Role of spinal cord neuropeptides in pain sensitivity and analgesia: Thyrotropin releasing hormone and vasopressin

308

Bram Research, 362 (1986) 308-317 Elsevier

BRE 11352

Role of Spinal Cord Neuropeptides in Pain Sensitivity and Analgesia: Thyrotropin Releasing Hormone and Vasopressin L R WATKINS, S.N SUBERG, C L THURSTON and E.S CULHANE Department of Ammal Phystology, Untverstty of Cahfornta at Davts, Darts, CA 95616 (U S A ) (Accepted May 7th, 1985) Key words" vasopressln - - thyrotropm-releasmg hormone - - morphine - - non-opmte antlnoc~ceptlon - endogenous opiate antagonist

The existence of a wide variety of neuropeptldes within the spinal cord dorsal horn raises the question of their possible roles in sensory processing The present series of behavioral experiments examined the effects of mtrathecal (IT) administration of two such neuropeptldes, thyrotropin-releasmg hormone (TRH) and vasopressm (VAS), on pare sensitivity and antmoclceptlon TRH exerted no marked effect on basal pain sensitivity over the dose range examined (0.25 ng-2.5 gg) However, a U-shaped dose-response effect on morphine antinocicept~on (3/~g, IT) was observed, wherein potent attenuation, moderate attenuation, or enhancement of morphmemdueed antmociception was observed following the various doses tested In contrast, VAS produced non-opiate antmociceptlon at the highest doses tested (25 ng and 250 rig) and none of the VAS doses (0 25 ng-250 ng) appeared to interact with IT morphine (3 gg) antinoctception Lastly, IT TRH was not observed to interact with IT VAS antmocicept~on These data provide ewdence that these neuropeptides exert strikingly different effects on pam sensitiwty and opmte antinoocept~on, and provide mmal ewdence that TRH may be included in the growing list of neuropeptides that can act like endogenous optate antagomsts w~thln the central nervous system

INTRODUCTION Within recent years, a large n u m b e r of n e u r o p e p tides have been identified within the central nervous system due to the advent of sensitive radioimmunoassay and immunohistochemistry techniques (for review, see ref. 41). M o r e specifically, many of these neuropeptides have been d e m o n s t r a t e d within cell bodies and/or nerve terminals of the spinal cord. Enkephalins, substance P, neurotensin, and somatostatin, among others (for review, see ref. 41), have been identified within neurons intrinsic to the cord. While the cell bodies of origin of many of the peptidergic nerve terminals within this region remain to be demonstrated, a growing h t e r a t u r e is identifying a variety of peptidergic supraspinal nuclei which send projections to the spinal cord. Supraspinal origins for spinal cord terminals containing cholecystokinin octapeptide (CCK) 44.45,58, thyrotropin-releasing h o r m o n e ( T R H ) 29, vasopressin (VAS) 11,50,53,56,61, substance

p29,58, and enkephalin2, among others 41, have now been demonstrated. The localization of diverse n e u r o p e p t l d e terminals and cell bodies within the spinal cord dorsal horn is intriguing since this neural area is critically involved in the modulation and transmission of pain 2. The question therefore arises as to what role, if any, these pepttdes play in pain modulation To date, behavioral investigations of the effect of intrathecally admmistered n e u r o p e p t l d e s on pain have largely been limited to substance p15 24,52 endogenous opiate peptides 2,64, A C T H 17, and CCK16. 30. Of these, substance p15.24 and various opiate pepttdes 2,64 have been demonstrated to produce naloxone-reverslble antinoclception following spinal administration. Intrathecal A C T H , in accordance with Its biochemical characterizatlon as a partial opiate a g o n i s t - a n t a g o n i s t 63, can produce elevations In pain thresholds yet antagonizes fl-endorphm induced antlnociceptlon 17. Spmal CCK, while not exerting measurable effects on basal pare

Correspondence L R Watkms, Department of Ammat Physiology, Umverstty of Cahforma at Davis, Davis, CA 95616, U S A 0006-8993/86/$03 50 © 1986 Elsevier Science Pubhshers B V (Biomedical Dwlslon)

309 sensitivity (refs. 16, 27, 69, cf. ref. 30), appears to selectively attenuate opiate forms of antinociception 16. Therefore, even at this very early stage of investigation, neuropeptides of the spinal cord dorsal horn appear to exert distinct and diverse effects on pain sensitivity. Before a clear understanding of peptidergic involvement in spinal pain processing can be attained, systematic investigations will be needed to characterize the effect of various neuropeptides on basal pain sensitivity as well as to identify whether these peptides modulate the function of either opiate or nonopiate antinociceptive systems. The aim of the present series of experiments was to provide an initial behavioral investigation of two neuropeptides, T R H and VAS, which have been identified within nerve terminals of the spinal cord dorsal horn. GENERAL METHODOLOGY Adult male Sprague-Dawley rats (350-500 g, Simonsen Laboratories) were used in all experiments. Prior to their use, all animals were implanted with lumbosacral intrathecal (IT) catheters (see ref. 70 for methodology) and received 5 daily habituation sessions (approximately 1 h each) to the Plexiglas restraining cylinders used during test sessions. During the behavioral assessments, a modification 1 of the tail flick (TF) test 13 was used to determine pain responsivity. Baseline latencies (BL) were measured (average of 3 TF trials, 2 min intertrial interval) prior to drug administration, with the bulb voltage adjusted to attain a mean BL of approximately 3.0-3.5 s; the same bulb voltage was used for all animals throughout each experiment. Following baseline testing, rats were randomly assigned to the various drug groups and received 2 IT injections delivered 10 min apart (void volume of catheter = 10 bd; total injection volume = 21.5/A; each of the 2 injections occurred over approximately 30 s). The drugs and dosages injected are detailed in each experimental protocol (see below). Beginning 5 min after the second IT rejection, TF latencies were recorded at 5 min intervals through 40 min; behavioral testing was performed blind with respect to the drug received. For all tests, the radiant heat was automatically terminated at 8.0 s if no TF occurred, in order to avoid tissue damage.

