| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Endocrine Research Unit (J.D.V.), Department of Internal Medicine, Mayo Medical School, Mayo School of Graduate Medical Education, General Clinical Research Center, Mayo Clinic, Rochester, Minnesota 55905; Department of Pediatrics (J.N.R.), State University of New York at Buffalo, Buffalo, New York 14214-3000; Pediatric Endocrinology (E.J.R.), National Children’s Hospital, San Jose 1000, Costa Rica; and Division of Endocrinology and Metabolism (C.Y.B.), Department of Internal Medicine, Tulane University Medical Center, New Orleans, Louisiana 70112-2699
Correspondence: Address all correspondence and requests for reprints to: Johannes D. Veldhuis, Endocrine Research Unit, Department of Internal Medicine, Mayo Medical School, Mayo School of Graduate Medical Education, General Clinical Research Center, Mayo Clinic, Rochester, Minnesota 55905. E-mail: Veldhuis.Johannes{at}mayo.edu
 |
Abstract |
|---|
 Top Abstract I. Regulation of the...II. Ensemble Regulation of...References |
|---|
|
I. Regulation of the GH and Gonadotropic Axes |
|---|
|
TopAbstractI. Regulation of the...II. Ensemble Regulation of...References |
|---|
Hypothalamic GHRH and somatostatin (SS) govern GH secretion by antagonistic and facilitative effects, depending upon the relative timing of their release (1, 7, 8, 9, 10, 11, 12, 13). Ghrelin, an acylated GH-releasing peptide (GHRP) synthesized in the stomach, anterior pituitary gland, hypothalamus, kidney, gonad, and placenta, induces GH secretion via combined hypothalamo-hypophyseal mechanisms, and concomitantly regulates appetite, insulin secretion, fat oxidation, and smooth-muscle contraction (14, 15, 16, 17, 18, 19, 20, 21, 22, 23). The foregoing three (minimal) peptides act as an ensemble, rather than individually, to supervise and respond adaptively to time-varying production of GH and IGF-I (1, 9, 10, 11, 12, 21, 24, 25, 26, 27) (Fig. 1
). The stimulatory efficacy of any one of GHRH, GHRP, or L-arginine (to deplete SS) depends upon gender, body composition, age, nutrient status, and underlying disease (1, 27, 28).
|
Rare mutations of the human GH gene encode dominant-negative proteins that block its processing, release, and/or effects, thereby leading to short stature (34). Biosynthetic analogs of GH have been designed to silence GH-receptor signaling (29, 35). The clinically available antagonist, pegvisomant, harbors 8-amino acid substitutions in high-affinity binding site 1 and one in dimerization site 2, thus disabling intracellular signaling despite binding, dimerizing, and internalizing the receptor (30, 36). Polyethylene glycol residues were affixed to reduce immunogenicity and extend the plasma half-life to 72 h compared with 16 min for native GH (37). Pegvisomant administration in individualized doses normalizes IGF-I concentrations in 95% of patients with acromegaly (38). GH-receptor antagonists also provide probes of the role of GH in physiology and pathophysiology. For example, pegvisomant unleashes GH secretion in young adults by feedback withdrawal without altering GH clearance (37, 39) (Fig. 2). The same agent limits the degree of diabetic glomerulopathy and inhibits growth of colorectal carcinoma in mice (36, 40).
|
Rare recessive mutations of the human GH receptor (Laron dwarfism) cause primary insensitivity to GH, IGF-I depletion, growth retardation, and hypersecretion of GH (43). Treatment with recombinant human (rh) IGF-I decreases GH concentrations and ameliorates certain features of the syndrome (44, 45). Perhaps unexpectedly, profound in utero GH deficiency restricts birth size only minimally by reducing longitudinal bone growth in the third trimester (46). Growth continues nearly normally for 6 to 10 months postnatally, and then declines progressively in childhood. Presumptively GH-independent stimuli in prenatal and early neonatal life include IGF-I, IGF-II, insulin, T4, glucocorticoids, and metabolic substrates (47, 48, 49, 50). Although IGFBP-1, -3, and -5 and novel IGFBP-related peptides that bind insulin both inhibit and stimulate cellular proliferation, their roles in human growth have not been elucidated (6, 51). These issues highlight major knowledge deficits in the GH-independent control of intrauterine and neonatal growth. In relation to this issue, Table 1 outlines inferred differences between GH and IGF-I actions observed experimentally.
|
and IL-1/-6, which induce proteases, stimulate cortisol synthesis, suppress GH secretion, inhibit IGF-I production via SOCS activation, and decrease gonadal and adrenal androgens (6, 55, 56, 57, 58). The consequence is marked catabolism without countervailing anabolism.
2. IGF and IGFBP system
a. Insulinomimetic peptides.
The IGF system comprises insulinomimetic peptides, binding proteins, receptors, and proteases (2, 6, 51, 59, 60, 61). IGF-I and IGF-II exert mitogenic, cytodifferentiative, and antiapoptotic actions in the fetus via receptors designated IGF type I, insulin isoform A, and hybrid IGF-I/insulin isoform A (47, 59, 60). In contrast, IGF type II receptors mediate sequestration, cellular uptake, and degradation of IGF-I and IGF-II. Excessive amounts of soluble IGF-II receptor inhibit hepatocyte proliferation in vitro and bladder smooth-muscle development in vivo (49).
