- Email: [email protected]
Naloxone reversible inhibition of reticular neurones in the rat caudal medulla produced by electrical stimulation of the periaqueductal grey matter
249
Pam, 15 (1983) 249-263 Elsevier Biomedical Press
Naloxone Reversible Inhibition of Reticular Neurones in the Rat Caudal Medulla Produced by Electrical Stimulation of the Periaqueductal Grey Matter R.G. Hill ‘, R. Morris and M.V. Sofroniew ’ Department
of F~arrnaeo~o~~,, Un~ver.~~t~of Bristol medical &hod, (Received
2t October
1981, accepted
Bristol BS8 1 TD (Great Britain)
3 August
1982)
Summary Chronic dorsal periaqueductal grey matter electrodes were implanted into adult rats under pentobarbitone anaesthesia. Stimulating these electrodes (25-300 PA) produced behavioural analgesia in 23 of 44 rats tested. In rats given the opiate antagonist naloxone attenuation of this analgesia was seen. In 14 rats displaying behavioural analgesia to periaqueductal grey matter stimulation acute electrophysiological experiments were performed under urethane anaesthesia. Microelectrode recordings were made from neurones, excited by noxious heat or pinch applied to the limbs and tail, and located in the reticular formation of the caudal medulla. Stimulation of the periaqu~uctal grey matter at an intensity sufficient to produce analgesia in the conscious animal produced direct inhibition of the firing of 62% of neurones tested, excited 23%, had no effect on 14% and attenuated the nociceptive responses of 66%. The inhibitions were characteristically long. Local application of n%loxone by microiontophoresis attenuated these long inhibitions in 11 out of 16 neurones tested. Immunohistochemical localization of &-endorphin containing structures in the vicinity of stimulating and recording sites suggested that the naloxone sensitive i~bition of nociceptive neuronal responses in caudal medulla reticular formation may be due to activation of /3-endorphin fibres descending through the periaqueductal area to the caudal medulla.
’ Correspondence
to R.G. Hill. Presgnt address: Pharmacology Laboratory, Parke-Davis Research Unit, Cambridge CB2 2QU, Great Britain. ’ Present address: Department of Human Anatomy, University of Oxford, South Parks Road, Oxford OX1 3QX, Great Britain.
0304-3959/83/0000-0000/$03.00
0 1983 Elsevier Science Publishers
Introduction
It is IIOWwellestablished that electrical stimulation of periayueductal or periventricular structures may produce analgesia in laboratory animals [33,34.41,45.55] and in man [ 1,5,24,42]. The neuronal mechanisms producing this analgesia are likely to be complex [34,45] and involve a number of different neurotransmitter systems notably the catecholamines [3.52] and Shydroxytryptamine [2,16]. In many cases analgesia, produced by stimulation of the periaqueductal grey matter (PAG). can be reversed by administration of the opiate antagonist. naloxone [ 1,4,19,36. but see also 541, suggesting the involvement of an endogenous opioid neur~~trailsmitter or neuromodulator in addition to the monoamines. Electrophysiological studies have shown that PAG and raphe stimulation will both inhibit the nociceptive responses of neurones in the dorsal horn of the spinal cord [I 1.13,14,17.25] and spinal trigeminal nucleus [3 I]. and. furthermore. in some cases there is no effect on responses of neurones to innocuous stimuli [ 13.14,25,31]. In contrast to the behavioural effects of stimulating these midbrain structures. the inhibitory effects on single spinal sensory neurones are in most cases naloxone resistant [ 11.13.17,19.25] suggesting that the involvement of the endogenous opioids may be at a more rostra1 level of the neuroaxis. As the axons of a large proportion of spinal secondary sensory neurones synapse within the reticular formation rather than passing directly to the thalamus [9.10] and many reticular neurones are powerfully excited by noxious peripheral stimuli [7,12.21.32.35,36,40] this area appeared worthy of further study. It has already been demonstrated that electrical stimulation of analgesia supporting mesencephalic sites will inhibit reticular neurones [21,36,40] but little is known about the neurotransmitters involved in this process. We have therefore made a study of the effects of PAG stimulation on the nociceptive responses of neurones in the reticular formation (CRF) of the rat caudal medulla [14,21,35]. In particular we wished to explore the possibility that an opioid operated descending system had a direct inhibitory effect upon these neurones as we have previously shown that their inhibition. by a variety of peripheral stimuli, is reversed by naloxone [ 18,353 and that their firing is readily depressed by microiontophoretic application of opioids [22]. The distribution of met-enkephalin in the caudal medulla has been well described f44.511, but previous studies on /3-endorphin have concentrated on more rostra1 structures [8]. There is now evidence that neurones in the hypothalamic arcuate nucleus, where /3-endorphin containing neurones are concentrated [8,46]. project to the caudal medulla [ 15,43,45,53]. We have therefore used immunocytochemical techniques to detect ~-endorphin containing fibres in the vicinity of our recording sites in CRF. The PAG is known to contain both ,&endorphin and enkephahn fibres [X,15] but it was considered worthwhile to re-examine the distribution of p-endorpmn to check for correlation with the location of our stimulation sites. The behavioural effects of PAG stimulation are complex and do not always include analgesia [29,30]. It was therefore considered essential to demonstrate that our PAG electrode pfacements would produce analgesia, before any attempt was made to examine the effects of PAG stimulation on the responses of single neurones.
251
The identification of mechanisms operated by endogenous opioids, at the present time, is dependent on the use of the opiate antagonist, naloxone [19]. It was therefore thought important to test both PAG evoked behavioural analgesia and the consequent inhibitory effects on single reticular neurones with this drug. In the single neurone studies naloxone was applied by microiontophoresis [19] to ensure that the action of the drug was restricted to the vicinity of the recorded neurone. This investigation has produced evidence that PAG stimulation, employing parameters that will sustain behavioural analgesia, produces a prolonged inhibition of caudal reticular formation neurones. This inhibition can be reversed by naloxone, applied microiontophoretically, and the distribution of fl-endorphin containing fibres within CRF is such that it is reasonable to suppose that the inhibition may be operated by synaptic release of /3-endorphin.
