Interactions of intrathecally administered ziconotide, a selective blocker ofneuronal N-type voltage-sensitive calcium channels, with morphine on

nociception in rats

Yong-Xiang Wang*, Da Gao, Mark Pettus, Cora Phillips, S. Scott Bowersox

Department of Pharmacology, Elan Pharmaceuticals, 3760 Haven Avenue, Menlo Park, CA 94025, USA

Received 13 February 1999; received in revised form 16 June 1999; accepted 13 August 1999

Abstract

Ziconotide is a selective, potent and reversible blocker of neuronal N-type voltage-sensitive calcium channels (VSCCs). Morphine is an

agonist of m-opioid receptors and inhibits N-type VSCC channels via a G-protein coupling mechanism. Both agents are antinociceptive when

they are administered intrathecally (spinally). The present study investigated the acute and chronic (7-day) interactions of intrathecally

administered ziconotide and morphine on nociception in several animal models of pain. In the acute study, intrathecal bolus injections of

morphine and ziconotide alone produced dose-dependent inhibition of formalin-induced tonic ¯inch responses and withdrawal responses to

paw pressure. The combination of ziconotide and morphine produced an additive inhibition of formalin-induced tonic ¯inch responses and a

signi®cant leftward shift of the morphine dose-response curve in the paw pressure test. After chronic (7-day) intrathecal infusion, ziconotide

enhanced morphine analgesia in the formalin test. In contrast, chronic intrathecal morphine infusion produced tolerance to analgesia, but did

not affect ziconotide antinociception. Antinociception produced by ziconotide alone was the same as that observed when the compound was

co-administered with morphine to morphine-tolerant rats. In the hot-plate and tail immersion tests, chronic intrathecal infusion of morphine

lead to rapid tolerance whereas ziconotide produced sustained analgesia with no loss of potency throughout the infusion period. Although

ziconotide in combination with morphine produced an apparent synergistic analgesic effects during the initial phase of continuous infusion, it

did not prevent morphine tolerance to analgesia. These results demonstrate that (1) acute intrathecal administrations of ziconotide and

morphine produce additive or synergistic analgesic effects; (2) chronic intrathecal morphine infusion results in tolerance to analgesia but does

not produce cross-tolerance to ziconotide; (3) chronic intrathecal ziconotide administration produces neither tolerance nor cross-tolerance to

morphine analgesia; (4) intrathecal ziconotide does not prevent or reverse morphine tolerance. q 2000 International Association for the

Study of Pain. Published by Elsevier Science B.V.

Keywords: Ziconotide; Voltage-sensitivity calcium channels (VSCCs); Blocker; Morphine; m-Opioid receptors; Interaction; Addition; Synergistism; Noci-

ception; Tolerance; Spinal cord; Intrathecal route

1. Introduction

Several sub-types of voltage-sensitive calcium channels

(VSCCs) (e.g. L-, N-, T-, P, Q-, and O-type) have been

identi®ed in the nervous system and other tissues (see

reviews by Olivera et al., 1994; Miljanich and Ramachan-

dran, 1995). De®ned as high af®nity binding sites for v -

conotoxin GVIA (Olivera et al., 1987), N-type calcium

channels are located in presynaptic nerve terminals where

they permit the calcium in¯ux necessary for transmitter

release from both central and peripheral nerves (see review

by Miljanich and Ramachandran, 1995). Ziconotide (SNX-

111), a 25-amino acid polypeptide, is the synthetic version

of the naturally occurring v -conotoxin MVIIA (see review

by Miljanich and Ramachandran, 1995). Ziconotide selec-

tively and reversibly binds to v -conotoxin GVIA binding

sites (N-type calcium channels) with an extremely high

potency (Kd < 8 pM) (Kristipati et al., 1994; Wang et al.,

1998), blocks calcium currents mediated by N-type VSCCs

in IMR-32 human neuroblastoma cells and other prepara-

tions (see Wang et al., 1998), and inhibits transmitter release

from peripheral and central neurons (Gaur et al., 1994;

Wang et al., 1998). Ziconotide or v -conotoxin MVIIA is

antinociceptive in animal models of acute, persistent and

Pain 84 (2000) 271±281

0304-3959/00/$20.00 q 2000 International Association for the Study of Pain. Published by Elsevier Science B.V.

PII: S0304-3959(99)00214-6

www.elsevier.nl/locate/pain

* Corresponding author. Tel.: 11-650-833-1231; fax: 11-650-614-

1002.

E-mail address: jwang@elanpharma.com (Y.-X. Wang)

neuropathic pain when it is administered intrathecally

(Chaplan et al., 1994; Malmberg and Yaksh, 1994, 1995;

Bowersox et al., 1996; Yamamoto and Sakashita, 1998), or

by other routes (Xiao and Bennett, 1995; White and

Cousins, 1998). Ziconotide is currently under clinical inves-

tigation for the treatment of acute and chronic pain

syndromes via an intrathecal route (Brose et al., 1997; Pres-

ley et al., 1998).

Morphine is a conventionally prescribed narcotic and a

spinal analgesic agent (Abram and Yaksh, 1993). It binds

mainly to m-opioid receptors coupled to GTP-binding

proteins (G-proteins) that initiate a variety of downstream

cellular effects, including reductions in intracellular cAMP

levels, increased potassium channel activity and decreased

calcium channel activity (Bourinet et al., 1996; see review

by Basbaum, 1996). These activities hyperpolarize neurons,

which results in inhibition of spike ®ring and decreased

neurotransmitter release (see review by Basbaum, 1996).

Recently, m-opioid receptors have been shown to be

coupled to the pore-forming a 1 subunits of N-type VSCCs

via G-proteins. Activation of m-opioid receptors leads to

inhibition of N-type calcium currents (Seward et al., 1991;

Bourinet et al., 1996; Toth et al., 1996; Zhang et al., 1996;

De Waard et al., 1997). Therefore, inhibition of N-type

calcium currents may be one of the mechanisms by which

morphine produces analgesia. Repeated administration

(including intrathecal administration) of morphine can

result in tolerance to analgesia, but the mechanism is not

well understood. It has been proposed that tolerance is due

to the uncoupling of opioid receptors and G-proteins

(Kennedy and Henderson, 1991, 1992; Tao et al., 1993;

see reviews by Basbaum, 1995, 1996). G-protein-coupled

effectors, such as calcium and potassium channel activities,

and cAMP levels are not altered in morphine tolerance

status (see reviews by Basbaum, 1995, 1996). However,

an increase in v -conotoxin GVIA binding sites (presumably

N-type VSCCs) has been observed in mouse brains after

chronic morphine treatment (Suematsu et al., 1993). In

another study, v -conotoxin GVIA partially inhibited KCl-

induced calcium in¯ux in synaptosomes prepared from

naive rats but not from morphine-tolerant rats (Welch and

Olson, 1991).