Both GH-dependent and GH-independent (organotypic) mechanisms regulate IGF-I peptide and receptor expression in all major organs (6, 51, 59, 60, 62, 63). Figure 3 illustrates the diversity of organotypic IGF-I regulation. For example, estradiol induces synthesis of IGF-I and cognate receptor in breast, uterus, ovary, pituitary, brain (neurons and glia), and bone; testosterone acts analogously in muscle and bone; PTH does so in bone; LH and FSH in the ovary and testis; and TSH in the thyroid gland (6, 51, 60, 64, 65). The effects of organotypic hormones on local IGF-I availability are not strictly independent of GH, because systemically delivered GH comodulates tissue concentrations of IGF-I, IGFBPs, and proteases.
|
Targeted disruption of hepatic IGF-I gene expression in two murine models reduces systemic IGF-I concentrations by 60–80% and elevates GH concentrations by 2- to 10-fold (68, 69). The transgenic Cre/loxP models exploited postnatal activation of the hepatic albumin and 
-interferon promoters driving the Cre recombinase enzyme to trigger scission of loxP (P1 bacteriophage) DNA sequences flanking the native IGF-I gene. Despite marked depletion of total IGF-I concentrations, transgenic animals achieve 94% of expected adult size. However, liver and kidney size are increased and decreased, respectively; osteopenia occurs in adulthood; insulin action is impaired in muscle; and hyperlipidemia develops despite decreased total body fat (70). Which of the foregoing outcomes reflect a primary reduction of total IGF-I vis a vis a secondary increase in GH concentrations has not been defined.
Genetic inactivation of either IGFBP-3 or ALS (acid-labile subunit contained in ternary IGF-I/IGFBP-3/ALS) lowers IGF-I concentrations by 65%, yet limits growth minimally and does not amplify GH secretion (71). Normal GH production may reflect the capability of free IGF-I, binary IGF-I/IGFBP-3, and binary IGF-I/IGFBP-1 complexes to undergo capillary endothelial transcytosis into tissues, thereby mediating negative feedback (51, 72). Conversely, overexpression of IGFBP-3, but not of nonlimiting ALS, retards somatic growth putatively by sequestering plasma IGF-I in a stable ternary complex (73) (Table 2).
|
). Salient inferences are that: 1) IGF-II stimulates early embryonic development and maintains placental size; 2) IGF-I and IGF-II together promote later organ growth; and 3) parental genetic imprinting controls IGF-II availability (59, 74). Although direct examination of these principles is not possible in the human, observational studies correlate amniotic-fluid and umbilical-vein total and free IGF-I (but not IGF-II) concentrations with live birth weight (2, 6). Murine data in aggregate suggest that approximately 14% of adult somatic size is determined by GH alone; 17% by each of IGF-I or IGF-II; 35% by IGF-I and GH acting jointly; and 17% by none of the three peptides (59, 68, 75) (Table 2
).
|
). Rare IGF-I receptor mutations have been recognized clinically in infants with intrauterine growth retardation (77, 78). A partial defect in IGF-I receptor signaling also exists in Efe pygmies in Zaire, whose mean height at age 6 yr is –4 SD values by Western-European predictions and in adulthood is 2 feet 4 in. (0.7 m) (79, 80). Disruption of murine genes encoding both type I IGF receptor and IGF-II peptide reduces birthweight by 70% (compared with respective individual reductions of 55 and 40%) (60). Inactivation of the insulin receptor does not decrease live birthweight in this species (81). In contrast, mutation of the human insulin receptor (classical isoform B) is associated with the syndrome of leprechaunism typified by marked insulin resistance, hyperlipidemia, growth impairment, and premature atherosclerosis (82). Combined mutation of the IGF-I and insulin receptors in mice decreases growth by 70% (compared with 50% and no growth failure, respectively) (60, 81). Experiments motivated by this disparity identified a hybrid IGF-I/insulin isoform A receptor, which transduces fetal growth-promoting actions of both IGF-I and IGF-II (47, 81). In mice, insulin-receptor deletion in both ß-cells and hepatocytes or neurons, but not in adipocytes or myocytes individually, impairs adult glucose tolerance, thus indicating partial redundancy in the mechanisms of glucose homeostasis (81, 83).
The type II IGF receptor does not signal growth, but rather directs lysosomal degradation of both IGF-I and IGF-II (84). Silencing the murine IGF-II receptor elevates birthweight by 16% and induces polydactyly, organomegaly, cardiac anomalies, hypoglycemia, edema, and malignancies putatively due to accumulation of insulinomimetic peptides. Severity of the fetal-overgrowth syndrome is ameliorated, but not abolished, by concomitant genetic disruption of either the IGF-I receptor or IGF-II peptide (84) (Table 2).
Transgenic knockout of insulin-receptor substrate (IRS)-1 reduces growth by about 50%, consistent with the fact that IRS-1 transduces cellular signaling by IGF-I, IGF-II, and insulin (81, 85). Significant residual growth motivated the discovery of other related growth factor-receptor signaling messengers, IRS-2, IRS-3, and IRS-4. IRS-2 mediates insulinomimetic peptide drive of satiety and ovarian folliculogenesis (86). In principle, genetic variations in the structure of growth-promoting peptides and their receptor-effector pathways may contribute to currently unexplained growth retardation and populational growth differences in stature (87).
c. Insulinomimetic peptides in puberty.
