The Journal of Pain, Vol 7, No 7 (July), 2006: pp 500-512

Available online at www.sciencedirect.com

Bee Venom Injection Significantly Reduces Nociceptive Behavior

in the Mouse Formalin Test via Capsaicin-Insensitive Afferents
Dae-Hyun Roh,* Hyun-Woo Kim,* Seo-Yeon Yoon,* Seuk-Yun Kang,*
Young-Bae Kwon,† Kwang-Hyun Cho,‡ Ho-Jae Han,§ Yeon-Hee Ryu,储 Sun-Mi Choi,储
Hye-Jung Lee,¶ Alvin J. Beitz,# and Jang-Hern Lee*
*Department of Veterinary Physiology, College of Veterinary Medicine and School of Agricultural Biotechnology,
Seoul National University, Seoul, South Korea.
†
Department of Pharmacology, and
‡
Department of Psychiatry, Institute for Medical Science, Chonbuk National University Medical School, Jeonju,
South Korea.
§
Hormone Research Center and College of Veterinary Medicine, Chonnam National University, Gwangju, South
Korea.
储
Department of Medical Research, Korea Institute of Oriental Medicine, Daejeon, South Korea.
¶
Department of Acupuncture and Moxibustion, College of Oriental Medicine, Kyunghee University, Seoul, South
Korea.
#
Department of Veterinary and Biomedical Sciences, College of Veterinary Medicine, University of Minnesota, St
Paul, Minnesota.

Abstract: Peripheral bee venom (BV) administration produces 2 contrasting effects, nociception and
antinociception. This study was designed to evaluate whether the initial nociceptive effect induced by
BV injection into the Zusanli acupoint is involved in producing the more prolonged antinociceptive
effect observed in the mouse formalin test, and whether capsaicin-sensitive primary afferents are
involved in these effects. BV injection into the Zusanli point increased spinal Fos expression but not
spontaneous nociceptive behavior. BV pretreatment 10 minutes before intraplantar formalin injection
dose-dependently attenuated nociceptive behavior associated with the second phase of the formalin
test. The destruction of capsaicin-sensitive primary afferents by resiniferatoxin (RTX) pretreatment
selectively decreased BV-induced spinal Fos expression but did not affect BV-induced antinociception.
Furthermore, BV injection increased Fos expression in tyrosine hydroxylase immunoreactive neurons
in the locus caeruleus, and this expression was unaltered by RTX pretreatment. Finally, BV’s antino-
ciception was blocked by intrathecal injection of 10 ␮g idazoxan, and this effect was not modified by
RTX pretreatment. These findings suggest that subcutaneous BV stimulation of the Zusanli point
activates central catecholaminergic neurons via capsaicin-insensitive afferent fibers without induc-
tion of nociceptive behavior. This in turn leads to the activation of spinal ␣2-adrenoceptors, which
ultimately reduces formalin-evoked nociceptive behaviors.
Perspective: This study demonstrates that BV acupuncture produces a significant antinociception with-
out nociceptive behavior in rodents, which is mediated by capsaicin-insensitive afferents and involves
activation of central adrenergic circuits. These results further suggest that BV stimulation into this acu-
puncture point might be a valuable alternative to traditional electrical or mechanical acupoint stimulation.
© 2006 by the American Pain Society
Key words: Bee venom, capsaicin-sensitive primary afferents, formalin test, fos, resiniferatoxin.

Received May 10, 2005; Revised February 3, 2006; Accepted February 4,
2006.
Supported by a grant (M103KV010009 03K2201 00940) from the Brain
Research Center of the 21st Century Frontier Research Program funded by
the Ministry of Science and Technology of the Republic of Korea and by
SRC program of KOSEF (R11-2005-014) as well as the Brain Korea 21
project.
Address requests for reprints to Jang-Hern Lee, DVM, PhD, Department of
Veterinary Physiology, College of Veterinary Medicine, Seoul National
University, Seoul 151-742, South Korea. E-mail: JHL1101@snu.ac.kr
1526-5900/$32.00
© 2006 by the American Pain Society
doi:10.1016/j.jpain.2006.02.002

500
ORIGINAL REPORT/Roh et al
B
ee venom (BV) injection can produce both an initial
nociceptive effect and a prolonged antinociceptive
effect. BV contains a number of potential pain-pro-
ducing substances including melittin, histamine, and
phospholipase A2, and therefore it is not surprising that
several reports described a nociceptive effect after in-
traplantar injection.3,18,20 Luo et al23 have also reported
that intraplantar BV injection significantly increases Fos
expression in the spinal cord dorsal horn of anesthetized
rats. In contrast, subcutaneous injection of diluted BV
into an acupoint, termed apipuncture, has been used
clinically in Oriental medicine to produce a potent anal-
gesic effect. In support of this alternative medicine ap-
proach recent experimental studies in our laboratories
have demonstrated that subcutaneous injection of BV
(0.01 to 1 mg/kg) into the Zusanli acupuncture point pro-
duces prominent antinociceptive and antihyperalgesic
effects in animal models of acute and persistent pain,
respectively.15,17,18,30 The above studies indicate that a
dichotomy exists with respect to the physiologic re-
sponse to subcutaneous injection of BV. On one hand,
intraplantar injection of BV and its major constituent,
melittin, produces robust nociceptive behaviors and hy-
persensitivity in rodents, whereas BV injection into the
Zusanli acupoint, on the other hand, produces little no-
ciceptive behaviors, but rather a significant antinocicep-
tive effect in a variety of animal pain models. Despite the
apparent conflicting data in the literature regarding the
consequences of BV injection, there have been no studies
that have examined a possible relationship between BV’s
nociceptive and antinociceptive effects, particularly with
respect to BV injection into an acupoint. To begin to
examine this issue, we evaluated whether the intensity
of the BV-induced nociception (as measured by sponta-
neous pain behavior) and BV-induced neuronal activa-
tion (as measured by spinal Fos expression) produced by
injection into the Zusanli acupoint is correlated with the
intensity of the BV-induced antinociception (BVAN) in
the mouse formalin test. Because both BV’s nociceptive
and antinociceptive effects appear to involve activation
of primary afferent fibers, we also explored whether pri-
mary afferent axons expressing the vanilloid receptor 1
(TRPV1) were involved in either of these effects.
