Effects of maternal social isolation on adult rodent offspring cognition

Abstract

Prenatal experiences can influence offspring physiology and behaviour through the lifespan. Various forms of prenatal stress impair adult learning and memory function and can lead to increased occurrence of anxiety and depression. Clinical work suggests that prenatal stress and maternal depression lead to similar outcomes in children and adolescents, however the long-term effects of maternal depression are less established, particularly in well controlled animal models. Social isolation is common in depressed individuals and during the recent COVID-19 pandemic. Accordingly, for this study we were interested in the effects of maternal stress induced via social isolation on adult offspring cognitive functions including spatial, stimulus-response, and emotional learning and memory that are mediated by different networks centered on the hippocampus, dorsal striatum, and amygdala, respectively. Tasks included a discriminative contextual fear conditioning task and cue-place water task. Pregnant dams in the social isolation group were single housed prior to and throughout gestation. Once offspring reached adulthood the male offspring were trained on a contextual fear conditioning task in which rats were trained to associate one of two contexts with an aversive stimulus and the opposing context remained neutral. Afterwards a cue-place water task was performed during which they were required to navigate to both a visible and invisible platform. Fear conditioning results revealed that the adult offspring of socially isolated mothers, but not controls, were impaired in associating a specific context with a fear-inducing stimulus as assessed by conditioned freezing and avoidance. Results from the water task indicate that adult offspring of mothers that were socially isolated showed place learning deficits but not stimulus-response habit learning on the same task. These cognitive impairments, in the offspring of socially isolated dams, occurred in the absence of maternal elevated stress hormone levels, anxiety, or altered mothering. Some evidence suggested that maternal blood-glucose levels were altered particularly during gestation. Our results provide further support for the idea that learning and memory networks, centered on the amygdala and hippocampus are particularly susceptible to the negative impacts of maternal social isolation and these effects can occur without elevated glucocorticoid levels associated with other forms of prenatal stress.

Figure 1

The control (n = 4) and maternal social isolation stress (MSIS (n = 4)) groups were evaluated on their distanced moved (top panel) and stereotypy time (middle panel) in activity boxes during Baseline (left) and at Lactation Day 6 (LD6) testing (right). Both groups had similar activity and exhibited a decrease in activity on LD6 compared to baseline. The amount of stereotypy time was also similar in both groups and did not change significantly between baseline and LD6. The bottom panel shows the results of the elevated plus maze test. At baseline testing, although both groups spent more time in the closed arms compared to open arms (**p < 0.01; ***p < 0.001), the control group spent significantly more time in the open arms than the MSIS group during baseline testing (*p < 0.05). There was no difference between the groups at LD6 testing in the amount of time they spent in the closed and open arms, as both groups had higher dwell times in the closed arms. Data is presented as ± SEM.

Figure 2

The control (n = 4) and MSIS (n = 4) dams had blood analysed at baseline (prior to pregnancy), G18 (gestational day 18), or LD1 (lactation day 1). (Left) Blood corticosterone levels (ng/ml) did not differ between the control or MSIS dams. There was a significant decrease in corticosterone between baseline to G18 and LD1 for the control (p < 0.05), but not MSIS group. (Right) The MSIS dams had significantly higher blood glucose (mmol/L) level than control dams on G18 and LD1 (*p < 0.05). Both groups had a significant decrease in blood glucose levels from baseline to G18 and LD1. Data is presented as ± SEM.

Figure 3

Maternal care and behaviour were scored in control (n = 4) and MSIS (n = 4) dams during the first 10 min of every hour over a 24-h period on post-natal day 1 (P1). Data is presented as ± SEM. (A) The percent time spent in the core nest was calculated as total time in the nest divided by total assessment time and multiplied by 100. (B) The nursing score percentage consisted of summing the scores (0–3) for each 10-min bin and dividing it by 24 bins. This average was then divided by 30 and multiplied by 100. No difference was observed between the control and MSIS groups in the amount of time they spent in the core nest or their nursing score. (C) A qualitative nursing score was also evaluated using the same nursing score except only the scores from 1 to 3 were used to represent the quality of the nursing given. The results revealed that the MSIS dams displayed significantly better nursing than the control dams (*p < 0.05). Other measures scored were the amount of time spent (D) self-grooming and (E) tail-chasing. No group differences were found on these measures. (F) A pup retrieval test was performed on P2 as an indication of the dam’s responsiveness to her pups. No significant differences were found between the groups. The higher mean time displayed by the control group resulted from a dam not retrieving any pups during the 5-min test.

