Neuroendocrinology of Reproduction

Hypothalamus

The hypothalamus is located at the base of the brain ( Fig. 1.2 ). Although small (approximately 10 g, less than 1% of total brain weight), it performs critical functions for maintenance of whole-organism homeostasis, including regulation of hunger and body weight, growth, various aspects of metabolism, thirst and renal water handling, body temperature, autonomic function, sleep, circadian rhythms, and emotion. Importantly, the hypothalamus is also a primary control center for reproduction and influences sexual behavior.

As an anatomic structure, the hypothalamus does not have discrete borders, but it generally forms the floor and inferior-lateral walls of the third ventricle ( Fig. 1.3 ). The medial portions of the hypothalamus are primarily made up of cell bodies, whereas the lateral portions are mostly composed of neuron fibers (axons), such as those connecting the medial hypothalamus to other areas of the brain. By convention, closely associated collections of neuron cell bodies are called nuclei; and the paraventricular, dorsomedial, ventromedial, and infundibular nuclei contain a majority of the neurons that secrete hypophysiotropic hormones into the portal circulation. (The human infundibular nucleus is the analogue to the arcuate nucleus in lower mammalian species.) GnRH cell bodies do not form discrete nuclei but are instead diffusely located throughout the preoptic area and the mediobasal hypothalamus ( Fig. 1.4 ); the latter is situated caudal to the preoptic area, extending from the retrochiasmatic area (i.e., the area located behind the optic chiasm) to the mammillary bodies, and including both the infundibular (arcuate) nucleus and the median eminence.

Median Eminence

Positioned at the base of the third ventricle, the median eminence is part of the anatomic link between the hypothalamus and anterior pituitary. The internal zone of the median eminence is located along the ventral floor of the third ventricle and is largely composed of axonal fibers from both magnocellular neurons (larger neurons that secrete vasopressin and oxytocin) and hypophysiotropic neurons as they travel from hypothalamic nuclei/areas to their final destinations; the neurohypophysis (posterior pituitary) and the external zone of the median eminence, respectively ( Fig. 1.5 ). The external zone contains hypophysiotropic neuron terminals, which release hypophysiotropic hormones into an extensive capillary plexus (the proximal end of the hypophyseal portal system). Some nerve terminals in this zone act on other nerve terminals to influence hormone release (e.g., kisspeptin neurosecretion at GnRH neuron terminals influences GnRH release).

The ependymal layer lining the third ventricle includes a population of specialized ependymal cells called tanycytes, which have a short process extending toward the ventricular surface and a long process extending into the median eminence toward areas around portal capillaries. The latter tanycyte projections envelop or retract from GnRH nerve terminals during episodes of low and high GnRH neuronal activity, respectively. Thus, tanycytes may influence GnRH secretion via the regulated process of physically isolating GnRH neuron terminals from portal capillaries. Tanycytes may also represent a link between cerebrospinal fluid and the external zone of the median eminence (e.g., by transporting substances from the third ventricle to portal blood).

The median eminence is among the so-called circumventricular organs, which lie adjacent to the ventricular system and represent openings in the blood-brain barrier. Although lipid-soluble molecules can diffuse in and out of the CNS relatively easily, and cellular transport mechanisms allow selective entry of ions, the blood-brain barrier functions to protect certain regions of the brain and hypothalamus from larger charged molecules, with physical protection provided by (1) tight junctions between endothelial cells and (2) neuron-capillary separation by both astrocyte foot processes and microglia. However, the CNS requires feedback signals, including hormonal, metabolic, and toxic cues via macromolecules of peripheral origin that would otherwise be excluded by the blood-brain barrier. Accordingly, capillaries of the circumventricular organs are fenestrated and permit transcapillary exchange of larger charged molecules (e.g., proteins, peptide hormones). Thus the median eminence represents a key access point for central sensing of peripheral cues. Similarly, fenestrated vessels readily allow entry of hypothalamic-releasing factors into portal blood.