TRPV1-expressing primary afferent neurons, termed
capsaicin-sensitive primary afferents (CSPAs), have been
recognized as nociceptive polymodal C-fibers whose cell
bodies are located in dorsal root ganglia. Functionally,
CSPAs are known to play a major role in nociceptive
transmission.2,37 Recent studies with a BV-induced pain
model have shown that CSPAs play a critical role in me-
diating both the thermal and mechanical hyperalgesia
induced by BV injection.4 On the other hand, capsaicin-
induced excitation of TRPV1 receptors has also been
shown to be involved in counter-irritation mechanisms
(ie, pain in one part of the body can be used to control
pain in another part) that are involved in inhibiting the
development of subsequent nociceptive behaviors and
inflammatory reactions at distant body sites in the
rat.1,33 On the basis of these studies, we hypothesized
that BV activation of CSPAs not only elicits a nociceptive
501
signal, but that activation of CSPAs can simultaneously
produce BVAN via a counter-irritation mechanism that
involves activation of the descending pain inhibitory sys-
tem (DPIS). To test this hypothesis we examined whether
the depletion of CSPAs by using resiniferatoxin (RTX)
pretreatment35 could modify BV-induced spinal Fos ex-
pression and BVAN in the formalin test.
BVAN involves activation of primary afferent fibers as
discussed above. We have previously reported that BVAN
is blocked by intrathecal pretreatment with the ␣2-adre-
noceptor antagonists idazoxan or yohimbine in several
different pain models.16,17,30 This implies that BVAN is
mediated by the activation of spinal ␣2-adrenoceptors,
which are known to be involved with the DPIS.24 In this
regard, we have recently shown that peripheral BV injec-
tion effectively increases brainstem catecholaminergic
neuronal activity including the activity of the locus caer-
uleus (LC).19 Therefore, the final objective of this study
was to evaluate whether RTX pretreatment also affects
BV-induced catecholaminergic neuronal activity in the
LC and subsequent spinal ␣2-adrenoceptor activation.
Materials and Methods
Animals
Experiments were performed on male ICR mice weigh-
ing 20 to 25 g. All experimental animals were obtained
from the Laboratory Animal Center of Seoul National
University. They were housed in colony cages with free
access to food and water and maintained in tempera-
ture- and light-controlled rooms (23°C ⫾ 2°C, 12/12-hour
light/dark cycle with lights on at 7:00 AM) for at least 1
week before the study. All of the methods used in the
present study were approved by the Animal Care and Use
Committee at Seoul National University and conform to
National Institutes of Health guidelines (NIH publication
no. 86-23, revised 1985). In addition, the ethical guide-
lines for investigating experimental pain in conscious an-
imals recommended by the International Association for
the Study of Pain were followed.43
BV Administration and RTX Pretreatment
To evaluate the effect of BV injection into the Zusanli
acupoint on spinal Fos expression and nociceptive behav-
iors as well as on BVAN in the formalin test in naive mice,
BV from Apis mellifera (Sigma, St Louis, Mo) was dis-
solved in physiologic saline (20 ␮L) at doses ranging from
0.001 to 10 mg/kg. A therapeutic dose of 0.005 to 0.5
mg/kg of BV is typically used to produce analgesia in
human patients and is considered to be safe because this
dose range does not appear to affect the central ner-
vous, cardiovascular, respiratory, and gastrointestinal
systems.14 Accordingly, the dose range of BV used in the
present study encompassed these clinically used doses.
Diluted BV was subcutaneously administered into the
Zusanli acupoint of the right hind limb located on the
lateral side of the stifle joint adjacent to the anterior
tubercle of the tibia as previously described.16 Animals in
the control group received an injection of vehicle into
502
the Zusanli acupoint. Five mice per individual group
were used for analysis of the effects of different doses of
BV on spinal cord Fos expression and on BV-induced no-
ciceptive behavior. Eight or more mice were included in
each BV treatment or control group for behavioral anal-
ysis in the formalin test.
To evaluate the potential role of CSPAs in BV-induced
spinal Fos expression and on BVAN in the formalin test,
only the high (10 mg/kg), middle (0.1 mg/kg), and low
(0.001 mg/kg) doses of BV were used in these experi-
ments. The extremely potent capsaicin analog RTX (0.2
mg/kg; Sigma) was dissolved in a mixture of 10% Tween
80, 10% ethanol, and 80% normal saline.12,27 Either RTX
or vehicle (10% Tween 80, 10% ethanol, and 80% nor-
mal saline; SHAM) was injected subcutaneously in a vol-
ume of 50 ␮L into the scruff of the neck of the mouse
anesthetized with 3% isoflurane in a mixture of N2O/O2
gas 2 weeks before performing BV-induced Fos immuno-
histochemistry and the formalin test. We waited 2 weeks
after RTX pretreatment to test the possible role of CSPAs,
which is based on the timeframe used in a previous
study.32 To confirm that RTX treatment destroyed CSPAs,
a diluted capsaicin solution (0.01%, dissolved in saline)
was dropped into cornea, and then the number of eye
wipes was counted for 1 minute on the day before BV-
induced Fos immunohistochemistry and formalin injec-
tion (SHAM, n ⫽ 24; RTX, n ⫽ 29). In addition to counting
capsaicin-induced eye wipes, TRPV1 immunohistochem-
istry was performed on both the dorsal root ganglion
(DRG) and the spinal cord at the completion of each
experiment to further confirm the depletion of CSPAs by
RTX treatment.2 TRPV1 immunoreactivity (Calbiochem,
San Diego, Calif; 1:100) was performed by using an im-
munohistochemistry procedure similar to that described
below for Fos immunostaining, except that a fluores-
cent-labeled secondary antibody was used. The number
of TRPV1-positive neurons in DRG and the area of TRPV1-
positive axons in spinal dorsal horn were calculated by
using an image analysis system. A total of 8 mice were
used for TRPV1 immunohistochemistry.
Spinal Fos Expression and Fos–Tyrosine
Hydroxylase Double Labeling in the LC
Immunohistochemistry
In the present study Fos immunohistochemistry was
performed on spinal cord tissue obtained 2 hours
post-BV injection because spinal cord Fos protein expres-
sion typically reaches peak values at approximately 2
hours after acute peripheral stimulation.9,40 Two hours
after each dose of BV or saline injection (n ⫽ 5, respec-
tively), animals were deeply anesthetized with 5% isoflu-
rane and perfused transcardially with calcium-free Ty-
rode’s solution followed by a fixative containing 4%
paraformaldehyde and 0.2% picric acid in 0.1 mol/L
phosphate buffer (pH 6.9). The spinal cord and brain
were removed immediately after perfusion, post-fixed in
the same fixative for 4 hours, and then cryoprotected in
30% sucrose in PBS for 48 hours (pH 7.4). Forty-microme-
ter thick transverse frozen sections were cut through the
Mechanism of Bee Venom–Induced Antinociception
spinal cord and brain by using a cryostat (Microm, Wall-
dorf, Germany).
