Neurosecretory Peptides

The concept that neurons would be capable of secretory function was first proposed by Scharrer in 1928 and was based on observations made in fish.¹ These observations were later extended to mammals in experiments that demonstrated that extracts from certain neurons contained bioassayable vasopressin activity. Because the initial identification of vasopressin, a growing number of neuropeptides, some of which directly influence reproductive function, have been identified.

Unlike the classic neurotransmitters that are secreted after enzymatic conversions at the nerve terminal, peptide synthesis requires transcription of mRNA from DNA, subsequent translation on ribosomes, and posttranslational processing. This synthesis occurs in the neuronal cell body rather than at the nerve terminals. The initial product is usually a larger molecular weight precursor molecule that is then transported down the axon and further processed by enzymatic cleavage into the active peptide before secretion (Fig. 1). From a phylogenetic perspective, the widespread distribution of brain peptides probably reflects evolutionary conservation of vital regulatory functions imposed by these peptides. For example, a vasopressin-like peptide occurs in the freshwater snail, in which it functions as a neurotransmitter. The importance of neuropeptides is further supported by the presence in higher animals of multiple subtypes of many of the neuropeptides and their receptors, with each having different physiologic roles.² Although it is likely that these multiple subtypes of neuropeptides and their receptors arose through genome duplication events, their retention and diversification into separate physiologic roles convey their importance and suggest that additional subtypes remain to be discovered.²

On the basis of evidence from many biologic systems, information transfer from the neurosecretory neuron can modulate biologic activity by four principal means: autocrine, paracrine, neuroendocrine, and neurotransmission. The most direct form of neurohumoral signaling is the autocrine mode, in which the release of a substance from the cell acts to autoregulate activity within that same cell. For example, gonadotropin-releasing hormone (GnRH) inhibits its own gene transcription in hypothalamic neurons.³ The paracrine mode of communication occurs when secretion of a substance controls activity on adjacent cells. This mode of regulation functions in the pituitary. When gonadotrophs (i.e. luteinizing hormone [LH]- and follicle stimulating hormone [FSH]-secreting cells) are exposed to GnRH, a paracrine product is secreted that allows adjacent lactotrophs to respond with prolactin release.⁴ A third form of informational transfer, the neuroendocrine mode, is characterized by a release of a neurohumoral product such as oxytocin into the general circulation to regulate distant tissues that have oxytocin receptors, such as the myoepithelial cells of the breast, causing these cells to eject milk.

Neurotransmission can be conceptualized as a special form of paracrine secretion in which there is cell-to-cell communication through typical synaptic contacts. The classic neurotransmitters include acetylcholine, norepinephrine, epinephrine, dopamine, histamine, serotonin, and γ-aminobutyric acid (GABA). In general, neurons that contain these transmitters are activated by an action potential above a threshold leading to membrane depolarization presynaptically. Ion fluxes then mobilize release of a small amount of the transmitter substance, which diffuses across the synaptic gap and causes excitatory or inhibitory postsynaptic potentials. The transmitter is then removed by enzymatic degradation or a reuptake mechanism. It is unclear whether all brain peptides are capable of synaptic transmission; however, it is estimated that the classic neurotransmitters only account for 40% of the synaptic contacts in the central nervous system.⁵ Peptide-secreting hypothalamic neurons make synaptic contact with each other (Fig. 1), and it is likely that neuropeptides play an important role in the modulation of synaptic transmission.

The development of a bioassay system in which the biologic effect of an unknown substance can be accurately and reproducibly measured is the first step in the characterization of new peptides. With conventional biochemistry techniques, tissue extracts are subjected to chemical manipulations such as differential solvent extraction and chromatographic procedures to purify the unknown factor. After each purification step, the biologic activity of the fractions is determined, and the fractions of interest are further purified to the final active material. Using this approach, Schally and Guillemin's laboratories were able to isolate gonadotropinreleasing factor, a process requiring more than 500,000 hypothalami.^6,7 Through similar meticulous extraction procedures, Guillemin's group also isolated somatostatin.⁸

Recombinant DNA techniques have been rapidly applied to this area and have been combined with conventional biochemical approaches. One series of isolation strategies is illustrated in Figure 2. After amino acid sequencing, antibodies against the new peptide can be raised, and immunocytochemical staining of tissue sections can provide an overall view of the localization and distribution of the new peptide. These techniques are responsible for the rapid progress in the discovery of new peptides. With immunocytochemistry, many peptides that were originally isolated in the gastrointestinal tract, such as vasoactive intestinal polypeptide (VIP), have also been localized to the central nervous system. However, neuropeptides such as oxytocin and β-endorphin have been identified in nonneural tissue such as the ovary, testes, and adrenal gland (Table 1).⁹

TABLE 1. Classification of Neuroactive Peptides

Adapted from Cooper JR, Bloom FE, Roth RH: The biochemical basis of neuropharmacology. In Cooper JR, Bloom FE, Roth RH (eds): Neuroactive Peptides, p 295. New York, Oxford University Press, 1982.

An important advance in our understanding of the physiology of neurosecretory peptides occurred with the development of transgenic technology, or the ability to precisely alter the genetic makeup of an experimental animal. These techniques result in the generation of animals, usually mice, with genetic alterations that include the introduction of human genes, an increase in the expression of specific single genes, and the complete lack of expression of specific genes, or gene knockouts. A detailed analysis of these animals can yield important information about the physiology of a neuropeptide that cannot be obtained in any other way.

The knockout mouse,¹⁰ in which the animal is completely normal except for the precise deletion of a single gene, is the most widely used of these transgenic techniques. With the exception of naturally occurring knockouts, such as Kallmann's syndrome, that are found for a variety of genes in a number of species, experimentally generated knockouts are only produced in mice. These mice are generated by a complex series of experiments that begins with the cloning of the gene of interest (Fig. 3). This information is used to generate a mutant gene that is inactive or irrelevant. The mutant gene is exchanged for the normal gene through a process known as homologous recombination, a normal cellular process allowing for the exchange of genetic information between chromosomes. This exchange is carried out in highly specialized cells known as embryonic stem cells, which have the unique property that they can be incorporated into mouse embryos. The mutant cells are then incorporated into these first-generation mice essentially at random. To generate the full knockout mouse, a large number of these mice must be screened to find one in which the mutant cells were correctly incorporated into germ cells. When this mouse reproduces, its offspring will carry the mutant gene. Further breeding can be performed to generate mice in which both copies of the gene are mutant, creating the full knockout mouse. This technology has been applied to a large number of neurosecretory peptides, as described in the following sections.