Methods Adult, male rats, of the MRC Porton strain, were anaesthetized with pentobarbitone (60 mg/kg) and implanted with PAG stimulating electrodes prepared from twisted, teflon coated stainless steel wire (Clark Electra-Medical SS-ST or SS-ST) with the insulation intact except at the electrode tips. The electrode pedestal was held in position by 10 BA stainless steel screws tapped into the skull and by fixation wires threaded through small bilateral burr holes, the whole implant being imbedded in dental acrylic (Simplex). Stereotaxic co-ordinates of the electrode tips were in the range of A 1.9 to P 0.5, L 0.0 to 1.0 and H -0.5 [26] and were broadly similar to those used by other workers [34,36,41,55] although it should be noted that the majority of our placements were aimed at PAG proper with very few in the vicinity of nucleus raphe dorsalis (e.g. contrast with [2]). Animals were caged individually and kept under a 12 h light, 12 h dark regime (dark 8 p.m.-8 a.m.). Testing was carried out 4-6 days postoperatively and took place between 9 a.m. and 8 p.m. The behaviour of the animals was observed in a circular open-field measuring 1.5 m in diameter. After a short period of habituation the animals showed little locomotor activity and tended to s& in one position, occasionally grooming. PAG stimulation (50 Hz, 0.5 msec duration unipolar pulses) was applied for 15-20 set at the following current steps: 25, 50, 100, 150, 200, 250 and 300 PA. The threshold and nature of any motor changes were noted. Current intensities above those causing a distinct motor arousal were only examined in a few individuaIs and in all cases were found to augment the motor behaviour. In all other animals the appearance of motor behaviour was used as the cut-off intensity for subsequent behavioural tests. Analgesia was assessed in two ways. The simplest method consisted of pinching the animal’s tail with a pair of artery forceps. The animal characteristically flicked its tail, oriented to the stimulus or tried to escape, and occasionally vocalized; all of which would take place within 2 set of the stimulus application. Analgesia was considered to have been produced by PAG stimulation if all components of the response were abolished.
The taif irnnl~rsi(~n test [XI] was used as a secctnd method of assessmp an;tlgc~i;~. In this test the animals were lightly restrained in a plastic tube and the end IO cm of tail immersed in water at 54°C. The end point was taken as a clear flick of the tail. If this did not occur within 15 set the tail was removed from the water to avoid tissue damage. The responses were used to calculate an index of analgesia [20,39]. PA<; stimulation was considered to have produced analgesia if it produced a figure abo\~ 30%. which was more than two standard deviations from the mean control response. Naloxone, for injection, was prepared as a solution of the hydrochloride salt (Endo) in 0.9% w/v sodium chloride solution. and doses are expressed as those of the salt. Some of those animals showing analgesia in response to PAG stimulation were further studied in acute. electrophysiological experiments under urethane anaesthesia ( 1.22 g/kg i.p.). Animals not showing analgesia were not used as a preGous study f39] found that in such animals there were no inhibitory effects on medulla neurones. Rats were mounted in a stereotaxic frame and prepared for recording from the caudal medulla as previously described [1X,35]. Extracellular action potentials were recorded from single neurones using the 4 M NaCl solution filled crntre barrel of a InultibarrelIed pipette with an overatl tip diameter of 4--6 pm {lx]. Outer barrels of the pipettes contained solutions of drugs for application b\; microiontophoresis selected from sodium I.-glutamate (Sigma) 0.5 M. pH 8.5, y-aminobutyric acid (GABA, Sigma) 0.5 M. pH 3.5. met-ei~kephaIil1 (WeIl~(~me), ~-end~~rphi~l ( Pel~illsL~la) I mM in 150 mM NaCl and naloxone hydrochloride (Endo) 50 mM. One barrel always contained 1 M NaCl solution and wab used for automatic balancing of the iontophoresis currents and another contained a solution of pontamine sky blue (Gurr) 2.5% w/‘v in sodium acetate buffer so that spots of dye (IO0 pm approx.) could be deposited during an experiment. This allowed the rrcomtruction of electrode tracks and the verification of recording positions from photographs of 100 pm transverse, frozen sections of the medulla. Action potentials were displayed on an oscilloscope and a trigger circuit was used to generate standard pulses from those action potentials crossing a voltage gate. These pulses were counted and displayed as an analogue rate on a multichannel chart recorder together with drug application event pulses. Noxious stimuli were applied to the hind limbs and. tail either by heating to between 45 and 55°C with a radiant heat lamp or by immersion in warm water, the temperature of the skin being monitored in both cases with a miniature thermocouple, or by pinching with haemostats. Innocuous stimuli used included tapping and gentle stroking of the skin with a camel hair brush. At the end of each electrophysiological experiment a lesion was made by passing a brief DC current through the PAG stimulating electrode. the animal was perfused through the aorta with 4% paraf~~rmaldeh~d~ solution and the brain was removed for subsequent histology. Immunohistochemical techniques used have recently been described [38,46-491 but briefly, an antiserum to P-endorphin was prepared using synthetic peptide (Serva) conjugated with thyroglobulin and injected into rabbits. Tissue fixed @-endorpbin was visualized by using antiserum at a dilution of 1 in 300 and sites of antiserum binding were located with the immunoperoxidase method. The specificity
253
of the antiserum used was tested by incubation with 50 pg/ml fi-endorphin or ACTH for 4 h prior to application to the tissue sections. Staining with the antiserum was completely abolished by pre-incubation with /3-endorphin but was unaffected when ACTH was used. Photographs of stained sections were taken and maps of P-endorphin containing structures were constructed using a camera lucida.
