|
|
|
|
Vol. 59, Issue 3, 612-618, March 2001
Institut für Experimentelle und Klinische Pharmakologie und Toxikologie, Universität Erlangen-Nürnberg, Erlangen, Germany (S.A., C.K., H.G., A.P., H.U.Z.); and Division of Cell Biology, Institute of Life Science, Kurume University, Fukuoka, Japan (H.T.)
| |
Abstract |
|---|
|
|
|---|
Nociceptin/orphanin FQ (N/OFQ) and nocistatin (NST) are two
neuropeptides derived from the same precursor protein that exhibit opposing effects on spinal neurotransmission and nociception. Here, we
have used whole-cell, patch-clamp recordings from visually identified
neurons in spinal cord dorsal horn slices of genetically modified mice
to investigate the role of the N/OFQ receptor (N/OFQ-R) in the
modulatory action of both peptides on excitatory glutamatergic and
inhibitory glycinergic and
-aminobutyric acid (GABA)-ergic synaptic transmission. In wild-type mice, N/OFQ selectively suppressed excitatory transmission in a concentration-dependent manner but left
inhibitory synaptic transmission unaffected. In contrast, NST reduced
only inhibitory but not
-amino-3-hydroxy-5-methylisoxazole-4-propionic acid (AMPA)
receptor-mediated excitatory synaptic transmission. N/OFQ-mediated
inhibition of excitatory transmission was completely absent in N/OFQ-R
receptor-deficient (N/OFQ-R
/
) mice and significantly
reduced in heterozygous (N/OFQ-R+/
) mice, whereas the
action of NST on inhibitory neurotransmission was completely retained.
To test for the relevance of these results for spinal nociception, we
investigated the effects of intrathecally injected N/OFQ in the mouse
formalin test, an animal model of tonic pain. N/OFQ (3 nmol/mouse)
induced significant antinociception in wild-type mice, but had no
antinociceptive effects in N/OFQ-R
/
mice. These results
indicate that the inhibitory action of N/OFQ on excitatory
glutamatergic synaptic transmission and its spinal antinociceptive
action are mediated via the N/OFQ receptor, whereas the action of NST
is independent of this receptor.
| |
Introduction |
|---|
|
|
|---|
Nociceptin/orphanin
FQ (N/OFQ; Meunier et al., 1995
; Reinscheid et al., 1995
) and
nocistatin (NST; Okuda-Ashitaka et al., 1998
) are two neuropeptides,
which have been implicated in several CNS functions including
nociception. N/OFQ is an endogenous agonist at the opioid receptor-like
1 receptor (Meunier et al., 1995
; Reinscheid et al., 1995
), which is
also called ROR-C, LC132, or, most recently, NOP, OP-4,
or N/OFQ receptor (Bunzow et al., 1994
; Fukuda et al., 1994
;
Mollereau et al., 1994
, Nishi et al., 1994
; Wang et al., 1994
). Both
pro- and antinociceptive effects of exogenously applied N/OFQ have
repeatedly been reported, mainly depending on the dose and site of
application (e.g., Inoue et al., 1999
; for a recent review, see Calo'G
et al., 2000
). Whereas pronociceptive (Meunier et al., 1995
; Reinscheid
et al., 1995
) and/or antiopioidergic effects (Mogil et al., 1996
)
dominate after intracerebroventricular injection, antinociception is
reliably observed after spinal application of nanomolar doses in a
variety of mouse and rat pain models (e.g., Xu et al., 1996
; Erb et
al., 1997
; Inoue et al., 1999
).
NST is derived from the same precursor protein as N/OFQ, preproN/OFQ
(ppN/OFQ; Saito et al., 1995
; Houtani et al., 1996
; Nothacker et al.,
1996
; Pan et al., 1996
), and has been demonstrated to antagonize
several effects of N/OFQ at the cellular and behavioral levels (Minami
et al., 1998
; Nicol et al., 1998
; Okuda-Ashitaka et al., 1998
; Zhao et
al., 1999
). Several lines of evidence suggest that NST is a
biologically active peptide per se that is involved in nociception at
the spinal level (Xu et al., 1999
; Nakano et al., 2000
;
Zeilhofer et al., 2000
).
