| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
The Journal of Neurophysiology Vol. 88 No. 4 October 2002, pp. 1707-1715
Copyright ©2002 by the American Physiological Society
1Instituto Venezolano de Investigaciones Científicas, Caracas 1020-A, Venezuela; and 2Washington State University Vancouver, Vancouver, Washington 98686
 |
ABSTRACT |
|---|
 TOP
 ABSTRACT
 INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
Tortorici, V. and M. M. Morgan. Comparison of Morphine and Kainic Acid Microinjections Into Identical PAG Sites on the Activity of RVM Neurons. J. Neurophysiol. 88: 1707-1715, 2002. The rostral ventromedial medulla (RVM) modulates nociception through changes in the activity of two classes of neuron, ON- and OFF-cells. The activity of these neurons is regulated, in part, by input from the periaqueductal gray (PAG). The objective of this study was to determine whether PAG-mediated antinociception is associated with excitation of both ON- and OFF-cells in the RVM. Microinjection of morphine into the ventrolateral PAG produced antinociception at 50% of the injection sites. This antinociception was associated with continuous activation of RVM OFF-cells and inhibition of both the spontaneous and reflex-related activity of RVM ON-cells. Microinjection of kainic acid into the same injection sites produced antinociception 92% (37/40) of the time. Although kainic acid directly excites PAG output neurons, the changes in ON- and OFF-cell activity associated with microinjection of kainic acid into the ventrolateral PAG were the same as when morphine was injected. That is, ON-cells were inhibited and OFF-cells were activated. These data indicate that the excitatory connection between the PAG and RVM is directed at RVM OFF-cells specifically. In addition, these data suggest that direct activation of PAG output neurons, as occurs with kainic acid, is much more likely to produce antinociception than disinhibition of output neurons as occurs following morphine administration.
| |
INTRODUCTION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
The antinociceptive
properties of opioids are mediated, at least in part, by their action
on supraspinal structures such as the periaqueductal gray (PAG) and the
rostral ventromedial medulla (RVM). These structures produce
antinociception via a descending pathway that runs from the PAG to RVM
to spinal and trigeminal dorsal horns (Basbaum and Fields
1984
). Microinjection of morphine into the PAG or RVM produces
antinociception (Jacquet and Lajtha 1974; Jensen
and Yaksh 1986; Morgan and Whitney 2000;
Morgan et al. 1998), and inactivation of the RVM
disrupts antinociception mediated by the PAG (Prieto et al.
1983; Sandkühler and Gebhart 1984).
The RVM is of particular interest because the neural circuitry
underlying antinociception has been described. There are three classes
of neuron in the RVM: ON-, OFF-, and neutral
cells (Fields et al. 1983a; Vanegas et
al. 1984b). OFF-cells show an abrupt cessation of
activity immediately prior to nociceptive reflexes. More importantly,
manipulations that cause OFF-cells to become continuously
active (i.e., prevent the OFF-cell pause) invariably produce antinociception (Barbaro et al. 1986;
Cheng et al. 1986; Heinricher and Drasner
1991; Heinricher et al. 1989, 1994;
Moreau and Fields 1986; Morgan and Fields
1993). In contrast, ON-cells show a burst of
activity prior to nociceptive reflexes, and selective activation of
ON-cells is associated with facilitation of nociception (Bederson et al. 1990; Heinricher et al.
1989; Kaplan and Fields 1991; Morgan and
Fields 1994; Ramirez and Vanegas 1989). Neutral cells show no consistent change in activity associated with nociceptive reflexes and their function remains unclear.
Anatomical, electrophysiological, and behavioral studies indicate
that an excitatory glutaminergic pathway transmits information from the
PAG to the RVM (Aimone and Gebhart 1986; van
Praag and Frenk 1990; Wiklund et al. 1988).
Electrical stimulation of the PAG has been reported to produce a
predominantly excitatory effect on the activity of RVM neurons
(Behbehani and Fields 1979). This excitation appears to
be directed at both ON- and OFF-cells
(Vanegas et al. 1984a). Antinociception also can be
produced by microinjecting opioids into the PAG. Unlike electrical
stimulation, microinjection of opioids into the PAG inhibits
ON-cell activity (Cheng et al. 1986;
Fang et al. 1989). OFF-cells become
continuously active in both situations, a finding that leads to the
conclusion that OFF-cells produce antinociception.
