Fig 2: Decrease in postsynaptic density (PSD) area in postsynaptic membrane in Nsf+/- mice. Reconstruction of dendritic spines from serial FIB-SEM images. PSD was defined as postsynaptic membrane specialization. (A) Representative FIB-SEM images of the hippocampal CA1 region of Nsf+/+ and Nsf+/- mice (spine head in orange and PSD in red). (B) Examples of 3D-reconstructed spines from Nsf+/+ and Nsf+/- mice (head in transparent orange and PSD in red). (C) Correlation between the PSD area and spine head volume in mice of each genotype (Spearman’s rank-order test: Nsf+/+ mice, n = 58 synapses, r = 0.862, p < 0.001; Nsf+/- mice, n = 65 synapses, r = 0.784, p < 0.001). (D) The average PSD areas and (E) head volumes in the hippocampus of Nsf+/+ and Nsf+/- mice (mean ± SEM. Spearman’s rank-order test, p = 0.074).

Fig 3: Decrease in serotonin transporter (SERT) expression at the membrane surface in Nsf+/- mice. (A) Total and biotinylated membrane protein levels in the raphe of mouse of each genotype were analyzed by immunoblotting using an anti-SERT antibody. (B) Stain-Free gels were activated by UV exposure and imaged using a ChemiDocTM MP imager as protein loading control. (C) Relative band densities of SERT were quantified using scanning densitometry and normalized to protein loading control Results are expressed as a ratio of Nsf+/+ mice expression, resulting in a Nsf+/+ mice ratio of 1. (n = 4 for each genotype, mean ± SEM. Student’s t-test, **p < 0.01).

Fig 4: Deficits in social interaction and communication in Nsf+/- mice. (A) Schematic representation of three-chamber social interaction tests. Session I: the subject mouse was placed in the middle chamber without any stimulants. Session II–V: stranger 1 was placed inside a cage located in one of the chambers. Then, the subject mouse was placed at the center of the middle chamber. The subject mouse investigated the same stimulant (stranger) mouse during each of the four sessions (II–V). Session VI: stranger 2, instead of stranger 1, was placed inside the same cage as session II-V. Then, the subject mouse was placed at the center of the middle chamber. The subject mouse moved freely in the three chambers for 5 min for each session. (B) Yellow represents the virtual interaction areas near cages. (C) Time spent around cages with “stranger 1 or 2” or “Empty” by Nsf+/+ mice during sessions I–IV (n = 13, mean ± SEM. Two-way repeated measures ANOVA, p = 0.004, Tukey’s post hoc test, ***p < 0.001, *p < 0.05). (D) Time spent around cages with “stranger 1 or 2” or “Empty” by Nsf+/- mice during sessions I–IV (n = 13, Mean ± SEM. Two-way repeated measures ANOVA, p = 0.376). (E) Representative mount-graph of each genotype during session II. (F) The preference index of the sociability (n = 13 for each genotype, mean ± SEM. Wilcoxon/Kruskal-Wallis test, *p < 0.05). (G) Number of ultrasonic vocalizations emitted by postnatal day 6 pups of each genotype during a 10 min separation from their mother (n = 26 in Nsf+/+ mice and n = 20 in Nsf+/- mice, mean ± SEM. Spearman’s rank-order test, *p < 0.05).

Fig 5: Schematic representation of possible mechanisms through which haploinsufficiency of Nsf causes neurophysiological and behavioral phenotypes similar to autism spectrum disorder via defective membrane trafficking of serotonin transporter (SERT) and alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA) receptor in mice.