In the spinal cord dorsal horn, which constitutes the first important
site of synaptic integration in the pain pathway (Yaksh and Malmberg,
1994
), both the precursor protein of N/OFQ, ppN/OFQ, and the
presumed targets of N/OFQ, N/OFQ receptors, are expressed (Narita et
al., 1999
; Houtani et al., 2000
). L-glutamate and glycine, together with GABA, serve as the major excitatory and inhibitory neurotransmitters in this CNS area, respectively (Doubell et al., 1999
). We have previously shown that N/OFQ selectively inhibits excitatory glutamatergic synaptic transmission in the rat spinal cord
dorsal horn via a presynaptic naloxone-insensitive mechanism (Liebel et
al., 1997
). In contrast, NST selectively reduced the release of the
inhibitory neurotransmitters glycine and GABA and left
-amino-3-hydroxy-5-methylisoxazole-4-propionic acid
receptor-mediated synaptic transmission unaffected (Zeilhofer et al.,
2000
). Although inhibition of excitatory synaptic transmission can be
competitively antagonized by the partial N/OFQ receptor antagonist
phe1
(CH2-NH)Gly2]-nociceptin-(1-13)NH2
(developed by Guerrini et al., 1998
), reduction of inhibitory synaptic
transmission by NST is insensitive to
phe1
(CH2-NH)Gly2]-nociceptin-(1-13)NH2
(Ahmadi et al., 2001
). These findings suggest that both peptides
act via different receptors and specifically target excitatory and
inhibitory transmitter release in the spinal cord dorsal horn.
The generation of mice deficient in N/OFQ receptors (Nishi et al.,
1997
) opened the possibility to directly investigate the role of the
N/OFQ receptor in the mediation of the cellular effects of N/OFQ and
NST. Here, we show that the inhibitory effect of N/OFQ on spinal
excitatory glutamatergic synaptic transmission is completely absent in
mice lacking N/OFQ receptors, whereas the suppression of inhibitory
synaptic transmission by NST is completely retained in these mice. In
addition, we show that in the mouse formalin test, N/OFQ induces
antinociception in wild-type mice after intrathecal injection, but has
no effects on the nociceptive behavior in
N/OFQ-R
/
mice.
| |
Experimental Procedures |
|---|
|
|
|---|
Animals.
Experiments were carried out on mutant mice lacking
the N/OFQ receptor (morcm1; Nishi et
al., 1997
), in wild-type mice and in mice heterozygous for this gene.
All mice used in our experiments were bred from heterozygous pairs of
mice (75% C57BL/6J, 25% 129 strain).
Slice Preparation and Electrophysiological Recordings.
Mice
of either sex that were 10 to 16 days old were killed in ether narcosis
by decapitation. Transverse slices of the lumbar spinal cord (250 µm
thick) were prepared as described previously (Liebel et al., 1997
;
Zeilhofer et al., 2000
). The tips of the mouse tails were stored at
70°C for post hoc genotyping. Whole-cell, patch-clamp recordings
were performed from neurons identified under visual control using the
infrared gradient contrast technique coupled to a video microscopy
system (Dodt and Zieglgänsberger, 1994
). Slices were completely
submerged and continuously superfused with external solution, which
contained 125 mM NaCl, 26 mM NaHCO3, 1.25 mM
NaH2PO4, 2.5 mM KCl, 2 mM
CaCl2, 1 mM MgCl2, and 10 mM glucose, pH 7.30, 315 mOsM) and was bubbled with 95%
O2/ 5% CO2. Patch pipettes
(4-5 M
) were filled with internal solution containing 130 mM
K-gluconate, 20 mM KCl, 2 mM MgCl2, 0.05 mM EGTA,
3 mM Na-ATP, 0.1 mM Na-GTP, and 10 mM Na-HEPES, pH 7.30. QX-314 (5 mM)
was added to the internal solution to block voltage-activated sodium
currents. Excitatory and inhibitory postsynaptic currents (EPSCs and
IPSCs) were evoked at a frequency of 0.1-0.07 Hz and recorded at a
holding potential of
80 mV at room temperature. Short hyperpolarizing
voltage steps to
90 mV were applied every minute to monitor input and
access resistance. EPSCs and IPSCs were elicited by ipsilateral
extracellular electrical stimulation (100 µs, 3-10 V) using a glass
electrode filled with 1 M NaCl. Peptide- or drug-containing solutions
were bath applied at a rate of 1 to 2 ml/min. Percentage inhibition of
postsynaptic currents by N/OFQ and NST was determined from the average
amplitude of 10 consecutive postsynaptic currents evoked immediately
before application of the peptides and when a steady state of
inhibition was reached, usually about 3 min after peptide application.