After elimination of endogenous peroxidase activity
with 3% hydrogen peroxide in PBS and preblocking with
3% normal goat serum and 0.3% Triton X-100 in PBS,
sections were incubated in polyclonal rabbit anti-Fos an-
tibody (Calbiochem, EMD Biosciences; 1:10000) over-
night at 4°C. After several washes, the tissue sections
were processed with the avidin-biotin method (Elite
ABC; Vector Laboratories, Burlingame, Calif). Finally, Fos
immunoreactive neurons were visualized by using a 3,3=-
diamino-benzidine (DAB; Sigma) reaction with 0.2%
nickel chloride intensification (yielding black-labeled
neuronal nuclei). For double labeling experiments to co-
localize Fos and tyrosine hydroxylase (TH, a marker of
catecholaminergic neuron as one of the catecholamine
synthesis enzymes)13 in the LC region, the Fos-reacted
sections were thoroughly rinsed and subsequently incu-
bated with rabbit anti-TH antibody (Biogenesis, Poole,
England; 1:2000). TH immunoreactivity was visualized by
using a DAB reaction (yielding brown-labeled neuronal
perikarya) as previously described.19
Image Analysis
All data analysis procedures were performed blindly
with respect to the experimental condition of the ani-
mal. For quantitative analysis of Fos-positive neurons in
the lumbar spinal cord (L2-3) and LC region, sections
were scanned, and then 5 spinal cord and 5 LC sections
with the greatest number of Fos immunoreactive neu-
rons were selected from each animal. Spinal cord tissue
sections were first examined by using dark-field micros-
copy (Zeiss Axioscope, Hallbergmoos, Germany) to de-
fine the individual spinal cord laminae according to the
gray matter landmarks. The sections were then exam-
ined under a bright-field microscope at 100⫻ to localize
and quantify Fos-positive neurons. The L2-3 segments of
the spinal cord were chosen for analysis in the present
study because these 2 segments receive primary afferent
input from the knee (Zusanli acupoint) area of the hind
limb.34 Moreover, in a preliminary study we found that
BV injection into the Zusanli acupoint selectively in-
creased Fos expression in the L2-3 spinal cord segments
rather than the L4-6 segments, which receive input from
the hind paw.
To specifically identify the brainstem LC cell group, we
used the nomenclature and nuclear boundaries defined
by Franklin and Paxinos in their stereotactic mouse brain
atlas. The region of the LC is located approximately
⫺1.50 mm to ⫺1.95 mm behind the interaural line of the
brainstem. The selected sections were digitized with
4096 gray levels by using a cooled CCD (Micromax Kodak
1317; Princeton Instrument, Trenton, NJ) equipped with
a computer-assisted image analysis system (Metamorph;
Universal Imaging Co, West Chester, Pa). To maintain a
constant threshold for each image and to compensate
for subtle variability of immunostaining, we counted
only neurons that were at least 30% darker than the
average gray level of each image after background sub-
traction and shading correction were performed. BV-in-
ORIGINAL REPORT/Roh et al
Figure 1. Photomicrographs (A-D) of representative L2-L3 spi-
nal cord sections illustrating Fos expression in the dorsal horn
after administration of different doses of BV. Injection of saline
(A) or a low dose (0.001 mg/kg) of BV (B) induces very little Fos
expression in the dorsal horn. In contrast, administration of an
intermediate dose (0.1 mg/kg) (C) or a high dose (10 mg/kg) (D)
of BV produced a significant increase in spinal cord Fos expres-
sion. Scale bar, 200 ␮m. (E, F) Graphs demonstrating the laminar
distribution of Fos immunoreactive neurons in the ipsilateral (E)
and contralateral (F) dorsal horn (L2-3) induced by injection of
different doses of BV (n ⫽ 5 for all groups). *P ⬍ .05, **P ⬍ .01
significantly different from the saline treatment group (SAL),
respectively. Total, entire spinal cord dorsal horn.
503
Figure 2. Graphs illustrating the log dose-response curves for
BV’s effect on (A) the total counts of BV-induced Fos expression
in the entire spinal cord dorsal horn and (B) on formalin-induced
nociceptive behavior during the second phase (10 to 30 minutes
after formalin injection) of the formalin test. The straight lines
are derived from the equation Y ⫽ 13.23logX ⫹ 60.10 of the
administered dose with R ⫽ 0.945 in (A) and Y ⫽ ⫺46.81logX ⫹
89.51 with R ⫽ 0.975 in (B). (C) A graph demonstrating the effect
of BV injection (0.001, 0.1, and 10 mg/kg) into Zusanli point on
spontaneous nociceptive behavior (0 to 60 minutes post-BV in-
jection) in animals that did not receive formalin injection.
504
Figure 3. Graphs illustrating the antinociceptive effect pro-
duced by injection of different doses (0.001-10 mg/kg) of BV on
formalin-induced nociceptive behavior for the entire 30-minute
time course (A) and during the first (0-10 minutes) and second
phases (10-30 minutes) of the formalin test (B). *P ⬍ .05, **P ⬍
.01 significantly different from saline treatment group (SAL),
respectively.
duced Fos staining was analyzed in the following 3 gray
matter regions on the basis of cytoarchitectonic criteria:
(1) the superficial dorsal horn (SDH, laminae I and II); (2)
the nucleus proprius (NP, laminae III and IV); and (3) the
neck region (NECK, laminae V and VI).
Neurons double-labeled for Fos and TH were quanti-
fied in the LC as previously described.19 Eight sections
through the LC were randomly selected from each ani-
mal and subsequently processed for Fos and TH double
labeling. The average number of immunoreactive neu-
rons from each animal was calculated from at least 5
representative sections. The percentage of Fos double-
labeled catecholaminergic (TH) neurons was calculated
as follows: Ratio of double labeling ⫽ Number of double-
labeled (Fos and TH) neurons/Number of TH-labeled neu-
rons ⫻ 100.
