Full text of DOI:10.1007/BF03019761

2
88
REGIO N AL AN EST H ESIA A N D P AI N
Synergistic analgesic
effects of intrathecal
midazolam and NMDA
or AMPA receptor
antagonists in rats
Tomoki Nishiyama MD PhD,*†
Laszlo Gyermek MD PhD ,†
Chingmuh Lee MD ,†
Sachiko Kawasaki-Yatsugi PhD,‡
Tokio Yamaguchi PhD‡
Purpose: To investigate the interaction of midazolam and N-methyl-D-aspartate (NMDA) receptor or -amino-
- hydroxy-5-methyl isoxazole-4-propionic acid (AMPA) receptor antagonist on the effects of persistent inflam-
3
matory nociceptive activation.
Methods: Male Sprague-Dawley rats were implanted with lumbar intrathecal catheters and were tested for their
responses to subcutaneous formalin injection into the hindpaw. Saline, midazolam (1 to 100 µg), AP-5 (1 to 30
µg), a NMDA receptor antagonist, or YM872 (0.3 to 30 µg), an AMPA receptor antagonist was injected intrathe-
cally 10 min before formalin injection. The combinations of midazolam and AP-5 or YM872 in a constant dose
ratio based on the 50% effective dose (ED ) were also tested and were analysed with an isobologram.
50
Results: Dose-dependent effects were observed with midazolam (ED50 was 1.34 µg and 1.21 µg in phase 1 and
of the formalin test, respectively), AP-5 (7.64 µg and 1.4 µg) and YM872 (0.24 µg and 0.21 µg). Synergistic
2
effects in both phases were obtained when combining midazolam with AP-5 or YM872. The ED50 of midazolam
decreased to 0.012 µg (phase 1) and 0.27 µg (phase 2) with AP-5 and to 0.09 µg (phase 1) and 0.35 µg (phase
2
) with YM872 (P < 0.01).
Conclusions: These results suggest a functional coupling of benzodiazepine—aminobutyric acid (GABA) receptor
A
with NMDA and AMPA receptors in acute and persistent inflammatory nociceptive mechanisms in the spinal cord.
Objectif : Examiner l’interaction du midazolam et d’un antagoniste du N-méthyl-D-aspartate (NMDA) ou d’un
antagoniste du récepteur de l’acide -amino-3-hydroxy-5-méthylisoxazole-4-propionique (AAMP) sur les effets de
l’activation nociceptive de l’inflammation persistante.
Méthode : On a implanté à des rats mâles Sprague-Dawley des cathéters intrathécaux lombaires et on a testé
leurs réactions à l’injection sous-cutanée de formaline dans la patte arrière. Une solution salée, du midazolam (1
à 100 µg), de l’AP-5 (1 à 30 µg), un antagoniste du récepteur NMDA ou le YM872 (0,3 à 30 µg), un antagoniste
du récepteur AMPA a été administré en injection intrathécale 10 min avant l’injection de formaline. Les combi-
naisons de midazolam et d’AP-5 ou de YM872 en dose au rapport constant, fondée sur la dose efficace 50 %
(
ED50) ont été aussi testées et analysées à l’aide d’un isobologramme.
Résultats : Des effets dépendants de la dose ont été observés avec le midazolam (ED50 a été de 1,34 µg et
,21 µg en phases 1 et 2 du test de formaline, respectivement), avec AP-5 (7,64 µg et 1,4 µg) et avec YM872
0,24 µg et 0,21 µg). Les effets synergiques des deux phases ont été obtenus en combinant le midazolam et AP-
ou YM872. La ED50 du midazolam a baissé à 0,012 µg (phase 1) et à 0,27 µg (phase 2) avec AP-5 et à 0,09
1
(
5
µg (phase 1) et à 0,35 µg (phase 2) avec YM872 (P < 0,01).
