Commentary
Signal transduction systems employed by synthetic GH-releasing
peptides in somatotrophs
C Chen, D Wu and I J Clarke
Introduction
Around the time when the endogenous hypothalamic
releasing factor for growth hormone (GH) was isolated, a
new range of synthetic peptides were also shown to be
specific releasors of GH. These peptides were originally
developed in the late 1970s by Bowers et al. (1980) who
synthesized the enkephalin analogue Tyr-D-Trp-Gly-
Phe-Met-NH2 and found that it specifically stimulated
GH secretion. Although this pentapeptide was a weak GH
secretagogue it was the first non-GH-releasing hormone
(GHRH) found to act on the anterior pituitary to specifi-
cally release GH and was used as a model to design
more potent forms of GH-releasing peptides (GHRPs)
(Momany et al. 1981). The most potent of the first
generation analogues was GHRP-6 (Bowers et al. 1984,
Momany et al. 1984), a hexapeptide which stimulates GH
release in a variety of animal species and in man (Bowers
et al. 1984, 1991, Wu et al. 1994a). Since then, a range of
compounds has been developed with varying potency
including GHRP-1 and GHRP-2. In spite of their
obvious eYcacy and specificity, no endogenous homologue
has been found for these synthetic peptides and the
receptor through which they exert their eVect has not
been identified. Nevertheless, some progress has been
made in defining the signal transduction systems that
mediate responses to GHRPs. Some of the peptides act in
a synergistic or additive way with endogenous GHRH and
this has provided a basis for their clinical use. Since the
GHRPs are short peptides (5–7 amino acid residues), they
are synthesized easily and are not as readily degraded in
plasma as GHRH; these features ameliorate their clinical
potential. Because of their chemical nature the GHRPs are
eYcacious when administered orally or intravenously
(Bowers et al. 1990, 1992).
If endogenous GHRPs do exist then how could they be
rationally integrated into the picture for the control of GH
that has been developed on the basis that GHRH stimulates
and somatostatin (SRIF) inhibits secretion (Brazeau
et al. 1973, Rivier et al. 1982, Guillemin et al. 1982)? A
regulatory model has been established in the rat which
dictates that GH release is the sum of the combined eVects
of GHRH and SRIF (Tannenbaum & Ling 1984). There
is some experimental evidence for this, since reciprocal
alterations in the hypophysial portal blood levels of GHRH
and SRIF seem to exist (Plotsky & Vale 1985) and relate
to the pulsatile secretion of GH. On the other hand,
studies in sheep (Frohman et al. 1990, Thomas et al. 1991,
Magnan et al. 1994) do not indicate such a close relationship
between GHRH, SRIF and GH, calling into question
the general applicability of the rat model. Perhaps
there is a role for an endogenous GHRP in the regulation
of GH secretion in sheep and other species.
Given the range of GHRPs that have now been
synthesized and studied, it seems timely to review what is
known about their mechanism of action. An excellent
recent review by Korbonits & Grossman (1995) discussed
in vivo and in vitro actions of GHRPs on GH secretion and
the possible clinical application of GHRPs. The present
article focusses on the intracellular signal transduction
systems in somatotrophs employed by GHRPs with a brief
overview of the site of action, chemical structure and
possible physiological roles.
Site of action of GHRPs
It is likely that GHRPs stimulate GH release by both
direct action on pituitary somatotrophs and hypothalamus
(Bowers et al. 1991, Fletcher et al. 1994, Guillaume et al.
1994, Fletcher et al. 1995).
Hypothalamic action GHRPs do not appear to inhibit
SRIF release (Hao et al. 1988, Guillaume et al. 1994,
Korbonits & Grossman 1995). The issue of whether or not
these peptides have any hypothalamic action on GHRH or
other factors is a little unclear. GHRP-6 has been shown
to activate neurones of the arcuate nucleus that project to
the median eminence (****son et al. 1995), but infusion in
the third ventricle had a minor eVect on the pulsatile
release of GHRH in conscious ewes and no eVect on SRIF
secretion (Fletcher et al. 1995). Moreover, GHRPs did not
acutely stimulate GHRH release with perifusion of rat
hypothalamus in vitro (Korbonits & Grossman 1995).