Results Behavioural experiments Forty-four animals were found to respond to PAG stimulation. Motor arousal, sometimes in the form of a co-ordinated running behaviour, was seen particularly with higher stimulus currents (up to 300 PA) and longer pulse trains (up to 30 set). Twenty-three of the 44 rats showed a clear analgesia to noxious thermal or mechanical stimulation of the tail or hind paws following electrical stimulation of the PAG. The analgesia usually outlasted the stimulus producing it, as described previously by others [41], but motor arousal terminated with the stimulus so that there was little difficulty in separating the phenomena in all 23 animals. In 11 rats the analgesia produced by PAG stimulation was tested after administration of either 2 or 10 mg/kg of naloxone i.p. Four of the 8 rats given 2 mg/kg naloxone showed reversal of the PAG evoked analgesia and all 3 rats given 10 mg/kg showed reversal. Lesions located in frozen sections of the brain stem showed that stimulating electrode tips in animals displaying analgesia were located within the confines of the central grey matter between A 1.6 and P 0.5, and this is illustrated in Fig. 1. Electrophysiological experiments Fourteen animals showing analgesia phase of the investigation. Micropipettes
after PAG stimulation were used in this were inserted perpendicular to the surface
PO.5 Fig. I. Transverse section of rat brain stem at intervals indicated by the stereotaxic co-ordinates beneath each section, showing stimulation sites capable of supporting behavioural analgesia in the 14 rats used in electrophysiological experiments. The symbols are placed at positions determined by lesions made through the implanted stimulating electrodes and indicate the stimulus current necessary for the production of analgesia, such that *, 25-100 PIA; 0, 100-200 pA; and A. 200-300 PA. AP zero was the intra-aural line.
of the medulla, within 1.5 mm of the midline and approximately 1 mm caudal to the obex. While the electrode was advanced. continuous microiontophoretic application of glutamate was used as a search stimulus. When a neurone was located it was tested in order to determine whether it would respond to noxious and/or innocuous peripheral stimuli. When a response to peripheral stimulation was established this was then challenged with PAG stimulation. using stimulus intensities earlier found adequate to produce analgesia in the same animal. The responses of a total of 62 neurones were examined in this way. Six of these were identified as dorsal column nucleus (DCN) neurones as they were located superficially within the medulla and responded only to innocuous stimuli applied to a restricted receptive field on the limbs. None of these DCN neurones showed any attenuation of their sensory responses by PAG stimulation. The remaining 56 neurones were all located deeper in the reticular part of the medulla, in the vicinity of nucleus reticularis ventralis (NRV) and responded in a characteristic manner, as previously described [7.18.2 1.351. No responses to tapping or brushing were seen but noxious stimuli readily evoked excitation, or in some cases inhibition, of these neurones. Receptive fields were large encompassing at least one hind limb and the tail and sometimes included the majority of the body surface. PAG stimulation directly inhibited 32 of these neurones with a further 3 neurones being inhibited after a brief excitation. thus. 62% of the CRF neurones tested were inhibited by analgesia supporting PAG stimulation and, most interestingly. the time course of the inhibition resembled that of the behavioural analgesia being prolonged beyond the stimulus and depressing the firing of the neurone for several minutes. A further 13 neurones (23%) were excited by PAG stimulation, and although no assumptions were made about the possible significance of this observation this group received no further study at this time. The remaining 8 neurones (14%) were not affected by PAG stimulation in any way. On 21 neurones with clear and reproducible excitatory responses to peripheral noxious stimuli. the ability of analgesia supporting PAG stimulation to attenuate these responses was tested. Following a series of control responses to noxious stimuli, a period of PAG stimulation was given and immediately followed by a noxious stimulus, which was then repeated at regular intervals until the neuronal response had regained its control size. Fourteen (66%) of the neurones studied in this way displayed either complete abolition of the neuronal nociceptive response or a reduction in its size. The reduction in response size characteristically outlasted the stimulus by several minutes and had a similar time course to the effects of the same PAG stimulus on the animal’s nociceptive behaviour. Microiontophoretic application of GABA or met-enkephalin readily depressed the firing of CRF neurones and attenuated their nociceptive responses as previously described [ 18,221. The actions of ,f%endorphin. applied with currents in the range 10-60 nA, were examined on 13 neurones, of which 8 showed a depression of firing rate, 2 showed a brief excitation followed by depression. 1 was excited and 2 were unaffected. Met-enkephalin (20-100 nA) depressed 16 of 17 neurones tested. Naloxone (2-40 nA) reversed the depression produced by P-endorphin on 2 neurones and that produced by met-enkephalin in a further 3 neurones but had little effect on that
255 I NALOX 2
L
.
D
[3
CJ .
I
.
c3
cl
P .
cl
I
i
D B-END I
*
20
GABA 10
Fig. 2. An analogue rate meter record of the firing of an NRV neurone, maintained by the continuous microiontophoretic application of glutamate 20 nA. GABA 10 nA (solid bars) and P-endorphin 20 nA (open bars) were both found to depress the firing of this neurone, although during the concurrent application of naloxone 2 nA (long open bar) the response to /3-endorphin was abolished. Partial recovery was seen 10 min after the end of the naloxone application. Time calibration is 2 min.
produced by GABA as illustrated in Fig. 2. The direct inhibitory effects of PAG stimulation on the spontaneous or glutamate maintained firing of CRF neurones were tested in the presence of naloxone on 13 neurones and on 9 of these the inhibitory effects were reduced. On a further 3 neurones naloxone was tested against the PAG induced depression of nociceptive responses and reduced the inhibition in 2 cases. The remaining result was equivocal as the control nociceptive response was enhanced by naloxone. In Fig. 3 an experiment on a CRF neurone excited by a continuous application of glutamate is shown. PAG stimulation depressed the firing of this neurone and the inhibitory effect was prolonged. After several minutes iontophoretic application of
I
I
I
I
Naloxone
20na I
PAG stim
Fig. 3. An analogue rate meter record of the firing of an NRV neurone maintained by the continuous application of glutamate 20 nA. Electrical stimulation applied to the PAG (indicated by solid bars) produced a prolonged depression of firing and after several minutes microiontophoretic application of naloxone 20 nA this inhibition was blocked. Recovery of the depressant effect of PAG stimulation was seen when the naloxone application was discontinued. The rat used in this study had previously shown analgesia to noxious tail pinch following PAG stimulation. This analgesia was reversed by 10 mg/kg of naloxone.
Tail pinch . . LO 7
.
.
.
.
.
.
Naloxone
.
.
.
.
.
I
.
.
.
.