It would initially appear that the development of a knockout mouse for a neuropeptide would provide specific information about the physiology of the peptide: what does this neuropeptide really do? Fortunately, this is often the case, but in many instances, the knockout mouse yields surprising and unanticipated information. For example, a knockout mouse may be completely normal, indicating that the peptide of interest does not subserve proposed functions, at least not in the mouse, or more likely that there are other peptides or systems that compensate for its loss. Another common result is that the knockout mouse is nonviable. When this occurs in the embryonic period, such findings support a critical role for the peptide in development but may yield little other useful physiologic information. Although the knockout mouse often provides extremely useful information about the physiologic roles of a neurosecretory peptide, there are also significant limitations to its use.

The ubiquitous distribution of neurosecretory peptides has raised many questions regarding their evolution, mechanism of action, function, and ultimate physiologic roles. In this chapter, neuropeptides are discussed with special emphasis on hypothalamic peptides and peptide derivatives of the opioidergic system and on their recognized or putative roles in reproduction.

There is a common structural organization of hypothalamic peptide neurons. Most of the neuronal cell bodies are concentrated in the arcuate, paraventricular, and periventricular nuclei (Fig. 1). With more refined immunocytochemical techniques, it is evident that more than one neuropeptide can be synthesized by a single group of neurons. For example, neurons that secrete corticotropin-releasing hormone (CRH) may also contain vasopressin or oxytocin.

The neurohypophysial hormones oxytocin and vasopressin are synthesized in the paraventricular and supraoptic nuclei of the hypothalamus. Each peptide is 9 amino acids long and contains an internal disulfide bridge at the 1 through 6 positions (Fig. 4). Their similarity to neurohypophysial hormones such as vasotocin in other species suggests that they evolved from a common precursor by a single point mutation. Both peptides are synthesized as part of a larger prohormone; oxytocin is associated with an estrogen-stimulated neurophysin, and vasopressin is associated with a nicotine-stimulated neurophysin.¹¹ Immunocytochemical studies of individual neurons have shown that they contain oxytocin, vasopressin, and their neurophysins.¹²

Clinical studies have shown that the vasopressin secretion is tightly controlled and governed primarily by plasma osmolarity and circulating blood volume. For example, serum osmolarity is kept constant within strict tolerances (+ 1.8%).¹³ Vasopressin secretion can also be stimulated by administration of angiotensin II.

Clearly established roles for oxytocin that is released into the circulation are milk ejection from the breast and stimulation of uterine contractions during the second stage of parturition. Controversy still exists concerning the role of oxytocin in labor in humans, because no appreciably higher levels are detected during early labor¹⁴ and because an oxytocin knockout mice deliver normally.¹⁵ Perhaps a more important factor is that the activity of circulating oxytocin is amplified by the increase in uterine-myometrial oxytocin receptors.¹⁶

Oxytocin and vasopressin have also been shown to affect the adrenocorticotropic hormone (ACTH)-adrenal axis. Vasopressin is capable of stimulating ACTH release and is considered to have weak corticotropin-releasing activity. After the isolation of ovine CRH, it was shown under experimental conditions that vasopressin acted synergistically with CRH to release fivefold more ACTH than CRH alone in humans.¹⁷ Alternatively, oxytocin has been reported to have effects opposite that of vasopressin, causing suppression of ACTH secretion.¹⁸ These potential physiologic interactions are reinforced by the findings of Bergland and Page that the neurohypophysis is vascularly interconnected with the anterior pituitary.¹⁹

Many other diverse actions of these nonapeptides have been hypothesized. Preliminary evidence in rats suggests that these two peptides are differentially released in response to different types of stress.²⁰ Moreover, an increase in oxytocin activity has been associated with a decrease in GnRH secretion. Both nonapeptides have been postulated to have effects on cognitive functions. Vasopressin or its synthetic analogs has been shown to enhance memory in normal and cognitive-impaired adults.²¹ Oxytocin appears to have the opposite effect, impairing memory functions,²² but in addition oxytocin can induce maternal behavior in rats.²³

A knockout mouse for oxytocin has been developed. Oxytocin-deficient mice cannot lactate and display mild behavioral changes. Surprisingly, these mice deliver normally and do not display problems in mating or maternal behavior. This finding supports a stronger role for prostaglandins in uterine contractions of parturition, and mice deficient in cyclooxygenase-1 (COX-1), the enzyme that synthesizes prostaglandins, have delayed labor and subsequent neonatal death. Mice deficient in oxytocin and COX-1 do deliver normally, an effect attributed to the luteotrophic action of oxytocin. A "natural" knockout model for vasopressin has been known for many years: the Brattleboro rat.²⁴ Although these rats display the expected diabetes insipidus of vasopressin deficiency, they also have several other central nervous system (CNS) defects, affecting behavior and cognition.²⁵

Thyrotropin-Releasing Hormone

The first hypothalamic-releasing factor isolated was the tripeptide thyrotropin-releasing hormone (TRH), which was discovered independently by Schally and Guillemin's group.^7,26 Like other neuropeptides, TRH is derived from a much larger 30-kd precursor (Fig. 5) and specifically stimulates release of thyroid-stimulating hormone (TSH). TRH also has prolactin-releasing properties, because lactotropes have receptors for TRH. TRH may play other roles besides its physiologic role in TSH release. It has been localized by immunocytochemical techniques to the cerebral cortex, spinal cord, paraventricular nucleus, pineal gland, neurohypophysis, and ventral horn motor cells.²⁷ A diagrammatic view of the circumventricular hypothalamic nuclei is shown in Figure 6. Neuropharmacologic studies implicate TRH as a neuromodulator with general stimulant activity. Experimentally observed properties of TRH are summarized in Table 2.²⁸

TABLE 2. Central Nervous System---Mediated Actions of Thyrotropin-Releasing Hormone

  Hypothermia
  Reversal of spinal shock
  Increased spontaneous motor activity
  Alteration in sleep patterns
  Anorexia
  Inhibition of avoidance behavior
  Reversal of depression
  Neuromodulator of classic transmitter action

  Releases norepinephrine and dopamine from synaptosomal preparations
  Enhances disappearance of norepinephrine from nerve terminals
  Potentiates excitatory actions of acetylcholine on cerebral cortical neurons

Vale W, Rivier C, Brown M: Regulatory peptides of the hypothalamus. Ann Rev Physiol 39:473, 1977.