381
Journal of Endocrinology (1996) 148, 381–386 ? 1996 Journal of Endocrinology Ltd Printed in Great Britain
0022–0795/96/0148–0381 $08.00/0
Hexarelin, a more potent peptide, elevated GHRH levels
in the hypophysial portal blood of rams (Guillaume et al.
1994) with no eVect on SRIF secretion. GHRPs are able
to potentiate GH release in response to a maximally
stimulating dose of GHRH (Bowers et al. 1990) and
anti-GHRH antibody infusion cannot abolish the eVect
of GHRP-6 on GH release from rat pituitary glands
(Locatelli et al. 1994). These latter observations suggest
action at the pituitary level or, possibly, through a non-
GHRH hypothalamic factor yet to be identified (Bowers
et al. 1990).
Pituitary action All GHRPs act directly on the pituitary
gland in vitro (Akman et al. 1993, Bowers 1993, Cheng
et al. 1993, Bowers et al. 1994, Wu et al. 1994a,b, Chen &
Clarke 1995b) and in vivo (rat; Mallo et al. 1993, sheep;
Fletcher et al. 1994) to increase GH secretion. Long-term
(days) GHRP-6 treatment of GHRH-deprived infant rats
stimulated GH gene expression which counteracted the
GHRH deprivation (Locatelli et al. 1994). This action of
GHRP-6 on GH synthesis is, however, abolished when
the pituitary is anatomically disconnected from the hypothalamus
which suggests an involvement of unknown
hypothalamic factors. The eVect of GHRPs on GH
synthesis requires further study before any conclusions can
be drawn.
Structure and potency of GHRPs
Since the development of the first peptide in the GHRP
series (GHRP-6), a range of more potent compounds
including non-peptidergic GH secretagogues have been
developed (Deghenghi et al. 1992, Akman et al. 1993,
Bowers 1993, Smith et al. 1993). Because it has been
demonstrated that non-peptidergic GH secretagogues act
in a similar way to GHRP-6 (Cheng et al. 1993, Smith
et al. 1993), these non-peptidergic compounds are not
separately considered below.
The structure and relative potencies of the most wellcharacterized
of the GHRPs is given in Table 1. GHRP-2
is the most potent, being equivalent to GHRH in terms of
its ability to stimulate GH release from sheep pituitary cells
in vitro (Wu et al. 1994b). GHRP-1 has been reported to
release GH in the rat with the same potency as GHRH
(Akman et al. 1993), but this is not true for ovine cells (Wu
et al. 1994a). GHRP-6 was found to stimulate GH
secretion in most human acromegalic tumours in vitro
whereas GHRH stimulated GH release in only some
(Renner et al. 1994). A mutation in the subunit of the Gs
protein has been reported in a proportion of such tumours
as being responsible for constitutive activation of adenylyl
cyclase activity in response to GHRH (Harris et al. 1992);
this may explain the latter finding. The adenylyl cyclase
pathway may not necessarily be stimulated by GHRPs (see
below), whereas this is a recognized second messenger
pathway for GHRH (Harwood et al. 1984).
Intracellular second messenger systems employed by GHRPs
Intracellular free Ca2+ and membrane ion channels
GH secretion is directly related to the intracellular free
calcium concentration ([Ca2+]i) (Lussier et al. 1991). Ca2+
influx through voltage-dependent Ca2+ channels is stimulated
by GHRH and reduced by SRIF (Chen et al.