As both ziconotide and morphine inhibit N-type VSCCs

and as both agents may be given intrathecally to produce

analgesia at the same or different times, it is important to

understand possible interactions between morphine and

ziconotide with respect to nociception and tolerance to

analgesia. The speci®c aims of the present study were to

determine whether (1) acute intrathecal administration of

ziconotide and morphine affect each other's antinociceptive

properties, (2) chronic (7-day) intrathecal morphine or zico-

notide infusion alter ziconotide or morphine analgesia,

respectively, and (3) intrathecal ziconotide treatment

prevents or reverses morphine tolerance to analgesia. Preli-

minary ®ndings have been published previously in abstract

form (Wang et al., 1997).

2. Materials and methods

2.1. Animals

Male, Sprague±Dawley rats weighing between 220 and

280 g (Simonson Laboratories, Gilroy, CA) were acclimated

to the laboratory environment for 5±7 days before entering

the study. While in the home cage environment, the animals

were allowed free access to water and a commercial rat diet.

Room temperature was maintained at 20±238C and room

illumination was on a 12/12 h light/dark cycle (07:00/

19:00). All experiments were carried out with the approval

of the Animal Care and Use Committee of Elan Pharmaceu-

ticals.

2.2. Drugs

Ziconotide tri¯uoroacetic acid (Elan Pharmaceuticals,

Menlo Park, CA) and morphine sulfate pentahydrate

(Sigma Chemical Co., St. Louis, MO) were dissolved and

diluted in sterile, preservative free 0.9% NaCl solution

(Phoenix Scienti®c, Inc., St. Joseph, MO). For the acute

study with the formalin test, three test article solutions

were prepared: ziconotide alone, morphine alone, and the

combination of ziconotide and morphine in a ratio of 1:10.

All three test article solutions were then diluted in a ratio of

1:3 or 1:3.33.

2.3. Spinal catheterization procedures

The animals were implanted with polyethylene catheters

(PE-10: 0.23 mm i.d. and 0.61 mm o.d., and with volume of

approximately 7 ml (Clay Adams, Parsippany, NJ)) ®lled

with heparinized (25 IU/ml) saline under halothane anesthe-

sia as described by Yaksh and Rudy (1976). The catheter

measured 8.5 cm in length and terminated at the lumbar

enlargement of the spinal cord. The external end of the

catheter was anchored to the adjacent muscle tissue where

it emerged from the cisterna magna with one stay suture. In

single treatment studies including the bolus injection of test

articles in the paw pressure test and continuous infusion of

test articles in the hot-plate and tail immersion tests (see

Experiment 3.1 and 3.3), compounds were administered

approximately 7 days after catheterization. In the second

experiment of 3.2 where chronic morphine infusion

followed by bolus injection of ziconotide, the catheter was

connected to an osmotic minipump (Alzet, Model 2001,

Alza Co., Palo Alto, CA). The minipump containing test

article solution was implanted subcutaneously between the

scapulae. Seven days after implantation, the minipump was

removed under halothane anesthesia, the catheter was

¯ushed with saline, and used for bolus injection of zicono-

tide 3 h later. In the ®rst experiment of 3.2 where chronic

intrathecal infusion of ziconotide followed by bolus injec-

tion of morphine, a `PE-5' catheter (0.006 `i.d. and 0.14'

o.d., Spectranetics, Colorado Springs, CO) connected with

PE-10 tubing was implanted, together with another PE-10

Y.-X. Wang et al. / Pain 84 (2000) 271±281272

tubing, into the spinal cord as described above for intrathe-

cal bolus injection.

2.4. Formalin test

Saline containing 5% formalin was injected subcuta-

neously into the right dorsal hindpaw. The rat was immedi-

ately placed in a 23 £ 35 £ 19 cm polycarbonate box

positioned in front of a mirror for behavioral observations.

Nociceptive behavior was quanti®ed by counting the

number of paw ¯exions in 1-min epochs. Measurements

were taken at 10-min intervals beginning immediately

after formalin injection and ending 90 min later.

2.5. Paw pressure test

The rat was placed in a cone restrainer that allowed its

feet to hang free. The mechanical nociceptive threshold for

the ¯exion re¯ex elicited by stimulation of the dorsal

surface of the right hindpaw was quanti®ed using an Ugo

Basile analgesymeter (Biological Research Apparatus,

Varese, Italy). This device generates a mechanical force,

with a dome-shaped plunger placing directly on the dorsum

of the rat's hindpaw, that increases linearly with time. The

mechanical nociceptive threshold was de®ned as the force

(in g) at which the rat withdrew its paw. The cut-off force

was 750 g.

2.6. Hot-plate test

The animal was placed on a hot-plate (Hot Plate Analge-

sia Meter, Model 39D, IITC, Inc., Woodland, CA) main-

tained at 52.58C (52.2±52.88C). The thermal nociceptive

threshold was de®ned as the time required to elicit either

a hindpaw lick or a jump. The cut-off time was 60 s.

2.7. Tail immersion test

The animal was placed in a cone restrainer and the end of

the tail (5 cm) was placed in a 508C water bath (49.5±

50.58C). The thermal nociceptive threshold was de®ned as

the time required to elicit a ¯ick of the tail. The cut-off time

was 30 s. As the hot-plate and tail immersion tests were

performed in the same session, the tail immersion test was

followed by the hot-plate test.

2.8. Data analysis

Nociceptive thresholds were expressed as means ^ SEM.

For the hot-plate and tail immersion tests, data were

converted to percent maximum possible effect (% MPE)

by the following formula: % MPE � (post-treatment

value 2 pre-treatment value)/( cut-off value 2 pre-treat-

ment value) £ 100. Antinociceptive ED50 values were calcu-

lated from dose-response curves of morphine, ziconotide, or

the combination of morphine and ziconotide by a program

(Drug Interaction Toolbox: Version 1.32, Tallarida and

McCary, 1998) based on the Graded Dose Response Method

of Tallarida and Murray (1987). The program also calcu-

lated the regression lines for individual dose-response

curves, theoretic additive curves, and combination (mixture)

curves according to the methods described by Tallarida et

al. (1997). Interactions between morphine and ziconotide

were analyzed by comparing the composite additive line

vs. the mixture regression line (F-test) and the theoretical

additive ED50 vs. the mixture ED50 (t-test). As this drug

interaction analysis approach is relatively new, the drug

interaction was also analyzed by the conventional isobolo-

graphical analysis (Tallarida et al., 1989). Statistical signif-

icance was also determined by repeated measures analysis

of variance (ANOVA) followed by post-hoc two-tailed t-

tests, and one-factor ANOVA via using a statistical analysis

program (StatView, Abacus Concepts, Berkeley, CA). P #0:05 was considered statistically signi®cant.

