- Email: [email protected]
Effects of midbrain, bulbar and combined morphine microinjections and systemic injections on orofacial nociception and rostral trigeminal stimulation: independent midbrain and bulbar opiate analgesia systems?
342
Brain Research, 215 (1981) 342-348 ~) Elsevier/North-Holland Biomedical Press
Effects of midbrain, bulbar and combined morphine microinjections and systemic injections on orofacial nociception and rostral trigeminal stimulation: independent midbrain and bulbar opiate analgesia systems?
J. PETER ROSENFELD and SUSAN STOCCO Cresap Neuroscience Laboratory, Department o] Psychology, Northwestern University, Evanston, IlL 60201 (U.S.A.)
(Accepted February 12th, 1981) Key words: opiate - - pain - - paragigantocellularis - - periaqueductal gray - - trigeminal nucleus
Microinjections of 0.35 pg and 0.7 pg morphine were made into rat nucleus reticularis paragigantocellularis and ventral central gray respectively.When injected simultaneously, the analgesic effect on orofacial thermal pain was significantly greater than the effect of either injection alone, suggesting a summation mechanism tor the two sites. No microinjection affected the threshold for aversive response to main sensory trigeminal nuclear stimulation. This threshold was, however, profoundly elevated by 10 mg/kg systemic injection.
Midbrain ventral central gray (VCG) and bulbar nucleus reticularis paragigantocellularis (PGC) can be electrically activated or opiate microinjected, either procedure resulting in antinociceptive effects 1,8,1°A1. Both structures connectS, 9 to the nucleus raphe magnus (NRM), another opiate substrate with established descending connections to dorsal horn which have been suggested to relay descending, opiate-induced inhibition of pain inputs 4. These data raise the question of whether V C G and P G C converge on N R M cells whose efferents would then act as a final common path for opiate analgesia, or if V C G and P G C have separate, independent descending inhibitory routes 17, since P G C and V C G may have independent connections to dorsal horn that do not go through N R M 2. Systemic morphine administration (10 mg/kg) produces profound analgesia for aversive trigeminal brain stimulation and for tail flick. However, when locally microinjected into either V C G or P G C in doses producing profound analgesic effects for facial pain (equivalent to those resulting from systemic injection), morphine has no effect on aversive caudal trigeminal nuclear stimulation, and an effect on tail flick significantly smaller than that of systemic injection10,11. These data suggest that systemic injections access multiple independent opiate analgesia substrates, only a fraction of which can be reached by local microinjections. While Rudy and Yeung 15 have shown summation effects of independent activation of opiate substrates in spinal cord and VCG, the present report supports an intracranial summation hypothesis by demonstrating that combined microinjections of VCG and P G C have reliably larger orofacial analgesic effects than do single injections of either structure.
343 This report also examines for the first time effects of systemic and microinjected opiates on aversive stimulation in the main trigeminal sensory nucleus, a structure which has been recently reconsidered as a possibly crucial integration area for central relay of orofacial pain 3,6,12. Fifteen male albino rats (500-600 g) were implanted with bipolar pairs of twisted, 78 # m nichrome wires for electrical stimulation as detailed elsewhere10,11,i3, 14. The target for the electrodes, (one pair per rat) was the center of the main sensory trigeminal nucleus (TNM). Each rat also had implanted two stainless steel 25 gauge outer guide cannulae with 31 gauge obturators. One of these cannulae was targeted toward the N P G region of medulla (as in ref. 11) and the other toward the ventrolateral quadrant of ventral central gray in caudal midbrain (as in ref. 10). The guide cannulae terminated 2 mm above the targets just named. During microinjections, the obturator was removed and replaced with 31 gauge patent inner cannulae which would protrude 2 mm below the end of the guides to reach the respective targets. The two cannulae and stimulation electrodes were all mutually ipsilateral. Installed on the rats' faces between eye and upper lip, bilaterally, were 10 f~, 4 mm × 1 mm diameter, 1/8 W resistors. These devices were in contact with the skin through facial fur. When a 150-250 mA, DC current is applied to such a resistor, it becomes aversively hot, producing a non-stereotyped, face-rubbing escape reaction within 10 sec. Previous work has established the reliability and validity of the face-rub response latency as a non-reflexive orofacial nociception index 12. These face rub latencies, in addition to paw lick latencies (on a 54 °C hot plate), comprised our peripheral analgesia test battery. There were two tests for aversiveness of stimulation in TNM. (1) We would give 200 Hz gradually rising, sinusoidal bursts of about 0.5 sec duration every 5 sec, increasing the peak-to-peak current of each successive burst