| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Endocrine Section (A.G.), Department of Internal Medicine, University of Brescia, 25125 Brescia, Italy; and Division of Endocrinology (J.D.V.), Department of Internal Medicine and National Science Foundation Center in Biological Timing, University of Virginia, Charlottesville, Virginia 22908
 |
Abstract |
|---|
 Top Abstract I. IntroductionII. Contemporary Tools for...III. Neuropeptide Regulation of...IV. Other GH-Regulating...V. Neurotransmitter Regulation...VI. Role of Metabolic...VII. Other Hormonal Regulators...VIII. Regulation of the...IX. Exercise-Induced Modulation...X. SummaryReferences |
|---|
|
I. Introduction |
|---|
|
TopAbstractI. IntroductionII. Contemporary Tools for...III. Neuropeptide Regulation of...IV. Other GH-Regulating...V. Neurotransmitter Regulation...VI. Role of Metabolic...VII. Other Hormonal Regulators...VIII. Regulation of the...IX. Exercise-Induced Modulation...X. SummaryReferences |
|---|
). After the
isolation and characterization of GHRH and the cloning of its receptor,
as well as the more recent availability of molecular probes for
somatostatin and its recently cloned receptor family, more detailed
studies of the pathophysiological neuroregulation of GH secretion have
been performed. In this review, we will update earlier discussions in
the Journal (1, 2) by presenting recent developments in our
understanding of neuroregulatory mechanisms and their relevance to
clinical alterations in GH control. Primary diseases of the pituitary
gland or hypothalamus (e.g., tumors originating at these
loci) resulting in excessive or deficient GH secretion and peripheral
actions of GH or insulin-like growth factor-I (IGF-I) on target tissues
will not be reviewed here, but are discussed recently elsewhere (3, 4, 5, 6, 7, 8, 9).
Where possible, major interspecies differences in neuroregulation of
the GH axis (rat, sheep, and human) will be distinguished to limit any
confusion on this basis.
|
|
II. Contemporary Tools for Neuroendocrinological Investigation of the GH Axis |
|---|
|
TopAbstractI. IntroductionII. Contemporary Tools for...III. Neuropeptide Regulation of...IV. Other GH-Regulating...V. Neurotransmitter Regulation...VI. Role of Metabolic...VII. Other Hormonal Regulators...VIII. Regulation of the...IX. Exercise-Induced Modulation...X. SummaryReferences |
|---|
). Several
valuable molecular models, namely, genetically impaired GH secretion in
the little (lit) mouse, the dwarf (dw) rat, the fatty (fa) rat, and the
high-growth (hg) mouse, were recently reviewed in detail by Frohman
(10), to which the reader is referred. Conversely, the GHRH-transgenic
mouse exhibits excessive GH secretion, pituitary somatotrope
hyperplasia, and increased plasma IGF-I levels (11) that are
antagonized by potent GHRH-receptor inhibitors (12). In addition,
transgenic (rodent) models of suppressed secretion have been developed,
e.g., utilizing hypothalamically targeted GH or GHRH
transgenes (13, 14, 15). The experimental notion of targeting GH gene
expression selectively to the hypothalamus, and thereby producing
deficiency of endogenous GH, has disclosed alterations in
non-rapid-eye-movement sleep in the induced hyposomatotropic state
(16). Targeting of the GH gene in the mouse to the hypothalamus was
accomplished via use of either the tyrosine hydroxylase or GHRH
promoters to drive topographically localized GH excess (13, 14, 15).
This approach thus allows study of hypothalamic-regulatory peptide
responses to localized overexpression of a selected gene. In the case
of GH gene overexpression targeted to the hypothalamus, the
pathophysiological responses include increased hypothalamic
somatostatin and decreased GHRH accumulation, as predicted by other
experiments of GH autofeedback (17, 18, 19). Conversely, the genetic model
of the Ames dwarf mouse, which has a recessive defect with a
hypocellular anterior pituitary gland lacking somatotrophs,
lactotropes, and thyrotrophs (20), shows, as predicted from
physiological principles, increased expression of GHRH peptide and mRNA
in the hypothalamus (21). This follows from the premise that GH
autofeedback normally suppresses GHRH and stimulates somatostatin
expression (19). This thesis has recently been corroborated via
implantable GH-secreting (GC) cells in primary genetic vs.
transgenic GH-deficient rat models (22).
|
A recent novel gene knockout model, STAT5b gene disruption, which inactivates a specific signal-transduction pathway evokes loss of multiple sexually differentiated growth and cellular responses to GH, akin to those expected if the sexually dimorphic (male) pattern of pituitary GH secretion were abolished (32). This reflects the presumptive intracellular transcriptional factor-signaling role of STAT5b in mediating the cellular and nuclear (gene-transcriptional) actions of pulsatile (but not continuous) GH delivery. Thus, the STAT5b knockout mouse, albeit not measurably GH deficient, appears to be GH pulse resistant as reflected by dwarfism, low plasma IGF-I concentrations, and obesity. These features also are recognized in human GH receptor molecular defects causing tissue resistance to GH actions, e.g., Laron-type dwarfism (6, 33, 34, 35), and in the GH receptor knockout mouse (36)
Another transgenic strategy, transgenic expression of a GH receptor antagonist, has confirmed an important role of GH in antagonizing insulin action (37). Moreover, coexpression of the bovine GH gene and a human GH-receptor antagonist minigene in mice shows that the GH receptor antagonist will oppose the development of severe renal glomerulosclerosis and the increase in body growth otherwise driven by excess GH (38). In contrast, IGF-I overexpression augments body growth without inducing glomerulosclerosis, thus distinguishing certain tissue actions of GH and IGF-I. Indeed, another selective role of GH, but not IGF-I, is to increase motor neuron size in the lumbar spine, as inferred in transgenic mice overexpressing GH vs. IGF-I peptide (39).