Isolated hyposomatotropism including GH-resistant Laron syndrome delays the onset and slows the tempo of sexual maturation (43, 44). Although the clinical pathophysiology is not clear, experimental studies indicate that IGF-I administration can: 1) advance the timing of puberty in the female monkey and male or female rat (88, 89, 90); 2) augment secretion by and proliferation of GnRH neurons (91); 3) potentiate GnRH-stimulated LH secretion in vitro (92, 93); 4) synergize with LH and FSH in driving the proliferation and cytodifferentiation of gonadal cells, albeit with wide species differences (94, 95); and 5) enhance human chorionic gonadotropin (hCG)-induced ovulation in the rabbit and oocyte cleavage in the mouse (94). Figure 5 highlights these loci and mechanisms. How such laboratory insights relate to human pubertal development is not established.
|
Biological actions of GH and IGF-I differ in several respects (Table 1). For example, both GH and IGF-I increase whole-body protein synthesis and inhibit protein degradation, albeit with different molar efficacies; GH but not IGF-I induces lipoprotein (a), ALS, and IGF-I production; represses muscle IGFBP-4 expression; increases IGFBP-3 protease activity; and stimulates lipolysis; and IGF-I but not GH represses insulin secretion acutely (2, 5, 43, 44, 45, 96, 99, 101).
d. Roles of IGFBP in modulating growth.
Six principal IGFBPs sequester and transport IGF-I and IGF-II and/or stimulate cellular proliferation directly (6, 51). Distinct characteristics confer diversity of control, viz.: 1) IGFBP-1, -2, -3, and -5 bind IGF-I especially, IGFBP-6 binds IGF-II preferentially, and IGFBP-4 binds IGF-I and IGF-II equally; 2) all IGFBPs can block IGF-I actions; 3) IGFBP-1, -3, and -5 stimulate cellular proliferation in vitro; 4) GH induces hepatic and skeletal IGFBP-2, -3, and -5 synthesis, inhibits IGFBP-1 gene transcription, and does not affect IGFBP-4 and -6 concentrations; 5) ovarian and endothelial cells express all except, respectively, IGFBP-6 and IGFBP-1; and 6) IGFBP-1 combines with IGF-I in a high-affinity binary complex, whereas IGFBP-3 and -5 associate with ALS and IGF-I in stable ternary complexes.
The sequestrative effect of IGFBP-3 may have therapeutic implications. An experiment in the ovariectomized rat showed that prolonged administration of binary IGF-I/IGFBP-3 complexes can increase skeletal mass and tensile strength (102); and a short-term clinical study in postmenopausal women reported positive trends in hip bone mineral density (103). Combined peptide injections in patients with type I diabetes mellitus decreased insulin requirements by 50% without increasing hypoglycemia (104). The long-term safety, efficacy, practicability, and indications for combined-protein administration are not known.
IGFBP-1 unlike IGFBP-3 is regulated rapidly (within 1–2 h) by insulin, GH, and glucocorticoids (inhibitory) and estrogen and IGF-I (stimulatory) (51, 105). Posttranslational phosphorylation of IGFBP-1 enhances its affinity for IGF-I and IGF-II, yielding a stable binary complex in plasma. IGFBP-1 reduces dialyzably free IGF-I concentrations in vitro and attenuates hypoglycemia and organ growth in vivo (106). Strongly reciprocal associations between the two proteins are reported in: 1) fetal life, puberty and obesity (low IGFBP-1 and high free IGF-I); and 2) fasting, renal failure, and type I diabetes mellitus (high IGFBP-1 and low free IGF-I). The fact that hyperinsulinemia represses IGFBP-1 production explicates several of the foregoing observations.
IGF-I, IGF-II, and insulin concentrations are also determined by systemic hormones (e.g., testosterone, estradiol, T4, and glucocorticoids) and nonendocrine factors (e.g., hepatic metabolism, renal filtration, the soluble IGF-II receptor, and endothelial IGFBPs) (107, 108, 109). Pregnancy illustrates such multifactorial control. Maternal concentrations of IGFBP-1 rise, and those of IGFBP-2, -3, -4, -5, and -6 fall, thereby increasing free but not total IGF-I concentrations due to high plasma protease activity, hyperestrogenism, insulin resistance, and secretion of placental variant GH (GH-V), which is a potent somatotropin (6, 51, 61).
e. Sex steroids and the IGF/IGFBP system.
Conjoint activation of the somatotropic and gonadal axes in puberty promotes growth synergistically via independent and concerted mechanisms (1, 27). Sex steroids govern the expression, reception, and degradation of GH and insulinomimetic signals and, conversely, GH and insulinomimetic peptides up-regulate androgen and estrogen receptors (AR and ER) in an organ-specific manner (109, 110) (Table 3). For example, in the rat estradiol 1) induces synthesis of IGF-I and its receptor and alters availability of selected IGFBPs in the central nervous system (CNS), pituitary gland, ovary, breast, uterus, bone, and preadipocytes while inhibiting IGF-I production by the liver; and 2) stimulates the hepatic (subtype 1) GH receptor but represses that in the hypothalamus (62, 64, 65, 111, 112). Testosterone and nonaromatizable androgens induce IGF-I peptide in bone, muscle, and skin, but suppress the type I IGF receptor in fat (109, 112). Thus, the type of sex steroid and the target tissue determine regulation of GH and IGF-I peptides and receptors.
|
Clinical and research endpoints of GH replacement include linear growth and pubertal progression in childhood; well-being, mood, physical vigor, sexual energy, and sleep quality; lean-body mass and isokinetic or isometric muscle strength; physical endurance and maximal oxygen consumption (aerobic capacity); epiphyseal maturation, bone mineral density, and serum markers of skeletal remodeling; total body fat and regional adiposity; blood pressure, arterial dilatation, cardiac output, and ventricular-wall thickness; extracellular fluid volume, renal size, and glomerular filtration rate; lipoproteins and triglycerides; glycated hemoglobin and glucose; total energy expenditure and basal metabolic rate; TSH, T4, and T3 concentrations, because GH facilitates T4
T3 conversion. The multiplicity of pragmatic and research endpoints underscores the fact that no one biochemical measure is uniquely suited to gauge the full repertoire of GH activity (87, 96, 98, 113, 114).