Hypophyseal Portal Circulation

No direct neuronal connections exist between the hypothalamus and the anterior pituitary. However, the hypophyseal portal circulation (hypothalamic-hypophyseal portal system, pituitary portal system) represents the functional connection between the median eminence and the anterior pituitary (see Fig. 1.4 ). The superior hypophyseal artery—a branch of the internal carotid artery—subdivides to form an extensive capillary network in the external zone of the median eminence, with loops that reach into the inner zone. Capillary blood then drains into sinusoids that converge into the hypophyseal portal veins. Traversing the pituitary stalk, the hypophyseal portal system forms the primary blood supply of the anterior pituitary. The direction of blood flow is primarily, but not exclusively, from the hypothalamus to the anterior pituitary; some retrograde flow allows for short-loop hypothalamic feedback.

Pituitary Gland (Hypophysis)

The pituitary gland appears as an extension at the base of the hypothalamus and resides cradled within the sella turcica, a saddle-like structure of the sphenoid bone (see Fig. 1.2 ). The adenohypophysis (anterior pituitary) is of ectodermal origin, derived from an upward invagination of pharyngeal epithelium (Rathke pouch) during embryologic development. The adenohypophysis is composed of primarily the anterior lobe (pars distalis), which contains specialized cell populations that produce specific hormones: gonadotropes (the gonadotropins LH and FSH), mammotropes (prolactin), corticotropes (adrenocorticotropic hormone [ACTH]), thyrotropes (thyroid-stimulating hormone [TSH]), and somatotropes (growth hormone). The intermediate lobe is vestigial in adult humans but includes a small population of cells (e.g., POMC cells) in contact with the posterior lobe; the pars tuberalis is a slender layer of tissue (e.g., LH-producing cells and TSH-producing cells) surrounding the infundibulum (the funnel-shaped connection between the hypothalamus and the posterior pituitary) and pituitary stalk.

In contrast to the adenohypophysis, the neurohypophysis (posterior pituitary) is composed of neural tissue and forms as a downward extension of neuroectodermal tissue from the infundibulum during embryologic development. It is thus a direct extension of the hypothalamus. The neurohypophysis includes the infundibular stalk and the pars nervosa (posterior lobe of the pituitary). The supraoptic and paraventricular nuclei include magnocellular neurons that produce oxytocin and arginine vasopressin (AVP; also known as antidiuretic hormone [ADH]), respectively; these axons project to the posterior lobe of the pituitary, where oxytocin and AVP are secreted into a capillary network that drains into the hypophyseal veins (i.e., directly into the systemic circulation). The posterior lobe also includes specialized glial cells called pituicytes, which envelop or retract from magnocellular nerve terminals during episodes of low and high neuronal activity, respectively.

Gonadotropin-Releasing Hormone Structure

GnRH (GnRH-1 in particular) is a decapeptide, with the amino acid structure (pyro)Glu-His-Trp-Ser-Tyr-Gly-Leu-Arg-Pro-Gly-NH ₂ . The amino acid structure of GnRH is identical in essentially all mammalian species; with the exception of the central Tyr-Gly-Leu-Arg segment, the amino acids of GnRH are highly conserved among vertebrate species. The GnRH-1 gene (GNRH1) is located on human chromosome 8 (8p11.2-p21) and produces a 92–amino acid precursor peptide called prepro-GnRH, which includes a signal sequence (23 amino acids), GnRH (10 amino acids), a proteolytic processing site (3 amino acids), and GnRH-associated peptide (56 amino acids) ( Fig. 1.6 ). The latter peptide can stimulate gonadotropin secretion and inhibit prolactin secretion, although its precise physiologic role, if any, remains unclear. The actions of GnRH are mediated through the GnRH type I receptor.