In relation to pituitary developmental mechanisms, transgenic ablation
(via coupling a relevant glycoprotein 
-subunit cDNA to diptheria
toxin A chain) produces loss of gonadotrophs and thyrotrophs, as
distinct from somatotrope and corticotroph cell lineages (40). Other
recent experiments indicate the potential to "trap" developmental
progenitor cells by using transgenic overexpression of regulatory
regions responsive to cell-specific transcription factors,
e.g., the homeodomain protein, GHF-1. The latter is believed
to be required for the developmental generation of somatotropes and
lactotrophs, and hence GH and PRL gene expression (41). This technique
of so-called developmental entrapment can evaluate the roles of GH
progenitor cells by constitutive overexpression of GHF-1, and hence
"immortalization" of GH or PRL progenitor cells, which results in
dwarfism. A pituitary-targeted transgenic mouse model expressing the
leukemia inhibitory factor has suggested the possible role of
altered pituitary gene transcription and cell replication in the
pathogenesis of Rathke’s cysts (42).
Another informative molecular strategy is experimentally controlled transgene expression. For example, exogenously triggered activation and inactivation of the GH transgene in mice will reversibly reduce and induce obesity in this species (43). GH antisense RNA-transgene transfer in rats has achieved dose-dependent inhibition of GH gene expression, thus reducing GH secretion to varying degrees in heterozygous and homozygous transgenic animals (44). Conversely, expression vectors designed to transfer the GH gene as directed by tissue-specific promoters (e.g., muscle) provide an exciting prospect for targeted delivery of GH to, and for regulated expression of GH in, specific cell types, such as myoblasts in vitro (45) or myocytes in vivo (46). Indeed, in the latter, in vivo context or single injection of a myocyte-targeted GHRH-minigene expression vector can stimulate GH secretion and increase serum IGF-I levels by 3- to 4-fold or more for 2 weeks.
In brief, a rich and expanding repertoire of genetic, molecular, and transgenic and knockout models of receptors, regulatory peptides, IGF-I or GH itself, and/or signaling molecules has emerged and continues to unfold. This resource will make more specific and compelling studies possible of the molecular mechanisms of neuroendocrine regulation of the hypothalamo-somatotrophic axis in experimental animals. In some circumstances, a clinical counterpart is identifiable, in which a mutation of the corresponding gene is discernible in an individual or family. For example, IGF-I gene deletion, GH receptor, and GHRH receptor mutations with dwarfism all are recognized in the human (6, 28, 33, 34, 35, 47, 48, 49). However, GHRH receptor mutations causing isolated GH deficiency in the human are not common (50).
B. Human GH rhythms
1. Introduction. In conjunction with the development of
increasingly specific pharmacological agents to block or activate
individual regulatory receptors, new neuroendocrine tools have been
developed that allow improved quantitative measures of hormone release
over short intervals (ultradian rhythms or episodic peaks), as well as
over 24 h (circadian rhythms), and in relation to the patterned
orderliness of the release process (e.g., approximate
entropy estimates). These novel technologies often aid in dissecting
the neurohormonal mechanisms that underlie pathophysiological states or
molecular models (above). Biomathematical advances have also enhanced
clinical neuroendocrinological investigation, since in the human direct
portal catheterization techniques implemented in experimental animals
to monitor hypothalamo-pituitary secretion of GH, GHRH, and
somatostatin individually and concurrently, e.g., in the rat
and sheep (29, 51, 52, 53, 54, 55, 56), are not practicable ethically.
2. Background. Even relatively direct studies are not always
free of controversy. For example, a portal vein sampling study in the
anesthetized male rat indicated that both somatostatin withdrawal and
GHRH release tend to coincide with a GH pulse (56). Other repetitive
portal blood sampling investigations in the unrestrained and
unanesthetized (ovariectomized or intact male) sheep indicate that GHRH
increments typically precede GH pulses, but somatostatin decrements do
not necessarily do so (51, 57, 58, 59, 60) (Fig. 2). Moreover, in sheep, somatostatin
release can actually rise before a GH pulse or remain unchanged (see
Table 2). Consequently, GHRH-somatostatin
interrelationships in this species may be variable and quite complex.
For example, stress appears to stimulate both GHRH and somatostatin
release into portal blood, although the net result is increased GH
secretion (61). In addition, technical differences (e.g.,
anesthesia, sampling frequency, etc.) may confound a simple mechanistic
model originated in the adult male rat consisting of reciprocal
somatostatin withdrawal and GHRH release in initiating a GH pulse.
Indeed, taken as a whole, available studies clearly allow for other
input into GH secretory-burst timing, such as 1) extra- or
intrapituitary feedback by IGF-I (62, 63, 64); 2) intrapituitary paracrine
factors; 3) other hypothalamic coregulators [e.g.,
endogenous GH-releasing peptides (GHRPs), galanin, neuropeptide Y
(NPY), etc.]; for example, either NPY or galanin can colocalize with
GHRH in the hypothalamus (65, 66); and/or 4) other as yet unexplained
technical factors (e.g., hydraulic variations in blood flow
or sample collection; inconsistencies in GHRH, somatostatin, or GH
immunoassays; etc.).
|
|
Considerable clinical effort and numerous studies have appraised the
endocrine, metabolic, and pharmacological regulation of episodic GH
release (82, 83, 84, 85, 86, 87, 88, 89). For example, in the case of the GH axis, a burst of
GH release presumably reflects an appropriate stimulus such as
endogenous GHRH secretion (with or without other putative
secretagogues) in the context of responsive somatotropes
(i.e., at least not maximally inhibited by somatostatin),
and adequate releasable GH pools within the anterior pituitary gland.
