Chapter 49:  Endocrine Effects on the Brain and Their Relationship to Behavior

The hypothalamic releasing factors regulate release of the anterior pituitary trophic hormones

As summarized in Figure 49-1, the releasing factors are produced in various neuronal groups within the hypothalamus and are transported to the median eminence for release into the portal circulation to the anterior pituitary. Neurons in the hypothalamus also produce the hormones oxytocin and vasopressin, which are released by the posterior pituitary into the blood. Therefore, it is not surprising that behavior and experience, which influence the hypothalamus, sometimes alter the secretion of these hypothalamic releasing factors and hormones.

Secretion of pituitary hormones is responsive to behavior and effects of experience

Consider, for example, the phenomenon of lactation, in which the sucking stimulus to the nipple triggers the release of oxytocin, which facilitates milk ejection, and of prolactin, which helps the mammary gland to replenish the supply of milk [2]. The phenomenon of stress also illustrates the behavioral and emotional control of anterior pituitary hormone secretion. Conditions associated with tissue injury and surgical trauma, as well as the so-called psychic stresses of fear, novelty and even joy, can activate the release of adrenocorticotropic hormone (ACTH), which in turn stimulates the secretion of adrenal glucocorticoids [2]. The behavioral—emotional stimuli are mediated by neural pathways that can be modified readily by learning.

In the female rabbit, copulation activates spinal reflex pathways that stimulate the secretion of luteinizing hormone (LH), which leads to ovulation [3]. In the male rabbit, copulation also activates the secretion of LH and increases plasma testosterone [3]. Social stimuli also modify gonadotropin secretion. Olfactory cues between female mice can interrupt normal estrous cycles and lead to pseudopregnancies or to periods of prolonged diestrus, termed the Lee-Boot effect; olfactory cues from male mice can shorten the estrous cycle and either cause rapid attainment of estrus, termed the Whitten effect, or terminate pregnancy in a newly impregnated mouse, termed the Bruce effect [3]. In male rhesus monkeys, sudden, decisive defeat by other males leads to prolonged reduction in plasma testosterone, which can be reversed in the defeated male by the introduction of a female companion [3]. In men, the anticipation of sexual intercourse has been reported to increase beard growth, a process under the control of circulating androgen [3], although this finding has not been documented.

Hormones secreted in response to behavioral signals act in turn on the brain and on other tissues

Functional changes caused by hormones secreted in response to behavioral signals include modifications of behavior. With the sex hormones, these changes strengthen and guide the reproductive process. Thus, aggressive encounters between male birds or mammals in defense of territory during the mating period stimulate gonadotropin and testosterone secretion. Increases in these hormones further increase readiness for sexual activity by enhancing supplies of sperm and seminal fluid. This analysis was taken further by Lehrman (see [1]), who showed that among doves the behavioral sequence of courtship, mating, nest building and parenting involves complex behavioral interplay between the partners that triggers further hormonal secretion, which unfolds the next phase of behavior and hormonal secretion.

Regarding the adrenal steroids, the behavioral activation of hormonal secretion in stress is part of a mechanism for restoring homeostatic balance. For example, an encounter with a predator may require rapid evasive action in which neural activity and rapidly mobilized hormones such as epinephrine play a role. Adrenal steroid secretion is slower, reaching a peak minutes after the stressful event, and as such is not expected to play a role in coping with the immediate situation. If the evasive action is successful and the animal survives, it will have to re-establish homeostasis; presumably, it also will learn from the experience to minimize the chances of another such encounter. Adrenal steroids facilitate such long-term adaptation; that is, they facilitate the extinction of a conditioned avoidance response [4]. Suppose an animal has learned to avoid a certain place where previously it was punished; it later discovers that being in that place no longer results in punishment. If, for example, that place also contains a food or water supply, it is in the best interest of the animal to extinguish the avoidance response in order to take advantage of the available food or water. Adrenal steroids have, in fact, been found to facilitate such extinction and, thus, can be said to facilitate a form of behavioral adaptation [4]. Adrenal steroids also appear to play a role in both selective attention and consolidation of a variety of learned information related to episodes or events in daily life [5]. Another aspect of adaptation in which stress-induced secretion of adrenal steroids participates concerns the ability of the organism to cope with a repeated stressful event through a variety of neurochemical changes [6].