Behavioral Testing.
The pro- and/or antinociceptive effects
of N/OFQ were analyzed in the formalin test (Dubuisson and Dennis,
1977
), which has been adapted to mice. Experiments were performed at
room temperature. Male mice, 7 to 9 weeks old, were briefly
anesthetized with isoflurane and intrathecally injected at the level of
the lower lumbar spine with N/OFQ or vehicle (0.9% ACSF) in a total
volume of 5 µl. A small amount of black ink (1% v/v) was added to
the peptide or vehicle containing solutions to allow post hoc
verification of proper intrathecal injection. Only mice with normal
postinjection behavior and unimpaired motor function were used.
Wild-type and N/OFQ-R
/
mice were randomly
assigned to the different treatment groups consisting of 10 animals
each. Formalin (10 µl, 5%) was injected subcutaneously into the
dorsal surface of the left hind paw 10 min after intrathecal injection
of N/OFQ or vehicle. At this time point, mice had completely recovered
from light isoflurane narcosis. Flinches of the injected paw were
counted at 1-min intervals for 60 min after formalin injection. After
the tests, mice were killed by CO2 inhalation and
proper intrathecal injection was verified by visual inspection after
laminectomy. All behavioral tests and the killing of the animals were
performed in accordance with the institutional guidelines of the
University of Erlangen-Nürnberg.
Genotyping. The genotype of the animals was determined with PCR using the following primer pairs: Wild-type N/OFQ-R gene, 5'-GCC CAT CGA GGT GTT CAT GTG CCT GT and 5'-GAC CCG CCT ACC TGA GGA TGA CAT AC; targeted N/OFQ-R gene, 5'-GCC CAT CGA GGT GTT CAT GTG CCT GT, and 5'-CAA TAT CGC GGC TCA GTT CGA GGT GC.
Peptides.
NST (bovine) and methionine-enkephalin (met-enk)
were purchased from Tocris (Bristol, UK). N/OFQ was obtained from Dr.
M. Herkert (Institut für Biochemie, Universität
Erlangen-Nürnberg, Germany) and from Tocris. Peptides
(purity > 95%) were dissolved in external recording solution and
stored in aliquots (1 mM) at
20°C. Fresh dilutions were made with
standard external solution on every experimental day.
Statistical Analysis.
All measurements are given as
mean ± S.E.M. Statistical testing for significant inhibition
against the null hypothesis "no inhibition" was done with the
one-sample sign test. Significant differences in the degree of
inhibition were evaluated using ANOVA (
= 0.05) followed by
Fisher's post hoc test.
| |
Results |
|---|
|
|
|---|
Effects of N/OFQ and NST on Excitatory and Inhibitory Synaptic
Transmission in the Mouse Spinal Cord Dorsal Horn.
The effects of
N/OFQ and NST on excitatory synaptic transmission were tested in
superficial dorsal horn neurons of the mouse spinal cord using
whole-cell, patch-clamp recordings. Excitatory synaptic transmission
was studied in the presence of the GABAA receptor
blocker bicuculline (10 µM) and the glycine receptor blocker
strychnine (2 µM). EPSCs evoked by extracellular electrical stimulation of the dorsal root entry zone and recorded at a holding potential of
80 mV were mediated by ionotropic glutamate receptors as
indicated by their sensitivity to a combination of the
N-methyl-D-aspartate and
non-N-methyl-D-aspartate receptor antagonists,
D-APV (50 µM) and CNQX (10 µM), respectively. At this
holding potential, EPSCs were almost exclusively mediated by
-amino-3-hydroxy-5-methylisoxazole-4-propionic acid receptors.
EPSC amplitude: 3.1 ± 3.4% (mean ± S.E.M.),
n = 10), but were reduced in amplitude by N/OFQ (10 µM) by 37.3 ± 5.4% (n = 10; Fig.