1989a,b, 1990, 1992, Lussier et al. 1991, Chen & Clarke
1995a). There is no clear evidence, however, that GHRH
or SRIF mobilizes Ca2+ from intracellular Ca2+ storage
sites. On the other hand, GHRP-6 causes the release of
intracellular Ca2+ as well as Ca2+ influx through the
cell membrane (Bresson-Bepoldin & Dufy-Barbe 1994,
Herrington & Hille 1994). In isolated rat somatotrophs,
GHRP-6 evoked dual-phase increases in [Ca2+]i; an
initial phase transient increase in [Ca2+]i due to Ca2+
release which was not blocked by the Ca2+ channel
blocker, and a second long-lasting phase that was due
to Ca2+ influx (Bresson-Bepoldin & Dufy-Barbe 1994,
Herrington & Hille 1994). In ovine pituitary cells,
GHRP-1 caused a subtle and transient increase in [Ca2+]i
even when extracellular Ca2+ was chelated to zero (K
Katoh, C Chen, D Wu, J Zhang, I J Clarke, C Y Bowers
& D Engler, unpublished observations). This probably
involves the generation of inositol trisphosphate (InsP3)
(Adams et al. 1995, Lei et al. 1995) but this has not yet
been proven. In spite of the mobilization of Ca2+ from
intracellular stores, the major contribution to the elevation
of [Ca2+]i is caused by Ca2+ influx via Ca2+ channels. It
would appear that this is an integral factor in the release of
GH in response to GHRPs because Ca2+ channel blockade
abolishes secretion (Akman et al. 1993, Wu et al.
1994a,b).
In somatotrophs, the major Ca2+ channels are the
voltage-gated T- and L-types (Chen et al. 1990, Chen &
Clarke 1995a). Studies of rat and sheep cells have respectively
shown that GHRP-6 and GHRP-2 depolarize the
cell membrane potential leading to the opening of these
channels (Herrington & Hille 1994, Chen & Clarke
1995b). Since this depolarization can only be recorded
with the nystatin-perforated patch clamp configuration
Table 1 Chemical structure and the relative potency of GHRPs in
terms of ability to stimulate GH release. Potency is presented in
relation to the potency of GHRH (unity) in rat and ovine pituitary
cells in vitro
Chemical structure
In vitro
potency
(relative to
GHRH)
Name
GHRP-6 His-D-Trp-Ala-Trp-D-Phe-Lys-NH2 0·01–0·1
GHRP-1 Ala-His-D-‚Nal-Ala-Trp-D-Phe-Lys-NH2 0·1–1
GHRP-2 D-Ala-D-‚Nal-Ala-Trp-D-Phe-Lys-NH2 1
Commentary 382
Journal of Endocrinology (1996) 148, 381–386
(which does not disturb intracellular second messenger
systems), this implies that an intact intracellular second
messenger signalling system is required. In addition, recent
data indicate that GHRP-2 acts on voltage-gated Ca2+
channels via second messenger systems, leading to an
increase in transmembrane L- and T-type Ca2+ currents
(Chen & Clarke 1995b). All these eVects of GHRPs on
electrophysiological properties of the membrane of somatotrophs
resemble the eVects of GHRH but are opposite to
the eVects of SRIF.
The ion channels that are involved in depolarization of
the somatotroph cell membrane are not defined. Studies to
date show that depolarization caused by GHRPs does not
involve the Na+ channel in a major way (Herrington &
Hille 1994, Chen & Clarke 1995b). It is thought that the
voltage-gated Ca2+ channel activation is partly responsible
for the depolarization caused by GHRP-2 (Chen & Clarke
1995b). K+ channels may also be involved but these have
not been fully characterized (Pong et al. 1991, Chen &
Clarke 1995b).
In summary, it is proposed that GHRPs first releases
intracellular Ca2+ and then causes Ca2+ influx by an
increase in membrane Ca2+ permeability. The latter is due
to membrane depolarization and the action of second
messengers on Ca2+ channel proteins. Possible second
messenger pathways involved in the action of GHRPs are
discussed in the following sections.
The cyclic AMP (cAMP)/protein kinase A (PKA)
pathway It is well established that GHRH activates the
cAMP/PKA pathway in somatotrophs and this is fundamental
to the release of GH (Harwood et al. 1984,
Frohman et al. 1992). Part of the eVect of SRIF is via its
inhibition on cAMP formation (Schonbrunn 1990). In
contrast, GHRP-6 and GHRP-1 have no direct eVect on
intracellular cAMP levels in rat and ovine somatotrophs
(Cheng et al. 1989, Akman et al. 1993, Wu et al. 1996).