The test latencies (TL) of each rat were expressed as a percent of maximal possible effect (%MPE), using the following equation21: %MPE = [(TL-BL)/(8.0-BL)] x 100 Statistical comparisons between drug groups were made using analysis of variance (ANOVA) to determine main effects. Paired t-tests were calculated for each group comparing their BL with latencies recorded after drug injections. EXPERIMENT 1 EFFECT OF IT TRH ON BASAL PAIN SENSITIVITY AND IT MORPHINE-INDUCED ANTINOCICEPTION Methods

The present experiment was designed to examine the effect of a range of IT T R H doses on pain sensitivity when IT T R H was delivered in the presence or absence of IT morphine. Six doses of T R H (delivered in 0.5/d saline; Beckman Laboratories) were examined: 0 (vehicle control), 0.25 ng, 2.5 ng, 25 ng, 250 ng and 2.5/~g. Each dose was delivered twice (10 min interval) and was then followed immediately by 3 ~g morphine sulfate or equivolume vehicle (0.5/A saline). This dose of morphine was chosen to produce submaximal antinociceptive effects in order to allow either enhancement or attenuation to be observed. Behavioral testing was according to the procedures detailed m General Methodology (see above). Results

When T R H was tested in the absence of morphine (Fig. 1A, B), neither 0.25 ng (n = 6), 2.5 ng (n = 9), 25 ng (n = 7) nor 250 ng (n = 8) T R H significantly elevated TF latencies above baseline level at any time tested. The highest dose of T R H tested (2.5/~g, n = 7) produced only a transient effect on pain sensitivity. The TF latencies of this group were significantly elevated above baseline only at 5 min after injection (paired t-test, P < 0.05). In contrast to the general lack of effect of T R H on basal pain sensitivity, T R H exerted marked effects on IT morphine analgesia (Fig. 1C, D). A U-shaped dose-response effect was observed with the lowest (0.25 ng, n = 10) and the highest (2.5/~g, n = 12) T R H doses producing the most profound attenuation

310

EFFECT OF TRH ON BASELINE

TRH DOSE • 2.5 ng o 25 ng

A 4O

B

TRH DOSE 2.5 pg o250 ng • 25 ng ®

40

hu2o 0

0

,

0

I0

20 TIME(MIN)

50

40

0

I0

J

20 TIME(MIN)

i

,

i

40

50

EFFECT OF TRH ON MORPHINE ANALGESIA TRH DOSE • 2.5 ng o .25 ng .Ong

I00

TRH DOSE

D

I0 0 _

.

80

L_--I " ~

60

• 2.5 #g

o 250 ng 25 ng

80

•

I

.Ong

60

,

T

I

I

3O

40

hJ o_ ~40

40 20

........f ......'l ........ ? .....T ........1 ....... 1 ........ 0

0 i

0

i

I0

i

i

i

20 TIME (MIN)

i

30

b

40

0

IO

2o TIME(MIN)

Fig. l. Effect of IT TRH on basal pain sensitivity (A, B, top) and IT morphine (3 pg) analgesia (C, D, bottom). The range of TRH doses tested IT (0.25 ng-2.5 #g) had no marked effect on tail-flick latencies in the absence of morphine (A, B). The lowest (0.25 ng, open circles in C) and highest (2.5 #g, half-filled circles in D) doses of TRH potently attenuated IT morphine-induced antinociception. The 25 ng and 250 ng TRH doses (filled triangles and open circles, respectively in D) moderately attenuated antinocicepfion whereas the 2.5 ng TRH dose (filled circles in C) consistently potentiated antinociception.

of morphine analgesia, 250 ng (n = 9) and 25 ng (n = 10) T R H producing a m o r e m o d e r a t e decrease in morphine analgesia, and 2.5 ng (n = 10) T R H producing a consistent p o t e n t i a t i o n of m o r p h i n e analgesia, c o m p a r e d to m o r p h i n e - i n j e c t e d controls (n = 9) ( A N O V A , P < 0.0001 for each group c o m p a r e d to controls). The attenuation p r o d u c e d by 25 ng and 250 ng T R H was not significantly different ( A N O V A , P > 0.7) but the attenuation p r o d u c e d by both of these groups was less than that observed following either 2.5 p g or 0.25 ng T R H ( A N O V A , P < 0.0005

for each comparison). C o m p a r i s o n of these latter two groups revealed that 2.5 p g T R H p r o d u c e d a m o r e p r o n o u n c e d attenuation of m o r p h i n e - i n d u c e d antinociception than 0.25 ng T R H ( A N O V A , P < 0.0001). In fact, in contrast to the m o r p h i n e - i n j e c t e d control group (no T R H ) in which T F latencies were significantly elevated at 10 and 2 0 - 4 0 min after injections (paired t-tests, P < 0.05 at each time point), the T F latencies of animals injected with 2 doses of 2.5 #g T R H plus 3 p g m o r p h i n e were not significantly elevated over baseline levels at any time tested (paired t-

311 tests). The TF latencies of the 0.25 ng T R H group were significantly (paired t-test, P < 0.05) elevated only at 5 min after completion of drug delivery. EXPERIMENT 2 EFFECT OF IT VAS ON BASAL PAIN SENSITIVITY AND IT MORPHINE-INDUCED ANTINOCICEPTION Methods