I
Hz
PAG “stir-n
Fig. 4. An analogue rate meter record of the firing of an NRV neurone induced by no~tous pinching of the tail (solid bars). Electrical stimulation of the PAG (hatched bar below the record) initially abolished the excitatory response to noxious pinch and produced attenuation outlasting the stimulus by several minutes. Iontophoretic application of naloxone 6 nA did not prevent the initial abolition of the response but considerably accelerated its recovery.
naloxone, the inhibitory effect of PAG stimulation was lost. and partial recovery of the inhibitory response was seen after termination of the naloxone application. In Fig. 4 an experiment of the second type is illustrated. Excitatory responses of the neurone were evoked by periodic noxious pinching of the tail and PAG stimulation was seen to produce an attenuation of this nociceptive response. When naloxone was applied iontophoretically it had no effect on the size of the control
Ftg. 5. A transverse section of the caudal medulla of the rat at a level corresponding to I mm caudal to the obex. Plotted on this section are the locations of I I CRF neurones (open triangles) which showed naloxone reversible inhibition following electrical stimulation of the PAG, and the location of fibres and terminals showing /I-endorphin immunoreactivity (dots). ntV. caudal trigeminal nucleus: gr. gracile nucleus; nts, nucleus of the solitary tract: nrv, nucleus reticularis ventralis: ap. area postrema: p, pyramidal tract
251
responses to noxious stimulation but did reduce the effectiveness of PAG stimulation. The most striking effect was on the time course of recovery of the nociceptive response as even during the application of naloxone, the first noxious stimulus after PAG stimulation was without effect. Immunohistochemical localization of /3-endorphin and reconstruction of recording and stimulation sites The location of P-endorphin containing fibres was plotted onto diagrammatic transverse sections of the rat brain at appropriate AP co-ordinates matching recording sites in the CRF and stimulation sites within the PAG. In Fig. 5 the distribution of P-endorphin in the caudal medulla is shown together with the
Fig. 6. Photomontage showing location of branching /3-endorphin and b correspond to the regions thus identified in Fig. 5.
containing
fibres within CRF. Panels a
Fig. 7. A transverse section of rat brain stem at A 1.6 showing the location (open diamonds) of the 5 PAG stimulation sites at this AP plane which produced analgesia with currents < 100 PA and the distribution of /3-endorphin immunoreactivity (dots). Note the correspondence of /3-endorphin fibres and stimulation sites. A further descending tract of &endorphin fibres runs ventrolateral to PAG. PAG. periaqueductal grey; LM, medial lemniscus; SNC, substantia nigra pars compacta.
locations of those 1 I neurones found to exhibit naloxone reversible inhibition following PAG stimulation. P-Endorphin fibres can be seen to be in a patchy distribution in most parts of the CRF but do not spread laterally into the caudal trigeminal nucleus to any significant extent. Although only scattered individual fibres were found, and no dense beds of terminals, these individual fibres seemed to give off a large number of short collateral or terminal branches and this is illustrated in Fig. 6. In Fig. 7 the distribution of ~-endorp~n fibres within PAG is shown together with the location of 5 stimuIation sites from animals with stimulus currents for analgesia that were clearly lower than the currents needed to evoke motor activity. It was thought that these experiments would provide reliable location of the actual structures stimulated as the effects produced would have the smallest chance of being attributable to stimulus spread. All of these stimulus sites were clearly within the PAG and overlapped with the main group of ,&endorphin containing fibres passing through this area.
In this study we have obtained evidence that a proportion of the reticular formation neurones within the caudal medulla of the rat, which are exclusively responsive to noxious inputs, are strongly inhibited by trains of electrical stimuli applied within the PAG. Furthermore, the stimuli producing these inhibitory effects
259
are sufficient to induce analgesia in the same animals, exhibiting a similar time course to the neuronal inhibitions. In view of the importance of the reticular formation as a relay in the ascending nociceptive system [9,10,12] it is tempting to relate these behavioural and neuronal responses to PAG stimulation, although it is noteworthy that reticular neurones have complex projections both descending [23] and ascending [9] the neuroaxis and are likely to be polyfunctional. The inhibitory effects of PAG stimulation were clearly seen on neurones whose firing was sustained by microiontophoretic application of the excitatory amino acid, glutamate, suggesting that the inhibitory mechanism was, at least in part, postsynaptic. The confirmation of this by intracellular recording would thus be of great interest particularly in view of the prolonged time course of these inhibitions. The reversal of a proportion of the inhibitory effects of PAG stimulation by microiontophoretic naloxone raises the possibility of an endogenous opioid substance acting as a neurotransmitter, and this has been inferred from the action of naloxone in reversing behavioural analgesia evoked by PAG stimulation [ 1,4,19,20]. Application of naloxone by microiontophoresis restricts the action of the drug to the immediate vicinity of the micropipette. Thus, any endogenous opioid acting to inhibit CRF neurones must be released locally. It has been suggested that naloxone may antagonize the actions of GABA in addition to those of opioids [19] and this observation is likely to be important in explaining the convulsant effects seen with toxic doses of the drug. However, we have recently described a short lasting inhibition of NRV neurones, evoked by single stimuli applied to PAG, which can be blocked by the GABA antagonist bicuculline but not by naloxone [37, see also 321 and which is quite distinct from the prolonged inhibition, evoked by trains of PAG stimuli, that we describe here. Many hypothalamic fibres project to the PAG and to the caudal medulla [15,43,47,48] and some of these contain