Transgenic mice deficient in TRH have been generated.²⁹ These mice develop normally and are fertile. The affected mice are hypothyroid with elevated TSH levels (i.e. tertiary hypothyroidism). The mice also display hyperglycemia and abnormal insulin secretion.²⁹

Because of its ability to reverse spinal shock in animals, TRH has been used experimentally in patients with spinal cord trauma as an adjunctive treatment with promising results. Because of a short half-life, additional clinical applications will require development of long-acting analogs.

Gonadotropin-Releasing Hormone

GnRH is a decapeptide that was also isolated and characterized by Schally and Guillemin's laboratories in 1971 from porcine and ovine hypothalami.^6,7 The structure of this peptide is identical in all mammals, including humans (Fig. 5). GnRH has been localized primarily to the hypothalamus in the interstitial nucleus (medial preoptic area) and has its highest concentrations in the arcuate nucleus adjacent to the median eminence³⁰ (Fig. 6). Small amounts of GnRH are also found in the circumventricular organs of the brain. There are an estimated 1,500 GnRH neurons in the hypothalamus.

GnRH, like many other neuropeptides, is synthesized as part of a much larger precursor peptide containing 92 amino acids.³¹ GnRH is cleaved from the precursor, leaving a 56-amino acid peptide known as GnRH-associated peptide (GAP).³² GAP has been shown to have potent activity to inhibit the release of prolactin. The placenta, which also secretes GnRH, forms a different precursor protein through the use of a different promoter sequence.³³

Because of the importance of GnRH in the control of reproduction, intensive efforts have been made to determine its physiologic role. When administered to humans, GnRH stimulates a rapid and large release of LH and a smaller release of FSH. In the absence of GnRH, as in patients with Kallmann's syndrome, there is virtually no secretion of LH and FSH, emphasizing the importance of GnRH in the reproductive axis. As with other knockouts, the unexpected finding of anosmia in these patients led to the discovery that GnRH neurons arise in the olfactory placode and migrate to the hypothalamus.³⁴

The half-life of GnRH in the peripheral circulation is 2 to 4 minutes.³⁵ Because GnRH is rapidly degraded by peptidases, analogs have been synthesized that have much longer half-lives (Fig. 5). By trial and error testing of thousands of GnRH analogs, it was discovered that substitution of amino acids at the 2 or 3 position results in analogs with antagonistic properties, whereas modifications at the 6 or 10 position result in analogs with agonist action.

Two unique functional properties of GnRH have permitted its widespread clinical application. When GnRH is administered as a continuous infusion, pituitary LH response is initially brisk but is followed by a decrease in LH secretion and decrease in GnRH receptors (i.e. downregulation). Agonistic analogs of GnRH promote this downregulation phenomenon and can be used to induce "medical castration" by shutting down the pituitary-gonadal axis. Several GnRH analogs are available clinically and are widely used for a number of diseases, including endometriosis and uterine leiomyomas. If GnRH is administered in a pulsatile fashion at 60- to 90-minute frequencies, LH receptors and LH response increase. Continuation of this mode of administration for prolonged periods results in stimulation of a normal menstrual cycle³⁶ or normal testosterone production and spermatogenesis in patients with a deficiency or disorder of GnRH secretion. A summary of the clinical applications for GnRH and its analogs is provided in Table 3.

TABLE 3. Clinical Application of Gonadotropin-Releasing Hormone and Its Analogs

  Activation of pituitary-gonadal axis

  Dynamic testing of pituitary
  Ovulation induction
  Induction of spermatogenesis
  Cryptorchidism

  Downregulation of pituitary-gonadal axis

  Contraception
  Gonadal steroid-dependent tumors

  Prostate cancer
  Breast cancer
  Endometrial cancer

  Endometriosis
  Polycystic ovarian disease
  Uterine leiomyoma
  Premenstrual syndrome
  Precocious puberty
  Controlled ovarian hyperstimulation

Besides its role in the control of reproduction, GnRH also has intriguing behavioral effects. When administered to ovariectomized, estrogen-primed female rats intracerebrally, GnRH induces lordotic posturing and mating behavior.³⁷ In contrast, administration of GnRH antibodies has suppressed sexual behavior.³⁸ This effect is a unique property of this peptide. From a phylogenetic perspective, these behavioral effects appear similar to a peptide that has been isolated from yeast, the α-mating factor. The α-mating factor has sequence homology with GnRH and can stimulate LH release from gonadotrophs.³⁹

The role of GnRH as a true neurotransmitter is not established. GnRH has been shown to elicit a slow excitatory synaptic potential in frog sympathetic ganglia.⁴⁰ Whether GnRH plays a modulatory role on other neurotransmitter systems remains to be determined.

Somatostatin

During the screening of multiple fractions from hypothalamic extracts, fractions with the ability to inhibit secretion of growth hormone (GH) from anterior pituitary cells were isolated. One factor subsequently isolated and characterized by Guillemin's group was a 14-amino acid peptide named somatostatin (Fig. 5).⁸ This peptide is derived from a 210-amino acid precursor.

Clinical studies of this peptide demonstrated that somatostatin inhibited GH, TSH, prolactin, glucagon, and insulin secretion in humans.⁴¹ Subsequently, somatostatin was found to be widely distributed in the gastrointestinal tract, the delta cells of the pancreas, and the thyroid-parafollicular cells. In the central nervous system, somatostatin has been localized to the hypothalamus, limbic system, septum, hippocampus, cortex, and medulla.⁴² The secretion of hypothalamic somatostatin is episodic and probably plays a role in the regulation of episodic GH secretion.⁴³ Because somatostatin is localized adjacent to other hormone-secretory cells, control of secretion can be affected by paracrine action. In the case of the thyroid, parafollicular cells that contain calcitonin and somatostatin regulation may occur through autocrine action.