Besides stress, adrenal steroids are secreted in varying amounts according to the time of day, and in this capacity they perform an important role in coordinating daily activity and sleep patterns with food-seeking and processing of information [3]. In nocturnally active animals, such as the rat, adrenal steroids are secreted at the end of the light period prior to onset of darkness. In humans and monkeys, adrenal steroid secretion precedes waking in the morning to begin daily activity. In both rats and primates, adrenal steroid secretion precedes the waking period and appears to contribute, during waking, to optimal synaptic efficacy in the hippocampus for long-term potentiation, a correlate of learning. It is this aspect of adrenal steroid action that contributes to enhanced attention and improved retention of episodic memories [5] (see Chap. 50). Moreover, adrenal steroid elevation prior to waking also increases food-seeking behavior and enhances appetite for carbohydrates [6].

Cyclic changes in hormonal secretion, which are under the control of daily and seasonal light—dark rhythms, are important not only for the adrenals but for the gonads as well. Estrous cycles, menstrual cycles and seasonal breeding patterns represent adaptations of individual species to climatic conditions of the environment. The feedback actions of gonadal and adrenal hormones, which are secreted in response to rhythmic output of hypothalamic and pituitary hormones, prime or activate the nervous system to perform the appropriate behavioral responses. It is important to stress that hormones themselves do not cause behaviors; rather, hormones induce chemical changes in particular sets of neurons, making certain behavioral outcomes more likely as a result of the strengthening or weakening of particular neural pathways.

Hormonal actions on target neurons are classified in terms of cellular mechanisms of action

Hormones act either via cell-surface or intracellular receptors. Peptide hormones and amino acid derivatives, such as epinephrine, act on cell-surface receptors that do such things as open ion channels, cause rapid electrical responses and facilitate exocytosis of hormones or neurotransmitters. Alternatively, they activate second-messenger systems at the cell membrane, such as those involving cAMP, Ca²⁺/calmodulin or phosphoinositides (see Chaps. 20–24), which leads to phosphorylation of proteins inside various parts of the target cell (Fig. 49-2A). Steroid hormones and thyroid hormone, on the other hand, act on intracellular receptors in cell nuclei to regulate gene expression and protein synthesis (Fig. 49-2B). Steroid hormones can also affect cell-surface events via receptors at or near the cell surface.

The various modes of hormonal action summarized in Figure 49-2 may be distinguished from each other by time course. The fastest effects, in both latency and duration, are those involving direct opening of ion channels and stimulation of exocytosis. Intermediate effects involve phosphorylation of enzymes, ion channels, receptors or structural proteins, which may last for minutes or even hours. The slowest and most enduring effects are those that alter gene expression and lead to induction or repression of enzyme or receptor proteins, growth responses and even the structural remodeling of tissues.

As summarized in Figure 49-3, steroid/thyroid hormone receptors bind to other proteins as well as to DNA [7–9]. In the simplest type of action (Fig. 49-3A), the steroid/thyroid hormone receptor becomes activated after the hormone binds to it; activation results in conformational changes that include shedding of other proteins, such as heat-shock proteins, and exposing the DNA-binding domain. The receptor then binds to the specific sequence of DNA, called a “hormone-response element” or enhancer, located on the coding strand of DNA; this enhances transcription by permitting other transcription factors as well as the RNA polymerase to bind to the promoter region [7–9]. A second scheme (Fig. 49-3B) is for the hormone receptor to bind with high affinity to another protein transcription factor, in this case the c-fos—c-jun complex, removing both protein complexes from binding to DNA [7–9]. Such a result also blocks the enhancement of transcription by either agent, although it also could reduce inhibitory effects produced by the hormone receptor through the scheme shown in Figure 49-3C. A variant on this theme, not shown in the figure, is “squelching,” in which multiple transcription factors compete with each other for a limited supply of soluble ligands that enhance their activities [7–9]. A third scheme, shown in Figure 49-3C, is for the steroid receptor to compete with another transcription factor for binding sites in the promoter regions. These other factors may be COUP or the cAMP-dependent transcription factors that bind to the adapter protein-1 (AP-1) response element. The result of this competition is inhibition of transcription since, in this situation, the other transcription factor enhances transcription, whereas the hormone receptor does not do so when it binds to the coding DNA strand [7–9].