Formalin-Induced Pain Behavior Test
Ten minutes after BV injection, 1% formalin in a
volume of 20 ␮L was injected subcutaneously into the
Mechanism of Bee Venom–Induced Antinociception
plantar surface of the right hind paw with a 30-gauge
needle. After formalin injection, the animals were im-
mediately placed in an acrylic observation chamber (30
cm in diameter and height), and nociceptive responses
in each animal were recorded by using a video camera
for a period of 30 minutes. The summation of time (in
seconds) spent licking and biting the formalin-injected
hind paw during each 5-minute block was measured as
an indicator of the nociceptive response. Two experi-
enced investigators who were blinded to the experi-
mental conditions measured these formalin-induced
behaviors. The duration of the responses during the
first 10-minute period and the subsequent 10- to 30-
minute period represents the first and second phases,
respectively, of the formalin test.
To evaluate the nociceptive response induced by sub-
cutaneous administration of different doses (0.001, 0.1,
and 10 mg/kg) of BV into the Zusanli acupoint in animals
without formalin injection, the duration of spontaneous
pain behavior was measured for a period of 60 minutes
after injection by using the same method that was used
for the formalin test.
Intrathecal Injection of ␣2-Adrenoceptor
Antagonist
To evaluate the potential involvement of spinal ␣2-
adrenoceptors on BVAN after RTX pretreatment, an ␣2-
adrenergic receptor antagonist, idazoxan (IDA, 10 ␮g/
mice28; Sigma) was injected intrathecally 10 minutes
before BV injection in both the SHAM and RTX-treated
groups. Six or 7 mice were randomly assigned to each BV
or control group, respectively. Intrathecal injections
were made by using a modification of the Hylden and
Wilcox technique.10 Briefly, a 30-gauge needle con-
nected to a 50-␮L Hamilton syringe with polyethylene
tubing was inserted into the skin and then through the
L5-L6 intervertebral space directly into the subarachnoid
space. A flick of the mouse’s tail provided a reliable indi-
cator that the needle had penetrated the dura, and 5 ␮L
of the drug was subsequently injected into the subarach-
noid space.
Statistical Analysis
One-way analysis of variance (ANOVA) was performed
to determine the overall effect of BV treatment on spinal
Fos expression and on nociceptive behaviors as well as on
the resultant Fos-TH double labeling in the LC. An un-
paired t test was used to determine the P value between
the vehicle (SHAM) and RTX-treatment groups, whereas
a Newman-Keuls test was used to determine the 95%
confidence interval among the BV treatment groups
when ANOVA indicated a significant group difference. A
P value ⬍.05 was considered statistically significant. All
values are expressed as the mean ⫾ standard error of the
mean.
ORIGINAL REPORT/Roh et al 505

Figure 4. Photomicrographs (A-D) and graphs (E, F) showing the effect of RTX treatment on capsaicin-sensitive neurons in a
representative section through a DRG (B) and on capsaicin-sensitive axons in a representative section from the dorsal horn (D). Many
TRPV1-ir neurons are evident in the DRG (A, E), whereas their central axonal processes are present in spinal dorsal horn (C, F) of
vehicle-treated mice (SHAM, n ⫽ 8). Immunostaining is absent in the DRG (B, E) and dorsal horn (D, F) of mice that were treated with
RTX (n ⫽ 8). Scale bar, 200 ␮m. (G, H) Graphs demonstrating the effect of RTX treatment on the capsaicin-induced eye-wiping test (G,
SHAM: n ⫽ 24; RTX: n ⫽ 29) and on formalin-induced pain behavior (H, n ⫽ 9, respectively). RTX treatment totally suppressed
capsaicin-induced eye-wiping behavior (**P ⬍ .01) and significantly reduced pain behavior during first phase (**P ⬍ .01), but not the
second phase, of the formalin test.
506
Figure 5. Graphs illustrating the effect of vehicle (SHAM) or
RTX pretreatment on the antinociceptive effect produced by BV
injection (0.001-10 mg/kg) on formalin-induced nociceptive be-
havior during the first phase (A, 0-10 minutes) and the second
phase (B, 10-30 minutes) of the formalin test (n ⫽7 vehicle; n ⫽
9 RTX). *P ⬍ .05 and **P ⬍ .01 as compared with saline treat-
ment, respectively.
Results
Relationship Between the Dose of BV
and Its Nociceptive and Antinociceptive
Effects
Injection of BV at doses ranging from 0.01 to 10 mg/kg
into the right hind limb resulted in a significant dose-de-
pendent increase in Fos immunoreactive (Fos-ir) neurons in
the ipsilateral (right half, Fig 1A-E and Fig 2A), but not the
contralateral (Fig 1F), dorsal horn of the lumbar spinal cord.
Injection of the 2 highest doses of BV (1 and 10 mg/kg)
evoked significant increases in Fos expression throughout
much of the ipsilateral dorsal horn including the SDH, NP,
and NECK regions, whereas the intermediate doses of BV
Mechanism of Bee Venom–Induced Antinociception
(0.01 and 0.1 mg/kg) selectively increased Fos expression
only in the SDH and NECK regions of the ipsilateral spinal
dorsal horn. Interestingly, all doses of BV (0.001, 0.1, and 10
mg/kg) injected into the Zusanli acupoint failed to evoke
significant nociceptive behaviors during the 60-minute ob-
servation period (Fig 2C). Thus BV injection did not produce
any detectable nocifensive behaviors in comparison with
the vehicle treatment group.
Similar to what was observed with BV-induced Fos ex-
pression, injection of the lowest dose of BV (0.001 mg/kg)
had no suppressive effect on nociceptive behavior (paw
licking and biting time) in either the first or second phase
of the formalin test (Fig 3A, B). Injection of the middle
doses of BV produced a weak, nonsignificant antinoci-
ceptive effect, whereas injection of the high dose of BV
produced a significant antinociceptive effect on paw
licking/biting time during the first phase of the formalin
test (Fig 3A, B). Although a dose-dependent effect was
not observed during the first phase, this could be an
artifact of the lower measures. In contrast, injection of
the middle and high doses of BV (0.01, 0.1, 1, and 10
mg/kg) potently suppressed the second phase of forma-
lin-induced pain as compared to the saline injection con-
trol group, with the highest dose of BV producing a sig-
nificantly more potent BVAN effect as compared to any
of the lower doses (Fig 2B and Fig 3A, B).
