| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Endocrine Unit of the Department of Metabolic Medicine, Imperial College School of Medicine, Hammersmith Hospital, London, United Kingdom W12 0NN; and Department of Histopathology, Ealing Hospital (M.F.), Southall, Middlesex, United Kingdom UB1 3HW
Address all correspondence and requests for reprints to: Dr. David M. Smith, Endocrine Unit of the Department of Metabolic Medicine, Imperial College School of Medicine, Hammersmith Hospital, London, United Kingdom W12 0NN. E-mail: d.m.smith{at}ic.ac.uk
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Binding sites for GLP-1 have been demonstrated on human
pancreatic ß-cells, rat insulinoma cell lines, lung, stomach, certain
brain regions, heart, and kidney (2). The complementary DNAs encoding
the human (9) and rat (10) pancreatic islet GLP-1 receptor
have been cloned. The rat GLP-1 receptor consists of 463
amino acids, and its structure includes seven transmembrane regions.
Rat and human cloned GLP-1 receptors are characterized by
a high affinity for the lizard venom peptide, exendin-4, and a potent
antagonist effect of the exendin fragment, exendin-(9–39) (11). These
receptors belong to the subclass of the G protein-coupled receptor
family that includes receptors for glucagon, secretin, vasoactive
intestinal peptide, pituitary adenylate cyclase-activating polypeptide
(PACAP), PTH, and calcitonin (10). All of these G protein-linked
receptors have been shown to be coupled, via Gs
(the stimulatory G protein), to the activation of membrane-bound
adenylyl cyclase. In cells transfected with the cloned rat or human
GLP-1 receptors, agonist binding also primarily stimulates
adenylyl cyclase (10, 12), but mechanisms involving increases in
intracellular calcium have also been described (13, 14).
GLP-1 has been shown to increase intracellular cAMP in
insulinoma cell lines (15). We have shown that GLP-1
increased cAMP levels in the GT1-7 mouse
hypothalamic LHRH-secreting cell line (16) and in the mouse 
TSH
pituitary thyrotroph-like cell line (17). However, the cloned
GLP-1 receptors may not be the only GLP-1
receptors. GLP-1 binding and effects have been
demonstrated in human and rat adipose tissue (18, 19) and skeletal
muscle (20) in the absence of messenger RNA for the cloned receptors.
Interestingly, GLP-1 has been associated with either no
change in cAMP levels (20, 21) or a decrease in cAMP levels (22, 23) in
both of these systems.
In the CNS, the presence of GLP-1-like immunoreactive (-IR) cell bodies has been demonstrated only in the nucleus of solitary tract and the ventral and dorsal parts of the medullary reticular nucleus (24). GLP-1-IR nerve fibers are found throughout the brain, but the highest concentrations are in the hypothalamus and brainstem (25). GLP-1-IR has not been detected in the pituitary (26). Recent studies indicate a physiological role for GLP-1 as a central neurotransmitter or neuromodulator. We and others have demonstrated that intracerebroventricular (icv) administration of GLP-1 in rats inhibits food intake (27, 28, 29). Intracerebroventricular injection of GLP-1 in rats also inhibits basal water intake and stimulates urinary excretion of water and sodium. In rats both icv and ip administrations of GLP-1 inhibit basal and angiotensin II-induced water drinking (28, 29).
These studies suggested that GLP-1 has a potential physiological role in the CNS control of drinking behavior. Specific binding sites for GLP-1 have been demonstrated in the CNS, in particular in the paraventricular nucleus, central nucleus of the amygdala, and anterodorsal thalamic nucleus (25, 27, 30). Goke et al. recently demonstrated specific GLP-1-binding sites on the posterior lobe of rat pituitary (31). The presence of high density GLP-1-binding sites in the rat posterior pituitary suggests that this peptide may play a direct (via circulating GLP-1) or an indirect (via hypothalamic GLP-1 and hypothalamic receptors) role in the control of posterior pituitary hormone secretion. The posterior pituitary consists mainly of glial cells, termed pituicytes, and of unmyelinated nerve fibers and axon terminals of neurons whose cell bodies reside primarily in supraoptic nucleus and paraventricular nucleus (32). These hypothalamo-neurohypophyseal fibers deliver the two primary posterior pituitary hormones, arginine vasopressin (AVP) and oxytocin (OT). In addition to AVP and OT, many other neuropeptides, including endothelin, neuropeptide Y, and PACAP, have been located within these neurons. PACAP has been shown to stimulate both cAMP formation and AVP secretion in rat posterior pituitary (33). Interestingly, direct in vitro addition of GLP-1 to isolated posterior pituitaries had no effect on AVP release, but slightly increased OT release (34). It might have been expected that GLP-1 would, like PACAP, increase cAMP and stimulate AVP release.
