Nongenomic effects of estradiol on aggression under short day photoperiods

Abstract

In several vertebrate species, the effects of estrogens on male aggressive behavior can be modulated by environmental cues. In song sparrows and rodents, estrogens modulate aggression in the nonbreeding season or winter-like short days, respectively. The behavioral effects of estrogens are rapid, which generally is considered indicative of nongenomic processes. The current study further examined the hypothesis that estradiol acts nongenomically under short days by utilizing a protein synthesis inhibitor, cycloheximide (CX). Mice were housed in either short or long day photoperiods, and treated with an aromatase inhibitor. One hour before resident-intruder testing mice were injected with either CX or saline vehicle, and 30 min later were treated orally with either cyclodextrin conjugated estradiol or vehicle. Under short days, mice treated with estradiol showed a rapid decrease in aggressive behavior, independent of CX administration. CX alone had no effect on aggression. These results show that protein synthesis is not required for the rapid effects of estradiol on aggression, strongly suggesting that these effects are mediated by nongenomic processes. We also showed that estradiol suppressed c-fos immunoreactivity in the caudal bed nucleus of the stria terminalis under short days. No effects of estradiol on behavior or c-fos expression were observed in mice housed under long days. Previously we had also demonstrated that cage bedding influenced the directional effects of estrogens on aggression. Here, we show that the phenomenon of rapid action of estradiol on aggression under short days is a robust result that generalizes to different bedding conditions.

Keywords: Aggression; Estradiol; Nongenomic; Photoperiod.

Figure 1

Plasma estradiol levels (A) 15 min following a 100 ug/kg injection of estradiol (n = 8) or 100 uL saline (n = 4) and (B) 15 and 30 minutes following oral administration of estradiol. Doses of oral administration were 100 ug/kg (15 min: n = 3; 30 min: n = 3), 500 ug/kg (15 min: n = 4; 30 min: n = 3), and 1000 ug/kg (15 min: n = 4; 30 min: n = 3). The dotted line represents plasma levels of estradiol following a 100 ug/kg injection of estradiol. Estradiol levels 30 min following a 500 ug/kg oral dose of estradiol are comparable to plasma levels 15 min following a 100 ug/kg s.c. injection.

Figure 2

Photomicrographs indicating AVP/cfosco localizations in the PVN at low (A) and high (B) magnification. The red nuclear staining indicates c-fos activity, while the green cytoplasmic staining indicates AVP. 3V = third ventricle. Bar = 100 um.

Figure 3

The number of c-fos-ir cells in the cBNST following exposure to clean bedding (control) or soiled male bedding (vehicle and drug conditions) following CX administration. Mice received either 3% ethanol in saline (control and vehicle), 2.5 mg/kg CX, 25 mg/kg CX, or 50 mg/kg CX. n = 3 for all groups. Bars with different letters are significantly different according to LSD post hoc tests (p < 0.05).

Figure 4

Latency to attack in (A) short days and (B) long days, and number of bites given in (C) short days and (D) long days following a resident-intruder test. Black bars represent data from experiments 2a and 3a testing the combined effect of estradiol and CX on aggressive behavior. White bars represent data from experiments 2b and 3b testing the effects of CX alone on aggressive behavior. On short days, estradiol increased attack latency (A) and decreased bite frequencey (C), and this effect was not blocked by CX. In contrast estradiol had no effect on aggression in long days (B & D). CX treatment alone had no effect on aggression whether mice were housed on short days (A& C) or long days (B & D). *planned comparison, p < 0.05. n = 8–11 per group.

Figure 5

Photomicrographs from short day mice of c-fos immunoreactivity in the rBNST (A–C), cBNST (D–F) and PVN (G–I) over three treatment conditions: saline (A, D & G), estradiol (B, E & H) and estradiol + CX (C, F & I). Mice treated with saline had significantly more c-fos-ir cells in the cBNST than mice receiving estradiol or estradiol + CX. rBNST and PVN cell counts were not significantly different between treatments. sm = stria medullaris, f = fornix, 3V = third ventricle. Bar = 100 um.

Figure 6

Correlations between the number of bites and AVP-ir cells in the cBNST (A & B) and latency to attack and AVP-ir cells in the PVN (C & D). In mice housed in long days, there was a significant positive correlation between the number of bites and AVP-ir cells in the cBNST. In mice housed in short days, there was a significant positive correlation between latency to attack and AVP-ir cells in the PVN.