Effect of RTX Pretreatment on BV-
Induced Spinal Fos Expression and BV-
Induced Antinociception
RTX treatment was found to dramatically suppress
eye-wiping behavior induced by dropping diluted cap-
saicin (0.01%) onto the cornea in the majority of RTX-
treated mice compared with non–RTX-treated mice
(Fig 4G; P ⬍ .01). Furthermore, TRPV1-ir neurons that
are evident in the DRG and spinal cord dorsal horn of
vehicle-treated mice (SHAM, Fig 4A, C) were not de-
tected in the RTX-treated group (Fig 4B, D), further
indicating that the RTX treatment was successful in
depletion of CSPAs (Fig 4E, F; P ⬍ .01). It was notable
that RTX pretreatment itself significantly suppressed
the first phase of formalin-induced pain behavior but
not the second phase of pain behavior (Fig 4H and Fig
5A; P ⬍ .01). This result was consistent with those of
other previous studies.31, 41
Spinal Fos expression induced by the intermediate
dose of BV (0.1 mg/kg) was not affected by RTX pretreat-
ment (Fig 6A, C and Fig 7). On the other hand, spinal Fos
expression induced by the high dose of BV (10 mg/kg) in
the SDH and NECK regions was selectively attenuated by
RTX treatment (Fig 6B, D and Fig 7; P ⬍ .01 and P ⬍ .05,
respectively), but the total number of Fos-ir neurons was
similar to that of the intermediate-dose BV group (Fig
7D). In addition, BV-induced Fos expression in the con-
tralateral spinal cord dorsal horn was not affected by
RTX pretreatment (Fig 7E).
On the other hand, RTX pretreatment did not reduce
the BVAN effect on the second phase of the formalin test
in either the middle- or high-dose BV (0.1 and 10 mg/kg)–
ORIGINAL REPORT/Roh et al 507

Figure 6. Photomicrographs of representative spinal cord sections from the vehicle (SHAM, A, B) and RTX (C, D) treatment groups
illustrating BV-induced Fos immunolabeling in the ipsilateral lumbar spinal cord dorsal horn. (A) Spinal Fos expression is illustrated in
a spinal cord section taken from an animal in the SHAM group that was treated with an intermediate dose (0.1 mg/kg) of BV. (B) Fos
expression from an animal in the SHAM group treated with a high dose of BV (10 mg/kg). (C) Spinal cord Fos expression in an animal
pretreated with RTX followed by an injection of BV (0.1 mg/kg). (D) Spinal cord Fos expression in an animal pretreated with RTX
followed by an injection of a higher dose of BV (10 mg/kg,). RTX pretreatment caused a significant reduction in the Fos expression
produced by the 10-mg/kg dose of BV. Scale bar, 200 ␮m.

treated groups (Fig 5B; P ⬍ .05 and P ⬍ .01). Similarly,
RTX pretreatment did not affect BVAN in the first phase
of the formalin test in the high-dose BV treatment group
(Fig 5A; P ⬍ .01).
Effect of RTX Pretreatment on the
Neuronal Mechanism of the BV-Induced
Antinociceptive Effect
In the vehicle groups (SHAM), BV treatments (0.1 and
10 mg/kg) significantly increased the number of Fos-
expressing neurons and the ratio of double-labeled
Fos-TH immunoreactive neurons in the LC region (Fig
8A-D; P ⬍ .01) as compared with the saline-treated
group. This indicated that more TH-positive neurons
co-contained Fos immunoreactivity after BV treat-
ment. This anatomic finding correlates well with the
increased antinociceptive effect produced by these
doses of BV (Fig 5). In the RTX pretreatment groups,
the number of Fos immunoreactive neurons and the
colocalization ratio between Fos and TH were not
changed in comparison to the SHAM groups (Fig 8C, D;
P ⬍ .01), indicating that RTX had no effect on BV-
induced Fos expression in the LC.
Intrathecal idazoxan pretreatment (IDA, 10 ␮g/mice) in
the SHAM group (IDA-SAL) did not affect formalin-in-
duced nociceptive behavior in comparison to intrathecal
saline treatment in the SHAM group (SAL-SAL). On the
other hand, IDA pretreatment blocked the development
of BVAN produced by injection of either 0.1 or 10 mg/kg
of BV (Fig 9). Importantly, the inhibitory effect produced
by intrathecal IDA on BVAN was not affected by RTX
pretreatment (Fig 9).
508 Mechanism of Bee Venom–Induced Antinociception

Figure 7. Graphs illustrating the effect of vehicle (SHAM) or RTX pretreatment on BV-induced Fos expression in (A) the SDH, (B) the
NP, (C) the NECK, and in (D) the entire dorsal horn (Total-ipsi) from the ipsilateral spinal cord (n ⫽ 5, respectively). (E) Graph showing
the effect of vehicle (SHAM) or RTX pretreatment on BV-induced Fos expression in the entire dorsal horn (Total-contra) of contralat-
eral spinal cord (n ⫽ 5). *P ⬍ .05 and **P ⬍ .01 as compared with saline treatment, respectively. The high dose of BV (10 mg/kg)-
induced spinal Fos expression was selectively attenuated by RTX pretreatment in the SDH and NECK regions (P ⬍ .01 and P ⬍ .05,
respectively).

Discussion present as well as in previous studies.21 This difference

Peripheral BV Stimulation–Induced Spinal
Fos Expression Without Nociceptive
Behavior Is Closely Related to BV’s
Antinociceptive Effect
It has been reported that intraplantar BV injection pro-
duces a set of nocifensive behaviors including licking,
biting, and flinching for a period of approximately 1
hour after injection.3,18 In contrast, we failed to detect
any observable nocifensive behavior when different
doses of BV were injected into the Zusanli point in the
could be due to the fact that we are injecting BV directly
into an acupoint as opposed to a non-acupoint in the
foot or to the fact that the subcutaneous tissue of the
hind paw has a greater innervation density than the area
near the stifle joint where the Zusanli acupoint is lo-
cated. In addition, there are anatomic and likely func-
tional differences between intraplantar glabrous skin
and the hairy skin where the Zusanli point is located in
rodents. Thus, these results indicate that BV stimulation
of the Zusanli acupoint evokes very little nociceptive be-
havior in the rodent. Although BV injection into the hu-
ORIGINAL REPORT/Roh et al 509

Figure 8. Photomicrographs illustrating single- and double-labeled Fos and TH immunoreactive neurons in the LC region (A, B) and
graphs (C, D) showing the effect of BV treatment on the number of Fos immunoreactive neurons in the LC region (C) and the ratio
of Fos co-expression with TH (D) in either vehicle (SHAM) or RTX-pretreated mice (n ⫽ 5, respectively). The number of Fos/TH
double-labeled neurons in animals treated with the high dose of BV (10 mg/kg) (B) was greater than that of saline-treated animals
(A). **P ⬍ .01 compared with saline treatment group. White arrowhead, TH immunoreactive neuron; black arrowhead, Fos-labeled
neuronal nuclei; white arrow, double-labeled neurons. Scale bar, 50 ␮m.