Here, using a combination of receptor autoradiography and membrane binding assays, we confirm the presence of GLP-1-binding sites in rat neurointermediate lobe membranes and demonstrate for the first time GLP-1 binding in human pituitary. We characterized these sites with respect to binding parameters, cAMP generation, and mol wt and compared them with the cloned GLP-1 receptors. We also examined the effect of GLP-1 on AVP or OT release in isolated rat neurointermediate lobes.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Iodination of GLP-1
GLP-1 was iodinated by the chloramine-T method, and
monoiodinated peptide was purified by HPLC as described previously
(17). Peak fractions were assayed for receptor-binding activity, and
those with high specific binding were aliquoted, freeze-dried, and
stored at -20 C. The specific activity of the label, determined by RIA
(4), was 50 ± 4 becquerels (Bq)/fmol (n = 6).
Animals and tissues
Adult male Wistar rats (BSU, Hammersmith Hospital, UK), weighing
200–250 g, were housed in wire-bottomed cages with ad
libitum access to food and water. All procedures were licensed by
the British Home Office under the Animals (Scientific Procedures) Act
1986. Animals were killed by CO2 asphyxiation,
and the brain, pituitary, and other tissues were rapidly removed and
either frozen in liquid nitrogen and stored at -80 C before membrane
preparation or used immediately for receptor autoradiography or static
incubation of pituitary. For rat hypothalamic membrane preparations the
rat hypothalamus was dissected before freezing as previously described
(35). For rat neurointermediate lobe preparations the posterior and
intermediate lobes were dissected free of the anterior lobe. It was not
considered important to separate the posterior and intermediate lobes,
as Goke et al. had already stated that the intermediate lobe
has no GLP-1-binding sites (31). Human whole pituitaries
were removed from cadavers during postmortem examination at Ealing
Hospital, frozen immediately in liquid nitrogen, and stored at -80 C
before membrane preparation or receptor autoradiography. The patients
(age, 70 ± 4 yr; age range, 44–90 yr; seven men and four women)
had no recorded pituitary or CNS disorders.
Membranes were prepared by the method of homogenization and differential centrifugation as described previously (36). Briefly, tissues were homogenized in ice-cold 50 mM HEPES buffer (pH 7.4; containing 0.25 M sucrose; 10 µg/ml soybean trypsin inhibitor; 0.5 µg/ml pepstatin, leupeptin, and antipain; 0.1 µg/ml benzamidine; 0.1 µg/ml bacitracin; and 30 µg/ml aprotinin), with an Ultra Turrax homogenizer (Merck & Co., Inc.). For the human pituitary membranes one pituitary was used for each preparation. The homogenate was centrifuged for 10 min at 1,500 x g, and the supernatant was centrifuged for 1 h at 100,000 x g at 4 C. The pellet was resuspended in the same buffer without sucrose and centrifuged for 1 h at 100,000 x g at 4 C. Using a hand-held glass-Teflon homogenizer, membranes were resuspended in the same buffer to a final protein concentration of 3–10 mg/ml, as measured by the biuret method and stored at -80 C.
In vitro receptor autoradiography
After removal from the skull, the whole rat pituitary gland was
placed onto a small piece of rat tongue to aid cutting of sections and
then embedded in OCT compound (Raymond Lamb, London, UK) on a cork
block. Sections were snap-frozen in isopentane (cooled in liquid
nitrogen for 20 sec) and stored at -80 C. For the human whole
pituitary, frozen tissue was embedded in OCT compound on a cork block
and placed into the cryostat (-20 C; cryostat microtome, Bright
Instrument Co., Huntingdon, UK) to enable the OCT compound to freeze.