Linear growth, energy expenditure, body composition, and pubertal development are also influenced by rare mutations and putative polymorphisms of an array of genes variously encoding: 1) receptors for androgen, estrogen, T4, LH, FSH, glucocorticoid, GH, IGF-I, ACTH, ghrelin, melanocortins, ß-adrenergic agonists, serotonin, insulin, and leptin; and 2) peptides such as GH, IGF-I, LH ß-subunit, ghrelin, leptin, and carboxypeptidase E (1, 115, 116, 117, 118, 119, 120). Thus, precise prediction of growth responses to GH, IGF-I, and/or sex steroids is difficult in individual patients.
B. The hypothalamo-pituitary-gonadal axis
1. In utero
a. Testis.
Human Leydig and Sertoli cells secrete testosterone and inhibin B, respectively, in the second month of fetal life (120). In mice, fetal androgen production is driven by placental hCG acting via Leydig-cell LH receptors, as modulated by insulinomimetic peptides, cytokines, and testis-specific factors (121). Activating mutations of the human LH receptor produce gonadotropin-independent male isosexual precocity, whereas inactivating mutations cause micropenis, hypospadias, genital ambiguity, or severe male pseudohermaphroditism (122, 123). In contrast, silencing the murine LH receptor does not impair prenatal Leydig-cell function (124). The critical role of maternal hCG is affirmed by clinical observations that hypogonadotropic hypogonadism (e.g., Kallmann’s syndrome), mutations of the GnRH receptor or LH ß-subunit gene, pituitary agenesis, and anencephaly do not impair in utero masculinization (120). However, such conditions are associated with a higher incidence of cryptorchidism, putatively due to reduced stimulation of gubernacular muscle by insulin-like 3 (relaxin-like factor) and/or less activation of innervating motoneurons by androgens (125).
Beyond genetic determination of the testis by SRY and SOX9 transcriptional factors (126), gonadal testosterone biosynthesis requires steroidogenic factor 1 (SF-1) and its transcriptional partner DAX (a dosage-sensitive sex-reversal gene product) to induce requisite oxidoreductase enzymes and sterol transport proteins; the Wilm’s tumor suppressor gene and stem-cell factor (c-kit ligand); AR, ER, LH ß-subunit and the LH and peripheral benzodiazepine receptors; selected cytokines and insulinomimetic peptides; and an enlarging array of transcriptional proteins (122, 127, 128, 129, 130, 131). Experimental studies suggest that intragonadal estrogen influences early Leydig-cell development, although this sex steroid inhibits steroidogenesis in the adult animal (132). Transgenic data illustrate that Muellerian-inhibiting substance, signaling via the eponymous Leydig-cell receptor, also modulates steroidogenesis (133).
Concentrations of LH and gonadal steroids in the fetus decline markedly before birth, but are higher in cord blood of boys than girls (120). Male but not female primates exhibit a prominent neonatal surge in LH, FSH, testosterone, and inhibin B production in the presence of a functional AR (134, 135). The surge begins within several hours of birth, peaks at 6 to 8 wk of age, and wanes gradually over 4 to 6 months (134, 136, 137). Testosterone concentrations often exceed 300 ng/dl (10.4 nmol/liter), and LH pulse frequency and amplitude are indistinguishable from values in young men (137) (Fig. 6). The clinical import of this event is not clear.
|
2. Adrenarche.
Adrenarche is defined phenotypically by the appearance of pubic and axillary hair and increased activity of apocrine sweat glands, and biochemically by increased adrenal synthesis of dehydroepiandrosterone, dehydroepiandrosterone-sulfate, and androstenedione but not cortisol (141, 142, 143). Expression of 3 ß-hydroxysteroid dehydrogenase type II declines in the zona reticularis concomitantly (144). Adrenarche is limited to the primate, thus stymying mechanistic investigations. However, a recent analysis associated mutation of a pituitary-specific transcription factor, PROP-1 (prophet of pit-1), with impaired progression of adrenarche (145).
Adrenarche occurs at or after age 7 yr and 8 yr in African-American and Caucasian girls, respectively, and typically at or after age 9 yr in boys (141, 146). Precocious adrenarche is more common in girls than boys and in Caribbean-Hispanic children. Epidemiological studies associate premature adrenarche with intrauterine growth retardation, prepubertally reduced ovarian volume, peripubertal onset of hyperandrogenism, elevated IGF-I concentrations, and insulin resistance in muscle (147, 148, 149). This constellation mimics features of polycystic ovarian syndrome (POCS). Administration of metformin or a thiazolidenedione in the adolescent and young adult with PCOS ameliorates hyperinsulinism, subfertility, oligo- and amenorrhea, and hyperandrogenism (150, 151). However, the long-term safety of insulin-sensitizing drugs in pubertal girls and the alexiteric impact of these agents on adult cardiovascular events are not known.
Certain polymorphisms of the insulin and insulin-receptor genes predict premature adrenarche and risk of pubertal hyperinsulinism, hyperthecosis, and PCOS (152, 153). Of relevance, IGF-I, insulin, and IGF-II all stimulate adrenal and ovarian androgen secretion in vitro (154, 155, 156). In relation to theca cells, IGF-I or insulin and LH act synergistically (94). Synergy between LH and insulin has import clinically in that hyperinsulinemia coexists with increased pulsatile LH secretion in most patients with PCOS, and in vitro synergism is blocked by troglitazone (156, 157, 158). In one transgenic mouse model, pituitary-directed overexpression of a hybrid LH/hCG gene resulted in persistent anestrus, concomitant adrenal LH-receptor expression, depletion of primordial ovarian follicles, and theca-cell hyperplasia (159). At present, the basic cause of excessive adrenal and ovarian androgen secretion and deficient muscle insulin signaling in PCOS remains undefined.