Conclusion : Ces résultats suggèrent un couplage fonctionnel des récepteurs NMDA ou AMPA avec les récep-
teurs benzodiazépine-acide -aminobutyrique (GABA) comme partie des mécanismes nociceptifs de l’inflamma-
A
tion aiguë et persistante dans la moelle épinière.
From the Department of Surgical Center,* Institute of Medical Science, University of Tokyo, Tokyo, Japan, the Department of
Anesthesiology,† Harbor-University of California, Los Angeles Medical Center, Torrance, California, USA, and the Institute for Drug
Discovery Research,‡ Yamanouchi Pharmaceutical Company, Ibaraki, Japan.
Address correspondence to: Dr. Tomoki Nishiyama, 3-2-6-603, Kawaguchi, Kawaguchi-shi, Saitama, 332-0015, Japan.
Phone: 81-3-5449-5356 (Dept.); Fax. 81-3-5449-5356 (Dept.); E-mail: nishiyam@ims.u-tokyo.ac.jp
Accepted for publication November 18, 2000.

Nishiyama et al.: M I D AZ O LAM A N D GLU TAMATE REC EPTO R
289
N the spinal cord, both the excitatory neuro-
Materials and methods
Animal preparations
1
transmitter, glutamate and the inhibitory neu-
2
rotransmitter, - aminobutyric acid (GABA) are
The protocol was approved by the Research and
Education Institute of H arbor - UCLA Medical Center.
Male Sprague-Dawley rats (280–300 g; BK Universal,
Fremont, CA, USA) were implanted with chronic lum-
bar intrathecal catheters under halothane (2%) anesthe-
I
involved in nociceptive mechanisms. The N-
methyl-D-aspartate (NMDA) receptors are thought
1
to mediate persistent nociceptive activation, while the
-
amino-3-hydroxy-5-methyl isoxazole-4- propionic
1
2
acid (AMPA) receptors are considered to have a role
sia according to the previously described method.
3
in acute nociception. The AMPA receptors also medi-
Briefly, an 8.5-cm polyethylene (PE-10; Clay Adams,
Parsippany, NJ, USA) catheter was advanced caudally
through an incision in the atlanto-occipital membrane,
to the thoracolumbar level of the spinal cord. The
external part of the catheter was tunnelled subcuta-
neously to exit on the top of the skull and plugged with
a 28G stainless steel wire. The position of the catheter
was checked by the aspiration of cerebrospinal fluid at
the implantation and was directly verified after sacrific-
ing the rat. Only rats with normal motor function and
behaviour five days after surgery were used. In total,
176 rats were used.
4
ate facilitated state of pain processing. Both GABAA
and GABA receptors also have roles in spinally medi-
B
5
ated nociception. The benzodiazepine binding site is
–
present on the GABA receptor (Cl channel) in the
A
6
spinal cord and benzodiazepines have spinally medi-
ated antinociceptive effects.7
The relations between different receptors involved
in spinal nociceptive mechanisms are studied primarily
with opioid receptors. Morphine, a µ-opioid receptor
agonist, possesses a synergistic antinociceptive interac-
8
9
tion with a 2 adrenergic agonist, midazolam, or
3
AMPA receptor antagonist. H owever, it did not show
synergy with NMDA receptor antagonist or NMDA
glycine site antagonist. The GABA and glutamate
Drugs and administration
Midazolam (a benzodiazepine-GABA receptor ago-
3
A
receptors may operate in concert to regulate nocicep-
nist; Sigma, St. Louis, MO, USA) 1, 3, 10, 30, and
100 µg, and AP-5 (2-amino-5-phosphonovaleic acid,
a NMDA receptor antagonist; Sigma, St. Louis, MO,
USA) 1, 3, 10, and 30 µg were dissolved in 10 µL
saline. YM872 {[2,3-Dioxo-7-(1H -imidazol-1-yl)-6-
nitro-1,2,3,4-tetrahydro-1-quinoxalinyl] acetic acid},
an AMPA receptor antagonist (Yamanouchi
Pharmaceutical Co. Ltd., Tsukuba, Japan) 10 mg was
dissolved in 0.97 mL distilled water with 30 µL, 1N
NaOH added to adjust pH to 7.3–7.5. Solutions of
0.3 (0.86), 1 (2.86), 3 (8.59), 10 (28.63), or 30
(85.89) µg (nMol) per 10 µL were made using nor-
mal saline. Normal saline, 10 µL, was used in the con-
trol group. After intrathecal drug or saline injection,
the catheter was flushed with an injection of 10 µL
normal saline to clear the dead space of the catheter (7
10
tive signals. There are few studies that address the
relation between GABA receptors and glutamate
receptors in spinal nociceptive mechanisms. We have
reported that midazolam, a benzodiazepine-GABAA
receptor complex agonist, exhibited synergistic anal-
gesia for thermally induced acute nociception both
1
1
with NMDA and AMPA receptor antagonists.