3. Results

3.1. Acute interaction of ziconotide and morphine in the

formalin test

To study the acute interactions of ziconotide and

morphine, dose-response curves for ziconotide, morphine,

and the combination of ziconotide and morphine were

measured. The ®rst ®ve groups of rats (n � 8 in each

group) received intrathecal bolus injections of 10 ml saline

or ziconotide (0.03, 0.1, 0.3, or 1 mg). The second six groups

of rats (n � 8 in each group) received intrathecal bolus

injections of 10 ml saline or morphine (0.1, 0.3, 1, 3 or 10

mg). The last ®ve groups of rats (n � 8 in each group)

received intrathecal bolus injections of 10 ml saline or

four doses of the combination of ziconotide and morphine

in a ®xed dose ratio of 1:10. The dose combinations were as

follows: 0.01 mg ziconotide 1 0:1 mg morphine, 0.03 mg

ziconotide 1 0:3 mg morphine, 0.1 mg ziconotide 1 1 mg

morphine, and 0.3 mg ziconotide 1 3 mg morphine. All

rats were treated 10 min before initiating the formalin test.

Subcutaneous injection of formalin in control rats receiv-

ing intrathecal saline produced a characteristic bi-phasic

¯inch response consisting of an initial, rapidly decaying

acute phase (within 10 min after formalin injection)

followed by a slowly rising and long-lived (10±90 min)

tonic phase. Intrathecal ziconotide up to 1 mg did not inhibit

¯inch responses in the acute phase, but produced a dose-

dependent inhibition of formalin-induced ¯inch responses in

the tonic phase (Fig. 1A). The areas under the ¯inch

response curve from 10 to 90 min (AUC10±90s) were calcu-

lated and plotted in Fig. 2A. The ED50 for ziconotide analge-

sia was 0:11 ^ 0:5 mg. Intrathecal bolus injection of

morphine in doses up to 10 mg did not signi®cantly inhibit

formalin-induced ¯inch responses in the acute phase, but

blocked the tonic phase responses in a dose-dependent

manner (Fig. 1B). The AUC10±90s were calculated and

plotted in Fig. 2A. The ED50 for morphine analgesia was

Y.-X. Wang et al. / Pain 84 (2000) 271±281 273

1:5 ^ 0:4 mg. Possible interactions between ziconotide and

morphine in the formalin test (the tonic phase) were

analyzed by comparing the composite additive and mixture

regression lines. The analysis showed that the theoretic

additive ED50 for ziconotide and morphine in the ®xed

ratio of 1:10 was 0:67 ^ 0:16 mg. The regression line calcu-

lated for the theoretic additive equivalent was:

y � 191:4 2 50x.

Intrathecal injections of the combination of ziconotide

and morphine blocked acute phase ¯inch responses, with

the combinations of 1 mg morphine 1 0:1 mg ziconotide

and 3 mg morphine 1 0:3 mg ziconotide signi®cantly differ-

ent from control (Fig. 1C). Considering that 1 and 3 mg

morphine did not signi®cantly suppress the acute phase

¯inch responses (see Fig. 1B), the results indicated that

ziconotide potentiated morphine analgesia in the acute

phase. The combination of ziconotide and morphine also

produced a dose-dependent inhibition of formalin-induced

¯inch responses in the tonic phase (Fig. 1C). The AUC10±90s

were calculated and plotted in Fig. 2A. The analgesic ED50

for the mixture of morphine and ziconotide was 0:70 ^ 0:22

mg, which was not signi®cantly different from the above

theoretic additive ED50 (t-test, P . 0:05). The regression

line for the morphine and ziconotide mixture dose-response

curve was: y � 234:6 2 64:8x, which was also not signi®-

cantly different from the above theoretic additive equivalent

(F-test, P . 0:05). A conventionally used isobolographical

analysis showed that the experimental mixture ED50 over-

lapped the theoretic additive ED50 (95% con®dent limit)

(Fig. 2). This con®rms that the level of analgesia produced

by spinal morphine and ziconotide together can be

explained by the additive contributions of these agents.

3.2. Acute interaction of ziconotide and morphine in the paw

pressure test

Ziconotide and morphine interactions were also exam-

ined in the paw pressure test. Five groups of rats (n � 8±9

in each group) received an intrathecal bolus injection of

saline (10 ml) or ziconotide (0.03, 0.1, 0.3 or 1 mg). Paw

withdrawal thresholds were measured prior to and 0.5, 1, 2,

4, 8, and 24 h post-injection. Pretreatment baseline paw

withdrawal thresholds were not signi®cantly different

Y.-X. Wang et al. / Pain 84 (2000) 271±281274

Fig. 2. Interaction between intrathecal bolus injections of morphine and

ziconotide in the rat hindpaw formalin test. (A) Dose-response curves for

morphine, ziconotide, and the combination of morphine and ziconotide,

n � 8 in each dose group. Data (means ^ SEM) are areas under 10±90

min ¯inch response curve (AUC10±90) calculated from the Fig. 1±3, respec-

tively. (B) Isobologram obtained from ED50s calculated from (A). Solid

circles on the abscissa and the ordinate represent ED50s for morphine and

ziconotide, respectively, when administered alone. Bars crossing the axes

indicate their 95% con®dence limits. The solid line connecting the separate

drug ED50s represents the additive interaction line. The open triangle repre-

sents ED50s obtained when morphine and ziconotide were co-administered.

The crosses through this point represent the 95% con®dence intervals for

the ED50s. The open circle drawn on the additive interaction line represents

the calculated theoretical additive ED50s and the crosses through this point

represent the 95% con®dence intervals for those ED50s. The overlap of the

experimental values with the theoretic additive values indicate an additive

relationship between morphine and ziconotide.

Fig. 1. Dose-response curves for intrathecal bolus injections of ziconotide

(A), morphine (B), and the combination of ziconotide and morphine (C) on

¯inch responses in the formalin test in rats (n � 8 in each group). 5%

formalin (50 ml) was subcutaneously injected 10 min after intrathecal

bolus injection. Data are means ^ SEM.

from among the ®ve groups of rats. Saline did not signi®-

cantly change paw withdrawal thresholds. Ziconotide

caused a long-lived and reversible inhibition of pressure-

induced paw withdrawal responses; the peak effect was

observed 1 h post-injection and was maintained for at

least 8 h before returning to pre-treatment levels within 24

h (Fig. 3A). The antinociceptive effect of ziconotide was

dose-dependent with an ED50 of 0.6 mg (95% con®dence

interval: 0.3±1.3 mg). Fig. 3B shows the dose-response

curve when the paw withdrawal thresholds 1 h post injection

were plotted.