To determine whether VAS affects basal pain sensitwity or IT morphine analgesia, a range of VAS doses were examined either in the absence or presence of IT morphme. Six doses of VAS (dehvered in 0.5/~1 sahne; arginine vasopressin, Sigma Chemicals) were examined: 0 (vehicle control), 0.0025 ng, 0.25 ng, 2.5 ng, 25 ng and 250 ng. Each dose was delivered twice (10 min interval) and followed immediately by 3/~g morphine sulfate or equivolume vehicle (0.5/~1 sahne). Additional procedures to those detailed in General Methodology were performed in this experiment since pilot studies revealed that high doses of VAS could affect motor function as well as pain sensitivlty. In accordance with these initial observations, all animals injected with VAS were observed for any signs of myoclomc twitches ('scratching bouts', see ref. 68 for details), flaccidity, respiratory distress, and alterations in normal body posture and locomotion. Naloxone (10 mg/kg, i.p.) was injected immediately after the 40 min behavioral test session whenever maximal (8.0 s) antinociception was observed in order to determine whether VAS produces antmooception through opiate pathways. TF latencies of naloxone injected animals were then recorded at 5 min intervals for an additional 30 min. Results

When VAS was injected in the absence of morphine (Fig. 2), dose-dependent 'antinociception' was observed. Following 0.0025 ng (n = 8), 0.25 ng (n = 10), and 2.5 ng (n = 10) VAS, these increases m TF latencies were generally quite small. Significant increases in TF latencies above baseline values (paired t-tests) were observed at 5, 10, and 30 min for the 0.0025 ng VAS group, at 5, 10, and 30 min for the 0.25 ng VAS group, and at no time tested for the 2.5 ng VAS group. For the two highest doses, 25 ng (n = 10) and 250 ng (n = 9) VAS, TF latencies were significantly elevated over baseline at all times tested.

Maximal (8.0 s) TF latencies were recorded throughout the 40 min test session for 4 of 10 rats in the 25 ng VAS group and 6 of 9 rats in the 250 ng VAS group. These animals received 10 mg/kg naloxone at the end of the 40 min test session and then observed for an additional 30 min (see methods above). At no time during this additional test session did any of these animals have a TF latency of less than 8.0 s. In addition to 'non-opiate antinoclception', other observations were made which suggest that IT VAS exerts effects on motor systems which may cloud the interpretation of the tail flick results. While no sign of motor disturbances was observed following the 3 lowest doses of VAS (0.0025, 0.25 and 2.5 ng), motoric effects were noted for the highest two doses. Of the 10 animals injected with 25 ng VAS, 4 exhibited 'scratching bouts' similar to the convulsive effects observed following high (i.e. 25/~g) doses of IT morphine 68. Of the 9 animals injected with 250 ng VAS, 5 exhibited 'scratching bouts', one exhibited transient difficulty in breathing, and one exhibited transient hindbody flaccidity in addition to transient difficulty in breathing. The 'scratching bouts', faccidity and respiratory effects appeared within 2-10 min after the first VAS injection and lasted 10-15 min. All animals appeared motorically normal by the time TF trials were initiated. It should be emphasized that no correlation could be seen between these dysfunctions and 'antinociception'. Of the 4 animals exhibiting 'scratching bouts' following 25 ng VAS, 2 animals exhibited maximal (8.0 s) TF latencies after injection and 2 exhibited no antinoctception. Of the remaining 6 animals in that group which exhibited no 'scratching bouts', 2 were not analgesic, 3 were moderately analgesic, and one was maximally analgesic. Of the 5 animals exhibiting 'scratching bouts' following 250 ng VAS, 2 animals exhibited maximal TF latencies after injection and 3 exhibited no or low levels of antinociception. Of the remaining 4 animals in that group, all 4 were maximally analgesic, including the 2 animals exhibiting transient breathing dysfunction and transient hlndbody flacodlty and the 2 animals exhibiting no observable motoric dysfunction. The 6 animals of the 250 ng VAS group exhibiting maximal TF latencies were briefly removed from their Plexiglas restraining cylinders at 11 and 31 min after their second VAS injection (i.e. after the 10 and 30 m m TF latencies were recorded). All animals demonstrated nor-

312 A

B

250 ng VAS I00

I00

80

80

60

60

ILl

25 ng VAS

;_ ....

i... I

5

I0

w

n 40

a_ 40

z~ 20

20

0

0

-20

-20 I

5

i

I0

I

i

15

I

20 25 TIME (MIN)

i

i

I

50

55

40

C

D

2.5 ng VAS

I00

15

20 25 TIME(MIN)

30

35

40

35

40

.25 ng VAS

I00 8O

8O

; .... .......

6O w

A-"t ......I

"'1"1

6O w

a_ 40

a_ 40 2O

o~ 20

0

0

-20

-20 5

I0

15

20 25 TIME (MIN)

50

.'.35 4 0

5

I0

15

20 25 TIME (MIN)

30

.0025 ng VAS

I00 80 60

o MOR+SAL

w

a_ 40 2; 20

•" MOR+ VAS r- .... r .... 1:..... 1 ..... f"

1

.L

35

40

• SAL+ VAS

0 -20 5

I0

15

20 25 TIME (MIN)

5(3

Fig. 2. Effect of IT VAS on basal pain sensitivity and IT morphine-induced antinociception. In contrast to IT morphine (3 l~g, open circles in A - E ) , low doses of IT VAS (0.25-2.5 ng, filled triangles in C - E ) did not markedly affect pain sensitivity. High doses of IT VAS (25 ng and 250 ng, filled triangles in A, B) produced significant elevations in tail-flick latencies. VAS and morphine antinociceptions were not additive in that the latencies observed for the VAS plus morphine groups (open triangles in A - E ) were not significantly different from either the morphine alone (open circles in A - E ) or the VAS alone (filled triangles in A - E ) groups, whichever demonstrated greater antinociception.

mal posture and locomotion at both times tested. At 11 min, 4 responded (vocalization and escape) to manual tail pinch; at 31 min, one animal responded to tail pinch. The remaining animals in this group gave no sign of perceiving this noxious stimulus.