Discrete pulse analysis provides information about the frequency and
amplitude of serum hormone concentration peaks (67, 68, 70) and can
disclose insights into neuroendocrine mechanisms that control this
ultradian pituitary release activity. For example, sampling blood every
30 sec during sleep in young men has revealed a strongly correlated
pattern of episodic GH release and the onset of slow-wave (stages III
and IV) sleep (90). Significantly, in some species such as the rat,
there is abundant evidence that the time profile of (pulsatile) GH
release strongly conditions the target tissue response anticipated,
e.g., induction of the LDL receptor and GH receptor (GHBP)
genes by a "feminized"/continuous GH release pattern, rather than
by a "masculinized"/pulsatile GH profile (91) (see Table 3). Hence, substantial physiological
effort has been applied to understand the neuroendocrine mechanisms
that supervise and dictate such pulsatile (GH) release patterns.
|
). This general class of analytical technology
thereby affords insights into the neuroendocrine control of the
underlying secretory events per se (104), which may be
regulated in principle by way of shape, duration, maximal amplitude,
and/or mass (106), with or without any attendant changes in hormone
half-life or distribution volume (73, 93, 94, 107). Indeed, altered GH
half-lives are likely only in subjects with severely impaired hepatic
or renal function [prolonged GH half-lives (108, 109, 110, 111)], and in
obesity [reduced GH life-life (112, 113, 114)]. Most recently, a
stochastic differential equation, random effects, maximum-likelihood
methodology based on feedback concepts within a neuroendocrine axis,
ultradian pulsatility coupled to circadian variations, and
biexponential kinetics has emerged to capture secretory dynamics more
fully (115, 116, 117).
|
0.020–0.05 µg/liter). This assay also revealed a pulse amplitude-
(and mass-) dependent mechanism for the gender distinction between GH
release in healthy middle-aged men and women (118). An even more
sensitive chemiluminescence-based GH assay (sensitivity of
0.002–0.005 µg/liter) confirmed small amounts of (<15%) basal
GH release in healthy men and women and illustrated that glucose
ingestion reduces serum GH concentrations typically to less than 0.7
µg/liter in women and to less than 0.07 µg/liter in men (119, 120, 121).
Thus, earlier "normal" glucose suppression of serum GH
concentrations assessed by RIA or IRMA has been grossly misestimated
and has also overlooked a strong gender difference. A high-sensitivity
enzyme-linked immunosorbent assay (ELISA) also corroborated very low
rates of basal GH release in both normal and hypopituitary adults
(122, 123, 124). How such basal secretion is generated or regulated is not
known (125). To our knowledge, detectable interpulse basal GH release
has not yet been measured in the male rat or other nonhuman species. Statistically correct treatment of assay data variability, especially at the low end, is also essential (126). Variability arises from uncertainties in the zero-dose tubes, among other replicates in the standard curve, due to the fit of the standard curve itself, and from the replicates of the unknown samples (127). Methods have been created recently to address these four joint/combined sources of within-assay experimental uncertainty (126, 128).
Clinical studies now indicate that visceral obesity, age, and other states of relative hyposomatotropism can be attributed mechanistically to decrease in GH-secretory burst mass, as estimated by the combined techniques of deconvolution analysis and ultrasensitive GH assays (121). Notably, as many as 97% of daytime samples for serum GH concentrations in older and/or obese subjects can be undetectable by conventional RIA or IRMA methods (125). Immunofluorometric, ELISA, and chemiluminescence GH measurement techniques, combined further in some cases with improved statistical assay analyses at the low end (126), have overcome some earlier limitations in studying dynamics of the human GH axis (125). For example, an ultrasensitive human GH assay has revealed that somatostatin infusion in young men suppresses both GH pulse amplitude and frequency (123). This is consistent with somatostatin’s putative inhibition not only of pituitary GH secretion but also of hypothalamic GHRH release (22, 129, 130). The reciprocal relationship between somatostatin and GHRH has also been inferred recently in in vitro studies in bovine hypothalamus (131).
Beyond and complementary to the pulsatile and 24-h (nyctohemeral) modes
of GH secretion, the subordinate (nonpulsatile) regularity or pattern
orderliness of hormone release over time can be quantified now by a
so-called approximate entropy statistic (132, 133). This is a
model-free and scale-invariant measure different from deterministic
chaos (nonlinear dynamical measures). An intuitive notion of the
entropy statistic is shown in Fig. 3B
. The orderliness of GH release
over time so quantified is markedly altered in certain conditions of
health or disease. For example, acromegalic patients secrete GH with a
high degree of quantifiable disorderliness, i.e., there is a
profound loss of or deterioration in regularity or reproducibility of
point-by-point subpatterns recurring across the 24-h GH profile,
whether defined by the approximate entropy statistic or a network-based
predictability measure (134, 135, 136, 137, 138). The distinctions between
GH-secretory patterns in active and remitted acromegaly as achieved via
the approximate entropy statistic are illustrated in Fig. 4.
