| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
W. M. Keck Foundation Center for Integrative Neuroscience (B.K.T., M.A.P., A.I.B.) and Departments of Anatomy (B.K.T., M.A.P., A.I.B.) and Physiology (S.F.A., M.F.D., A.I.B.), University of California, San Francisco, San Francisco, California 94143-0452
Address all correspondence and requests for reprints to: Brad Taylor, Ph.D., Assistant Research Professor, Department of Anatomy, University of California, San Francisco, Box 0452, San Francisco, California 94143-0452. E-mail: brad{at}phy.ucsf.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Although there is considerable information regarding stress-induced activation of the pituitary-adrenal system, the contribution of pain-induced activation of this system to the control of nociception and inflammation remains unclear. Indeed, it is still not clear whether the nociceptive component of the formalin stimulus (independent of other components such as arousal, the stress associated with handling, and stimulation of high-threshold primary somatosensory afferents) is an adequate stimulus of the pituitary-adrenal system. Although many studies suggest that noxious stimuli activate the hypothalamo-pituitary-adrenocortical (HPA) axis, (9, 10, 11, 12, 13, 14, 15, 16, 17, 18), Bereiter and colleagues found that neither electrical stimulation of the tooth pulp nor thermal stimulation of the cornea increased the plasma concentration of ACTH (although they did show that the noxious stimuli potentiated the ACTH release produced by hemorrhage) (19, 20). Because the use of general anesthetics in these studies may have eliminated a direct effect of noxious stimulation on the release of pituitary-adrenocortical hormones, we have directly tested the hypothesis that the nociceptive component of a painful stimulus is sufficient to activate the pituitary-adrenocortical system in the awake animal. To this end, we evaluated the effects of noxious stimulation (injection of formalin into the hindpaw) on pain-related behavior, plasma ACTH, and B in the awake rat.
Unlike traditional reflex tests of nociception (e.g. tail-flick, hot-plate), pain produced by the hindpaw injection of formalin results from persistent tissue damage and, thus, more closely resembles clinical pain conditions (21). Another important feature of the formalin test is its biphasic nature; the first and second phases of the formalin response are thought to represent acute and persistent nociception, respectively (1, 2). Both phases are readily quantifiable and are conserved across several measures, including pain-related behavior (22), increases in arterial pressure (23), and neuronal activity of primary afferent axons (24) and spinal cord neurons (25). In the present study, we obtained multiple measurements of ACTH and B to evaluate pituitary-adrenocortical activity during both acute and persistent phases of the formalin response. Next, because B can reduce the inflammation that is frequently associated with peripheral hyperalgesia and pain (26, 27) via a negative feedback regulation, further studies evaluated the contribution of formalin-evoked increases in B to the magnitude of behavioral and cardiovascular responses.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Catheterization
To minimize pituitary-adrenocortical activation associated with
handling, we used remote iv catheters to sample the plasma for ACTH and
B and to measure blood pressure and heart rate. Venous jugular, carotid
arterial, and femoral arterial catheters were constructed by
heat-fusing a 4.2-, 3.2-, or 4.0-cm length of PE-10 polyethylene tubing
(Becton Dickinson, Sparks, MD) to a 6.0-, 6.0-, or 17.5-cm length of
PE-50 tubing, respectively. The catheters were inserted under
pentobarbital anesthesia (50–60 mg/kg). Vessels were isolated by blunt
dissection, with care taken to avoid injury to the sciatic or vagus
nerves. Through a small slit cut into the right jugular vein, the left
carotid artery, or the femoral artery, a catheter (prefilled with 100
IU/ml heparin) was advanced proximally to the vena cava, aortic arch,
or renal bifurcation of the abdominal aorta, respectively. After
securing the catheter to the vessel with 4–0 suture, the PE-50 end of
the catheter was tunneled under the skin, exteriorized at the nape, and
sutured to the dorsal neck muscles. After recovery from anesthesia, the
animals were returned to their cages and allowed to recover for a
minimum of 3 days before testing.