H owever, there are no reports of the interaction
between these receptors in other types of nociceptive
mechanisms. The purpose of the present study was to
investigate the relationships of the benzodiazepine-
GABA receptor complex and the NMDA receptor or
A
the AMPA receptor in the mechanisms of persistent
inflammatory nociceptive activation in the formalin
test of intrathecally catheterized rat model.
T ABLE E D5 0 (50% effective dose) values (µg)
Each single agent
AP-5 + Midazolam
YM872 + Midazolam
Phase 1
Phase 2
Phase 1
Phase 2
Phase 1
Phase 2
ED5 0 of
Midazolam
AP-5
1.34
(0.29–6.19)
7.64
1.21
(0.19–7.63)
1.4
0.012*
(0.001–0.13)
0.014*
0.27*
(0.15–0.49)
0.32*
0.09*
(0.027–0.30)
N / A
0.35*
(0.21–0.58)
N / A
(
4.3–13.6)
0.24
0.08–0.75)
(0.05–35.86)
0.21
(0.06–0.73)
(0.001–0.13)
N/ A
(0.18–0.57)
N / A
YM872
0.015*
0.058*
(
(0.005–0.05)
(0.035–0.097)
In the parenthesis, 95% confidence intervals are shown
P <0.01 vs the values at each single agent
N/ A=not available
*

2
90
C AN AD I AN JO U RN AL O F AN ESTH ESIA
FIGU RE 1 Time course of the effects of intrathecal midazolam (A), AP-5 (NMDA receptor antagonist, B), YM872 (AMPA receptor
antagonist, C), AP-5 + midazolam (D), and YM872 + midazolam (E) on the formalin test. Each point represents mean ± SEM of eight
animals. AP: AP-5, Mid: midazolam, YM: YM872. The numbers after symbols, AP, Mid and YM are the doses expressed in µg.

Nishiyama et al.: M I D AZ O LAM A N D GLU TAMATE REC EPTO R
291
FIGU RE 2 Dose response curves of the effects of intrathecal midazolam, AP-5 (NMDA receptor antagonist), YM872 (AMPA receptor
antagonist) (A, B), AP-5 + midazolam, and YM872 + midazolam (C, D) on phase 1 and 2 of the formalin test. Each point represents
mean ± SEM of eight animals. AP: AP-5, Mid: midazolam, YM: YM872.
±
0.4 µL, mean ± standard error). Micro injector
placed in an open Plexiglas chamber and observed for
60 min. Quantification of pain behaviour was made by
counting the incidence of spontaneous flinches/ shak-
ing of the injected paw at 1–2 min, 5–6 min and at 5-
min intervals during 10- to 60-min periods after
formalin injection. The animals were then sacrificed
with an overdose of halothane. As previously
described, two distinct phases were observed after for-
malin injection: phase 1, during 0–6 min after injection,
and phase 2, beginning about 10 min after injection.
syringes were used for all injections. In each dose
group, eight randomly selected rats were used.
Nociceptive test
Rats were kept in an open Plexiglas chamber with a
diameter of 40 cm for 30 min before drug injection.
