Sex differences in resilience at puberty depend upon divergent effects of a stress steroid at α4βδ GABAA receptors

Abstract

Resilience is a critical skill that lessens the adverse effects of stress which are increasingly reported in adolescents, where sex differences are noted. It is not known how this process develops in adolescence when unique changes in neuronal properties occur. For this study, pubertal mice were tested for their coping ability following 2-week restraint. We show here that this predictable stress produced resilience in pubertal female mice where time immobile in the forced swim test (FST) decreased by ∼50 % (P < 0.0001), an effect that extended into adulthood, and increased escape behavior 8-fold (P = 0.01). This effect was not seen in pubertal male or adult female mice. This process required the stress steroid 3α-OH,5α-pregnan-20-one (THP, allopregnanolone) and its primary target, α4βδ GABA_A receptors (GABARs). These receptors emerge at puberty in prelimbic prefrontal cortex (PL PFC) and basolateral amygdala (BLA), which play a pivotal role in resilient behavior. Stress-induced release of THP decreased anxiety-like behavior (increasing open arm time in the elevated plus maze) and enhanced PFC-dependent learning (temporal order recognition) after 1d restraint in pubertal female, but not male, mice while differentially altering α4βδ expression in PL and BLA. This divergent THP-induced effect ultimately increased and decreased mushroom spine density in PL and BLA, respectively, to produce a circuit optimized for resilient behavior in the pubertal females. These findings demonstrate a novel mechanism for the development of resilience unique to the pubertal period. The results from the present study may suggest therapeutic strategies for adolescent stress which would impact mental health in adulthood.

Keywords: Adolescence; Allopregnanolone; Alpha-4; GABA(A) receptor; Prefrontal cortex; Resilience; Stress.

Fig. 1.. Chronic predictable stress during adolescence generates resilience only in pubertal female mice.

Inset, Timeline for restraint – 14 d beginning at puberty (lower bar) or adulthood (upper bar) and testing (1–2 d later or in adulthood). PND, post-natal day; Restr, restraint; Con, control. In this and the following figures, data are represented as diamonds (pubertal females), up triangles (pubertal males) or down triangles (adult females). A, Average time immobile (% of total) on the forced swim test (FST) following pubertal restraint stress for female (left) and male (right) mice. F(3,71) = 18.83, P < 0.0001, sex, F(1,71) = 8.92, P = 0.0039, stress, F(1,71) = 35.4, P < 0.0001, sex x stress interaction, F (1,71) = 7.43, P = 0.0081: *P < 0.0001 versus Con, N = 16–21 mice/group. B, Time immobile for female mice tested as adults, following restraint at puberty (left) or restraint as adults (right). *t(34) = 4.54, P = 0.00007 vs. Con, N = 17–19 mice/group. C, Distance scaled (% of total) on the escape device (plastic straw) presented after the 6 min FST for pubertal female (left), pubertal male (center) and adult female (right) mice after restraint versus control. *Mann Whitney U test: U; 31, Z: 2.41, *P: 0.016. N = 5–7 mice/group. D, Open arm time (% of total) on the elevated plus maze (EPM) for female and male mice after pubertal restraint. F(3,72) = 3.4, P = 0.02; sex, F(1,72) = 4.37, P = 0.04. There were no significant interactions, q value = 0.53, P = 0.98, female con vs. restraint; q value = 3.34, P = 0.09, male con vs. restraint. N = 18–21 mice/group. E, Open arm time, EPM for adults restrained as adults. N = 14 mice/group.

Fig. 2.. Blockade of THP formation and knock-down of α4 prevent the development of resilience in female mice at puberty.

Inset, Timeline for restraint and injections – 14 d beginning at puberty or adulthood and testing (1–2 d later). Upper bar, WT mice: restraint + FIN or VEH; lower bar, α4 knock-out (KO) mice: restraint or no restraint control. (FIN, finasteride; VEH, vehicle. A-C, Time immobile, forced swim test (FST), restraint plus FIN or VEH (left) or after α4 knock-out (KO, right) in pubertal female (A), pubertal male (B) or adult female (C) mice. Pubertal female, FIN, *t(28) = 4.98, P < 0.0001 vs. VEH, N = 15 mice/group (A); adult female, *t(16) = 2.13, P = 0.049 (C). D-F, Open arm time, elevated plus maze (EPM), restraint plus FIN or VEH (left) or after α4 knock-out (KO, right) in pubertal female (D), pubertal male (E) or adult female (F) mice. Pubertal female, FIN, *t(28) = 2.52, P < 0.018 vs. VEH, N = 15 mice/group (D); KO, *t(28) = 3.24, P = 0.0031, N = 15 mice/group (D).