by about 5 #A, until an aversive reaction was seen. This reaction was defined to involve a non-stereotyped movement of all 4 limbs in an apparent escape attempt as agreed by two independent observers 95 ~ of the time (as in refs. 10, 11, 13, 14). Defecation or urination, squealing, face rubbing, and teeth gnashing accompanied this reaction 75 ~o of the time. The current level just producing this reaction was recorded as the aversive reaction threshold. We have previously reported that this judged threshold will sustain goal-directed, passive instrumental avoidance13; this validates the judged aversive property of the threshold current. (2) The preceding procedure was repeated with 20 Hz stimulation. (These stimulation parameters and procedures are explained in ref. 14.) Two weeks after recovery from surgery, rats were subjected to the following 8 tests (which were run every other day): (1)first baseline, peripheral and central aversive reactions were obtained with no drug given; (2) I/'CG microinjection, 0.7 #g morphine sulphate in 0.5 #1 solution was microinjected in 2 min and behavioral tests were given 10 min later; (3) second baseline, repeat first baseline; (4) PGC microinjection, 0.35/~g morphine sulphate in 0.5 #1 was microinjected in 2 min and behavioral tests were given 5 min later; (5) third baseline; (6) conjoint microinjections, VCG was injected as above and 5 min later PGC was injected as above. Five more minutes later behavioral tests were given; (7) fourth baseline; (8) systemic morphine, 10 mg/kg, was
344 injected (s.c.) and behaviors were tested 45-75 min later. In our previous report, the controls run were saline microinjections which also controlled for effects of injection procedure. In the present study since the major comparisons are between different injections and since the repeated injections necessary here plus saline injections could produce lesions, saline controls are not used. Moreover, in the earlier work, there were no significant differences between baseline (as above) and saline data reported for face rubs and trigeminal stimulation10,11, using the same VCG and P G C targets. Behavioral data were either latencies in seconds or aversive stimulation threshold currents in #A. For all rats the average of the 4 baseline control days was taken as the denominator of a fraction whose numerator would be a value obtained on an experimental drug day. This fraction × 100 gives the drug effect expressed as per cent of control; 100 ~ ~-- no change, ~ 100 ~ = analgesia, < 100 ~ -- hyperalgesia. If under any drug, no reaction to a heat stimulus occurred after a 3 0 0 ~ of control latency expired or if 25 sec lapsed, whichever came first, the heat was stopped and the value 300 ~ or the ~ change at 25 sec was recorded. This procedure was used to avoid injurious heat levels and yields a conservative bias in calculated means. One/~g morphine microinjected into P G C or VCG blocks the face-rub reactions11; i.e. elevates the latency over 3 times the baseline (our cut-off criterion). We needed to find lower doses which would produce analgesia but not total block, otherwise it would be impossible to interpret a total block produced by conjoint microinjection. The differential doses and delays noted above satisfied these requirements for the face-rub reaction, which is our lowest variance nociception index. That P G C is more sensitive than V C G is consistent with other reports 16. The doses used here for face rub have no effect on tail flick (which wasn't used here). Histological verification of all placements concluded the studies. Immediately after rats lost consciousness due to the sacrificing injection, 0.5 #1 of saturated Fast Green dye was injected for 2 min through both cannulae. Brains were removed within 10 rain. Histological study of the brains of rats yielding usable data showed 14 P G C cannula tip placements distributed in (n~--4) and above (n--10) the rostral border area of the inferior olivary nucleus with dense dye stain seen in PGC. In 4 of these, dye was observed to reach nucleus raphe magnus. In two of these and in one other case the dye crossed midline by about 0.5 mm. The VCG tip loci were distributed in and just below the ventrolateral quadrant of VCG. All but two of their tracks were 0.5 mm or more lateral to the aqueduct. None of these off-target placements (i.e. through aqueduct for VCG or where dye was seen in nucleus raphe magnus for PGC) was associated with any special single microinjection effects. For example, only 1 of the 4 PGC placements involving raphe produced total blocks of the face-rub. However, the 3 other blocks were produced by laterally situated cannulae well away from raphe. Similarly, of the 3 VCG placements yielding total block, only 1 was associated with the 2 cases of aqueduct damage. The dye in all cases occupied a spherical volume of diameter 1.7 mm. Systemic opiate effects on T N M stimulation. Only 10 rats had effectively aversive T N M stimulating electrodes, and histological follow-up showed that only 7 of these were well localized in the T N M . The 3 misses were medial, in the reticular formation.