Ten minutes after the intrathecal administration of the
agent, 50 µL formalin 5% were injected subcutaneous-
ly into the dorsal surface of the right hindpaw with a 30
G needle. Immediately after injection, the rat was
1
3

2
92
C AN AD I AN JO U RN AL O F AN ESTH ESIA
Data analysis and statistics
The ED5 0 values were calculated using the % maxi-
mum possible effect (% MPE) by a computer program
made in the Anesthesiology Laboratory of University
3
,11
of California, San Diego.
The % MPE was the %
against the value in the control group.
To describe the magnitude of interaction between
the agents, a total fractional dose value was calculated as
follows: [(ED5 0 dose of drug 1 in combination) /
(
2
ED5 0 value for drug 1 alone)] + [(ED5 0 dose of drug
in combination) / (ED50 value for drug 2 alone)].
The values were normalised by assigning the ED5 0 val-
ues of the agents given alone as 1. Values near to 1 sug-
gest an additive interaction, values >1 imply an
antagonistic interaction, while values <1 indicate a syn-
ergistic interaction.
Student’s t test was used to compare the calculated
E D5 0 values with the theoretical additive values. A P
value <0.05 was considered statistically significant.
Results
Analgesic effects of each agent
Midazolam, AP-5 and YM872 dose-dependently
decreased the number of flinches in both phase 1 and
phase 2 (time course was shown in the Figures 1A, 1B,
and 1C, and dose response was shown in the Figures
2A and 2B). The ED5 0 values are shown in the Table.
FIGU RE 3 Isobolograms for the intrathecal interaction of AP-5
or YM872 and midazolam in both phase 1 and 2 of the formalin
test. Horizontal and vertical bars indicate 95% confidence interval.
The oblique lines between the x axis and y axis are the theoretical
additive lines. The points in the middle of this line are the theoret-
ical additive points calculated from separate ED5 0 values. The
experimental points lie far below the additive line , indicating very
marked significant synergism (P <0.01 in phase 1 (A) and P=0.038
in phase 2 (B) of AP-5 + midazolam, P <0.01 in phase 1 (C) and
P=0.029 in phase 2 (D) of YM872 + midazolam).
Analgesic interaction between midazolam and AP-5
or YM872
Intrathecal administration of the combination of
midazolam and AP-5 produced dose-dependent anal-
gesic effects on both phase 1 and phase 2 of the for-
malin test (time course was shown in the Figure 1D,
and dose response was shown in the Figures 2C and
2D). The ED5 0 values for midazolam and AP-5
decreased in the combination compared with each sin-
gle agent (Table). Total fractional dose values were
0.01 in phase 1 and 0.45 in phase 2. Isobolographs are
shown in the Figure (Figures 3A, 3B).
Experimental paradigm
The first series of experiments were performed to
determine the dose-dependency of the antinociceptive
effects of intrathecally administered midazolam, AP-5,
and YM872 on the formalin test.
To investigate the interaction between midazolam
1
4
and AP-5 or YM872, an isobolographic analysis was
used. The method is based on comparisons of the dose
ratios that are determined to be equieffective. First, the
respective 50% effective dose (ED ) values were deter-
The combination of midazolam and YM872 also
induced dose-dependent decreases of flinches in both
phase 1 and phase 2 of the formalin test (time course was
shown in the Figure 1E, and dose response was shown in
the Figures 2C and 2D). The ED values for midazolam
5
0
mined from the dose-response data of each agent alone.
Subsequently, dose-response data were obtained by
coadministration of the two drugs in a constant dose
ratio based on the ED5 0 values of the phase 2 of the sin-
gle agents; i.e., combinations of each ½ ED , 1/ 4 ED ,
5
0
and YM872 decreased in the combination compared with
each single agent (Table). Total fractional dose values
were 0.07 in phase 1 and 0.57 in phase 2. Isobolographs
are shown in the Figure (Figures 3C, 3D).
