Endocrine Reviews 18 (5): 621-645
Copyright © 1997 by The Endocrine Society
Peptidomimetic Regulation of Growth Hormone Secretion
Roy G. Smith,
Lex H. T. Van der Ploeg,
Andrew D. Howard,
Scott D. Feighner,
Kang Cheng,
Gerard J. Hickey,
Matthew J. Wyvratt, Jr.,
Mike H. Fisher,
Ravi P. Nargund and
Arthur A. Patchett
Merck Research Laboratories, Rahway, New Jersey 07065
 |
Abstract
|
|---|
- I. Introduction
- II. Identification of Peptidomimetic GH Secretagogues
- A. Mechanism of action of GHRH, GHRP-6, and somatostatin
- B. In vitro assays
- III. Molecular Design by Medicinal Chemistry
- A. Benzolactams and L-692,429
- B. Spiroindanes and MK-0677
- C. Isonipecotic acid peptidomimetics
- IV. Characterization of the MK-0677 Receptor
- A. Pituitary gland
- B. Hypothalamus
- V. Signal Transduction Pathway
- VI. Cloning the GH Secretagogue Receptor
- A. Chromosomal localization
- VII. Action of the Peptidomimetic GH Secretagogues in the Central
Nervous System
- VIII. Peptidomimetic GH Secretagogues in Vivo
- A. Animal models
- B. Clinical studies in humans
- IX. Regulation of Pulsatile GH Release
- A. GHRH and somatostatin
- B. The role of GHS-R
- X. Concluding Comments
|
I. Introduction
|
|---|
HORMONES are generally released episodically, yet to our
knowledge no attempt has been made to design drugs for treating hormone
deficiencies that imitate or amplify the endogenous oscillators
governing pulsatile hormone release. For example, although GH is
normally secreted in a pulsatile manner with peaks occurring
approximately every 3 h, GH deficiency is treated by injecting
recombinant GH either once daily or every 3 days. Clearly, this dosing
regimen does not result in a physiological GH profile and urges the
development of a more natural approach for the treatment of GH
deficiency. For convenience to the patient, a pathway was sought that
could be activated by a drug given orally once daily. By design,
conventional peptides were excluded, because the molecular template had
to be readily manipulable for optimization of oral absorption and
pharmacokinetics. A receptor that controlled GH release by interaction
with an endogenous small molecule was unknown, necessitating a
heterodox approach to drug discovery. This reverse pharmacology
involved: 1) establishing assays for the functional endpoint,
stimulation of episodic GH release; 2) understanding at the cellular
level how known peptides controlled GH release so that new specific
targets of intervention could be selected; 3) identifying selective GH
secretagogues of appropriate structure for optimization of
pharmacokinetic properties; 4) characterizing and cloning the receptor
for these ligands; and 5) identifying the receptor’s natural ligand.
This review provides a summary of points 1–4, focusing on the
discovery and properties of new GH secretagogues. These molecules were
subsequently shown to be peptide mimetics of the GH-releasing peptides
(GHRPs) first identified by the pioneering work of Bowers, Momany, and
colleagues (1, 2, 3); for a comprehensive review of GHRPs the reader is
referred to Refs. 4–9. The natural ligand for the receptor activated
by GHRPs and its peptidomimetics remains elusive, but now that the
receptor has been cloned and distributed, we anticipate its ligand will
soon be discovered.
|
II. Identification of Peptidomimetic GH Secretagogues
|
|---|
A. Mechanism of action of GHRH, GHRP-6, and somatostatin
The objective was to somehow amplify the GH-secretory pathway with
a small orally active molecule. This required a detailed knowledge of
the mechanisms governing GH release at the cellular biochemistry level
as well as understanding the interplay between the central nervous
system and the anterior pituitary gland. Two known hypothalamic
hormones, GHRH and somatostatin, are key regulators of GH release from
somatotrophs in the pituitary gland; GHRH stimulates GH secretion, and
somatostatin is inhibitory (10, 11, 12, 13). In addition to GHRH, a synthetic
hexapeptide,
His-D-Trp-Ala-Trp-D-Phe-Lys-NH2
(GHRP-6), first described by Bowers’ group, is also a potent GH
secretagogue (1, 2). The GHRPs were first described in 1981 (3), before
GHRH (14, 15), but until recently their receptor had not been defined
(16). The obvious approach was to identify nonpeptide mimetics of
either GHRH or GHRP-6; however, in contrast to antagonists, the
probability of finding nonpeptide agonists of peptides was considered
to be very low. The only precedent was morphine and its congeners,
which are nonpeptide mimetics of the opioid peptides. Therefore, from a
medicinal chemistry perspective, a more reasonable approach to amplify
GH release with a small molecule was to design a nonpeptide antagonist
of somatostatin. However, the potential for extrapituitary effects of a
nonselective somatostatin antagonist made somatostatin a less than
ideal target; moreover, before the cloning of somatostatin receptor
subtypes, screening for selectivity was not an option.
In spite of the historical difficulties of identifying peptidomimetics,
GHRH and GHRP-6 were excellent drug targets. A priori, a
peptidomimetic of GHRH seemed the most obvious choice because it had
been studied extensively. However, structure-activity relationships
indicated that the size of the molecule could not be reduced below 29
amino acids without a significant loss in activity (17). The peptide
GHRP-6 was of ideal size, but because its receptor had not been
identified, and cell lines responsive to GHRP-6 were unknown, high
volume screening for a peptidomimetic was impractical. Based on these
considerations, investigators modified the structure of GHRP-6 and
identified more potent peptides (4, 5, 6, 18). For example, activity was
enhanced by replacing D-Trp2 by D-2-(2-napthyl)alanine and
His by D-alanine to furnish GHRP-2
(D-Ala-D-2
Nal-Ala-Trp-D-Phe-Lys-NH2) (5). However, the
peptides still had low oral bioavailability.
The development of the GHRPs had primarily emerged from in
vivo studies and before attempting to identify peptidomimetics, it
was essential to understand the cellular mechanisms involved in the
action of GHRP-6. In 1989, the signal transduction pathway activated by
GHRP-6 was reported (19). It was established that GHRP-6 acted directly
on somatotrophs to cause GH release and to potentiate the effects of
GHRH. In contrast to GHRH, which increases cAMP in somatotrophs, GHRP-6
alone had no effect on intracellular cAMP, but when combined with GHRH,
the hexapeptide amplified the effects of GHRH on cAMP production (19).
GHRP-6 was subsequently shown to activate L-type Ca2+
channels, to depolarize the plasma membrane of somatotrophs by
inhibiting K+ channels, and behave as a functional
antagonist of somatostatin (8, 20, 21, 22). In contrast to GHRH, which
stimulates GH release through the kinase A pathway, GHRP-6 apparently
transduced its signal through protein kinase C (23). Phloretin, an
inhibitor of protein kinase C, inhibited GHRP-6-stimulated GH release
(23). Also, prolonged exposure of pituitary cells to phorbol esters
before GHRP-6 treatment markedly attenuated the action of GHRP-6
without affecting GH release induced by GHRH (23). Subsequent studies
showed that GHRP-6 stimulated IP3 turnover, activated
protein kinase C, and caused the release of intracellular stores of
Ca2+ (24, 25, 26). Collectively, these data provided evidence
that GHRP-6 acted through a receptor distinct from that of the GHRH
receptor and were consistent with the notion that the GHRP-6 receptor
was G-protein coupled.