As we have noted, second-messenger systems, through phosphorylation of nuclear proteins, can influence gene expression. There is evidence that even the classical steroid receptors are subject to regulation by phosphorylation and that phosphorylations promoted by a neurotransmitter such as dopamine (Chap. 12) are able to cause nuclear translocation of a steroid receptor in the absence of the steroid [10].

So far, the best understood examples of genomic regulation of neuronal function stem from the actions of gonadal and adrenal steroids and thyroid hormone, and many of these actions are involved in the plasticity of behavior that results from hormonal secretion, such as changes in aggressive and reproductive behavior and adaptation to repeated stress. In fact, hormonal actions that involve the genome are pervasive throughout the life cycle.

We can distinguish four major types of hormonal actions on the nervous system: (i) developmental actions, such as those involved in sexual differentiation and the effects of thyroid hormone and retinoic acid; (ii) reversible, and often cyclical, effects on the structure and function of neurons and glial cells that underlie corresponding cyclical changes in behavior, such as in reproduction and the daily rhythms of sleep and waking; (iii) experiential effects involving environmentally induced changes in hormonal secretion that evoke adaptive or maladaptive changes in the brain, as in stress; and (iv) effects that protect neurons or potentiate damage and lead to cell death.

It will be seen below that, in addition to genomic actions, there are nongenomic effects of steroids that modulate neurotransmitter release as well as ion traffic across the cell membrane and do so frequently in coordination with genomic actions.

Steroid hormones are divided into six classes, based on physiological effects: estrogens, androgens, progestins, glucocorticoids, mineralocorticoids and vitamin D

As shown in Figures 49-2 and 49-3, steroid hormone action on the brain and on other target tissues involves intracellular receptor sites that interact with the genome [1]. There are also important metabolic transformations of certain steroids, occurring in the nervous system, that either generate more active metabolites or result in the production of less active steroids. Such transformations are particularly important for the actions of androgens, of lesser importance for estrogens and progestins and of practically no importance for glucocorticoids and mineralocorticoids. For vitamin D, the principal transformation to an active metabolite occurs in the kidney and liver [11]. Some metabolites, such as allopregnanolone and allotetrahydrodeoxycorticosterone, produce nongenomic effects on the GABA_A receptor.

Some steroid hormones are converted in the brain to more or less active products that interact with receptors

The brain, like the seminal vesicles, is able to reduce testosterone to 5α-dihydrotestosterone (DHT); and, like the placenta, the brain aromatizes testosterone to estradiol (Fig. 49-4). Neither conversion occurs equally in all brain regions. The aromatization reaction is discussed below. Regional distribution of 5α-reductase activity toward testosterone in rat brain reveals that the highest activity is found in midbrain and brainstem, intermediate activity is found in the hypothalamus and thalamus and the lowest activity is found in the cerebral cortex [12]. The pituitary has higher α-reductase activity than any region of the brain, and its activity is subject to changes as a result of gonadectomy, hormone replacement and postnatal age [3,12]. 5α-DHT has been implicated in the hypothalamus and pituitary as a potent regulator of gonadotropin secretion, but it is relatively inactive toward male rat sexual behavior [3]. Labeled metabolites with Rf values of 5α-DHT have been detected in extracts of hypothalamic and pituitary tissue after [³H]-testosterone administration in both adult and newborn rats. It is interesting that progesterone inhibits 5α-reductase activity toward [³H]testosterone and that [³H]progesterone is converted to [³H]5α-dihydroprogesterone (Fig. 49-4). Progesterone competition for 5α-reductase may explain some of the antiandrogenicity of this steroid [3,12]. 5β-DHT is a metabolite of testosterone formed in the avian CNS, as is 5αDHT. 5β-DHT is inactive toward sexual behavior and is believed to represent an inactivation pathway for testosterone.