oxytocin, vasopressin and related peptides [48]. We have recently shown that oxytocin will depress the firing of CRF neurones 1381 and thus this substance remains a transmitter candidate for those descending inhibitions which are resistant to both bicuculline and naloxone. Descending fibres also passing close to PAG contain catecholamines [52] and Shydroxytryptamine [2,50] and these too are worthy of consideration as transmitter candidates in this system. It is perhaps relevant that depletion of Shydroxytryptamine with parachlorophenylalanine will reduce the analgesia produced by stimulation of ventral PAG sites, close to the dorsal raphe, but does not reduce the analgesia evoked from more dorsal sites, such as those used in the present study [2j. Enkephalin containing fibres are found on the dorsal and lateral borders of NRV [44,51] but it is difficult to explain how they might be stimulated by pathways activated by electrodes within the PAG. There is evidence for an excitatory connection from PAG to nucleus raphe magnus and neurones in this region have descending enkephalin containing fibres [23]. However, these do not seem to terminate in the CRF as we have shown that horseradish peroxidase injected to the CRF does not undergo retrograde transport to raphe magnus although transport is seen direct to PAG [unpublished observations and 61. Our immunohistochemical studies suggest that P-endorphin may be a better candidate for the transmitter of naloxone sensitive
inhibition in the (‘RF as demonstration of the presence of /I-endorphrn 1’1hres within this region has been made [thla paper and i5] and injection of horseradish peroxidase to the (‘RF results in the retrograde labelling of neuronal cell bodies in the arcuate nucleus [47]. believed to be the origin of most P-endorphin containing axons in the central nervous hystem [X,15.46]. The stimulus sites which ue ~‘OLIIK! to have the lowest current threshold for producing beha\,ioural analgesia Lvere clearly located over the /!I-endorphin containing fibres passing through PAG at the same AP co-ordinates and it is tempting to attribute the naloxone sensiti\,c a&g&c and neuronal inhibitorv effects of PACT stimulation to activation of these fibres. ‘This is also consistent with the observations that /%endorphin can be released into the cerebrospinal fluid by analgesia producing pericentricular or PAG .stimulation [24] and that microinjection of ,&endorphin into midbrain reticular formation produces profound analgesia [27]. It seems unlikely that a /3-endorphin operated inhibition of caudal reticular neurones constitutes an adequate neuronal substrate for analgesia produced by PAG stimulation. However. when this is combined with a similar inhibition of more rostra] reticular neurones [36] and the likelihood that significant amount of /I-endorphin would be released in the vicinity of PAG and periventricular thalamus neurones thought to have a role in nociception [9,10,12,29] then a powerful net analgesic effect may well result. It may be possible to activate such an opioid operated descending inhibitory system by intense peripheral stimuli. as has been suggested by Besson and his colleagues [28]. for we have found that the inhibition of caudal reticular neurones by noxious thermal or mechanical stimuli [1X.35] or by vaginal distention [1X.20] is also naloxone reversible.
Acknowledgements
We are grateful to Endo for the gift of naloxone. Skilled technical assistance was provided by C. Allen, Catherine Crozier and P. Darby. This work is supported by grants from the Wellcome Trust, Medical Research Council and Royal Society.
References I Adams, J.E.. Naloxone reversal of analgesia produced hv brain stimulation in the human. PaIn. 2 (1976) 161-166. 2 Akll. H. and Mayer. D.J.. Antagonism of stimulation-produced analgesia hy p-CPA. a \erotonin bynthesis inhibitor. Brain Res.. 44 (1972) 692-697. 3 Akil. H. and Liebeskind. J.C.. Monoaminergic mechanism5 of stimulation-produced analge%l, Brain Res.. 94 (1975) 279-296. 4 Akil. H.. Mayer. D.J. and Liebeskind. J.C.. Antagonism of stimulation-produced analgesia bv naloxone, a narcotic antagonist, Science. I91 (1976) 96 I --962. material &bated in 5 Akil, H., Richardson. D.E., Hughes. J. and Barchas. J.A.. Enkephalin-like ventricular 46X-465.
cerebrospinal
fluid of pain patients
after analgetic
focal stimulation.
Science.
201 (1978)
261
6 Benz, A.J., Ascending
and descending
Abstr.. 5 (1979) Abstr. 2053. 7 Benjamin, R.M., Single neurones
projections
in the rat medulla
to the periaqueductal responsive
gray of the rat, Neurosci.
to nociceptive
stimulation.
Brain Res.,
24 (1970) 525-529. 8 Bloom, F.. Battenberg, E., Rossier, J., Ling, N. and Guillemin, R.. Neurones containing /3-endorphin in rat brain exist separately from those containing enkephalin: immunocytochemical studies. Proc. nat. Acad. Sci. (Wash.). 75 (1978) 1591-1595. 9 Bowsher, D., Collateralization of reticulothalamic
projections
and its possible
significance
for pain
sensation, Adv. Pain Res. Ther., I (1976) 245-249. 10 Bowsher, D., Role of the reticular formation in responses to noxious stimulation, pain, 2 (1976) 361-378. I1 Carstens, E., Klumpp. D. and Zimmerman, M., The opiate antagonist, naloxone. does not affect descending inhibition from midbrain of nociceptive spinal neuronai discharges in the cat. Neurosci. Lett., 11 (1979) 323-327. 12 Casey, K.L., Reticular formation and pain: toward a unifying concept, Res. Publ. Ass. nerv. ment. Dis., 58 (1980) 93-105. 13 Duggan. A.W. and Griersmith, B.T., Inhibition of the spinal transmission of nociceptive information by supraspinal stimulation in the cat, Pain, 6 (1976) 149- 161. 14 Fields, H.L., Basbaum, A.I., Clanton, C.H. and Anderson, SD.. Nucleus raphe magnus inhibition of spinal cord dorsal horn neurons, Brain Res., 126 (1977) 441-453. 15 Finley, J.C.W., Lindstrom. P. and Petrusz. P. Immunocytochemicat localization of @-endorphin containing neurons in the rat brain, Neur~nd~rinoiogy, 33 (1981) 28-42. 16 Hayes. R.L., Newlon, P.G.. Rosecrans, J.A. and Mayer, D.J., Reduction of stimulation-produced analgesia by lysergic acid diethylamide. a depressor of serotonergic activity, Brain Res., 122 (1977) 367-372. 17 Hayes, R.L.. Price, D.D., Ruda, M. and Dubner, R., Suppression of nociceptive responses in the primate by electrical stimulation of the brain or morphine administration: behavioural and electrophysiological comparisons, Brain Res., 167 (1979) 418-421. 