Another form of GnRH (GnRH-2) and its receptor have been identified in a variety of animal species, including humans. GnRH-2 is a decapeptide with a similar structure to GnRH-1: (pyro)Glu-His-Trp-Ser- His -Gly- Trp - Tyr -Pro-Gly-NH2 (italicized amino acids denote differences compared with GnRH-1). However, the gene for GnRH-2 is located on human chromosome 20 (20p13). GnRH-2 is widely expressed in the CNS and extra-CNS tissues, and it may contribute to reproductive behavior regulation in some species. In lower animals, GnRH-2 can act via its own receptor, which is structurally and functionally distinct from the GnRH type I receptor. Although a homologue of the GnRH-2 receptor gene has been detected in humans, it includes a frameshift and premature stop codon. Thus the physiologic role of GnRH-2 in humans remains unclear.

Anatomy of Gonadotropin-Releasing Hormone-Secreting Neurons

GnRH neurons are a heterogeneous population of hypothalamic neurons. They are relatively few, numbering approximately 7000 in humans, and the majority of GnRH neuronal cell bodies are located in the infundibular (arcuate) nucleus—part of the mediobasal hypothalamus—and the medial preoptic area. ^, GnRH neurons in the infundibular (arcuate) nucleus appear to be requisite for gonadotropin secretion. For example, selective radiofrequency ablation of the arcuate nucleus in adult female monkeys obliterates gonadotropin secretion. Although GnRH neurons are rather loosely affiliated anatomically, they are functionally integrated and form a complex, interconnected network with extensive connections to other neuronal populations. GnRH neurons extend projections through the tuberoinfundibular tract to the median eminence, where neuron terminals gain access to the hypophyseal portal system.

Recent work in mice suggests that the GnRH neuron fibers extending from the cell body to the median eminence are morphologically atypical: although they do not exhibit many of the molecular markers classically associated with axons or dendrites, they demonstrate morphologic and functional characteristics of both axons and dendrites, including functional synaptic inputs along the fiber. ^, Accordingly, the term dendron has been used for such projections. In mice, the distal portion of such dendrons exhibit a particularly high density of dendritic spines and synaptic inputs, beyond which the dendron branches into multiple short axons at the median eminence.

The physiologic function of other GnRH neurons, which arise from the anterior and posterior hypothalamus and project to the limbic system and posterior pituitary, respectively, remains unclear, although some of these circuits may possibly be involved with various behavioral responses.

Embryologic Development of the Gonadotropin-Releasing Hormone Neuronal Network

The ontogeny of GnRH neurons in vertebrate species is unique among neuronal systems of the CNS: nascent GnRH neurons are initially identified outside of the CNS in the nasal placode (sometimes called the olfactory placode). However, GnRH cells migrate during embryologic development, as directly observed in embryonic nasal explant cultures and in murine embryonic head slices. The specific migratory pathway of GnRH neurons was first demonstrated in mice by documenting the presence of GnRH-immunoreactive cells in different areas at different stages of embryonic development ( Fig. 1.7 ). Specifically, GnRH expression is first observed within the nasal placode circa embryonic day 10 or 11. By embryonic day 13, GnRH-expressing cells are primarily located around the cribriform plate, and GnRH-expressing cells begin to reach the hypothalamus by embryonic day 14, approaching their final positions around embryonic day 16. This migratory pathway has been confirmed in both nonhuman primates and humans.

Successful migration of GnRH neurons is inextricably intertwined with olfactory system development, perhaps reflecting the close functional relationship between reproduction and the olfactory system (e.g., pheromones) in mammalian phylogeny. The nasal placode gives rise to nasal epithelium and olfactory sensory neurons, the latter of which extends axonal projections to the olfactory bulb. Vomeronasal neurons are a subset of olfactory neurons believed to be involved with pheromone detection; these axons originate in the vomeronasal organ and largely extend to the accessory olfactory bulb. At the level of the cribriform plate, some olfactory (vomeronasal) axons separate and form a branch that extends caudally into the forebrain. Of great importance, migrating GnRH neurons maintain adhesion to these axons; thus, these olfactory neurons form a critical guidance track for GnRH neuronal migration across the nasal epithelium and through the forebrain toward the hypothalamus. ^, After reaching the hypothalamus, GnRH neurons detach from olfactory nerve axons and may disperse further before resting. A critical step is the extension of GnRH neuronal projections to the median eminence, where GnRH gains access to the hypophyseal portal system.

