Am J Physiol Cell Physiol 289: C811–C818, 2005.
First published June 15, 2005; doi:10.1152/ajpcell.00504.2004.
Immature osteoblastic MG63 cells possess two calcitonin gene-related
peptide receptor subtypes that respond differently to [Cys(Acm)2,7]
calcitonin gene-related peptide and CGRP8 –37
Tomoyuki Kawase,1 Kazuhiro Okuda,2 and Douglas M. Burns3
1
Division of Cellular Pharmacology, Department of Signal Transduction Research, and 2Division of Periodontology,
Department of Oral Biological Science, Graduate School of Medical and Dental Sciences, Niigata University,
Niigata, Japan; and 3Medical Research Service, Kansas City Department of Veterans Affairs Medical Center, and
Department of Biochemistry and Molecular Biology, University of Kansas Medical Center, Kansas City, Missouri
Submitted 14 October 2004; accepted in final form 4 May 2005
cAMP; MAP kinase; preosteoblasts; Schild plot
THE NEUROPEPTIDE calcitonin gene-related peptide (CGRP) is
widely distributed throughout the tissues of peripheral systems,
including the cardiovascular and skeletal systems, where it
produces multiple actions. It has been proposed that these
diverse actions result from the activation of two CGRP receptor (CGRPR) subtypes: a subtype 1 CGRP receptor (CGRPR1)
and a putative subtype 2 receptor (CGRPR2) (8, 24, 34). A
Address for reprint requests and other correspondence: T. Kawase, Division
of Cellular Pharmacology, Dept. of Signal Transduction Research, Graduate
School of Medical and Dental Sciences, Niigata Univ., 2-5274 Gakkocho-dori,
Niigata 951-8514, Japan (e-mail: [email protected]).
http://www.ajpcell.org
molecular biological approach has revealed that CGRPR1 is a
heterodimer composed of the calcitonin receptor-like receptor
protein (CRLR or CLR) and receptor activity-modifying
protein 1 (RAMP-1) (23). CGRPR1 is associated with the
formation of cAMP through activation of adenylyl cyclase.
In contrast, the putative CGRPR2 has been identified only in
functional studies, in which receptor-mediated CGRP action is
insensitive to the well-known antagonist of CGRPR1, CGRP8 –
2,7
37, but is instead mimicked by [Cys(Acm) ]CGRP (8, 34).
Therefore, to our knowledge, the biochemical identity of a
putative CGRPR2 has not yet been demonstrated using protein
isolation, protein cross-linking studies, or molecular cloning.
A regulatory role for CGRP in skeletal tissue was originally
suggested by the existence of prominent CGRP-positive peptidergic nerve fibers that permeate the marrow, lining, and
mineralized areas of bone (3, 7, 12, 13, 33). Recent ultramorphological data have demonstrated close contact between these
fibers and metaphyseal osteoblasts and osteoclasts in areas of
active bone repair (14), and additional studies have suggested
that CGRP-positive nerve fibers correlate with local centers of
bone formation (4, 28). A recent study targeted transgenic
expression of CGRP to osteoblasts and demonstrated a 27–
40% increase in bone density at various skeletal sites and
associated increases in cancellous bone formation and osteoblastic parameters (2). Most recently, a study with CGRPspecific knockout mice has demonstrated that generally decreased bone density and widespread osteopenia are due to
decreased bone formation and concluded, “CGRP is a physiologic activator of bone formation” (27). In agreement with the
latest findings, we have demonstrated in a series of in vitro
studies examining dose-dependent activation of intracellular
signaling pathways (6, 17, 18, 19, 22) that (immature) osteoblastic cells are one of the major targets for CGRP in skeletal
tissue.
However, it is still unknown which CGRPR subtype mediates CGRP’s bone-forming activity. In our previous study (22),
we demonstrated for the first time that a type of CGRPR1 is
expressed in immature osteoblastic MG63 cells, but that this
receptor did not display all of the properties expected for
CGRPR1. Therefore, we could not exclude the possibility that
another receptor subtype responsive to [Cys(Acm)2,7]CGRP
also was expressed in these cells. To further examine this
The costs of publication of this article were defrayed in part by the payment
of page charges. The article must therefore be hereby marked “advertisement”
in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
C811
Downloaded from http://ajpcell.physiology.org/ by 10.220.33.1 on June 18, 2017
Kawase, Tomoyuki, Kazuhiro Okuda, and Douglas M. Burns. Immature osteoblastic MG63 cells possess two calcitonin gene-related peptide
receptor subtypes that respond differently to [Cys(Acm)2,7]calcitonin generelated peptide and CGRP8 –37. Am J Physiol Cell Physiol 289: C811–C818,
2005. First published June 15, 2005; doi:10.1152/ajpcell.00504.2004.—
Calcitonin gene-related peptide (CGRP) is clearly an anabolic factor
in skeletal tissue, but the distribution of CGRP receptor (CGRPR)
subtypes in osteoblastic cells is poorly understood. We previously
demonstrated that the CGRPR expressed in osteoblastic MG63 cells
does not match exactly the known characteristics of the classic
subtype 1 receptor (CGRPR1). The aim of the present study was to
further characterize the MG63 CGRPR using a selective agonist of the
putative CGRPR2, [Cys(Acm)2,7]CGRP, and a relatively specific
antagonist of CGRPR1, CGRP8 –37. [Cys(Acm)2,7]CGRP acted as a
significant agonist only upon ERK dephosphorylation, whereas this
analog effectively antagonized CGRP-induced cAMP production and
phosphorylation of cAMP response element-binding protein (CREB)
and p38 MAPK. Although it had no agonistic action when used alone,
CGRP8 –37 potently blocked CGRP actions on cAMP, CREB, and p38
MAPK but had less of an effect on ERK. Schild plot analysis of the
latter data revealed that the apparent pA2 value for ERK is clearly
distinguishable from those of the other three plots as judged using the
95% confidence intervals. Additional assays using 3-isobutyl-1-methylxanthine or the PKA inhibitor N-(2-[p-bromocinnamylamino]ethyl)5-isoquinolinesulfonamide hydrochloride (H-89) indicated that the
cAMP-dependent pathway was predominantly responsible for CREB
phosphorylation, partially involved in ERK dephosphorylation, and
not involved in p38 MAPK phosphorylation. Considering previous
data from Scatchard analysis of [125I]CGRP binding in connection
with these results, these findings suggest that MG63 cells possess two
functionally distinct CGRPR subtypes that show almost identical
affinity for CGRP but different sensitivity to CGRP analogs: one is
best characterized as a variation of CGRPR1, and the second may be
a novel variant of CGRPR2.