In general, somatostatin injected intraventricularly in animals acts to decrease spontaneous motor activity and displays a sedative effect. The half-life of somatostatin in the peripheral circulation is only 2 to 4 minutes.⁴⁰

Mice overexpressing somatostatin have been generated.⁴⁴ These mice displayed motor hyperactivity, especially in females, supporting a role for somatostatin as a neurotrophic factor. Mice deficient in somatostatin receptor type 2 do not respond to the negative feedback of GH⁴⁵ and have high gastric acid secretion.⁴⁶

Corticotropin-Releasing Hormone

CRH was the first hypothalamic-releasing hormone to be recognized. However, it was not until 1981 that Vale and his colleagues first reported the isolation and characterization of ovine CRH as a 41-amino acid peptide.⁴⁷ Through genetic engineering techniques, the structure of human CRH was determined to be a 41-amino acid peptide that differed from the ovine hormone in 7 positions and was derived from a 196-amino acid precursor.⁴⁸ Using immunocytochemical techniques, CRH was found to be highly concentrated in the hypothalamus (i.e. paraventricular nucleus and parvocellular neurons) and pituitary stalk. Immunoreactive CRH was found in the posterior pituitary, thalamus, cerebral cortex, cerebellum, pons, medulla oblongata, and spinal cord.⁴⁹ CRH neurons are also capable of synthesizing and storing oxytocin and vasopressin. There is evidence that CRH, like GnRH, is released in regular episodic pulses from a pulse generator intrinsic to the hypothalamus (Fig. 7).⁵⁰ The physiologic significance of the pulsatile nature of CRH remains unknown.

In initial studies, intracerebroventricular injection of CRH in rats activated the sympathetic nervous system, which increased plasma levels of epinephrine, norepinephrine, glucose, glucagon, and insulin.⁵¹ These responses suggest that CRH is a principal mediator of ACTH/cortisol secretion during stress in addition to its ability to activate the sympathetic nervous system. CRH may also play a role in depression, because elevated concentrations of CRH-like peptides have been reported in the cerebrospinal fluid of depressed patients.⁵² CRH also can suppress secretion of GnRH-LH release in vitro⁵³ and in vivo.^54,55 This effect can be reversed by a CRH antagonist⁵³ or naloxone.⁵⁶ This effect remains controversial, because other studies have found no effect of CRH on LH pulsatility.⁵⁷

Outside the central nervous system, CRH has been localized to pancreas, stomach, duodenum, liver, lung, placenta, and adrenal gland. The molecular weight of immunoreactive CRH from these tissues suggests that it is similar to hypothalamic CRH.⁴⁹ The physiologic role of CRH in these tissues is unknown.

In clinical studies, the half-life of human CRH is 5 minutes when injected into the circulation.⁵⁸ When injected into humans, the ACTH-releasing effects of CRH are synergistically potentiated by vasopressin.¹⁷ These responses suggest that ACTH release may be modulated by peptide-peptide interaction.

Transgenic technology has been used to generate mice that overexpress CRH and mice deficient in CRH and its receptor. Mice overexpressing CRH show elevated ACTH and glucocorticoid and display the signs of Cushing's syndrome, including female infertility.⁵⁹ The CRH knockout mice are viable, reproduce, and have normal longevity despite severe glucocorticoid deficiency.⁶⁰ Vasopressin and oxytocin expression were not increased. Knockout mice have a decreased glucocorticoid response to stress, with males much more severely affected. The most interesting finding in CRH-deficient mice was the effect of CRH deficiency on the offspring; lack of lung maturation led to neonatal death within 12 hours of life. The mice could be rescued by the administration of glucocorticoid to the maternal drinking water.⁶¹ This is reminiscent of the beneficial effect of glucocorticoid on lung maturation to women in preterm labor. Mice deficient in CRH receptor type 1 also show glucocorticoid deficiency but display increased exploratory activity and decreased anxiety, suggesting a role for CRH in stress and the anxiety response.

CRH administration is limited to research and diagnostic applications. CRH may be useful in the differentiation between pituitary and nonpituitary Cushing's syndrome.⁶² The therapeutic potential for CRH lies in its involvement in memory, obesity, and Alzheimer's disease.⁶³ Additional diagnostic and therapeutic applications await development of long-acting analogs.

A new member of the CRH family, urocortin, was cloned from rat brain.⁶⁴ It was later described in humans, sharing 95% identity with the rat molecule.⁶⁵ This 40-amino acid peptide shares 45% identity with CRH and stimulates ACTH release in vitro and in vivo.⁶⁶ Urocortin has greater ability to bind and activate type 2 CRH receptors than CRH, suggesting that it may be the endogenous ligand for the type 2 CRH receptor.⁶⁴ When administered intravenously to rats, urocortin produces a significant reduction in blood pressure,⁶⁷ similar to the activity of urotensin, a closely related, vasoactive hormone in fish.

Growth Hormone-Releasing Hormone

The most recent hypothalamic-releasing factor to be characterized, sequenced, and synthesized is growth hormone-releasing hormone (GHRH). Because many brain peptides have been localized outside the central nervous system, it was not surprising that GHRH was isolated from two patients with pancreatic tumors that induced acromegaly. From these tumors, two peptides (40 and 44 amino acids long) were isolated that possessed GH-releasing properties.^68,69 Both have been shown to be equally potent and specific for GH release that is inhibited by somatostatin.^70,71 Subsequent reports indicate that both peptides are derived from a 108-amino acid precursor and are localized predominantly to the arcuate nuclei.

Studies in patients with acromegaly have demonstrated no significant different responses to GHRH,⁷² whereas in patients with hyperprolactinemia and galactorrhea responses to GHRH are blunted.⁷³ The discovery of this peptide has obvious clinical applications. Studies that use GHRH for the treatment of GH-deficient children are already in progress.⁷⁴ As with other hypothalamic peptides, long-acting analogs of GHRH may be an alternative to the use of synthetic human GH in the treatment for this condition. There are now several nonpeptide analogues of GHRH, such as MK-677, in clinical development for the stimulation of the GH axis.⁷⁵

Neuropeptide Y, Galanin, Substance P, and Angiotensin II

A number of other neuropeptides that are localized to the circumventricular nuclei have been shown to modulate release of prolactin, GH, and GnRH. To review the interaction of these peptides is beyond the scope of this chapter; however, four of the better-known neuropeptides are discussed.