The different effects of PAG stimulation and morphine microinjection on the activity of RVM ON-cells (activation and inhibition, respectively) suggest there is an excitatory input to RVM ON-cells that PAG opioids do not activate. An alternative hypothesis is that electrical stimulation of the PAG activates ON-cells by exciting fibers of passage or antidromically. The objective of the present study is to determine whether activation of PAG output neurons excites or inhibits RVM ON-cells. This was accomplished by comparing the activity of RVM neurons before and after sequential microinjections of morphine and kainic acid, an excitatory amino acid, into the PAG.
| |
METHODS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
Male Sprague-Dawley rats (250-350 g, B and K Universal, Kent, WA) were initially anesthetized with pentobarbital sodium (60 mg/kg ip, Abbott Laboratories, North Chicago, IL). Guidelines of the Society for Neuroscience and the International Association for the Study of Pain regarding animal experimentation were followed throughout.
A PE-20 catheter (Intramedic, Sparks, MD) was inserted into an external
jugular vein, and the animals were placed in a stereotaxic frame (David
Kopf Instruments, Tujunga, CA). Two small craniotomies allowed drugs to
be injected into the ventrolateral PAG and neurons to be recorded in
the RVM. Injections were made through a stainless steel guide cannula
(25 gauge, 12 mm length, Small Parts, Miami Lakes, FL) aimed
stereotaxically (Paxinos and Watson 1998) at the
ventrolateral PAG but remaining 2 mm above the target area. The cannula
was attached with dental cement to one screw placed in the skull. Body
temperature was maintained at 37°C using a circulating water pad
(K-MOD 100, Baxter, Deerfield, IL).
Following surgery, animals were allowed to recover partially from the
initial anesthetic dose so that it was possible to elicit a tail flick
reflex to noxious heat. Once this anesthetic plane was reached,
methohexital sodium (Eli Lilly, Indianapolis, IN) was infused at a
constant rate (15-30
mg·
1kg·1h)
through the jugular catheter using an infusion pump (model 355, Sage
Instruments, Boston, MA). This maintained the animals in a lightly
anesthetized state with a steady baseline tail flick latency of around
4-5 s and prevented any sign of discomfort or spontaneous movement.
The tail flick test consisted of focusing a projector lamp (Winston Electronics, San Francisco, CA) on the blackened ventral surface of the tail. Heat was applied to areas 3, 4, and 5 cm from the caudal tip of the tail on successive tests. Each trial consisted of a linear increase in temperature (1.8°C/s) from a holding temperature of 35°C until the tail flick occurred or to a maximum of 53°C if no tail flick occurred within 10 s. A thermistor probe in contact with the surface of the tail provided feedback control of the temperature. The tail was attached to a mechanoelectric bridge transducer for automatic detection of tail movement.
Extracellular unit activity was recorded with a stainless steel
electrode (FHC, Bowdoinham, ME) stereotaxically placed in the medulla
(Paxinos and Watson 1998). Unit activity was amplified, filtered, and continuously monitored on an oscilloscope (TDS 410A, Tektronix, Portland, OR). The electrode was advanced through the medulla in steps of 3 µm using a hydraulic microdrive (TrentWells, Coulterville, CA) until a neuron was identified. Neurons were characterized following the classification system applied to cells of
the RVM by Fields et al. (1983). Briefly,
OFF-cells showed an abrupt cessation of activity just prior
to the occurrence of the tail flick. ON-cells were
identified by a sudden burst of activity beginning just before the tail
flick. Neurons that did not display either of these characteristics
(i.e., neutral cells) were not studied. Spike waveform, spike time, and
time of tail flick occurrence were monitored and stored for off-line
analysis (Datawave Systems, Longmont, CO).