Six groups of rats (n � 8 in each group except for 20 mg

morphine group where n � 4) received an intrathecal bolus

injection of saline (10 ml), morphine (0.5, 2 and 20 mg), and

the combinations of ziconotide (0.1 mg) and morphine (0.5

and 20 mg). Paw withdrawal thresholds were measured prior

to and 0.25, 0.5, 1, 2, and 4 h post-injection. Pretreatment

baseline paw withdrawal thresholds were not signi®cantly

different among the six groups of rats. Morphine increased

the mechanical thresholds in a time-dependent manner.

Ziconotide potentiated the antinociceptive effects of

morphine at both 0.5 and 2 mg (Fig. 4A). Fig. 4B shows

the dose-response curves for morphine in the absence or

presence of ziconotide when mechanical nociceptive thresh-

olds 1 h post injection were plotted. Morphine produced

dose-dependent analgesia. Ziconotide signi®cantly shifted

the analgesic dose-response curve for morphine to the left.

The ED50s for morphine analgesia in the absence and

presence of ziconotide (0.1 mg, which by itself did not

signi®cantly increased mechanical threshold, see Fig. 3)

were 1.3 (95% con®dence interval: 0.5±3.9 mg) and 0.2

mg (95% con®dence interval: 0.02±0.3 mg), respectively,

and were signi®cantly different from each other (P , 0:05).

3.3. Effects of chronic intrathecal ziconotide infusion on

morphine analgesia in the formalin test

To test whether chronic (7-day) intrathecal ziconotide

affected morphine analgesia in the formalin test, 44 rats

were in four groups (n � 10±12 in each group) and each

group received two treatments. They were (1) saline 1saline; (2) ziconotide 1 saline; (3) saline 1 morphine; and

(4) ziconotide 1 morphine. The ®rst treatment was contin-

uous, constant-rate, intrathecal infusion of saline (1 ml/h) or

ziconotide (0.1 mg/h) for 7 days; the second treatment was

intrathecal bolus injection of saline (10 ml) or morphine (1

mg) using different sets of intrathecal tubings after 7-days of

Y.-X. Wang et al. / Pain 84 (2000) 271±281 275

Fig. 3. Time-course (A) and dose-response curve (B) of the antinociceptive

effect of ziconotide (0.03±1 mg) administered intrathecally in the paw

pressure test in rats (n � 8±9 in each group). Data are means ^ SEM.

Data in B were pressure nociceptive thresholds at 1 h after intrathecal

bolus injection from A. ED50 of ziconotide analgesia was 0.6 mg (95%

con®dence interval: 0.3±1.3 mg).

Fig. 4. Time-course (A) and dose-response curve (B) of the antinociceptive

effect of morphine (0.5±20 mg) administered intrathecally, with or without

ziconotide, (0.1 mg) in the paw pressure test in rats (n � 8±9 in each group

except for 20 mg morphine where n � 4). Data are means ^ SEM. Data in

B were pressure nociceptive thresholds at 1 h after intrathecal bolus injec-

tion from A. ED50s of morphine analgesia in the absence and presence of

ziconotide were 1.3 mg (95% con®dence interval: 0.5±3.9 mg) and 0.2 mg

(95% con®dence interval: 0.02±0.3 mg), respectively.

the ®rst treatment. After 10 min 5% formalin (50 ml) was

subcutaneously injected into the right hindpaw.

Neither treatment (chronic ziconotide infusion, bolus

morphine injection, or the combination of bolus morphine

injection and chronic ziconotide infusion) altered formalin-

induced ¯inch responses in the acute phase. However,

chronic intrathecal ziconotide infusion and intrathecal

morphine injection inhibited formalin-induced ¯inch

responses in the tonic phase, chronic intrathecal ziconotide

infusion signi®cantly increased morphine analgesia in the

tonic phase (Fig. 5A). Fig. 5B displays the results using the

AUC10±90s for the four treatment groups. The results indi-

cated that after chronic intrathecal infusion, ziconotide

enhances morphine analgesia presumably in an additive

manner.

3.4. Effects of chronic intrathecal morphine infusion on

ziconotide analgesia in the formalin test

To test whether chronic intrathecal morphine infusion

affected ziconotide analgesia in the formalin test, 50 rats

were assigned into 6 groups (n � 8±9 in each group) and

each group received two treatments. They were (1) saline 1saline; (2) saline 1 morphine; (3) morphine 1 morphine;

(4) saline 1 ziconotide; (5) morphine 1 ziconotide; and

(6) morphine 1 the combination of morphine 1 ziconotide.

The ®rst treatment was continuous, constant-rate, intrathe-

cal infusion of saline (1 ml/h) or morphine (15 mg/h) for 7

days. After 7 days of infusion, the rats were anesthetized

under halothane, the infusion pumps were removed, and the

intrathecal tubings were ¯ushed with saline followed by the

second treatment 3 h later. The second treatment was

intrathecal bolus injection of saline (10 ml), morphine (20

mg), ziconotide (0.03 mg), or the combination of morphine

(20 mg) 1 ziconotide (0.03 mg). Ten min later the rats

received a subcutaneous injection of formalin.

Intrathecal bolus injections of morphine and ziconotide

suppressed tonic phase ¯inch responses in rats that received

a 7-day saline infusion. However, intrathecal morphine was

not effective in blocking ¯inch responses in rats that

received 7-day morphine infusion. In contrast, intrathecal

ziconotide produced inhibition of ¯inch responses in the

tonic phase in rats that received a 7-day morphine infusion,

which was comparable to that observed in rats that received

a chronic saline infusion (P . 0:05). Moreover, co-admin-

istration of morphine and ziconotide in rats that received

chronic morphine infusion did not produce a further

decrease in ¯inch responses compared to that produced by

ziconotide alone (Fig. 6A). Fig. 6B shows the results using

the AUC10±90s after formalin administration for the six treat-

ment groups. The results indicate that (1) chronic intrathecal

morphine infusion produces tolerance to morphine analge-

sia, but does not produce cross-tolerance to ziconotide

analgesia, and (2) ziconotide does not reverse morphine

tolerance to analgesia.

3.5. Effects of co-administration of intrathecal ziconotide

with morphine on morphine tolerance to analgesia in the

hot-plate and tail immersion tests

To test whether concurrent infusion of intrathecal zicono-

tide with morphine prevented morphine tolerance to analge-

sia, 28 rats in four groups (n � 7 in each group) received

continuous, constant-rate intrathecal infusion of saline (1

ml/h), ziconotide (0.03 mg/h), morphine (15 mg/h), or the

combination of ziconotide (0.03 mg/h) and morphine (15

mg/h) for 7 days. Nociceptive responses to noxious heat

stimuli were ®rst measured in the tail immersion test and

then the hot-plate test. Measurements were made once daily

at approximately 09:00 between 1 day prior to and 7 days

after the start of infusions except during the ®rst two days of

infusion when measurements were made twice daily at

approximately 09:00 and 18:00, respectively. Pretreatment

thermal nociceptive thresholds in the hot-plate and tail

immersion tests were not statistically different among the

four groups.