Nevertheless, GHRP-6 may synergize with GHRH to
increase intracellular cAMP levels (Cheng et al. 1989).
Similar results were also found for the non-peptidergic
analogue L-692,429 (Cheng et al. 1993), but synergy
could not be demonstrated with GHRP-1 and GHRH
(Akman et al. 1993). This discrepancy suggests that an
increase in cAMP levels is not the primary signalling
pathway for GHRP-6 and GHRP-1.
GHRP-2 increased intracellular cAMP levels in ovine
(but not rat) somatotrophs (Wu et al. 1996). GHRP-2-
stimulated GH secretion was also blocked in these cells by
an inhibitor of adenylyl cyclase and a cAMP-binding
antagonist. Thus, in sheep cells at least, GHRP-2 activates
adenylyl cyclase leading to an increase in cAMP levels and
activation of cAMP-dependent PKA. Thus, PKA could
phosphorylate transmembrane Ca2+ channels to modify
their properties in the manner observed by electrophysiological
studies (see above); further work is needed to
clarify this point. A significant species diVerence is seen
in sheep and rat somatotrophs in terms of GH release in
response to GHRP-2 (Bowers 1993, Wu et al. 1994b).
The increase in adenylyl cyclase activity is obtained in
ovine pituitary cells with GHRP-2 but not GHRP-6 or
GHRP-1 (Wu et al. 1996). We propose therefore that, in
ovine somatotrophs, the pathway employed by GHRP-2,
resulting in an increase in cAMP, is diVerent from that
employed by GHRP-6 or GHRP-1 in rat or ovine
pituitary cells (Cheng et al. 1989, 1993, Akman et al. 1993,
Wu et al. 1994a).
The mechanism of activation of adenylyl cyclase by
GHRP-2 is not clear. Although some subtypes of adenylyl
cyclase can be activated by Ca2+ (Cooper et al. 1995), the
subtype of adenylyl cyclase that is involved in the response
to GHRP-2 may not be stimulated by Ca2+ because
blockade of Ca2+ influx did not aVect the increase in
cAMP levels in response to GHRP-2 (Wu et al. 1996).
Although it is clear that GHRH elevates cAMP levels in
ovine somatotrophs, it may act through a cyclase which is
diVerent from that used by GHRP-2, since GHRP-2 and
GHRH have an additive eVect on both cAMP accumulation
and GH secretion when both secretagogues are
applied at maximal doses (Wu et al. 1994b, 1996).
In summary, GHRP-2 promotes cAMP accumulation
in ovine but not in rat somatotrophs via activation of
adenylyl cyclase. This appears to be the pathway responsible
for stimulation of GH secretion by GHRP-2.
GHRP-6 and GHRP-1 do not, however, increase cAMP
levels in ovine, rat or human somatotrophs but do amplify
the cAMP response to GHRH in rat pituitary cells. This
amplification may explain the synergistic action of
GHRP-6 or GHRP-1 and GHRH on GH secretion in
this species.
Protein kinase C (PKC) It has been suggested that the
action of GHRP-6 and the non-peptidergic analogue
L-692,429 to stimulate GH release from rat pituitary cells
is mediated by PKC (Cheng et al. 1991, 1993). The
synergistic eVect of GHRP-6 and GHRH on cAMP
accumulation and GH secretion in rat pituitary cells may
also be mediated by PKC (Cheng et al. 1991). It should be
noted, however, that the specificity of the inhibitor
(phloretin) used in the latter study has not been widely
tested and its eVect on other kinase systems is not defined.
In particular, over the same dose range (10–200 µM),
phloretin increased the opening probability of Ca2+-
activated K+ channels (Koh et al. 1994) which can
hyperpolarize cell membrane potential and prevent the
stimulation of GHRP-6 on GH secretion. Downregulation
of PKC with long-term treatment by
phorbol,12-myristate,13-acetate (PMA; 1 µM) partially
blocked the eVect of GHRP-6 on GH secretion (Cheng
et al. 1991), suggesting some involvement of PKC. It was
shown by Akman et al. (1993) however, that GHRP-1
causes GH release following maximal stimulation of cells
with PMA. In ovine pituitary cells, GHRP-6 does not
Commentary 383
Journal of Endocrinology (1996) 148, 381–386
cause PKC translocation (Wu et al. 1995). Furthermore,
down-regulation of PKC with PMA does not block GH
release in response to GHRP-6 whereas PMA-stimulated
GH release is totally abolished by the same treatment (Wu
et al. 1995).