To further evaluate differences between GHRH and GHRP-6 receptors, the
kinetics of desensitization and resensitization of pituitary cells
exposed to GHRP-6 and GHRH (27) were compared. These studies clearly
showed that pituitary cells were desensitized very rapidly by GHRP-6
compared with GHRH, and that complete resensitization required
interruption of exposure to GHRP-6 for at least 1 h (27). Having
established the kinetics of desensitization, tachyphylaxis was
deliberately induced to GHRH and GHRP-6 by prolonged continuous
perifusion of each secretagogue. The cells were then challenged with
GHRH and GHRP-6 to test for stimulation of GH secretion.
Cross-desensitization was not evident, confirming that discrete
receptors for GHRH and GHRP-6 were involved in the GH release pathways
(27). The rapid rate of tachyphylaxis observed with GHRP-6 explained
why some early studies failed to demonstrate an effect on GH release.
In contrast to studies with the longer acting GHRH, where increases in
GH in the medium of cultured pituitary cells are generally measured
over 2–3 h, GHRP-6 culture medium must be sampled within 10–15 min of
treatment for optimal results.
B. In vitro assays
Primary cultures of rat pituitary cells were used to screen for
small molecules that selectively caused GH release. Because it was
impractical to efficiently screen thousands of compounds in a primary
cell assay, the structural classes to be assayed were rationally
selected based on the key structural features of GHRP-6. However, to
maximize the probability of finding a small molecule GH secretagogue,
assay conditions were chosen that would identify compounds acting on
GHRP-6, GHRH, or ion channel pathways (28). Selectivity for activation
of somatotrophs, rather than mammotrophs or corticotrophs, was
determined by assaying the pituitary cell culture medium for PRL and
ACTH as well as GH.
The mechanism through which each active compound stimulated GH release
was ascertained. For example, each compound was evaluated to determine
whether it caused increases in cAMP or whether the protein kinase C
pathway was involved in signal transduction. The actives were also
assayed in the presence of specific GHRH and GHRP-6 antagonists to
determine whether they were GHRH or GHRP-6 mimetics (29). Their effects
on Ca2+, K+, and Na+ channels and
on membrane potential were also investigated. Finally, after
identification of somatotrophs with the reverse hemolytic plaque assay
(30), electrophysiology and Ca2+ imaging studies were
applied to more carefully determine effects on the target cell (29).
Having established the signal transduction pathway involved, pituitary
cells were treated with each active compound in combination with GHRH
or GHRP-6 to test for amplification of the respective pathways.
|
III. Molecular Design by Medicinal Chemistry
|
|---|
A. Benzolactams and L-692,429
Based on the structure of GHRP-6, and the knowledge that
benzodiazepine and related templates mimic small peptides,
approximately 120 compounds were selected for their ability to
stimulate GH release. A benzolactam 1 (Fig. 1
) with an EC50 of 3
µM for GH release in the rat pituitary cell assay was
identified (29, 31). Replacement of the carboxylic acid moiety with a
tetrazole and resolution of the enantiomers led to the discovery of
L-692,429 (Fig. 1, compound 2), a nonpeptide GH secretagogue
with an EC50 of 60 nM (29, 31, 32, 33). L-692,429
was the first example of a potent nonpeptide GH secretagogue (29).
L-692,428, the S enantiomer of L-692,429, was inactive,
demonstrating the stereoselectivity of the response (Fig. 2). Using fluorescent ratio imaging,
L-692,429 was shown to increase free intracellular Ca2+ in
somatotrophs (Fig. 3); L-692,428 was
ineffective (29). A structurally related nonpeptide antagonist of
L-692,429 was identified and used in conjunction with peptide
antagonists of GHRP-6 and GHRH to define the receptor target of
L-692,429 (29). L-692,429 synergized with GHRH but, in the presence of
a maximally effective concentration of GHRP-6, exhibited no additive
effect on stimulating GH release from rat pituitary cells. The activity
of L-692,429 was blocked by both peptide and nonpeptide antagonists of
GHRP-6, but not by an antagonist of GHRH (29). These results were
consistent with the notion that L-692,429 was a peptidomimetic of
GHRP-6.
View larger version (12K):
[in this window]
[in a new window]
|
Figure 2. Stereoselectivity of a nonpeptide GH secretagogue.
A comparison of the effects of L-692,429 (R-enantiomer), •, and
L-692,428 (S-enantiomer),  , on inducing GH release from cultured rat
pituitary cells (29). [Reprinted with permission from R. G. Smith
et al.: Science 260:1640–1643 (29). © 1993
American Association for the Advancement of Science.]
|
|
View larger version (70K):
[in this window]
[in a new window]
|
Figure 3. Fluorescent ratio imaging showing the effect of
L-692,429 on cytoplasmic free Ca2+ in a rat somatotroph.
Images of a somatotroph at 340 nm and 380 nm are shown as a function of
time after addition of L-692,429. The concentration of L-692,429
selected was 33-fold the EC50 for GH release, and the free
intracellular Ca2+ increased from approximately 100 to 780
nM (29). [Reprinted with permission from R. G. Smith
et al.: Science 260:1640–1643 (29). © 1993
American Association for the Advancement of Science.]
|
|
The specificity of L-692,429 was evaluated in rat pituitary cells to
determine whether L-692,429 stimulated corticotrophs and lactotrophs in
addition to somatotrophs. ACTH secretion was not affected; however,
small increases in PRL were observed in some instances (34). The
increase in PRL was more pronounced in pituitary cells derived from
lactating female rats, suggesting that L-692,429 might be acting on
somatomammotrophs (34, 35, 36). In contrast to the effects of L-692,429,
when GnRH, CRF, and TRH were used as controls, profound increases in
LH, ACTH, and PRL were measured in the pituitary cell culture medium;
however, no clear synergism between L-692,429 and GnRH, CRF, or TRH was
evident (K. Cheng and R. G. Smith, unpublished observations). To
further investigate whether the effects on PRL release might be
explained by activity on somatomammotrophs (35, 36), pituitary cells
were treated sequentially with L-692,429 and TRH and monitored for
changes in Ca2+ flux by fluorescent ratio imaging.
Approximately 7% of the cells responded to both L-692,429 and TRH,
indicating that somatomammotrophs as well as somatotrophs respond to
L-692,429 (S.-S. Pong and R. G. Smith, unpublished results).