When VAS was combined with morphine, a complex effect was observed. The most striking observation can be seen in Fig. 2; that is, 'antinociception' induced by VAS and antinociception induced by morphine were not additive. For every dose of VAS

313 tested, the data recorded for the VAS plus morphine group closely paralleled the data of either the morphine alone group or the VAS alone group, whichever exhibited the greater increases in TF latencies. Specifically, no significant differences were observed between the morphine control (no VAS) group and either 0.0025 ng VAS plus morphine ( A N O V A , P > 0.05, Fig. 2E), 0.25 ng VAS plus morphine (ANOVA, P > 0.5, Fig. 2D), 2.5 ng VAS plus morphine (ANOVA, P > 0.9, Fig. 2C), or 25 ng VAS plus morphine (ANOVA, P > 0.05, Fig. 2B). With the exception of the 25 ng VAS groups ( A N O V A , P > 0.1), each of these VAS plus morphine groups were significantly more analgesic than the group receiving an equal dose of VAS without morphine (ANOVA, P < 0.0001 for each comparison). Additionally, the 250 ng VAS (no morphine) group was not significantly different from the 250 ng VAS plus morphine group (ANOVA, P > 0.35, Fig. 2A). This latter group (250 ng VAS plus morphine) did exhibit enhanced 'antinociception' compared to morphine alone (ANOVA, P < 0.0001). EXPERIMENT 3. E F F E C T OF I T TRH ON NON-OPIATE ANTINOCICEPTION INDUCED BY I T VAS Methods

In the first two experiments, T R H and VAS were found to exert very different effects following IT adIOO T

SALINE

ministration. T R H was observed to either enhance (2.5 ng T R H ) or abolish (2.5/~g T R H ) IT morphineinduced antinociception, yet have no marked effect on basal pain sensitivity in the absence of morphine. In contrast, VAS (25 ng) appeared to produce nonopiate analgesia without as marked motoric 'sideeffects' as 250 ng VAS. The present experiment was designed to determine whether analgesia induced by 25 ng VAS would be affected by either 2.5 ng or 2.5 /~g TRH. Three groups of rats were tested. All animals received 2 doses of 25 ng VAS in 0.5/A saline (10 rain interval). Each dose was immediately followed by either 2.5 ng T R H (n = 10), 2.5 ttg T R H (n = 9) or equivolume saline (0.5/4, n = 9). The total injection volume delivered to each rat was 22 M. Behavioral testing was in accordance to the procedures described in General Methodology. Results

In accordance with experiment 2, 25 ng VAS produced reliable analgesia following IT administration (Fig. 3) with TF latencies being significantly elevated over baseline at all times tested (paired t-tests, P < 0.05 for each comparison). However, in contrast to the effects of T R H on IT morphine-induced antinociception (Fig. 1), neither 2.5 ng T R H nor 2.5~g T R H significantly affected VAS-induced antinociception (ANOVA, P > 0.2 for both comparisons; see Fig. 3). DISCUSSION

C~. 0 0 2 5 pg TRH

80

• 2.5 pg TRH

6o LLI

a_ 4 0 2O

-20 5

10

15 20 25 30 55 40 TIME(MIN) Fig. 3. Effect of two doses of IT TRH on IT VAS antinociception. Since the doses of TRH which potentiated (0.0025~g) and abolished (2.5 #g) morphine-induced antinociception had no effect on VAS-induced antinociception, it appears that TRH does not influence this form of non-opiate antinociception. Symbols: open circles = 25 ng IT VAS plus saline vehicle; open squares = 25 ng IT VAS plus 0.0025 ~g TRH; filled circles = 25 ng IT VAS plus 2.5/~g TRH.

The present series of experiments indicate that VAS and T R H can exert very different behavioral effects following intrathecal (IT) administration onto the lumbosacral spinal cord of awake rats. T R H generally exerted no observable effect on basal pain sensitivity when injected IT in log doses ranging from 0.25 ng to 2.5/~g (each dose delivered twice, 10 min apart). However, these same T R H doses exerted striking effects on analgesia induced by morphine (3 /~g IT). A U-shaped dose-response effect was noted in that morphine-induced antinociception was profoundly attenuated by 0.25 ng and 2.5/~g T R H , moderately attenuated by 25 ng and 250 ng T R H , and enhanced by 2.5 ng TRH. In contrast to the effects induced by T R H , VAS can alter responsiveness to radiant heat stimuli in the absence of morphine. While