Adrenalectomy (ADX)
Bilateral (n = 7) or sham ADX (n = 9) was performed
just before femoral arterial catheterization. To maintain normal plasma
corticosterone levels until testing, and thus reduce the long-term
effects of ADX [e.g. changes in neuropeptide content
(28)], adrenalectomized rats were supplied with corticosterone (25
µg/ml) and 0.5% NaCl in their drinking water. This procedure permits
the normal circadian rise in B that occurs during the dark cycle, when
rats consume 70–75% of their daily fluid intake (29). The drinking
water was removed on the morning of testing, at least 3 days after
surgery. Testing began approximately 2 h after removal of the
drinking water, sufficient time for B levels to decline to low and
steady levels (30).
Dexamethasone (DEX) pretreatment
At least 2 days after femoral arterial catheterization (the
evening before formalin injection), rats were sc injected with either
DEX (250 µg/kg, n = 12) or saline (n = 12). A second
injection of DEX (250 µg/kg) or saline was administered 2 h
before formalin injection.
Experimental protocol
Because adaptation to the test environment not only allows
hormone levels to reach baseline, but also decreases the variability
associated with behavioral measurement in the formalin test (21), each
animal was transferred to a bedded 10 x 10 x 10-inch
Plexiglas box in the laboratory, with food and water (or 0.5% NaCl)
provided ad libitum, at least 16 h before testing
(
1700 h). After this acclimation period, the arterial and/or venous
catheters were connected, via PE-50 tubing, to a pressure transducer
(Kobe, Arvada, CO) and/or an infusion pump (CMA/100, Carnegie-Mellon).
Cardiovascular recording began at least 20 min later; this time period
allows MAP and heart rate to return to a resting state after the
sympathetic activation produced by handling (31). Next, animals
received a 50 µl sc injection of either sterile saline or formalin
(37%, wt/wt, formaldehyde, diluted in 0.9% saline) with a 30-gauge
needle into the plantar surface of the right hindpaw (23). Each animal
was used only once, i.e. neither saline nor formalin
injection was repeated in the same animal.
Recording of formalin response data
Both pain-related behavior and cardiovascular responses are
reliable measures of the central transmission of nociceptive signals in
the formalin test (21). To quantify flinching and licking behavior
during phase 1, the number of flinches or the number of seconds spent
licking during the second and third minute after injection were counted
as previously described (32). From 8–90 min, flinches and time spent
licking were counted for 2 min at 5-min intervals. These numbers were
divided by 2 to yield values per min. With this method, behavior in two
animals was simultaneously recorded by one observer at 1–2, 2–3, and
then 8–10, 13–15 .. 88–90 min after the formalin injection. To
quantify cardiovascular responses, arterial pressure waveforms were
first amplified and then conditioned by a digital blood pressure
analyzer (Micromed, Inc., Louisville, KY) to yield mean arterial
pressure (MAP) and heart rate. Data points were collected at 1-min
intervals, each of which represents 5 seconds of processed information.
Inflammation was assessed 90 min after the injection of formalin, by
measuring paw thickness with a microcaliper (Mitsutoyo, Kanagawa,
Japan), which is accurate to 0.01 mm.
Remote collection of blood
To obtain baseline measurement of ACTH and B after overnight
acclimation to the test environment, 140-µl samples of blood were
remotely withdrawn, using a venous jugular or carotid arterial
catheter, 15 and 5 min before the injection of saline (n = 8) or
5.0% formalin (n = 11). To estimate the stimulus-intensity
dependence of the hormonal responses, we also evaluated the effects of
1.25% in four animals. To determine the temporal profile of
formalin-evoked changes in hormone levels, additional 140-µl samples
were taken at 2, 10, 15, 30, 45, 60, and 75 min after formalin
injection. After each sample was collected, the volume was replaced
with saline.
Analysis of ACTH and B
Samples were centrifuged, and the plasma was collected and
frozen at -80 C. After thawing, ACTH was determined by RIA (33), with
antiserum R7 generously provided by Dr. W. Engeland. The antiserum is
used on unextracted plasma (5–50 ul) at a working dilution of
1:165,000. The assay incorporates the use of late trace addition of
125I (Incstar, Stillwater, MN), second antibody separation
(1:10 goat antirabbit globulin; Antibodies, Inc., Davis, CA), and a
buffer wash immediately before centrifugation. The ACTH assay is linear
between 2.2–22.0 pg/tube, with inter- and intraassay coefficients of
variation of 4.7% and 8.4%, respectively.