5
0
5 0
1
/ 8 ED , or 1/ 16 ED doses. From the dose-response
5 0 5 0
data of the combined drugs, the ED5 0 value of the mix-
ture was calculated (calculated ED5 0 values). The theo-
retical additive values were the ½ ED s of the single
Discussion
Intrathecally administered midazolam (benzodi-
5
0
agents. Eight rats were used for each dose combination.
azepine-GABA receptor agonist), AP-5 (NMDA
A

Nishiyama et al.: M I D AZ O LAM A N D GLU TAMATE REC EPTO R
293
H unter and Singh,1 3 YM872 used in our present
study, and other compounds in the study of Simmons
receptor antagonist) and YM872 (AMPA receptor
antagonist) produce dose-dependent decreases in the
number of flinches in both phase 1 and phase 2 of the
formalin test in rats. Midazolam showed synergistic
suppressive effects with AP-5 and YM872 on the
number of flinches of both phase 1 and phase 2. The
synergism of midazolam was more pronounced with
AP-5 than with YM872 in phase 1 of the formalin test.
4
et al. have different affinities to the AMPA and
4
,22
kainate subtypes of ionotropic glutamate receptors.
Further experiment to compare the affinity of these
compounds to different subtypes in the same study is
necessary to discuss the discrepancies of the analgesic
effect furthermore.
1
1
The GABA receptors are proposed to exist at the
As we already reported, midazolam and NMDA
receptor antagonist or AMPA receptor antagonist had
synergistic antinociception for acute thermal stimula-
tion. Similarly in the present study, midazolam was
very markedly synergistic with NMDA receptor antag-
onist and AMPA receptor antagonist in suppressing
phase 1 response of the formalin test. Curiously, AP-
5, which is usually less effective for acute pain,
enhanced the effect of midazolam in phase 1 of the
formalin test more than YM872. The phase 1 response
is due to the direct stimulation of nociceptors by for-
malin or tissue damage and is thought to be an acute
A
primary afferent terminal in the spinal cord; and the
GABAergic system plays an important role in the
1
3
presynaptic inhibition of primary afferents.
Benzodiazepine receptors are concerned with pain
7
transmission in the dorsal horn of the spinal cord.
Benzodiazepine receptor agonists increase the intrin-
sic efficacy of GABA at the GABA receptor coupling
A
6
with benzodiazepine receptor in the spinal cord by
increasing Cl conductance. Midazolam, a benzodi-
azepine derivative, has spinally mediated anal-
–
5
7
, 1 1
1 6
gesic and anesthetic effects in animal experiments.
In the present study, we could show analgesic effects
of intrathecally administered midazolam on both for-
malin induced acute nociception and persistent noci-
ceptive activation. Goodchild et al. reported that
intrathecal midazolam caused spinally mediated
antinociception in rats by a mechanism involving opi-
oid receptor activation.1 7 We did not use any antago-
nists for GABA or opioid receptors in the present
study. Therefore, further studies are necessary to veri-
fy the detailed mechanism of the analgesic action of
midazolam in the spinal cord.
2
3
pain reaction. In addition, in the phase 2 response of
the formalin test, midazolam had synergistic antinoci-
ception with NMDA and AMPA receptor antagonists
in the present study. The phase 2 response is due to
subsequent inflammation after formalin injection and
2
4
central sensitization related to C fibre activity.
Therefore, in the spinal cord, benzodiazepine-GABAA
receptor and NMDA or AMPA receptor might have
some functional coupling in both acute and persistent
inflammatory nociception.
–
Benzodiazepines increase the Cl conductance of
Glutamates also exist in primary afferents and in
interneurons.1 8 The NMDA receptors are involved in
persistent nociceptive input of deep dorsal horn
GABA on the primary afferents hyperpolarizing the
A
6
afferent, therefore, reducing the release of glutamate in
2
5
the spinal cord. The glutamate receptor antagonists
inhibit the binding of glutamate post-synaptically
reducing the depolarization of the postsynaptic cell.