|
). Unlike deterministic
chaos measures, which require 105 or more observations (and
theoretically noise-free data), approximative entropy values can be
calculated with good statistical replicability in hormone series
containing as few as 50–300 samples (118, 140, 141, 142). Such a small
sample size confers significant utility to this new measure.
|
In addition to the above available tools, other recent studies have begun to model the network or feedback-control linkages within the GH and other axes, to provide a more quantitative basis for articulating, testing, and revising specific neuroendocrine hypotheses (76, 154, 155). Considerable additional research will be required to refine the correct conceptual basis and dynamic features of the neuroendocrine component of the GH axis in appropriate biomathematical models.
|
III. Neuropeptide Regulation of the GH Axis: Somatostatin and GHRH |
|---|
|
TopAbstractI. IntroductionII. Contemporary Tools for...III. Neuropeptide Regulation of...IV. Other GH-Regulating...V. Neurotransmitter Regulation...VI. Role of Metabolic...VII. Other Hormonal Regulators...VIII. Regulation of the...IX. Exercise-Induced Modulation...X. SummaryReferences |
|---|
A. Mechanism of somatostatin actions and its receptors
Somatostatin binds to a family of specific receptors and inhibits
adenylyl cyclase via Gi, with additional actions to reduce
net calcium influx. Somatostatin inhibits GH release but not its
biosynthesis. This concept is important, since it may explicate
"rebound" GH secretion after somatostatin priming and withdrawal in
both the rat and human (157, 160). In addition, somatostatin may play
potentially (dual) inhibitory or stimulatory roles in controlling GH
secretion by acting on two distinct somatotrope cell populatins at
least in porcine pituitary (161).
Five somatostatin receptor subtypes have been cloned and characterized to date (162, 163, 164, 165). The somatostatin receptor is regulated in a subtype- and tissue-specific manner (166, 167). Furthermore, the cloning of somatostatin receptors and the development of somatostatin-receptor scintigraphy (168, 169, 170) hold significant clinical implications for identifying a variety of neuroendocrine and gastroenteropancreatic tumors. The clinical tool of somatostatin receptor scintigraphy has been reviewed in detail elsewhere (168).
Somatostatin peptide and somatostatin receptors are probably important in mediating many feedback and regulatory actions of neurohormones on the GH axis. For example, glucocorticoids likely influence both hypothalamic GHRH and somatostatin activity (171). In addition, GH autonegative feedback at the hypothalamic level involves both the GH receptor and hypothalamic somatostatin expression, inasmuch as treatment with antisense RNA to the GH receptor amplifies GH pulsatility (i.e., by abrogating GH receptor-mediated autonegative feedback) and decreases hypothalamic somatostatin gene expression in the rat (130). Other neuronal pathways may also participate in GH autonegative feedback, such as neuropeptide Y and galanin (172, 173). Indeed, the GH receptor also is expressed in NPY neurons in the arcuate nucleus (174). Moreover, the GH receptor gene in the hypothalamus is modulated by sex steroid and glucocorticoid hormones under several conditions (175).
Somatostatin receptor subtypes are expressed in individual rat pituitary cells, as defined by double-labeling in situ studies. For example, somatostatin subtypes II and V are present in all five major pituitary cell types in the rat (176), with the somatotropes expressing especially subtype V and to a lesser degree II, while thyrotrophs predominantly express subtype IV (177). In the rat hypothalamus, somatostatin receptor subtypes I and II may modulate GHRH and somatostatin release (178).
Somatostatin itself regulates somatostatin receptor subtype expression in some pituitary cell lines, e.g., GH-3 cells with differential up-regulation of subtype I vs. subtype II. In addition, food-deprived and diabetic rats show differential pituitary and hypothalamic somatostatin subtype gene expression/regulation (179). The molecular pharmacology of somatostatin receptors is reviewed further in Ref. 180 .
Somatostatin receptors also colocalize with GHRH-secreting neurons in the rat arcuate nucleus (181, 182). This important synaptic connectivity allows for intrahypothalamic neurohormonal interactions in GH regulation (58, 183), as reviewed further below (see GHRH). Somatostatin receptors in the locus ceruleus of male rats, seemingly paradoxically, stimulate GH secretion (184), thus illustrating the role of central nervous system (CNS) topography of somatostatin receptor distribution in mediating differential actions.
B. Human somatostatin receptor
Various somatostatin receptor subtypes are expressed within
pituitary adenomas, e.g., subtypes II and V in acromegalic
tumors that are responsive to somatostatin (185, 186, 187). This observation
has significant clinical implications, since available long-acting
somatostatin analogs (e.g., octreotide and lanreotide) are
known to interact especially with the subtype II somatostatin receptor
(187).
All five human somatostatin receptor subtypes are expressed in pituitary tumors as well as in normal (fetal) pituitary tissue (13, 180). In culture, the human pituitary gland predominantly expresses subtypes II and V whether in tumoral or normal pituitary cells (176, 188, 189). The human somatostatin receptor subtype genes are localized on different chromosomes, e.g., 14, 17, and 22, with simple tandem repeat DNA polymorphisms in subtypes I and II (190). All five cloned human somatostatin receptors subtypes are functionally coupled to adenylyl cyclase (191), and subtype I additionally stimulates inositol phosphate accumulation (192). The type V receptor also can mediate inhibition of GH secretion from acromegalic tumor cells in vitro (189). Consequently, the development of highly specific somatostatin receptor subtype agonists by structure/activity-based methods, such as a lanthionine octapeptide with high affinity for this receptor subtype, may have clinical application (193). Other studies of pituitary adenomas confirm expression of multiple subtypes, e.g., as assessed by PCR or other molecular methods (185, 194, 195).