C812
CGRP RECEPTOR SUBTYPES IN OSTEOBLASTIC MG63 CELLS
possibility, we have pharmacologically characterized receptormediated cellular responses in MG63 cells in more detail. Our
results provide pharmacological evidence that at least two
distinct subtypes of CGRPR with almost identical affinity to
CGRP are expressed in MG63 cells.
MATERIALS AND METHODS
RESULTS
Effects of [Cys(Acm)2,7]CGRP and CGRP8 –37 on cAMP
production. We discovered in our previous study (22) that
MG63 cells produce much greater increases in cAMP production in response to human CGRP (Fig. 1A) than are observed
for many other osteoblastic cell lines (e.g., UMR106, SaOS-2).
CGRP (10 pM–100 nM) dose dependently stimulated cAMP
formation at 15 min in the presence of 1 mM IBMX, with a
maximal 30- to 35-fold stimulation observed at 10 nM and with
an EC50 of ⬃70 pM (Fig. 1A). The putative CGRPR2 agonist
[Cys(Acm)2,7]CGRP used alone weakly stimulated cAMP production sixfold at 1 ␮M, while the CGRPR1 antagonist
CGRP8 –37 (up to 1 ␮M) had no effect when used alone.
Coaddition of [Cys(Acm)2,7]CGRP (1 ␮M) effectively inhibited CGRP-induced cAMP production (Fig. 1A). As shown in
Fig. 1B, comparative plots of CGRP8 –37’s inhibition of CGRP
at three selected concentrations (20 nM, 200 nM, and 2 ␮M)
vs. CGRP alone demonstrated the expected shift to the right of
the original dose-response curve. When this inhibitory effect
was analyzed by performing more detailed Schild plot analysis,
the pA2 (i.e., negative log of the antagonist concentration
required to produce a twofold increase in the EC50 of the
agonist) and the slope values for these linear regressions were
calculated as 8.736 and 1.687, respectively.
Effects of [Cys(Acm)2,7]CGRP and CGRP8 –37 on CREB
phosphorylation. CGRP (0.1–100 nM) dose dependently phosphorylated cAMP response element-binding protein (CREB)
(Fig. 2, A–D), with an EC50 of ⬃260 pM (Fig. 2E), a value that
is not as consistent as might have been expected with conclusions that CREB phosphorylation is regulated primarily by
cAMP (25). [Cys(Acm)2,7]CGRP alone did not significantly
stimulate CREB phosphorylation (Fig. 2, A and E). Coaddition
of [Cys(Acm)2,7]CGRP (1 ␮M) substantially attenuated
CGRP-induced CREB phosphorylation (Fig. 2B), but this antagonism was less complete than that observed for cAMP
production (Fig. 2E vs. Fig. 1A). CGRP8 –37 alone (up to 1 ␮M)
Fig. 1. A: dose-dependent effects of calcitonin gene-related peptide
(CGRP) and peptide analogs on cAMP production in MG63 cells.
Cells were treated for 15 min with the indicated concentrations of
CGRP alone or in combination with 1 ␮M [Cys(Acm)2,7]CGRP,
[Cys(Acm)2,7]CGRP alone, or CGRP8 –37 alone in the presence of
3-isobutyl-1-methylxanthine (IBMX). cAMP production was evaluated as described in MATERIALS AND METHODS. B: dose-dependent
inhibition of CGRP-induced cAMP production by CGRP8 –37. Cells
were treated for 15 min with CGRP alone in the presence of IBMX
(E) or with coaddition of CGRP8 –37 at 20 nM (‚), 200 nM (F), or 2
␮M (■). Each symbol and vertical bar represents the mean ⫾ SD of
data from four or more independent experiments. Inset: Schild plot
analysis of data obtained from detailed functional inhibition studies
using CGRP8 –37. Resulting data were transformed into linear Schild
plots using GraphPad Prism software.
AJP-Cell Physiol • VOL
289 • OCTOBER 2005 •
www.ajpcell.org
Downloaded from http://ajpcell.physiology.org/ by 10.220.33.1 on June 18, 2017
Cell and cell culture. Human osteosarcoma-derived immature osteoblastic MG63 cells were obtained from Dainippon Pharmaceutical
(Osaka, Japan). MG63 cells were characterized in part by their
pronounced cAMP response to CGRP and by providing a low cAMP
response to parathyroid (PTH) hormone-related protein (22). Cells
were maintained in Dulbecco’s modified Eagle’s medium (DMEM;
Life Technologies, Grand Island, NY) supplemented with 10% fetal
bovine serum (FBS; Hyclone, Logan, UT) in humidified 5% CO295% air at 37°C.