Serial 15-µm cryostat sections were thaw-mounted on gelatin/chrome
alum-subbed slides (37). These were dried under vacuum at 4 C, wrapped
in cling-film, and stored at -20 C. On the day of use, stored slides
were allowed to equilibrate at room temperature before removal of
cling-film. Sections were preincubated in ice-cold 25 mM
HEPES assay buffer (pH 7.4) containing 2 mM
MgCl2, 0.1% bacitracin, 0.05% Tween-20, 0.1
mM diprotin A (Ile-Pro-Ile, Sigma), and 1%
BSA for 20 min. Excess liquid was drained from the slides, which were
then incubated in assay buffer containing
[125I]GLP-1 (1000 Bq/ml) for 90
min at 22 C. Nonspecific binding was determined in the presence of 200
nM unlabeled GLP-1. After incubation, the
slides were washed four times for 30 sec each time in ice-cold assay
buffer, rinsed once in ice-cold distilled water, and dried overnight at
4 C under vacuum. This was followed by dry apposition to
3H-labeled Hyperfilm (Amersham International) in a light-proof cassette that was stored at -80
C for 7–10 days before developing. Hematoxylin and eosin stain (H&E)
was employed to distinguish between the anterior and posterior portions
of the pituitary.
GLP-1 receptor binding studies
GLP-1 binding studies were essentially carried out
as previously described (16). In brief, rat neurointermediate lobe (50
µg) or anterior lobe (100 µg) membranes were incubated for 90 min
in silanized polypropylene tubes together with
[125I]GLP-1 (750 Bq, 30
pM) at room temperature (20 C) in binding buffer for
autoradiography in a final assay volume of 0.5 ml. Bound and free label
were separated by rapid filtration of the entire incubation mixture
through a glass-fiber filter (GF/B, Whatman, Clifton, NJ), presoaked
for 24 h in 0.3% polyethylenimine, under reduced pressure using a
Brandel M24 cell harvester (Biomedical Research Instruments, Inc., Gaithersburg, MD). The filters were washed six times with
2 ml ice-cold 50 mM Tris-HCl buffer (pH 7.4) and dried, and
the radioactivity bound to the filter was measured using a 
-counter.
Specific binding was calculated as the difference between the amount of
[125I]GLP-1 bound in the absence
(total) and presence (nonspecific) of 200 nM unlabeled
peptide. To determine the time taken to achieve equilibrium binding,
assays were terminated at varying time points (from 2–240 min), and
specific binding was calculated. Equilibrium competition curves were
constructed with increasing amounts of unlabeled GLP-1
(0–200 nM). The exendin peptides, exendin-3, exendin-4,
and exendin-(9–39), were also investigated for competition at
GLP-1-binding sites at a single concentration of 200
nM.
GLP-1 binding studies of human whole pituitary were essentially carried out as described above with the following minor modifications. Pituitary membranes (100 µg) were incubated with [125I]GLP-1 (1,000 Bq, 40 pM) at 20 C for 90 min. The binding site-[125I]GLP-1 complex was separated from free tracer by centrifugation at 15,600 x g for 2 min. Nonspecific binding was defined in the presence of 1 µM unlabeled peptide. Equilibrium competition curves were constructed with the concentration of unlabeled peptide [GLP-1, exendin-4, and exendin-(9–39)] varied from 0–1 µM. Analysis of all binding data, human and rat pituitary, was carried out by nonlinear regression using the ReceptorFit program (Lundon Software, Cleveland, OH).