Somatostatin itself is expressed in the brain and in the periphery in two principal forms; namely, somatostatin-14 and somatostatin-28 (196) (the latter preferentially binds to the subtype V receptor) (168). Somatostatin influences not only GH secretion but also that of numerous other hormones, as well as cognitive and behavioral processes, and impacts the gastrointestinal tract, the cardiovascular system, and tumor growth (156, 158). Thus, somatostatin and its receptor subtypes are widely distributed and regulated throughout the body, rather than solely in the hypothalamus and pituitary gland. Considerable complexity exists in the domain of somatostatin receptor subtype regulation in health and disease. This should represent a valuable purview for further incisive clinical and basic studies.
C. GHRH
1. Isolation, actions, and neuronal distribution. Human GHRH
was isolated originally from two pancreatic tumors in patients with
acromegaly from the United States and France as 44- and 40-amino acid
forms (197, 198). These bioactive peptides are derived from either of
two larger polypeptide precursors (pre-pro GHRH 107 and 108) (199, 200). The human GHRH gene resides on chromosome 20. The naturally
occurring variants of GHRH (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40) and the synthetic analog
1–29 are biologically equipotent on a molar basis in their capacity to
stimulate GH release in humans (201). The biological half-life of GHRH
1–44 is about 3–6 min. This is because GHRH is rapidly inactivated by
a plasma dipeptidylaminopeptidase, producing a more stable metabolite,
GHRH 3–44, which is about 1,000 times less potent than the parent
compound (202).
Intravenous administration of GHRH in humans evokes dose-related
release of GH from the pituitary gland. Approximately 1.0 µg/kg is a
maximally stimulating dose (203, 204, 205, 206, 207). GH secretion is detectable
within 5 min following GHRH injection, becomes maximal at 15–45 min,
and returns to baseline by 90–120 min (203). GH release induced by
intravenous GHRH injection in adults is highly variable
interindividually (208) and also in the same volunteer stimulated on
different occasions (207). This may be explained in part by the
diversity of hypothalamic somatostatin and/or cosecretagogue tone that
likely characterizes different individuals, and also by time-variable
somatostatin (or cosecretagogue) release and/or action in the same
subject (209). The relevance of variable hypothalamic somatostatin
secretion per se to the nonuniformity of the GH response to
single-bolus GHRH injection is suggested by experiments showing that a
functional somatostatin antagonist, i.e., the cholinergic
agent pyridostigmine, is able to cause a significant leftward
(increased sensitivity) and upward (increased efficacy) shift of the
GHRH dose-response curve in young men (206) (Fig. 6).
|
Age is a critical factor in influencing the GH response to GHRH in humans. Neonates, children, and young adults have similar GH responses to GHRH, whereas GH release in most studies is reduced in older individuals (212). Cells from the human fetal pituitary gland become responsive to GHRH in vitro before the appearance of either GHRH-containing neurons or the hypophysial portal vasculature (213). Human fetal pituitary responsiveness in vitro to GHRH matures further during gestation (214), which suggests that GH secretion from the human pituitary gland is not dependent on hypothalamic GHRH’s availability from weeks 7–24 of fetal life (213).
Sex steroids may also influence the GH-secretory response to GHRH in the human. However, not all workers have observed consistent differences between men and women (205), in young women after ovariectomy (215), during progestin-opposed estrogen replacement in the postmenopausal woman (216), or at different stages of the menstrual cycle (210). This is discussed further in Section VII.B.2.
Prolonged continuous infusion or repeated intravenous bolus administration of GHRH leads to a modest decrease in the subsequent GH responses to GHRH especially in the rat in vitro (217, 218) and to a lesser extent in the human in vivo (219, 220). Prolonged continuous GHRH infusion in normal humans causes acute maximal GH release; thereafter, pulsatile GH release continues but tends to decline (albeit remaining above basal values) by 5 h despite uninterrupted GHRH administration (220). Continued pulsatility of GH secretion occurs despite an unvarying GHRH stimulus, which suggests intermittent somatostatin action and withdrawal in normal humans or intermittent cosecretagogue release (221, 222, 223).
Treatment with GHRH in the rat leads to a fall in pituitary GH content. Concomitant administration of somatostatin with GHRH in the human or rat largely prevents this attenuation (human) or loss (rat) of somatotrope responsiveness, which occurs in vivo in normal human subjects (224) and in vitro in rat anterior pituitary cells (225). Thus, depletion of a GHRH-sensitive (and somatostatin-antagonized) releasable pool of GH is one factor contributing to the loss of GH responsiveness to GHRH in the rodent. Other biochemical mechanisms are also suggested, however, since prior treatment with GHRH in vitro in the presence or absence of somatostatin causes a decrease in the cAMP response to a subsequent GHRH stimulus as well as an increase in the ED50 for the stimulation of both cAMP accumulation and GH secretion by GHRH (226). Whereas GHRH pretreatment in vitro leads to a 48% decrease in the number of GHRH receptors on somatotropes, maximal GH responses to GHRH can be obtained by occupancy of only 10–20% of the total number of anterior pituitary GHRH receptors. Thus, the physiological relevance of in vitro receptor desensitization to GHRH observed in the rat is not easily translatable to in vivo down-regulation in the human.