Attempts were made to improve the efficacy of L-692,429. A comparison
of the biological activity of a series of six and eight-member lactam
analogs of L-692,429 showed that the seven-member ring was preferred
(31). Substitution with heterocyclic analogs of the benzolactam nucleus
resulted in diminished activity (37). Continued exploration of
structures related to L-692,429, focusing on refining
structure-activity relationships in the amino acid side chain, revealed
that the basic amine was an essential pharmacophore for GH-releasing
activity (38). A systematic investigation of this dimethyl-ß-alanine
side chain led to the identification of L-692,585, a
2-(R)-hydroxypropyl analog (Fig. 1
, compound 3),
which was 20-fold more potent than L-692,429 (38, 39). Comparison of
the binding data for L-692,585 [inhibition constant (Ki) =
0.8 nM] vs. L-692,429 (Ki = 63
nM) strongly suggested that the 2-hydroxypropyl moiety in
L-692,585 makes an additional binding interaction with the GH
secretagogue receptor. Although L-692,585 had much improved potency,
and subsequent studies in beagle dogs showed it had highly reproducible
oral activity, its oral bioavailability was unacceptably low (
4%)
for clinical development. Replacement of the central phenyl ring of the
biphenyl moiety in L-692,429 and L-692,585 was evaluated. A
cyclohexenyl analog of L-692,585 showed similar activity to its parent,
showing that the aromaticity of the central ring was not critical for
bioactivity and suggested that this ring may serve to orient the
benzolactam and phenyltetrazole pharmacophore (40). However, this
structural change provided no improvement in oral bioavailability.
To investigate replacements for the 2'-tetrazole moiety of L-692,429, a
variety of 2'-carboxamides and 2'-biphenyl analogs were evaluated. A
2'-carboxamide and N-2-hyroxypropyl tetrazole were found to
have similar potency to the acidic tetrazole; however,
N-alkyl tetrazoles, sulfonamides, and acyl sulfonamides were
generally less potent replacements (41, 42). The primary and secondary
carboxamides were potent GH secretagogues, and L-700,653 (Fig. 1, compound 4) had improved oral bioavailability in dogs and in
swine (42, 43). However, in spite of excellent potency, selectivity,
and tolerability in animals, the relatively low bioavailability
remained an issue with the benzolactam structural class; hence a
different structural lead was sought (44, 45, 46).
B. Spiroindanes and MK-0677
A new structural class of GH secretagogues was discovered by
screening compounds from a project to prepare derivatized privileged
structures for broad testing in receptor assays. The term "privileged
structures" refers to structural units that are found on a recurring
basis in receptor ligands. Their recognition and derivatization have
been proposed as a useful way to prepare receptor agonists and
antagonists (47). In the current instance, the successful strategy was
to derivatize a spiroindanylpiperidine with capped amino acids. This
piperidine derivative was considered a privileged structure since it
was present in 
- (48) and oxytocin receptor ligands (49) and, in
fact, was also present in a camphor sulfonamide lead 1
(L-368, 112, Fig. 4), which showed weak
GH secretagogue activity (50). Thus, it could be inferred that the
spiroindanylpiperidine would be an appropriate core from which to
elaborate ligands for the putative GH secretagogue receptor. Its
derivatization afforded L-252,564 (Fig. 4, compound 2) whose
activity as a secretagogue (EC50 = 50 nM) was
remarkable since it was an unseparated mixture of four
diastereoisomers. The other components of compound 2 are
tryptophan and a quinuclidinylurea. To account for the high activity of
this lead, it was noted that the quinuclidene group was also present in
an unpublished Merck GH secretagogue lead, and tryptophan is a key
amino acid in the GHRP-6 structure (50).
Although compound 2 was not orally active in dogs at 5
mg/kg, good oral bioavailability was achieved by replacing the
quinuclidinylurea with one of the amino acid side chains that had been
discovered earlier in the benzolactam program. The bioavailability of
this analog 3 (L-162,752) in dogs was more than 40% (51).
Also, its selectivity for GH release was exemplified by
IC50 values that were greater than 1 µM in
more than 24 in vitro assays (51). Specificity is sometimes
difficult to achieve in privileged structure derivatives, and this
property, coupled with excellent oral bioavailability, gave
considerable impetus to the lead’s development.
Potency enhancement was achieved by the introduction of a carbonyl or
hydroxyl substituent at the indane benzylic position. The oral activity
of these compounds was disappointing; however, it was restored by the
introduction of a methanesulfonamide group in this position. In
addition, replacement of D-Trp by
O-benzyl-D-serine further improved oral
bioavailability. The resultant compound 4 (L-163,191) was
active in the rat pituitary cell assay (EC50 = 1.3
nM) and was a specific GH secretagogue when counterscreened
in more than 50 in vitro assays (50). These included
oxytocin, enkephalin, cholinergic, adrenergic, serotonin, neurokinin,
and galanin receptors. Its pharmacokinetic properties in rats included
an oral bioavailability of between 6–22% (52), and in beagle dogs its
bioavailability was greater than 60% with a terminal half-life of
between 5–6 h (50, 52). In beagles, after oral administration of 1
mg/kg, GH was elevated for more than 6 h (50, 53). Insulin-like
growth factor I (IGF-I) levels were also increased significantly. Most
importantly, L-163,191 was the first GH secretagogue in its class
demonstrated to provide a sustained increase in IGF-I levels for up to
24 h after a single oral dose (53, 54). Based on these properties,
L-163,191 entered safety studies and then clinical studies as MK-0677.
A series of analogs was also prepared (55, 56) to evaluate more
extensively interactions with the MK-0677 receptor using site-directed
mutagenesis studies.
C. Isonipecotic acid peptidomimetics
The Genentech group used a multidisciplinary approach in their
discovery of a new series of small molecule GH secretagogues (Fig. 5). Toward the goal of determining the
topographical requirements for the GH-releasing activity of the GHRPs,
G-7203 (EC50 = 0.43 ± 0.11 nM), a cyclic
analog of the linear hexapeptide GHRP-2, was developed (57, 58).
Nuclear magnetic resonance studies showed that G-7203 was structured in
water. Furthermore, the D-2-Nal-Ala-Trp-D-Phe
fragment adopts a compact conformation with nested hairpin turns
initiated at D-Lys1 and Ala3. Other less active cyclic
GHRP-2 analogs did not readily adopt this conformation, suggesting that
a precise arrangement of the three aromatic side chains was crucial for
GH-releasing activity.
Peptidomimetic GH secretagogue G-7502 (EC50 = 10.6 ±
6.2 nM) was developed from extensive medicinal chemistry
studies on GHRP-6 (57, 58). An interesting finding from their studies
leading to the discovery of G-7039 (EC50 = 0.18 ±
0.04 nM) was that the
D-Nal-Ala-Trp-D-Phe triaromatic core of GHRP-6
could be presented as D-2-Nal-D-2-Nal-Phe.
Furthermore, an isonipecotic acid amino side chain was identified as a
satisfactory replacement for the N-terminal histidine of GHRP-6.
Further optimization of G-7039 by removal of the Phe residue led to a
series of small molecule GH secretagogues exemplified by G-7502. These
studies lend further support to proposals made by the Merck group that
the essential GH secretagogue pharmacophore is comprised of only a
diaromatic core and a basic amine. All of the Genentech GH
secretagogues, including G-7039 and G-7502, released GH from rat
primary pituitary cells in a dose-dependent manner,
synergized with GHRH but not GHRP-6, and demonstrated homologous
desensitization after prolonged exposure. Furthermore, coincubation
with somatostatin completely suppressed the GH-releasing activity of
these compounds. At 100 nM concentration, G-7030 and G-7502
were relatively specific for releasing GH except for small increases in
PRL. These results suggest that these new peptidomimetics exert their
action through the GHRP pathway (58).
|
IV. Characterization of the MK-0677 Receptor
|
|---|
A. Pituitary gland
Based on binding studies using [3H]naloxone as a
ligand, it was clear that although GHRP-6 and its analogs were derived
from met-enkephalin, neither GHRP-6 nor the peptidomimetics bind to
opiate receptors. Moreover, their GH-releasing activity was not blocked
by coadministration of naloxone (34). Extensive evaluation of L-692,429
and MK-0677 showed that they lacked activity on different opiate
receptor subtypes and in more than 50 different receptor assays (29, 50, 59). Attempts to define a receptor in the pituitary gland using
radiolabeled GHRP-6 failed to show a correlation between high-affinity,
low-capacity binding and biological activity of both peptide and
peptidomimetic GH secretagogues (60).