Chemical cross-linking of receptor-ligand complexes
Chemical cross-linking and SDS-PAGE were performed essentially
as previously described (38). In brief, membranes (200 µg) were
incubated with 10,000 Bq (0.3 nM)
[125I]GLP-1with or without unlabeled
GLP-1 (1 µM) as described above, but bound
label was separated from free label by centrifugation at 15,000 x
g for 2 min at 4 C. Pellets were washed with 0.5 ml binding
buffer at 4 C. The cross-linking reaction was initiated by the addition
of BSOCOES in Me2SO to a final concentration of 3
mM. After incubation for 30 min at 4 C, the
reaction was quenched by washing the membranes twice with 500 µl
ice-cold 50 mM Tris-HCl (pH 7.4). Membranes were
recovered by centrifugation, and the pellet was taken up in 40 µl
Laemmli sample buffer (39). Protein was separated on a 4% acrylamide
stacking-10% acrylamide resolving gel. After drying, the gels were
exposed to Kodak X-Omat film (IBI, Cambridge,
UK) for up to 2 weeks at -80 C using an image-intensifying screen.
Effects of GLP-1 on cAMP in the rat neurointermediate
lobe
Adult male Wistar rats (200–250 g) were killed by
CO2 asphyxiation, and the neurointermediate lobes
were immediately dissected free of the anterior pituitaries. The
neurointermediate lobes were transferred to a Krebs-Ringer (KRB)
bicarbonate solution, pH 7.4 (120 mM NaCl, 5 mM
KCl, 2.6 mM CaCl2, 1.2 mM
KH2PO4, 0.7 mM
MgSO4, 22.5 mM
NaHCO3, 1.8 g/liter BSA, and 0.1 g/liter ascorbic
acid). Before starting the experiment, the neurointermediate lobes were
singly immersed in wells containing 1 ml fresh KRB and maintained at 37
C for 60 min in 95% O2-5%
CO2. Neurointermediate lobes were washed twice in
KRB and exposed for 15 min to KRB buffer containing 2 mM
isobutylmethylxanthine (IBMX) alone or with GLP-1 (1–1000
nM) or forskolin (100 µM). Alternatively, to
investigate any inhibitory effect that GLP-1 might have on
cAMP formation, neurointermediate lobes were exposed for 15 min to KRB
buffer containing 2 mM IBMX alone or with PACAP-(1–38)
(100 and 1000 nM) or PACAP-(1–38) (100 nM)
plus GLP-1 (10 and 100 nM). PACAP-(1–38) is a
known stimulator of cAMP in the posterior pituitary (33). The
neurointermediate lobes were then individually homogenized in 500 µl
75% ethanol containing 16 mM HCl, using a hand-held
glass-Teflon homogenizer and extracted overnight at -20 C. The samples
were dried by rotary evaporation, and cAMP concentrations were measured
using a cAMP RIA kit (NEN Life Science Products,
Steven-age, UK).
Effect of GLP-1 on neurohypophyseal hormone secretion
in isolated neurointermediate lobes
Neurointermediate lobes were dissected from adult male Wistar
rats (200–250 g) as described above. They were then singly immersed in
wells containing 1 ml fresh KRB and maintained at 37 C in 95%
O2-5% CO2. Before starting
the experiments, the neurointermediate lobes were equilibrated for 60
min in KRB. Tissues were then exposed for 15 min to KRB (control) or
for 15 min to KRB together with GLP-1 at the required
concentration (1–1000 nM). To check that the
neurointermediate lobes could be stimulated, the tissues were exposed
for 15 min to KRB containing 56 mM KCl and 69
mM NaCl to maintain medium osmolarity (high
K+ depolarization treatment). Medium was
collected after each period of stimulation and stored at -20 C until
AVP and OT levels were measured by specific RIA (reagents and methods
from Biogenesis, Poole, UK). The effects of various doses
of GLP-1 on basal and 56 mM
K+-evoked release of AVP and OT in each
experiment were estimated by comparing the mean secretions from six
neurointermediate lobes.
