Effects of neural estrogen receptor beta deletion on social and mood-related behaviors and underlying mechanisms in male mice

Abstract

Estradiol derived from neural aromatization of testosterone plays a key role in the organization and activation of neural structures underlying male behaviors. This study evaluated the contribution of the estrogen receptor (ER) β in estradiol-induced modulation of social and mood-related behaviors by using mice lacking the ERβ gene in the nervous system. Mutant males exhibited reduced social interaction with same-sex congeners and impaired aggressive behavior. They also displayed increased locomotor activity, and reduced or unaffected anxiety-state level in three paradigms. However, when mice were exposed to unescapable stress in the forced swim and tail suspension tests, they spent more time immobile and a reduced time in swimming and climbing. These behavioral alterations were associated with unaffected circadian and restraint stress-induced corticosterone levels, and unchanged number of tryptophan hydroxylase 2-immunoreactive neurons in the dorsal raphe. By contrast, reduced mRNA levels of oxytocin and arginine-vasopressin were observed in the bed nucleus of stria terminalis, whereas no changes were detected in the hypothalamic paraventricular nucleus. The neural ERβ is thus involved to different extent levels in social and mood-related behaviors, with a particular action on oxytocin and arginine-vasopressin signaling pathways of the bed nucleus of stria terminalis, yet the involvement of other brain areas cannot be excluded.

Figure 1

Sociability and social interaction in control (ERβ^fl/fl) and mutant (ERβ^NesCre) male mice. (A–C) Sociability: Time spent in the empty and stimulus compartments of the three-chamber test (A), number of entries and interactions in the stimulus compartment (B), and time spent in olfactory investigation of the stimulus (C). Data are expressed as means ± S.E.M (n = 12–15 males per genotype). ^bp < 0.0001: effect of compartment with males more interested in the stimulus compartment. (D) Social interaction: Time spent in total, anogenital and nose-to-nose interaction. ^*p < 0.05 versus the control group.

Figure 2

Effects of neural ERβ knockout on locomotor activity, anxiety- and despair-like behaviors in male mice. (A) Left – Locomotor activity per 20 min recorded for males (n = 12–15 males per genotype). ^ap < 0.001: effect of time; ^bp < 0.05: effect of genotype with mutants more active than their control littermates. Right – Total activity measured during the 140 min of the test. ^*p < 0.05 versus control littermates. (B-C) Left – Number of entries in the open arms of the O-maze (B), or light compartment of the dark-light box (C). Right – Time spent in the open arms of the O-maze (B), or light compartment of the dark-light box (C). (D) Left – Time spent in immobility over two consecutive days in the forced swim test (n = 12–15 males per genotype). Two-ANOVA shows an effect of time (^ap < 0.001); post hoc analyses show increased time in immobility for mutants at day 1 (^ap < 0.001 versus control) and day 2 (^**p < 0.01 versus control). Middle – Time spent in climbing. Two-ANOVA shows an effect of time (^ap < 0.05); post hoc analyses show reduced time in climbing for mutants at day 1 (^*p < 0. 05 versus control). Right – Time spent in swimming. Two-way ANOVA shows an effect of time (^ap < 0. 0.001); post hoc analyses show reduced time in swimming for mutants at days 1 and 2 (^*p < 0. 05 versus control).

Figure 3

Effects of neural ERβ knockout on sucrose preference in male mice. Control and mutant males (n = 11 per genotype) were subjected to sucrose preference test with the habituation phase (A) and testing sessions (B). (A) Water consumption over two consecutive days 1 and 2. Two-way ANOVA shows an effect of location (^ap < 0. 05) at day 1 but not at day 2. (B) Liquid consumption with the bottle containing 2% sucrose either at the right or left position. Two-way ANOVA shows an effect of solution for the two tests (^ap < 0.001), with a preference of sucrose regardless of the genotype.

Figure 4

Aggressive behavior in control and mutant males. (A) Percentage of males (n = 9 males per genotype) exhibiting aggressive behavior towards intruder males over four consecutive days. ^*p < 0.05 versus controls. (B) Latency to the first attack at Test day; ^**p < 0.01 versus controls. (C) Total duration of aggressive behavior at Test day; ^**p < 0.01 versus controls.

Figure 5

Corticosterone levels in control and mutant male mice. Corticosterone levels were measured in plasma samples collected in the afternoon and the following morning in control and mutant mice (n = 8–9 per genotype). They were then subjected to restraint stress and plasma was collected and measured for hormones 20 min and 150 min after the stress. ^ap < 0. 001 versus afternoon levels; ^bp < 0.001 versus morning levels.

Figure 6

Anxiety- and despair-like behaviors in females, and TPH2 immunoreactivity in males and females. (A) Time spent in the light compartment of the dark-light box by control and mutant females at the proestrous/estrous (PE-E) or diestrous (D1-D2) stage (n = 18–29 females per genotype). Two-way ANOVA shows an effect of cycle (^ap < 0.01); posthoc analyses indicate a reduced time spent in the light compartment for mutant females at the P-E stage (p < 0.01 versus controls at the same stage). (B) Time spent in immobility in the forced swim test for females at the P-E or D1-D2) stage. Two-way ANOVA shows an effect of cycle (^ap < 0.05); posthoc an^alyses indicate an increased time spent in immobility for mutant females at the P-E stage (p < 0.01 versus controls at the same stage). (C) Time spent swimming in the forced swim test for females at the P-E or D1-D2) stage. Two-way ANOVA shows an effect of cycle (^ap < 0.05); posthoc analyses indicate a reduced time spent in swimming for mutant females at the P-E stage (p < 0.01 versus controls at the same stage). (D,E) Number of TPH2 neurons counted in the dorsal raphe of females (D) and males (E). Data are means ± S.E.M. of 6–8 animals per sex and per genotype; ^**p < 0.01 versus controls.

Figure 7

Effects of neural ERβ deletion on OT and AVP expression in male mice. (A) Levels of AVP (Left) and OT mRNAs (Right) normalized to GAPDH expression in the bed nucleus of stria terminalis of males (n = 9 animals per genotype). ^*p < 0.05 or ^**p < 0.01 versus controls. (B) Levels of AVP (Left) and OT (Right) expression normalized to GAPDH in the hypothalamic paraventricular nucleus.