B was determined by RIA (34) or with an RIA kit (ICN Pharmaceuticals, Costa Mesa, CA) used at half-volume. These assays yield similar standard curves [ED 80 = 65; ED 50 = 311; and ED 20 = 1554 ng/tube (35)].
In the first case, B antiserum (Endocrine Sciences B3–163) was used with 125I-corticosterone (ICN Pharmaceuticals). The B assay is linear between 0.065–1.44 ng/tube, with inter- and intraassay coefficients of variation of 6.9% and 7.1%, respectively.
Data analysis and statistics
After the animals had acclimated to the test environment, five
sequential steady-state baseline cardiovascular values were recorded.
The average of these baseline values was then subtracted from each
poststimulus value, to yield changes in MAP or heart rate at each of
the timepoints after the formalin injection. Formalin-evoked flinching,
cardiovascular, and hormonal responses were analyzed by two-way
repeated-measures ANOVA, with Group as the between-subjects variable
and Time as the repeated measure. Behavioral and cardiovascular
responses during phase 1 (min 1–5), interphase (min 11–15), and phase
2 (min 20–60) were separately evaluated. When significant, these
analyses were followed by post hoc t-tests. Baseline
(preformalin) cardiovascular and hormonal values were compared between
groups with unpaired, independent t tests.
Materials
Stock solutions of formalin (aqueous solution of 37%, wt/wt,
formaldehyde; Fisher, Fair Lawn, NJ) were freshly diluted in 0.9%
isotonic sterile saline (Baxter Healthcare Corp., Deerfield, IL).
Pentobarbital was obtained from Abbott Laboratories (North Chicago,
IL). DEX sodium phosphate was obtained from American Reagant Labs
(Shirley, NY).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
As illustrated in Fig. 1A
, sc injection
of dilute formalin into the hindpaw evokes an early-onset behavioral
response [termed: phase 1 (22)] that lasts approximately 5 min and,
after a brief quiescent interphase, a second prolonged behavioral
response (termed: phase 2) that lasts approximately 50 min. Injection
of saline into the hindpaw did not produce flinching or licking
behavior. As illustrated in Fig. 1B, however, saline injection produced
a peak increase in ACTH (to 113 ± 18 pg/ml) at 2 min, which
gradually returned to baseline values. In contrast, 5% formalin
produced significantly greater increases in ACTH [F(1, 25)
= 11.3, P < 0.005], with significant post
hoc differences at 2, 30, 45, and 60 min (P <
0.01). Fig. 1C also illustrates that saline injection produced a peak
increase in B (to 12.1 ± 3.5 µg/dl), at 15 min, that gradually
returned to baseline values. In contrast, 5% formalin produced
significantly greater and longer duration increases in B
[F(1, 17) = 5.7, P < 0.05], with
significant post hoc differences at 30, 45, and 60 min
(P < 0.05).
|
shows that all of
the groups displayed mean baseline values of ACTH and B that were below
100 pg/ml and 2.0 µg/dl, respectively. These values indicate that the
groups experienced minimal stress before the injection.
|
, both
concentrations of formalin produced greater responses in all measures,
compared with baseline values or saline injection. Fig. 2, A and B,
also illustrate that the average magnitude of ACTH and B responses did
not increase as stimulus intensity was raised from 1.25% to 5.0%
formalin. In contrast, Fig. 2, C and E, illustrate that the magnitude
of flinching, licking, and inflammatory responses increased with
formalin concentration (P < 0.05).
|
shows that thymus weight of the Sham and ADX + B-water groups did not
differ, indicating that the latter group was eucorticoid and that
corticosteroid replacement was successful. As illustrated in Fig. 3, ADX + B did not significantly change
flinching behavior, licking behavior, MAP responses, or heart rate
responses during phase 1 or phase 2 (P > 0.05). Table 1 also shows that ADX + B did not significantly change formalin-evoked
inflammation.
|
shows that DEX pretreatment
decreased thymus and adrenal weights. This indicates that the DEX group
was hypercorticoid (38). Table 1 also shows that DEX decreased the
inflammation (paw thickness) associated with formalin injection
(P < 0.05). As illustrated in Fig. 4, DEX did not change flinching or
licking behavior or MAP responses (P > 0.05).