510
Figure 9. Graph illustrating the effect of intrathecal (i.t.) saline
and IDA (10 ␮g/mouse) on the BV (0.1 and 10 mg/kg)-induced
antinociceptive effect on the second phase (10-30 minutes) of
formalin-induced pain behaviors in both vehicle (SHAM) and
RTX-pretreated groups (n ⫽ 6 and 7, respectively).
man Zusanli acupoint might represent a more effective
acupuncture stimulation paradigm, it remains to be de-
termined whether BV injection into the human Zusanli
acupoint is painful.
On the other hand, spinal Fos expression, which served
as a marker of the neuronal activity induced by BV stim-
ulation of the Zusanli acupoint, was dose-dependently
increased particularly in the SDH region (Fig 1), with
doses ranging from 0.001 mg/kg (which is typically not
painful in human subjects) to 10 mg/kg (which is approx-
imately equivalent to the dose received in one honey-
bee’s sting). Generally it is well-accepted that increases in
nociceptive stimulus intensity produce increases in dorsal
horn Fos expression.8,11 However, the present study
demonstrates that the increase in spinal Fos expression
induced by BV injection into the Zusanli acupoint did not
correlate with BV-induced nociceptive behavior. There
are 3 possible explanations for this discrepancy between
spontaneous nociceptive behavior and spinal Fos expres-
sion induced by BV injection into the Zusanli point. First,
it is possible that the BV-induced Fos expression repre-
sents a response to non-nociceptive stimulation at the BV
injection site, because not only noxious stimuli but also
innocuous stimuli can produce spinal Fos expression.6,36
Second and perhaps more likely, BV injection into the
Zusanli point does in fact produce nociception, but the
level of nociception, although great enough to evoke
spinal Fos expression, is not adequate to evoke detect-
able pain behaviors. Finally, it is possible that BV injected
into the Zusanli point does not produce observable no-
ciceptive behaviors (flinching, licking, or biting) as it does
in the hind paw, and although nociception was present,
it was not measurable with the behavioral assays used in
the present study. We believe this latter explanation is
the most likely because we have also injected 1% forma-
lin together with BV into the Zusanli point and found
that this combination failed to produce detectable noci-
ceptive behaviors (flinching, licking, or biting). Although
this might be due to differences in sensitivity between
Mechanism of Bee Venom–Induced Antinociception
the hind paw and the subcutaneous tissue around the
stifle joint, it is also possible that chemical activation of
the Zusanli acupoint produces a profound antinocicep-
tion without detectable nociception.
We also showed that BVAN on pain behaviors associ-
ated with the second phase of the formalin test is also
dose-dependent and is produced by the same doses of
BV (0.01, 0.1, 1, and 10 mg/kg) that produce spinal cord
Fos expression (Fig 1B). These findings indicate that the
magnitude of the BV-induced spinal neuronal activation
can be correlated with the magnitude of BVAN, suggest-
ing that BVAN might result from a counter-irritation
mechanism activated by peripheral stimulation. Al-
though counter-irritation is thought to be mediated by
central diffuse noxious inhibitory controls related to no-
ciceptive input,29 our findings importantly demon-
strated that the BVAN can be produced without the in-
duction of spontaneous nociceptive behavior.
Differential Roles of Capsaicin-Sensitive
Afferents and Capsaicin-Insensitive
Afferents on BV-Induced Spinal Fos
Expression and Antinociception
Although RTX pretreatment did not alter the
amount of spinal Fos expression produced by either
the low or middle doses of BV, it selectively attenuated
the increase in Fos expression evoked by the high dose
of BV in both the SDH and NECK regions of the dorsal
horn. (Figs 6 and 7) Importantly, RTX treatment did not
affect the number of BV-induced Fos-ir neurons in the
NP region of the spinal cord dorsal horn. In addition,
RTX had no significant effect on BVAN (Fig 5). This
would suggest that the high dose of BV either directly
or indirectly stimulates CSPAs, which in turn activate
neurons in the SDH and NECK regions. This is consis-
tent with previous anatomic studies demonstrating
that the central terminals of CSPAs primarily innervate
the SDH and NECK regions of the spinal cord, and that
their activation by noxious heat or chemical stimuli
results in the induction of Fos protein in neurons in the
SDH and NECK, but not the NP, regions of the dorsal
horn.12,26 This is also consistent with the present find-
ings demonstrating that RTX pretreatment failed to
alter the increase in Fos immunoreactive neurons in NP
region induced by the high dose of BV. This result
indicates that the high dose of BV probably activates
large-diameter, low-threshold primary afferent neu-
rons (mostly A␤ fibers) in addition to small and me-
dium afferents, and it also might serve to explain the
potent analgesic effect of high-dose BV stimulation
that is thought to occur via the activation of spinal
GABAergic inhibitory interneurons.7
With respect to the type of primary afferent that is
stimulated, it has been reported that the threshold of
neuronal activation to electrical stimulation differs ac-
cording to the type of primary afferent fiber (A␤, A␦,
or C). For example, the minimum stimulus intensities
and durations required to activate C and A␦ fibers
were 110 ␮A, 0.4 millisecond and 34 ␮A, 0.1 millisec-
ORIGINAL REPORT/Roh et al
ond, respectively, in an in vitro rat spinal cord prepa-
ration.42 Furthermore, it has been shown that activa-
tion of A␦ fibers is most effective in producing
prolonged inhibition of spinothalamic tract cells, al-
though significant additional effects were produced
by stimulation of A␣, A␤, and C fibers. These data,
together with the findings of Uchida et al39 showing
that electroacupuncture stimulation causes an in-
crease in spinal cord Fos expression via capsaicin-insen-
sitive primary afferent A␦ fibers, suggest that the most
effective way to produce analgesia by peripheral
nerve stimulation would be by high-frequency stimu-
lation with an intensity strong enough to activate A␦
fibers.5,22 This concept is compatible with our results
showing that activation of capsaicin-insensitive pri-
mary afferents (CIPAs) by peripheral chemical stimula-
tion with diluted BV is most likely involved in BVAN.