The aromatization of testosterone to form estradiol, and of androstenedione to form estrone (Fig. 49-4), has been described in brain tissue in vitro and in vivo [12]. Aromatization is higher in hypothalamus and limbic structures than in cerebral cortex or pituitary gland, and, in noncastrated animals, it is higher in the male than in the female brain. Aromatization has been found in reptile and amphibian brain as well as in mammalian brain [12]. The capacity to aromatize testosterone and related androgens, therefore, may be a general property of vertebrate brains. The functional role of aromatization has been studied most extensively in the rat. Male sexual behavior is facilitated by estradiol [13], and testosterone facilitation of male sexual behavior can be blocked by a steroidal inhibitor of aromatization [3,12,13]. There are indications that a similar situation exists in birds, amphibians and reptiles; that is, testosterone and estradiol can stimulate male and female heterotypical sexual behavior. Curiously, not all mammals are like the rat; for example, male sexual behavior of guinea pig and rhesus monkey is restored by the nonaromatizable androgen DHT [3,13].

Both aromatization and 5α-reductase are regulated by gonadal steroids. In mammals such as the rat, it is principally the neural aromatase activity that is upregulated by androgens acting via neural androgen receptors [14]. In birds, both neural aromatase and 5α-reductase are induced by testosterone, and this regulation provides a way by which androgens can regulate CNS hormone sensitivity without regulating receptor number [15].

Both estrogens and glucocorticoids appear to act on brain cells without first being metabolized because both [³H]estradiol and [³H]corticosterone are recovered unchanged from their cell nuclear binding sites in brain [3]. However, estradiol is subject to conversion to the catechol estrogen 2-hydroxyestradiol, and this metabolite is both a moderately potent estrogen via intracellular estrogen receptors as well as an agent capable of interacting with cell-surface receptors such as those for catecholamines, albeit at fairly high concentrations [16] (see Chap. 12).

Vitamin D, prior to acting in the brain, is converted to an active metabolite, 1,25-dihydroxy vitamin D₃, by enzymes in liver and kidney [11] (see Fig. 49-5). The nervous system also is capable of cleaving the side chain from cholesterol to generate the same initial series of steroids [17] that are produced by the adrenals and gonads, namely, pregnenolone, dehydroepiandrosterone and progesterone (Fig. 49-6). In addition, neural tissue converts progesterone via reduction of the Δ4–5 double bond and reduction of the 3 keto group to 3α,5α-pregnanolone, which is active on the chloride channel of the GABA_A receptor [18] (Fig. 49-7). Glial cells are believed to be the primary sites of both cholesterol side-chain cleavage and generation of pregnanolone from progesterone [17]. While steroids generated in the brain have been referred to as “neurosteroids,” a more useful term is “neuroactive steroids” to refer to all steroids that affect brain function via any mechanism and irrespective of site of formation. The term neuroactive steroid also has been used to describe neuroactive steroid drugs.

Genomic receptors for steroid hormones have been clearly identified in the nervous system

The detection of intracellular, DNA-binding steroid receptors became possible with the introduction of tritium-labeled steroid hormones of high specific radioactivity: 20 to 25 Ci/mmol at each labeled position. The limited number of these sites had escaped detection using steroids labeled with ¹⁴C at a much lower specific radioactivity. Tritium labeling also permitted histological localization of steroid receptors because the high spatial resolution of ³H, 1 to 2 μm in light-microscopic autoradiography, allows cellular and even cell nuclear localization of the radioactivity.

Cell fractionation procedures were fundamental to the biochemical identification of steroid and thyroid hormone receptors in brain as well as in other tissues. Isolation of highly purified cell nuclei from small amounts of tissue from discrete brain regions generally is accomplished with the aid of a nonionic detergent, such as Triton X-100 [19]. Cytosolic fractions of brain tissue, prepared by centrifugation of homogenates at 105,000 g for 60 min, contain the soluble steroid hormone-binding proteins, and a variety of methods intended to separate bound from unbound steroid have been used for measuring their binding activity [3,19]. The most commonly employed are gel filtration chromatography and sucrose density gradient centrifugation. Dextran-coated charcoal or Sephadex LH20 are frequently used because they effectively absorb unbound steroid and leave intact the complexes between steroid and putative receptor. Other methods, such as gel electrophoresis and precipitation of putative receptor material with protamine sulfate, have more restricted uses.

The objective of such studies is to measure the affinity, capacity and specificity of the hormone—receptor interaction [3,19]. Measurements of affinity and capacity are accomplished with equilibrium binding analysis. Specificity is based on competition between the labeled and various unlabeled ligands for binding sites.