Measurement of intracellular cAMP. Cells were seeded onto 24multiwell plates at a density of 4 ⫻ 104 cells/well and cultured for 2
days. Cells were preincubated overnight in 1% FBS-containing medium and then treated with CGRP (Peptide Institute, Osaka, Japan)
or [Cys(Acm)2,7]CGRP (Bachem, Torrance, CA) for 15 min
in the presence or absence of CGRP8 –37 (Peptide Institute) or
[Cys(Acm)2,7]CGRP in Hanks’ balanced salt solution (HBSS) containing 1 mM 3-isobutyl-1-methylxanthine (IBMX; Sigma, St. Louis,
MO) at 37°C. Intracellular cAMP was extracted and assessed using
the cAMP enzyme immunoassay (EIA) system (Amersham Pharmacia Biotech, Little Chalfont, UK) as described previously (21).
Western blot analysis for MAPK phosphorylation. As described
previously (20, 21), cells were treated with CGRP or [Cys(Acm)2,7]CGRP in
combination with CGRP8 –37, [Cys(Acm)2,7]CGRP, or IBMX. In some
cases, cells pretreated with N-(2-[p-bromocinnamylamino]ethyl)-5isoquinolinesulfonamide hydrochloride (H-89, 10 ␮M; Calbiochem)
were used for CGRP stimulation. Cells were then lysed in Laemmli
sample buffer and subjected to SDS-PAGE (9% linear gel) and
immunoblot analysis using the following primary antibodies: polyclonal anti-phosphorylated-ERK antibody (1:1,000 dilution; New England BioLabs, Beverly, MA), polyclonal anti-ERK antibody (1:1,000
dilution; Transduction Laboratories or New England BioLabs), polyclonal anti-phosphorylated p38 MAPK antibody (1:1,000 dilution;
New England BioLabs), or polyclonal anti-p38 MAPK antibody
(1:3,000 dilution; Santa Cruz Biotechnology). Probed blots were then
reacted with horseradish peroxidase (HRP)-conjugated protein A
(1:10,000 dilution; Zymed, South San Francisco, CA), followed by
visualization using the Amersham ECL system (Amersham Pharmacia Biotech). For reprobing with a different primary antibody, blots
were stripped with Restore Western blot stripping buffer (Pierce,
Rockford, IL).
The developed films were scanned using an Epson scanner (GTX700; Epson, Tokyo, Japan) and quantitated using Scion Image (version 4.02; Frederick, MD) or Totallab (Nonlinear Dynamics, Newcastle upon Tyne, UK). The intensity of each phosphorylated protein
band was normalized to the intensity of each total protein band
observed in the lane.
Statistical analysis. Unless otherwise stated, results are expressed
as means ⫾ SD of at least three independent experiments. Using
GraphPad Prism software (version 4 for Windows; GraphPad, San
Diego, CA), we performed Schild plot analysis and calculated the
values for half-maximal effective concentration (EC50) using nonlinear regression analysis. The 95% confidence interval was used for
evaluation of statistically significant differences. In Fig. 5, for example, statistical analyses were performed using Student’s t-test. P ⬍
0.05 values were considered significant.
CGRP RECEPTOR SUBTYPES IN OSTEOBLASTIC MG63 CELLS
C813
did not show significant effects (Fig. 2, C and E), and as
observed for cAMP formation, increasing concentrations of
this antagonist effectively shifted the CGRP dose-response
curve to the right (Fig. 2F). As shown, more detailed Schild
plot analyses generated a pA2 value of 8.841 and a slope value
of 1.258.
Effects of [Cys(Acm)2,7]CGRP and CGRP8 –37 on p38
MAPK phosphorylation. Findings very similar to those obtained for the phosphorylation of CREB were obtained for p38
MAPK phosphorylation. CGRP (0.1–100 nM) dose dependently phosphorylated p38 MAPK (Fig. 3, A–D), with an EC50
of ⬃180 pM (Fig. 3, E or F). [Cys(Acm)2,7]CGRP alone failed
to have an effect; however, coaddition of [Cys(Acm)2,7]CGRP
potently but incompletely antagonized this action of CGRP
(Fig. 3, A, B, and E). CGRP8 –37 did not itself act as an agonist
(up to 1 ␮M) and again effectively shifted the CGRP doseresponse curve to the right at selected concentrations of 20 nM,
200 nM, and 2 ␮M (Fig. 3F). More detailed Schild plot
analyses yielded a pA2 of 8.803 and a slope value of 1.212.
Effects of [Cys(Acm)2,7]CGRP and CGRP8 –37 on ERK dephosphorylation. ERK dephosphorylation demonstrates a distinctly different pattern of peptide-receptor interaction. Both
CGRP and [Cys(Acm)2,7]CGRP induced ERK dephosphorylation with similar maximal effects, although CGRP was approximately fourfold more potent than [Cys(Acm)2,7]CGRP (Fig. 4,
A and E), with EC50 values of 210 and 820 pM, respectively.
Fig. 3. A–D: dose-dependent effects of CGRP and peptide analogs, alone or in combination, on the phosphorylation of p38 MAPK in MG63 cells. Cells were treated
for 15 min with CGRP alone (A–D), [Cys(Acm)2,7]CGRP
alone (A), or CGRP8 –37 alone (C). Otherwise, cells were
treated with CGRP in the presence of 1 ␮M
[Cys(Acm)2,7]CGRP (B) or 2 ␮M CGRP8 –37 (D). Experiments were repeated more than three times, and the
results shown are representative of all of these data. E:
dose-dependent effects of CGRP and CGRP analogs on
CREB phosphorylation. F: dose-dependent inhibition by
CGRP8 –37 of CGRP-induced p38 MAPK phosphorylation. Cells were treated for 15 min with CGRP in the
absence (E) or presence of CGRP8 –37 at 20 nM (‚), 200
nM (F), or 2 ␮M (■). Each symbol and vertical bar
represents the mean ⫾ SD of data from more than three
independent experiments. Inset: Schild plot analysis of
data obtained from detailed functional inhibition studies
using CGRP8 –37.