After characterization of the cells, three baseline tail flick trials were carried out at 5-min intervals. If the latency of the tail flick was stable (<1 s between each baseline trial), morphine sulfate (5 µg/0.4 µl, RBI, Natick, MA) or 0.4 µl of saline (SAL) was microinjected into the ventrolateral PAG using a stainless steel microinjection cannula (31 gauge, 14 mm length, Small Parts). The injection cannula was attached to a 1-µl syringe (Hamilton, Reno, NV) with a length of PE-20 tubing (Intramedic) and extended 2 mm beyond the guide cannula. Drug injections were made by hand over a period of 2 min. Kainic acid (8.5 ng/0.4 µl, Sigma, St. Louis, MO) was injected into the same PAG location 25 min after the morphine injection. For control purposes, kainic acid was injected before morphine in 16 of 40 trials. There was no difference in the data regardless of the order of the injections, therefore these conditions were combined for data analysis (see Tables 1 and 2).
|
|
At the end of the experiment, a lesion was made at the recording site
(20 µA DC anodal current for 20 s). The injection site was
marked with a microinjection of cresyl violet (0.4 µl, Sigma). The
animal then received a lethal injection of pentobarbital sodium (120 mg/kg, Abbott Laboratories) and the brain was removed and placed in
10% formalin for 1 wk. Microinjection and lesion sites were located on
50-µm coronal sections using the stereotaxic atlas of Paxinos
and Watson (1998).
Tail flick latency and neuronal activity, both spontaneous and tail flick-evoked, were measured at 5-min intervals throughout the experiment. Spontaneous activity was measured as the mean firing rate over the 30 s immediately preceding tail heating. Evoked activity was measured as the mean firing rate for the 2 s immediately preceding the tail flick reflex. This time period covers the range when the ON-cell burst and OFF-cell pause occur. In cases in which the tail flick reflex was inhibited, mean activity during the last 2 s of the trial was used. All data are presented as means ± SE. Rats were separated into two groups based on whether morphine microinjection inhibited the tail flick reflex. Groups were compared across trials using a two-way ANOVA (groups × trial). Post hoc analysis (Tukey's HSD) was used where appropriate. P values < 0.05 were considered to be statistically significant.
| |
RESULTS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
Data were collected from 40 RVM neurons in 36 rats. Several rats in which microinjection of morphine into the PAG did not produce antinociception were tested again 20- to 25-min later with a more ventral injection and a different RVM neuron. All microinjections were in or immediately adjacent to the ventrolateral PAG. There was no clear difference in the locations of the morphine injections that inhibited the tail flick reflex and those that did not (Fig. 1, top). In addition, there was no difference in the location of ON- and OFF-cells in the RVM (Fig. 1, bottom).
|
Tail flick reflex
Microinjection of morphine into the PAG produced antinociception at 50% (20/40) of the injection sites. In the analgesic rats, the increase in tail flick latency began within 5 min and reached the cutoff value in all of these rats by 15-20 min. Microinjection of kainic acid inhibited the tail flick reflex in all of the rats in which morphine microinjection produced antinociception and in 17 of 20 rats in which morphine was ineffective. Taken together, microinjection of kainic acid into the PAG produced antinociception 92% of the time (37/40). The onset for kainic acid-induced inhibition of the tail flick reflex was rapid (within 5 min) and short lived (the reflex was evident 10 min after the injection in most rats).
Spontaneous neural activity
Microinjection of kainic acid into the PAG inhibited the spontaneous activity of RVM ON-cells similar to that produced by microinjection of morphine. In almost all cases, microinjection of morphine or kainic acid caused a complete inhibition of ongoing ON-cell activity (Fig. 2A). The spontaneous activity of ON-cells following morphine and kainic acid microinjection into the PAG was significantly lower than baseline activity (F(2,34) = 27.70; P < 0.01).