Because the nervous system is highly heterogeneous from the standpoint of many neurochemical characteristics, including steroid and thyroid hormone receptors, the most useful techniques for mapping these receptors have been histochemical. Steroid autoradiography was the first such method. With purification of receptors and generation of antibodies, immunocytochemistry has been added as a tool. Cloning of steroid and thyroid receptors has opened the way to mapping of receptor mRNAs via in situ hybridization histochemistry.

Several criteria determine whether a steroid hormone-binding site is a putative receptor. First, the steroid hormone-binding site must be present in hormone-responsive tissues or brain regions and absent from nonresponsive ones. Second, it should bind steroids that are either active agonists or effective antagonists of the hormone effect and should not bind steroids that are inactive in either sense.

There is an ongoing controversy regarding intracellular localization of steroid and thyroid hormone receptors in cells. With the exception of thyroid hormone receptors, which are exclusively nuclear in location, cell fractionation studies have revealed that in the absence of hormone, steroid receptors are extracted in the soluble or cytosolic fraction. However, when steroid is present in the cell, many occupied receptors are retained by purified cell nuclei. Histological procedures, such as immunocytochemistry, have confirmed the largely nuclear localization of occupied receptors, but they also have revealed a nuclear localization of receptors in the absence of hormone in the tissue. This is true for estrogen, progestin and androgen receptors, but the mineralocorticoid and glucocorticoid receptors show a cytoplasmic localization in the absence of hormone. Thus, steroid receptors may exist in nuclei in a loose association that is disrupted during cell fractionation. This is not an uncommon situation for many constituents of cell nuclei [19].

Steroid hormone receptors are phosphoproteins that have a DNA-binding domain and a steroid-binding domain

All steroid receptors have a molecular weight of 55,000 to 120,000. The state of phosphorylation appears to influence functional activity.

Estradiol. The first neuroactive steroid receptor type to be recognized was that for estradiol [1]. In vivo uptake of [³H] estradiol and binding to cell nuclei isolated from hypothalamus, pituitary and other brain regions revealed steroid specificity closely resembling that of the uterus, where steroid receptors were first discovered [1]. Cytosolic estrogen receptors isolated from pituitary and brain tissue closely resemble those found in uterus and mammary tissue. A hallmark of the estrogen receptor is its existence as an aggregate of subunits that dissociate during steroid-induced transformation to the DNA-binding nuclear form of the receptor. This part of the estrogen receptor complex was cloned from human breast cancer cells and consists of a 65- to 70-kDa hormone and DNA-binding subunit [6,20]. The dissociation constant of estradiol binding is approximately 0.2 nM.

Estrogen receptors are found in the adult pituitary, hypothalamus, preoptic area and amygdala. They are principally in neurons, although glial cells also may express these receptors in some brain regions [1]. The developing rat brain expresses estrogen receptors in cerebral cortex and hippocampus, but these receptors largely disappear as the brain matures [21]. A newly described second form of the estrogen receptor, the β-estrogen receptor, is similar to the α form in affinity and specificity but different in tissue distribution [22].

Progesterone receptors in brain were detected using the synthetic progestin R5020 (promegestrone; 17α, 21-dimethyl-19-nor-pregna-4,9-diene-3,20-dione), which has a high affinity for the progestin receptor, with a K_d of 0.4 nM [23]. The progestin receptor, cloned from chick oviduct, consists of a steroid- and DNA-binding subunit of 108 kDa, although one 79-kDa subunit has also been described [6,20]. Progestin receptors with similar properties are found in pituitary, reproductive tract and most estrogen receptor-containing brain regions; these receptors are inducible by estrogen treatment [23]. There are also progestin receptor sites in brain areas lacking estrogen receptors, such as the cerebral cortex of the rat; these receptors are not induced by estradiol treatment. Nevertheless, such receptors resemble those induced by estradiol [23]. Another inducer of progestin receptors in brain is testosterone, which works through its conversion to estradiol via aromatization [23]. Progesterone acts rapidly to induce feminine sexual behavior, termed lordosis, in female rats that have been primed with estradiol to induce progestin receptors [23]. The principal site of estradiol and progesterone action is the ventromedial nucleus of the hypothalamus [1].