Because most A␦ afferents are insensitive to capsa-
icin,25 it is likely that BV is producing its antinocicep-
tive effect by activation of A␦ fibers at the site of in-
jection.
Role of Capsaicin-Insensitive Afferents in
the Central Neuronal Mechanisms
Underlying BV-Induced Antinociception
We have recently demonstrated that peripheral BV
injection increases Fos expression in rat brainstem cat-
echolaminergic neurons including many neurons in
the LC.19 We have further shown that the activation of
spinal ␣2-adrenoceptors, but not opioid receptors, is
critically involved in the BV-induced antinociceptive
and antihyperalgesic effects observed in rodent mod-
els of visceral pain, inflammatory pain, and neuro-
pathic pain.16,17,30 As an extension of this work, the
present study demonstrated that chemical stimulation
References
1. Carlton SM, Zhou S, Kraemer B, Coggeshall RE: A role for
peripheral somatostatin receptors in counter-irritation-in-
duced analgesia. Neuroscience 120:499-508, 2003
2. Caterina MJ, Schumacher MA, Tominaga M, Rosen TA,
Levine JD, Julius D: The capsaicin receptor: A heat-activated
ion channel in the pain pathway. Nature 389:816-824, 1997
3. Chen J, Luo C, Li HL: The contribution of spinal neuronal
changes to development of prolonged, tonic nociceptive
responses of the cat induced by subcutaneous bee venom
injection. Eur J Pain 2:359-376, 1998
4. Chen J, Chen HS: Pivotal role of capsaicin-sensitive pri-
mary afferents in development of both heat and mechanical
hyperalgesia induced by intraplantar bee venom injection.
Pain 91:367-376, 2001
5. Chung JM, Lee KH, Hori Y, Endo K, Willis WD: Factors
influencing peripheral nerve stimulation produced inhibi-
tion of primate spinothalamic tract cells. Pain 19:277-293,
1984
6. Coggeshall RE: Fos, nociception and the dorsal horn. Prog
Neurobiol 77:299-352, 2005
511
of the Zusanli acupoint by BV leads to activation of
peripheral CIPAs fibers, which in turn evokes cat-
echolaminergic neuronal activation in the LC region.
Because the coeruleospinal pathway plays an impor-
tant role in the descending pain inhibitory system,38
activation of the LC by BV stimulation of CIPAs likely
plays a key role in BV’s antinociceptive effects on for-
malin-induced nociceptive behaviors. The fact that
RTX pretreatment failed to alter the number of Fos-ir
neurons or the ratio of Fos-TH double-labeled neurons
in the LC induced by BV injection would support our
hypothesis that BV injection activates the central nor-
adrenergic system via stimulation of peripheral CIPAs.
This is further supported by the finding that intrathe-
cal pretreatment with the ␣2-adrenoceptor antagonist
IDA abolished BV’s antinociceptive effect on formalin-
induced pain behavior, whereas RTX pretreatment
failed to alter IDA’s inhibitory effect on BVAN. Collec-
tively these data indicate that BV injection into the
Zusanli acupoint stimulates peripheral CIPAs, which in
turn activate catecholaminergic neurons in the LC. The
LC then stimulates spinal cord ␣2-adrenoceptors via
descending noradrenergic pathways, and this leads to
BV’s antinociceptive effect on formalin-induced pain
behaviors.
In conclusion, there are 2 important findings that stem
from the data obtained in this study: (1) BV stimulation
of the Zusanli acupoint produces a significant antinoci-
ceptive effect in the second phase of the formalin test
that involves spinal neuronal transmission without de-
tectable nociceptive behavior, and (2) peripheral CIPAs
are primarily involved in activating central catecholamin-
ergic pathways associated with BV’s antinociceptive ef-
fect.
7. De Koninck Y, Henry JL: Prolonged GABAA-mediated in-
hibition following single hair afferent input to single spinal
dorsal horn neurones in cats. J Physiol 476:89-100, 1994
8. Doyle CA, Hunt SP: Substance P receptor (neurokinin-1)-
expressing neurons in lamina I of the spinal cord encode for
the intensity of noxious stimulation: a c-Fos study in rat.