Neuropeptide Y (NPY) is the most abundant neuropeptide in the brain. It is one of several gastrointestinal peptides that are also localized in the CNS. NPY shares many similarities with the pancreatic family of peptides, showing a high degree of homology to peptide YY and pancreatic peptide (Table 1). There is considerable evidence that NPY plays a significant role in the regulation of appetite, body weight, and reproduction. In this capacity, NPY may be an important factor in the complex interplay between nutrition and reproduction.

NPY is a 36-amino acid peptide and is the only member of the pancreatic peptide family with localization in the CNS. NPY has a CNS distribution that is similar to somatostatin and is localized to the many hypothalamic nuclei, including the arcuate nucleus, median eminence, amygdala, septum, cortex, and the hippocampus.^76,77 NPY receptors are also widely distributed in the CNS, especially the hypothalamus and median eminence.⁷⁷ Pharmacologic and molecular biologic studies have confirmed the presence of at least four subtypes of NPY receptor.⁷⁸

Considerable evidence supports a major role for NPY in the control of reproductive function. Initial studies demonstrated that NPY stimulated LH release from intact rats but inhibited LH release from ovariectomized animals. The stimulatory effect of NPY on LH release was later shown to be mediated by an increase in the hypothalamic release of GnRH, in estrogen-treated animals or tissues. NPY is secreted in an episodic manner that corresponds to GnRH pulsatility. There is evidence that NPY is involved in the pulsatile secretion of GnRH and in the burst of GnRH seen at the time of the LH surge.⁷⁹ There is also evidence that NPY influences LH secretion at the pituitary level. NPY has a pronounced inhibitory effect on sexual behavior.⁸⁰

NPY is recognized as the most potent appetite-stimulating substance in the brain. Injection of NPY into the brains of satiated rats consistently stimulates ongoing feeding behavior and the development of obesity. Evidence that NPY is an endogenous appetite stimulator came from the neutralization of NPY by administration of NPY antibody, which eliminated the normal feeding pattern of rats. It is likely that NPY interacts with other neurosecretory peptides, such as galanin and β-endorphin, to regulate feeding behavior.⁷⁹

It was expected that a mouse deficient in NPY would display alterations in feeding behavior and body weight, but this is not the case. NPY-deficient mice maintain a normal body weight and have normal reproductive function.⁸¹ In addition, the response of these animals to the newly discovered obesity hormone, leptin, is normal. Surprisingly, NPY-deficient mice have an increased propensity for seizure activity.⁸¹ It is likely that other neuropeptides, such as galanin, play a compensatory role in feeding and reproductive function in these animals.⁸²

Galanin is an important neurosecretory peptide. Originally isolated in 1983 in extracts of porcine intestine, galanin is composed of 29 amino acids in all species except humans, in whom it is 30 amino acids long.⁸³ The first 14 amino acids, which are associated with biologic activity, are fully conserved across 13 species in which galanin has been characterized.⁸³ The name galanin was derived from the first and last amino acids of the sequence: glycine and alanine. Galanin is recognized as a neurosecretory peptide, with wide distribution in the central nervous system. Most important is its secretion from the hypothalamus, which also has a high concentration of galanin receptors. There is evidence that hypothalamic galanin has physiologic roles involving feeding, cognition, gastrointestinal function, and regulation of hypothalamic and pituitary hormone secretion.

The galanin gene was cloned in 1987 from rat pituitary and hypothalamus.⁸⁴ Analysis of the galanin peptide suggests that it is unrelated to any other known families of neuropeptides, making it unique among neurosecretory peptides. Like all other neurosecretory peptides, the galanin receptor (GALR) is a member of the G protein-coupled receptors with seven transmembrane domains.⁸⁵ Galanin receptors are found predominantly in the brain in three regions: cerebral cortex, amygdala, and hypothalamus. The first GALR was cloned from a human melanoma in 1994.⁸⁶ There are three known GAL receptors, designated GALR1, GALR2, and GALR3, which differ considerably in primary structure and distribution.⁸⁵ Based on the effects of different galanin agonists and antagonists, it is likely that there are other galanin receptors.⁸³

Galanin has a well-recognized role in reproductive function. One of the first indications of this action was the demonstration that galanin expression is markedly stimulated by estrogen.⁸⁷ There is evidence that galanin is co-released with GnRH and that it modulates LH secretion from the pituitary. In this regard, galanin secretion is modulated by estrogen and testosterone levels and varies with the estrus cycle, peaking in proestrus with the LH surge.^88,89 Galanin stimulates the release of GnRH from the hypothalamus in a dose-dependent manner.⁹⁰ Based on these findings, it has been proposed that galanin expression may modulate the pulsatility of GnRH release.⁸⁸ The pituitary gland is another important site of galanin secretion, with expression stimulated by estrogen.⁸⁴ Galanin was also shown to stimulate prolactin release and proliferation of prolactin-secreting cells (i.e. lactotrophs).⁹¹

Galanin, in conjunction with NPY and other hypothalamic neuropeptides, has a well-characterized role in the regulation of feeding and body weight.^92,93 Hypothalamic galanin secretion is modulated by circulating hormones, such as insulin and leptin, and nutrient ingestion. There is evidence that galanin secretion from one particular site within the hypothalamus, the anterior paraventricular nucleus, mediates these effects.⁹⁴ Galanin responds to these metabolic signals by regulating energy balance through its effects on behavior, hormone secretion, and metabolism.⁹⁵ Galanin has orexigenic action, or stimulation of eating behavior. In this regard, galanin may act through β-endorphin, which also stimulates appetite.⁹⁶ Galanin is specifically linked to ingestion of fats, in contrast to NPY, which stimulates the ingestion of fats and carbohydrates.⁹⁵ Galanin not only stimulates feeding behavior, but also has effects on metabolism, acting to reduce energy expenditure. These effects tend to increase body weight, especially of adipose tissue.⁹³

A knockout mouse for the galanin gene has been developed and characterized.⁹⁷ The mice develop normally and can reproduce, indicating that galanin does not play critical roles in these physiologic processes. The mice have a normal body weight compared with wild-type littermates. The mice do demonstrate a significant reduction in pituitary prolactin content, supporting an important role of galanin in the development or maintenance of the lactotroph cell. Mammary development was also delayed. Affected mothers cannot lactate, so that the pups die shortly after birth if not transferred to a foster mouse.⁹¹ The stimulatory response of the lactotroph cell to estrogen is also abolished. The mice also demonstrate several neurologic abnormalities, including important changes in the responses of sensory neurons to injury and pain.⁹⁷

NEUROSECRETORY PEPTIDES AND THE LINK BETWEEN NUTRITION AND REPRODUCTION.