345 TABLE !
Baseline avers±re reaction threshold (ART) currents in ttA and % o f control ARTs due to systemic (subcutaneous) morphine injection for avers±re brain stimulation at accurate and medially off-target placements o f stimulating electrodes in the main sensory trigeminal nucleus Stimulation Jrequency
Accurate placements (n -- 7)
Medial placements (n = 3)
200 Hz
20 Hz
200 H z
20 H z
26.3 ± 14.6
25.3 ± 4.6
45.3 ± 16.6
Baseline ART current,
pA ± 1 S.D. 13.7 i 8.5 Subcutaneous morphine effect, ~ ± 1 S.D. 196.7± 65
272.9 ± 99
126.3 ± 42
131.7 -+- 60
Table I shows that subcutaneous morphine, 10 mg/kg, elevated aversive reaction thresholds for stimulation of T N M to about double (200 Hz stimulation) or more (20 Hz) the baseline values. When the electrodes, though aversive, were just medial to the T N M , effects (at about 130 ~ ) were not significantly different than baseline. Withinsubject, two-tailed t-tests confirmed ( a t p < 0 . 0 5 ) the 200 Hz vs 20 Hz differences seen in Table I for both baseline and opiate-treated aversive reaction thresholds, as well as the baseline versus systemic opiate-treatment difference. The low baseline threshold value seen in Table I for 200 Hz stimulation of T N M , 13.7 #A, is notable. In a recent study in the same lab. 11 using Trigeminal Nucleus Caudalis stimulation, the comparable mean was 47.8 4- 16.9 # A for 6 rats. An independent-groups, two-tailed, t-test comparing this value with the T N M mean was significant (P<0.01). Microinjection effects. On T N M aversive reaction thresholds, there were no effects of either single or conjoint opiate microinjections; ~ of control means were > 9 0 ~ or ( l l 0 ~ . As noted above, in 4 cases, injection into the P G C alone produced maximum analgesia as measured by the face rub test. The blockade was bilateral in 3 of these animals and unilateral in the other. There were also 3 instances of complete blockade of the face rub response after injections into the V C G alone. The blockade was bilateral in one rat and unilateral in the other two. Averaged ~ change scores with and without data from rats in which injections into the P G C or the V C G alone produced complete block of the face rub response are given in Table II. When all cases are considered, the mean effect of conjoint injections (205.6 ~ ) was significantly greater than the effect produced by injections into either site alone (two-tailed, correlated ttest; P < 0.02). These differences were still significant after removal of the data from rats in which an injection into P G C or V C G alone produced complete block of the face rub response (P < 0.01). It should be noted that with the data from these animals removed, there were 19 opportunities for observation of summation effects within individual rats, summation being defined as a conjoint effect at least (1) 40 ~o larger than the average of the effects produced by injections into each site individually, (2) 25 larger than the larger of the associated pair of single effects. Summation was observed in 15 of the 19 cases (79 ~). In 6 of the 15 instances of summation, no analgesia ( < 120 ~ ) was produced by injection into each locus separately, whereas an analgesia score of greater than 150 ~ was produced by conjoint injection.
346 TABLE II Effects o f the two single and one conjoint morphine microinjections on per cent-of-control face rub ( FR) lateneies Jbr all available data (mMdle column), and for those 19 cases in which rats showing total bloek after an injection into the PGC or the VCG alone were eliminated (rightmost colunln). Behavioral measures and
% o f F R controllatencies ± S.D.
microinjection
All cases (n
PGC (single) VCG (single) PGC ? VCG (conjoint)
152.1 ::~ 84.7 138.7 ±: 77 205.6 ± 93.8
28)
Cases with no single block effects (n
19)
114.6 ~: 44.5 108.5 ~: 10.4 198.8 ~ 81.2
H o t p l a t e data. The mean effects of PGC, VCG, conjoint and subcutaneous injections of morphine on 12 hot-plate paw-lick (% of control) latencies were 205.5 99 %, 181.9 ~ 108 %, 149.2 ~ 87.3 %, and 172.7 ± 73 % respectively. High variance obviated significant effects in comparison of these means. D i s c u s s i o n . The most significant finding here is that for facial pain, conjoint morphine microinjections of sites separated by at least 6 mm produced significantly larger effects than did injection of either site alone. Given previous evidence 7 and our own dye dispersion data that microinjected materials diffuse neither from PGC to VCG nor from VCG to PGC, the present results suggest that these two sites have independent influences on pain inputs. There are the following explanations for such summation, all based on documented pathways cited above 2,5,9 : (a) spatial inhibitory summation of PGC cell terminals with terminals from raphe magnus cells at the primary afferent synapse, with the raphe cells simply relaying descending effects from VCGg; (b) summation of PGC cell terminals with VCG terminals descending directly to primary afferent synapsesZ; (c) spatial summation of two pathways from raphe to the primary afferent synapse; one pathway would involve VCG-to-raphe cell connections 9independent of and distinct from the other path, a PGC-to-raphe cell connection .~; (d) spatial summation on common raphe cells of PGC-to-raphe and VCG-to-raphe terminals. These non-mutually exclusive hypotheses cannot be distinguished by the present data, nor is the list suggested to be exhaustive. There remains the possibility of complex interactions involving unknown sites and presynaptic effects not considered here. There are, however, two hypotheses on which the present data bear which could also explain the present results: (e) PGC-injected opiates diffused to raphe, driving opiate-sensitive cells there 4 which also received descending drive from VCG-to-raphe connections; (f) PGC-injected opiates diffused to and activated raphe cells distinct from another population of raphe cells receiving the descending terminals from VCG. However these two distinct raphe cell populations could give rise to raphe-trigeminal pathways which spatially summate at the primary afferent synapse (as in c) above). These possibilities are not likely because our dye studies showed only 4 of 14 cases of dye diffusion to raphe magnus. Since the dye does not bind to opiate receptors, it probably diffused further than the opiates did. Moreover, two of the largest single PGC injection effects were in rats where the PGC placement was most dorsolateral, i.e. well away from raphe.