Reduction of glutamate release and binding would
decrease transmission and, therefore, pain perception.
1
9
cells. Accordingly, NMDA receptor antagonists are
1
effective in phase 2 of the formalin test rather than on
3
acute nociception. This is consistent with our present
results in which the ED5 0 value of AP-5 was smaller in
phase 2 than in phase 1 of the formalin test.
It is also reported that a GABA receptor agonist inhib-
A
The AMPA receptors are localized in superficial
laminae of the dorsal horn pre- and postsynaptically.
ited the behavioural effects of NMDA, quisqualic acid
2
0
1 0
and kainic acid. Thus, GABA receptors might have
A
These receptors alter acute afferent evoked excitation.
Intrathecal application of AMPA receptor antagonists
some functional coupling with glutamate receptors in
the spinal cord with regard to nociceptive transmission.
Regarding side effects, we did not investigate any
behavioural effects in this study. H owever, in our pre-
vious study, the combinations of midazolam with
AP-5 or YM872 decreased side effects compared to
each single agent. Therefore, we also expected
decreased side effects in the same combinations in the
present study.
are commonly thought to decrease the responses to
acute noxious stimulus.1
1,21
H owever, on the formalin
1
1
test of the animals, H unter et al. reported that the
AMPA receptor antagonist inhibited only phase 1, but
not phase 2 responses.1 3 In contrast, Simmons et al.
reported that the AMPA receptor antagonist reduced
4
phase 2 response but not phase 1. In our present
study, YM872 inhibited both phase 1 and phase 2
responses. The discrepancies in these results might be
due to the different compounds used: NBQX used by
In conclusion, intrathecal coadministration of
midazolam (a benzodiazepine-GABA receptor ago-
A
nist) with AP-5 (a NMDA receptor antagonist) or

2
94
C AN AD I AN JO U RN AL O F AN ESTH ESIA
midazolam with YM872 (an AMPA receptor antago-
nist) produced synergistic analgesia on formalin
induced acute nociception and persistent nociceptive
activation. These results suggest a functional coupling
1 1 Nishiyama T, Gyermek L, Lee C, Kawasaki-Yatsugi S,
Yamaguchi T. Analgesic interaction between intrathecal
midazolam and glutamate receptor antagonists on ther-
mal-induced pain in rats. Anesthesiology 1999; 91: 531–7.
1 2 Yaksh TL, Rudy TA. Chronic catheterization of the spinal
subarachnoid space. Physiol Behav 1976; 17: 1031–6.
13 Hunter JC, Singh L. Role of excitatory amino acid
receptors in the mediation of the nociceptive response to
formalin in the rat. Neurosci Lett 1994; 174: 217–21.
14 Tallarida RJ, Porreca F, Cowan A. Statistical analysis of
drug-drug and site-site interactions with isobolograms.
Life Sci 1989; 45: 947–61.
of benzodiazepine-GABA receptor with NMDA and
A
AMPA receptors in various nociceptive mechanisms in
the spinal cord. These synergisms can reduce dose
requirements (increase therapeutic index) in conse-
quence can reduce toxicity when projected to possible
clinical application.
Acknowledgements
We thank to Dr. Ang Ji, Dr. Young-moon Cho, and
Nguyen B. Nguyen, Department of Anesthesiology,
H arbor / University of California, Los Angeles, for
their assistance.
15 Sluka KA, Willis WD, Westlund KN. Inflammation-
induced release of excitatory amino acids is prevented
by spinal administration of a GABA but not by a
A
GABA receptor antagonist in rats. J Pharmacol Exp
B
Ther 1994; 271: 76–82.
References
16 Bahar M, Cohen ML, Grinshpon Y, Chanimov M. Spinal
anaesthesia with midazolam in the rat. Can J Anaesth
1997; 44: 208–15.
1
2
3
Näsström J, Karlsson U, Post C. Antinociceptive actions
of different classes of excitatory amino acid receptor
antagonists in mice. Eur J Pharmacol 1992; 212: 21–9.