|
Microinjection of morphine and kainic acid caused a significant increase in the spontaneous activity of RVM OFF-cells compared with baseline activity levels on trials in which antinociception was produced (F(2,38) = 131.20; P < 0.01). Baseline activity in drug-naive rats was 15.6 ± 1.9 Hz preceding morphine administration and 11.9 ± 0.6 Hz preceding kainic acid administration. Within 5 min of morphine administration, the spontaneous firing rate of OFF-cells had risen to 28.3 ± 3.3 Hz. A peak firing rate of 44.2 ± 1.8 Hz occurred 20 min following morphine administration (Fig. 2B). OFF-cells reached a peak firing rate of 58.9 ± 2.1 Hz within 5 min of microinjection of kainic acid in these same rats. This increase in ongoing OFF-cell activity following kainic acid administration was significantly greater than following morphine administration (Tukey test, P < 0.05). The rapid and short-lived increase in OFF-cell activity following kainic acid microinjection into the PAG corresponded to the time course for tail flick inhibition. Microinjection of kainic acid into the PAG produced a similar increase in OFF-cell activity (55.7 ± 5.9 Hz) in rats in which morphine administration did not produce antinociception. Moreover, the effect of microinjecting kainic acid into the PAG on the spontaneous activity of RVM ON- and OFF-cells did not vary when the injections preceded or followed morphine administration (Table 1).
Tail flick-related neural activity
ON-cells are defined by a burst of activity immediately prior to the occurrence of the tail flick reflex (see Fig. 3). Microinjection of morphine into the PAG inhibited this burst of activity on trials in which the tail flick reflex was inhibited, but not on trials in which the tail flick reflex was not inhibited (Figs. 3 and 4). The difference in ON-cell activity in these two conditions was statistically significant (F(1,17) = 16.09, P < 0.01). Microinjection of kainic acid into these same sites inhibited the tail flick reflex in 18 of 19 rats and inhibited the ON-cell burst regardless of whether morphine administration had an effect. For example, Fig. 3 shows a ratemeter record of an ON-cell following an injection of morphine into the PAG that failed to inhibit the tail flick reflex or the ON-cell burst. Subsequent microinjection of kainic acid into the same PAG site inhibited the tail flick reflex and ON-cell burst of activity. The rapid onset and short duration of this inhibition was characteristic of kainic acid microinjections into the PAG.
|
|
OFF-cells are defined by a pause in activity immediately prior to the occurrence of the tail flick reflex (Fig. 5). On trials in which microinjection of morphine inhibited the tail flick reflex, this pause in activity was eliminated. In contrast, the OFF-cell pause in activity continued to occur on trials in which the tail flick reflex was not inhibited (Fig. 6). Microinjection of kainic acid inhibited both the tail flick reflex and the OFF-cell pause in 19 of 21 rats. These kainic acid-induced inhibitory effects occurred regardless of the effect of the morphine microinjection (see Figs. 5 and 6). Elimination of the OFF-cell pause caused OFF-cells to be significantly more active following microinjection of morphine or kainic acid compared with baseline activity (F(2,38) = 95.46, P < 0.05). The effect of microinjecting kainic acid into the PAG on the tail flick-related activity of RVM ON- and OFF-cells did not vary when the injections preceded or followed morphine administration (Table 2).
|
|
| |
DISCUSSION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
The present data demonstrate that the antinociception produced by
morphine or kainic acid administration into the ventrolateral PAG is
associated with inhibition of RVM ON-cells and
activation of OFF-cells. Although an excitatory
response from PAG output neurons to RVM OFF-cells
has been described previously (Cheng et al. 1986;
Fang et al. 1989; Vanegas et al. 1984a),
the effect of PAG output neurons on RVM ON-cells
was unclear. The present findings demonstrate that inhibition, not
excitation is the net effect of PAG output neurons on RVM
ON-cells.
Previous research showed that activation of the PAG, whether by
microinjection of glutamate or electrical stimulation, excited RVM
neurons indiscriminately (Behbehani and Fields 1979;
Vanegas et al. 1984a). Unfortunately, these studies were
inconclusive because either ON- and
OFF-cells were not characterized
(Behbehani and Fields 1979)-ON-
and OFF-cells were first characterized in 1983 (Fields et al. 1983)-or electrical stimulation was used
(Vanegas et al. 1984a). Electrical stimulation activates
both PAG neurons and fibers of passage, making it impossible to tell
whether excitation of ON-cells is mediated by
neurons within the PAG. In addition, electrical stimulation of the PAG
has been shown to antidromically activate RVM
ON-cells (Gao et al. 1997) so
previous reports of excitation may not be physiologically relevant. The
present study overcame these problems by recording the activity of RVM
ON- and OFF-cells following
microinjection of kainic acid, an excitatory amino acid, into the PAG.
