Unstable maternal environment, separation anxiety, and heightened CO2 sensitivity induced by gene-by-environment interplay

Abstract

Background: In man, many different events implying childhood separation from caregivers/unstable parental environment are associated with heightened risk for panic disorder in adulthood. Twin data show that the occurrence of such events in childhood contributes to explaining the covariation between separation anxiety disorder, panic, and the related psychobiological trait of CO(2) hypersensitivity. We hypothesized that early interference with infant-mother interaction could moderate the interspecific trait of response to CO(2) through genetic control of sensitivity to the environment.

Methodology: Having spent the first 24 hours after birth with their biological mother, outbred NMRI mice were cross-fostered to adoptive mothers for the following 4 post-natal days. They were successively compared to normally-reared individuals for: number of ultrasonic vocalizations during isolation, respiratory physiology responses to normal air (20%O(2)), CO(2)-enriched air (6% CO(2)), hypoxic air (10%O(2)), and avoidance of CO(2)-enriched environments.

Results: Cross-fostered pups showed significantly more ultrasonic vocalizations, more pronounced hyperventilatory responses (larger tidal volume and minute volume increments) to CO(2)-enriched air and heightened aversion towards CO(2)-enriched environments, than normally-reared individuals. Enhanced tidal volume increment response to 6%CO(2) was present at 16-20, and 75-90 postnatal days, implying the trait's stability. Quantitative genetic analyses of unrelated individuals, sibs and half-sibs, showed that the genetic variance for tidal volume increment during 6%CO(2) breathing was significantly higher (Bartlett χ = 8.3, p = 0.004) among the cross-fostered than the normally-reared individuals, yielding heritability of 0.37 and 0.21 respectively. These results support a stress-diathesis model whereby the genetic influences underlying the response to 6%CO(2) increase their contribution in the presence of an environmental adversity. Maternal grooming/licking behaviour, and corticosterone basal levels were similar among cross-fostered and normally-reared individuals.

Conclusions: A mechanism of gene-by-environment interplay connects this form of early perturbation of infant-mother interaction, heightened CO(2) sensitivity and anxiety. Some non-inferential physiological measurements can enhance animal models of human neurodevelopmental anxiety disorders.

Conflict of interest statement

Competing Interests: The authors have declared that no competing interests exist.

Figures

Figure 1. Scheme of the respiratory protocol.

During ‘baseline’ and ‘recovery’ periods, subjects inhaled normal air. During ‘challenge’ periods subjects were exposed to one type of air mixture: 6% CO2-enriched air, or 10%O2 air, or normal air.

Figure 2. Maternal care and offspring behavioural indices in standard rearing (CT) vs. repeated cross-fostering (RCF) conditions.

Sum of 2 daily observations of maternal behaviour: A) nursing behaviour, encompassing ‘arch-back’+‘blanket’ postures, and B) grooming/licking (GP/L) behaviour towards adoptive (RCF n = 10) and own (CT n = 8) litter, measured during PND1–PND7. Nursing decreased significantly in time, and was comparable in RCF and CT pups across the PND1–PND7 time span. Grooming/licking (GP/L) did not vary significantly in time, and RCF and CT pups received comparable GP/L (see Methods and Results sections for details). Pups' behaviour: C) Mean number of ultrasonic vocalisations (USVs) emitted by 8-day old RCF and CT pups. Pups were isolated and exposed for 5′ to fresh clean bedding (clean) and own-cage bedding (own). ANOVA showed that the postnatal treatment (RCF vs. CT) yielded a significant effect (F1,73 = 4.24, p = 0.04) while the condition (‘clean’ vs. ‘own’ bedding) did not exert a significant effect (F1,73 = 0.84, p = NS); there was no significant postnatal treatment-by-condition effect (F1,71 = 1.29, p = NS). D) Percentage of time spent during 5 minutes by pups in a compartment containing own-cage vs. fresh clean bedding (own vs. clean), or own-cage vs. an alien dam's bedding (own vs. alien dam). RCF pups spent less time in the compartment with own-cage bedding than controls in both conditions (F1,64 = 7.46, p<0.01).

Figure 3. Behavioural and Endocrinological Phenotypes among RCF and CT subjects.

A: Serum concentration of corticosterone. In both RCF and CT subjects corticosterone was significantly heightened (F1,9 = 49.71 p = 0.0001) after 20 minutes of 6% CO2 breathing compared to the basal values obtained in room air breathing; neither ‘postnatal treatment’ (RCF vs. CT) nor ‘postnatal treatment-by-air mixtures’ revealed differences for corticosterone serum concentrations (respectively F1,9 = 0.2 p = NS, and F1,7 = 0.08, p = NS). B: Place avoidance/preference towards environments with heightened CO2 concentration (6% CO2 air mixture): during the first five minutes of the test session, RCF individuals showed significantly higher tendency to avoid the compartment that had been previously paired with 6% CO2. (ANOVA-R : postnatal manipulation x time interval effect: F1,22 = 4.51, p<0.05, Tukey HSD post-hoc test p<0.02). C: Free exploratory test. The percentage of time spent at the centre of an arena was significantly influenced by the interaction of postnatal treatment-by-air mixtures (F1,74 = 4.03, p = 0.048) whereby RCF subjects spent significantly less time than CT subjects after exposure to 20% CO2. Neither the ‘postnatal treatment’ (RCF/CT), nor the ‘air mixtures’ (normal air/20% CO2) variables showed significant effects alone (respectively F1,76 = 1.90 p = NS, F1,76 = 0.13 p = NS). D: Latency to fall from the Rotarod. Analysis of variance showed that the performance at the Rotarod was significantly worse among RCF than CT subjects (F1,67 = 5.08 p = 0.03), in that the former showed significantly shorter latency of fall from the rod.

Figure 4. Respiratory responses to air, 10%O2, or 6%CO2 in CT and RCF subjects at different ages.

Percentage of tidal volume changes from baseline (ΔTV%) for: a) 16–20 day-old pups in response to normal air, 10% O2, or 6% CO2. The ANOVA-R carried out on two consecutive respiratory challenges (as depicted in Figure 1) indicated a significant effect of: 1) treatment (type of air mixture): F2,58 = 91.30, p = 0.000001, 2) time: F1,58 = 4.34, p<0.05, and 3) an interaction effect of postnatal manipulation-by-type of air mixture: F2,58 = 9.99, p<0.0002); Tukey HSD post-hoc test p<0.001; b) 75–90 day-old adult mice in response to normal air or 6% CO2. The ANOVA-R carried out on two consecutive respiratory challenges (as depicted in Figure 1) indicated a significant effect of: 1) treatment (type of air mixture): F1,71 = 184.83, p = 0.00001, 2) time: F1,72 = 35.12, p = 0.00001 and 3) an interaction effect of postnatal manipulation-by-type of air mixture F1,71 = 6.60, p = 0.012). Tukey HSD post-hoc test p<0.001. Sample sizes varied between 9 and 13 animals per group. Only the responses to the first of two consecutive challenges performed for each subject with the same air mixture (air/10% O2/6% CO2) are shown in Figure 4 for the sake of conciseness.