three-Chamber Sociability Test: 120 × 80 × 40 cm apparatus under 10 lux; 15-min habituation then 10-min test with one unfamiliar male rat (6 weeks) in a barred cylinder versus an empty cylinder; entries, time in chambers, cylinder contacts and duration measured. Tracking: Shanghai Jiliang system. Primary neuron culture and in vitro knockdown: PFC dissected from E18 rat embryos; neurons plated on poly-L-lysine in DMEM + 10% horse serum, then switched to Neurobasal + 2% B27. Lentiviral shRNA against NRXN1 (MOI 10; titer 1 × 10^8 TU/ml) used for knockdown; groups: blank, negative control (Nc), Sh-NRXN1. Media changed 12 h post-infection; assays at 72 h. Knockdown verified by qRT-PCR, Western blot, and IF. Molecular assays: qRT-PCR using SYBR Green; 2^-ΔΔCt; exon-spanning primers (Table S1). Western blot: RIPA lysis, BCA quantification, SDS-PAGE (4–12%), PVDF transfer, primary antibodies against NRXN1 and GAPDH, HRP secondaries, ECL detection, ImageJ densitometry. IF: Fix 4% PFA, permeabilize 0.3% Triton X-100, block 5% serum, stain MAP2 (1:500) and secondary Alexa Fluor 488; Hoechst nuclear counterstain. Neurite quantification: Images from ≥5 random fields per well; ≥3 coverslips from different wells; 30 cells per group; ≥3 biological replicates. ImageJ with NeuronJ for total neurite length, longest neurite length, number of primary neurites; Sholl analysis for branching. Immunohistochemistry: Perfusion with saline then 4% PFA; postfix, 30% sucrose cryoprotection; 25 μm coronal sections; NRXN1 primary (1:30); DAB visualization. Proteomics (TMT-MS): Six-plex TMT labeling (KD in duplicate: TMT-129,130; Nc: TMT-126,127,128; 30 μg peptides/sample). LC-MS/MS searched against Rat UniProt (2015; 79,228 entries). Parameters: precursor tolerance 20 ppm, trypsin specificity, ≤2 missed cleavages, static mod on peptide N-termini and Lys, variable Met oxidation. Target-decoy approach; Percolator filtering; protein-level FDR 1%. Reporter ion quantification within 0.02 m/z. Normalization across channels; statistics via Friedman test with Benjamini-Hochberg correction (P < 0.00134) in Scaffold 4.5. Functional analyses: GO/KEGG, GSEA, Ingenuity Pathway Analysis (IPA). Automated capillary Western (Wes): Validation of selected proteins (ANXA1, ANXA4, GRB2) per manufacturer (ProteinSimple). Statistics: GraphPad Prism 7; tests for normality (Kolmogorov-Smirnov) and homogeneity (Levene's). Unpaired t-tests for pairwise; one-way ANOVA with Tukey post hoc for multiple comparisons. Significance P < 0.05. Data reported as mean ± SD.