AJP-Cell Physiol • VOL
289 • OCTOBER 2005 •
www.ajpcell.org
Downloaded from http://ajpcell.physiology.org/ by 10.220.33.1 on June 18, 2017
Fig. 2. A–D: dose-dependent effects of CGRP and peptide analogs, alone or in combination, on the phosphorylation of cAMP response element-binding protein
(CREB) in MG63 cells. Cells were treated for 15 min with
CGRP alone (A–D), [Cys(Acm)2,7]CGRP alone (A), or
CGRP8 –37 alone (C). In addition, cells were treated with
CGRP in the presence of 1 ␮M [Cys(Acm)2,7]CGRP (B) or
2 ␮M CGRP8 –37 (D). Phosphorylated CREB (p-CREB)
and total CREB were fractionated by performing SDSPAGE and then visualized using Western blot analysis.
Each experiment was repeated at least four times, and the
results shown are representative of all of these data. E:
dose-dependent effects of CGRP and CGRP analogs on
CREB phosphorylation. Target bands were scanned,
quantitated, and plotted as described in MATERIALS AND
METHODS. F: dose-dependent inhibition of CGRP-induced CREB phosphorylation by CGRP8 –37. Cells were
treated for 15 min with CGRP in the absence (E) or
presence of CGRP8 –37 at 20 nM (‚), 200 nM (F), or 2
␮M (■). Each symbol and vertical bar represents the
mean ⫾ SD of data from more than four independent
experiments. Inset: Schild plot analysis of data obtained from detailed functional inhibition studies using
CGRP8 –37.
C814
CGRP RECEPTOR SUBTYPES IN OSTEOBLASTIC MG63 CELLS
The potency and magnitude of CGRP-induced ERK dephosphorylation appeared almost identical to its effect on CREB or
p38 MAPK phosphorylation. However, as predicted from earlier studies, in the presence of 1 ␮M [Cys(Acm)2,7]CGRP (the
maximally effective concentration), CGRP could not produce a
significant additional effect until concentrations of 100 nM
were reached (Fig. 4E). Thus, because both peptides stimulate
ERK dephosphorylation, it is difficult to evaluate the power of
[Cys(Acm)2,7]CGRP to block CGRP action. Only at 1 ␮M and
higher concentrations did CGRP8 –37 alone slightly but significantly stimulate ERK dephosphorylation (Fig. 4E). Furthermore, the antagonistic actions of CGRP8 –37 were substantially
different in the case of ERK dephosphorylation than those
observed for the three previously studied parameters: substantial antagonism did not occur until 2 ␮M CGRP8 –37, and 1 ␮M
CGRP still produced ⬎50% of its maximal effect on ERK
dephosphorylation in the presence of 2 ␮M CGRP8 –37. Detailed Schild plot analyses of the partial inhibition produced by
CGRP8 –37 generated pA2 and slope values of 8.105 and 1.184,
respectively.
Comparison of computer-calculated 95% confidence intervals for each data plot demonstrated the pA2 value of ERK
dephosphorylation to be statistically distinguishable from that
obtained for the other three plots. Upon initial inspection, the
slope of the ERK Schild plot appeared substantially different
from that of all of the other plots; however, analysis using
the 95% confidence interval comparison revealed that this
value was significantly different from only the slope of the
cAMP Schild plot.
Effects of IBMX and H-89 on CGRP action. To examine
further whether different receptor subtypes are involved in
each response, pharmacological inhibitors were applied to this
experimental system. As shown in Fig. 5, A and C, coaddition
of 1 mM IBMX, a level sufficient to inhibit phosphodiesterasedependent degradation of cAMP in these cells (Kawase T,
Okuda K, and Burns DM, unpublished observation), to 15-min
AJP-Cell Physiol • VOL
CGRP treatments clearly augmented CGRP-induced CREB
phosphorylation and slightly enhanced CGRP-induced ERK
dephosphorylation. However, IBMX had no appreciable effect
on CGRP-induced p38 MAPK phosphorylation. In support of
this finding, when cells were pre- or coincubated with 10 ␮M
H-89, a relatively specific inhibitor of cAMP-dependent protein kinase, CGRP-induced CREB phosphorylation was substantially blocked after 5 min but p38 MAPK phosphorylation
was unaffected (Fig. 5, B and D). Despite having observed
little effect for added IBMX, coaddition of H-89 significantly
attenuated CGRP-induced ERK dephosphorylation.
DISCUSSION
How many functional CGRPR subtypes are expressed in
MG63 cells? Until now, it has been widely accepted that
CGRPR1 is the predominant CGRPR subtype expressed in
osteoblastic cells. In support of this concept, our previous study
(22) demonstrated in radioligand receptor binding assays, followed by both directly binding kinetic and Scatchard analyses,
that only one type of CGRP-binding site is expressed on MG63
cells, and in immunocytochemical assays that both the main
protein components of a CGRPR1 (i.e., CLR and RAMP-1) are
present in MG63 cells. Therefore, it seemed likely that MG63
cells predominantly express a CGRPR1 (or at least a variation
of this receptor subtype). Nonetheless, we also observed that a
200-fold excess of CGRP8 –37, a specific antagonist of
CGRPR1, competes for only 70 – 80% of membrane-bound
fluorescein-conjugated CGRP (22) and that 1 ␮M CGRP8 –37
antagonizes only lower concentrations (⬃1 nM) of CGRP in
cellular dephosphorylation of ERK (22). We speculated that
incomplete competition could be attributed to the large difference in binding affinities between these individual peptides,
but we could not completely rule out the possibility that a
second receptor with a very similar CGRP-binding affinity
might be expressed in MG63 cells.
