The androgen receptor is selectively involved in organization of sexually dimorphic social behaviors in mice

Abstract

It is well established that sexually dimorphic neural regions are organized by steroid hormones during development. In many species, neonatal males are exposed to more testosterone than their female littermates, and ultimately it is the estradiol, produced by aromatization of testosterone, that affects sexual differentiation. However, the androgen receptor also plays an important role in the masculinization of brain and behavior. Here we tested the hypothesis that sexually dimorphic social and odor preference behaviors can be differentiated by a nonaromatizable androgen during development by treating female mice on the day of birth (PN0) with dihydrotestosterone (DHT). Control mice received a single vehicle injection on PN0. Adults were gonadectomized, treated with estradiol, and tested for social behaviors. In contrast with control females, females treated on PN0 with DHT, like male controls, exhibited a preference for female-soiled vs. male-soiled bedding, a preference to investigate a female vs. a male and reduced c-Fos-immunoreactivity (ir) in several neural areas after exposure to male-soiled bedding. However, females treated with DHT on PN0 had normal female-typical sexual behavior. The number of calbindin-ir cells in the preoptic area is sexually dimorphic (males more than females), but females given DHT on PN0 had intermediate numbers of calbindin-ir neurons, not significantly different from control males or females. Our data demonstrate that organization of social and olfactory preferences in mice can be affected by perinatal DHT and lends support to the role of androgen receptor in organization of sexual differentiation of brain and behaviors.

Figure 1

A, Mean (±sem) time spent sniffing soiled bedding from hormone-primed females minus time spent sniffing male-soiled bedding. Groups included mice treated on PN0 with vehicle, males (n = 10), and females (n = 8), and females treated with DHT (n = 12). At the time of testing, all animals were gonadectomized and treated with an E2 implant to provide equivalent levels of activational hormone. B, Mean (±sem) time in seconds spent investigating the Y maze arm containing a hormone-primed stimulus female, minus time spent in the arm containing a stimulus male. Groups contained mice treated on PN0 with either vehicle (males, n =10; females, n =11) or females treated with DHT (n = 13). At the time of testing, all animals were gonadectomized and treated with an E2 implant to provide equivalent levels of activational hormone. *, Significantly different from both other groups (P < 0.05).

Figure 2

Mean (±sem) LQ exhibited by female mice treated on PN0 with either DHT (•, n = 11) or vehicle (▿, n = 9). Males given vehicle on PN0 were also tested (▪, n = 6). Adults were gonadectomized and after priming with EB and progesterone given a series of six sexual behavior tests. *, Significantly lower lordosis was exhibited by males on trials 2–6, as compared with both groups of females (P < 0.05).

Figure 3

Mean (±sem) time spent sniffing soiled bedding from hormone-primed females minus time spent sniffing male-soiled bedding. Groups were composed of males treated on PN0–2 with vehicle (n = 10), females that received vehicle (n = 7), and females that received E2 on PN0–2 (n = 10). As adults, animals were gonadectomized and treated with an E2 implant to provide equivalent levels of activational hormone. *, Significantly different from the other two groups (P < 0.05).

Figure 4

Mean (±sem) number of Fos-ir cells present in the mPOA (A) and the BNST (B) after exposure to either clean (black histograms) or male-soiled bedding (gray histograms). Males (n = 6 in clean bedding and n = 10 in male-soiled bedding) and females (n = 8 in clean bedding and n = 10 in male-soiled bedding) were treated with vehicle on PN0. Another group of females was treated on PN0 with DHT (n = 6 in clean bedding and n = 8 in male-soiled bedding). All mice were gonadectomized and treated with an E2 implant as adults to provide equivalent levels of activational hormone. *, Significantly different from PN0 vehicle-treated females exposed to clean bedding (P ≤ 0.05).

Figure 5

Mean (±sem) number of Fos-ir cells present in the mPOA (A) and BNST (B) after exposure to either clean (black histograms) or male-soiled (gray histograms) bedding. Mice were treated on PN0–2 with either E2 or vehicle. The control group (black histograms) exposed to clean bedding had subjects from all three neonatal/sex conditions (n = 10). Adults exposed to male-soiled bedding included three groups: males treated with neonatal vehicle (n = 6), females similarly treated with vehicle (n = 5), and females given E2 from PN0–2 (n = 6). At the time of testing, all adults were gonadectomized and treated with an E2 implant to provide equivalent levels of activational hormone. *, Significantly different from the controls exposed to clean bedding (P < 0.05).

Figure 6

A–C, Representative photomicrographs of calbindin D28K-ir cells in adult mouse brains. These cells are observed in the boundary between the anterior preoptic area and the anterior hypothalamus in each experimental group. A, A male treated on PN0 with vehicle. B, An identically treated female. C, The calbindin-ir in a female that received DHT on PN0. Scale bar, 100 μm. D, The mean (±sem) numbers of calbindin-ir cells present in the mPOA. In this study we used brains from five control males, nine brains from vehicle-treated control females, and 10 females treated with DHT on PN0. At the time the animals were killed, all adults were gonadectomized and had been without hormone treatment for at least 2 wk. *, Significantly different from the vehicle-treated female group (P < 0.05).