Surprisingly, DEX increased the magnitude of formalin-evoked heart rate
responses during phase 2 [F(1, 13) = 5.3, P
< 0.05]; however, the fact that DEX also reduced baseline heart rate
(Table 1) shows that absolute heart rate (but not change in heart rate)
during formalin testing was not different between the saline and DEX
groups.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Our study is the first to evaluate plasma ACTH and B, during both acute and persistent nociception, in a widely accepted model of inflammatory pain, the formalin test. Although other studies have shown that formalin increases the plasma concentration of ACTH (16, 17, 18), formalin was injected at unusual concentrations (i.e. 10%) (16) or at unusual sites [such as the ip space (18), or the thigh (17)], and ACTH measurements were not ascertained before 30 min (i.e. during the first phase of the formalin response) (16, 17, 18). Surprisingly, Aloisi et al. (16) reported that the intraplantar injection of 10% formalin did not change the plasma concentration of B. Because control levels of B were much lower in our study (<2.0 ug/dl) than in the report of Aloisi et al. (16) (>7.0 ug/dl), we believe that a marked increase in baseline adrenocortical activity in the latter study may have contributed to this difference. Based on our results and on clinical studies showing that plasma levels of ACTH and cortisol are elevated during the pain associated with surgery (10, 11, 12, 13), trigeminal neuralgia (14), or cluster headache (15), we conclude that nociceptive input activates the pituitary-adrenocortical system.
In light of our results, it is a conundrum that Bereiter and colleagues (19, 20) found that neither electrical stimulation of the tooth pulp nor thermal stimulation of the cornea increased the plasma concentration of ACTH. The discrepancy between these results and ours may be related to the different types of noxious stimuli used and to the use of general anesthesia in the Bereiter study. In fact, because general anesthetics not only decrease the transmission of nociceptive signals (23), but also may alter the release of ACTH and corticosterone after stimulation with insulin (40) or ether (41), it is possible that anesthesia reduces noxious stimulus-evoked changes in pituitary-adrenocortical activity.
At least two physiological mechanisms could underlie the increase in plasma ACTH and B after intraplantar formalin injection. First, similar to an adjuvant-induced arthritis model of chronic inflammation (9, 42), Aloisi et al. (16) found that formalin increased the blood-borne release of interleukin-6. In addition to other cytokines (such as interleukin-1), interleukin-6 is capable of increasing ACTH and corticosterone via several mechanisms. For example, cytokines can directly stimulate hypothalamic corticotropin-releasing factor release from neuronal cell bodies in the paraventricular nucleus of the hypothalamus or from terminals in the median eminence into the hypophysial-portal circulation (42, 43). Second, the formalin stimulus robustly activates ascending pain pathways, which may, in turn, activate the HPA axis. For example, at the level of the primary afferent fiber, interruption of nociceptive transmission with neonatal capsaicin treatment (which destroys most C-fibers) decreases the ability of ip formalin (18), or the intraplantar injection of the inflammatory agent, complete Freund’s adjuvant (4), to increase ACTH. Also, interruption of nociceptive transmission at the spinal level decreases the ACTH response to noxious stimuli in dogs and humans (10, 12, 13). These studies indicate that both the behavioral and pituitary-adrenal responses may be mediated by formalin-induced activation of primary afferent C-fibers and ascending nociceptive spinal cord neurons. Because many of the above studies were performed in the anesthetized subject, however, future experiments should address the mechanisms by which noxious stimuli stimulate the HPA axis in the awake subject.
The present results suggest that the behavioral and pituitary-adrenal systems differ in their sensitivity to formalin concentration. We found that the maximal stimulus intensity for the behavioral/inflammatory responses was at least 5.0% formalin, whereas that of the pituitary-adrenal response was only 1.25%. Because other types of stressors, such as restraint, produce much larger plasma levels of ACTH (39) and B (17, 39, 44) than were found in the current study, the formalin effect is certainly submaximal. We speculate that a ceiling effect precludes the ability of higher formalin concentrations to evoke maximal pituitary-adrenal responses.