The dependence of GnRH neuronal migration on normal olfactory system development is exemplified by Kallmann syndrome, a form of isolated hypogonadotropic hypogonadism accompanied by an absent or reduced sense of smell (anosmia and hyposmia, respectively). In this syndrome, faulty development of the olfactory system renders an inadequate guidance infrastructure for migrating GnRH neurons, leading to absent or incomplete GnRH neuron migration to the hypothalamus. The first identified cause of Kallmann syndrome was a deletion of ANOS1 (formerly the Kallmann syndrome 1 sequence [ KAL1 ] gene), which is located on the X chromosome (Xp22.3) and encodes anosmin-1, a secreted matrix glycoprotein expressed in the presumptive olfactory bulb. Although precise mechanisms are unclear, anosmin-1 is believed to be important for the formation of olfactory elements that provide migratory guidance to GnRH neurons as they move out of the nasal placode. Evaluation of a 19-week-old human fetus with X-linked Kallmann syndrome demonstrated GnRH-immunoreactive cells within a tangle of olfactory and vomeronasal nerves at the dorsal surface of the cribriform plate, along with the absence of olfactory tracts and bulbs. In a second human fetus (16 weeks) with X-linked Kallmann syndrome, GnRH was detected along terminal nerve fascicles in the nasal mucosa only. This form of Kallmann syndrome illustrates that without the guidance framework provided by the olfactory neuronal system, GnRH neurons do not appropriately migrate into the hypothalamus and thus cannot release GnRH into the hypophyseal portal system.

A number of additional single-gene defects have been associated with Kallmann syndrome, including variants in the genes for prokineticin 2 (PROK2) and its receptor (PROKR2) , fibroblast growth factor 8 (FGF8) and its receptor fibroblast growth factor receptor 1 (FGFR1) , ^, NMDA receptor synaptonuclear signaling and neuronal migration factor ( NSMF ; formerly nasal embryonic LH-releasing hormone factor [ NELF ]), chromodomain helicase DNA binding protein 7 (CHD7), semaphorin 3A ( SEMA3A ), and SRY-box 10 ( SOX10 ). The importance of these genes in GnRH neuronal development is corroborated by mouse studies. For example, in fetal mice lacking either Prok2 or Prokr2 , GnRH neurons are trapped in a tangled web of olfactory/vomeronasal axons, with few, if any, reaching the forebrain. Although such gene products are clearly important for GnRH neuron migration, their precise roles remain unclear.

Of interest, patients with specific Kallmann syndrome-related gene variants can exhibit variable penetrance and different phenotypic expressions (even within families), suggesting the importance of other factors such as gene-gene interactions (oligogenicity). Some Kallmann syndrome-related gene variants (e.g., PROK2 , PROKR2 , FGF8 , FGFR1 , NSMF , CHD7 ) are also associated with normosmic hypogonadotropic hypogonadism. In addition, a minority of patients with some Kallmann syndrome-related gene variants (e.g., ANOS1 , PROKR2 , FGFR1 , NSMF , or CHD7 ) may demonstrate partial or full recovery of reproductive function ( reversal of hypogonadotropic hypogonadism) in later life, ^, suggesting plasticity of the GnRH neuronal network and, perhaps, gene-environment interactions. Mechanisms accounting for such phenomena remain unclear.

