Intergenerational Sex-Specific Transmission of Maternal Social Experience

Abstract

The social environment is a major determinant of individual stress response and lifetime health. The present study shows that (1) social enrichment has a significant impact on neuroplasticity and behaviour particularly in females; and (2) social enrichment in females can be transmitted to their unexposed female descendants. Two generations (F0 and F1) of male and female rats raised in standard and social housing conditions were examined for neurohormonal and molecular alterations along with changes in four behavioural modalities. In addition to higher cortical neuronal density and cortical thickness, social experience in mothers reduced hypothalamic-pituitary-adrenal (HPA) axis activity in F0 rats and their F1 non-social housing offspring. Only F0 social mothers and their F1 non-social daughters displayed improved novelty-seeking exploratory behaviour and reduced anxiety-related behaviour whereas their motor and cognitive performance remained unchanged. Also, cortical and mRNA measurements in the F1 generation were affected by social experience intergenerationally via the female lineage (mother-to-daughter). These findings indicate that social experience promotes cortical neuroplasticity, neurohormonal and behavioural outcomes, and these changes can be transmitted to the F1 non-social offspring in a sexually dimorphic manner. Thus, a socially stimulating environment may form new biobehavioural phenotypes not only in exposed individuals, but also in their intergenerationally programmed descendants.

Conflict of interest statement

The authors declare no competing interests.

Figures

Figure 1

Experimental design. F0 and F1 generations (males and females) were exposed to either standard and social housing conditions. Experiment 1(F0): Male and female pups gathered from 5 different litters were randomly assigned to four experimental groups (n = 12/group) in two housing conditions at postnatal day 21. While standard animals were kept for 96 days in standard housing in groups of 2–3, social animals were raised in groups of 12 for the same period of time. On days 97–101 behavioural assessments were completed. Five rats per group were maintained for pairing, and 7 rats were used for tissue collection. Mating: Five standard males and 5 social females, and also 5 social males and 5 standard females were paired for mating. Experiment 2(F1): Forty pups from F0 dams who were previously raised in either standard or social housing conditions were randomly selected for Experiment 2 (F1). Based on their mothers’ housing conditions, the F1 offspring were then split into males and females (n = 9–11/group) and housed in groups of 2–3 for 93 days. All F1 animals were euthanized for histological assessments after behavioural testing was completed on days 94–98.

Figure 2

(A,B) Changes in circulating CORT in F0 and F1 generations. Social experience for 96 days significantly reduced plasma CORT levels in F0 rats (n = 12/group). F1 offspring that did not experience social interaction during postnatal development inherited the reduced level of HPA axis activity from their social mother. No significant difference was found between F0 social males and females. (C,D) Effects of social experience on body weight in F0 and F1 generations. Only F0 social females (n = 12) showed lower body weight when compared with other groups. (E&F) Brain weight and housing conditions in F0 and F1 generations. Despite a slight increase in brain weight, the F0 socially-housed animals were not significantly different when compared with standard rats. Asterisks indicate significant differences: *p ≤ 0.05; One-Way ANOVA. Error bars show ± SEM.