347 The failure of the hot plate test to show summation is questionable for two reasons: (1) there were several occasions where apparently aversive behavior was noted but which did not include paw licks; (2) microinjection produced an extremely variable, apparently dichotomous distribution of hot plate ~o change scores. For single V C G injections, 6/12 (of 12) scores were < 130 ~ , 5/12 were > 280 ~ , and 1 was middling at 137 ~ . For PGC, 4/12 were < 106 ~ , 6/12 were > 270 ~o, 1 was 217 ~ , and 1 was 138 ~o. It is also possible that hot plate tests involving front and hind paws (i.e. upper and lower body) are affected differently by particular microinjections than are face-rub tests. Evidence of such possible somatotypy has been presented elsewherO0,11. We have emphasized the high reliability (i.e. low variance) of the face-rub test 12. In this study, our parameters were tailored to it and the summations seen for facial analgesia are probably more reliable than the lack of summation for hot-plate tests. (Tail flicks could not be used as noted above.) There may be either a denser concentration of nociceptive fibers in T N M than in subnucleus caudalis or the firing thresholds of fibers in the latter may be significantly higher than those of the more rostral locus. This notion is based on the profoundly different thresholds for aversive reaction characterizing the two loci, and complements our related findings of larger effects of rostral vs caudal lesions on orofacial nociception3,12. Nevertheless, T N M resembles caudalis in that localized P G C microinjections in doses producing some analgesia for orofacial nociception was without effect on aversive T N M stimulation thresholds: for the face-rub latency, 0.35 #g morphine produced a 152.1% of control score here; 1.0/tg produced > 270 ~ of control previously 11. Moreover 10 mg/kg subcutaneous injection at least doubles the aversive threshold for stimulation of both T N M and caudalisla,Xa, 14. Two related hypotheses consistent with these effects are: (1) our P G C locus inhibits a T N M (or caudalis, ref. 11) locus different than where we were stimulating; (2) systematic injection accesses multiple analgesia substrates of which only a fraction are accessed by local microinjection. These notions are consistent with the summation effects also reported here. The lack of effect of systemic morphine on aversive thresholds to brain stimulation central (medial) to the main sensory nucleus is entirely consistent with and is dis cussed in our previous reports13,14. Supported by N I H Grants GM23696 and DE05204.
1 Akaike, A., Shibata, T., Satoh, M. and Takagi, H., Analgesia induced by microinjection of morphine into, and electrical stimulation of, the nucleus reticularis paragigantocellularis of rat medulla oblongata, Neuropharmacology, 17 (1978) 775-778. 2 Basbaum, A. I. and Fields, H. L., The origin of descending pathways in the dorsolateral funiculus of the spinal cord of the cat and rat: further studies on the anatomy of pain modulation, J. Comp. Neurol., 187 (1979) 513-532. 3 Broton, J. and Rosenfeld, J. P., The necessity of rostral (vs caudal) trigeminal thalamic projections for orofacial nociception, .4mer. Pain Soc., Abstr. (1980) 17. 4 Dickenson, A. D., Oliveras, J.-L. and Besson, J.-M., Role of the nucleus raphe magnus in opiate analgesia as studied by the microinjection technique in the rat, Brain Research, 170 (1979) 95-111. 5 Gallager, D. W. and Pert, A., Afferents to brain stem nuclei (brain stem raphe, n. reticularis
348
6 7 8 9 10
I1
12 13 14
15 16
17
pontis caudalis, and n. gigantocellularis) in the rat as demonstrated by microiontophoretically applied horseradish peroxidase, Brain Research, 144 (1978) 257-275. Khayyat, G., Yu, Y. and King, R., Response pattern to noxious and non-noxious stimuli in rostral trigeminal relay nuclei, Brain Research, 97 (1975) 47-60. Lomax, P., The distribution of morphine following intracerebral microinjection, Experientia (Basel), 22 (1966) 249-250. Mayer, D. J. and Price, D. D., Central nervous system mechanisms of analgesia, Pain, 2 (1976) 379~,04. Pomeroy, S. L. and Behbehani, M. M., Physiologic evidence for a projection from periaqueductal gray to nucleus raphe magnus in the rat, Brain Research, 176 (1979) 143-147. Rosenfeld, J. P. and Keresztes-Nagy, P., Differential effects of intracerebrally microinjected enkephalin analogs on centrally versus peripherally induced pain, and evidence for a facial versus lower body analgesic effect, Pain, 9 (1980) 171-181. Rosenfeld, J. P. and Stocco, S., Differential effects of systemic versus intracranial injection of opiates on central, orofacial and lower body nociception: somatotypy in bulbar analgesia systems, Pain, 9 (1980) 307-318. Rosenfeld, J. P., Clavier, R. M. and Broton, J. G., Bilateral and unilateral antinociceptive effects of rostral trigeminal nuclear complex lesions in rats, Brain Research, 157 (1978) 147-152. Rosenfeld, J. P. and Holzman, B. S., Differential effect of morphine on stimulation of primary versus higher order trigeminal terminals, Brain Research, 124 (1977) 367-372. Rosenfeld, J. P. and Vickery, J. L., Differential effect of morphine on trigeminal nucleus versus reticular aversive stimulation: independence of negative effects from stimulation parameters, Pain, 2 (1976) 405-416. Rudy, T. A. and Yeung, J. C., Profound potentiation of morphine's analgetic potency by concurrent intrathecal and intraventricular administration, Neurosci. Abstr., 5 (1979) 615. Satoh, M., Akaike, A. and Takagi, H., Excitation by morphine and enkephalin of single neurons of nucleus reticularis paragigantocellularis in the rat : a probable mechanism of analgesic action of opioids, Brain Research, 169 (1979) 406-410. Satoh, M., Akaike, A., Nakazawa, T. and Takagi, H., Evidence for involvement of separate mechanisms in the production of analgesia by electrical stimulation of the nucleus reticularis paragigantocellularis and nucleus raphe magnus in the rat, Brain Research, 194 (1980) 525-529.