To characterize the MK-0677 receptor, high-specific activity
radiolabeled MK-0677 (800–1100 Ci/mmol) was synthesized by
substituting 35S for 32S in the molecule (60, 61). [35S]MK-0677 bound with high affinity [dissociation
constant (Kd) = 140 pM] and limited capacity
(Bmax = 6.4 fmol/mg protein) to porcine pituitary membranes
(Fig. 6). This concentration of binding
sites in pig pituitary is remarkably low, but in rat pituitary
membranes the concentration is even lower (2 fmol/mg protein).
[35S]MK-0677 binding was displaced by L-692,429,
L-692,585, and by the peptide GH secretagogue GHRP-6, but not by GHRH
or somatostatin (60). The Ki values in the binding assay
correlated with the EC50 values for stimulating GH release
in the rat pituitary cell assay (Table 1). Consistent with
[35S]MK-0677 binding to a G protein-coupled receptor,
binding was dependent upon Mg2+ (5 mM) and
displaced by GTP-
-S (10 nM) but not by ATP--S (59, 60). Remarkably, competition binding studies with L-692,429 and GHRP-6
showed that these two structurally distinct molecules were competitive
inhibitors of MK-0677 binding (59). By contrast, as anticipated for a G
protein-coupled receptor, GTP--S was an allosteric inhibitor (59, 60). To further explore the specificity of [35S]MK-0677
binding, met-enkephalin, GnRH, TRH, galanin, gastrin releasing peptide,
substance P, MSH, isoproterenol, dopamine, bromocriptine, propanolol,
and clonidene were tested in the binding assay at a concentration of 1
µM; none competed for [35S]MK-0677 binding
(60).
View larger version (22K):
[in this window]
[in a new window]
|
Figure 6. Scatchard analysis of [35S]MK-0677
binding to pig pituitary membrane showing a single class of
high-affinity binding sites (60). [Reproduced with permission from
S.-S. Pong et al.: Mol Endocrinol 10:57–61, 1996
(60). © The Endocrine Society.]
|
|
View this table:
[in this window]
[in a new window]
|
Table 1. Specificity of [35S]MK-0677 binding
to rat pituitary membranes and correlation with GH-releasing activity
on rat pituitary cells (60)
|
|
Digitonin was used to solubilize the MK-0677 receptor from porcine
anterior pituitary membranes. The receptor was recovered as a
receptor-[35S]MK-0677-G protein complex. The apparent
molecular mass of the complex determined by gel filtration under native
conditions was 255 kDa (62). Attempts to isolate the receptor before
labeling with MK-0677 were unsuccessful. Treatment of the solubilized
receptor in the digitonin micelle with GTP--S caused dissociation of
MK-0677 from the receptor with an EC50 of 5 nM;
ATP--S was ineffective even at 10 µM, consistent with
the MK-0677 receptor being coupled to a G protein in the soluble
complex (62).
To determine whether the binding to pituitary membranes was localized
to somatotrophs, a biotinylated analog of MK-0677, L-164,683, was
prepared as a suitable ligand for immunofluorescence studies (63).
L-164,683 was an excellent competitor for [35S]MK-0677
binding (IC50 = 0.2 nM) and stimulated GH
release with an EC50 of 2.5 nM. Primary
cultures of rat pituitary cells were treated with L-164,683 for 3 min
at 37 C and treated with avidin-Texas Red. GH-containing cells were
labeled with fluorescein-conjugated goat anti-rabbit IgG. Dual
fluorophor labeling for GH and the MK-0677 receptor by confocal
microscopy showed that about half of the GH-containing cells also
expressed the MK-0677 receptor (Fig. 7).
L-164,683 binding was confined to cells that contained GH (63). While
only about half of the GH-containing cells showed binding of the
MK-0677 analog, based on limits of sensitivity of the assay, this
proportion should be considered a low estimate. More quantitative
estimates have been attempted by localization using MK-0677 receptor
antibodies, and preliminary results suggest that all GH-containing
cells in the pituitary gland express the MK-0677 receptor (R. G.
Smith, A. D. Howard, S. D. Feighner, and J. W. Woods,
unpublished results).
View larger version (34K):
[in this window]
[in a new window]
|
Figure 7. Dual fluorophore localization of MK-0677-binding
sites and GH in rat pituitary primary cultures (63). Left
panel, Identification of rat GH with a rabbit anti-rat GH
antibody and fluorescein-conjugated goat anti-rabbit antibodies.
Right panel, Identification of GH secretagogue-binding
sites on the same cells with biotinylated MK-0677 (L-164,683; 1
µM; 3 min 37 C) and avidin Texas red. Cells were first
reacted with L-164,683 (15 min 4 C) and avidin, after which the cells
were fixed and permeabilized and treated for GH detection. Note that
approximately 50% somatotrophs are labeled for MK-0677-binding sites,
and all cells labeled with the MK-0677 derivative contain GH.
|
|
B. Hypothalamus
Hypothalamic membranes were isolated from rats to test for the
presence of the receptor in the hypothalamus.
[35S]MK-0677 bound with high affinity (Kd =
170 pM), and the concentration of binding sites
(Bmax = 8 fmol/mg protein) was higher than measured in rat
anterior pituitary membranes (64). Binding to rat hypothalamic
membranes was Mg2+ dependent and inhibited by
nonhydrolyzable analogs of GTP such as GTP--S and
guanyl-imidodiphosphate (64). These properties are consistent with
MK-0677 binding to a G protein-coupled receptor in the hypothalamus.
Membranes isolated from liver, thalamus, cerebral cortex, medulla,
pons, and posterior pituitary membranes were assayed, but
high-affinity [35S]MK-0677 binding was not detected,
demonstrating tissue specificity of MK-0677 binding (64). In common
with binding to rat anterior pituitary membranes, binding to
hypothalamic membranes was highly selective for GHRP-6, GHRP-2,
MK-0677, L-692,429, and L-692,585. Moreover, the relative
IC50 values for displacement of [35S]MK-0677
binding was highly correlated with activity in stimulating GH release
from cultured rat pituitary cells (64). Thus the hypothalamic receptor
has identical characteristics to the MK-0677 receptor identified in the
anterior pituitary gland.
|
V. Signal Transduction Pathway
|
|---|
MK-0677, in common with L-692,429 but in contrast to GHRH, does
not increase intracellular cAMP levels in pituitary cells (50, 59).