In the human, continuous intravenous infusion of GHRH in normal subjects for 2 weeks evokes a marked increase in nocturnal GH pulses and plasma IGF-I levels (227). Moreover, 3 days of pulsatile GHRH administration (0.33 µg/kg every 90 min iv) to men of varying ages and body compositions augments and maintains pulsatile GH secretion and its nyctohemeral rhythm (151, 228). Lastly, GHRH-secreting tumors achieving sustained high serum GHRH concentrations elicit pulsatile GH release (229). These clinical data indicate the robustness of the human (in contradistinction to the rat) somatotrope to unvarying GHRH stimulation and also suggest the feasibility of treating selected patients with GHRH on a continuous basis, e.g., via a repository form of GHRH.
Immunoreactive GHRH exists in high concentrations in neurons of the
median eminence and arcuate nuclei of mammalian and nonmammalian
species. GHRH is also expressed in the anterior hypothalamic region as
well as the dorsomedial and ventromedial nuclei (230, 231). GHRH is
colocalized with other neuropeptides (e.g., galanin or NPY)
in the hypothalamus (232). Moreover, hypothalamic GHRH neurons receive
significant afferents from somatostatin neuronal nuclei (181, 233) that
are hypothesized to be important in the GHRH-somatostatin interplay
that presumptively directs ultradian GH pulsatility (Refs. 22, 129, 130, 131 and Fig. 7). GHRH has also been
identified both in the secretory granules and the nuclei of
somatotropes (234). Appreciable quantities of immunoreactive material
are detected in plasma (235), duodenum (236), and placenta (237). Thus,
circulating GHRH levels are not usually helpful to the endocrinologist
clinically, except in the rare condition of ectopic GHRH secretion with
attendant acromegaly (229).
|
The human GHRH receptor was cloned from an acromegalic pituitary cDNA library (240). The porcine and rat GHRH receptors were cloned independently (245, 246). The GHRH receptor is homologous with that for secretin and vasoactive intestinal polypeptide. Expression of the pituitary GHRH receptor is developmentally regulated in the rat (247). Moreover, albeit initially unexpected, both the GHRH gene and peptide are expressed in the gonad, e.g., rat ovary (248). There is tissue heterogeneity of GHRH receptor expression in the human, e.g., kidney compared with pituitary gland (249).
GHRH receptors are critical for GH release and growth, since point
mutations of the extracellular peptide-binding domain of the GHRH
receptor in the dwarf lit/lit mouse (50) disrupt body growth and
abrogate GHRH receptor function (250). GH deficiency also occurs in
corresponding mutations of the human GHRH-receptor gene (see Table 1).
Moreover, in the human and the rat, overnight GH secretion as well as
GH release stimulated via a number of secretagogues can be blocked by a
selective GHRH antagonist (251, 252), thus supporting a role for GHRH
in endogenous pulsatile GH secretion.
3. GHRH receptor regulation. The GHRH receptor, its signal transduction mechanisms, and mediation of GHRH-stimulated gene expression were reviewed recently (243, 244). In principle (although not yet described), activating mutations of this receptor might lead to constitutive cAMP overproduction, e.g., in patients with GH-secreting tumors and acromegaly. Alternatively spliced mRNA species encoding truncated nonfunctioning GHRH receptors can be identified in human pituitary adenomas (253). The receptor for GHRH is regulated physiologically by glucocorticoids and estrogen (254, 255). Steroid hormone regulation of GHRH receptor activity will thus require further study.
Lesioning of GHRH neurons or GHRH antagonist administration diminishes growth length and weight in the experimental animal, decreases GHRH receptor number, and reduces pituitary GH content (252). The receptor is clearly distinguishable from that of GHRP (below), somatostatin, or pituitary adenylate cyclase-activating peptide (PACAP), etc. GHRH receptor mRNA is expressed in the pituitary gland, as well as in the periventricular, arcuate, and ventromedial nuclei of the hypothalamus, thus suggesting that these areas are sites of intrahypothalamic GHRH action in the rat (256). Moreover, GH autofeedback likely involves, in part, suppression of hypothalamic GHRH expression, since mice expressing a human GH transgene targeted to the hypothalamus exhibit markedly reduced expression of GHRH peptide and mRNA in the hypothalamus, of GH in the anterior pituitary gland, and of IGF-I in blood (15).
D. Feedback regulation of GH secretion
Somatostatin, GHRH, GH itself, and its nearly ubiquitous tissue
mediator, IGF-I, are maintained homeostatically in hypothalamic loci,
the pituitary gland, and the circulation by a complex interplay of
feedback signals involving the 4 (poly-) peptides themselves (see Fig. 7). So-called GH autofeedback can be demonstrated in normal subjects,
since GH injection reduces the subsequent GH secretory response to a
GHRH stimulus (257, 258). This inhibitory effect is observed as early
as 3 h after intraperitoneal GH administration before any rise in
the plasma total IGF-I concentration occurs (259, 260). GH autofeedback
can be relieved by pyridostigmine pretreatment (260), which likely acts
in part to reduce hypothalamic somatostatin secretion and increase GHRH
release (261, 262). Thus, such clinical experiments support the view
that GH autonegative feedback stimulates somatostatin release from the
hypothalamus in vivo. In addition, a less important direct
pituitary inhibitory effect and/or concomitant GHRH withdrawal in
response to a GH stimulus cannot be excluded. In the rat, gender
differences in GH autonegative feedback exist; there is continuing
responsiveness to repeated GHRH stimuli (less sensitivity to GH’s
negative feedback) in the female rat, but not in the male animal (see
below).