When PAG neurons are activated by kainic acid, RVM
ON-cells are inhibited and
OFF-cells are activated.
Direct comparison of the effect of microinjecting morphine and
kainic acid into the same PAG site demonstrated that both drugs affect
the activity of RVM neurons in the same way: RVM
OFF-cells become continuously active and
ON-cells are inhibited. This finding is
consistent with many other reports showing that administration of
opioids into the PAG inhibits RVM ON-cells and
excites OFF-cells (Barbaro et al.
1986; Cheng et al. 1986; Fang et al.
1989; Fields et al. 1983; Morgan et al.
1992). Although excitatory amino acids (EAAs) activate PAG
output neurons directly, whereas opioids disinhibit output neurons
(Chieng and Christie 1994; Depaulis et al.
1987; Moreau and Fields 1986; Vaughan and
Christie 1997; Vaughan et al. 1997), the present
findings suggest that EAAs and opioids produce antinociception through
the same output neurons. Disinhibition of PAG output neurons by
microinjection of the GABA antagonist bicuculline also inhibits
ON-cells and increases the activity of
OFF-cells (Moreau and Fields
1986). Previous studies comparing the antinociceptive effects
of microinjecting morphine and EAAs into the PAG also suggest that both
classes of drug produce antinociception through the same mechanism
(Jensen and Yaksh 1989). However, a report showing that
EAA antagonists in the RVM block the antinociception produced by PAG
administration of opioids, but not glutamate suggests there may be
subtle differences (van Praag and Frenk 1990), although this difference could be caused by differences in antinociceptive efficacy, not mechanism.
The present data are consistent with anatomical,
electrophysiological, and behavioral studies indicating that an
excitatory glutaminergic pathway transmits information from the PAG to
RVM (Aimone and Gebhart 1986; van Praag and Frenk
1990; Wiklund et al. 1988). Analysis of
conduction velocities suggests that this excitatory connection running
from the PAG to the RVM is monosynaptic (Shah and Dostrovsky
1980; Wiklund et al. 1988). Our data indicate that this direct excitatory connection from the PAG is to RVM OFF-cells specifically. Activation of PAG output
neurons causes RVM ON-cells to be inhibited, not
excited. It is not known whether this inhibition is the result of a
direct or indirect connection. One possibility is that
OFF-cells inhibit ON-cells.
In contrast, it is unlikely that ON-cell
inhibition is necessary for OFF-cell activation.
Recent work shows that microinjection of a glutamate antagonist
prevents the ON-cell burst without preventing the
OFF-cell pause (i.e., the
ON-cell burst is not necessary for the
OFF-cell pause) (Heinricher and
McGaraughty 1998).
An understanding of the relationship between the PAG and RVM is
complicated by regional differences within the PAG. Although antinociception can be produced at sites throughout the PAG
(Jacquet and Lajtha 1974; Jensen and Yaksh
1986; Morgan and Whitney 2000; Morgan et
al. 1998), the characteristics of this antinociception differ
between ventrolateral and lateral/dorsal locations in the PAG
(Cannon et al. 1982; Fardin et al. 1984;
Morgan 1991; Morgan and Liebeskind 1987;
Morgan et al. 1988, 1989; Prieto et al.
1983; Thorn et al. 1989; Tortorici et al.
1999). The data reported here derive only from microinjections
into the ventrolateral PAG. The previously reported differences between
the ventrolateral and lateral/dorsal PAG suggest there are differences
in the output of these regions. Whether this difference includes unique
connections to RVM ON- and
OFF-cells is not known. Thus it is possible that output neurons from the lateral PAG may excite RVM
ON-cells.