18 Hill, R.G., Opioid peptides as inhibitory neurotransmitters acting on sensory neurones in the rat medulla. In: C. Ajmone Marsan and W.Z. Traczyk (Eds.), Neuropeptides and Neural Transmission, Raven Press, New York, 1980, pp. 199-208. 19 Hill, R.G., The status of naloxone in the identification of pain control mechanisms operated by endogenous opioids, Neurosci. Lett., 21 (1981) 217-222. 20 Hill, R.G. and Ayfiffe, S.J.. The antin~i~ptive effect of vaginal stimulation in the rat is reduced by naloxone. Pharmacol. Biochem. Behav.. 14 (1981) 631-632. 21 Hill, R.G. and Morris, R., Responses of nucleus reticularis ventralis neurones of the rat to noxious stimuli and to electrical stimulation of the periaqueductal grey and raphe dorsalis, J. Physiol. (Lond.), 301 (1980) 3%39P. 22 Hill, R.G. and Morris, R., Responses of neurones in medial caudal medulla to iontophoretically applied opioids, Brit. J. Pharmacol., 70 (1980) 147P. 23 Hokfelt, T., Terenius, L., Kuypers, H.G.J.M. and Dann, O., Evidence for enkephalin immunoreactive neurons in the medulla oblongata projecting to the spinal cord, Neurosci. Lett., 14 (1979) 55-60. 24 Hosobuchi, Y.. Rossier. J., Bloom, F.E. and Guillemin. R., Stimulation of human periaqueductal gray for pain rehef increases immunoreactive /3-endorphin in ventricular fluid, Science, 203 (1979) 279-281. 25 Jurna. I.. Effect of stimulation in the periaqueducta~ grey matter on activity in ascending axons of the rat spinal cord: selective inhibition of activity evoked by afferent AS ad C fibre stimulation and failure of naloxone to reduce inhibition, Brain Res., 196 (1980) 33-42. 26 Konig. J.F.R. and Klippel, R.A., The Rat Brain. a Stereotaxic Atlas of the Forebrain and Lower Parts of the Brain Stem, Williams and Wilkins, Baltimore, Md., 1962. 27 Kuraishi. Y., Satoh. M., Harada, Y.. Akaike, A.. Shibata, T. and Takagi, H., Analgesic action of intrathecal and intracerebral P-endorphin in rats: a comparison with morphine, Europ. J. Pharmacol.. 67 (1980) 43-146. 28 Le Bars. D., Chitour, D., Kraus, diffuse noxious inhibitory controls
E.. Dickenson, A.H. and Besson, J.M.. Effect of naloxone (DNIC) in the rat, Brain Res.. 204 (1981),387-402.
upon
29 Ltebeskind. J.C. and Mayer, DJ Sornato~enmr~ txorhd rcqxmm in (he mcsencrph~ltc ccn~r;il gr;t\ matter ~>fthe rat. ISraIn l&b.. 27 (lY71) 133 151. 30 Liebman. J.M.. Majer. D.J. and Lleheskind. J.C.. Mesencrphalic central gray lesions and fear-rn
J.H., Inhibitory effects of nucleus raphe magnus nucleus to nocicepti\r compared with non-nocicepti\c
on neuronal tnputs. Pain.
32 Lovick. T.A. and Wolstencroft. J.H.. Inhibitton from nucleus raphe magnua of tooth pulp responses in medial reticular neurnnes of the cat can be antagonized h! hicuculline. Neurosci. Lett.. 19 (1980) 3255330. 33 Mayer. D.J. and Liebeskind. J.C.. Pam reduction by focal electrical stimulatton of the hrain: dn anatomical and behavioural analysis. Brain Res.. 68 (1974) 73-93. 34 Mayet-. D.J. and Price. D.D.. Central nervous system mechanisms of analgesia. Pain, 2 (1976) 379-404. 35 Mayer. M.L. and Hill, R.G.. The effects of intravenous fentanyl. morphine and naloxone on nociceptive responses of neurones m the rat caudal medulla. Neuropharmacology, 17 (1978) 533-539. 36 Mohrlands. J.S. and Gebhart. G.F.. Effects of focal electrical stimulation and morphine mtcroinjecttons m the periaqueductal gray of the rat mesencephalon on neuronal activity in the medullary reticular formation. Brain Res.. 201 (1980) 23-37. 37 Morris. R. and Hill, R.G.. Inhibitory mechanisms in the nucleus reticular& ventralis (NRV) of the rat. Neurosci. Lett.. Suppl. 5 (1980) S389. 38 Morris. R.. Salt. T.E.. Sofroniew, M.V. and Hill, R.G.. Actions of microiontophoretically applied nxytocin and immunohistochemical localization of oxytocin. vasopressin and neurophyain in the rat caudal medulla, Neurosci. Lett.. IX (1980) 163- 168. 39 Morris. R., Hill, R.G.. Cahuaac. P.M.B. and Salt, T.E.. The involvement of endogenous opioids in the descending inhibition of caudal trigeminal and reticular nociceptive neurones evoked by analgesia producing periaqueductal grey atnnulation m the rat. In: B. Mattheus and R.G. Hill (Eda.). Anatomical, Physiological and Pharmacological Aspects of Trigeminal Pam. Excerpta Medica. Amsterdam. 1982, pp. 195-206. 40 Morrou. T.J. and Casey, K.L.. Analgesia produced by mesencephalic stimulation: effect on hulhoreticular neurons. Adv. Pain Rea. Ther.. 1 (1976) 50% 5 10. 41 Reynolds. D.Y.. Surgery in the rat during electrical analgesia induced by focal brain stimulation. Science, 164 (1969) 444-445. 42 Richardson, D.E.. Central gray stimulation for control of cancer pain. Adv. Pain Res. Ther.. 2 (1979) 487 -492. 43 Saprr. C.B., Loewy, A.D.. Swanson, L.W. and Cowan. W.M.. Direct hypothalamus-autonomic connections, Brain Rea.. 117 (1976) 3055312. 44 Sar. M.. Stumpf. W.E., Miller. R.J.. Chang. R.-J. and Cuatrecasas. P.. Immunohistochemical localization of enkephalin m rat brain and spinal cord. J. camp. Neurol.. I83 (1978) 17-38. 45 Sherman. J.E. and Liebeskind. J.C.. An endorphinergic. centrifugal substrate of pam modulation: Res. Publ. Ass. nerb. ment. Dis.. 58 (1980) recent findings. current concepts and complexities. 191.-204. 46 Sofroniew. M.V.. Immunoreactive P-endorphin and ACTH in the same neurons of the hypothalamic arcuate nucleus in the rat. Amer. J. Anal., 154 (1979) 283.-289. M.V. and Schrell. U.. Hypothalamic neurons proJectmg to the rat caudal medulla 47 Sofroniew, oblongata, examined by immunoperoxidase staining of retrogradely transported horseradish peroxidase. Neurosci. Lett.. I9 (1980) 2577263. 48 Sofroniew, M.M. and Schrell. U.. Evidence for a direct proJection from oqtocm and vasopressm neurons in the hypothalamic paraventricular nucleus to the medulla oblongata: tmmunohistochemical visualization of both the horseradish peroxidase transported and the peptide produced by the same neurons, Neurosci. Lett., 22 (1981) 211-217. 49 Sofroniew, M.V. and Schrell, C.. Long term storage and regular repeated use of diluted antisera in glass staining jars for increased sensitivity, reproducibility and convenience of single- and tw-color light microscopic immunocytochemistry. J. Histochem. Cytochem.. 30 (I 982) 504-5 11.