As shown in Fig. 7A, thermal nociceptive thresholds in

Y.-X. Wang et al. / Pain 84 (2000) 271±281276

Fig. 5. (A) Effects of intrathecal bolus injection of saline (10 ml) or

morphine (1 mg) on chemically-evoked nociceptive responses (¯inches)

in the formalin test in rats (n � 10±12 in each group) after 7 days of

continuous, constant-rate, intrathecal infusions of saline (1 ml/h) or zicono-

tide (0.1 mg/h). The ®rst treatment was intrathecal infusion of saline (1 ml/

h) or ziconotide (0.1 mg/h) for 7 days; the second treatment was intrathecal

bolus injection of saline (10 ml) or morphine (1 mg) using different intrathe-

cal tubings 10 min before formalin challenge. Data are means ^ SEM. (B)

Summed results in the tonic phase of the formalin test. AUC10±90s were

calculated from data from the tonic phase presented in (A). aDenotes statis-

tically signi®cantly difference from saline 1 saline control (P , 0:05);bDenotes statistically signi®cant difference from the combination of zico-

notide and morphine (P , 0:05).

the saline-treated rats remained stable throughout the 7-day

infusion period in the hot-plate test. Intrathecal ziconotide

infusion produced small but signi®cant increases in thermal

nociceptive thresholds; this antinociceptive effect (22 ^ 1%

MPE) was observed on the ®rst day after initiating zicono-

tide infusion and was maintained over the 7-day infusion

period. Intrathecal morphine infusion (15 mg/h) was moder-

ately antinociceptive (27 ^ 10% MPE) during the ®rst day

of infusion. However, thermal nociceptive thresholds

declined to control levels after 2 days of treatment (biolo-

gical t1=2 < 1 day). Intrathecal infusion of ziconotide and

morphine produced marked antinociceptive effects

(91 ^ 9% MPE) during the ®rst day of treatment. Thermal

nociceptive thresholds thereafter declined until the antino-

ciceptive effects of ziconotide and morphine were similar to

those produced by ziconotide alone after 4 days of treat-

ment.

The results in the tail immersion test were similar to those

obtained in the hot-plate test (Fig. 7B). Ziconotide adminis-

tered at a rate of 0.03 mg/h had no effect on thermal noci-

ceptive thresholds. Morphine increased thermal nociceptive

thresholds to 28 ^ 7% MPE observed on the ®rst day during

morphine infusion but tolerance developed after approxi-

mately 4 days of treatment (biological t1=2 < 3 days).

Intrathecal infusion of ziconotide and morphine also

produced marked antinociception (91 ^ 3% MPE) during

the ®rst day of infusion. Thereafter, thermal nociceptive

thresholds declined towards control values. Thermal noci-

ceptive thresholds of animals that received ziconotide and

morphine in combination were comparable to those of

animals that received saline or ziconotide alone after 5

days of treatment.

Fig. 8 shows a further analysis of the time-effect curves

obtained in the above study. Open circles represent the

effects of ziconotide and morphine administered in combi-

nation whereas the solid circles represent the sum of the

effects of ziconotide and morphine when each was adminis-

tered alone. As shown, the effects of co-administrating

morphine and ziconotide in combination greatly exceeded

the sum of those produced by either treatment alone during

the ®rst 4±5 day infusion period and the two effects were

comparable after that period. These results suggest that

ziconotide and morphine interact apparent synergistically

in the hot-plate and tail immersion tests in contrast to the

additive effect observed in the formalin test (the tonic

phase). The results also demonstrate that the co-administra-

Y.-X. Wang et al. / Pain 84 (2000) 271±281 277

Fig. 6. (A) Effects of intrathecal (spinal) bolus injections of saline (10 ml)

or morphine (20 mg), ziconotide (0.03 mg), or the combination of zicono-

tide (0:03 m g�1 morphine (20 mg) on chemically-induced ¯inch

responses in the formalin test in rats (n � 8±9 in each group) after 7 days

of continuous, constant-rate, intrathecal infusions of saline (1 ml/h) or

morphine (15 mg/h). The second treatment was done 3 h after the infusion

pumps were removed and the intrathecal tubings were ¯ushed with saline.

Shown are means and SEM. (B) Summed ¯inch counts in the tonic phase.

AUC10±90s were calculated from data of tonic phase presented in (A).aDenotes statistically signi®cantly difference from saline 1 saline control

(P , 0:05); bDenotes statistically signi®cant difference from saline 1

morphine group (P , 0:05).

Fig. 7. Effects of intrathecal infusions of saline (1 ml/h), ziconotide (0.03

mg/h), morphine (15 mg/h) or a combination of ziconotide (0.03 mg/h) and

morphine (15 mg/h) on thermally-evoked (heat) nociceptive responses in

the hot-plate test (A) and the tail immersion test (B) in rats (n � 7 in each

group). Test/control articles were administered by continuous, constant-rate

infusion for seven days. Data are means ^ SEM.

tion of intrathecal ziconotide does not prevent morphine

tolerance to analgesia.

4. Discussion

Acute intrathecal administration of ziconotide or v -cono-

toxin MVIIA has been reported to block mechanical allo-

dynia (Chaplan et al., 1994; Bowersox et al., 1996), cold

allodynia (Wang et al., 1998; unpublished data), heat hyper-

algesia (Yamamoto and Sakashita, 1998) in neuropathic

rats, and persistent pain (tonic phase) in the formalin test

(Malmberg and Yaksh, 1994, 1995; Bowersox et al., 1996).

The present study con®rms that ziconotide is a potent

blocker of persistent pain as measured by tonic phase ¯inch

responses in the formalin test. Our data also demonstrate

that ziconotide produces mild analgesia in acute (nocicep-

tive) pain. (1) Our extensive experience show that intrathe-

cal ziconotide up to 1 mg, which completely blocks

persistent pain in the tonic phase, does not block the acute

phase ¯inch responses (Bowersox et al., 1996; present

study). This contrasts with an earlier study that ziconotide

blocked acute phase responses in the formalin test (Malm-

berg and Yaksh, 1994). (2) Ziconotide has mild analgesic

effects on thermal (heat) nociception in the tail immersion

and hot-plate tests (Malmberg and Yaksh, 1995; present

study). (3) Ziconotide increases the mechanical thresholds

in the paw pressure test in a dose-dependent manner. This is

consistent with recent ®ndings that v -conotoxin GVIA, a

selective but irreversible blocker of N-type VSCCs

increased mechanical nociceptive thresholds in the paw

pressure test (Omote et al., 1996). However, the analgesia

produced in the paw pressure test is less potent (ED50 � 0:6

mg) than observed in neuropathic pain and persistent pain

(ED50s � 0:1 mg) with ziconotide (Bowersox et al., 1996;

present study). Neuropathic pain and persistent pain are

thought to be mediated by injury-induced spinal sensitiza-

tion of dorsal horn neurons (i.e. central sensitization)

(Woolf, 1983; Woolf et al., 1994; Sotgiu et al., 1998). Our

results suggest that ziconotide may interfere with central

sensitization and subsequent neuropathic pain and persistent

pain. In contrast, morphine produces little or no antinocicep-

tion in rat models of painful neuropathy (Courteix et al.,

1994; Mao et al., 1995; Nichols et al., 1995). Results from

human studies are consistent with these ®ndings with

morphine (Arner and Meyerson, 1988).