289 • OCTOBER 2005 •
www.ajpcell.org
Downloaded from http://ajpcell.physiology.org/ by 10.220.33.1 on June 18, 2017
Fig. 4. A–D: dose-dependent effects of CGRP and
peptide analogs, alone or in combination, on ERK phosphorylation in MG63 cells. Cells were treated for 15 min
with CGRP alone (A–D), [Cys(Acm)2,7]CGRP alone
(A), or CGRP8–37 alone (C). Otherwise, cells were
treated with CGRP in the presence of 1 ␮M
[Cys(Acm)2,7]CGRP (B) or 1 ␮M CGRP8 –37 (D).
Experiments were repeated more than three times,
and the results shown are representative of all of these
data. E: dose-dependent effect of CGRP and CGRP
analogs on ERK phosphorylation. F: dose-dependent
inhibition of CGRP-induced ERK dephosphorylation
by CGRP8 –37. Cells were treated for 15 min with
CGRP alone (E) or with coadditions of CGRP8 –37 at
20 nM (‚), 200 nM (F), or 2 ␮M (■). Each symbol
and vertical bar represents the mean ⫾ SD of data
from at least four independent experiments. Inset:
Schild plot analysis of data obtained from detailed
functional inhibition studies using CGRP8 –37.
CGRP RECEPTOR SUBTYPES IN OSTEOBLASTIC MG63 CELLS
C815
The more complete functional assays performed in our
present study have provided evidence sufficient to modify the
understanding of the interaction of CGRP8 –37 and CGRP.
CGRP8 –37 potently antagonized CGRP-induced phosphorylation of CREB and p38 MAPK as observed in cAMP production, but could not produce complete inhibition of ERK dephosphorylation induced by higher concentrations (⬃1 ␮M) of
CGRP. Schild plot analyses clearly showed two distinct sets of
interactions between CGRP8 –37 and CGRPR subtypes: among
four plots describing inhibition of CGRP-induced effects by
CGRP8 –37, the pA2 value of the Schild plot regression for the
partial inhibition of ERK dephosphorylation appeared significantly different. The Schild plot for cAMP formation is statistically different from that for ERK dephosphorylation in both
pA2 and slope values.
Complementary findings were obtained in assays using
[Cys(Acm)2,7]CGRP. As described in the introduction,
[Cys(Acm)2,7]CGRP has been widely accepted as an agonist of
AJP-Cell Physiol • VOL
putative CGRPR2 (8, 34) and its antagonism of what appeared
to be CGRPR1-mediated effects was not expected (15). In the
present study, [Cys(Acm)2,7]CGRP did not act exactly as data
in the literature would have predicted. [Cys(Acm)2,7]CGRP
produced substantial agonistic effects only on dephosphorylation of ERK and instead potently antagonized CGRP-induced
cAMP production and the phosphorylation of CREB and p38
MAPK. Taken together with the data obtained from CGRP8 –37,
therefore, we must revise our previous conclusions and now
propose a novel dual receptor explanation that immature osteoblastic MG63 cells possess two similar but distinguishable
subtypes of CGRPR. Presumably, CGRP’s effects on cAMP,
CREB, and p38 MAPK are associated primarily with CGRPR1
(or a CGRPR1 variant), while CGRP-induced ERK dephosphorylation is associated with a CGRPRx (see Fig. 6).
Skepticism about the traditional classification of CGRPR
subtypes. It should be noted that our treatment of these topics
is based primarily on classically published data suggesting the
289 • OCTOBER 2005 •
www.ajpcell.org
Downloaded from http://ajpcell.physiology.org/ by 10.220.33.1 on June 18, 2017
Fig. 5. Ability of IBMX or N-(2-[p-bromocinnamylamino]ethyl)-5-isoquinolinesulfonamide hydrochloride (H-89) to inhibit or enhance CGRPinduced changes in p38 MAPK, CREB, or ERK
phosphorylation. A: MG63 cells were treated with
CGRP (0, 1, or 10 nM) for 15 min in the presence of
IBMX (1 mM). B: cells were preincubated with H-89
(10 ␮M) for 30 min and then treated with CGRP (0,
1, or 10 nM) for 5 min. Experiments were repeated at
least four times, and the results shown are representative of all of these data. C and D: each target
protein was quantitated and plotted. Each column and
vertical bar represents the mean ⫾ SD of the data
from more than three independent experiments. Significant differences were observed in CREB phosphorylation in the cases of both inhibitors and in
ERK phosphorylation after preincubation with H-89
(aP ⬍ 0.05, bP ⬍ 0.02, cP ⬍ 0.002, dP ⬍ 0.001 vs.
cultures treated with CGRP at same concentrations
alone).
C816
CGRP RECEPTOR SUBTYPES IN OSTEOBLASTIC MG63 CELLS
widely accepted general concept that two or more versions of
a CGRP receptor do exist. However, a series of recent and
extensive studies by Abel and coworkers (26, 32) raised a
cautionary note arguing against those criteria established for
the classification of CGRPR subtypes (8, 24, 34). In CGRPR1transfected human embryonic kidney (HEK)-293 cells (26),
[Cys(Acm)2,7]CGRP binds to this receptor and displaces
[125I]CGRP with a Ki of 28 –32 nM as did another putative
CGRPR2 agonist, [Pro14]CGRP (Ki of 5.5–9 nM), and this
agonist is capable of stimulating cAMP production. For these
puzzling effects, therefore, Abel and colleagues offered possible explanations that the effectiveness of various CGRPR2
agonists in different tissues (8, 24, 34) is actually a function of
the quantity of excess CGRPR1 being held in cellular reserve
(26, 32), and an explanation for tissue-to-tissue differences in
the effects of agonists and antagonists is that this behavior
reflects the local complement of proteases present to degrade
the unprotected peptides (26). In addition, the affinities of
agonists and antagonists are possibly affected by whether the
peripheral membrane receptor component protein (10) is
present in the receptor complex (26).
