Sociability and social novelty preference test. (A) Sociability test. No group difference was found. (B) Social novelty preference test. VPA-exposed male rats spent less time with S2 than did control male rats. Data are represented as mean ± SEM. VPA, valproic acid; O, object; S1, stranger 1; S2, stranger 2; SEM, standard error of the mean. **P < 0.01, ***P < 0.001.

Spatial distribution of independent components. Each component is shown in sagittal (left), transverse (middle), and coronal (right) planes. In each component, regions with intensities greater than two standard deviations from the mean are highlighted. Red indicates regions with positive values, and blue indicates regions with negative values. Independent components included in the connectivity estimation are listed in the following. IC1, positive: cerebellum, both; IC2, positive: medulla, both; IC3, positive: thalamus, both; IC4, positive: inferior colliculus, both; IC4, negative: motor cortex, left, and orbitofrontal cortex, left; IC5, positive: visual cortex, both; IC6, positive: anterodorsal hippocampus, both; IC7, positive: ventral tegmental area, both, and hypothalamus, both; IC8, positive: somatosensory cortex, both; IC8, negative: cerebellum, both; IC9, positive: somatosensory cortex, both; IC10, positive: orbitofrontal cortex, both, and cerebellum, both; IC11, positive: superior colliculus and inferior colliculus, left, and retrosplenial cortex, both; IC12, positive: auditory cortex, both; IC13, positive: somatosensory cortex, both; IC14, positive: olfactory bulb, right, and caudoputamen, right; IC15, positive: cingulate cortex, both; IC15, negative: cerebellum, left; IC16, positive: somatosensory cortex, both; IC16, negative: superior colliculus, left; IC17, positive: retrosplenial cortex, both; IC18, positive: entorhinal cortex, both; IC19, positive: amygdala, left, caudoputamen, left, and visual cortex, left; IC19, negative: thalamus, left; IC20, positive: cingulate cortex, both; IC21, positive: caudoputamen, both; IC22, positive: thalamus, both; IC23, positive: nucleus accumbens, right, and cerebellum, right; IC23, negative: somatosensory cortex, right, orbitofrontal cortex, left, and insular cortex, left; IC24, positive: nucleus accumbens, left; IC25, positive: nucleus accumbens, right; IC25, negative: orbitofrontal cortex, right; IC26, positive: orbitofrontal cortex, right, and caudoputamen, right; IC27, positive: orbitofrontal cortex, both, and entorhinal cortex, right; IC27, negative: olfactory bulb, right, and medial prefrontal cortex, right; IC28, positive: caudoputamen, both; IC28, negative: olfactory bulb, left; IC29, positive: caudoputamen, left; IC30, negative: somatosensory cortex, both; IC31, positive: olfactory bulb, both, and entorhinal cortex, right; IC31, negative: inferior colliculus, right; IC32, positive: motor cortex, right; IC33, positive: insular cortex, both; IC33, negative: retrosplenial cortex, right; IC34, positive: medial prefrontal cortex, left, cingulate cortex, left, and orbitofrontal cortex, right; IC34, negative: hypothalamus, left; IC35, positive: insular cortex, both, and motor cortex, both; IC36, positive: olfactory bulb, both; IC37, positive: olfactory bulb, both; IC38, positive: olfactory bulb, both. IC, independent component.

Activity changes induced by prenatal VPA exposure and the interaction with sex. (A) Regions with decreased activity in the group exposed to prenatal VPA. (B) Regions with increased activity in the group exposed to prenatal VPA. (C) Regions where activity decreased more in males than in females. (D) Regions where activity increased more in males than in females. Each component is shown in transverse plane. Red indicates regions with positive values, and blue indicates regions with negative values. VPA, valproic acid; IC, independent component; V, valproic-acid treated; C, control; M, male; F, female. **P < 0.01, ***P < 0.001.

Connectivity changes induced by prenatal VPA exposure and the interaction with sex, as estimated by sparse inverse covariance estimation. (A) The main effect of VPA exposure: the blue line indicates the connections weakened by VPA treatment, while the red line indicates the connections strengthened by VPA treatment. (B) The main effect of sex: the blue line indicates connections that were weaker in males than females. (C) The interaction effect between VPA exposure and sex: the blue line indicates areas with less pronounced connectivity increases in males when compared to females following VPA exposure, while the red line indicates more pronounced increases in connectivity in males when compared to females following VPA treatment. The activity in the IC1* (the cerebellum) and IC35* (the insula) were correlated with autistic-like behavior in Fig. . Each component is shown in transverse plane. Line thickness and the corresponding number indicate the relative strength (, estimated partial correlation coefficient) of the connections. Only the connections with a FDR ≤ 0.05 were shown. VPA, valproic acid; IC, independent component.

Spatial patterns associated with normal social behavior revealed by sparse canonical correlation analysis. Normal social behavior was positively correlated with activity in the insular cortex and motor cortex (IC35), as well as that in the ventral tegmental area and hypothalamus (IC7), and negatively correlated with activity in the cerebellum (IC1). The relative contribution of each behavior or IC is shown in numbers. The number in red indicates positive correlation, and the number in blue indicates negative correlation. Each component is shown in sagittal (left), transverse (middle), and coronal (right) planes. IC, independent component.

Activity and connectivity changes induced by prenatal VPA exposure, and brain regions associated with normal social behavior (A) and the effect of VPA exposure on sex (B). (A) The main effects of VPA exposure and the results from sparse canonical correlation analysis are shown. The blue circle indicates the regions with diminished activity in the group exposed to prenatal VPA, while the red circle indicates the regions with increased activity in the VPA exposed group. The blue line indicates the connections weakened in the group exposed to VPA, while the red line indicates the connections strengthened in the VPA exposed group. Line thickness indicates the relative strength of the connections. The blue dashed circle indicates the regions positively correlated with normal social behavior, while the red dashed circle indicates the regions negatively correlated with normal social behavior. (B) Interaction on connectivity between VPA exposure and sex are displayed. The red line indicates the connections more strengthened in males than in females when exposed to VPA, while the blue line indicates the connections less increased in males than in females by VPA exposure. VPA, valproic acid; IC, independent component; NAc, nucleus accumbens; CB, cerebellum; CG, cingulate cortex; CP, caudoputamen; HA, anterodorsal hippocampus; HY, hypothalamus; IC, inferior colliculus; INS, insular cortex; MB, midbrain; OB, olfactory bulb; OF, orbitofrontal cortex; RSP, retrosplenial cortex; SC, superior colliculus; SS, somatosensory cortex; TH, thalamus; VIS, visual cortex; VTA, ventral tegmental area.