Gonadotropin-Releasing Hormone Neuronal Firing and Gonadotropin-Releasing Hormone Secretion

The activity of many GnRH neurons is marked by bursts of action potentials (burst firing), the patterns and rates of which change across time. Such changes in GnRH neuron firing rates may relate to changes in pulsatile GnRH release at the median eminence, although this hypothesis has not been experimentally verified. Variable firing rate patterns (e.g., times of high and low firing rates) appear to be intrinsic to GnRH neurons, but they can also be altered by neurotransmitters and neuromodulators (e.g., glutamate, GABA, kisspeptin). Although sex steroids can markedly influence GnRH neuronal firing rates, GnRH neurons lack the primary receptors mediating sex steroid feedback (i.e., estrogen receptor alpha, progesterone receptor, androgen receptor). However, many studies suggest that sex steroid actions on GnRH neuronal activity are mediated primarily via afferent neurons (e.g., those secreting glutamate, GABA, kisspeptin).

GnRH neuron cell bodies are relatively scattered across the mediobasal hypothalamus and preoptic area, yet GnRH is secreted into the hypophyseal portal system in a coordinated, pulsatile fashion. Specifically, GnRH secretion is marked by episodic bursts of hormone release into the portal system, as demonstrated in rats, sheep, and monkeys. After being released into the portal vascular compartment, GnRH is rapidly degraded via enzymatic proteolysis, and the half-life of GnRH in the blood is very short—approximately 2 to 4 minutes. Thus GnRH presentation to gonadotrope cells is intermittent.

Pulsatile GnRH secretion is an absolute requirement for long-term stimulation of gonadotropin synthesis and secretion, and there is a relatively narrow window of GnRH pulse frequency and amplitude that will optimally stimulate gonadotropin secretion. Intermittent GnRH stimulation of gonadotrope cells can increase (or maintain) GnRH receptors on gonadotropes (the self-priming or autopriming effect). Thus intermittent GnRH stimulation facilitates or maintains gonadotrope responsiveness to GnRH. However, more frequent exposure to GnRH pulses can reduce gonadotropin responses to GnRH. In classic experiments involving rhesus monkeys with hypothalamic lesions that abolished GnRH secretion, once-hourly exogenous GnRH administration restored pituitary gonadotropin secretion. However, changing from once-hourly pulses to twice-hourly pulses reduced LH and FSH secretion by 50% to 60%, while 3 to 5 pulses per hour profoundly suppressed plasma LH and FSH concentrations. Marked desensitization of gonadotropin release is also observed when changing from once hourly (intermittent) to continuous GnRH administration ( Fig. 1.8 ). Although reduced GnRH receptor expression on gonadotropes (i.e., receptor downregulation) plays a role in desensitization, additional mechanisms contribute to the uncoupling of GnRH receptor agonism and gonadotropin synthesis.

The foregoing phenomenon can be exploited therapeutically with the use of long-acting GnRH receptor agonists. Such agonists are peptides with structures very similar to that of GnRH but with amino acid substitutions that enhance receptor binding affinity, increase resistance to proteolytic degradation, or both ( Fig. 1.9 ), thus providing continuous GnRH receptor stimulation. Although initial GnRH receptor agonism temporarily (for 1 to 2 weeks) increases gonadotropin release (gonadotropin “flare”), continued agonism leads to desensitization of gonadotropin secretion with accompanying reductions of gonadal sex steroid concentrations to castrate levels (“medical oophorectomy,” “medical castration,” “pseudomenopause”), typically over several weeks. These agents can be useful in the therapy of gonadotropin-dependent disorders such as central precocious puberty, endometriosis, and prostate cancer.

Peptide GnRH receptor antagonists are also available for clinical use. These antagonists reversibly bind to, but do not stimulate, the GnRH receptor (i.e., competitive antagonism). Thus these agents do not initially stimulate gonadotropin release, and they reduce gonadotropins more rapidly than GnRH agonists—usually within 24 to 72 hours.