Neuroscience 89:17-28, 1999
9. Fukura M, Kashima K, Maeda S, Shiba R: Changes in bite
force and muscle forces in the upper extremities after
counter irritation. Cranio 22:45-49, 2004
10. Hylden JL, Wilcox GL: Intrathecal morphine in mice: A
new technique. Eur J Pharmacol 67:313-316, 1980
11. Iwata K, Takahashi O, Tsuboi Y, Ochiai H, Hibiya J,
Sakaki T, Yamaguchi Y, Sumino R: Fos protein induction in
the medullary dorsal horn and first segment of the spinal
cord by tooth-pulp stimulation in cats. Pain 75:27-36, 1998
12. Jinks SL, Simons CT, Dessirier JM, Carstens MI, Antognini
JF, Carstens E: C-fos induction in rat superficial dorsal horn
following cutaneous application of noxious chemical or me-
chanical stimuli. Exp Brain Res 145:261-269, 2002
13. Jones SL: Descending noradrenergic influences on pain.
Prog Brain Res 88:381-394, 1991
512
14. Kim HW, Kwon YB, Ham TW, Roh DH, Yoon SY, Lee HJ,
Han HJ, Yang IS, Beitz AJ, Lee JH: Acupoint stimulation using
bee venom attenuates formalin-induced pain behavior and
spinal cord fos expression in rats. J Vet Med Sci 65:349-355,
2003
15. Kim HW, Kwon YB, Ham TW, Roh DH, Yoon SY, Kang SY,
Yang IS, Han HJ, Lee HJ, Beitz AJ, Lee JH: General pharma-
cological profiles of bee venom and its water soluble frac-
tions in rodent models. J Vet Sci 5:309-318, 2004
16. Kim HW, Kwon YB, Han HJ, Yang IS, Beitz AJ, Lee JH:
Antinociceptive mechanisms associated with diluted bee
venom acupuncture (apipuncture) in the rat formalin test:
Involvement of descending adrenergic and serotonergic
pathways. Pharmacol Res 51:183-188, 2005
17. Kwon YB, Kang MS, Han HJ, Beitz AJ, Lee JH: Visceral
antinociception produced by bee venom stimulation of the
Zhongwan acupuncture point in mice: Role of alpha(2) ad-
renoceptors. Neurosci Lett 308:133-137, 2001
18. Kwon YB, Lee JD, Lee HJ, Han HJ, Mar WC, Kang SK,
Beitz AJ, Lee JH: Bee venom injection into an acupuncture
point reduces arthritis associated edema and nociceptive
responses. Pain 90:271-280, 2001
19. Kwon YB, Han HJ, Beitz AJ, Lee JH: Bee venom acupoint
stimulation increases Fos expression in catecholaminergic
neurons in the rat brain. Mol Cells 17:329-333, 2004
20. Lariviere WR, Melzack R: The bee venom test: Compari-
sons with the formalin test with injection of different ven-
oms. Pain 84:111-112, 2000
21. Lee JH, Kwon YB, Han HJ, Mar WC, Lee HJ, Yang IS, Beitz
AJ, Kang SK: Bee venom pretreatment has both an antino-
ciceptive and anti-inflammatory effect on carrageenan-in-
duced inflammation. J Vet Med Sci 63:251-259, 2001
22. Lee KH, Chung JM, Willis WD Jr: Inhibition of primate
spinothalamic tract cells by TENS. J Neurosurg 62:276-287,
1985
23. Luo C, Chen J, Li HL, Li JS: Spatial and temporal expres-
sion of c-Fos protein in the spinal cord of anesthetized rat
induced by subcutaneous bee venom injection. Brain Res
806:175-185, 1998
24. Mansikka H, Lahdesmaki J, Scheinin M, Pertovaara A:
Alpha2A adrenoceptors contribute to feedback inhibition
of capsaicin-induced hyperalgesia. Anesthesiology 101:185-
190, 2004
25. Nagy JI, van der Kooy D: Effects of neonatal capsaicin
treatment on nociceptive thresholds in the rat. J Neurosci
3:1145-1150, 1983
26. Nakatsuka T, Furue H, Yoshimura M, Gu JG: Activation
of central terminal vanilloid receptor-1 receptors and alpha
beta-methylene-ATP-sensitive P2X receptors reveals a con-
verged synaptic activity onto the deep dorsal horn neurons
of the spinal cord. J Neurosci 22:1228-1237, 2002
27. Pan HL, Khan GM, Alloway KD, Chen SR: Resiniferatoxin
induces paradoxical changes in thermal and mechanical sen-
sitivities in rats: Mechanism of action. J Neurosci 23:2911-
2919, 2003
28. Pericic D: Swim stress inhibits 5-HT2A receptor-medi-
ated head twitch behaviour in mice. Psychopharmacology
167:373-379, 2003
Mechanism of Bee Venom–Induced Antinociception
29. Roby-Brami A, Bussel B, Willer JC, Le Bars D: An electro-
physiological investigation into the pain-relieving effects of
heterotopic nociceptive stimuli: Probable involvement of a
supraspinal loop. Brain 110:1497-1508, 1987
30. Roh DH, Kwon YB, Kim HW, Ham TW, Yoon SY, Kang SY,
Han HJ, Lee HJ, Beitz AJ, Lee JH: Acupoint stimulation with
diluted bee venom (apipuncture) alleviates thermal hyper-
algesia in a rodent neuropathic pain model: involvement of
spinal alpha 2-adrenoceptors. J Pain 5:297-303, 2004
31. Shibata M, Ohkubo T, Takahashi H, Inoki R: Modified
formalin test: Characteristic biphasic pain response. Pain 38:
347-352, 1989
32. Szallasi A, Joo F, Blumberg PM: Duration of desensitiza-
tion and ultrastructural changes in dorsal root ganglia in
rats treated with resiniferatoxin, an ultrapotent capsaicin
analog. Brain Res 503:68-72, 1989
33. Szolcsanyi J, Pinter E, Helyes Z, Oroszi G, Nemeth J: Sys-
temic anti-inflammatory effect induced by counter-irrita-
tion through a local release of somatostatin from nocicep-
tors. Br J Pharmacol 125:916-922, 1998
34. Takahashi Y, Chiba T, Kurokawa M, Aoki Y: Dermatomes
and the central organization of dermatomes and body sur-
face regions in the spinal cord dorsal horn in rats. J Comp
Neurol 462:29-41, 2003
35. Tender GC, Walbridge S, Olah Z, Karai L, Iadarola M,
Oldfield EH, Lonser RR: Selective ablation of nociceptive
neurons for elimination of hyperalgesia and neurogenic in-
flammation. J Neurosurg 102:522-525, 2005
36. Tominaga M, Caterina MJ, Malmberg AB, Rosen TA, Gil-
bert H, Skinner K, Raumann BE, Basbaum AI, Julius D: The
cloned capsaicin receptor integrates multiple pain-producing
stimuli. Neuron 21:531-543, 1998
37. Torsney C, Meredith-Middleton J, Fitzgerald M: Neona-
tal capsaicin treatment prevents the normal postnatal with-
drawal of A fibres from lamina II without affecting fos re-
sponses to innocuous peripheral stimulation. Brain Res Dev
Brain Res 121:55-65, 2000
38. Tsuruoka M, Maeda M, Nagasawa I, Inoue T: Spinal
pathways mediating coeruleospinal antinociception in the
rat. Neurosci Lett 362:236-239, 2004
39. Uchida Y, Nishigori A, Takeda D, Ohshiro M, Ueda Y,
Ohshima M, Kashiba H: Electroacupuncture induces the ex-
pression of Fos in rat dorsal horn via capsaicin-insensitive
afferents. Brain Res 978:136-140, 2003
40. Williams S, Evan GI, Hunt SP: Changing patterns of c-fos
induction in spinal neurons following thermal cutaneous
stimulation in the rat. Neuroscience 36:73-81, 1990
41. Wismer CT, Faltynek CR, Jarvis MF, McGaraughty S: Dis-
tinct neurochemical mechanisms are activated following ad-
ministration of different P2X receptor agonists into the
hindpaw of a rat. Brain Res 965:187-193, 2003
42. Yajiri Y, Yoshimura M, Okamoto M, Takahashi H, Hi-
gashi H: A novel slow excitatory postsynaptic current in sub-
stantia gelatinosa neurons of the rat spinal cord in vitro.
Neuroscience 76:673-688, 1997
43. Zimmermann M: Ethical guidelines for investigations of
experimental pain in conscious animals. Pain 16:109-110,
1983