1, A and C). In contrast, IPSCs, which
were recorded in the presence of CNQX (10 µM) and APV (50 µM),
remained almost unchanged when N/OFQ was applied (
IPSC amplitude,
1.3 ± 4.8%; n = 10) but were significantly
reduced by 41.6 ± 3.9% in the presence of NST (10 µM; Fig. 1,
B and C).
|
Lack of Spinal EPSC Inhibition by N/OFQ in N/OFQ-R
/
Mice.
We next compared the effects of N/OFQ (10 µM) on
excitatory glutamatergic transmission in wild-type, heterozygous and
N/OFQ-R
/
mice. N/OFQ-mediated inhibition was
statistically significant in wild-type mice (40.8 ± 3.8%;
n = 19; p
0.001) and heterozygous mice (23.2 ± 4.1%; n = 10; p
0.001), but not in N/OFQ-R
/
mice (4.1 ± 2.1%;
p = 0.77; Fig. 2A)
indicating a pronounced gene dose effect of N/OFQ receptor expression
in the mouse spinal cord dorsal horn (Fig. 2 B).
|
/
littermates. Even at a concentration of 10 µM, no significant reduction was seen (
EPSC amplitude: 5.5 ± 2.52%;
n = 6).
Further experiments were performed to demonstrate that the observed
lack of modulation of excitatory transmission by N/OFQ in the
N/OFQ-R
/
mice was indeed caused by the lack
of N/OFQ receptors and not by a disturbance of downstream modulatory
elements. In these experiments, we used met-enk (10 µM), which
suppresses spinal glutamatergic synaptic transmission via receptors
different from the N/OFQ receptor (Liebel et al., 1997
and µ opioid receptors (Goldstein and Naidu, 1989
|
Inhibition of Spinal IPSCs by NST is retained in
N/OFQ-R
/
Mice.
In contrast to EPSC inhibition by
N/OFQ, inhibition by NST of IPSCs was completely retained in
N/OFQ-R
/
mice (Fig.
4A). Statistically significant inhibition
by NST (10 µM) was achieved in all three genotypes (
IPSC amplitude
35.5 ± 4.0%; 44.8 ± 3.5%, and 49.4 ± 6.1%;
n = 7-19; p
0.05-0.001). No
statistically significant differences were found in the degree of
inhibition by 10 µM NST (ANOVA, at
= 0.05; Fig. 4B) between the different genotypes. When NST was applied in different
concentrations ranging from 0.1 to 10 µM, no significant difference
in the concentration dependence was detected (Fig. 4C).
|
The Antinociceptive Effect of N/OFQ Is Absent in
N/OFQ-R
/
Mice.
The inhibitory effects of N/OFQ on
spinal glutamatergic synaptic transmission are similar to those of
classical opioids like morphine and may well underlie the spinal
antinociceptive effects of N/OFQ seen after intrathecal injection of
nanomolar doses. Because the cellular effects of N/OFQ on spinal
neurotransmission were absent in N/OFQ-R
/
mice, we tested whether these mice also lacked the antinociceptive effects of spinally applied N/OFQ. When N/OFQ (3 nmol/mouse in 5 µl
of ACSF) was injected intrathecal to the lumbar spinal cord of
wild-type mice, a statistically significant reduction of the number of
flinches as compared with vehicle (5 µl of ACSF)-injected wild-type
mice was observed during both phases of the formalin test (Fig.
5, A and B). By contrast,
N/OFQ-R
/
mice exhibited no change in
nociceptive behavior after injection of N/OFQ compared with
vehicle-injected mice (Fig. 5, C and D).
|
/
mice showed a nearly identical number of flinches/min during phase I
(1-10 min after formalin injection) of the formalin test (8.29 ± 1.33 versus 9.20 ± 1.75; n = 10). However, during
phase II (20-60 min), N/OFQ-R
/
exhibited
markedly increased nociceptive behavior (5.80 ± 1.49 versus
3.45 ± 0.46 flinches/min; n = 10). No signs of
motor dysfunction were observed after intrathecal injection of N/OFQ.
| |
Discussion |
|---|
|
|
|---|
In previous reports, we have shown that NST selectively reduces
the amplitudes of GABAergic and glycinergic IPSCs in the rat spinal
cord, whereas N/OFQ interferes only with excitatory glutamatergic synaptic transmission (Liebel et al., 1997
; Zeilhofer et al., 2000
). In
the present study, we provide direct evidence that the inhibitory
effect of N/OFQ on spinal glutamatergic synaptic transmission as well
as the antinociceptive effects of spinally administered N/OFQ are
mediated by N/OFQ receptors, whereas the functional N/OFQ antagonist
NST acts via a different receptor.