3. Gonadarche.
Gonadarche requires orderly activation of the hypothalamo-pituitary-gonadal axis (120, 160, 161). Although the fundamental initiators remain unknown, adrenarche and gonadarche can proceed independently, viz.: 1) adrenarche unfolds in Turner syndrome and under GnRH analog therapy when gonadarche is repressed; and 2) gonadarche progresses with minimal delay in children with adequately treated primary Addison’s disease (141, 142).
Accurate quantitation of gonadal sex-steroid and gonadotropin secretion in childhood requires sensitive assays of estrogen, testosterone, LH, and FSH. A recombinant estradiol receptor-based bioassay documented higher estrogen concentrations in girls than boys prepubertally (162). Ultrasensitive gonadotropin assays have unveiled pulsatile, orderly (low-entropy), and 24-h rhythmic release of LH in prepubertal children (161, 163, 164). In one study, combined application of immunofluorometry, overnight (15-min) blood sampling, and deconvolution analysis disclosed 30-fold amplification of LH secretory-burst mass and 1.8-fold acceleration of LH pulse frequency in late-pubertal compared with prepubertal boys (164). In another analysis, joint use of immunochemiluminometry, daytime (5-min) sampling, and deconvolution analysis delineated 4.2- and 7.5-fold augmentation of LH secretory-burst mass in late- vs. prepubertal boys and girls, respectively, with no detectable increase in LH pulse frequency (163) (Fig. 7). Thus, the amount of LH secreted per burst rather than pulse frequency is the primary locus of pubertal control.
|
). Silencing or stimulation of KiSS1/GPR54 mutes or activates, respectively, GnRH outflow to gonadotropes in the rodent and primate (168, 169, 170). Undernutrition and sex steroids decrease KiSS1 gene transcripts and peptide, indicating the operation of both adaptive and feedback regulation (171, 172). Although activated in primate puberty, what initially drives or disinhibits the KiSS1-GPR54 pathway in early gonadarche is not known.
|
Molecular cloning has identified more than 16 isoforms of GnRH represented among diverse organisms (182). GnRH II (sequenced initially in the chicken) is expressed in the rodent, monkey, and human, and increases in puberty in the primate hypothalamus (183, 184). Transcripts for GnRH-II and cognate receptor are also detected in gonadotropes, the limbic system, ovary, placenta, prostate gland, kidney, and bone marrow (183, 185). GnRH II stimulates LH and FSH secretion in vivo in the primate (186). Actions of GnRH II on limbic neurons may mediate sexual arousal (184, 187, 188), whereas those on granulosa-luteal cells may subserve steroidogenic inhibition (185). GnRH III (first identified in lamprey) may release FSH preferentially over LH in the rat (182).
b. LH secretion in childhood.
In the rodent, sheep, and monkey, hypothalamic GnRH outflow rises in early gonadarche, remains elevated in adulthood, and declines gradually in senescence (166, 174). In the human, LH concentrations reach a nadir at 6 to 8 yr of age before increasing multifold in puberty (161, 163, 164). Postulated bases of developmentally amplified GnRH outflow include the "gonadostat" and "centrally mediated" models.
c. Sex-steroid control of GnRH and LH outflow: gonadostat model.
The gonadostat theory posits that a decline in hypothalamo-pituitary sensitivity to gonadal-steroid negative feedback initiates puberty (Fig. 8). Albeit relevant to the rodent and ruminant, the gonadostat notion does not apply to the monkey and human. In particular, pathological depletion of sex steroids fails to unleash adult castrate-like LH secretion in children with inactivating mutations of LH, FSH, estrogen, or ARs or aromatase or 5 -reductase type 2 enzymes, congenital Leydig-cell agenesis, bilateral anorchia, or gonadal dysgenesis (120, 122, 189, 190, 191). Moreover, in the primate heightened resistance to sex-steroid restraint appears only after the onset of puberty is established (90, 166, 191, 192). The emergence of feedback-induced orderliness of LH secretion after the onset of puberty was verified recently in boys by approximate-entropy analysis (161) (Fig. 9).
|
-amino butyric acid (GABA)] receptors; 3) modulate hypothalamo-pituitary expression of AR and ER and steroid-metabolizing enzymes (194, 196, 199); and 4) govern transcription of genes encoding GnRH (negatively and positively by estrogen), GnRH receptor (positively by estrogen), and LH ß-subunit (negatively by both androgen and estrogen) (200, 201, 202).
Molecular studies establish that single GnRH neurons in the adult rodent contain gene transcripts for ER- (>65% of neurons) and ER-ß (>8% of cells) (198). Germ-line mutations of ER- (-ER knockout) or steroid-receptor coactivator proteins (SRC-1 and SRC-3) elevate LH concentrations by 2- to 3-fold in the adult male mouse (202, 203). Inactivation of the murine ER-ß gene does not affect LH concentrations, but impairs fertility due to defective ovulation and obstructed efferent spermatogenic ductules (197). Nonetheless, facile interpretation of LH hypersecretion in the -ER knockout model is confounded by: 1) unabated or amplified production of sex steroids by the gonad, adrenal gland, and CNS; 2) unopposed activities of AR and ER-ß; 3) loss of interactions between ER- and AR- or ER-ß; 4) altered metabolism of sex steroids; and 5) secondary neuronal and glial synaptic changes (189, 193, 194, 196, 204, 205). These issues highlight the challenge of valid clinical integration of multiple mechanisms of in vivo control of GnRH outflow.
d. Neurally mediated (central) hypothesis of gonadarche.