Brain Research, 215 (1981) 342-348 ~) Elsevier/North-Holland Biomedical Press
Effects of midbrain, bulbar and combined morphine microinjections and systemic injections on orofacial nociception and rostral trigeminal stimulation: independent midbrain and bulbar opiate analgesia systems?
J. PETER ROSENFELD and SUSAN STOCCO Cresap Neuroscience Laboratory, Department o] Psychology, Northwestern University, Evanston, IlL 60201 (U.S.A.)
(Accepted February 12th, 1981) Key words: opiate - - pain - - paragigantocellularis - - periaqueductal gray - - trigeminal nucleus
Microinjections of 0.35 pg and 0.7 pg morphine were made into rat nucleus reticularis paragigantocellularis and ventral central gray respectively.When injected simultaneously, the analgesic effect on orofacial thermal pain was significantly greater than the effect of either injection alone, suggesting a summation mechanism tor the two sites. No microinjection affected the threshold for aversive response to main sensory trigeminal nuclear stimulation. This threshold was, however, profoundly elevated by 10 mg/kg systemic injection.
Midbrain ventral central gray (VCG) and bulbar nucleus reticularis paragigantocellularis (PGC) can be electrically activated or opiate microinjected, either procedure resulting in antinociceptive effects 1,8,1°A1. Both structures connectS, 9 to the nucleus raphe magnus (NRM), another opiate substrate with established descending connections to dorsal horn which have been suggested to relay descending, opiate-induced inhibition of pain inputs 4. These data raise the question of whether V C G and P G C converge on N R M cells whose efferents would then act as a final common path for opiate analgesia, or if V C G and P G C have separate, independent descending inhibitory routes 17, since P G C and V C G may have independent connections to dorsal horn that do not go through N R M 2. Systemic morphine administration (10 mg/kg) produces profound analgesia for aversive trigeminal brain stimulation and for tail flick. However, when locally microinjected into either V C G or P G C in doses producing profound analgesic effects for facial pain (equivalent to those resulting from systemic injection), morphine has no effect on aversive caudal trigeminal nuclear stimulation, and an effect on tail flick significantly smaller than that of systemic injection10,11. These data suggest that systemic injections access multiple independent opiate analgesia substrates, only a fraction of which can be reached by local microinjections. While Rudy and Yeung 15 have shown summation effects of independent activation of opiate substrates in spinal cord and VCG, the present report supports an intracranial summation hypothesis by demonstrating that combined microinjections of VCG and P G C have reliably larger orofacial analgesic effects than do single injections of either structure.
343 This report also examines for the first time effects of systemic and microinjected opiates on aversive stimulation in the main trigeminal sensory nucleus, a structure which has been recently reconsidered as a possibly crucial integration area for central relay of orofacial pain 3,6,12. Fifteen male albino rats (500-600 g) were implanted with bipolar pairs of twisted, 78 # m nichrome wires for electrical stimulation as detailed elsewhere10,11,i3, 14. The target for the electrodes, (one pair per rat) was the center of the main sensory trigeminal nucleus (TNM). Each rat also had implanted two stainless steel 25 gauge outer guide cannulae with 31 gauge obturators. One of these cannulae was targeted toward the N P G region of medulla (as in ref. 11) and the other toward the ventrolateral quadrant of ventral central gray in caudal midbrain (as in ref. 10). The guide cannulae terminated 2 mm above the targets just named. During microinjections, the obturator was removed and replaced with 31 gauge patent inner cannulae which would protrude 2 mm below the end of the guides to reach the respective targets. The two cannulae and stimulation electrodes were all mutually ipsilateral. Installed on the rats' faces between eye and upper lip, bilaterally, were 10 f~, 4 mm × 1 mm diameter, 1/8 W resistors. These devices were in contact with the skin through facial fur. When a 150-250 mA, DC current is applied to such a resistor, it becomes aversively hot, producing a non-stereotyped, face-rubbing escape reaction within 10 sec. Previous work has established the reliability and validity of the face-rub response latency as a non-reflexive orofacial nociception index 12. These face rub latencies, in addition to paw lick latencies (on a 54 °C hot plate), comprised our peripheral analgesia test battery. There were two tests for aversiveness of stimulation in TNM. (1) We would give 200 Hz gradually rising, sinusoidal bursts of about 0.5 sec duration every 5 sec, increasing the peak-to-peak current of each successive burst by about 5 #A, until an aversive reaction was seen. This reaction was defined to involve a non-stereotyped movement of all 4 limbs in an apparent escape attempt as agreed by two independent observers 95 ~ of the time (as in refs. 10, 11, 13, 14). Defecation or urination, squealing, face rubbing, and teeth gnashing accompanied this reaction 75 ~o of the time. The current level just producing this reaction was recorded as the aversive reaction threshold. We have previously reported that this judged threshold will sustain goal-directed, passive instrumental avoidance13; this validates the judged aversive property of the threshold current. (2) The preceding procedure was repeated with 20 Hz stimulation. (These stimulation parameters and procedures are explained in ref. 14.) Two weeks after recovery from surgery, rats were subjected to the following 8 tests (which were run every other day): (1)first baseline, peripheral and central aversive reactions were obtained with no drug given; (2) I/'CG microinjection, 0.7 #g morphine sulphate in 0.5 #1 solution was microinjected in 2 min and behavioral tests were given 10 min later; (3) second baseline, repeat first baseline; (4) PGC microinjection, 0.35/~g morphine sulphate in 0.5 #1 was microinjected in 2 min and behavioral tests were given 5 min later; (5) third baseline; (6) conjoint microinjections, VCG was injected as above and 5 min later PGC was injected as above. Five more minutes later behavioral tests were given; (7) fourth baseline; (8) systemic morphine, 10 mg/kg, was