Statistical analysis
Results are shown as the mean ± SE. For the
cAMP and posterior pituitary hormone secretion experiments data were
compared by repeated measures ANOVA with subsequent post-hoc
Tukey’s tests (Systat, Evanston, IL) between control and experimental
groups, with P < 0.05 considered to be statistically
significant. For binding data, analysis of one-site vs.
two-site competition curves was performed using the F test, with
two-component fits considered significant at P <
0.05.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
shows specific binding for
[125I]GLP-1 in the rat pituitary
gland. Low specific binding of radiolabeled GLP-1 was
found in the anterior pituitary, whereas high density specific binding
sites were clearly visible in the posterior pituitary. There was no
binding of [125I]GLP-1 in the
presence of 200 nM unlabeled peptide (Fig. 1C).
Investigation into the distribution of
[125I]GLP-1-binding sites in the
human pituitary revealed dense localized binding in one discrete
region, with low specific binding in the remainder of the tissue (Fig. 1, E and F). Histological staining of the same slices (Fig. 1, A and D)
using H&E confirmed that the area displaying high specific binding
sites for [125I]GLP-1 was indeed
the neural tissue of the posterior lobe, at least in the rat. Analysis
of the stained sections revealed that hematoxylin had stained the
nuclei of the anterior pituitary cells, whereas no such staining was
visible in the nerve endings of the posterior pituitary (Fig. 1, A and
D). The intermediate lobe can be seen on the rat H&E-stained section in
between the posterior and anterior pituitary and does not show any
GLP-1 binding (Fig. 1, A and B). In the human pituitary,
the relatively small intermediate lobe is not clearly visible (Fig. 1D).
|
). All
future experiments were carried out employing a 90-min incubation at 20
C. Figure 2B shows the results from the equilibrium competition studies
in rat neurointermediate lobe membranes. Analysis of
[125I]GLP-1 and competition by
unlabeled GLP-1 indicated one class of binding site with
an estimated Kd of 0.2 ± 0.01
nM (n = 3) and a binding capacity
(Bmax) of 600 ± 33 fmol/mg protein (n
= 3). In addition, exendin-3, exendin-4, and exendin-(9–39) were all
able to completely inhibit
[125I]GLP-1 binding at a
concentration of 200 nM (results not shown). Figure 3 shows the equilibrium competition
curves in human pituitary membranes constructed using
GLP-1, exendin-4, and the antagonist, exendin-(9–39).
Analysis of [125I]GLP-1 and
competition by unlabeled GLP-1 indicated one class of
binding site with an estimated Kd of 0.82 ±
0.05 nM and a Bmax of 680 ± 93
fmol/mg protein (n = 3). Specific radioligand binding was
inhibited in a concentration-dependent manner by all unlabeled
peptides. The rank order of potency was as follows: GLP-1
(Kd = 0.82 ± 0.05 nM; n =
4) = exendin-4 (Ki = 2.6 ± 0.37
nM; n = 3) > exendin-(9–39)
(Ki = 48 ± 7.1 nM; n = 3).
The structurally related peptides, glucagon, vasoactive intestinal
peptide, and PACAP, and the unrelated peptides, calcitonin gene-related
peptide and neuropeptide Y, in concentrations up to 1 µM
were not able to inhibit
[125I]GLP-1 binding, thereby
establishing the specificity of binding (results not shown).
|
|
Chemical cross-linking of
[125I]GLP-1 to rat and human
pituitary membranes
Incubation of human whole pituitary membranes with
[125I]GLP-1 followed by covalent
attachment of the radiolabel to the binding protein component
identified a single ligand-binding protein complex with a relative
Mr of 73,100 ± 1,400 (n = 4; Fig. 4). Subtracting 3,300 as the
Mr of GLP-1 gives a
Mr of 69,800 for the binding site. This band was
not detectable, and therefore specific, when the incubations were
carried out in the presence of 1 µM unlabeled
GLP-1. As shown in the representative autoradiograph (Fig. 4), single labeled bands were also detected in samples from rat
neurointermediate lobe, hypothalamic, and lung membranes, which were
analyzed in parallel. Cross-linking with rat neurointermediate lobe and
hypothalamic membranes showed apparently lower Mr
than that of human whole pituitary membranes, with the ligand-binding
site complexes centered at Mr of 59,300 ±
100 and 58,100 ± 900, respectively (n = 3; Fig. 4A),
indicating Mr for the binding site of 56,000 and
54,800, respectively. The Mr of the
ligand-binding site complex in rat lung was also lower than that in
human lung (Mr = 56,600 ± 900; n = 3;
Fig. 4B), giving a Mr of 53,300 for the binding
site alone.