314 low doses of VAS (0.0025, 0.25 and 2.5 ng; each VAS dose delivered twice, 10 min apart) had no observable effect on pain responsiveness, 25 and 250 ng VAS produced marked increases in TF latencies which were not affected by 10 mg/kg naloxone 0.P.). Additionally, increases in TF latencies induced by VAS and morphine were not additive. For every dose of VAS tested, the TF latencies of the VAS plus morphine group closely approximated the latencies of either the morphine (no VAS) group or the VAS (no morphine) group, whichever exhibited the greater latency increases. Since the doses of T R H which enhanced (2.5 ng) and abolished (2.5/zg) morphineinduced antinociception had no effect on the TF latency increases produced by VAS, these data provide initial evidence that T R H may selectively interact with opiate antinoclceptive systems at the level of the spinal cord. Taken together, these experiments provide new insights into the behavioral effects which may be exerted by neuropeptldes of the spinal cord dorsal horn. While 25 ng and 250 ng VAS were observed to produce marked increases in TF latencies, a cautionary note is warranted with regard to the nature of this effect. Similar to previous reports using intracerebroventricular VAS administration40. 41, hlndhmb 'scratching bouts' were observed m about half (9/19) of the animals rejected with these doses which resembled myoclomc twitches 68 Of the 9 ammals rejected with the highest VAS dose, one exhibited respiratory distress and one exhibited hindbody flaccidity in addition to respiratory difficulty. These results are similar to a recent report by Millan et al.50 that i> 20 ng VAS produced hmdlimb flaccidity and respiratory disturbances (but, apparently, no 'scratching bouts') m some rats after lumbosacral IT injection. In the work of Millan et al. 50. a perfect correlation was observed between flaccidity (which lasted throughout the testing period) and increased TF latencles, leading to their conclusion that VAS doses they tested do not produce antlnociceptlon Although the bases for the &screpancies between the present study and that of Millan et al. 50 remain unclear, our data do not support the conclusion drawn by Millan et al. 50. In the present study, all of the observed motorIc effects were transient, lastmg 15 min or less after the first VAS mjectLon. Since behavioral testing began 15 rain after the first injection and continued for an addmon-

al 40 min, testing was conducted during a period when no motoric dysfunctions were observable. Additionally, tail pinch 'antinociception' appeared later than maximal 'antlnociception' to radiant heat in 4 of 6 rats tested, indicating that increases in TF latency could not be accounted for by motor suppression. From these indications, and the lack of correlation between rats exhibiting increased TF latencies and transient motor disturbances, it appears that, in the present study, the increases in TF latencies by VAS may indeed be attributable to non-opiate mediated alterations m pain sensitivity rather than generalized motor suppression. The fact that VAS ~s concentrated in the substantia gelatinosa as well as being found within the ventral horns of the spinal cord may potentially account for the sensory and motoric effects observed following IT VAS administration. Since destruction of the paraventricular nuclei, a major though not sole source of spinal cord VAS8. 50, alters neither basal pain responsivity nor gross motor function8. 50, VAS from this hypothalamic region does not appear critical for tonic modulation of sensory or motor spinal cord function. Although environmental stimuli are known to induce release of peripheral and/or central vasopressin 33-36.39, it remains to be determined whether spinal cord VAS Is released by pharmacological agents or environmental stimuh71 which produce antinociception. However, in this respect, it is interesting to note that cold water swims induce s~gnificantly less nonopmte antinociceptlon m Brattleboro rats (genetically VAS deficient) than in normal ammals 9. The observation of non-opmte analgesia following IT VAS comphments and extends previous studies of the effects of systemic and mtracerebroventncular VAS administration on pain responsivlty. Antmoclcept~on has been reported following VAS administration by e~ther of these routes 3-5,38-40, effects which are neither naloxone-reverslble3.4.3s.39, nor cross-tolerant with morphine s. Although no opiate-link has been identified m antinoclception induced by VAS, VAS has been lmphcated in the reinforcing propertles of opiates 67 and (controversially)49. 57 in the development of opiate tolerance and dependence 14,19,42.66 Whde the bases of these latter effects are unknown, the influence of VAS on learning and memory has often been cited to account for these observations (for review, see refs. 37, 65)

315 As in the case for VAS, a link between T R H and opiate systems has been controversially proposed28, 51. Although Holaday and co-workers 25 never observed T R H to antagonize opiate analgesia, the naloxone-like effects of T R H on other parameters of their elegant hemorrhagic and endotoxic shock models were so striking as to lead these investigators to term T R H a 'physiological antagonist'. This term was used since, to date, T R H has not been found to bind to opiate receptors attained from brain homogenates 26.46,62. However, since TRH-opiate binding interactions have not yet been examined in spinal cord preparations, the current findings that T R H can either antagonize or enhance IT morphme-induced antinociception, dependent upon dose, suggests that further biochemical investigations of spmal T R H opiate interactions may be warranted. In agreement with the conclusions of Holaday et al. 2s26, Osbahr et al. 54 reported that intracisternal T R H did not block fl-endorphin antinociceptlon in mice and Kasson and George 31 observed no antagonism of systemic morphine-induced antinociception by intracerebroventricular T R H . On the other hand, T R H has been observed to inhibit development of tolerance to the antinociceptive effect of opiates 7 and to suppress naloxone-induced hyperalgesia 55. Additionally, T R H has been reported to potentiate nonopiate footshock-induced analgesia 12, and to produce antinociception following microinjectlon into the periaqueductal gray20.72 or cerebral ventricleslO. 55. Whether 70 or noO0. 55 these latter effects are consistently modified by opiate antagonists is m dispute. The use of one or, at most, a very narrow range of T R H doses in all previous studies severely hinders the conclusions that currently can be drawn regarding T R H actions. In the current study, log doses of 0.25 ng to 2.5/tg T R H were tested IT and found to either enhance (2.5 ng), attenuate (25 ng, 250 ng) or