However, in combination with GHRH, MK-0677 amplifies the GHRH-induced
increase in cAMP and potentiates GH release (59). At doses of MK-0677
that maximally stimulated GH release in vitro, additional
stimulation is not observed in the presence of maximally stimulating
concentrations of L-692,429. Similar observations were made when GHRP-6
was substituted for L-692,429 indicating, in agreement with binding
studies, that GHRP-6 and these peptidomimetics all mediate their
effects through the same receptor. The synergistic effects of the GH
secretagogues on the GHRH pathway were mimicked by phorbol myristic
acetate. Preincubation of rat pituitary cells with phorbol myristic
acetate for 24 h, before treatment with MK-0677, profoundly
attenuated the stimulation of GH release by MK-0677 (50). These
results, and the cation and nucleotide dependence of MK-0677 binding,
suggested that MK-0677 interacts with a G-protein coupled receptor that
activates phospholipase C (59, 60). Consistent with activation of this
pathway were the earlier observations that other ligands for this
receptor, such as L-692,429 and GHRP-6, increase IP3
turnover (65), increase free intracellular Ca2+ through an
IP3 pathway, and cause translocation of protein kinase C
(24, 25, 26, 65).
The role of L-type Ca2+ channels in the transduction
pathway was confirmed by fluorescence ratio imaging in somatotrophs
after treatment with either L-692,429 or MK-0677 (29, 59). Nifedipine
and 
-agatoxin IIIA, but not conotoxin, were shown to block increases
in intracellular Ca2+, consistent with activation of L-type
Ca2+ channels (66, 67, 68). Electrophysiology studies showed
that the GHRP-6 peptidomimetics blocked K+ currents in
somatotrophs, resulting in depolarization and electrical spiking to
enhance Ca2+ entry through voltage gate channels (21, 69, 70). Modulation of these channels was evident when perforated patch
clamp or on-cell single-channel recording techniques were used, but not
when the cells were dialyzed using the whole cell voltage clamp
configuration, suggesting that a soluble second messenger was involved.
The depolarizing effects were confirmed using the membrane-sensitive
dye bisoxanol (29). Depolarizing agents such as the potassium channel
blockers tetraethylammonium and 4-amino pyridine and the sodium channel
agonist veratridine had no additive effects on GH secretion induced by
the peptidomimetics; rather, they mimicked the peptidomimetics by
amplifying the effects of GHRH (8). The fact that the peptidomimetics
cause depolarization explains their functional antagonism of
somatostatin, since somatostatin hyperpolarizes somatotrophs by
increasing potassium conductance (71). The magnitude of GH release
caused by depolarizing agents alone is less than that induced by the
peptidomimetics, emphasizing the importance of other aspects of the
signal transduction pathway. The MK-0677 receptor via phospholipase C
activates Ca2+ and inhibits K+ channels to
facilitate GH secretion (Fig. 8A). The
IP3-mediated redistribution of intracellular
Ca2+ alone does not explain all the effects of these
secretagogues because somatostatin does not prevent the
Ca2+ redistribution but does inhibit GH release (30);
rather, it is speculated that IP3 facilitates GH release by
synchronizing docking of GH-secretory granules to the plasma membrane
(72, 73).
View larger version (70K):
[in this window]
[in a new window]
|
Figure 8. A, Signal transduction pathway activated by
ligands that interact with the MK-0677 receptor. B, GHRH and MK-0677
act on discrete receptors and transduce their signal through different
pathways.
|
|
The GHRH receptor is also G protein coupled (74) but activates a
different signal transduction pathway to that of the MK-0677 receptor.
These differences are illustrated in Fig. 8B. GHRH stimulates adenylate
cyclase, resulting in an increase in cAMP (75), whereas GHRP-6,
L-692,429, and MK-0677 activate phospholipase C (23, 25, 26, 59, 65).
Apparently, through cross-talk in the signal transduction pathways,
these ligands synergize with GHRH to further increase intracellular
cAMP (8, 59). This augmentation might be mediated by interactions
between the Gß subunits associated with the MK-0677
receptor and G
s of the GHRH receptor complex (76). A
potential bonus of the synergy of GHRH and MK-0677 is that increases in
cAMP have been associated with increased GH synthesis (77). Based on a
knowledge of the signal transduction pathway that results in amplifying
the activity of GHRH and functionally antagonizing somatostatin, it
becomes clear why GHRP-6 and the peptidomimetics are so effective in
inducing GH release in vivo. The intriguing properties of
these synthetic GH secretagogues force us to speculate that such ideal
characteristics are shared by an undiscovered natural hormone that
plays a key role in the physiological regulation of pulsatile GH
release.
|
VI. Cloning the GH Secretagogue Receptor
|
|---|
Having established from binding data and biochemical studies that
MK-0677 probably binds to a G protein-coupled receptor that signals
through phospholipase C, an expression-cloning strategy was developed
for the MK-0677 receptor by microinjecting size-fractionated poly
A+ RNA from pig pituitaries into Xenopus oocytes
(16). Expression of the RNA from one fraction resulted in an
MK-0677-induced calcium activated chloride current, but the assay
reproducibility was low. A more robust cloning strategy was developed,
which coupled receptor expression in Xenopus oocytes to a
functional endpoint that measured MK-0677-induced Ca2+
release (16). The Ca2+-sensitive bioluminescent reporter
protein aequorin was used successfully as a signal for positive
coupling, but responses to MK-0677 were still weak and variable. It was
speculated that expression of a specific G protein essential for
receptor coupling might be limiting; therefore, cRNAs encoding a series
of G proteins (G11, Gq,
G16, G13, Gi1,
Gi3, Go) were individually coinjected
into Xenopus oocytes together with pituitary gland poly
A+ RNA and aequorin cRNA. Only G11 provided
a highly reproducible Ca2+-mediated luminescence signal in
response to MK-0677 (16). This robust expression system was used to
screen pools of cRNA from a pig pituitary gland cDNA library for an
MK-0677-inducible signal. Approximately 2 x 106
individual cDNAs from a pig pituitary library were screened in pools of
10,000. Stepwise fractionation of a single positive pool resulted in
the isolation of a single cDNA clone that conferred both
MK-0677-activated aequorin luminescence and an inward chloride current
in Xenopus oocytes. Interestingly, supplementation with
G11 was unnecessary when cRNA pool complexity dropped
below 50 clones.
The nucleotide sequence of the full-length swine MK-0677 receptor
cDNA-1a predicted a protein of 366 amino acids with seven transmembrane
(7-TM)-spanning domains, three intra- and extracellular loops, and a G
protein-coupled receptor triplet signature sequence (16). The human
receptor was subsequently cloned, and its predicted topology is
represented in Fig. 9. Genomic analysis
by Southern blotting was consistent with a single highly conserved gene
in human, chimpanzee, bovine, rat, and mouse (16). Sequence alignments
showed that the swine MK-0677 receptor was 93% identical and 98%
similar at the amino acid level to the human receptor (Fig. 10). Additional cDNAs clones were
obtained from pig and human libraries that encode a shorter form of the
MK-0677 receptor (16). Receptor 1b cDNA encodes a polypeptide of 289
amino acids that lacks transmembrane domains 6 and 7 of the 1a receptor
(16). This truncated receptor is identical to the 1a receptor from the
translation initiation codon to Leu-265 beyond which the cDNA is fused
to a short contiguous reading frame of 24 amino acids followed by a
translation stop codon. This 24-amino acid sequence is highly conserved
in both the pig and human MK-0677 receptor genes (16). A similarly
truncated mRNA has been reported for the neuropeptide Y1 (NPY1)
receptor (78). Inspection of the amino acid sequence of the human 1a
receptor revealed a G protein-coupled receptor signature [ERY142], a
series of potential N-glycosylation sites, protein kinase C, casein
kinase II phosphorylation sites, a cAMP/cGMP-dependent phosphorylation
site [346–349 RKLS], myristoylation sites, and an amidation site
(235–238 IGRK).