Two neurosecretory peptides, NPY and galanin, play major physiologic roles in nutrition and reproduction, suggesting that these peptides may be involved in the reproductive changes seen in disorders such as anorexia nervosa and the chronic anovulation of obesity. The most important part of this puzzle was the discovery of the obesity hormone leptin.⁹⁸ This important discovery provided the missing piece of the puzzle: how does the hypothalamus know how much fat tissue is present in the organism? Produced by fat tissue, circulating levels of leptin correlate directly with the amount of adipose tissue present.⁹⁹ Receptors for leptin are located in the hypothalamus, where leptin acts to regulate energy expenditure and body weight. In this model (the "lipostat" hypothesis of body weight regulation), leptin serves as the feedback signal to the hypothalamus that allows the animal to maintain body weight in a narrow range, despite daily fluctuations in food consumption and energy expenditure.¹⁰⁰

It is likely that the primary mechanism by which leptin modulates appetite and energy balance is through its effects on NPY, galanin, and possibly other neurosecretory peptides, such as melanin-concentrating hormone (MCH), opiates, and CRH.^100,101 Leptin receptor is coexpressed in NPY neurons,¹⁰² where leptin acts to decrease NPY mRNA and neural signaling.^103,104 Leptin also inhibits the expression of galanin in the hypothalamus.¹⁰¹ Although not yet proved, it is likely that these effects of leptin on NPY, galanin, and other neurosecretory peptides modulate the reproductive axis.⁹⁰ The requirement for both NPY and galanin is demonstrated in studies showing markedly decreased GnRH pulsatility when the activity of both neuropeptides is inhibited but only partial effect when only one is blocked.¹⁰⁵ This study helps to explain why mice deficient in either neuropeptide have normal weight. Taken together, the interaction of leptin, NPY, galanin, and other neuropeptides is beginning to shed light on the long-recognized link between nutritional status and reproductive function.

Substance P was one of the first nonhypothalamic-releasing hormones shown to affect pituitary function. It is related to the tachykinin family of neuropeptides and has been localized to the brain stem, amygdala, cortex, medulla, and hippocampus.¹⁰⁶ Receptors for substance P-like peptides are found in brain, pituitary lactotrophs (i.e. prolactin-releasing cells), and gonadotrophs.¹⁰⁷ In animal studies, administration of substance P to rats stimulates prolactin secretion and inhibits GH release.¹⁰⁸ Mice deficient in substance P or its receptor are viable and fertile but do show altered responses to pain.¹⁰⁹

Although angiotensin II is well known for its regulation of salt-water balance, this octapeptide is also found in the paraventricular and supraoptic nuclei. It is often colocalized with another osmoregulating hormone, vasopressin. With regard to its modulatory role on pituitary function, angiotensin II has been shown to stimulate LH release and inhibit prolactin release.¹¹⁰ Mice deficient in angiotensin II are fertile and show the predicted alterations in salt balance and blood pressure.¹¹¹ Male mice lacking angiotensin-converting enzyme show altered sperm mobility and decreased fertility.¹¹²

For many years, the existence of endogenous opiate compounds similar to morphine was postulated. However, it was not until 1975 that two endogenous opiate-like substances, methionine enkephalin and leucine enkephalin, were discovered by Hughes and coworkers.¹¹³ Subsequently, a much larger peptide with potent opiate activity was isolated by Li and was named β-endorphin.¹¹⁴ A comparison of the structure of this peptide indicated that β-endorphin was identical with the carboxyl-terminal sequence of β-lipotropin (β-LPH), a previously isolated protein that did not appear to have a specific function.¹¹⁵

These observations led to a series of elegant experiments by Eipper and Main.¹¹⁶ Using the pulse-chase technique for differential labeling of peptides, they were able to establish that β-endorphin and ACTH were major intermediate peptide products of a larger precursor molecule with a molecular weight of 31 kd. This precursor was subsequently named proopiomelanocortin (POMC). The sequence for POMC was determined by genetic engineering techniques.¹¹⁷ Further examination of the POMC mRNA revealed that differential processing of this precursor occurred in different tissues. For example, in the rat anterior pituitary lobe the POMC molecule is processed predominantly to ACTH and β-LPH. In the pituitary intermediate lobe, the hypothalamus, and placenta, ACTH and β-LPH are further cleaved to form α-melanocyte-stimulating hormone (α-MSH) and β-endorphin.¹¹⁸ A schematic representation of the POMC molecule and its known processed products is shown in Figure 7.

Since the discovery of the POMC family of peptides, three other classes of opioid peptides have been characterized: enkephalins, dynorphins, and the newest opioid peptide, nociceptin or orphanin FQ. The endogenous opiate peptides are all derived from posttranslational processing of larger precursor peptides (i.e. POMC, proenkephalin A, proenkephalin B, and pronociceptin), each of which are encoded by separate genes (Fig. 8). Preproenkephalin A is processed to yield four met-enkephalins, one leu-enkephalin, and two other modifications of met-enkephalin. Preprodynorphin, or preproenkephalin B, gives rise to dynorphins A and B, and neoendorphin. In a similar manner, nociceptin and perhaps other uncharacterized peptides are derived from pronociceptin.¹¹⁹

Opioid Receptors

The primary action of the opioidergic peptides is through binding to specific opioid receptors. In the past, pharmacologic studies suggested the presence of many different opiate receptor subtypes. Molecular biologic techniques have determined four major classes of opioid receptors, each arising from a separate gene.¹²⁰ The three major classes of opiate receptor are classified as μ, δ, and κ. A fourth opiate receptor, known as ORL-1, was cloned in 1994.¹²¹ It is likely that subtypes exist within these four major groups.

Originally localized to the vas deferens (hence the δ designation), the first opioid receptor to be cloned and definitively established was the δ receptor, also known as DOR and OP1.¹²² The δ receptor displays preferential binding of enkephalins. The highest δ receptor density is found in the olfactory bulb, neocortex, caudate putamen, and nucleus accumbens.¹²³ It is likely that there are two subtypes of δ receptors. Proposed functions of the δ receptor include analgesia, olfaction, motor function, cognition, and mood.