344 injected (s.c.) and behaviors were tested 45-75 min later. In our previous report, the controls run were saline microinjections which also controlled for effects of injection procedure. In the present study since the major comparisons are between different injections and since the repeated injections necessary here plus saline injections could produce lesions, saline controls are not used. Moreover, in the earlier work, there were no significant differences between baseline (as above) and saline data reported for face rubs and trigeminal stimulation10,11, using the same VCG and P G C targets. Behavioral data were either latencies in seconds or aversive stimulation threshold currents in #A. For all rats the average of the 4 baseline control days was taken as the denominator of a fraction whose numerator would be a value obtained on an experimental drug day. This fraction × 100 gives the drug effect expressed as per cent of control; 100 ~ ~-- no change, ~ 100 ~ = analgesia, < 100 ~ -- hyperalgesia. If under any drug, no reaction to a heat stimulus occurred after a 3 0 0 ~ of control latency expired or if 25 sec lapsed, whichever came first, the heat was stopped and the value 300 ~ or the ~ change at 25 sec was recorded. This procedure was used to avoid injurious heat levels and yields a conservative bias in calculated means. One/~g morphine microinjected into P G C or VCG blocks the face-rub reactions11; i.e. elevates the latency over 3 times the baseline (our cut-off criterion). We needed to find lower doses which would produce analgesia but not total block, otherwise it would be impossible to interpret a total block produced by conjoint microinjection. The differential doses and delays noted above satisfied these requirements for the face-rub reaction, which is our lowest variance nociception index. That P G C is more sensitive than V C G is consistent with other reports 16. The doses used here for face rub have no effect on tail flick (which wasn't used here). Histological verification of all placements concluded the studies. Immediately after rats lost consciousness due to the sacrificing injection, 0.5 #1 of saturated Fast Green dye was injected for 2 min through both cannulae. Brains were removed within 10 rain. Histological study of the brains of rats yielding usable data showed 14 P G C cannula tip placements distributed in (n~--4) and above (n--10) the rostral border area of the inferior olivary nucleus with dense dye stain seen in PGC. In 4 of these, dye was observed to reach nucleus raphe magnus. In two of these and in one other case the dye crossed midline by about 0.5 mm. The VCG tip loci were distributed in and just below the ventrolateral quadrant of VCG. All but two of their tracks were 0.5 mm or more lateral to the aqueduct. None of these off-target placements (i.e. through aqueduct for VCG or where dye was seen in nucleus raphe magnus for PGC) was associated with any special single microinjection effects. For example, only 1 of the 4 PGC placements involving raphe produced total blocks of the face-rub. However, the 3 other blocks were produced by laterally situated cannulae well away from raphe. Similarly, of the 3 VCG placements yielding total block, only 1 was associated with the 2 cases of aqueduct damage. The dye in all cases occupied a spherical volume of diameter 1.7 mm. Systemic opiate effects on T N M stimulation. Only 10 rats had effectively aversive T N M stimulating electrodes, and histological follow-up showed that only 7 of these were well localized in the T N M . The 3 misses were medial, in the reticular formation.
345 TABLE !
Baseline avers±re reaction threshold (ART) currents in ttA and % o f control ARTs due to systemic (subcutaneous) morphine injection for avers±re brain stimulation at accurate and medially off-target placements o f stimulating electrodes in the main sensory trigeminal nucleus Stimulation Jrequency
Accurate placements (n -- 7)
Medial placements (n = 3)
200 Hz
20 Hz
200 H z
20 H z
26.3 ± 14.6
25.3 ± 4.6
45.3 ± 16.6
Baseline ART current,
pA ± 1 S.D. 13.7 i 8.5 Subcutaneous morphine effect, ~ ± 1 S.D. 196.7± 65
272.9 ± 99
126.3 ± 42
131.7 -+- 60
Table I shows that subcutaneous morphine, 10 mg/kg, elevated aversive reaction thresholds for stimulation of T N M to about double (200 Hz stimulation) or more (20 Hz) the baseline values. When the electrodes, though aversive, were just medial to the T N M , effects (at about 130 ~ ) were not significantly different than baseline. Withinsubject, two-tailed t-tests confirmed ( a t p < 0 . 0 5 ) the 200 Hz vs 20 Hz differences seen in Table I for both baseline and opiate-treated aversive reaction thresholds, as well as the baseline versus systemic opiate-treatment difference. The low baseline threshold value seen in Table I for 200 Hz stimulation of T N M , 13.7 #A, is notable. In a recent study in the same lab. 11 using Trigeminal Nucleus Caudalis stimulation, the comparable mean was 47.8 4- 16.9 # A for 6 rats. An independent-groups, two-tailed, t-test comparing this value with the T N M mean was significant (P<0.01). Microinjection effects. On T N M aversive reaction thresholds, there were no effects of either single or conjoint opiate microinjections; ~ of control means were > 9 0 ~ or ( l l 0 ~ . As noted above, in 4 cases, injection into the P G C alone produced maximum analgesia as measured by the face rub test. The blockade was bilateral in 3 of these animals and unilateral in the other. There were also 3 instances of complete blockade of the face rub response after injections into the V C G alone. The blockade was bilateral in one rat and unilateral in the other two. Averaged ~ change scores with and without data from rats in which injections into the P G C or the V C G alone produced complete block of the face rub response are given in Table II. When all cases are considered, the mean effect of conjoint injections (205.6 ~ ) was significantly greater than the effect produced by injections into either site alone (two-tailed, correlated ttest; P < 0.02). These differences were still significant after removal of the data from rats in which an injection into P G C or V C G alone produced complete block of the face rub response (P < 0.01). It should be noted that with the data from these animals removed, there were 19 opportunities for observation of summation effects within individual rats, summation being defined as a conjoint effect at least (1) 40 ~o larger than the average of the effects produced by injections into each site individually, (2) 25 larger than the larger of the associated pair of single effects. Summation was observed in 15 of the 19 cases (79 ~). In 6 of the 15 instances of summation, no analgesia ( < 120 ~ ) was produced by injection into each locus separately, whereas an analgesia score of greater than 150 ~ was produced by conjoint injection.