After cloning of the GH receptor gene, studies indicate that GH feeds back to suppress the hypothalamic expression of the GH receptor itself (263). It is noteworthy that GH receptor antisense administration centrally [intracerebroventricularly (icv)] in the rat augments spontaneous GH pulsatility and diminishes hypothalamic somatostatin expression. This supports a primary role of the GH receptor and secondarily of somatostatin in mediating physiological GH autonegative feedback (130).
In addition to the network-like feedback actions of GH on hypothalamic
somatostatin, GHRH, and the GH receptor in the rat, greater complexity
arises since hypothalamic GHRH and somatostatin can each negatively
regulate its own secretion and reciprocally control secretion of its
counterpart (Figs. 1 and 7). In particular, GHRH inhibits its own
release but increases somatostatin release in vitro (264).
Conversely, somatostatin inhibits its own secretion and that of GHRH
in vitro (131, 265). Intrahypothalamic interactions are also
indicated by intracerebroventricular administration of somatostatin or
GHRH to rats, which elicits, respectively, an increase or a decrease in
blood GH concentrations (266). Moreover, hypothalamic GHRH and
somatostatin neuronal systems are anatomically coupled (183),
e.g., somatostatin neurons from the periventricular nuclei
synapse on GHRH neurons in the arcuate nucleus (267). Thus, GHRH and
somatostatin may play opposing roles in the control of GH secretion not
only on the pituitary gland but also at the hypothalamic level by
acting as neuromodulators.
Presumptively, functional reciprocal intrahypothalamic linkages between GHRH and somatostatin release also make in vivo studies with GHRH or somatostatin antagonists more difficult to interpret. For example, in the human, a predominant GHRH antagonist with some agonist properties can inhibit spontaneous pulsatile GH secretion (251, 268), as well as that stimulated by various secretagogues, e.g., GHRH, sleep, insulin, pyridostigmine, and L-arginine (269). Such inhibition might be explicable not only if endogenous GHRH mediates the actions of these secretagogue, but also if the GHRH antagonist serves to block GHRH-stimulated somatostatin release.
Feedback actions of IGF-I (and IGF-II) also are inferred at the hypothalamic-pituitary levels in the rat and human (270). Although somewhat controversial (271), a dominant feedback action of IGF-I administered peripherally in the human, sheep, or rat probably occurs directly on the pituitary gland (62, 63, 64), whereas rapid feedback by GH itself is primarily central (except perhaps in the sheep) (272, 273). Evidence also exists for hypothalamic actions of IGF-I (271, 274, 275), and hence this issue is not fully established. In earlier studies in the rat, partially purified preparations of IGF-Is administered centrally initially suggested major inhibitory effects of IGF-I and IGF-II each at hypothalamic loci (270, 276). However, more recent experiments using recombinant human IGF-I or IGF-II indicate that either IGF-I or IGF-II acting alone is not strongly inhibitory of the GH axis when administered centrally or in hypothalamic cultures in vitro (275, 277, 278). Coexposure to IGF-I and -II suppresses the GH axis. Hypothalamically targeted IGF-I gene and/or IGF-I receptor antisense mRNA expression would likely help clarify this important conceptual issue.
GH autofeedback in the rat may also decrease GHRP receptor expression in the arcuate and ventromedial hypothalamic nuclei (279). Whereas a putative GHRP-like endogenous ligand has not been identified, its demonstration would allow the conjecture that the (endogenous) GHRP-pathway participates in GH autonegative feedback.
|
IV. Other GH-Regulating Neuropeptides |
|---|
|
TopAbstractI. IntroductionII. Contemporary Tools for...III. Neuropeptide Regulation of...IV. Other GH-Regulating...V. Neurotransmitter Regulation...VI. Role of Metabolic...VII. Other Hormonal Regulators...VIII. Regulation of the...IX. Exercise-Induced Modulation...X. SummaryReferences |
|---|
A. GHRPs
1. Introduction. Potent GH-releasing synthetic oligopeptides
(so-called GHRPs) were developed mostly in the laboratory of Bowers
(280, 281, 282) via conformational energy calculations, solid-phase
synthesis, and screening for biological activity. These compounds were
first synthesized in the early 1980s as enkephalin derivatives and
modified subsequently to nonpeptidyl analogs (283, 284, 285, 286, 287). Although
unproven, GHRP oligopeptides likely mimic endogenous effector molecules
of currently unknown structure and identity, since they release GH via
specific receptors expressed in the hypothalamus and pituitary gland
(below) in a dose-related and specific manner both in vivo
and in vitro in several species (280, 288, 289).
Furthermore, chronic administration of GHRP to rats, dogs, or humans
can promote an increase in body weight, GH secretion, and plasma IGF-I
or BP-3 concentrations (290, 291, 292, 293, 294, 295, 296, 297). The mechanisms of GHRP actions are
at least partially independent of those of GHRH, since the combined
effects of GHRP and GHRH in vivo (but not usually in
vitro) on GH release are typically synergistic at maximal
concentrations (282, 298). GHRPs also release in vivo
hypothalamic GHRH into portal blood in the sheep (299, 300), and act
directly in vitro on functional GHRP receptors in human
(fetal) pituitary gland (214). In the rat pituitary in
vitro, GHRP but not GHRH receptor antagonists block GHRP(-2)
actions (301). Available data thus indicate (plausibly) multiple sites
of GHRP actions (Refs. 289, 302, 303, 304 and see below).