The second opioid receptor to be established was the κ receptor, named for its binding to a synthetic opiate ketocyclazocine.¹²⁴ Also known as KOR and OP2, this receptor is localized to the cerebral cortex and other sites within the CNS. Pharmacologic data support the existence of three subtypes of the κ receptor. Proposed functions include pain, appetite, diuresis, and neuroendocrine regulation.

The third major class of opioid receptor is designated μ for its affinity for morphine.¹²⁵ Other designations for the μ receptor are OP3 and MOR. Pharmacologic data are largely derived from studies on the guinea pig ileum, which has a high density of these receptors. The widely used opioid antagonist naloxone has a high affinity for the μ receptor, whereas the orally active naltrexone is less selective. The μ receptor is localized throughout the CNS, spinal cord, and peripheral nervous system, especially the gut. Binding studies support the existence of two subtypes of the μ receptor.¹²³ Believed to mediate the actions of morphine and its derivatives, the μ receptor generally exerts a depressant effect on many functions, including respiration, cardiac function, gastrointestinal function, locomotor activity, and memory.

In the search for opioid receptor subtypes, a fourth opioid receptor, initially designated an "orphan opiate receptor" or ORL-1, was identified.¹²¹ Despite a high degree of homology with the other opioid receptors, ORL-1 showed poor binding to known opioid compounds. Later, the endogenous ligand for ORL-1 was found and given the names nociceptin¹²⁶ and orphanin FQ.¹²⁷

Distribution

The POMC-derived peptides have a ubiquitous distribution and have been identified in the placenta, gastrointestinal tract, and brain (Table 4). Within the brain, the highest concentrations of β-endorphin have been localized to the basomedial and basolateral hypothalamus¹²⁸ from which neuronal processes project to the anterior hypo thalamus. Other areas where β-endorphin is distributed include the locus ceruleus (pons), periventricular nucleus, amygdala, and periaqueductal gray matter.

TABLE 4. Distribution of Proopiomelanocortin-Derived Peptides

  Brain

  Globus pallidus
  Nucleus accumbens
  Mesencephalon
  Amygdala
  Medial hypothalamus
  Locus ceruleus

  Pituitary gland
  Male reproductive tract

  Vas deferens
  Epididymis
  Seminal vesicles

  Female reproductive tract

  Ovary (corpus luteum)

  Placenta
  Gastrointestinal tract

  Stomach
  Pancreas
  Intestine

  Lung

Adapted from Krieger DT: The multiple faces of proopiomelanocortin, a prototype precursor molecule. Clin Res 31:342, 1983.

The enkephalin family of peptides is also widely distributed in the brain, with the concentrations of met-enkephalin exceeding leu-enkephalin by a 5:1 ratio. Areas that have high concentrations include medial hypothalamus, amygdala, mesencephalon globus pallidus, and the substantia gelatinosa of the spinal cord.¹²⁹

The third major opioid peptide group, dynorphin, has been localized to the supraoptic and paraventricular nucleus of the hypothalamus and the posterior pituitary.¹³⁰ Dynorphins also are colocalized with vasopressin, oxytocin, and CRH and are found within pituitary gonadotrophs.^131,132 The presence of all three groups of peptides in the brain reflects the complexity of the opioid system.

Nociceptin and its receptor are widely distributed in the central nervous system, where it has a primarily inhibitory effect. Nociceptin is highly expressed in the hippocampus, stria terminalis, amygdaloid nucleus, and selective thalamic nuclei.¹³³ The nociceptin receptor, known as ORL-1, is expressed in the diencephalon, mesencephalon, pons/medulla, and spinal cord.¹³⁴

Measurement in Biologic Fluids

Because POMC-derived peptides share similar amino acid sequences, antisera that are generated to individual peptides such as β-endorphin and β-LPH often cross-react with other POMC peptides. Measurement of these peptides requires confirmation by rigorous chromatographic characterization.¹³⁵ Peptidases are present in biologic fluids such as plasma, and further cleavage of the peptide to other opiate fragments can potentially take place. Current approaches to measurement of β-endorphin and β-LPH use extraction techniques such as C-18 minicolumns before chromatographic separation of peptides and radioimmunoassay.

In general, measurements of peripheral levels of β-endorphin and β-LPH parallel that of ACTH in pathologic states such as Cushing's syndrome or Nelson's syndrome. Under CRH stimulation, simultaneous elevations of β-endorphin, β-LPH, and ACTH are also observed.¹³⁶ Interpretation of changes in peripheral β-endorphin levels under basal conditions is difficult. Because opiate peptides do not cross the blood-brain barrier, changes in peripheral β-endorphin levels may not reflect central opioidergic activity. Wehrenberg and coworkers demonstrated that large changes in the concentration of β-endorphin in the portal blood are not accompanied by concomitant changes in peripheral β-endorphin levels.¹³⁷

Behavioral Effects

Beyond the analgesic effects of opiates, the behavioral effects in humans have not been well characterized. After one dose of morphine, human volunteers report a positive euphoric effect, mild nausea, giddiness, and a subjective sense of mental clouding. On the basis of animal studies, β-endorphin is 50 to 100 times more potent than morphine in achieving this effect. In contrast, injection of nociceptin into mice increases their sensitivity to pain (hence the name nociceptin).¹³⁸

Animal studies suggest that the opioidergic system may influence memory, blood pressure, temperature, feeding, and sexual activity. These findings are not surprising, because it appears that the opioidergic system can modulate other major neuron systems.¹²⁸

Opioid Modulation of Anterior Pituitary Function

Two approaches have been used to study the effects of opiates. In initial studies, morphine or opiate analogs were administered to animals or humans, and corresponding changes in hormonal levels were measured. These experiments were primarily pharmacologic and probably did not reflect physiologic changes that might be mediated by opiates. Later studies took a more physiologic approach using naloxone, a nonspecific opioid receptor antagonist, to assess the action of endogenous opiates. The effects of opiates can be determined with greater confidence using both approaches.

In early in vivo studies, β-endorphin or morphine was shown to stimulate GH, prolactin, and corticosteroid release when administered intracerebrally or intravenously. These effects were abolished by naloxone.¹²⁹ Under conditions in which morphine is given chronically, a tolerance to these effects is observed. From these findings, it is unclear whether opiates play a role in modulation of basal secretion of these hormones. The observed effects of opiates on anterior pituitary function in humans are summarized in Table 5.