Our data could prove to be consistent with these interpretations, because our previous (22) and present studies have
demonstrated pharmacologically distinguishable CGRPR subtypes in MG63 cells that show a very similar affinity to
[125I]CGRP. Nonetheless, we have not yet obtained direct
evidence to strongly support or demonstrate this explanation.
Therefore, a considerable body of additional experiments is
needed to provide evidence that distinguishes one of several
candidates thought to be expressed in MG63 cells that may
AJP-Cell Physiol • VOL
serve as a functional CGRPR2 (or “CGRPRx” as shown in
Fig. 6).
How are receptor subtypes associated by signaling pathways with cellular responses in MG63 cells? It seems possible
to classify these cellular responses by the degree to which the
cAMP-dependent pathways are involved. IBMX augmented
CGRP-induced CREB phosphorylation but not p38 MAPK
phosphorylation. Furthermore, H-89 blocked CGRP-induced
CREB phosphorylation and ERK dephosphorylation but not
p38 MAPK phosphorylation. Therefore, these findings suggest
that both CREB and p38 MAPK phosphorylation are apparently mediated by the same receptor subtype (i.e., a CGRPR1
variant that stimulates cAMP production), but only CREB
phosphorylation is located downstream of the cAMP-mediated
pathway. The other receptor subtype (shown as CGRPRx in
Fig. 6) on which [Cys(Acm)2,7]CGRP acts as an agonist
weakly activates adenylyl cyclase to produce cAMP and
strongly stimulates ERK dephosphorylation through a mechanism that is primarily cAMP independent.
Furthermore, we recently reported that CGRP induces biphasic increases in intracellular free Ca2⫹ concentration
([Ca2⫹]i): an initial transient increase and a subsequent slow
incremental increase in MG63 cells (6). The slow increment in
[Ca2⫹]i is completely blocked by a cAMP antagonist, Rp
diastereomer of adenosine 3⬘,5⬘-cyclic monophosphothioate
Rp-cAMPS, or verapamil, while the transient initial peak is
sensitive only to thapsigargin. These data primarily suggest
that the CGRPR is associated with two independent components, such as a cAMP-operated Ca2⫹ channel and a component responsible for Ca2⫹ discharge from intracellular stores
289 • OCTOBER 2005 •
www.ajpcell.org
Downloaded from http://ajpcell.physiology.org/ by 10.220.33.1 on June 18, 2017
Fig. 6. Proposed mechanism of interaction between CGRP and its specific receptor subtypes to produce the four major cellular responses addressed in this paper.
Our best explanation for the data in this report is that CGRP acts on two distinct CGRP receptor (CGRPR) subtypes expressed by MG63 cells, but these receptor
subtypes share some intracellular signaling pathways. 1) cAMP is produced mainly through activation of a CGRPR1 variant, but the agonistic action of
[Cys(Acm)2,7]CGRP indicates the additional involvement of a second CGRPR subtype (CGRPRx) in this response. 2) Because CGRP-induced p38 MAPK
phosphorylation was potently blocked by CGRP8 –37, as was cAMP production, this response is produced predominantly through activation of CGRPR1.
However, because cAMP-related inhibitors (i.e., IBMX, H-89) had no appreciable effect, this response is not mediated by cAMP. 3) Because CREB
phosphorylation, which is known to depend predominantly on cAMP (25), was potently inhibited by CGRP8 –37 and weakly stimulated by [Cys(Acm)2,7]CGRP,
this response was mediated by a cAMP-dependent mechanism mainly through activation of CGRPR1 and secondarily by activation of CGRPRx. 4) Because
CGRP-induced ERK dephosphorylation could not be inhibited effectively by high levels of CGRP8 –37, which potently inhibited CGRP-induced cAMP
production and p38 MAPK or CREB phosphorylation in parallel experiments, this response is mediated primarily through activation of CGRPRx. However, this
response was also slightly and significantly influenced by IBMX and H-89, respectively, and thus we cannot exclude the possible additional involvement of
CGRPR1. (Note that the thicker of any two arrows represents the major pathway.)
CGRP RECEPTOR SUBTYPES IN OSTEOBLASTIC MG63 CELLS
GRANTS
This study was supported in part by a Grant-in-Aid for Scientific Research
from the Ministry of Education, Science, Sports, and Culture of Japan, by a
AJP-Cell Physiol • VOL
Merit Review Grant and Research Career Scientist Award from the U.S.
Department of Veterans Affairs, and by a Lied Bridge Funds Grant from the
University of Kansas Medical Center Research Institute.
REFERENCES
1. Bai X, Lu D, Bai J, Zheng H, Ke Z, Li X, and Luo S. Oxidative stress
inhibits osteoblastic differentiation of bone cells by ERK and NF-␬␤.
Biochem Biophys Res Commun 314: 197–207, 2004.
2. Ballica R, Valentijn K, Khachatryan A, Guerder S, Kapadia S,
Gundberg C, Gilligan J, Flavell RA, and Vignery A. Targeted expression of calcitonin gene-related peptide to osteoblasts increases bone
density in mice. J Bone Miner Res 14: 1067–1074, 1999.
3. Bjurhölm A, Kreicbergs A, Brodin E, and Schultzberg M. Substance
P- and CGRP-immunoreactive nerves in bone. Peptides 9: 165–171, 1988.
4. Bjurhölm A, Kreicbergs A, Dahlberg I, and Schultzberg M. The
occurrence of neuropeptides at different stages of DBM-induced heterotropic bone formation. Bone Miner 10: 95–107, 1990.
5. Burns DM and Kawase T. Calcitonin gene-related peptide, amylin, or
parathyroid hormone stimulates in vitro biomineralization (Abstract).
J Bone Miner Res 12, Suppl 1: F392, 1997.