Androgen receptors have a steroid-binding subunit estimated to be 120 kDa [6,20]. The estimated K_d for active androgens is approximately 1 to 2 nM [1]. Androgen receptors are widely distributed in brain and pituitary tissue, although highest concentrations are found in hypothalamus, preoptic area and limbic brain tissue [24]. Androgen receptors are deficient in the androgen-insensitivity (Tfm) mutation, and animals with this mutation show defects in sexual behavior, juvenile rough-and-tumble play behavior and certain aspects of neuroendocrine function, thus indicating the actions of testosterone that are mediated by androgen receptors, as opposed to those mediated by aromatization of testosterone to estradiol (see above) and estrogen receptors [12].

Glucocorticoid. Adrenal steroid receptors have been subdivided into two categories, one of which is the classical glucocorticoid receptor [4]. This receptor, cloned from human and rat sources, consists of a steroid- and DNA-binding subunit of 95 kDa [6,20]. Such receptors, which have dissociation constants of 4 to 5 nM for glucocorticoids, are widely distributed across brain regions and are found in neurons and glial cells [4].

Mineralocorticoid. The other type of glucocorticoid receptor is similar to the mineralocorticoid receptor originally described in the kidney [4]. In the brain, receptors of this type bind the glucocorticoid corticosterone with high affinity, having a K_d of approximately 1 nM, and they are responsible for the high uptake of tracer levels of [³H]-corticosterone by the hippocampus [4]. These corticosterone receptors, which are found in high concentrations in the hippocampus but are also widely distributed in other brain regions at lower concentrations, may be involved in mediating the effects of diurnally varying concentrations of corticosterone [5]. Uptake of [³H]aldosterone by brain tissues reveals two types of binding sites: those in the hippocampus, which can be occupied preferentially by corticosterone, and those more diffusely distributed in the brain, which appear to retain [³H]aldosterone preferentially in the presence of the normally higher concentrations of corticosterone [4]. The reasons for this selectivity of an enzyme, 11β-hydroxysteroid dehydrogenase, are that, at least in the kidney-collecting tubules, it converts corticosterone to an inactive metabolite and allows aldosterone access to the mineralocorticoid receptors [24].

Vitamin D is a steroid hormone, production of which by the body requires the action of light. Therefore, it is often necessary to provide some vitamin D in the diet [11] (see Chap. 33). Moreover, vitamin D is converted by the kidney and liver to the active metabolite 1,25-dihydroxyvitamin D₃ (Fig. 49-5) [11]. Vitamin D₃ receptors consist of a hormone- and DNA-binding subunit of 55 kDa [7,20]. Receptor sites for 1,25-dihydroxyvitamin D₃ are found in pituitary and brain, especially in the forebrain, hindbrain and spinal cord neurons [25]. In the brain, one site containing vitamin D₃ receptors, the bed nucleus of the stria terminalis, responds to exogenous 1,25-dihydroxyvitamin D₃ with an induction of choline acetyltransferase, even though the calcium-binding protein that is regulated by vitamin D₃ in the intestine is not regulated by this hormone in the brain [26]. Moreover, vitamin D₃ also corrects deficiencies in serum testosterone and LH in vitamin D-deficient male rats [26]. It is not clear, however, whether this represents a major effect of vitamin D₃ in the pituitary gland or brain or both.

During development, steroid hormone receptors become evident in target neurons of the brain

These receptors appear within several days of final cell division [13]. Whether some receptors are also present in dividing neuronal precursors is not clear. After they have appeared, these receptors mediate a variety of developmental actions. For example, glucocorticoids direct differentiation of adrenergic/cholinergic neurons of the autonomic nervous system to develop in the adrenergic direction [4]. Glucocorticoids also increase the number of epinephrine-containing, small, intensely fluorescent cells, often referred to as SIF, in autonomic ganglia and are required for the normal postnatal ontogeny of serotonin neurons in the forebrain [4]. These effects may not all be direct but may involve hormonal modulation of growth factors produced by other cells surrounding the developing neurons.

Gonadal hormones, however, are involved in sexual differentiation of the reproductive tract and brain [13]. Mammals, among which animals have X and Y chromosomes, undergo sexual differentiation under the impetus of testosterone secreted by the testes during a period of perinatal life; in humans, this period is in mid-gestation, while in rats it is from the end of gestation into neonatal life. Key features of sexual behavior in birds are determined in the reverse manner, in keeping with the fact that the female has the chromosomal heterogeneity: females produce either estradiol or testosterone, either of which feminizes the brain, which otherwise would develop a masculine pattern in the absence of gonadal steroids [13,31].