- Knockdown efficiency: NRXN1 mRNA and protein significantly reduced in primary neurons 48–72 h post-transfection (qRT-PCR, WB, IF). In vivo PFC knockdown confirmed by GFP distribution and IHC; NRXN1 protein reduced in virus-affected PFC regions (one-way ANOVA F(2,6)=51.59, P<0.001). - Open-Field Test: No differences in total distance (F(2,24)=2.46, P>0.05) or average speed (F(2,24)=1.91, P>0.05). Sh-NRXN1 rats showed increased rearing frequency (F(2,24)=4.42, P=0.02) and reduced time in center (F(2,24)=8.23, P=0.002), indicating heightened anxiety-like behavior. Grooming increased: number of bouts (F(2,24)=4.58, P=0.02) and grooming time (F(2,24)=4.61, P=0.02), suggesting enhanced repetitive behaviors. - Three-Chamber Sociability: Time in social chamber differed among groups (F(2,24)=3.5, P=0.048); no change in center chamber (F(2,24)=1.058, P>0.05); empty chamber (F(2,24)=3.49, P=0.048). Sh-NRXN1 rats had increased entries into the social chamber (F(2,24)=4.16, P=0.028) and more contacts with the social cylinder (F(2,24)=5.2, P=0.013), but no increase in interaction duration with the social cylinder (F(2,24)=1.01, P=0.36). Pattern suggests more frequent but not prolonged social interactions, possibly reflecting repetitive/reciprocating behavior. - Neurite outgrowth (primary PFC neurons): NRXN1 knockdown reduced total neurite length (F(2,87)=7.05, P=0.001) and longest neurite length (F(2,87)=4.12, P=0.019); no difference in number of primary neurites (F(2,87)=1.36, P=0.261); decreased branch points (F(2,87)=3.29, P=0.042), indicating reduced neurite complexity. - Proteomics: Identified 3,296 proteins; 130 significantly altered by NRXN1 knockdown (57 upregulated, 73 downregulated). Top upregulated: Usp13, Setdb1, Hist1h1a, Spock3, Ror2. Top downregulated: Dpp9, Rlf, Grk3, NRXN1, Epas1. Enrichment analyses (GO/KEGG, GSEA, IPA) highlighted extracellular matrix, cell membrane, integrin/ECM receptor interactions, and morphology-related pathways; IPA emphasized links to mental diseases. - Validation: Automated capillary Western confirmed decreased ANXA1 (t=14.216, df=4, P<0.001), GRB2 (t=10.805, df=4, P<0.001), and ANXA4 (t=10.118, df=4, P=0.009) in Sh-NRXN1 neurons, supporting membrane/cytoskeletal pathway involvement.

The findings demonstrate that selective NRXN1 knockdown in juvenile rat medial PFC evokes anxiety-like behavior, increased grooming and rearing (repetitive/exploratory behaviors), and an atypical social pattern characterized by increased frequency of social approaches without prolonged interactions. These phenotypes align with core behavioral dimensions observed in NDDs and with prior reports linking NRXN1 dysfunction to anxiety and altered social behaviors, while also explaining discrepancies among mouse studies by differences in species, developmental stage (juvenile vs adult), genetic background, and regional specificity (PFC-focused knockdown). The observed deficits in neurite outgrowth and branching provide a cellular substrate for disrupted PFC circuitry, which may contribute to altered cognition, social behavior, and anxiety regulation. Proteomic alterations converge on extracellular matrix and membrane-associated pathways and key adaptor proteins (e.g., ANXA1, ANXA4, GRB2), suggesting that NRXN1 loss perturbs protein-protein interaction networks and membrane/cytoskeletal dynamics critical for neurite architecture and synaptic organization. Together, the behavioral, cellular, and proteomic data indicate that NRXN1 supports PFC network development and function, and its deficiency may drive NDD-related phenotypes via impaired neuritogenesis and membrane-associated signaling.

NRXN1 depletion confined to the medial PFC of juvenile rats induces anxiety-like and abnormal social behaviors, and impairs neurite outgrowth and complexity in primary PFC neurons. Proteomic profiling reveals 130 differentially expressed proteins with enrichment for extracellular matrix, membrane, and morphology-related pathways; decreased ANXA1, ANXA4, and GRB2 were validated, implicating membrane/cytoskeletal mechanisms. These results reinforce NRXN1 as a key NDD risk gene, expand the NRXN1 interactome, and suggest that disruptions in neurite and membrane-associated processes may underlie behavioral abnormalities. Further work is needed to delineate causal molecular mechanisms and therapeutic targets.

Explicit limitations were not extensively discussed. The authors note that further research is required to elucidate molecular mechanisms. Knockdown efficiency differed in vitro (~80%) versus in vivo (~50%), which may influence effect sizes. The study focused on juvenile male rats and PFC-specific knockdown, which may limit generalizability to other brain regions, sexes, species, and developmental stages.