The major ®ndings in the present study are that (1) acute

intrathecal ziconotide and morphine do not interfere each

other's analgesia and produce either additive or apparent

synergistic antinociception depending on the nociceptive

test employed; (2) chronic intrathecal morphine administra-

tion causes tolerance to analgesia but does not produce

cross-tolerance to ziconotide antinociception; (3) chronic

intrathecal ziconotide administration does not produce

either tolerance to analgesia or cross-tolerance to morphine

analgesia; (4) the intrathecal co-administration of zicono-

tide neither prevents nor reverses morphine tolerance to

analgesia.

4.1. Acute spinal interaction between ziconotide and

morphine

Activation of m-opioid receptors by morphine and other

opioids leads to inhibition of N-type calcium channels via a

G-protein (Seward et al., 1991; Bourinet et al., 1996; Toth et

al., 1996; Zhang et al., 1996; De Waard et al., 1997) by

changing channel gating (Lipscombe et al., 1989) and alter-

ing ion permeation (Kuo and Bean, 1993). Since both zico-

notide and morphine inhibit N-type calcium channel

activity, it is necessary to think that they may act synergis-

tically to produce analgesia. Indeed, our results suggest

there is a synergistic analgesic effect between morphine

and ziconotide during the early phase after infusion in the

hot-plate and tail immersion tests or after acute administra-

tion in the formalin and paw pressure tests. (1) At doses that

do not inhibit acute ¯inch responses in the formalin test,

ziconotide signi®cantly potentiates morphine analgesia

(see Fig. 1). (2) Ziconotide and morphine potentiate each

other's analgesia in the initial infusion phase of the hot-plate

and tail immersion tests, although the dose-response analy-

sis is not applied and there is an unlikely possibility that the

markedly increased responses produced by the combination

ziconotide and morphine are due to steep slopes (high Hill

coef®cients) of the dose-response curves for morphine and/

or ziconotide alone. (3) At a dose that does not produce

Y.-X. Wang et al. / Pain 84 (2000) 271±281278

Fig. 8. Thermal nociceptive responses recorded after co-administration of

ziconotide (0.03 mg/h) and morphine (15 mg/h) compared to the sum of the

values for ziconotide (0.03 mg/h) and morphine (15 mg/h) administered

alone in the hot-plate test (A) and tail immersion test (B) in rats with raw

data from Fig. 7.

signi®cant analgesia by itself, ziconotide potentiates

morphine antinociception in the paw pressure test. These

synergistic effects on thermal and mechanical nociception

are supported by recent ®ndings that v -conotoxin GVIA,

another blocker of N-type VSCCs, has a naloxone-reversi-

ble synergistic effect with morphine in the tail-¯ick and paw

pressure tests by using isobolographical analysis (Omote et

al., 1996), and that v -conotoxin GVIA left-shifts morphine

analgesia dose-response curve in the tail-¯ick test (Wei et

al., 1996).

In contrast, there is an additive rather than synergistic

relationship between ziconotide and morphine analgesia in

the tonic phase of the formalin test. This result was obtained

using to different methods of drug interaction analysis: i.e.

comparing the composite additive and mixture regression

line approach and isobolographical analysis approach. Simi-

lar results were obtained with spinal ziconotide and the

selective a 2-adrenoceptor agonist clonidine, or the selective

GABA-B receptor agonist baclofen, in the tonic phase

responses in the formalin test (Wang et al., 1998; unpub-

lished data). The mechanisms for the difference are not

known.

4.2. Intrathecal morphine and ziconotide do not produce

cross-tolerance to each other's analgesia

A signi®cant increase of v -conotoxin GVIA binding sites

in brain was observed after chronic morphine treatment

(Suematsu et al., 1993), and v-conotoxin GVIA was

found to inhibit KCl-induced calcium in¯ux in synapto-

somes prepared from naive rats but not from morphine-

tolerant rats (Welch and Olson, 1991). These results suggest

that ziconotide antinociception might be altered in

morphine-tolerant rats. Our results indicate that the antino-

ciceptive effect of was the same in morphine-tolerant rats as

in naive rats. The inability of morphine to produce cross-

tolerance to ziconotide antinociception suggests that the

electrophysiological properties and function of N-type

calcium channels remain the same after morphine tolerance.

Indeed, chronic exposure to morphine reduced morphine-

induced N-type calcium currents but does not change the

electrophysiological properties of N-type calcium channels

(Kennedy and Henderson, 1991, 1992). It has been proposed

that morphine tolerance develops as opioid receptors are

uncoupled from their G-proteins (rather than reduction of

m-opioid receptors) and consequently from their effectors

such as N-type VSCCs (Kennedy and Henderson, 1991,

1992; Tao et al., 1993; see reviews by Basbaum, 1995,

1996). Our results with ziconotide in morphine-tolerant

rats are consistent with this model.

That ziconotide maintains the same antinociception ef®-

cacy in morphine tolerant animals makes it a unique anti-

nociceptive agent. Morphine is known to produce cross-

tolerance to other opioids and many other analgesic agents.

For example, cross-tolerance between intrathecal morphine

and a 2-adrenoceptor agonists have been observed (Solomon

and Gebhart, 1988; Roerig, 1995). Therefore, ziconotide

would not be expected to have reduced analgesic ef®cacy

when used in morphine-tolerant patients. This is an impor-

tant clinical implication as ziconotide is under investigation

for the treatment of chronic pain and many of these patients

have developed morphine tolerance to analgesia or

morphine-resistant (Brose et al., 1997; Presley et al., 1998).

One of the advantages of ziconotide is that long-term

administration does not produce tolerance to antinocicep-

tion in the formalin test (Malmberg and Yaksh, 1995;

Bowersox et al., 1996), hot-plate test (Malmberg and

Yaksh, 1995), and in neuropathic rats (Bowersox et al.,

1996; Wang et al., 1998, unpublished data). The present

study supports these ®ndings by showing that the antinoci-

ceptive response to ziconotide in the hot-plate test is main-

tained during the period of infusion. In addition, ziconotide

potentiates morphine analgesia after 7 days of infusion at

the same level as found with acute treatment. Therefore,

ziconotide produces neither tolerance to analgesia nor

cross-tolerance to morphine analgesia. Previous reports on

the effect of pretreatment with v -conotoxin GVIA on

morphine analgesia are con¯icting. It was reported that

pretreatment (24 h earlier) with intracerebroventricular

administration of v -conotoxin GVIA potentiated morphine

analgesia (Spampinato et al., 1994) but the same treatment

was reported to block morphine analgesia (Basilico et al.,

1992).