REFERENCES 1 AkIl, H and Mayer, D.J , Antagonism of stimulation produced analgesia by p-CPA, a serotonin synthesis inhibitor, Brain Research, 44 (1972) 692-697 2 Basbaum, A.I. and Fields, H L, Endogenous pain control systems' brainstem spinal pathways and endorphin circuitry, Ann. R ev. Neurosct , 7 (1984) 309-338. 3 Berkowitz, B.A and Sherman, S, Characterization of vasopressin analgesia, J Pharmacol. Exp. Ther., 220 (1982) 329-334

virtually abolish (0.25 ng, 2.5 pg) IT morphine analgesia. This procedure was based on recent investigations demonstrating that non-linear dose-response effects are produced by a variety of compounds including MIF-1 (ref. 6), substance p22, somatostatin 23,31, CCK 16, and the putative CCK antagonist, proglumide 6°.69. Such results emphasize the need to examine broad dose ranges before conclusions are drawn regarding neuropeptide effects. The demonstration that T R H can antagonize morphine analgesia gives initial evidence that T R H may be added to the growing list of neuropeptldes which can act like endogenous antagonists of opiate analgesia systems. To date, evidence has accumulated for opiate antagonistic actions by MIF-1 (ref. 6, 32), aMSH 18,73, A C T H 17, CCK 16,60,69, and now TRH. Further investigations will be needed to clarify: (a) the selectivity of these neuropeptides for opiate vs nonopiate analgesia systems; (b) the specificity of these effects as regards modulation of various types of somatic stimuli; (c) the neuroanatomical loci for neuropeptide modulation of antinociceptive systems; and (d) the underlying mechanisms by which these interactions occur. Such systematic investigations should yield important contributions to our concept of pain modulation and may provide insights into new methods of controlling pain.

ACKNOWLEDGEMENTS We would hke to gratefully acknowledge Dr. Richard Bodnar for his invaluable comments on the manuscript, and Paula Buchignani for preparation of the manuscript. Naloxone hydrochloride was a gift from Endo Labs. This work was supported m part by a gift from the A.H. Robins Company of Richmond, VA.

4 Berntson, G G and Berson, B S., Antinociceptive effects of Intraventncular or systemic administration of vasopressm in the rat, Life Sci., 26 (1980) 455-459. 5 Berson, B.S, Berntson, G.G., Zlpf, W , Toreelo, M W. and Kirk, W.T., Vasopressin-induced antlnoclceptlon' an investigation into its physiological and hormonal basis, Endocnnology, 113 (1983) 337-343. 6 Bhargava, H N, Effect of peptides on the development of tolerance to buprenophme, a mixed opiate agonist-antagonlst analgesic, Eur. J Pharmacol, 79 (1982) 117-123. 7 Bhargava, H N, Dissociation of tolerance to the analgesic

316 and hypothermlc effects of morphine by using thyrotropln releasing hormone, Life Scz, 29 (1981) 1015-1020 8 Bodnar, R J , Treusdell, C , Haldar, J , Kordower, J H and Ntlaver, G , Elimination of vasopressin analgesia following lesions placed In the rat hypothalamlc paraventncular nucleus, Soc Neurosct. Abstr., 10 (1984) 676 9 Bodnar, R . J , Zimmerman, E A , Ndaver, G , Mansour, A , Thomas, L W , Kelly, D D and Glusman, M , Dissociation of cold-water swim and morphine analgesm in Brattleboro rats with dmbetes mslpidus, Life Sct, 26 (1980) 1581-1590 10 Boschi, G , Deslles, M , Reny, V., Raps, R and Wrigglesworth, S , Antmooceptive properties of thyrotropin releasing hormone in mice: comparison with morphine, Br J Pharmacol, 79 (1983) 85-92 11 BUlJS, R M , Intra- and extrahypothalamlc vasopressin and oxytocin pathways in the rat Pathways to the limblc system, medulla oblongata and spinal cord, Cell Tissue Res, 192 (1978) 423-435 12 Butler, P D and Bodnar, R J., Potentiation of foot shock analgesm, by thyrotropm releasing hormone, Pepudes, 5 (1984) 13 D'Amour, F E and Smith, D L , A method for determining loss of pare sensaUon, J. Pharmacol Exp Ther, 72 (1941) 74-79. 14 De Wled, D and Glspen, W H , Impmred development of tolerance to morphine analgesia m rats with hereditary diabetes msipldus, Psychopharmacology, 46 (1976) 27-29 15 Doi, T. and Jurna, I., Intrathecal substance P depresses the tall flick response - - antagonism by naloxone, Arch Pharmacol, 317 (1981) 135-139 16 Fans, P.L., Komlsaruk, B , Watklns, L R and Mayer, D.J., Ewdence for the neuropepUde cholecystokmm as an antagomst of opiate analgesia, Scwnce, 219 (1983) 310-312 17 Fields, H L , Belcher, G. and Smock, T , ACTHI_z4 antagonizes opioid-mduced analgesia in the rat spinal cord, Soc Neurosct Abstr, 7 (1981) 604 18 Gispen, W H., Greidanus, T B V W , Waters-Ezrm, C , Ztmmerman, E., Knvoy, W.A and DeWied, D , Influence of pepUdes on reduced response of rats to electnc footshock after acute admimstratlon of morphine, Eur J Pharmacol, 33 (1975) 99 19 Greldanus, T B V W., Fat-Bronstein, H.T K and Van Ree, J M., Antisera to pltmtary hormones modulate development of tolerance to morphine. In J M Van Ree and L Terenms (Eds), Characteristics and Funcnon of Optotds, Elsevier, North-Holland, Amsterdam, 1978, pp 73-74 20 Gnffiths, E C , Slater, P and Webster, V.A D , Behav~oural interactions between neurotensm and thyrotrophtnreleasing hormone in rat central nervous system, J Phystol (London), 320 (1981) 90P 21 Grumbach, L., The prediction of analgesic actwlty in man by ammal testing. In R S Knighton and P.R Dunke (Eds), Pam, Little Brown, Boston, 1966, pp 163-182 22 Hall, M E and Stewart, J M , Substance P and anttnoclceptlon, Peptides, 4 (1983) 31-35 23 Havhcek, V., Rezek, M , Leybin, L and Fnesen, H., Analgesic effect of cerebroventrIcular admimstraUon of somatostatm (SRIF), Fed. Proc Fed A m Soc. Exp Btol, 36 (1977) 363 24 Hayes, A.G and Tyres, M B , Effects of lntrathecal and lntracerebroventncular mlect~ons of substance P on nooception in the rat and mouse, Br J. Pharmacol, 66 (1979) 488