View larger version (45K):
[in this window]
[in a new window]
|
Figure 9. Predicted membrane topology and amino acid
sequence of the human type 1a GHS-R. TM denotes transmembrane domains
1–7.
|
|
View larger version (73K):
[in this window]
[in a new window]
|
Figure 10. Predicted amino acid sequences for the
human and swine type 1a and 1b GHS-R. Identities are highlighted in
the boxed amino acid sequences (16). Conserved cysteine
residues and the GPC-R signature sequence (ERY142) are shaded
gray. N-Linked glycosylation sites and potential
phosphorylation sites are highlighted with arrowheads
and asterisks, respectively. The transmembrane (TM)
domains are overlined and numbered sequentially. The
numbers on the left and right refer to
the sequentially numbered amino acid residues (1–366 for the type 1a
and 1–289 for the type 1b receptors). [Reprinted with permission from
A. D. Howard et al.: Science 273:974–977 (16).
© 1996 American Association for the Advancement of Science.]
|
|
The pharmacological properties of MK-0677 receptors 1a and 1b were
investigated with functional assays using aequorin bioluminescence and
electrophysiology in Xenopus oocytes (16). In addition,
after transient transfection of HEK293 or COS cells with 1a and 1b
cDNAs, aequorin bioluminescent assays (in HEK293 cells) and
[35S]MK-0677 competition binding assays were performed.
Induction of a Ca2+-activated Cl- current in
response to MK-0677 was observed in oocytes injected with 1a cRNA but
not with 1b cRNA. Similarly, in COS-7 cells transiently expressing 1a
or 1b, aequorin bioluminescence was only induced by MK-0677 with the 1a
clone. Transfection of clone 1a into COS-7 cells resulted in
high-affinity binding of [35S]MK-0677 to plasma
membranes. The binding was inhibited by GHRP-6 and GHRP-2 but not by
GHRH, somatostatin, TRH, or galanin (Fig. 11). Because this receptor binds GHRPs
in addition to MK-0677, it was designated the "GH secretagogue
receptor" (GHS-R). In contrast to GHS-R1a, high-affinity saturable
binding was not evident in membranes from cells expressing GHS-R1b
cDNA. Although no direct function has yet been assigned to receptor 1b,
it may play a regulatory role in the context of modifying the function
of a related G protein-coupled receptor. For example, it has been shown
that inactive truncated forms of related G protein-coupled receptors
can be coexpressed to rescue their function (79).
View larger version (20K):
[in this window]
[in a new window]
|
Figure 11. Competition binding assay showing displacement of
[35S]MK-0677 (.24 nM) from COS cell membranes
isolated from cells transiently expressing GHS-R1a. [Reprinted with
permission from A. D. Howard et al.: Science273:974–977 (16). © American Association for the Advancement of
Science.]
|
|
The closest identities of the GHS-R with other related G
protein-coupled receptors were to the neurotensin receptor (NT-R) and
thyroid releasing hormone receptor (TRH-R) with 35% and 29%
identity, and 59% and 56% similarity, respectively (values are for
the human GHS-R open reading frame compared with human NT-R and TRH-R).
A dendrogram of related G protein-coupled receptors suggest that the
GHS-R presents a new family of the NT-R, TRH-R branch of the
phylogenetic tree (Fig. 12).
View larger version (38K):
[in this window]
[in a new window]
|
Figure 12. Dendrogram of GHS-R and other G protein-coupled
receptors. Database searches (Genbank 92, EMBL 43, Swiss-Prot 31, PIR
45, dEST (Gbest 92), Prosite 12), sequence alignments, and analysis of
the GHS-R nucleotide and protein sequences were carried out using the
GCG Sequence Analysis Software Package (Madison, WI; pileup, peptide
structure and motif programs), FASTA and BLAST search programs, the
PC/Gene software suite from Intelligenetics (San Francisco, CA; protein
analysis programs) and Lasergene software (DNA Star, Madison, WI). The
amino acid sequence of representative members (51 sequences) for all
known classes (Families I-IV and pheromone) of G protein-coupled
receptors were used to construct the dendrogram using the clustal
method (PAM-250; gap and length penalty = 10). The length of each
pair of branches represents the distance between sequence pairs. The
scale below the tree measures the distance between
sequences. Units indicate the percent of substitution events. All
receptor sequences aligned to human GHS-R type Ia and type 1b cDNAs
were human unless otherwise noted, and accession numbers are for the
SwissProt database, GenBank database (designated with "G"), or PIR
database ("P"): 5HT-2A, serotonin, P28223; RDC-1, orphan receptor,
A39714; somatostatinR4, somatostatin, A47457; V1a, vasopressin, A53046;
V1b, vasopressin, A55089; A2b, adenosine, P29275; M1, muscarinic;
P11229; APJ, orphan, P11229; C5a, chemotactic, P21730; CASR,
extracellular calcium-sensing; P41180; CB1, cannabinoid, P21554; CRF,
corticotropin-releasing factor, P34998; FSHR, FSH, P23945; CC-BR,
gastrin/cholecystokinin type B, P32239; GRFR, GH-releasing, Q02643;
GnRH, gonadotropin-releasing, P30968; CALR, calcitonin, X69920 (G);
GlucR, glucagon, U03469 (G); OR- , opioid, U10504 (G); Mel-1a,
melatonin, U14108 (G); SecR, secretin, U20178 (G); PTHCaR, parathyroid
cell calcium-sensing, U20759 (G); Br1, bradykinin, U22346 (G); 1AD,
alpha-1 adrenergic, L31772 (G); GalR, galanin, L34339 (G); H1R,
histamine, D14436 (G); TRHR, TRH, D16845 (G); MetGlu5a, metabotropic
glutamate, JC2132 (P); 1AD, alpha-1 adrenergic, JC2331 (P); Br2,
bradykinin, JH0712 (P); NK-1, substance P, P25103; NM-BR, neuromedin-B,
P28336; NT-R, neurotensin, P30989; NPY1R, neuropeptide Y, P25929;
OR-µ, opioid, P35372; RHO-fish, rhodopsin, P35356; EP-1,
prostaglandin E1, P34995; PTH-R, PTH, Q03431; OR- , opioid-rat,
S39015 (P); ORL-1, opioid-orphanin FQ, S43087 (P); ET-B, endothelin,
S44866 (G); D2, dopamine, S62137 (G); STE2-yeast, pheromone factor,
P06842; SSTR2, SSTR3, SSTR5, SSTR2A-mouse, SSTR2B, SSTR1, somatostatin,
P30874, P32745, P35346, P30875, P30934, P30872, respectively; PACAP-R,
pituitary adenylate cyclase-activating peptide, D17516 (G); TR-R,
thrombin, P25116; TSH-R1, TSH, S49816 (G).
|
|
It has been suggested from pharmacology studies that a
subtype of the GHS-R may be expressed in the pituitary gland. For
example, GHRP-2, in contrast to GHRP-6, has been reported to increase
cAMP levels in cultured pituitary cells rather than activating
phospholipase C (80). However, since both GHRP-2 and GHRP-6 bind with
high affinity to GHS-R, these differences might be explained by each
ligand conferring a different conformation on the GHS-R. As a
consequence, the receptor could couple to different G proteins and
hence signal through alternative signal transduction pathways. This
behavior is illustrated by the octopamine/tyramine receptor, where the
presence of a single hydroxyl group on the ligand results in
differential coupling to different second messenger systems (81).