N/OFQ-Mediated Inhibition of Spinal Glutamatergic Transmission
Occurs via N/OFQ Receptors.
The inhibition of EPSCs by N/OFQ in
the rat spinal cord dorsal horn occurred at relatively high
concentrations, with an EC50 value in the range
of 400 to 500 nM (Liebel et al., 1997
). N/OFQ has been reported to be
significantly more potent in other preparations. In the rat CNS,
activation of G protein-coupled inwardly rectifying potassium currents
(Meis and Pape, 1998
) and inhibition of Ca2+
currents (Connor et al., 1999
) by N/OFQ occur with
EC50 values in the low nanomolar range. A
possible explanation for these apparent differences in potency might be
that N/OFQ acts via more than one receptor type. There is indeed
evidence for two different binding sites with different affinities for
N/OFQ (Mathis et al., 1997
). Furthermore, in a recent report, it has
been proposed that an eleven amino acid N/OFQ fragment (N/OFQ[1-11])
might specifically bind to a low affinity N/OFQ binding site, which is
different from the N/OFQ receptor found in Chines hamster ovary cells
transfected with the N/OFQ receptor (opioid receptor-like 1) gene
(Letchworth et al., 2000
).
/
mice, excludes the
possibility that the disruption of the N/OFQ receptor gene not only
prevented modulation by N/OFQ, but also affected modulation of
excitatory synaptic transmission in general. This is further rendered
unlikely by our in vivo finding that the baseline sensitivity to
nociceptive stimuli, as assessed in the first phase of the formalin
test, was nearly identical in N/OFQ-R
/
and
wild-type mice.
Our results are therefore in line with previous results by Inoue et al.
(1999)
/
mice
lack the inhibitory effect of intrathecal-injected N/OFQ on substance
P-induced nociceptive responses. They do not, however, exclude the
possibility that the effects of N/OFQ occur via receptor subtypes
different from the "classical" N/OFQ receptor, which might
originate from post-transcriptional modifications. The existence of
different splice variants of the mouse N/OFQ receptor has indeed been
demonstrated (Wang et al., 1994N/OFQ in the Formalin Test.
In our experiments, we have tested
the effects of N/OFQ in a tonic pain model that allows monitoring of
spontaneous nociceptive behavior for 1 h. In this test, spinally
applied N/OFQ (3 nmol/mouse) exhibited antinociceptive effects that
were completely absent in N/OFQ-R
/
mice,
indicating that the antinociceptive action of N/OFQ is mediated by the
N/OFQ receptor.
/
mice have mainly been evaluated in
rather acute pain models, such as the tail-flick test; in other cases,
nociceptive behavior has been followed for only relatively short
periods of time (e.g., 10 min in the mouse abdominal constriction;
Nishi et al., 1997
/
mice was nearly identical during
phase I (1-10 min). However, during phase II (20-60 min),
N/OFQ-R
/
mice showed an increased flinching
behavior, which may suggest a role of endogenous N/OFQ in pain
modulation. These findings support those earlier reported by Tian et
al. (1998)
/
mice. The
lack of a nociceptive phenotype in these mice might again be explained
by the use of an acute versus a tonic pain model. Another explanation
might be that these mice lack not only N/OFQ but also other
biologically active peptides potentially released from the same
precursor peptide, such as NST or orphanin FQ 2 (Rossi et al.,
1998Effects of N/OFQ and NST Occur via Different Receptors.
Although the inhibitory effect of N/OFQ on EPSCs was completely
absent in N/OFQ-R
/
mice and strongly reduced
in N/OFQ-R+/
mice, inhibition of IPSCs by NST
remained unchanged. These results clearly show that N/OFQ and NST act
via separate receptors and demonstrate that the synaptic effects of NST
do not depend on the presence of functional N/OFQ receptors. Our
results therefore confirm those of others who have demonstrated that
NST does not compete with N/OFQ for N/OFQ receptor binding
(Okuda-Ashitaka et al., 1998
) and that NST neither mimics nor blocks
N/OFQ receptor-mediated Ca2+ current inhibition
(Connor et al., 1999
). Together with our previous finding that the
synaptic effects of NST depend on the activation of pertussis
toxin-sensitive G proteins (Zeilhofer et al., 2000
), these results
indicate that NST is a biologically active peptide per se, which
activates a membrane receptor that has yet to be identified.