|
; data
expressed as a percentage of the control). The basal production of cAMP
in the unstimulated rat neurointermediate lobe was 29 ± 4
pmol/neurointermediate lobe·15 min (n = 4). GLP-1
at concentrations up to 1000 nM failed to stimulate cAMP
formation in isolated rat neurointermediate lobes (1 nM,
33 ± 6; 10 nM, 35 ± 5; 1000 nM,
32 ± 3 pmol/neurointermediate lobe·15 min; P =
NS; n = 4 for all). Forskolin (100 µM),
used as a positive control, significantly stimulated cAMP production
12-fold to 352 ± 47 pmol/neurointermediate lobe·15 min
(P < 0.001 vs. basal; n = 4). Having
established that the GLP-1-binding site in the
neurointermediate lobe does not appear to be functionally coupled to
increases in cAMP, the effect of GLP-1 on intracellular
cAMP levels increased by PACAP-(1–38), a known stimulator of cAMP in
the posterior pituitary, was investigated in a second series of studies
(Fig. 5). In these studies the basal production of cAMP in the
unstimulated neurointermediate lobe was 14 ± 1
pmol/neurointermediate lobe·15 min (n = 5). As shown in Fig. 5, GLP-1 alone failed to stimulate or inhibit cAMP formation
in the neurointermediate lobe (100 nM
GLP-1, 16 ± 1; 1000 nM
GLP-1, 15 ± 5 pmol/neurointermediate lobe·15 min;
P = NS; n = 5 for both). PACAP-(1–38) stimulated
cAMP levels at both concentrations tested (100
nM, 80 ± 4; 1000 nM,
149 ± 12 pmol/neurointermediate lobe·15 min; P
< 0.01; n = 5 for both). Incubation of neurointermediate lobes
with 100 nM PACAP-(1–38) and increasing
concentrations of GLP-1 produced no significant
potentiation or inhibition of PACAP-(1–38)-stimulated cAMP production
(100 nM PACAP plus 10 nM
GLP-1, 84 ± 9; plus 100 nM
GLP-1, 81 ± 7; plus 1000 nM
GLP-1, 91 ± 11 pmol/neurointermediate lobe·15 min;
P = NS; n = 5 for all).
|
). There was no
significant effect of GLP-1 at any of the tested
concentrations on AVP (Fig. 6A) or OT (Fig. 6B) compared with
unstimulated controls (control AVP release, 199 ± 24
fmol/neurointermediate lobe·15 min; control OT release, 174 ±
18 fmol/neurointermediate lobe·15 min). To determine whether release
of posterior pituitary hormones could be stimulated, we depolarized the
tissue using 56 mM K+. In both
experiments hormone release was greatly and significantly stimulated
[AVP, 8,694 ± 725 fmol/neurointermediate lobe·15 min (40 times
basal); OT, 12,200 ± 1,300 fmol/neurointermediate lobe·15 min
(70 times basal); P < 0.001; n = 6].