Figure 4

Maternal care and behaviour were assessed in control (n = 4) and MSIS (n = 4) dams during the first 10 min of every hour over a 24-h period on post-natal day 3 (P3). Data is presented as ± SEM. The measures evaluated were (A) Percent time spent in core nest, (B) Nursing score percentage, (C) Qualitative nursing score, (D) Self-grooming, and (E) Tail-chasing. There were no differences between the groups on any measure of maternal care nor other behaviours scored.

Figure 5

The offspring of the control (control-O (n = 12)) and maternal social isolation stress (MSIS-O (n = 12)) groups were evaluated on their (A) distanced moved, and (B) stereotypy time in activity boxes. Both groups had similar activity and stereotypy time. (C) Shows the results of the elevated plus maze test. Both groups spent more time in the closed arms compared to open arms (***p < 0.001). Data is presented as ± SEM.

Figure 6

The control-O (n = 5) and MSIS-O (n = 6) had blood extracted for analysis. The results presented here represent the female littermates of the males used in the current study. (left) Blood corticosterone levels (ng/ml) did not differ between the control-O or MSIS-O groups. (right) Blood glucose levels (mmol/L) also did not differ between the groups. Data is presented as ± SEM.

Figure 7

A depiction of the different phases of the discriminative fear conditioning to context task. (A) Pre-Exposure. The two contexts differed on multiple dimensions including shape, brightness, and smell. The chambers were connected by an alley which allowed the animals to freely explore both contexts for 10 min during pre-exposure. (B) Paired Training. Foot-shocks were administered within the paired context. (C) Unpaired training. No foot-shocks were presented in the unpaired context. The paired and unpaired context was counterbalanced. Over 8 training days, rats were exposed to the paired and unpaired contexts on alternating days. (D) Freezing Testing. Two assessment sessions occurred, one in the paired and the other in the unpaired context during which time spent freezing was recorded. (E) Preference Testing. The connecting alley was opened which allowed the subjects to freely move between contexts [Reprinted from Trow JE, Jones AM, McDonald RJ. (2019) Comparison of the effects of repeated exposures to predictable or unpredictable stress on the behavioural expression of fear in a discriminative fear conditioning to context task. Physiol Behav. 208, 112556].

Figure 8

The control-O (n = 9) and MSIS-O (n = 9) were trained on the discriminative fear conditioning to context task. Data is presented as ± SEM. (A) Pre-exposure to the contexts indicated no initial difference in the dwell time in each context. (B) Following 8 days of fear conditioning the rats were given tests in which the amount of time spent freezing was scored. The results revealed that the control-O group froze more in the paired vs. unpaired context (***p < 0.001), whereas the MSIS-O group did not show this discrimination (p = 0.25). The MSIS-O group also displayed less freezing than the control-O group in the paired context (**p < 0.01). (C) The preference test showed that the control-O group learnt the discrimination as evidenced by more time spent in the unpaired context vs. paired context (***p < 0.001), but that the MSIS-O group did not display a difference in dwell time between the contexts (p = 0.34).

Figure 9

A depiction of the cue-place water task used in the present experiments. Briefly, rats were trained for three days to swim from one of four cardinal start points to a visible platform located in the same place in the pool. On the fourth day, an invisible platform is put in place of the visible platform. This sequence is repeated thrice for a total of 12 training days in which each subject receives nine visible platform and three invisible training days [Reprinted from Gruber AJ, McDonald RJ. (2012) Context, emotion and the strategic pursuit of goals: interactions among multiple brain systems controlling motivated behaviour. Front. Neurosci, 6, 50].

Figure 10

The control-O (n = 12) and MSIS-O (n = 12) groups were trained on the cue-place water task in which they were trained to a visible platform for three days followed by an invisible platform located in the same location, repeated thrice. The closed markers represent the visible training days, and the open markers represent the invisible days. Data is presented as ± SEM. Post-hoc analysis revealed that the control-O group performed better than the MSIS-O group on Day 8 and Day 12 at locating the invisible platform (*p < 0.04, *p < 0.05, respectively). No differences were obtained on the visible platform training days as both groups quickly learned to swim to the platform.