346 TABLE II Effects o f the two single and one conjoint morphine microinjections on per cent-of-control face rub ( FR) lateneies Jbr all available data (mMdle column), and for those 19 cases in which rats showing total bloek after an injection into the PGC or the VCG alone were eliminated (rightmost colunln). Behavioral measures and
% o f F R controllatencies ± S.D.
microinjection
All cases (n
PGC (single) VCG (single) PGC ? VCG (conjoint)
152.1 ::~ 84.7 138.7 ±: 77 205.6 ± 93.8
28)
Cases with no single block effects (n
19)
114.6 ~: 44.5 108.5 ~: 10.4 198.8 ~ 81.2
H o t p l a t e data. The mean effects of PGC, VCG, conjoint and subcutaneous injections of morphine on 12 hot-plate paw-lick (% of control) latencies were 205.5 99 %, 181.9 ~ 108 %, 149.2 ~ 87.3 %, and 172.7 ± 73 % respectively. High variance obviated significant effects in comparison of these means. D i s c u s s i o n . The most significant finding here is that for facial pain, conjoint morphine microinjections of sites separated by at least 6 mm produced significantly larger effects than did injection of either site alone. Given previous evidence 7 and our own dye dispersion data that microinjected materials diffuse neither from PGC to VCG nor from VCG to PGC, the present results suggest that these two sites have independent influences on pain inputs. There are the following explanations for such summation, all based on documented pathways cited above 2,5,9 : (a) spatial inhibitory summation of PGC cell terminals with terminals from raphe magnus cells at the primary afferent synapse, with the raphe cells simply relaying descending effects from VCGg; (b) summation of PGC cell terminals with VCG terminals descending directly to primary afferent synapsesZ; (c) spatial summation of two pathways from raphe to the primary afferent synapse; one pathway would involve VCG-to-raphe cell connections 9independent of and distinct from the other path, a PGC-to-raphe cell connection .~; (d) spatial summation on common raphe cells of PGC-to-raphe and VCG-to-raphe terminals. These non-mutually exclusive hypotheses cannot be distinguished by the present data, nor is the list suggested to be exhaustive. There remains the possibility of complex interactions involving unknown sites and presynaptic effects not considered here. There are, however, two hypotheses on which the present data bear which could also explain the present results: (e) PGC-injected opiates diffused to raphe, driving opiate-sensitive cells there 4 which also received descending drive from VCG-to-raphe connections; (f) PGC-injected opiates diffused to and activated raphe cells distinct from another population of raphe cells receiving the descending terminals from VCG. However these two distinct raphe cell populations could give rise to raphe-trigeminal pathways which spatially summate at the primary afferent synapse (as in c) above). These possibilities are not likely because our dye studies showed only 4 of 14 cases of dye diffusion to raphe magnus. Since the dye does not bind to opiate receptors, it probably diffused further than the opiates did. Moreover, two of the largest single PGC injection effects were in rats where the PGC placement was most dorsolateral, i.e. well away from raphe.