In contrast, GHRP-2 stimulates PKC translocation
from cytosol to membrane in ovine somatotrophs in
primary culture (Wu et al. 1995). PKC inhibitors (calphostin
C, chelerythrine, staurosporine) and down-regulation
of PKC by phorbol-12,13-dibutyrate causes partial attenuation.
It seems likely therefore, that PKC is at least partially
involved in the action of GHRP-2 (but not GHRP-6) in
sheep cells. It is interesting to note that GHRH also causes
PKC translocation in ovine somatotrophs (Wu et al.
1995). GH secretion in response to GHRH and GHRP-2
can be accounted for by activation of the cAMP/PKA
pathway in ovine somatotrophs, and the concomitant
activation of PKC may play a role in the regulation of GH
synthesis.
InsP3 GHRP-6 and GHRP-1 increase [Ca2+]i through
mobilization of intracellular stores through a mechanism
that probably involves InsP3 (Bresson-Bepoldin & Dufy-
Barbe 1994, Herrington & Hille 1994). Indeed, it has been
reported that GHRP-6 and non-peptidergic GH secretagogues
increased phosphatidylinositol (PI) turnover via
an activation of phospholipase C (PLC) in human acromegalic
tumour cells (Adams et al. 1995, Lei et al. 1995).
Activation of PLC will produce both InsP3 which leads to
the release of Ca2+ from intracellular Ca2+ storage sites,
and diacyglycerol (DAG) which activates PKC. Whether
PLC (or an increase in PI turnover) is activated in ovine or
rat somatotrophs by any GHRP is still an open question.
Investigation of this issue would be useful given the
diVerence in the response of sheep cells to GHRP-2 and
other versions of GHRPs.
Receptor/s for GHRPs
It seems likely that a novel endogenous receptor exists
for GHRPs although there is, as yet, no indication of
its/their identity. It is certainly clear that GHRPs do not
act through the GHRH receptor. Evidence for this is
as follows. (1) A GHRH receptor antagonist inhibits
GHRH-stimulated GH release but not GHRP-6- or
GHRP-1-stimulated GH release (Thorner et al. 1994, Wu
et al. 1994a). A putative GHRP receptor antagonist does
not aVect GH release in response to GHRH (Cheng et al.
1989, Thorner et al. 1994). (2) Use of a radioreceptor assay
for GHRH indicates that GHRP-6 does not compete with
the GHRH-binding sites (Thorner et al. 1994). (3) Functionally,
there is an additive eVect on GH release when
both GHRH and the GHRPs are co-administered at a
maximal dosages (Wu et al. 1994a,b). (4) There is no
cross-desensitization between GHRH and GHRPs
whereas homologous desensitization occurs (Wu et al.
1994a,b). Because GHRPs stimulated GH release in vitro
in the absence of SRIF, it is unlikely that GHRPs act as
an inverse agonist on SRIF receptor. All of this evidence
strongly suggests that there is a novel receptor for
GHRPs.
It should, however, be emphasized that the GHRH
antagonist ([Ac-Tyr1,D-Arg2]-GHRH1–29) did prevent
GH release in response to GHRP-2 in sheep pituitary cells
(Wu et al. 1994b) which suggests some relationship
between the GHRP-2-binding sites and the GHRH
receptor. Previously, GHRP-6 was reported to interact
with a novel low-aYnity GHRH-binding site in rat
anterior pituitary cell membranes (Lau et al. 1991). Meanwhile,
Sethumadhavan et al. (1991) have identified two
classes of GHRP-binding sites using [125I]Tye-Ala-
GHRPs as a ligand. It seems possible therefore that a
receptor exists that has a low aYnity for GHRH and a high
aYnity for GHRP-2. Whether there is some homology
between receptors for GHRH and GHRPs is not known.