Figure 3

Behavioural consequences of social experience in F0 and F1 generations. (A) Illustration of the corridor field task (CFT). To assess novelty-seeking behaviour, rats were individually allowed to freely explore the environment (165 × 165 × 35 cm) for 8 minutes. Two variations were used: CFT without the central object, and CFT with a central object. (B) Representative search paths of the F0 generation in the different CFT variations. F0 social animals, particularly females, explored the open and central zones more than their standard housing counterparts. (C–E) Time (seconds) spent in each zone of the CFT (corridor, open, central) by F0 generation. Note that social animals spent significantly more time in the open and central zones in the no-central-object task variation. Social females spent less time in the corridor zone and explored the open and central zones more than their social male siblings (n = 11–12/group). Exploration in the central-object variation also showed that F0 social females spent significantly less time in the corridor zone, and more time in the open and central zones when compared with social males and standard groups. (F) Representative path trajectories of the F1 generation in both CFT variations. Only standard females born to social mothers spent more time exploring the open and central zones. (G–I) Time (seconds) spent in each zone of the F1 generation in CFT. Only standard female offspring born to social mothers spent less time in the corridor and more time in open and central zones than other groups. Like their social mothers, F1 female offspring spent significantly less time in the corridor and more time searching the open zones in the central-object CFT version (n = 9–11/group). (J) Occupancy plots of paths during exploration of the central zone as an indicator for novelty-seeking behaviour. Each plot compiled from individual tracks represents the percent time spent in the central zone (n = 6–8/group). (K) Insignificant effect of Litter in F0 and F1 rats in terms of the time spent in different zones of the CFT. Litters in the F0 and F1 generations indicated a similar pattern of search preference in different zones of the CFT with and without a central object. (L) Representative path trajectories in the F0 generation in the elevated plus maze. Social housing rats explored the open arms more than the standard housing rats. (M,N) Time (seconds) spent in open arms for F0 and F1 generations. Social F0 animals (n = 8/group) spent more time in open arms than standard animals. Also, standard housing female F1 offspring born to social mothers explored open arms significantly more than their standard female counterparts born to standard mothers (n = 7–8/group). Circles in the bar graphs represent individual animals. Asterisks indicate significant differences: *p ≤ 0.05; One-way and Repeated-measures ANOVA. Error bars show ± SEM.

Figure 4

Neuromorphological alterations by social experience in F0 and F1 generations. (A,B) Quantitative cytoarchitectonics in F0 generation. Densitometry based on absolute gray value index (GVI) shown by approximate plane (plane 16, ~0.48 mm anterior to bregma) of stained brain sections. White squares represent three regions of interest (M2&M1, S1DZ, S1ULp) in cortical regions. Panel B: F0 social animals, particularly females, displayed an increased GVI (n = 6–7/group). (C,D) Cortical thickness in the F0 generation. Four points (medial, central, lateral, and ventrolateral) on three coronal brain sections were selected. Panel C shows a stained section and an atlas plate (plane 17, ~−0.20 mm posterior to bregma) from the right hemisphere. Panel D indicates increased thickness in all four cortical regions in F0 social animals (n = 6–7/group). (E) i: Photomicrograph of a Golgi-Cox stained neuron as seen through the camera lucida. Sholl analysis was employed for morphological changes including the dendritic length, dendritic branching and spine density. ii: Representative Golgi-Cox stained coronal sections (AP ~3.70 mm and AP ~2.70 mm; 200 µm thickness). A total of 12 prefrontal pyramidal neurons (6 per hemisphere) were traced per animal at 200× magnification. iii & iv: Computer-assisted reconstructions of Golgi-Cox stained neurons (layers II-III) of mPFC. (sub-panels, right) High-power representation of dendritic segments with dendritic spines (magnification 1000×) demonstrating increased number of spines in F1 non-social females born to F0 social mothers (IV-right) compared to other groups (n = 5–6/group; Scale bar = 20 μm). (F,G) Dendritic branching in F1 generation. Apical and basilar measurements indicated that F1 standard males and females born to social mothers had more dendritic branching than their standard housing counterparts (n = 5–6/group). Daughters of social mothers displayed more basilar branching than their male siblings. (H,I) Dendritic length in F1 generation. Only in females was apical dendritic length influenced by their mothers’ social experience. Basilar dendritic length was significantly affected in both sexes, particularly females, born to social mothers (n = 5–6/group). (J,K) Spine density in F1 generation. Apical spine density was highest in non-social F1 females born to social mothers. Basilar spine density was not affected by maternal social experience (n = 5–6/group). (L) mPFC brain-derived neurotrophic factor (BDNF) expression in the F1 generation. Representative coronal plate (AP 2.70 mm) and region of interest (ROI-gray rectangular) of mPFC. i-iv: BDNF expression (scale bar = 500 μm). (M,N) Standard animals, particularly females, born to social mothers expressed more BDNF mRNA and protein in the mPFC. Circles in the bar graphs represent individual animals. Asterisks indicate significant differences: *p ≤ 0.05; One-way and Repeated-measures ANOVA. Error bars show ± SEM.