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
TSH cells (28 pM), which is the next highest affinity
for GLP-1 receptor binding. We used a concentration of 30
pM for the radioligand in our binding assays, making it
technically difficult to detect a 3-pM affinity site;
certainly, a label of extremely high specific activity would be
required to achieve this sensitivity. The affinity of the rat
neurointermediate lobe GLP-1-binding site is low compared
with those detected by us in TSH (28 pM) and
GT1-7 (70 pM) cells (16, 17),
but lies within the range (0.3–2.5 nM for high affinity
sites) reported in the literature for the other
GLP-1-binding sites (40, 41). The density of the rat
neurointermediate lobe GLP-1-binding site is considerably
higher than those detected in the literature (40, 41) or in TSH (85
fmol/mg protein) and GT1-7 cells (160 fmol/mg
protein) and is approximately equal to that detected in the rat lung,
which expresses the highest previous density of binding (560 fmol/mg
protein) (40). For the [125I]GLP-1-binding sites in human pituitary membranes, analysis of competition for [125I]GLP-1 binding by unlabeled GLP-1 indicated a single class of binding site. The affinity of the human pituitary GLP-1 binding site is a Kd of 0.82 nM, and the density is a Bmax of 680 fmol/mg protein. We used whole human pituitary membrane because it is very difficult to distinguish by eye the posterior lobe from the anterior lobe in the frozen human pituitaries. Thus, the actual density of receptors, even taking into account receptors on, for example, thyrotroph cells, would be much higher in pure posterior lobe membranes. Certainly, if the human is similar to the rat in this study, there would be approximately 16 times higher binding in the neurointermediate lobe than in the anterior lobe. Specific binding of GLP-1 in human pituitary membranes is also inhibited in a concentration-dependent manner by exendin-4 and exendin-(9–39). The affinity of exendin-4 (Ki = 2.6 nM), is similar to that of GLP-1 at the human pituitary binding site, and exendin-(9–39) (Ki = 48 nM) displays lower affinity as expected (11). These data are compatible with both the rat and human pituitary receptors being similar in binding properties to the cloned receptors.
Chemical cross-linking studies of human pituitary membranes for the first time reveal a single ligand-binding protein complex with a Mr of 73,000. This value is higher than those previously reported for GLP-1 receptors on human ß-cell membranes (Mr = 63,000) (42). Single labeled bands are also detected in the rat neurointermediate lobe (Mr = 59,300), hypothalamic (Mr = 58,100), and lung membranes (Mr = 56,600). These values are apparently lower than those in human pituitary membranes. GLP-1 receptor messenger RNA transcripts have been identified in human pancreas, lung, kidney, stomach, heart, and brain by ribonuclease protection analyses, and GLP-1 receptors in human pancreas, heart, and brain (which is the site of synthesis of posterior pituitary receptors on neurons) have the same amino acid sequences (43). The most likely explanation for the species and tissue variations in receptor Mr in cross-linking experiments is differential glycosylation (44). To determine this, deglycosylation protocols need to be employed in the cross-linking experiments and then Mr values compared (45). We have been able to show that the deglycosylated rat neurointermediate lobe binding site has a Mr of 50,000, but could not generate similar data for the human pituitary binding site (results not shown).
To investigate whether this GLP-1-binding site is functionally coupled to a second messenger system, we first measured cAMP accumulation in the rat neurointermediate lobe, as Gs coupling is the preferred coupling of the cloned receptor in both man and rat (9, 10). Previously, it was demonstrated that GLP-1 stimulates cAMP generation in tissue homogenates of rat whole pituitary gland, although in this study the GLP-1-36 amide peptide was used, which is now thought to be inactive (46). On the contrary, in the present study GLP-1 at concentrations up to 1 µM failed to stimulate or inhibit cAMP production in the rat neurointermediate lobe. This may imply that the effects of GLP-1 on cAMP generation are restricted to the anterior pituitary. Hence, it appears that in the neurointermediate lobe the GLP-1-binding site is not functionally coupled to increases in cAMP and adenylyl cyclase activity, so this binding site could be functionally different from that of the cloned receptors. Furthermore, GLP-1 had no effect on PACAP-stimulated cAMP production in the rat neurointermediate lobe. The lack of an inhibitory effect on basal cAMP also differs from some reports of novel GLP-1 receptors in adipose and muscle tissues (22, 23), but is in agreement with others (20, 21).
It is of course possible that the GLP-1 receptors described here are nonfunctional clearance receptors. However, neuroendocrine tissue would be a very unusual site for such clearance receptors, and although the density of binding is very high, the small size of the neurointermediate lobe would make such a role problematical compared with, for example, lung, which expresses both high density and high numbers of GLP-1 receptors. Thus, the neurointermediate lobe receptors may be functional, but their signaling pathways were beyond the scope of this study. GLP-1 signaling in primary islet cultures, ß-cell lines, cells transfected with the GLP-1 receptor complementary DNA, and the brain has been shown to be functionally coupled to both activation of adenylyl cyclase (47, 48) and phospholipase C pathways (14, 49). GLP-1 binding is associated with an increase in cytosolic free calcium (50) and may increase intracellular calcium via activation of a prolonged cAMP-sensitive inward current, leading to membrane depolarization and increases in intracellular calcium (13). It remains to be investigated whether neurointermediate lobe GLP-1 receptors are coupled to decreases in cAMP in the absence of IBMX or increases in calcium and phospholipase C activity. However, effects on intracellular calcium levels are usually seen together with cAMP increases (13, 14, 49).