6. Burns DM, Stehno-Bittel L, and Kawase T. Calcitonin gene-related
peptide elevates calcium and polarizes membrane potential in MG-63 cells
by both cAMP-independent and -dependent mechanisms. Am J Physiol
Cell Physiol 287: C457–C467, 2004.
7. Deftos LJ and Roos BA. Medullary thyroid carcinoma and calcitonin
gene expression. In: Bone and Mineral Research, edited by Peck WA.
New York: Elsevier, 1989, vol. 6, chapt. 8, p. 276 –316.
8. Dennis T, Fournier A, St. Pierre S, and Quirion R. Structure-activity
profile of calcitonin gene-related peptide in peripheral and brain tissues:
evidence for receptor multiplicity. J Pharmacol Exp Ther 251: 718 –725,
1989.
9. Engelbrecht Y, de Wet H, Horsch K, Langeveldt CR, Hough FS, and
Hulley PA. Glucorticoids induce rapid up-regulation of mitogen-activated
protein kinase phosphatase-1 and dephosphorylation of extracellular signal-regulated kinase and impair proliferation in human and mouse osteoblast cell lines. Endocrinology 144: 412– 422, 2003.
10. Evans BN, Rosenblatt MI, Mnayer LO, Oliver KR, and Dickerson IM.
CGRP-RCP, a novel protein required for signal transduction at calcitonin
gene-related peptide and adrenomedullin receptors. J Biol Chem 275:
31438 –31443, 2000.
11. Guicheux J, Lemonnier J, Ghayor C, Suzuki A, Palmer G, and
Caverzasio J. Activation of p38 mitogen-activated protein kinase and
c-Jun-NH2-terminal kinase by BMP-2 and their implication in the stimulation of osteoblastic cell differentiation. J Bone Miner Res 18: 2060 –
2068, 2003.
12. Hares KA and Foster GA. Immunohistochemical analysis of the ontogeny of calcitonin gene-related peptide-like immunoreactivity in the rat
central nervous system. J Chem Neuroanat 4: 187–203, 1991.
13. Hill EL and Elde R. Distribution of CGRP-, VIP-, D-␤-hydroxylase-, SP-,
and NPY-immunoreactive nerves in the periosteum of the rat. Cell Tissue
Res 264: 469 – 480, 1991.
14. Imai S, Rauvala H, Konttinen YT, Tokunaga T, Maeda T, Hukuda S,
and Santavirta S. Efferent targets of osseous CGRP-immunoreactive
nerve fiber before and after bone destruction in adjuvant arthritic rat: an
ultramorphological study on their terminal-target relations. J Bone Miner
Res 12: 1018 –1027, 1997.
15. Jansen-Olesen I, Kaarill L, and Edvinsson L. Characterization of
CGRP1 receptors in the guinea pig basilar artery. Eur J Pharmacol 414:
249 –258, 2001.
16. Kakita A, Suzuki A, Ono Y, Miura Y, Itoh M, and Oiso Y. Possible
involvement of p38 MAP kinase in prostaglandin E1-induced ALP activity
in osteoblast-like cells. Prostaglandins Leukot Essent Fatty Acids 70:
469 – 474, 2004.
17. Kawase T, Howard GA, Roos BA, and Burns DM. Diverse actions of
calcitonin gene-related peptide on intracellular free Ca2⫹ concentrations in
UMR 106 osteoblastic cells. Bone 16, Suppl 1: 379S–384S, 1995.
18. Kawase T, Howard GA, Roos BA, and Burns DM. Calcitonin generelated peptide rapidly inhibits calcium uptake in osteoblastic cell lines via
activation of ATP-sensitive potassium channels. Endocrinology 137: 984 –
990, 1996.
19. Kawase T and Burns DM. Calcitonin gene-related peptide stimulates
potassium efflux through ATP-sensitive potassium channels and produces
membrane hyperpolarization in osteoblastic UMR106 cells. Endocrinology 138: 3492–3502, 1998.
289 • OCTOBER 2005 •
www.ajpcell.org
Downloaded from http://ajpcell.physiology.org/ by 10.220.33.1 on June 18, 2017
such as inositol 1,4,5-trisphosphate. In consideration of the
finding that the slow increment is also sensitive to CGRP8 –37,
these data also support the concept that in addition to the
known CGRPR1, MG63 cells possess a distinct second receptor subtype (i.e., CGRPRx as shown in Fig. 6).
Reevaluation of the biological significance of CGRP on
osteoblastic bone formation. With regard to CGRP’s action on
osteoblastic cells, we have demonstrated in several publications (5, 22) that CGRP stimulates biomineralization and
expression of type I collagen mRNA in cultures of osteogenic
cells. As described in the introduction of this report, recent
molecular biological approaches using transgenic animals have
demonstrated in vivo anabolic actions of CGRP (2, 27). Therefore, there can no longer be real doubt that CGRP plays an
important role in bone formation.
The biological significance of the intracellular signaling
pathways invoked by CGRP in osteogenic cells is suggested by
several recent studies. Bone morphogenetic protein-2 activates
p38 MAPK to induced osteoblastic differentiation (11). Prostaglandin E1 enhances alkaline phosphatase activity through
activation of p38 MAPK and JNK by a cAMP-dependent
mechanism in osteoblastic MC3T3-E1 cells while ERK is
dephosphorylated (16). Oxidative stress inhibits osteoblastic
differentiation of calvarial osteoblasts through activation of
ERK (1), and glucocorticoids, useful cofactors to stimulate the
osteogenic differentiation of marrow mesenchymal stem cells,
rapidly stimulate ERK dephosphorylation and consequently
impair the proliferation of osteoblastic cells (9). In addition,
cAMP production, which has classically been used as a marker
response of osteoblastic cells to PTH, recently has been shown
to be required for c-fos expression, which is thought to have
many pleiotropic and essential effects in bone, through phosphorylation of CREB (30). Therefore, we think it very likely
that this complex of the individual signaling pathways may
prove central to understanding CGRP action in osteogenic
bone cells and bone formation.