Adrenal activity does not mediate or modulate nociception in the
formalin test
Although B exerts antiinflammatory actions (45), and there is
evidence that inflammation contributes to nociception in the formalin
test (21), the present studies indicate that the formalin-evoked
release of endogenous corticosterone is not antinociceptive. First,
prevention of acute corticosteroid responses by ADX did not
significantly increase the flinching, licking, or MAP responses evoked
by formalin. Second, despite our demonstration that DEX acted on GRs to
decrease thymus and adrenal weight and formalin-evoked inflammation,
this treatment did not inhibit the flinching, licking, MAP, or heart
rate responses. Although some studies found that DEX decreases
formalin-evoked behavior, this effect only occurred at very high (6–30
mg/kg) doses (46, 47). Third, in preliminary studies, we have provided
evidence that chronic stress (exposure to 4 C for 4 h/day for 7 days),
which produces a central facilitation of the HPA axis (i.e.
exaggerated stimulus-evoked increases in corticosterone), does not
decrease flinching, licking, MAP, or heart rate responses after
formalin injection (48). We conclude that formalin-evoked release of
corticosterone does not feed back and reduce nociception.
Consistent with our results, lesions of the paraventricular nucleus of the hypothalamus, a key input to the pituitary, did not alter pain-related behavior during phase 2 of the formalin test (49, 50). However, because glucocorticoids seem to modulate bradykinin-associated inflammation (4, 51), neuropeptide content in doral root ganglia (28, 52) and noxious stimulus-evoked expression of c-fos within the spinal trigeminal nucleus, it remains possible that glucocorticoids contribute to the magnitude of nociceptive responses in models of pain other than the formalin test. Furthermore, even though glucocorticoids do not seem to modulate nociceptive processing in response to a single noxious stimulus (such as the injection of formalin), it is still possible that glucocorticoids regulate nociception in longer-duration models of chronic pain.
The present studies indicate that the formalin-induced increase in blood pressure and heart rate are not mediated by adrenomedullary catecholamine release. Consistent with this result, Casto et al. (53) found that ADX did not prevent a nonnoxious airpuff stimulus from evoking increases in blood pressure and heart rate. Further studies are needed to evaluate the spinal and supraspinal sites that contribute to the cardiovascular responses induced by noxious stimuli in awake animals.
In summary, the nociceptive component of intraplantar formalin injection activates the pituitary-adrenocortical system in the awake, freely-moving rat. However, the ensuing release of corticosterone does not feed back and reduce nociceptive processing.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received September 29, 1997.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  K. Vissers, H. Adriaensen, R. De Coster, C. De Deyne, and T. F. Meert A Chronic-Constriction Injury of the Sciatic Nerve Reduces Bilaterally the Responsiveness to Formalin in Rats: A Behavioral and Hormonal Evaluation Anesth. Analg., August 1, 2003; 97(2): 520 - 525. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Le Bars, M. Gozariu, and S. W. Cadden Animal Models of Nociception Pharmacol. Rev., December 1, 2001; 53(4): 597 - 652. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Pacak and M. Palkovits Stressor Specificity of Central Neuroendocrine Responses: Implications for Stress-Related Disorders Endocr. Rev., August 1, 2001; 22(4): 502 - 548. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. L. Yaksh, G. Ozaki, D. McCumber, M. Rathbun, C. Svensson, S. Malkmus, and M. C. Yaksh An automated flinch detecting system for use in the formalin nociceptive bioassay J Appl Physiol, June 1, 2001; 90(6): 2386 - 2402. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Piva, G. M. Horowitz, and K. L. Sharpe-Timms Interleukin-6 Differentially Stimulates Haptoglobin Production by Peritoneal and Endometriotic Cells in Vitro: A Model for Endometrial-Peritoneal Interaction in Endometriosis J. Clin. Endocrinol. Metab., June 1, 2001; 86(6): 2553 - 2561. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