The obvious question posed by these results is what is the role of the
extremely high density of GLP-1 receptors on the
neurointermediate lobe? The results of Zueco et al. (34)
indicate a very small effect of GLP-1 on basal OT release
(
25% increase) and no effect on basal AVP release. We performed
similar static culture secretion experiments, and in agreement with
Zueco et al. we were unable to show any direct effect of
GLP-1 (1–1000 nM) on AVP secretion.
However, we could also show no effect on OT secretion. Both hormones
were released by K+ (56 mM)
stimulation. GLP-1 has been found to alter water intake
and can have effects on plasma AVP when injected icv. However, it does
seem unlikely that these effects are mediated via a direct interaction
with neurointermediate lobe GLP-1 receptors.
We have successfully demonstrated the presence of high density, specific binding sites for GLP-1 on the posterior/neurointermediate lobe of human and rat pituitary gland. These sites have similar binding characteristics and mol wt as other reported GLP-1 receptors even though there are some minor differences among them. Although we have been unable to assign a second messenger system or function to these GLP-1-binding sites, it seems highly likely that they do exist. The similarity in binding properties to the cloned receptors but differences in G protein coupling could indicate a modified C-terminal part of the receptor, perhaps even a splice variant. The autoradiographical methods presented do not reveal a morphological localization of the GLP-1-binding sites, so it remains to be determined whether they are associated with nerve terminals, pituicytes, or capillaries. Perhaps these GLP-1-binding sites have a role to play at the level of the pituicytes, modulating neuron-glia interaction.
|
Footnotes |
|---|
Received July 22, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-TSH cells. Endocrinology 137:4130–4138[Abstract]
This article has been cited by other articles:
 |
|
 |
|
  Y. Li, T. Perry, M. S. Kindy, B. K. Harvey, D. Tweedie, H. W. Holloway, K. Powers, H. Shen, J. M. Egan, K. Sambamurti, et al. GLP-1 receptor stimulation preserves primary cortical and dopaminergic neurons in cellular and rodent models of stroke and Parkinsonism PNAS, January 27, 2009; 106(4): 1285 - 1290. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Korner, M. Stockli, B. Waser, and J. C. Reubi GLP-1 Receptor Expression in Human Tumors and Human Normal Tissues: Potential for In Vivo Targeting J. Nucl. Med., May 1, 2007; 48(5): 736 - 743. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K K Sidhu, R C Fowkes, R H Skelly, and J M Burrin Exogenous expression of glucagon-like peptide 1 receptor and human insulin in AtT-20 corticotrophs confers cAMP-mediated gene transcription and insulin secretion J. Endocrinol., December 1, 2005; 187(3): 419 - 427. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Perry, N. J. Haughey, M. P. Mattson, J. M. Egan, and N. H. Greig Protection and Reversal of Excitotoxic Neuronal Damage by Glucagon-Like Peptide-1 and Exendin-4 J. Pharmacol. Exp. Ther., September 1, 2002; 302(3): 881 - 888. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. L. Ward, C. J. Small, K. G. Murphy, A. R. Kennedy, M. A. Ghatei, and S. R. Bloom The Actions of Tuberoinfundibular Peptide on the Hypothalamo-Pituitary Axes Endocrinology, August 1, 2001; 142(8): 3451 - 3456. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y.-H. Choi and G. H. Anderson An Interaction between Hypothalamic Glucagon-Like Peptide-1 and Macronutrient Composition Determines Food Intake in Rats J. Nutr., June 1, 2001; 131(6): 1819 - 1825. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