263
50 Steinbusch, H.W.M., Distribution of serotonin-immunoreactivity in the central neurone system of the rat cell bodies and terminals, Neuroscience, 6 (1981) 557-618. 51 Uhl, G.R., Goodman, R.R., Kuhar, M.J., Childers, S.R. and Snyder, S.H., Immunohistochemical mapping of enkephalin-containing cell bodies, fibres and nerve terminals in the brain stem of the rat, Brain Res., 166 (1979) 75-94. 52 Watson, S.J., Akil, H. and Barchas, J.D., A possible role for the dorsal periventricular catecholamine bundle in stimulation produced analgesia: a behavioural and histochemical study, Brain Res., 130 (1977) 335-342. 53 Watson, S.J., Richard, V.W. and Barchas, J.D., Adreno-corticotrophin in rat brain, Immunocytochemical localization in cells and axons, Science, 200 (1978) 1180-1182. 54 Yaksh, T.L., Yeung, J.C. and Rudy, T.A., An inability to antagonize with naloxone the elevated nociceptive thresholds resulting from electrical stimulation of the mesencephalic central gray, Life Sci., 18 (1976) 1193-1198. 55 Yeung, J.C.. Yaksh, T.L. and Rudy, T.A., Concurrent mapping of brain sites for sensitivity to the direct application of morphine and focal electrical stimulation in the production of antinociception in the rat, Pain, 4 (1977) 23-40.
Pam, 15 (1983) 249-263 Elsevier Biomedical Press
Naloxone Reversible Inhibition of Reticular Neurones in the Rat Caudal Medulla Produced by Electrical Stimulation of the Periaqueductal Grey Matter R.G. Hill ‘, R. Morris and M.V. Sofroniew ’ Department
of F~arrnaeo~o~~,, Un~ver.~~t~of Bristol medical &hod, (Received
2t October
1981, accepted
Bristol BS8 1 TD (Great Britain)
3 August
1982)
Summary Chronic dorsal periaqueductal grey matter electrodes were implanted into adult rats under pentobarbitone anaesthesia. Stimulating these electrodes (25-300 PA) produced behavioural analgesia in 23 of 44 rats tested. In rats given the opiate antagonist naloxone attenuation of this analgesia was seen. In 14 rats displaying behavioural analgesia to periaqueductal grey matter stimulation acute electrophysiological experiments were performed under urethane anaesthesia. Microelectrode recordings were made from neurones, excited by noxious heat or pinch applied to the limbs and tail, and located in the reticular formation of the caudal medulla. Stimulation of the periaqu~uctal grey matter at an intensity sufficient to produce analgesia in the conscious animal produced direct inhibition of the firing of 62% of neurones tested, excited 23%, had no effect on 14% and attenuated the nociceptive responses of 66%. The inhibitions were characteristically long. Local application of n%loxone by microiontophoresis attenuated these long inhibitions in 11 out of 16 neurones tested. Immunohistochemical localization of &-endorphin containing structures in the vicinity of stimulating and recording sites suggested that the naloxone sensitive i~bition of nociceptive neuronal responses in caudal medulla reticular formation may be due to activation of /3-endorphin fibres descending through the periaqueductal area to the caudal medulla.
’ Correspondence
to R.G. Hill. Presgnt address: Pharmacology Laboratory, Parke-Davis Research Unit, Cambridge CB2 2QU, Great Britain. ’ Present address: Department of Human Anatomy, University of Oxford, South Parks Road, Oxford OX1 3QX, Great Britain.
0304-3959/83/0000-0000/$03.00
0 1983 Elsevier Science Publishers
Introduction
It is IIOWwellestablished that electrical stimulation of periayueductal or periventricular structures may produce analgesia in laboratory animals [33,34.41,45.55] and in man [ 1,5,24,42]. The neuronal mechanisms producing this analgesia are likely to be complex [34,45] and involve a number of different neurotransmitter systems notably the catecholamines [3.52] and Shydroxytryptamine [2,16]. In many cases analgesia, produced by stimulation of the periaqueductal grey matter (PAG). can be reversed by administration of the opiate antagonist. naloxone [ 1,4,19,36. but see also 541, suggesting the involvement of an endogenous opioid neur~~trailsmitter or neuromodulator in addition to the monoamines. Electrophysiological studies have shown that PAG and raphe stimulation will both inhibit the nociceptive responses of neurones in the dorsal horn of the spinal cord [I 1.13,14,17.25] and spinal trigeminal nucleus [3 I]. and. furthermore. in some cases there is no effect on responses of neurones to innocuous stimuli [ 13.14,25,31]. In contrast to the behavioural effects of stimulating these midbrain structures. the inhibitory effects on single spinal sensory neurones are in most cases naloxone resistant [ 11.13.17,19.25] suggesting that the involvement of the endogenous opioids may be at a more rostra1 level of the neuroaxis. As the axons of a large proportion of spinal secondary sensory neurones synapse within the reticular formation rather than passing directly to the thalamus [9.10] and many reticular neurones are powerfully excited by noxious peripheral stimuli [7,12.21.32.35,36,40] this area appeared worthy of further study. It has already been demonstrated that electrical stimulation of analgesia supporting mesencephalic sites will inhibit reticular neurones [21,36,40] but little is known about the neurotransmitters involved in this process. We have therefore made a study of the effects of PAG stimulation on the nociceptive responses of neurones in the reticular formation (CRF) of the rat caudal medulla [14,21,35]. In particular we wished to explore the possibility that an opioid operated descending system had a direct inhibitory effect upon these neurones as we have previously shown that their inhibition. by a variety of peripheral stimuli, is reversed by naloxone [ 18,353 and that their firing is readily depressed by microiontophoretic application of opioids [22]. The distribution of met-enkephalin in the caudal medulla has been well described f44.511, but previous studies on /3-endorphin have concentrated on more rostra1 structures [8]. There is now evidence that neurones in the hypothalamic arcuate nucleus, where /3-endorphin containing neurones are concentrated [8,46]. project to the caudal medulla [ 15,43,45,53]. We have therefore used immunocytochemical techniques to detect ~-endorphin containing fibres in the vicinity of our recording sites in CRF. The PAG is known to contain both ,&endorphin and enkephahn fibres [X,15] but it was considered worthwhile to re-examine the distribution of p-endorpmn to check for correlation with the location of our stimulation sites. The behavioural effects of PAG stimulation are complex and do not always include analgesia [29,30]. It was therefore considered essential to demonstrate that our PAG electrode pfacements would produce analgesia, before any attempt was made to examine the effects of PAG stimulation on the responses of single neurones.