The neurally mediated hypothesis purports that gonad-independent activation of neuronal GnRH outflow initiates puberty (166). The categories of mechanisms thereby subsumed include at least: 1) enhanced coupling among excitable GnRH neurons; 2) potentiation of noradrenergic and excitatory amino acid (e.g., N-methyl-D,L-aspartate, NMDA) drive to GnRH neurons; and 3) muting of GABAergic and neuropeptide Y (NPYergic) inhibition of GnRH neurons (206, 207, 208, 209, 210). In a simplified model, augmented GnRH output would reflect facilitation of GnRH-neuronal coupling, amplification of excitatory inputs, and quenching of inhibitory inputs (Fig. 10).
|
GABA and NPY enforce either inhibition or excitation depending upon receptor subtype, neuroanatomic site, species, age, and sex-steroid milieu (206, 207, 209). GABA-A receptors transduce inhibition of the migration and secretion of GnRH neurons in the rat, guinea pig, sheep, and monkey (120, 206, 207, 208, 212). In the rodent and primate, suppressive effects of GABA and NPY recede, whereas excitative actions of the same effectors emerge, in early puberty (206, 208, 209, 210, 212). The fundamental factors that govern this counterpoise in neurotransmitter outflow and action are not established. One swich-like locus is the enzyme glutamic-acid decarboxylase (GAD67), which converts an activating neurotransmitter, glutamic acid, to an inhibitory ligand, GABA (206). Once elevated pubertally, sex steroids modulate GABA, glutamine, and NPYergic signaling (166) (Fig. 10). For example, feedforward by estrogen and progesterone before the LH surge represses, whereas feedback by testosterone in the adult male rat augments, GABA-A inhibition. In some children, anticonvulsant treatment with the GABA-A agonist, valproic acid, is accompanied by delayed puberty or clinical features of PCOS (214, 215). The pathophysiologies underlying these associations are not clear.
Proopiomelanocortin (POMC)-derived peptides, such as ß-endorphin, met-enkephalin, and -MSH, suppress GnRH neurons only in the adult human, monkey, rodent, goat, and sheep. Relief of opiatergic restraint thus does not mediate pubertal onset (216). Testosterone and estradiol enforce POMCergic inhibition of GnRH secretion after puberty, presumptively via ER- expressed in POMC neurons (217).
e. Integrative issues.
GnRH outflow is modulated by an array of countervailing signals, in addition to sex steroids, excitatory amino acids, GABA, NPY, and POMC derivatives. Other effectors include glucose, T4, glucocorticoids, leptin, ghrelin, TRH, neurotrophic peptides like TGF-, the cocaine and amphetamine-regulated transcript (CART), dopamine, norepinephrine, serotonin, CRH, urocortins, nitrous oxide, melanin-concentrating hormone (MCH), hypocretin (orexin), arginine vasopressin (AVP), vasoactive intestinal polypeptide, galanin, neurotensin, neurokinin B, insulin/IGF-I/IGF-II, ciliary neurotrophic factor, certain cytokines, and prostaglandin E2 (92, 120, 166, 218, 219, 220, 221). Their aggregate actions have not been parsed. In addition, how a burgeoning repertoire of intrapituitary regulators supervises pubertal development is largely unknown (222).
f. Late puberty and early adulthood.
Menarche occurs at Tanner stage 3 and early stage 4 of breast development in girls of Western European descent, corresponding to a chronological age of approximately 12.7 (± 1.1) yr. Adult reproductive function is not achieved immediately. A recent 4-yr longitudinal study in 112 girls showed that 10 or more ovulatory menstrual cycles per 12 months develop within 1 and 3 yr of menarche in 65 and 90% of adolescents, respectively (223). Normal luteal-phase progesterone concentrations also emerge only several years after onset of menses. Female fecundity and fertility require feedback-sensitive regulation of LH and FSH secretion, cyclical follicular development, surge-like secretion of LH, rupture of a dominant follicle, oocyte maturation, and adequate estradiol and progesterone secretion to ensure endometrial receptivity to blastocyst implantation (165). Given such ensemble complexity, an unresolved question is whether subtle disruption of integrative control underlies the pathophysiology of syndromes like inadequate luteal phase, luteinized unruptured follicle, stress-associated amenorrhea, and hyperandrogenemic oligovulation (37, 224).
Gene-silencing experiments in mice establish that aromatizable androgens imprint adult masculine behavioral traits via ER--dependent pathways, conditional on the SRC-1 and SRC-3 proteins (196, 203). The impact of the neonatal sex-steroid milieu on sexual behavior in humans is far less clear, given the importance of culture-specific psychosocial cues (225, 226). One 46 XY patient with an inactivating mutation of the aromatase enzyme had normal male gender identity and heterosexual orientation (227). Genetic defects in the AR (testicular-feminization syndrome) may alter childhood gender orientation subtly (226). In the Dominican Republic, 46 XY children with 5 -reductase type 2 deficiency, albeit reared as girls due to pseudohermaphroditism, may adopt a male gender role at puberty (225).
g. Timing and tempo of puberty.
An incompletely defined composite of genetic, environmental, nutritional, metabolic, and growth-related factors supervises the onset and progress of puberty. The gradual decline in age of menarche over the last century in the Western hemisphere illustrates such interactions. Observations that nutrient sufficiency allows and caloric deprivation impedes age-specific outflow of reproductive hormones sparked the prediction that somatic signals communicate the status of energy balance (the difference between availability and utilization of metabolic substrates) to integrative CNS loci. Brain signals in turn regulate behavior, appetite, energy expenditure, growth, and sexual maturation (228, 229, 230, 231, 232, 233). The notion is subsumed thematically under the somatometer hypothesis.
h. Somatometer hypothesis: somatic signals trigger pubertal onset.