Buckett WR. Induction of analgesia and morphine
potentiation by irreversible inhibitors of GABA-
transaminase. Br J Pharmacol 1979; 68: 129–30.
Nishiyama T, Yaksh TL, Weber E. Effects of intrathecal
NMDA and non-NMDA antagonists on acute thermal
nociception and their interaction with morphine.
Anesthesiology 1998; 89: 715–22.
17 Goodchild CS, Guo Z, Musgreave A, Gent JP.
Antinociception by intrathecal midazolam involves
endogenous neurotransmitters acting at spinal cord delta
opioid receptors. Br J Anaesth 1996; 77: 758 –63.
18 Battaglia G, Rustioni A. Coexistence of glutamate and
substance P in dorsal root ganglion neurons of the rat
and monkey. J Comp Neurol 1988; 277: 302–12.
19 Dickenson AH, Sullivan AF. Differential effects of exci-
tatory amino acid antagonists on dorsal horn nocicep-
tive neurones in the rat. Brain Res 1990; 506: 31–9.
20 Furuyama T, Kiyama H, Sato K, et al. Region-specific
expression of subunits of ionotropic glutamate recep-
tors (AMPA-type, KA-type and NMDA receptors) in
the rat spinal cord with special reference to nocicep-
tion. Mol Brain Res 1993; 18: 141–51.
21 Dougherty PM, Palecek J, Paleckova V, Sorkin LS, Willis
WD. The role of NMDA and non-NMDA excitatory
amino acid receptors in the excitation of primate
spinothalamic tract neurons by mechanical, chemical,
thermal, and electrical stimuli. J Neurosci 1992; 12:
3025–41.
22 Kohara A, Okada M, Tsutsumi R, et al. In-vitro charac-
terization of YM872, a selective, potent and highly
water-soluble -amino-3-hydroxy-5-methylisoxazole-4-
propionate receptor antagonist. J Pharm Pharmacol
1998; 50: 795–801.
4
Simmons RMA, Li DL, Hoo KH, Deverill M, Ornstein
PL, Iyengar S. Kainate GluR5 receptor subtype medi-
ates the nociceptive response to formalin in the rat.
Neuropharmacology 1998; 37: 25–36.
5
6
Sawynok J. GABAergic mechanisms of analgesia: an
update. Pharmacol Biochem Behav 1987; 26: 463–74.
Unnerstall JR, Kuhar MJ, Niehoff DL, Palacios JM.
Benzodiazepine receptors are coupled to a subpopula-
tion of -aminobutyric acid (GABA) receptors: evidence
from a quantitative autoradiographic study. J
Pharmacol Exp Ther 1981; 218: 797–804.
7
8
Goodchild CS, Serrao JM. Intrathecal midazolam in the
rat: evidence for spinally-mediated analgesia. Br J
Anaesth 1987; 59: 1563–70.
Monasky MS, Zinsmeister AR, Stevens CW, Yaksh TL.
Interaction of intrathecal morphine and ST-91 on
antinociception in the rat: dose-response analysis,
antagonism and clearance. J Pharmacol Exp Ther
1
990; 254: 383–92.
23 Tjølsen A, Berge O-G, Hunskaar S, Rosland JH, Hole K.
The formalin test: an evaluation of the method. Pain
1992; 51: 5–17.
24 Li J, Simone DA, Larson AA. Windup leads to charac-
teristics of central sensitization. Pain 1999; 79: 75–82.
25 Haefely WE. Benzodiazepines. Int Anesthesiol Clin
1988; 26: 262–72.
9
Yanez A, Sabbe MB, Stevens CW, Yaksh TL. Interaction
of midazolam and morphine in the spinal cord of the
rat. Neuropharmacology 1990; 29: 359–64.
0 Aanonsen LM, Wilcox GL. Muscimol, -aminobutyric
acid receptors and excitatory amino acids in the mouse
1
A
spinal cord. J Pharmacol Exp Ther 1989; 248: 1034–8.