4.3. Intrathecal concurrent ziconotide does not prevent or

reverse spinal morphine tolerance to analgesia

Co-intrathecal infusion of ziconotide and morphine

produce synergistic antinociceptive effects in the hot-plate

and tail immersion tests during the initial phase of infusion.

However, this markedly increased effect disappears as the

rats develop tolerance to morphine analgesia. In addition,

co-intrathecal bolus injection of ziconotide and morphine

produce the same antinociception as ziconotide alone in

established morphine-tolerant rats. These results indicate

that ziconotide does not prevent or reverse morphine toler-

ance to analgesia. In contrast, it has been shown that NMDA

receptor antagonists such as MK-801 block morphine toler-

ance to analgesia (Trujillo and Akil, 1991; Gutstein and

Trujillo, 1993) in the spinal cord (Gutstein and Trujillo,

1993). NMDA receptors antagonists are known to block

voltage-dependent calcium currents including N-type

currents (Church et al., 1994; Biton et al., 1994). Our results

with ziconotide suggest that inhibition of spinal N-type

calcium channels is not involved in the effect of NMDA

receptor antagonists on morphine tolerance.

Acknowledgements

The research was supported by funds from Elan Pharma-

ceuticals. We thank Dr Sandra Roerig at Louisiana State

University and Mr Jeffrey McCary at Temple University

Y.-X. Wang et al. / Pain 84 (2000) 271±281 279

for their kind help with drug interaction analysis. We also

thank Dr Bill Hopkins at Elan Pharmaceuticals for his care-

ful review of the manuscript.

References

Abram SE, Yaksh TL. Morphine, but not inhalation anesthesia, blocks post-

injury facilitation. The role of preemptive suppression of afferent trans-

mision. Anesthesiology 1993;78:713±721.

Arner S, Meyerson BA. Lack of analgesic effect of opioids on neuropathic

and idiopathic forms of pain. Pain 1988;33:11±23.

Basbaum AI. Insights into the development of opioid tolerance. Pain

1995;61:349±352.

Basbaum AI. Memories of pain. Sci Med 1996;3:22±31.

Basilico L, Parolaro D, Rubino T, Gori E, Giagnoni G. In¯uence of v-

conotoxin on morphine analgesia and withdrawal syndrome in rats. Eur

J Pharmacol 1992;218:75±81.

Biton B, Granger P, Carreau A, Depoortere H, Scatton B, Avenet P. The

NMDA receptor antagonist eliprodil (SL 82.0715) blocks voltage-oper-

ated Ca21 channels in rat cultured cortical neurons. Eur J Pharmacol

1994;257:297±301.

Bourinet E, Soong TW, Stea A, Snutch TP. Determinants of the G protein-

dependent opioid modulation of neuronal calcium channels. Proc Natl

Acad Sci USA 1996;93:1486±1491.

Bowersox SS, Gadbois T, Singh T, Pettus M, Wang YX, Luther RR. Selec-

tive N-type neuronal voltage-sensitive calcium channel blocker, zico-

notide, produces spinal antinociception in rat models of acute, persistent

and neuropathic pain. J Pharmacol Exp Ther 1996;279:1243±1249.

Brose WG, Gutlove DP, Luther RR, Bowersox SS, McGuire D. Use of

intrathecal ziconotide, a novel, N-type, voltage-sensitive, calcium chan-

nel blocker, in the management of intractable brachial plexus avulsion

pain. Clin J Pain 1997;13:256±257.

Chaplan SR, Pogrel JW, Yaksh TL. Role of voltage-dependent calcium

channel subtypes in experimental tactile allodynia. J Pharmacol Exp

Ther 1994;269:1117±1123.

Church J, Fletcher EJ, Abdel-Hamid K, MacDonald JF. Loperamide blocks

high-voltage-activated calcium channels and N-methyl-d-aspartate-

evoked responses in rat and mouse cultured hippocampal pyramidal

neurons. Mol Pharmacol 1994;45:747±757.

Courteix C, Bardin M, Chantelauze C, Lavarenne J, Eschalier A. Study of

the sensitivity of the diabetes-induced pain model in rats to a range of

analgesics. Pain 1994;57:153±160.

De Waard M, Liu H, Walker D, Scott VE, Gurnett CA, Campbell KP.

Direct binding of G-protein betagamma complex to voltage-dependent

calcium channels. Nature 1997;385:446±450.

Gaur S, Newcomb R, Rivnay B, Bell JR, Yamashiro D, Ramachandran J,

Miljanich GP. Calcium channel antagonist peptides de®ne several

components of transmitter release in hippocampus. Neuropharmacol-

ogy 1994;33:1211±1219.

Gutstein HB, Trujillo KA. MK-801 inhibits the development of morphine

tolerance at spinal sites. Brain Res 1993;626:332±334.

Kennedy C, Henderson G. Mu-opioid receptor inhibition of calcium

current: development of homologous tolerance in single SH-SY5Y

cells after chronic exposure to morphine in vitro. Mol Pharmacol

1991;40:1000±1005.

Kennedy C, Henderson G. Chronic exposure to morphine does not induce

dependence at the level of the calcium channel current in human SH-

SY5Y cells. Neuroscience 1992;49:937±944.

Kristipati R, Nadasdi L, Tarczy-Hornoch K, Lau K, Miljanich GP, Rama-

chandran J, Bell JR. Characterization of the binding of omega-conopep-

tides to different classes of non-L-type neuronal calcium channels. Mol

Cell Neurosci 1994;5:219±228.

Kuo CC, Bean BP. G±protein modulation of ion permeation through N±

type calcium channels. Nature 1993;365:258±262.

Lipscombe D, Kongsamut S, Tsien RW. Alpha-adrenergic inhibition of

sympathetic neurotransmitter release mediated by modulation of N-

type calcium-channel gating. Nature 1989;340:639±642.

Malmberg AB, Yaksh TL. Voltage-sensitive calcium channels in spinal

nociceptive processing: blockade of N- and P-type channels inhibits

formalin-induced nociception. J Neurosci 1994;14:4882±4890.

Malmberg AB, Yaksh TL. Effect of continuous intrathecal infusion of

omega-conopeptides N-type calcium-channel blockers, on behavior

and antinociception in the formalin and hot-plate tests in rats. Pain

1995;60:83±90.