As for the mechanisms of sexual differentiation, we must consider the metabolism of the hormone receptor types involved and the primary receptor-mediated events. Testosterone, as noted above, is a prohormone that is converted into either 5α-DHT or estradiol within the brain; these products exert effects on brain sexual differentiation via androgen and estrogen receptors, respectively [13,31]. Masculinization of sexual and aggressive behavior involves either 5α-DHT alone or a combination of 5α-DHT and estradiol acting on different cells. Besides masculinization, there is in some mammals a process of defeminization, in which feminine responses that would develop in the absence of testosterone are suppressed by its presence during the critical period. Conversion to estradiol appears to be involved in this process [13,31]. Progesterone plays no major role in brain sexual differentiation, but it does have the ability to antagonize actions at both androgen and estrogen receptors and, thus, can moderate the degree of masculinization and defeminization.

As to the primary developmental actions of testosterone, growth and differentiation appear to be involved. Testosterone or estradiol stimulates outgrowth of neurites from developing hypothalamic neurons that contain estrogen receptors [13]. This is believed to be one of the principal aspects of testosterone action that increases the number and the size of neurons within specific hypothalamic nuclei in males, compared to females [32]. 5α-DHT may have a similar effect on androgen-sensitive neurons. Differentiation of target neurons also occurs; in adult brain tissue, hormones like estradiol can evoke responses that differ between adult male and female rats [1,32].

The response of neural tissue to damage involves some degree of structural plasticity, as in development

Collateral growth and reinnervation of vacant synaptic sites is facilitated in some cases by steroid hormones [32]. In the hypothalamus, estrogen treatment after knife cuts that destroy certain inputs promotes increases in the number of synapses. In the hippocampus, glucocorticoid treatment promotes homotypical sprouting of serotonin fibers to replace damaged serotonin input. It has also been noted that androgens enhance the regrowth of the severed hypoglossal nerve [32]. One interpretation of these steroid effects is that injury reactivates programs of steroid-responsive genomic activity that normally operate during the phase of synaptogenesis in early development [32].

Another aspect of steroid action is stabilization of neurons against death and replacement. In the dentate gyrus of both the neonate and the adult rat, neurons are born and die; rates of both birth and death are increased by adrenalectomy, and these increases are prevented by low doses of adrenal steroids acting via mineralocorticoid receptors [4,6,33] (Fig. 49-9). Regulation of the turnover rate of dentate gyrus neurons may provide the hippocampal formation of the adult with the potential to increase and decrease its volume and functional capacity, as occurs in relation to seasonal or other long-term changes in the environment. This process also occurs in the developing dentage gyrus [33].

Activation and adaptation behaviors may be mediated by hormones

Hormonal secretion by the adrenals and gonads is controlled by endogenous oscillators that can be entrained by environmental cues such as light and dark. The actions of cyclically secreted hormones on behavior and brain function are referred to as activational effects. In addition to the cyclic mode, there is another mode of secretion initiated by such experiences as stress, fear and aggressive and sexual encounters. In this case, actions of adrenal steroid hormones secreted in response to experience lead to adaptive brain responses, which help the animal cope with stressful situations [4,6]. The activational and adaptational effects are largely reversible and involve a variety of neurochemical changes, most of which appear to be initiated at the genomic level. For example, estradiol is secreted cyclically during the estrous cycle in the female rat and triggers the surge of LH, which induces a surge of progesterone from the ovary. Progesterone, in turn, stimulates sexual responsiveness to the male rat [1].

Estradiol action to promote feminine sexual behavior in the rat involves a cascade of induced protein synthesis in specific hypothalamic neurons accompanied by morphological changes indicative of increased genomic activity [1]. Among the induced proteins are receptors for progesterone (see above), crucial for activating sexual behavior; receptors for acetylcholine and oxytocin that are active in enabling the hypothalamic neurons to respond to afferent input; proteins that are axonally transported from the hypothalamus to the midbrain, where they may be involved in neurotransmission; and structural proteins that contribute to formation of new synapses that come and go during the estrous cycle [1] (Fig. 49-10).