347 The failure of the hot plate test to show summation is questionable for two reasons: (1) there were several occasions where apparently aversive behavior was noted but which did not include paw licks; (2) microinjection produced an extremely variable, apparently dichotomous distribution of hot plate ~o change scores. For single V C G injections, 6/12 (of 12) scores were < 130 ~ , 5/12 were > 280 ~ , and 1 was middling at 137 ~ . For PGC, 4/12 were < 106 ~ , 6/12 were > 270 ~o, 1 was 217 ~ , and 1 was 138 ~o. It is also possible that hot plate tests involving front and hind paws (i.e. upper and lower body) are affected differently by particular microinjections than are face-rub tests. Evidence of such possible somatotypy has been presented elsewherO0,11. We have emphasized the high reliability (i.e. low variance) of the face-rub test 12. In this study, our parameters were tailored to it and the summations seen for facial analgesia are probably more reliable than the lack of summation for hot-plate tests. (Tail flicks could not be used as noted above.) There may be either a denser concentration of nociceptive fibers in T N M than in subnucleus caudalis or the firing thresholds of fibers in the latter may be significantly higher than those of the more rostral locus. This notion is based on the profoundly different thresholds for aversive reaction characterizing the two loci, and complements our related findings of larger effects of rostral vs caudal lesions on orofacial nociception3,12. Nevertheless, T N M resembles caudalis in that localized P G C microinjections in doses producing some analgesia for orofacial nociception was without effect on aversive T N M stimulation thresholds: for the face-rub latency, 0.35 #g morphine produced a 152.1% of control score here; 1.0/tg produced > 270 ~ of control previously 11. Moreover 10 mg/kg subcutaneous injection at least doubles the aversive threshold for stimulation of both T N M and caudalisla,Xa, 14. Two related hypotheses consistent with these effects are: (1) our P G C locus inhibits a T N M (or caudalis, ref. 11) locus different than where we were stimulating; (2) systematic injection accesses multiple analgesia substrates of which only a fraction are accessed by local microinjection. These notions are consistent with the summation effects also reported here. The lack of effect of systemic morphine on aversive thresholds to brain stimulation central (medial) to the main sensory nucleus is entirely consistent with and is dis cussed in our previous reports13,14. Supported by N I H Grants GM23696 and DE05204.
1 Akaike, A., Shibata, T., Satoh, M. and Takagi, H., Analgesia induced by microinjection of morphine into, and electrical stimulation of, the nucleus reticularis paragigantocellularis of rat medulla oblongata, Neuropharmacology, 17 (1978) 775-778. 2 Basbaum, A. I. and Fields, H. L., The origin of descending pathways in the dorsolateral funiculus of the spinal cord of the cat and rat: further studies on the anatomy of pain modulation, J. Comp. Neurol., 187 (1979) 513-532. 3 Broton, J. and Rosenfeld, J. P., The necessity of rostral (vs caudal) trigeminal thalamic projections for orofacial nociception, .4mer. Pain Soc., Abstr. (1980) 17. 4 Dickenson, A. D., Oliveras, J.-L. and Besson, J.-M., Role of the nucleus raphe magnus in opiate analgesia as studied by the microinjection technique in the rat, Brain Research, 170 (1979) 95-111. 5 Gallager, D. W. and Pert, A., Afferents to brain stem nuclei (brain stem raphe, n. reticularis
348
6 7 8 9 10
I1
12 13 14
15 16
17
pontis caudalis, and n. gigantocellularis) in the rat as demonstrated by microiontophoretically applied horseradish peroxidase, Brain Research, 144 (1978) 257-275. Khayyat, G., Yu, Y. and King, R., Response pattern to noxious and non-noxious stimuli in rostral trigeminal relay nuclei, Brain Research, 97 (1975) 47-60. Lomax, P., The distribution of morphine following intracerebral microinjection, Experientia (Basel), 22 (1966) 249-250. Mayer, D. J. and Price, D. D., Central nervous system mechanisms of analgesia, Pain, 2 (1976) 379~,04. Pomeroy, S. L. and Behbehani, M. M., Physiologic evidence for a projection from periaqueductal gray to nucleus raphe magnus in the rat, Brain Research, 176 (1979) 143-147. Rosenfeld, J. P. and Keresztes-Nagy, P., Differential effects of intracerebrally microinjected enkephalin analogs on centrally versus peripherally induced pain, and evidence for a facial versus lower body analgesic effect, Pain, 9 (1980) 171-181. Rosenfeld, J. P. and Stocco, S., Differential effects of systemic versus intracranial injection of opiates on central, orofacial and lower body nociception: somatotypy in bulbar analgesia systems, Pain, 9 (1980) 307-318. Rosenfeld, J. P., Clavier, R. M. and Broton, J. G., Bilateral and unilateral antinociceptive effects of rostral trigeminal nuclear complex lesions in rats, Brain Research, 157 (1978) 147-152. Rosenfeld, J. P. and Holzman, B. S., Differential effect of morphine on stimulation of primary versus higher order trigeminal terminals, Brain Research, 124 (1977) 367-372. Rosenfeld, J. P. and Vickery, J. L., Differential effect of morphine on trigeminal nucleus versus reticular aversive stimulation: independence of negative effects from stimulation parameters, Pain, 2 (1976) 405-416. Rudy, T. A. and Yeung, J. C., Profound potentiation of morphine's analgetic potency by concurrent intrathecal and intraventricular administration, Neurosci. Abstr., 5 (1979) 615. Satoh, M., Akaike, A. and Takagi, H., Excitation by morphine and enkephalin of single neurons of nucleus reticularis paragigantocellularis in the rat : a probable mechanism of analgesic action of opioids, Brain Research, 169 (1979) 406-410. Satoh, M., Akaike, A., Nakazawa, T. and Takagi, H., Evidence for involvement of separate mechanisms in the production of analgesia by electrical stimulation of the nucleus reticularis paragigantocellularis and nucleus raphe magnus in the rat, Brain Research, 194 (1980) 525-529.