Activation of alternative pathways would also depend on the relative
concentration of specific G proteins. Therefore, although we cannot
rule out the existence of different pituitary receptor subtypes for
GHRPs being involved in the control of GH release, the experimental
observations can be rationalized by the same receptor coupling to
different G proteins.
A. Chromosomal localization
The human GHS-R was mapped by fluorescence in situhybridization to band 3Q26.2 (82). Genes whose deficiencies affect
GH release do not map to this region. Interestingly, however, this
location is close to the map position reported for the
Brachmann-de-Lange Syndrome, a pre- and postnatal growth deficiency
(83, 84, 85). Mapping of the Brachmann-de-Lange Syndrome is based on
chromosome duplication and translocation mutants, which always include
region 3q26 (interval 3q26.31–q27.3). Given the close proximity
between the GHS-R gene and the presumed Brachmann-de-Lange location, it
will be important to determine whether these subjects respond to
MK-0677 treatment and whether they have alterations in the gene
encoding the GHS-R.
|
VII. Action of the Peptidomimetic GH Secretagogues in the Central
Nervous System
|
|---|
The central effects of GHRP-6 and the peptidomimetics are inferred
by the demonstration in sheep and guinea pigs that the doses required
to cause GH release are at least 10- to 100-fold lower when injected
into the third ventricle than when the compounds are administered
peripherally (86, 87). Intravenous administration of GHRP-6 and the
peptidomimetics L-692,429, L-692,585, and MK-0677 into conscious rats
and mice increased c-fos expression in the arcuate nucleus
at concentrations consistent with their relative biological activites
in releasing GH (86, 88, 89, 90). In situ hybridization
histochemistry studies with little (lit/lit) mice and
dwarf (dw/dw) mice treated with MK-0677 (Fig. 13) also showed pronounced increases in
c-fos expression (89). Lit/lit mice have a
reduced capacity to release GH because they lack a functional GHRH
receptor (74, 91, 92, 93), and dw/dw mice lack GH (94, 95).
Therefore, activation of fos by GHRP-6 and the
peptidomimetics cannot be explained by indirect effects on GHRH
receptors or through GH (86, 88, 89, 96), but rather as a direct
consequence of a central action of GHS-R ligands. In rats treated with
GHRP-6, approximately 25% of neurons showing an increase in
c-fos activity express GHRH mRNA, and 51% express NPY mRNA
(97). The activation of GHRH-containing neurons suggests that GHRP-6
and the peptidomimetic secretagogues stimulate the release of GHRH.
Indeed, increases in GHRH have been measured in
hypothalamic-pituitary portal vessels of sheep after systemic treatment
with GHRPs (98, 99). The finding that GHRP-6 also activates
NPY-containing neurons potentially explains the increased feeding
behavior seen in rats treated with GHRP-6 (100) and the observed
increases in corticosteroid and ACTH release (101, 102). Clearly,
activation of Fos immunoactivity or c-fos mRNA expression
should not be considered a direct and sole marker since other neurons
in the brain that do not express Fos may be activated, and of course
those expressing Fos might be activated indirectly as a result of
derepression.
View larger version (87K):
[in this window]
[in a new window]
|
Figure 13. Activation of c-fos gene
expression in the central nervous system of control and
dw/dw mice after intraperitoneal MK-0677 administration
(89). Wild type C57 BL/6J mice and dw/dw mice were
injected intraperitoneally with either PBS or MK-0677 (50 mg/kg) and
killed 40 min later. In situ hybridizations for
c-fos mRNA were performed on coronal brain sections.
Images have been artificially colored: blue represents
background; white e represents the most intense signals
(23).
|
|
Experiments were designed to measure effects of GHRP-6 and the
peptidomimetics on individual neurons projecting to the median
eminence. Recordings from electrodes implanted in the arcuate nucleus
of anesthetized rats showed that intravenous administration of GHRP-6
and L-692,585 stimulated electrical activity in secretory neurons that
had been identified antidromically as projecting to the median eminence
(88). Use of the retrograde tracer Fluorogold showed that 68–82% of
cells excited by GHRP-6 project outside the blood-brain barrier (103).
Interestingly, a subset of arcuate neurons that did not project to the
median eminence was inhibited by GHRP-6 and L-692,585(88). Since
previous studies have shown that somatostatin-containing cells in the
arcuate nucleus also do not project to the median eminence, it is
tempting to speculate that GHS-R ligands inhibit somatostatin release
from these cells. Reducing somatostatin tone on GHRH neurons would
facilitate GHRH release (104).
In situ hybridization studies using a cDNA probe selective
for GHS-R 1a to sections of rhesus monkey and rat brains demonstrate
that the receptor is expressed in the arcuate nucleus (16).
Localization of expression in this area of the hypothalamus is
consistent with electrophysiology experiments and c-fos
expression, suggesting that these molecules act on GHRH-containing
arcuate neurons (97). More complete localization studies have been
completed in rat brain and pituitary gland and show that GHS-R is
expressed in the anterior pituitary gland and in regions of the brain
outside those generally considered to be involved in GH release (Fig. 14 and 105 . Indeed, expression of
GHS-R is seen in the anterior hypothalamus, suprachiasmatic nucleus,
supraoptic nucleus, ventromedial hypothalamus, arcuate nucleus, dentate
gyrus, tuberomamillary nucleus, pars compacta of substantia nigra, the
ventral tegmental area, dorsal raphe nuclei, and median raphe nuclei
(105).
View larger version (112K):
[in this window]
[in a new window]
|
Figure 14. In situ hybridization of GHS-R
mRNA in rat brain (105). 33P-Labeled rat GHS-R
oligonucleotide probes were used for in situ
hybridization of coronal rat brain sections. Signals were highly
specific and could be displaced by 100-fold molar excess of cold probe.