Adrenal steroids secreted in the diurnal cycle are reponsible for reversibly activating exploratory activity, food-seeking behavior, carbohydrate appetite and synaptic efficacy. These appear to do so by acting on the hippocampus, where there are many mineralocorticoid and glucocorticoid receptors (see above). Adaptational effects of adrenal steroids that result from stress appear to operate via the classical glucocorticoid receptor found not only in hippocampus but also in other brain regions [3,4]. Changes in synaptic vesicle proteins, high-affinity GABA transport, neurotransmitter-stimulated cAMP formation and central serotonin and noradrenergic sensitivity accompany repeated glucocorticoid elevations [3,4]. One view of these changes induced by glucocorticoids is that they counter-regulate some of the immediate and persistent neural effects of stress and are, therefore, part of the mechanism of adaptation [3,4,6]. This and other aspects of adaptation protect the organism; without adrenal secretions, the body likely would not survive many events in daily life.

The price of adaptation often includes some wear and tear and damage, called allostatic load [34]. Allostasis, meaning to achieve stability through change, is another word to describe adaptation. The wear and tear is the result of the inefficiency or overactivity of allostatic systems, including those like the adrenal cortex and autonomic nervous system, which respond to challenge and promote adaptation. This takes three forms: (i) repeated activation by many stressful events; (ii) failing to shut off after the challenge is over; and (iii) inability to be activated adequately, allowing other systems that are normally counter-regulated to become overactive, for example, inflammatory cytokines. In each of these situations, there is a cumulative as well as an immediate effect on many organs, including the heart and brain, as well as on the immune system.

Enhancement of neuronal atrophy and cell loss during aging by severe and prolonged psychosocial stress are examples of allostatic load

Repeated psychosocial stress in primates and rodents causes pathological changes in various body organs, including neuronal loss in the hippocampus [6,35]. Shorter exposure of rats to restraint stress, psychosocial stress or corticosterone causes atrophy of neurons in the hippocampus, particularly in the CA3 region, which receives heavy input from the dentate gyrus mossy fiber system (Fig. 49-9). The mossy fiber input is also responsible for kainic acid- and seizure-induced damage of the CA3 region and strongly suggests the importance of excitatory amino acids.

Indeed, the stress-induced atrophy is blocked by NMDA receptor blockers and by an anticonvulsant drug, phenytoin. The presence of elevated glucocorticoids at the time of hypoxic damage or kainic acid lesions (Chaps. 34 and 35) to the brain potentiates the necrosis produced by these treatments, especially in the hippocampus. Adrenalectomy reduces such damage and retards loss of neurons with age [6,35]. Thus, adrenal steroids operate in conjunction with neural excitability to produce damage, and it has been suggested that they do so by compromising the ability of the brain to obtain nutrients to support ATP generation [35].

The concept of allostatic load implies that there is a paradox in the actions of adrenal steroids: they exert protection in the short run and have the potential to cause damage in the long run if the allostatic, that is, the adaptation-promoting, responses are not managed efficiently. “Good stress” is therefore the efficient management of an allostatic response, whereas “bad stress” involves the persistence or otherwise inefficient operation of these normally adaptive responses.

There are gradients of health status across socioeconomic status that are not explained by access to health care or other simple explanations [36]. Therefore, it may be of great relevance in the future to understand the role of such factors as sense of control, helplessness, persistent fear and anxiety, diet, exercise and the impact of the living and social, for example, family and work, environments in regulating the allostatic systems; these factors could cause allostatic systems to operate inefficiently.

Gonadal hormones, like adrenal steroids, can produce both protection and damage. High-dose estrogen treatment of adult female rats induces hypothalamic disconnection syndrome by activating low-level persistent ovarian estrogen secretion. This persistent ovarian secretion induces morphological changes in the hypothalamus associated with dysfunction of the cyclic gonadotropin-releasing mechanism [37]. Persistent effects of estrogen secretion occurring naturally throughout the life span are believed to facilitate termination of cyclic hypothalamic function with regard to ovulation [37]. In the opposite sense, the suppression of gonadal function by repeated stress is likely to deprive the brain and cardiovascular system of protective actions of gonadal hormones and, thus, increase allostatic load [38,39].