Fig. 3.. Restraint stress-induced release of THP reduces open arm time and improves PFC-dependent learning only in pubertal female mice

Inset, Timeline, Upper bar (A, B, E), day 1: 2 h restraint stress (or no restraint control), preceded by finasteride (FIN) or vehicle (VEH) injections; day 2: FIN or VEH followed by testing (EPM (A, B) or temporal order recognition (TOR) task (E)) 2 h later. Lower bar (C, D), day 1: 2 h restraint stress (or no restraint control); day 2: TOR task. A, Open arm time on the EPM after 1 d restraint plus FIN or VEH for pubertal female (left) and male (right) mice. F(3,35) = 10.79, P < 0.0001, sex, F(1,35) = 7.26, P = 0.011, drug, F(1,35) = 7.71, P = 0.009, sex x drug interaction, *F(1,35) = 19.8, P < 0.0001: female, q value = 6.73, *P = 0.00024 versus VEH, N = 8–10 mice/group. B, Open arm time on the EPM after FIN or VEH without restraint for pubertal female (left) and male (right) mice. No comparisons were significant. N = 8–10 mice/ group. C, D, Discrimination ratio on the TOR task after 1 d restraint for female (C) and male (D) mice. Female, *t(16) = 2.94, P = 0.01 vs. Con. N = 10 mice/group (B). E, Discrimination ratio on the TOR task after 1 d restraint plus FIN or VEH (left) or FIN/VEH without restraint (right) to female mice, Restraint, FIN vs. VEH, *t(16) = 7.73, P < 0.0001, N = 8–10 mice/group.

Fig. 4.. Restraint stress alters expression of the α4 GABAR subunit in PL and BLA of pubertal female mice.

Inset, Timeline of IHC assessment of α4 expression after 2, 7 or 14 D of restraint. A, PL, Averaged data, fluorescence intensity of α4 immunostaining as a percent of control, 2D, *t(8) = 2.44, P = 0.04, 7D, *t (8) = 2.58, P = 0.03. B, Representative images, PL, Con (top) and Restr (bottom), 2D restraint. Boxes, soma magnified in C, 2D. C, Representative soma, 2, 7 and 14 D restraint. D, BLA, Averaged data, Fluorescence intensity as a percent of control, 2D, *t(8) = 2.4, P = 0.04, 7D, *t(10) = 2.23, P = 0.04. E, Representative images, BLA, Con (top) and Restr (bottom), 2D restraint. Boxes, soma magnified in F, 2D. F, Representative soma, 2, 7 and 14 D restraint. N = 5–6 mice/group, 3–4 neurons/mouse.

Fig. 5.. Blockade of THP formation prevents the changes in α4 expression produced by 2 d of restraint stress of pubertal female mice.

Inset, timeline of vehicle (VEH) or finasteride (FIN) injection and α4 assessment. A, PL, Averaged data, fluorescence intensity of α4 immunostaining as a percent of control, *t(8) = 2.45, P = 0.04. B, Representative images, PL, Con (top) and Restr (bottom), Boxes, soma magnified in I. C, Representative soma. N = 5 mice, 4–6 neurons/mouse. D, BLA, Averaged data, fluorescence intensity as a percent of control, *t(8) = 3.7, P = 0.006. E, Representative images, BLA, Con (top) and Restr (bottom). Boxes, soma magnified in L. F, Representative soma. N = 5 mice/group, 3–8 neurons/mouse.

Fig. 6.. Restraint stress during adolescence alters mushroom spine density in the PL and BLA of pubertal female mice.

A, B, Averaged densities for thin, mushroom, stubby and total spines/10 μm in the PL (A) and BLA (B) of control or restrained (2 weeks) female mice. A, PL, Averaged data, mushroom spines, *t(65) = 4.6, P < 0.0001 vs. Con. B, BLA, Averaged data, mushroom spines, *t(65) = 3.99, P = 0.0018 vs. Con; total spines, *t(65) = 2.29, P = 0.04 vs. Con. C-D, Representative images for PL (C) and BLA (D). N = 7 mice, 4–5 dendrites/mouse/group.

Fig. 7.. Blockade of THP formation prevents the changes in mushroom spine density produced by restraint stress in pubertal female mice.

A, B, Averaged densities for thin, mushroom, stubby and total spines/10 μm in the PL (A) or BLA (B) of female mice injected with finasteride (FIN) or vehicle (VEH) during the 2-week restraint. A, PL, Averaged data, mushroom spines, *t(48) = 4.7, P < 0.0001 vs. Con. B, BLA, Averaged data, mushroom spines, *t(42) = 12.58, P < 0.0001 vs. Con; total spines, *t(42) = 3.71, P = 0.0006 vs. Con. C,D, Representative images for PL (C) and BLA (D). N = 5 mice, 4–7 dendrites/mouse/group.

Fig. 8.. Restraint stress does not alter dendritic spine density of pubertal male mice.

A, B, Averaged densities for thin, mushroom, stubby and total spines/10 μm in the PL (A) and BLA (B) for control or restrained (2 weeks) male mice. C, D, Representative images for PL (C) and BLA (D). N = 5 mice, 3–4 dendrites/mouse/group.