Mao J, Price DD, Mayer DJ. Experimental mononeuropathy reduces the

antinociceptive effects of morphine: implications for common intracel-

lular mechanisms involved in morphine tolerance and neuropathic pain.

Pain 1995;61:353±364.

Miljanich GP, Ramachandran J. Antagonists of neuronal calcium channels:

structure, function, and therapeutic implications. Annu Rev Pharmacol

Toxicol 1995;35:707±734.

Nichols ML, Bian D, Ossipov MH, Lai J, Porreca F. Regulation of

morphine antiallodynic ef®cacy by cholecystokinin in a model of neuro-

pathic pain in rats. J Pharmacol Exp Ther 1995;275:1339±1345.

Olivera BM, Cruz LJ, de Santos V, LeCheminant GW, Grif®n D, Zeikus R,

McIntosh JM, Galyean R, Varga J, et al. Neuronal calcium channel

antagonists. Discrimination between calcium channel subtypes using

omega-conotoxin from Conus magus venom. Biochemistry

1987;26:2086±2090.

Olivera BM, Miljanich GP, Ramachandran J, Adams ME. Calcium channel

diversity and neurotransmitter release: the v-conotoxins and v-agatox-

ins. Annu Rev Biochem 1994;63:823±867.

Omote K, Kawamata M, Satoh O, Iwasaki H. Spinal antinociceptive action

of an N-type voltage-dependent calcium channel blocker and the syner-

gistic interaction with morphine. Anesthesiology 1996;84:636±643.

Presley R, Charapata S, Ferrar-Brechner T, Yearwood T, Staats P, Wallace

MS, Ordia J, Gaeta R, Follett KA. Chronic, opioid-resistant, neuro-

pathic pain: marked analgesic ef®cacy of intrathecal ziconotide. Am

Pain Soc Abstr. 1998;17:697.

Roerig SC. Decreased spinal morphine/clonidine antinociceptive syner-

gism in morphine-tolerant mice. Life Sci. 1995;56:PL115±PL122.

Seward E, Hammond C, Henderson G. Mu-opioid-receptor-mediated inhi-

bition of the N-type calcium-channel current. Proc R Soc Lond B Biol

Sci 1991;244:129±135.

Solomon RE, Gebhart GF. Intrathecal morphine and clonidine: antinoci-

ceptive tolerance and cross-tolerance and effects on blood pressure. J

Pharmacol Exp Ther 1988;245:444±454.

Sotgiu ML, Biella G, Firmi L, Pasqualucci V. Topical axonal transport

blocker vincristine prevents nerve injury-induced spinal neuron sensi-

tization in rats. Neurotrauma 1998;15:1077±1082.

Spampinato S, Speroni E, Govoni P, Pistacchio E, Romagnoli C, Murari G,

Ferri S. Effect of omega-conotoxin and verapamil on antinociceptive,

behavioural and thermoregulatory responses to opioids in the rat. Eur J

Pharmacol 1994;254:229±238.

Suematsu M, Ohnishi T, Shinno E, Maeda S, Matsumoto K, Sakuda M,

Saito K. Effect of prolonged administration of clonidine on [3H]PN 200-

110 and [125I]omega conotoxin binding in mouse brain. Neurosci Lett

1993;163:193±196.

Tallarida RJ, Murray RB. Manual of pharmacologic calculations, New

York: Springer, 1987. pp. 26±31.

Tallarida RJ, Porreca F, Cowan A. Statistical analysis of drug-drug and site-

site interactions with isobolograms. Life Sci 1989;45:947±961.

Tallarida RJ, Stone Jr DJ, Raffa RB. Ef®cient designs for studying syner-

gistic drug combinations. Life Sci 1997;61:417±425.

Tao PL, Lee CR, Law PY, Loh HH. The interaction of the mu-opioid

receptor and G protein is altered after chronic morphine treatment in

rats. Naunyn-Schmiedebergs Arch Pharmacol 1993;348:504±508.

Toth PT, Shekter LR, Ma GH, Philipson LH, Miller RJ. Selective G-protein

regulation of neuronal calcium channels. J Neurosci 1996;16:4617±

4624.

Trujillo KA, Akil H. Inhibition of morphine tolerance and dependence by

the NMDA receptor antagonist MK-801. Science 1991;251:85±87.

Y.-X. Wang et al. / Pain 84 (2000) 271±281280

Wang YX, Pettus M,Luther RR, Bowersox SS.Effects of the selective N-type

calcium channel blocker, ziconotide, on the development of tolerance to

spinal morphine analgesia in rats. Neurosci Soc Abstr 1997;23:1121.

Wang YX, Bezprozvannaya S, Bowersox SS, Nadasdi L, Miljanich G,

Mezo G, Silva D, Tarczy-Hornoch K, Luther RR. Peripheral versus

central potencies of N-type voltage-sensitive calcium channel blockers.

Naunyn-Schmiedebergs Arch Pharmacol 1998;357:159±168.

Wei ZY, Karim F, Roerig SC. Spinal morphine/clonidine antinociceptive

synergism: involvement of G proteins and N-type voltage-dependent

calcium channels. J Pharmacol Exp Ther 1996;278:1392±1407.

Welch SP, Olson KG. Opiate tolerance-induced modulation of free intra-

cellular calcium in synaptosomes. Life Sci 1991;48:1853±1861.

White DM, Cousins MJ. Effect of subcutaneous administration of calcium

channel blockers on nerve injury-induced hyperalgesia. Brain Res

1998;801:50±58.

Woolf CJ, Shortland P, Sivilotti LG. Sensitization of high mechanothres-

hold super®cial dorsal horn and ¯exor motor neurones following

chemosensitive primary afferent activation. Pain 1994;58:141±155.

Woolf CJ. Evidence for a central component of post-injury pain hypersen-

sitivity. Nature 1983;306:686±688.

Xiao WH, Bennett GJ. Synthetic omega-conopeptides applied to the site of

nerve injury suppress neuropathic pains in rats. J Pharmacol Exp Ther

1995;274:666±672.

Yaksh TL, Rudy TA. Chronic cauterization of the spinal subarachnoid

space. Physiol Behav 1976;17:1031±1036.

Yamamoto T, Sakashita Y. Differential effects of intrathecally administered

N- and P-type voltage-sensitive calcium channel blockers upon two

models of experimental mononeuropathy in the rat. Brain Res

1998;794:329±332.

Yamamoto T, Sakashita Y. Differential effects of intrathecally administered

N- and P-type voltage-sensitive calcium channel blockers upon two

models of experimental mononeuropathy in the rat. Brain Res

1998;794:329±332.

Zhang JF, Ellinor PT, Aldrich RW, Tsien RW. Multiple structural elements

in voltage-dependent Ca21 channels support their inhibition by G

proteins. Neuron 1996;17:991±1003.

Y.-X. Wang et al. / Pain 84 (2000) 271±281 281