In one study in the rat, estrogen (or aromatizable androgen) augmented the maximal effect of GHRP in stimulating GH release (305), and in older women oral estradiol treatment amplified the steepness of the GH secretory response (sensitivity change) to increasing doses of GHRP-2 (306), suggesting sex steroid regulation of this pathway. The maximal GH-secretory response to GHRP (hexarelin) rises 3-fold in mid-late puberal children and falls in elderly individuals (307). In contrast, postpartum hyperprolactinemia and Cushing’s syndrome are associated with markedly attenuated responsiveness to GHRP (308).
In the human, an intact hypothalamo-pituitary unit is required for
maximal GHRP actions (309, 310). Propranolol and clonidine (see
Section V.B on catecholamines) do not greatly modify
the GH-secretory response driven by GHRP (311). Combined GHRP and GHRH
administration typically evokes marked GH release in the human (312, 313), and GHRP plus GHRH may be a nearly age-independent test of GH
reserve (314). Most recently, novel GHRP mimetics, such as the
intranasally or orally active nonpeptide (spiropiperidine) analog,
MK-0677 (283, 315, 316, 317), can stimulate and maintain pulsatile GH
release and increase plasma IGF-I concentrations during oral treatment
over hours to weeks in young, older, and obese men and women (290),
thereby suggesting possible GHRP mimetic use in GH adjuvant therapies
(see Fig. 8). Intranasal GHRP stimulates
linear growth in children (318). Indeed, both injectable and
noninjectable (intranasal or oral) forms of GHRP agonists and mimetics
hold promise for enhancing GH release, even in certain conditions of GH
axis pathophysiology, e.g., obesity (296), starvation (297),
critical illness (319, 320, 321), or aging (290, 295, 322). Repeated GHRP or
mimetic administration in the human is only sparingly (or, in some
cases, seemingly not at all) down-regulating (290, 293, 296, 297, 323, 324, 325, 326, 327, 328, 329), unlike a proclivity to down-regulation in the rat (330).
|
Further studies on the mechanisms of action of GHRP compounds (283, 287, 324), as well as eventual isolation of the putative endogenous GHRP-like ligands, will clearly be important.
2. GHRP receptor(s) and actions. GHRPs constitute a distinguishable family of synthetic oligopeptides (e.g., tri-, penta-, hexa-, and heptapeptides) that act presumptively at the hypothalamic as well as (probably to a lesser degree) pituitary levels to drive GH release (340, 341, 342). No natural ligand is yet known, but the receptor family has been cloned (340, 342, 343, 344, 345, 346, 347). The type 1a (but not 1b) GHRP receptor is biologically active (345). Stimulation by these novel peptides does not absolutely require endogenous GHRH activity, since they remain active after pituitary desensitization to GHRH (348) and can stimulate GH secretion by cells not expressing the GHRH receptor, such as GH1 cells (349). GHRPs act negligibly on GH secretion in mice with an 80% reduction in somatotrope cell number associated with a mutated GHRH receptor, e.g., lit/lit mouse (289). However, in this mutant animal, the arcuate nucleus c-fos genomic response to GHRP is preserved, indicating that brainstem neuronal activation by GHRP does not require GH or, for that matter, the GHRH receptor (350). GHRP, unlike GHRH or PACAP-38, does not activate the pituitary-specific transcription factor, GHF-1/Pit-1, in rat anterior pituitary cells in vitro (351), further distinguishing its biochemical actions from those of GHRH.
The cloned receptor families for GHRP and GHRH are distinct (352). Purification and later cloning of the GHRP receptor in pig and rat anterior pituitary membranes revealed a magnesium-dependent, GTP-binding, non-GHRH, non-somatostatin receptor that is G-protein coupled and apparently mediates the action of multiple GHRPs, e.g., GHRP-6, hexarelin, and the nonpeptidyl (spiropiperidine) GHRP analog, MK-0677 (340, 342, 343, 353, 354, 355, 356). GHRP receptors estimated by binding and functional reactivity have high affinity [e.g., dissociation constant (Kd) = 0.7 nM for MK-0677 (356)], and are expressed in the normal pituitary gland, pituitary tumors, the hippocampus, and ventromedial and arcuate nuclei in the hypothalamus (214, 340, 342, 344, 345, 345). GH autonegative feedback reduces (and GH deficiency increases) hypothalamic (ventromedial and arcuate nuclear) as well as hippocampal GHRP receptor expression, thus suggesting multisite feedback control mechanisms in regulating the putative endogenous GHRP receptor system (279).
GHRP’s intracellular signaling mechanisms in the rat include the phospholipase C-phosphoinositide pathway (357, 358, 359). GHRP also stimulates delayed calcium influx in rat and sheep somatotropes (360, 361), activates protein kinase C (357, 362), heightens GHRH signaling (the latter via cAMP) (302), and depolarizes somatotrope cells (343, 363). The mechanisms of GHRP action are distinctly nonopiate and non-GHRH receptor dependent (280, 298, 302, 352, 364, 365). However, effects of GHRP are partially antagonized by somatostatin in the rat and human (339, 366, 367, 368, 369), by centrally (icv) infused somatostatin in the rat (370), and (partially) by a GHRH receptor antagonist in the rat and human (371), but not by an opiate receptor antagonist in the human (372). GHRP peptides are active in vitro and in viv
, Acromegalic males;
, acromegalic females; 
, normal fed males; 
, normal fasted
males; •, normal fed females; 
, normal fasted females; 
,
acromegalic males in remission; 
, acromegalic females in remission.
[Redrawn with permission from M. L. Hartman et
al.: J Clin Invest 94:1277–1288, 1994
(