Somatic inputs to the CNS include (but are not limited to) adipocyte-derived factors, such as leptin (elevated in obesity), adiponectin (reduced in obesity), and perilipin or resistin (antagonists of insulin action); hypothalamo-pituitary-pancreatico enterogastric peptides, such as ghrelin (a potent orexigenic, direct lipotrophic factor, and GH secretagogue), and orexin A (hypocretin-1) (respectively decreased and increased in obesity); glucose, hydrogen ions, branched-chain amino acids, and free fatty acid (FFA); and insulin, IGF-II, and IGF-I (free IGF-I concentrations are higher in puberty and obesity) (228, 230, 231, 232, 234, 235, 236, 237, 238). Intersecting hypothalamic circuits in turn govern appetite, physical activity, sleep, sympathoadrenal outflow, insulin sensitivity, resting energy expenditure, and reproduction (228). An implicit theoretical expectation is that incremental signal exchange must proceed bidirectionally, viz., somaCNS and CNSsoma. The bases for the developmentally defined orchestration of such reciprocity of signaling are only sparingly understood.
i. Critical body weight.
The somatometer conjecture arose from the prescient epidemiological insight that attainment of a so-called critical body mass accompanies the onset of puberty in mammals (239). An important contemporary distinction is that positive energy balance (rather than absolute weight) is necessary but not sufficient for normal onset of puberty (230). Current research further distinguishes whether somatic, metabolic, and/or neural signals: 1) trigger or delay the initiation of puberty; 2) impede ongoing sexual maturation; 3) impair established adult reproductive function; and/or 4) reflect multifactorial effects of type I diabetes mellitus, inanition, systemic inflammation, environmental foxius, sepsis, burns, trauma, surgery, or organ-system failure.
5. Leptin pathophysiology in the human.
Although facile unification of mechanisms is thwarted by species differences, one uniform inference in mammals is that sustained negativity of energy balance (wherein caloric utilization exceeds availability) inhibits GnRH release, suppresses LH secretion, reduces gonadal steroidogenesis, and impairs fertility (230, 231). The adipocyte-derived glycoprotein, leptin, is one recognized mediator of the effects of caloric extremes on reproduction in the rodent, and constitutes a necessary but not sufficient centrally mediated signal for timely pubertal onset and adult LH pulsatility in the human (238, 240).
In most but not all clinical studies, puberty is associated with higher mean serum leptin concentrations than those in childhood (241, 242, 243, 244). However, causal relevance is difficult to appraise, because no rigorous biochemical markers of the earliest stages of puberty are established to use for comparison of relative timing (1, 3, 160, 165). Moreover, single leptin determinations are confounded by: 1) a strongly positive correlation between leptin and sc adiposity, which also increases in early puberty (3, 245, 246); 2) higher leptin values in the female than male in utero, infancy, childhood, puberty, and adulthood (247, 248); 3) the capability of estradiol to stimulate and testosterone to suppress leptin production in young adults (245, 246); 4) short-term fluctuations in serum leptin concentrations; 5) reduced leptin values during fasting and after fat ingestion (249, 250, 251); and 6) collateral regulation by insulin, cortisol, and cytokines (238).
Inactivating mutations of leptin peptide or receptor in the mouse and human are associated with obesity, insulin resistance, hypertriglyceridemia, hypogonadotropic hypogonadism, low basal oxygen consumption, TSH deficiency, and relative hypothermia (238, 252). Occasional patients have exhibited a less striking phenotype, which receded in adulthood. In a clinical study of three genetically leptin-deficient individuals, administration of rh leptin for 1.5 yr reduced body mass index by about 50%, improved insulin action, decreased hyperlipidemia, and reversed hypogonadism (240). On the other hand, administration of rh leptin in the obese but otherwise normal human reduces weight only initially and modestly (253). Fasting for 3 d reduces pulsatile LH and TSH secretion in healthy young adults, which changes are reversed by leptin injections (250) (Fig. 11). Leptin administration also elevates basal energy expenditure and thyroid-hormone concentrations in subjects with sustained weight reduction (254). Accordingly, leptin availability appears to be necessary but not sufficient to maintain pubertal and young-adult GnRH-LH and TRH-TSH outflow, a nonobese phenotype, and normal insulin sensitivity.
|
a. Leptinergic impact on appetite and reproduction.
Immunoreactive leptin receptors and gene transcripts abound in cerebral cortex, hippocampus, thalamus, hypothalamus, gonadotropes, somatotropes, and corticotropes (238, 257, 258). Hypothalamic leptin receptors localize on neurons expressing AVP, oxytocin, CRH, NPY/agouti-related protein (AGRP), POMC, GHRH, TRH, neurotensin, dopamine, and MCH, albeit notably not GnRH. Inferentially, therefore, leptin enhances GnRH secretion indirectly. Supplementation with or transgenic knockin of leptin stimulates TRH, GnRH, POMC, MCH, and CART and represses SS, CRH, NPY, and AGRP gene expression in fasting normal or ob/ob (leptin mutation) mice (238, 258, 259) (Table 4). Outcomes encompass increased production of LH, gonadal steroids, TSH, and GH and decreased secretion of ACTH and glucocorticoids (238).
|
b. Leptinergic adaptations to caloric depletion.