Another possibility that we considered was that GHRPs
could bind to a site on the GHRH receptor diVerent from
that employed by GHRH. This was tested using a GC cell
line with over-expression of the GHRH receptor that was
kindly provided by Professor S Melmed (C Chen, J
Zhang, P Farnworth, B Canny, S Petersenn & I J Clarke,
unpublished observations). In these cells, GHRH increased
cAMP levels but GHRP-2 or GHRP-6 did not do
Figure 1 The proposed signalling pathways for GHRPs in
somatotrophs. The binding of GHRPs to a putative receptor
activates the PLC and adenylyl cyclase pathways via G-proteins,
leading to an increase in InsP3 and the activity of PKC and PKA.
InsP3 then releases Ca2+ from the InsP3-sensitive Ca2+ pool and
protein kinases phosphorylate ion channels to increase Ca2+
influx. All of these events would lead to an increase in [Ca2+]i and
GH secretion. G=GTP-binding proteins; PIP2=phosphatidylinositol
(4,5)-bisphosphate.
Commentary 384
Journal of Endocrinology (1996) 148, 381–386
so; the GHRH eVect was blocked by a GHRH receptor
antagonist. This suggests that GHRPs do not act on
GHRH receptors. Nevertheless, it remains possible that
there is more than one GHRP receptor, type one
(GHRP-R1) that is not blocked by the GHRH antagonist
([Ac-Tyr1,D-Arg2]-GHRH 1–29) and type two (GHRPR2)
that may be blocked. There is a possibility that
GHRP-R1 has a similar binding aYnity for GHRP-6,
GHRP-1 and GHRP-2 whereas GHRP-R2 would have
a higher aYnity for GHRP-2 than for GHRP-6 or
GHRP-1. GHRP-R2 may also have a low aYnity for
GHRH-binding which would explain the eVects of
blockade with [Ac-Tyr1,D-Arg2]-GHRH 1–29.
The GHRP receptors (both GHRP-R1 and -R2) are
most likely to be coupled to G-proteins based on the
evidence of activation of adenylyl cyclase by GHRP-2
(Wu et al. 1996), activation of PKC (Cheng et al. 1991,
Wu et al. 1995), release of intracellular Ca2+ (Bresson-
Bepoldin & Dufy-Barbe 1994, Herrington & Hille 1994)
and an increase in PI turnover by GHRP-6 or nonpeptidergic
secretagogues (Adams et al. 1995, Lei et al.
1995). Because both adenylyl cyclase and phospholipase C
(PLC) are activated by GHRPs, Gs and Gq are most likely
to be involved in the response.
Conclusion
As discussed above, two major signal pathways have been
identified in somatotrophs employed by diVerent forms of
GHRP in diVerent animal species. The diVerences in
signalling systems may reflect two subtypes of receptor for
GHRPs. Because of cross-talk between diVerent signalling
systems, Fig. 1 illustrates the signalling pathways employed
by both possible GHRP receptors with GHRP-R1 mainly
coupled to PKC pathway and GHRP-R2 mainly to PKA
pathways.
The therapeutical use of GHRP is quite obvious but its
physiological role is still a mystery. Although the roles of
GHRH and SRIF are well established as endogenous
regulators of GH, the model of reciprocal secretion that
was proposed for the rat (Tannenbaum & Ling 1984)
cannot fully explain the pattern of GH secretion pattern in
sheep and human. Studies on the mechanism of the action
of synthetic GHRPs has revealed a novel regulatory system
for the manipulation of GH secretion. The potency of
GHRPs to release GH is higher in human and sheep than
in the rat. DiVerences in signal transduction pathways have
also been demonstrated between sheep and rat. A heterogeneity
of function is therefore suggested among diVerent
species. It seems unlikely that nature would create a
redundant system or a system that abets therapeutical use,
so we must conclude that an endogenous ligand exists for
GHRP receptors. This endogenous ligand for GHRP
receptors may play an important role in the regulation of
GH secretion in concert with GHRH and SRIF. The
identification of endogenous GHRPs ligands and receptors
provides a challenge for the future.