Conclusion and Implications.
The observation that
N/OFQ-R
/
mice lack both the inhibitory effect
of N/OFQ on glutamatergic neurotransmission and the antinociceptive activity of spinally applied N/OFQ strengthens the possibility that the
synaptic action of N/OFQ underlies the antinociceptive action observed
in vivo. In this respect, the action of N/OFQ resembles that of
classical opioid peptides at the spinal cord level. In higher brain
areas, however, where most of the unwanted opioid effects arise,
activation of N/OFQ receptors by N/OFQ or by nonpeptide N/OFQ receptor
agonists seems to exert anti-opioidergic (Mogil et al., 1996
) and
anxiolytic (Jenck et al., 2000
) effects. The combination of opioid-like
analgesia at the spinal cord level and anti-opioidergic (Mogil et al.,
1996
) and anxiolytic activity (Köster et al., 1999
; Jenck et al.,
2000
) in higher brain areas makes the N/OFQ system a promising target
for the development of new analgesics possibly devoid of the unwanted
side effects of classical opioids.
| |
Acknowledgments |
|---|
We thank Dr. Rachel Jurd for critical reading of the manuscript and Susanne Gabriel and Beate Layh for excellent technical assistance.
| |
Footnotes |
|---|
Received August 9, 2000; Accepted December 1, 2000
This work was supported by a grant from the Deutsche Forschungsgemeinschaft to H.U.Z. and A.P. (SFB 353/A8).
This work was supported by a grant from the Deutsche Forschungsgemeinschaft to H.U.Z. and A.P. (SFB 353/A8).
Send reprint requests to: Dr. H.U. Zeilhofer, Institut für Pharmakologie und Toxikologie, Universität Zürich, Winterthurerstrasse 190, CH-8057 Zürich, Switzerland. E-mail: zeilhofe{at}pharma.unizh.ch
| |
Abbreviations |
|---|
NOFQ, nociceptin/orphanin FQ;
NST, nocistatin;
ppN/OFQ, prepronociceptin/orphanin FQ;
GABA,
-aminobutyric acid;
CNS, central nervous system;
EPSC, excitatory postsynaptic current;
IPSC, inhibitory postsynaptic current;
ACSF, artificial cerebrospinal
fluid;
met-enk, methionine-enkephalin;
ANOVA, analysis of
variance;
CNQX, 6-cyano-7-nitroquinoxaline-2,3-dione;
APV, 2-amino-5-phosphonovaleric acid;
N/OFQ-R, N/OFQ receptor.
| |
References |
|---|
|
|
|---|
-nociceptin1-13, nocistatin and prepronociceptin154-181 effects on calcium channel currents and a potassium current in rat locus coeruleus in vitro.
Br J Pharmacol
128:
1779-1787[Medline].
S binding and immunohistochemistry.
Br J Pharmacol
128:
1300-1306[Medline].This article has been cited by other articles:
![]() |
J. W. Lynch Molecular Structure and Function of the Glycine Receptor Chloride Channel Physiol Rev, October 1, 2004; 84(4): 1051 - 1095. [Abstract] [Full Text] [PDF] |
||||
![]() |
H. U. Zeilhofer and G. Calo Nociceptin/Orphanin FQ and Its Receptor--Potential Targets for Pain Therapy? J. Pharmacol. Exp. Ther., August 1, 2003; 306(2): 423 - 429. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Inoue, T. Kawashima, R. G. Allen, and H. Ueda Nocistatin and Prepro-Nociceptin/Orphanin FQ 160-187 Cause Nociception through Activation of Gi/o in Capsaicin-Sensitive and of Gs in Capsaicin-Insensitive Nociceptors, Respectively J. Pharmacol. Exp. Ther., July 1, 2003; 306(1): 141 - 146. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. D. Moran and P. A. Smith Morphine-3beta -D-glucuronide Suppresses Inhibitory Synaptic Transmission in Rat Substantia Gelatinosa J. Pharmacol. Exp. Ther., August 1, 2002; 302(2): 568 - 576. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||