Caloric restriction suppresses leptin synthesis not only in adipose tissue but also in arcuate nuclei (ArC), where the glycoprotein normally reinforces signals that stimulate satiety and repress those that promote appetite (238) (Tables 4 and 5). Thus, in the rodent, nutritional or genetic leptin depletion enhances food intake, depresses LH, GH, and TSH secretion, stimulates ACTH and glucocorticoid output, and enhances oxidation of adipocyte triglycerides (238, 252). Fasting concomitantly augments gastric secretion of ghrelin (20). Ghrelin stimulates GH secretion and potentiates certain metabolic effects of leptin withdrawal, viz. induction of hypothalamic genes encoding NPY, orexin, and AGRP (central appetitive signals) and suppression of SS (an inhibitor of TSH and GH secretion) (264, 265, 266, 267, 268, 269). The reciprocal effects of leptin depletion and ghrelin repletion during nutritional deficiency thus drive food intake and minimize substrate diversion to reproduction and anabolism. The adipostat concept provides a collective term for complex interactions among central and systemic peptides that jointly mediate metabolic adaptations to hypocaloric stress (Fig. 12).
|
|
-MSH (melanocyte-stimulating hormone) receptors (265, 266). NPY restrains or enhances GnRH and SS secretion depending upon receptor subtype expression, topography, developmental age, gender, and species, thus controlling both LH and GH secretion (270, 271, 272). For example, in the rat NPY-Y1 pathways mediate inhibition whereas NPY-Y2 and NPY-Y5 subserve stimulation of GnRH secretion (209, 210). The foregoing distinctions are specific, because NPY-Y1 and NPY-Y5 both transduce appetitive drive (272, 273). Receptor subtypes NPY-Y1/2 may subserve GH feedback-induced SS release, which suppresses GH secretion (274, 275). The existence of neuroanatomically localized and functionally distinct receptor subtype-specific NPYergic pathways extends the interpretation of earlier less specific NPY gene-silencing studies (276).
Leptin induces CART, a potent anorexigenic hypothalamic signal that also stimulates GnRH release (259, 277). CART colocalizes with MCH (satiety factor) and TRH (a thermogenic peptide) in the rat and human ArC, and it promotes satiety even in NPY-infused animals (278, 279). Thus, withdrawal of leptinergic drive to CART during fasting would further reduce GnRH outflow, stimulate food intake, and impair thermogenesis (Table 5 and Fig. 12). CART and GnRH gene transcripts increase in the adult male rodent, suggesting a possible role in pubertal LH secretion.
The POMC gene driven by CRH encodes potent opiatergic peptides, such as enkephalins, -MSH, and ß-endorphin, which diminish feeding behavior via multiple mechanisms (228, 266, 270) (Table 5). Leptin promotes satiety in part by stimulating anorexigenic -MSH, which acts centrally via melanocortin-receptor (MCR) types 3 and 4. Mutations and certain polymorphisms of MCRs are associated with obesity in the human and mouse.
Food restriction elicits CRH-POMC-glucocorticoid outflow (120, 230, 231), inhibits GnRH-LH-androgen production (280), and increases GH secretion in the human, monkey, sheep, and guinea pig (1). In contrast, caloric deprivation suppresses GH secretion via CRH- and leptin-modulated SSergic mechanisms in the rat (238, 281). The tripartite response of elevated glucocorticoids, decreased androgens, and reduced GH/IGF-I in the fasting rodent predicts marked catabolism (238, 282). Catabolism is averted in part in other mammals by increased GH secretion.
7. Metabolic substrates.
Energy depletion limits GnRH output in the human, monkey, sheep, hamster, mouse, and rat (228, 231, 238, 280, 283, 284). An array of mechanisms pertains. Nutrient restriction accentuates negative feedback by sex steroids in the rodent and ruminant (220, 230, 285). Glucoprivation reduces GnRH secretion via brainstem noradrenergic drive of opiatergic restraint in the rat and sheep and via paraventricular CRH/urocortin and/or AVP-mediated inhibition in the monkey (228, 280). Increased central CRH/POMC outflow rather than the moderate elevation of circulating glucocorticoids in fasting seems to suppress GnRH secretion (231). Augmented ghrelin output associated with caloric restriction may also decrease LH pulsatility via central mechanisms (286, 287).
Food ingestion or infusion of glucose, but not of insulin or IGF-I alone, rapidly reinstates pulsatile GnRH/LH release in the nutrient-depleted animal and human (230, 231, 284, 288). More sustained exposure to insulinomimetic peptides can promote proliferation of immortalized GnRH neurons and enhance GnRH-stimulated LH secretion by pituitary cells (92). In addition, IGF-I supplementation advances the time of first ovulation slightly in the immature female rat and monkey (89, 90). Conversely, transgenic muting of the CNS insulin receptor or insulin-receptor substrate (which mediates actions of both IGF-I and insulin) disrupts murine estrous cyclicity without affecting male reproductive function (81, 86). Thus, insulinomimetic peptides exert long-term trophic effects on GnRH neurons.
|
II. Ensemble Regulation of GH Secretion |
|---|
|
TopAbstractI. Regulation of the...II. Ensemble Regulation of...References |
|---|
). The result is a complex pattern of GH outflow, defined by pulsatile, nycthemeral (24-h rhythmic), and entropic (degree of orderliness) (Fig. 13A). Interactive control is endowed by mechanistic complementation, which is exemplified by: 1) unique individual properties of GHRH, SS, and GHRP and cognate receptors; 2) partially intersecting cellular signaling systems; 3) reciprocal connectivity among cognate neurons; 4) feedback by GH and IGF-I on peptidergic neurons and receptors; and 5) developmentally timed emergence and quenching of pathways (1, 289, 290, 291, 292, 293, 294, 295, 296, 297, 298, 299, 300, 301). The organized activation of GH/IGF-I production in puberty inferentially requires adaptive control of each of the foregoing relationships.
|