Abbreviations are as follows. Top left panel: AHA,
anterior hypothalamic area; Sch, suprachiasmatic nucleus; SO,
supraoptic nucleus. Top right panel: ARC, arcuate
nucleus; VMH, ventromedial hypothalamus. Bottom left
panel: DG, dentate gyrus; CA2/CA3, CA2 and CA3 regions; TM,
tuberomammillary nucleus. Bottom right panel: SNC, pars
compacta of substantia nigra; VTA, ventral tegmental area. [Reprinted
from X.-M. Guan et al.: Mol Brain Res 48:23–29,
1997, with kind permission from Elsevier Science-NL, Sara
Burgerhartstraat 25, 1055 KV Amsterdam, The Netherlands.]
|
|
The presence of GHS-R RNA in regions of the brain outside the arcuate
nucleus is remarkable. The hippocampus is enriched with
neurotransmitter systems and has been implicated in learning and memory
(106, 107). The substantia nigra and ventral tegmental areas are main
centers for dopaminergic cell bodies that are involved in many
biological functions such as motor control and reinforcement behavior
(108). The dorsal and median raphe nuclei are centers for serotonergic
neurons that project to different parts of the central nervous system
and are implicated in a variety of functions including nociception,
affective behaviors, and feeding (109). Expression of the GHS-R in
brain regions not generally associated with GH release is intriguing
and suggests a broader physiological significance for the role of the
natural ligand of the GHS-R.
|
VIII. Peptidomimetic GH Secretagogues in Vivo
|
|---|
A. Animal models
Experiments with cultured pituitary cells demonstrated that
in vitro desensitization to the peptidomimetic GH
secretagogues occurred rapidly. However, animal studies with a compound
having a relatively long half-life suggested that the effects on GH
release were sustained (43). In guinea pigs, when GH was monitored at
10-min intervals during a constant infusion of L-692,585, a sustained
amplification of episodic GH release was observed, similar to that seen
with GHRP-6 (7, 87). Intriguingly, L-692,585 initiated GH pulsatility,
suggesting that this class of compounds has the capacity to reset the
ultradian rhythm of GH release (87).
The action of the peptidomimetic GH secretagogues, L-692,429 and
L-692,585 (5–100 µg/kg), on GH dynamics was evaluated in a crossover
design with four male and four female beagle dogs (110, 111). Peak GH
levels were recorded within 5–15 min after dosing, and L-692,585,
consistent with its in vitro potency, was effective at
approximately 1/20th of the dose compared with L-692,429 (111). No
sex-related difference was detected and, apart from small transient
increases in ACTH and cortisol, both compounds were very selective at
stimulating GH release (89, 103). In a more extensive chronic study in
which L-692,585 was given once daily for up to 14 days, desensitization
to repeated dosing was not observed. Increases in IGF-I were evident
6 h after dosing, but the increase was transient and IGF-I levels
returned to baseline within 24 h. PRL, insulin, and T4
levels were unaltered over the course of the study (111).
Pharmacokinetic oral bioavailability measurements had established that
MK-0677 was a viable candidate for once-daily oral dosing (50, 52). To
determine whether the effects on GH release could be sustained during
repeated oral treatment, dogs were treated with MK-0677 (1 mg/kg) for 4
days. On days 1 and 4, blood was collected at 15-min intervals for
8 h and assayed for GH. MK-0677 treatment resulted in sustained
amplification of the pulsatile profile of GH. However, because the
magnitude of GH release caused by MK-0677 decreased markedly by the
fourth day (59), a second study of 14 days duration was implemented to
determine whether complete desensitization would occur during more
prolonged treatment. Although reduced amplitude of the GH response to
MK-0677 was again observed by day 4, treatment for up to 14 days did
not result in further attenuation. A sustained increase in serum IGF-I
levels accompanied the reduced amplitude of GH release (54).
Interestingly, a similar study with the shorter acting secretagogue
L-692,585 gave neither a reduced GH response nor a sustained increase
in IGF-I levels during repeated daily treatment (111). When dogs were
dosed with MK-0677 on alternate days for up to 9 days, reduced GH
responses to repeated dosing was not evident. This particular dosing
regimen allowed IGF-I to return to basal levels before dogs
received the next dose of MK-0677. Based on these observations it was
speculated that tachyphylaxis might be associated with increases in
IGF-I rather than desensitization of the GHS-R. To address this
possibility, dogs were treated with MK-0677 on day 1; on days 2 and 3
the dogs were dosed with porcine GH; and on day 4 the dogs were treated
with MK-0677. At the time of treatment on day 4, GH concentrations had
returned to basal levels but IGF-I was still elevated. The GH response
to MK-0677 on day 4 was markedly attenuated compared with day 1. Thus
reduced responsiveness to chronic MK-0677 treatment is not necessarily
due to desensitization of the GHS-R but might be explained by reduced
responsiveness of the GH/GHRH axis caused by a sustained increase in
IGF-I (54). Perhaps chronically, IGF-I feeds back on arcuate neurons to
reduce the expression of both GHRH and the GHS-R. The demonstration
that the magnitude of the stimulatory effect of MK-0677 on GH release
is limited by negative feedback is very attractive clinically because
it prevents hyperstimulation of the GH/IGF-I axis.
Like GHRP-6, L-692,429 and L-692,585 also cause transient increases in
ACTH and cortisol (110, 111). The effect on ACTH is probably mediated
by action of the secretagogues on the hypothalamus, because in cultured
pituitary cells the peptidomimetics do not significantly increase ACTH
release. MK-0677 also increased cortisol levels when given acutely;
however, during chronic administration the stimulatory effect on
cortisol became insignificant (53, 54). When MK-0677 was given
chronically to dogs on alternate days, just as observed with GH, the
cortisol response was not attenuated (54). These results suggest that
the GH and cortisol responses are linked and perhaps similarly
attenuated during repeated daily treatment through an IGF-I-mediated
pathway.
Interesting anabolic activities were reported for the Genentech
compound G-7039. Body weight gain in 150-day-old female Sprague Dawley
rats was significantly greater when G-7039 was administered by
subcutaneous minipump twice daily for 14 days as compared with the body
weight gain after subcutaneous minipump infusion of G-7039 for 14 days
(58). The implication is that optimal growth requires intermittent
administration of these secretagogues rather than continuous treatment.
It remains to be tested whether this applies to species other than
rats.
B. Clinical studies in humans
GH treatment may prove to be beneficial to a variety of subjects
other than GH-deficient children. Most individuals over the age of 60
yr might be considered GH deficient according to young adult standards
(112, 113, 114), and it has been suggested that this relative deficiency is
responsible for decreases in bone and muscle mass and increased
adiposity during aging (115). Since GH increases bone turnover, GH in
combination with an inhibitor of bone resorption may have great benefit
in severe osteoporosis. GH deficiency is associated with lipid profiles
that favor the likelihood of atherosclerosis and mortality due to
cardiovascular disease (116, 117, 118, 119). Depressed subjects present a
significant decrease in 24-h GH production (120). Whether the decreased
amplitude of the GH pulses reported in major depressive illness is due
to changes in the activity of neurotransmitters or is related to GH
itself remains to be established.
Perhaps the greatest potential for GH replacement is in the frail
elderly population. Although pulsatile GH secretion declines during
aging, rodent studies show that the pituitary gland continues to
synthesize GH and remains responsive to GHRH (121, 122). The advantage
of increasing GH in the elderly is suggested by the recent
demonstration that in a population of healthy men over 60 yr old,
once-daily GH treatment over a 3-month period increased lean body mass,
muscle mass, and thigh strength measured by isokinetic dynamometry
(123). Therefore, identification of a compound that rejuvenates the
GH/IGF-I axis provides a way to optimally evaluate the potential
clinical benefits of reversing GH deficiency in the musculoskeletally
impaired elderly
Top
Abstract
I. Introduction