25 Holaday, J W , D'Amato, R J , Ruvio, B A and Faen, A I , Action of naloxone and TRH on the autonomic regulation of circulation. In E Costa and M Trabucchi (Eds), Regulatory Pepttdes From Molecular Biology to Function, Raven Press, New York, 1982, pp 353-361 26 Holaday, J W , Tseng, G - F , Loh, H.H and LI, C H , Thyrotropln releasing hormone antagomzes fl-endorphm hypothermla and catalepsy, Life Sct, 22 (1978) 1537-1544 27 Itoh, 8 , Katsuura, G and Maeda, Y , Caerulein and cholecystoktnm suppress fl-endorphin-induced analgesia In the rat, Eur J Pharmacol, 80 (1982) 421-425 28 Jackson, I . M . D , Thyrotropln-releasmg hormone, N Engl. J Med , 306 (1982) 145-155 29 Johansson, O , Hokfelt, T , Pernow, B , Jeffcoate, S L., White, N., Stelnbusch, H W M , Verhofstad, A A J , Emson, P C and Spmdel, E , Immunohlstochemical support for three putative transmitters in one neuron, coexistence of 5-hydroxytryptamlne, substance P, and thyrotropln releasing hormone like lmmunoreactlvlty in medullary neurons projecting to the spinal cord, J Neurosct, 6 (1981) 1857-1881 30 Jurna, I and Zetler, G , Antlnoocept~ve effect of centrally administered caeruleln and cholecystokinln octapeptide (CCK-8), Eur J Pharmacol, 83 (1981) 323-331 31 Kasson, B G and George, R , Endocnne Influences on the actions of morphine III Responses to hypothalamlc hormones, Neuroendocrmology, 37 (t983) 416-420 32 Kastin, A J , Olson, R . D , Ehrensmg, R H , Berzas, M C , SchaUy, A V and Coy, D H , MIF-I's differential acUons as an opiate antagonist, Pharmacol Btochem Behay., 11 (1979) 721-723 33 Knepel, W., Nutto, D and Anhut, H , fl-endorphm controls vasopressm release during foot shock-Induced stress in the rat, Regulat Peptutes, 7 (1983) 9-19 34 Knepel, W , Nutto, D , Anhut, H. and Herttmg, G , Naloxone promotes stimulus-evoked vasopressm release in VlVO, Eur J Pharmacol, 65 (1980) 449-450 35 Knepel, W , Nutto, D., Anhut, H and Herttmg, G , Vasopressln and fl-endorphm release after osmotic and non-osmotic stimuli" effect of naloxone and dexamethasone, Eur J Pharmacol, 77 (1982) 299-306 36 Knepel, W , Nutto, D. and Hertting, G , Evidence for Inhibition by fl-endorphm of vasopressm release during footshock-reduced stress in the rat, Neuroendocrmology, 34 (1982) 353-356 37 Koob, G F and Bloom, F E., Behavioral effects of neuropeptldes' endorphins and vasopressin, Ann Rev. Physiol, 44 (1982) 571-582 38 Kordower, J H and Bodnar, R J , Vasopressln analgesia' specificity of action and non-opioid effects, Peptides, 5 (1984) 747-756. 39 Kordower, J H , Bodnar, R J , Manning, M M and Sawyer, W . H , Vasopressin analgesm: speofictty of action and uon-oplold effects, Soc Neurosct Abstr, 9 (1983) 202 40 Kordower, J H , SIkorszky, V and Bodnar, R J , Central antinoclceptive effects of lysme-vasopressIn and an analogue, Peptides, 3 (1982) 613-617 41 Krieger, D . T , Brain peptides what, where and why?, Science, 222 (1983) 975-985 42 Knvoy, W A., Zimmerman, E and Lande, S , Facilitation of development of resistance to morphine analgesm by desglycmamlde3-1ysme vasopresstn, Proc Natl Acad Sct U S.A , 71 (1974) 1852-1856 43 Llghton, C , Marsden, C A and Bennett, G W , The elF-