To our knowledge, it is still widely accepted, without much
in the way of direct convincing evidence, that CGRPR1 is the
predominant functional receptor for CGRP in osteoblastic cells
throughout their differentiated pathways. Several research
groups have suggested the possibility that osteoblastic
CGRPR1 gradually disappears with the maturity of cell differentiation (29, 31). On the basis of our indirect data from years
of functional assays in a variety of cell lines, we also advanced
a similar possibility (17). Although the situation may ultimately prove more complex, we would now simply suggest
that the second CGRP subtype is important to CGRP’s bone
anabolic effects. We hope that our present data provide some
guidance for reevaluating the biological significance of CGRP
in osteoblastic bone formation.
Our present pharmacological data support the possibility that
a second subtype of CGRPR is expressed in preosteoblastic
MG63 cells. However, previously published results from our
[125I]CGRP binding assay indicate that this additional receptor
subtype is not identical to the classic CGRPR2 described in
other tissues and cell types and may instead represent a variant
of CGRPR2.
C817
C818
CGRP RECEPTOR SUBTYPES IN OSTEOBLASTIC MG63 CELLS
AJP-Cell Physiol • VOL
27. Schinke T, Liese S, Priemel M, Haberland M, Schilling AF, CatalaLehnen P, Blicharski D, Rueger JM, Gagel RF, Emeson RB, and
Amling M. Decreased bone formation and osteopenia in mice lacking
␣-calcitonin gene-related peptide. J Bone Miner Res 19: 2049 –2056,
2004.
28. Suva LJ, Seedor G, Endo N, Quartuccio HA, Thompson DD, Bab I,
and Rodan GA. Pattern of gene expression following rat tibial marrow
ablation. J Bone Miner Res 7: 379 –388, 1993.
29. Togari A, Arai M, Mizutani S, Koshihara Y, and Nagatsu T. Expression of mRNAs for neuropeptide receptors and ␤-adrenergic receptors in
human osteoblasts and human osteogenic sarcoma cells. Neurosci Lett
233: 125–128, 1997.
30. Tyson DR, Swarthout JT, and Partridge NC. Increased osteoblastic
c-fos expression by parathyroid hormone requires protein kinase A phosphorylation of the cyclic adenosine 3⬘,5⬘-monophosphate response element-binding protein at serine 133. Endocrinology 140: 1255–1261, 1999.
31. Uzan B, de Vernejoul MC, and Cressent M. RAMPs and CRLR
expressions in osteoblastic cells after dexamethasone treatment. Biochem
Biophys Res Commun 321: 802– 808, 2004.
32. Waugh DJ, Bockman CS, Smith DD, and Abel PW. Limitations in
using peptide drugs to characterize calcitonin gene-related peptide receptors. J Pharmacol Exp Ther 289: 1419 –1426, 1999.
33. Wimalawansa SJ. Calcitonin gene-related peptide and its receptors:
molecular genetics, physiology, pathophysiology, and therapeutic potentials. Endocr Rev 17: 533–585, 1996.
34. Wisskirchen FM, Burt RP, and Marshall I. Pharmacological characterization of CGRP receptors mediating relaxation of the rat pulmonary
artery and inhibition of twitch responses of the rat vas deferens. Br J
Pharmacol 123: 1673–1683, 1998.
289 • OCTOBER 2005 •
www.ajpcell.org
Downloaded from http://ajpcell.physiology.org/ by 10.220.33.1 on June 18, 2017
20. Kawase T, Okuda K, Wu HC, Yoshie H, Hara K, and Burns DM.
Calcitonin gene-related peptide acts as a mitogen for human Gin-1
gingival fibroblasts by activating the MAP kinase signalling pathway. J
Periodontal Res 34: 160 –168, 1999.
21. Kawase T, Okuda K, Momose M, Kato Y, Yoshie H, and Burns DM.
Enamel matrix derivative (EMDOGAIN) rapidly stimulates phosphorylation of the MAP kinase family and nuclear accumulation of smad2 in both
oral epithelial and fibroblastic human cells. J Periodontal Res 36: 367–
376, 2001.
22. Kawase T, Okuda K, and Burns DM. Immature human osteoblastic
MG63 cells predominantly express a subtype 1-like CGRP receptor that
inactivates extracellular signal response kinase by a cAMP-dependent
mechanism. Eur J Pharmacol 470: 125–137, 2003.
23. McLatchie LM, Fraser NJ, Main MJ, Wise A, Brown J, Thompson N,
Solari R, Lee MG, and Foord SM. RAMPs regulate the transport and
ligand specificity of the calcitonin-receptor-like receptor. Nature 393:
333–339, 1998.
24. Mimeault M, Fournier A, Dumont Y, St-Pierre S, and Quirion R.
Comparative affinities and antagonistic potencies of various human calcitonin gene-related peptide fragments on calcitonin gene-related peptide
receptors in brain and periphery. J Pharmacol Exp Ther 258: 1084 –1090,
1991.
25. Montminy MR, Gonzalez GA, and Yamamoto KK. Characteristics of
the cAMP response unit. Metabolism 39, Suppl 2: 6 –12, 1990.
26. Rorabaugh BR, Scofield MA, Smith DD, Jeffries WB, and Abel PW.
Functional calcitonin gene-related peptide subtype 2 receptors in porcine
coronary arteries are identified as calcitonin gene-related peptide subtype
1 receptors by radioligand binding and reverse transcription-polymerase
chain reaction. J Pharmacol Exp Ther 299: 1086 –1094, 2001.