Finally, microinjection of kainic acid was much more likely to produce
antinociception than microinjection of morphine. This difference is
unlikely to be the result of dose or methodology because our morphine
dose was quite high (5 µg/0.4 µl) and our microinjection procedure
resulted in a 92% success rate when kainic acid was injected into the
same sites. Our relatively low success rate in producing
antinociception following morphine microinjections into the PAG (50%)
is surprising given previous reports stating that there is no
difference in efficacy following microinjections of morphine and EAAs
(Jensen and Yaksh 1989). However, several studies report
success rates similar to ours (Mohrland and Gebhart 1980; Morgan et al. 1998). In fact, in vitro
recordings from PAG neurons revealed that morphine is surprisingly
ineffective in altering the postsynaptic activity of PAG neurons
(Vaughan et al. 1997). One possible explanation for this
difference in efficacy is that opioids activate PAG output neurons by
disinhibition whereas EAAs activate output neurons directly. It should
be noted, however, that efficacy could be reduced by anesthetic effects
of morphine (see Heinricher et al. 1994).
In summary, our data indicate that the excitatory connection between the ventrolateral PAG and RVM is directed at OFF-cells specifically. Moreover, direct excitation of this pathway as occurs with microinjection of kainic acid is much more likely to produce antinociception than following microinjection of morphine. These data also highlight the close relationship between changes in the activity of RVM neurons and changes in nociception measured behaviorally.
| |
ACKNOWLEDGMENTS |
|---|
We thank C. S. Robbins for excellent technical assistance and E. González from the Instituto Venezolano de Investigaciones Cientificas photography department for cooperation.
This work was supported in part by funds provided for medical and biological research by the State of Washington Initiative Measure No. 171 to M. M. Morgan. V. Tortorici was supported by a postdoctoral fellowship from the Venezuelan Consejo Nacional de Investigaciones Científicas y Tecnológicas (CONICIT).
| |
FOOTNOTES |
|---|
Address for reprint requests: V. Tortorici, Centro de Biofísica/Bioquímica, 8424 NW 56 Street, Suite CCS 00202, Miami, FL 33166 (E-mail: victort{at}cbb.ivic.ve).
Received 4 February 2002; accepted in final form 3 June 2002.
| |
REFERENCES |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
This article has been cited by other articles:
|
|
 |
|
  S. Maione, K. Starowicz, L. Cristino, F. Guida, E. Palazzo, L. Luongo, F. Rossi, I. Marabese, V. de Novellis, and V. Di Marzo Functional Interaction Between TRPV1 and {micro}-Opioid Receptors in the Descending Antinociceptive Pathway Activates Glutamate Transmission and Induces Analgesia J Neurophysiol, May 1, 2009; 101(5): 2411 - 2422. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Maione, T. Bisogno, V. de Novellis, E. Palazzo, L. Cristino, M. Valenti, S. Petrosino, V. Guglielmotti, F. Rossi, and V. D. Marzo Elevation of Endocannabinoid Levels in the Ventrolateral Periaqueductal Grey through Inhibition of Fatty Acid Amide Hydrolase Affects Descending Nociceptive Pathways via Both Cannabinoid Receptor Type 1 and Transient Receptor Potential Vanilloid Type-1 Receptors J. Pharmacol. Exp. Ther., March 1, 2006; 316(3): 969 - 982. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
) following morphine microinjection and
those not producing tail flick inhibition (
). All
injections were on the right side. There was no difference in the
location of effective and ineffective injection sites.
Bottom: location of the rostral ventromedial medulla
(RVM) ON- (
) and
OFF-cells (
).
ON-cells are presented on the right and
OFF-cells on the left even though the neurons were
evenly distributed. Filled triangles represent cell recordings
accompanied by tail flick inhibition following morphine administration.
Distance from the interaural line is indicated below each figure
(Paxinos and Watson 1998
) or not (
TF) and the
ON-cell burst. Bottom: single
oscilloscope traces clearly show the change in the
ON-cell burst and tail flick reflex (bottom bar,
arrow) to noxious radiant heat before and after administration of MOR
and KA. Heat onset is indicated by the vertical bar at the beginning of
the tail flick trace. Sweep time is 20 s.
