Rapid Effects of Steroid Hormones on Animal Behavior

A widely used method for determining whether steroid hormones, like testosterone, affect aggression is to conduct long-term manipulations. Manipulations, such as castration and/or applying hormone implants, are effective approaches to determining whether steroid hormones regulate behavior. When students are taught about the mechanisms of steroid hormones, they usually learn about how steroid hormone receptors act as transcription factors, regulating the expression of genes within the nucleus of a cell (Figure 1). This is a process that occurs over hours or days, and the scientific community's focus on these mechanisms began in the 1960s with the discovery of steroid receptors in the nucleus (Pfaff & Levine 2008). This change in focus shifted attention away from more rapid mechanisms of steroid action for several decades. Rapid mechanisms can occur within seconds or minutes of a change in steroid concentrations, and are generally considered to be independent of receptors acting in the nucleus (Figure 1). However, there has been a revival of interest in rapid actions of steroids on behavior (Vasudevan & Pfaff 2006). As outlined below, recent discoveries have shown that steroid hormones can influence a wide variety of behaviors within minutes, presumably through these so-called nongenomic mechanisms. Detecting rapid effects of steroid hormones requires different approaches to complement the widely used experiments using castration and home implant manipulations.

Glucocorticoids (cortisol and corticosterone) are secreted by vertebrates in response to unpredictable and noxious stimuli. One function of glucocorticoids is to mobilize energy, which is typically diverted away from processes such as reproduction into processes required for more immediate survival (Sapolsky et al. 2000). For example, snowshoe hares (Lepus americanus) have increased corticosterone levels during periods of high predation by the lynx (Lynx canadensis), which is associated with fewer offspring (Boonstra et al. 1998). These hormones are frequently measured as an index of relative condition or health of individuals and populations. However, recent evidence suggests that the relationship between glucocorticoids and estimates of fitness can vary not only across populations but also over time within a single population (Bonier et al. 2009). This may be because an acute elevation of glucocorticoids can have different effects on behavior and physiology than a chronic elevation of glucocorticoids. A comprehensive review of the effects of glucocorticoids on behavior is beyond the scope of this paper (see McEwen and Wingfield 2003). Here we will consider how glucocorticoids can rapidly affect behavior.

A seminal study in Frank Moore's lab showed that acute exposure to either endogenous or exogenous corticosterone exposure caused male rough-skinned newts (Taricha granulosa) to exhibit reduced clasping of females, a courtship behavior that typically lasts hours or days (Moore & Miller 1984). Importantly, the inhibition of this behavior occurs eight minutes after corticosterone exposure (Orchinik et al. 1991). Membrane-bound corticosterone receptors, which do not migrate to the nucleus, were also discovered in the newt brain (Orchinik et al. 1991, 1992). These findings are important because they show that reproductive behavior can be rapidly downregulated following the release of corticosterone. Many mating and courtship behaviors are conspicuous, which increases the risk of predation (Cooper 1999). This may explain why stress-response systems that are activated by predators quickly downregulate reproductive activities in the face of threats. In female sheep, cortisol rapidly renders the pituitary insensitive to gonadotropin releasing hormone released by the hypothalamus (Breen & Karsch 2004). This is important because hormones released by the pituitary facilitate female mating behaviors (Blaustein 2010).

Studies across a variety of species have demonstrated that individuals frequently vary in their behavioral responses to stress (Sih et al. 2004). For example, active responses typically engage a stressful event or stimulus directly (e.g., aggression, escape) whereas reactive responses are more passive (e.g., submissive behavior, vigilance). Active coping responses are usually associated with lower glucocorticoid secretion while passive coping responses are usually associated with higher glucocorticoid secretion (Koolhaas et al. 1999, Carere et al. 2003). Coping responses can be measured using the elevated plus maze (Crawley 2007; Figure 2), both by activity patterns and by recording specific vigilance postures during this test (Figure 3). The elevated plus maze measures the conflict between the natural tendency of mice to explore a novel environment versus the tendency to avoid exposed areas. Mice showing higher exploratory behavior spend more time investigating the open arms whereas mice showing low exploratory behavior spend more time in the closed arms. In a correlational study, individuals showing higher vigilance behavior had higher corticosterone levels. Subsequently it was shown that an injection of corticosterone increased vigilance behavior within two minutes and that this response was not blocked by a protein synthesis inhibitor (Mikics et al. 2005). This is a key result because changes in gene expression must be manifested in changes in protein levels for behavior to be affected. The rapid effects of corticosterone on vigilance behavior must be occurring independent of altered gene expression Thus, corticosterone must be altering brain activity and/or neurotransmitter release, processes that are likely mediated by corticosterone binding to receptors in the cell membrane.