317 fects of 5,7-dlhydroxytryptamme and p-chlorophenylalanine on thyrotrophm-releasmg hormone m regions of the brain and spinal cord of the rat, Neuropharmacology, 23 (1984) 55-60 44 Maclewlcz, R., Phlpps, B S , Grenler, J. and Polettl, C E., Edlnger-Westphal nucleus, cholecystokmm lmmunocytochemistry and projections to spmal cord and trlgemmal nucleus in the cat, Bram Research, 299 (1984) 139-145 45 Mantyh, P W and Hunt, S P., Evidence for cholecystoktnm-hke immunoreactlve neurons m the rat medulla oblongata which project to the spinal cord, Brain Research, 291 (1984) 49-54 46 Martin, B R., Dewey, W L , Chau-Pham, T and Prange, A J , Interactions of thyrotropm releasing hormone and morphine sulfate in rodents, Life Sct, 10 (1977) 715-720 47 Melsenberg, G , Short-term behavioral effects of neurohypophyseal hormones' pharmacological characteristics, Neuropharmacology, 21 (1982) 309-316. 48 Melsenberg, G and Simmons, W . H , Behaworal effects of mtracerebroventncularly administered neurophyseal hormone analogs in mice, Pharmacol Btochem Behav, 16 (1982) 819-825 49 Mello, N.K and Mendelson, J H , Effects of the neuropeptide DG-AVP on morphine and food self-administration by dependent rhesus monkey, Pharmacol Btochem. Behav , 10 (1979) 415-419 50 Mlllan, M . J , Schmauss. C., Mdlan, M H and Herz, A , Vasopressm and oxytocln in the rat spinal cord' analysis of their role m control of noclcepUon, Bram Research, 309 (1984) 384-388 51 Morley, J E , Mmirevlew extrahypothalamlc thyrotropln releasing hormone (TRH) - - its dlstribuuon and its functions, Life Sct , 25 (1979) 1539-1550. 52 Nemeroff, C B , Kahvas, P W , Golden, R N and Prange, A J , Behavioral effects of hypothalamtc hypophyslotroplc hormones, neurotensm, substance P and other neuropeptides, Pharmacol. Ther, 24 (1984) 1-56. 53 Nllaver, G , Zlmmerman, E A., Wdklns, J., Mlchaels, J., Hoffman, D and Sdverman, A.-J., Magnocellular hypothalamlc proJections to the lower bramstem and spinal cord of the rat, Neuroendocrmology, 30 (1980) 150-158. 54 Osbahr, R . J , Nemeroff, C . B , Luttmger, D , Mason, G A and Prange, A J , Neurotensm-mduced antmoctceptlon in mice: antagonism by thyrotropin-releaslng hormone,J Pharmacol Exp. Ther, 217 (1982) 645-651 55 Rips, R , Reny, V and Deslles, M., Suppression of naloxone hyperalgesla by TRH, IRCS Med Scl, 11 (1983) 110-111 56 Sawchenko, P E and Swanson, L W , Immunohlstochemlcal identification of neurons in the paraventncular nucleus of the hypothalamus that project to the medulla or to the spinal cord in the rat, J Comp Neurol, 205 (1982) 260-272 57 Schmldt, W K , Holaday, J.W., Loh, H H and Way, E L., Failure of vasopressln and oxytocin to antagonize acute morphine antmoclceptlon or facilitate narcotic tolerance development, Life Sct, 23 (1978) 151-158. 58 Sklrboll, L , Hokfelt, T., Dockray, G , Rehfeld, J , Brownstem, M and CueUo, A.C , Evidence for permque-

ductal cholecystokmm-substance P neurons projecting to the spinal cord, J. Neurosci., 3 (1983) 1151-1157 59 Sofroniew, M.V and Weindl, A , Central nervous system distribution of vasopressin, oxytocm and neurophysm In J L Martmez, R A. Jensen, R.B Messing, H Rlgter and J.L McGough (Eds), Endogenous Pepttdes and Learnmg and Memory Processes, Academic Press, New York, 1981, pp 327-369. 60 Suberg, S.N., Culhane, E S., Carstens, E and Watkms, L.R., Behavioral and electrophysiological investigations of oplate/cholecystokinm interactions In H.L. Fields, R Dubner and F Cervero (Eds.), Advances m Pare Research and Therapy, Vol 9, Raven Press, New York, 1985, m press. 61 Swanson, L W , Sawchenko, P E , Wlegand, S J and Place, J L . , Separate neurons in the paraventncular nucleus project to the median eminence and to the medulla or spinal cord, Brain Research, 198 (1980) 190-195 62 Tache, Y , Lls, M and Collu, R., Effects of thyrotropm-releasing hormone on behavioral and hormonal changes induced by fl-endorphin, Life Sct, 21 (1977) 841-846. 63 Terenms, L , Somatostatm and ACTH are pepudes with partial antagomst-hke selectivity for opiate receptors, Eur. J Pharmacol, 38 (1976) 211-213. 64 Tung, A S and Yaksh, T L , In vwo evidence for multiple opiate receptors mediating analgesia in the rat spinal cord, Bram Research, 247 (1982) 75-83. 65 Van Ree, J.M., Bohus, B., Versteeg, D H G and DeWled, D., Neurohypophyseal principles and memory processes, Btochem. Pharmacol, 27 (1978) 1793-1800. 66 Van Ree, J M and DeWled, D., Effect of neurohypophyseal hormones on morphine dependence, Psychoneuroendocrmology, 2 (1977) 35-41. 67 Van Ree, J.M. and DeWted, D , Heroin self-admimstratlon is under control of vasopressm, Life Sci., 21 (1977) 315-320 68 Watkms, L . R , Frenk, H., Miller, J. and Mayer, D . J , Cataleptlc effects of opiates following mtrathecal admlmstratlon, Brain Research, 299 (1984) 43-49. 69 Watkins, L R., Kanscheck, I.B. and Mayer, D J , Potentlauon of opiate analgesia and apparent reversal of morphine tolerance by proglumide, Science, 224 (1984) 395-396. 70 Watklns, L.R. and Mayer, D J., Involvement of spinal oplold systems m footshock-induced analgesia' antagonism by naloxone is possible only before induction of analgesia, Brain Research, 242 (1982) 309-316 71 Watkins, L.R and Mayer, D.J., The neural organization of endogenous opiate and non-opiate pain control systems, Science, 216 (1982) 1185-1192. 72 Webster, V.A.D , Griffiths, E.C and Slater, P., Antmoclceptlve effects of thryotropln-reteasmg hormone and Its analogues in the rat penaqueductal gray region, Neurosct. Lett., 42 (1983) 67-70. 73 Zlmmerman, E and Krwoy, W.A., Antagonism between morphine and the polypeptldes ACTH, ACTH1_24, and aMSH in the nervous system In E Zlmmerman, W.H. Glspen, B H Marks and D DeWled (Eds.), Drug Effects on Neuroendocrme Regulatton, Progress tn Bram Research, Vol 39, North-Holland, Amsterdam, 1973, pp 383