TABLE 5. Proposed Actions of Opioid Peptides on Pituitary Function

  Acute stimulation of growth hormone release
  Acute stimulation of prolactin release
  Decrease in corticotropin release
  Decrease in thyroid-stimulating hormone secretion
  Inhibition of gonadotropin-releasing hormone release
  Inhibition of suckling-induced oxytocin release

Morley SE: The endocrinology of the opiates and opioid peptides. Metabolism 30:195, 1981.

Opioid Modulation of Neurotransmitter Function

The long-term effects of opiates have been characterized at the neuronal level. In general, the biologic activity of opiate peptides is inhibitory. Opiates have been shown to suppress the firing of adrenergic neurons presynaptically in morphineaddicted animals. Because of the decrease in adrenergic input at the synapse, a compensatory increase of adrenergic receptors occurs postsynaptically. If opiate inhibition is interrupted, normal firing of adrenergic neurons resumes in the face of increased sensitivity postsynaptically, resulting in overactivation of the sympathetic nervous system.¹³⁹ Observed effects include tremors, restlessness, irritability, sweating, and other symptoms typical of heroin withdrawal.

Opiate peptides also exert similar effects on the dopaminergic system. In isolated dopaminergic neurons, opiates inhibit depolarization.¹⁴⁰ In vivo studies by Gudelsky and Porter show a reduction in dopamine concentration in rat pituitary stalk plasma when morphine or β-endorphin is administered.¹⁴¹ These observations suggest that the increases in prolactin levels after exposure to βendorphin or morphine reflect opioid inhibition of dopamine activity, resulting in increased prolactin release.

Opiates in the Control of Reproduction

As in other anterior pituitary hormones, the predominant effect of opiate peptides on gonadotropin secretion is inhibitory. When morphine, human β-endorphin, or enkephalin analogs are administered to women, acute suppression of LH levels is observed.^142,143 If naloxone is used to block the action of endogenous opiates during the late follicular and mid-luteal phases of the menstrual cycle, increases in LH, the opposite effect,¹⁴⁴ occur. What then governs the activity of these endogenous opiate peptides?

Evidence suggests that the gonadal steroid environment may influence endogenous opiate activity. Measurement of β-endorphin concentrations in the pituitary portal blood in monkeys during the menstrual cycle show an increase in release of hypothalamic β-endorphin during late follicular and mid-luteal phases of the menstrual cycles.¹³⁷ In ovariectomized monkeys, increased hypothalamic β-endorphin release was observed only in monkeys with estrogen and/or progesterone replacement.¹⁴⁵ These experiments suggest that opioid activity may be influenced by the gonadalsteroidal environment.

In an analogous situation in humans, after naloxone administration during the menstrual cycle, there appears to be a progressive increase in LH release from the late follicular to the mid-luteal phase of the cycle, suggesting indirectly that endogenous opiate activity was suppressing gonadotropin levels.¹⁴⁴ In low-estrogen environments such as in menopausal women, naloxone infusions have no effects on LH levels. However, replacement with progesterone¹⁴⁶ or with estrogen and progesterone result in reestablishing a gonadotropin rise in response to naloxone.¹⁴⁷ Collectively, these observations suggest that endogenous opiate inhibition is mediated in part by the gonadal-steroidal environment.

At what point in development does functional coupling between the opiate and GnRH system occur? If we examine the hypothalamic-pituitarygonadal axis of the fetus, it is evident that the human fetal pituitary gland can secrete gonadotropins during early life in sufficient amounts to stimulate production of gonadal steroids. If fetal hypothalami are exposed to naloxone, increased GnRH release occurs, indicating that the suppressive effect of endogenous opiates is already established by 20 weeks' gestation.¹⁴⁸

On the basis of these observations, it is clear that the opioidergic system is a major regulator of GnRH secretion and may be involved in several clinical hypogonadotropic states. One example in which endogenous opiates have been implicated is the postpartum period. The puerperium is characterized by an extended interval of low gonadotropin levels. Administration of naloxone during the second postpartum week reverses this situation and is accompanied by significant elevations of LH and FSH, suggesting that endogenous opiate activity is in part responsible for gonadotropin suppression.¹⁴⁹ Two other clinical conditions in which hypogonadotropinism occur are anorexia nervosa and hypothalamic amenorrhea. In both conditions, blockage of endogenous opiate activity with naloxone results in elevations of gonadotropin.^150,151

Transgenic Mice and the Opioidergic System

Many of the opioid peptides and their receptors have been investigated by gene knockout experiments, providing new information on the physiologic role of opiates. For example, mice deficient in β-endorphin are viable and reproduce.¹⁵² They do not show abnormalities in the adrenal axis. These mice do respond to morphine but show an abnormal analgesic response to stress. However, mice deficient in the μ opiate receptor do not respond to morphine with an analgesic response,¹⁵³ and mice lacking the κ receptor¹⁵⁴ or the δ receptor¹⁵⁵ respond to morphine normally. This evidence supports a role for the μ receptor in mediating the actions of this drug.¹⁵³ Affected mice develop normally and reproduce. Altered pain responses under basal conditions in these mice support a role of opioidergic peptides in the endogenous response to pain. Other findings in the μ receptor knockout mice were changes in hematopoiesis, motor function, and sexual behavior.¹⁵⁶ Alterations in the function of the κ and δ receptor in mice deficient in the μ receptor support a physiologic interaction among the opioid receptors, as suspected from pharmacologic data.¹⁵⁵ Enkephalin-deficient mice are healthy, are fertile, and show normal maternal behavior. The mice are, however, anxious, aggressive, and show abnormal responses to pain.¹⁵⁷ Mice lacking the nociceptin receptor (ORL-1) have been developed. Surprisingly, these mice appear to have a normal response to pain, but have primarily hearing and memory deficits.¹⁵⁸

Although the genetically altered mice have provided important insights into our understanding of opioidergic peptides, these results certainly underscore the vast amount of opiate physiology that remains to be defined. It is apparent that we are just beginning to unravel the complex interactions of the opioidergic system. From these early studies, it appears that opiates play an inhibitory role in the control of gonadotropin secretion. This activity appears to be functionally coupled by gonadal steroids and is active during early fetal life. With our enhanced ability to isolate and characterize new peptides, the answers should be forthcoming.