Courtship behaviors often involve dynamic displays and require complex behavioral interactions with potential mates. Many studies have documented that testosterone, acting over several days, promotes male mating behaviors (Valenstein & Young 1955, Beach & Inman 1965). But recent work has highlighted that courtship behaviors that lead to mating are affected by more rapid changes in steroid hormones. Zebra finches form monogamous mating pairs, and males use song to attract females. Parts of the brain controlling bird song production show a rapid increase in estradiol when males are exposed to females (Remage-Healey et al. 2008; Figure 4). Estradiol is a form of estrogen. Other studies have shown that estradiol promotes male courtship song (Walters & Harding 1988), suggesting that a rapid increase in estradiol in the brain could facilitate an increase in courtship behaviors, including song. Estradiol has also been observed to act rapidly to increase male mating behaviors in both quail (Cornil et al. 2006) and rats (Cross & Roselli 1999).

In zebra finches, estradiol is primarily synthesized in the brain. However, steroids from the gonads can also exert rapid effects on behavior. Male goldfish (Carassius auratus) show a rapid rise in plasma testosterone (an androgen hormone) upon encountering ovulating females (Kobayashi et al. 1986). This increase in testosterone is important because other studies demonstrate that an injection of testosterone increases the frequency that males approach females (Lord et al. 2009). Injections of estradiol had similar effects on male behavior. This observation is important because the aromatase enzyme converts testosterone into estradiol. This study also showed that an aromatase inhibitor, which blocks the conversion of testosterone into estradiol, prevented the positive effect of testosterone on male approach behavior. (Lord et al. 2009). Together these data indicated that testosterone released by the testes is converted into estradiol within the brain, which then rapidly alters behavior. Circulating androgens also act rapidly to increase male advertisement vocalizations in the Gulf toadfish (Remage-Healey & Bass 2006). Thus, it appears that androgens released in the presence of females can act rapidly to modulate male courtship behavior, which could presumably have important effects on reproductive success.

In addition to their roles in modulating courtship and reproductive behaviors, estradiol has important effects on aggressive behaviors. Interestingly, estrogen-dependent mechanisms of aggressive behavior appear to be very sensitive to the environment. In many rodents, aggression levels are mediated by photoperiod (Jasnow et al. 2002, Gutzler et al. 2009), a cue that reliably predicts seasonal changes in temperature and food availability. The effects of estradiol on aggression in particular depend on photoperiod. In mice of the genus Peromyscus, estradiol increases male aggression under winter short days but decreases aggression under summer long days (Trainor et al. 2007). These changes do not appear to be due to changes in the number of cells expressing estrogen receptors in the brain. Instead, the differential effects of estradiol on behavior appear to be mediated by changes in how these hormones are acting in the brain. On short days (but not long days), estradiol acts within fifteen minutes to increase aggression (Trainor et al. 2007, 2008; Figure 5). This suggests that changes in gene expression are not responsible for enhancing aggression levels.

Field studies on song sparrows (Figure 6) have demonstrated that estrogens increase male aggression in the winter (Soma et al. 2000b), which is consistent with the laboratory data from Peromyscus. Furthermore, inhibition of estradiol synthesis reduced aggression within 24 hours (Soma et al. 2000a), which is a relatively fast effect for field studies. In the winter, male song sparrow gonads atrophy and testosterone levels are extremely low. However, in the winter the adrenal gland produces elevated levels of dehydroepiandrosterone (DHEA)(Soma & Wingfield 2001), which is an androgen precursor. The brain can convert DHEA to androgens, and this process is specifically enhanced during aggressive challenges (Pradhan et al. 2010). Interestingly, DHEA is elevated during the winter in territorial red squirrels (Boonstra et al. 2008) and aggressive encounters increase DHEA levels in hamsters (Scotti et al. 2009). These data suggest that DHEA could be an important hormone regulating aggression in a wide array of species. It has been hypothesized that androgen synthesis in the brain allows individuals to engage in territorial behaviors that are mediated by androgens (and their estrogenic metabolites) while avoiding the metabolic costs of high testosterone levels.

Steroid hormones can modify behavior through protein synthesis pathways, which is a potent and enduring approach to responding to a changing environment. However, as we have shown, there are circumstances wherein more rapid, short-term responses are needed. Recent experiments have demonstrated that steroid hormones can indeed exert rapid effects on behavior, although the mechanisms for these effects differ from pathways involving long-term changes in gene expression. This could be beneficial, because some steroids, such as testosterone, are known to exert detrimental side effects. For example, long-term increases in plasma testosterone increase energy expenditure (Marler et al. 1995) and may suppress the immune system (Mills et al. 2010). Rapid nongenomic pathways permit steroid hormones to regulate behavior in response to sudden and short-lived environmental or social change. In this manner, a sudden encounter with a predator can prevent a rough-skinned newt from initiating a protracted mating ritual that leaves it vulnerable, as a corticosterone rapidly alters brain activity to suppress mating behavior As the literature builds, it is becoming clearer that steroid hormones are important in both short- and long-term regulation of behaviors.