251
The identification of mechanisms operated by endogenous opioids, at the present time, is dependent on the use of the opiate antagonist, naloxone [19]. It was therefore thought important to test both PAG evoked behavioural analgesia and the consequent inhibitory effects on single reticular neurones with this drug. In the single neurone studies naloxone was applied by microiontophoresis [19] to ensure that the action of the drug was restricted to the vicinity of the recorded neurone. This investigation has produced evidence that PAG stimulation, employing parameters that will sustain behavioural analgesia, produces a prolonged inhibition of caudal reticular formation neurones. This inhibition can be reversed by naloxone, applied microiontophoretically, and the distribution of fl-endorphin containing fibres within CRF is such that it is reasonable to suppose that the inhibition may be operated by synaptic release of /3-endorphin.
Methods Adult, male rats, of the MRC Porton strain, were anaesthetized with pentobarbitone (60 mg/kg) and implanted with PAG stimulating electrodes prepared from twisted, teflon coated stainless steel wire (Clark Electra-Medical SS-ST or SS-ST) with the insulation intact except at the electrode tips. The electrode pedestal was held in position by 10 BA stainless steel screws tapped into the skull and by fixation wires threaded through small bilateral burr holes, the whole implant being imbedded in dental acrylic (Simplex). Stereotaxic co-ordinates of the electrode tips were in the range of A 1.9 to P 0.5, L 0.0 to 1.0 and H -0.5 [26] and were broadly similar to those used by other workers [34,36,41,55] although it should be noted that the majority of our placements were aimed at PAG proper with very few in the vicinity of nucleus raphe dorsalis (e.g. contrast with [2]). Animals were caged individually and kept under a 12 h light, 12 h dark regime (dark 8 p.m.-8 a.m.). Testing was carried out 4-6 days postoperatively and took place between 9 a.m. and 8 p.m. The behaviour of the animals was observed in a circular open-field measuring 1.5 m in diameter. After a short period of habituation the animals showed little locomotor activity and tended to s& in one position, occasionally grooming. PAG stimulation (50 Hz, 0.5 msec duration unipolar pulses) was applied for 15-20 set at the following current steps: 25, 50, 100, 150, 200, 250 and 300 PA. The threshold and nature of any motor changes were noted. Current intensities above those causing a distinct motor arousal were only examined in a few individuaIs and in all cases were found to augment the motor behaviour. In all other animals the appearance of motor behaviour was used as the cut-off intensity for subsequent behavioural tests. Analgesia was assessed in two ways. The simplest method consisted of pinching the animal’s tail with a pair of artery forceps. The animal characteristically flicked its tail, oriented to the stimulus or tried to escape, and occasionally vocalized; all of which would take place within 2 set of the stimulus application. Analgesia was considered to have been produced by PAG stimulation if all components of the response were abolished.
The taif irnnl~rsi(~n test [XI] was used as a secctnd method of assessmp an;tlgc~i;~. In this test the animals were lightly restrained in a plastic tube and the end IO cm of tail immersed in water at 54°C. The end point was taken as a clear flick of the tail. If this did not occur within 15 set the tail was removed from the water to avoid tissue damage. The responses were used to calculate an index of analgesia [20,39]. PA<; stimulation was considered to have produced analgesia if it produced a figure abo\~ 30%. which was more than two standard deviations from the mean control response. Naloxone, for injection, was prepared as a solution of the hydrochloride salt (Endo) in 0.9% w/v sodium chloride solution. and doses are expressed as those of the salt. Some of those animals showing analgesia in response to PAG stimulation were further studied in acute. electrophysiological experiments under urethane anaesthesia ( 1.22 g/kg i.p.). Animals not showing analgesia were not used as a preGous study f39] found that in such animals there were no inhibitory effects on medulla neurones. Rats were mounted in a stereotaxic frame and prepared for recording from the caudal medulla as previously described [1X,35]. Extracellular action potentials were recorded from single neurones using the 4 M NaCl solution filled crntre barrel of a InultibarrelIed pipette with an overatl tip diameter of 4--6 pm {lx]. Outer barrels of the pipettes contained solutions of drugs for application b\; microiontophoresis selected from sodium I.-glutamate (Sigma) 0.5 M. pH 8.5, y-aminobutyric acid (GABA, Sigma) 0.5 M. pH 3.5. met-ei~kephaIil1 (WeIl~(~me), ~-end~~rphi~l ( Pel~illsL~la) I mM in 150 mM NaCl and naloxone hydrochloride (Endo) 50 mM. One barrel always contained 1 M NaCl solution and wab used for automatic balancing of the iontophoresis currents and another contained a solution of pontamine sky blue (Gurr) 2.5% w/‘v in sodium acetate buffer so that spots of dye (IO0 pm approx.) could be deposited during an experiment. This allowed the rrcomtruction of electrode tracks and the verification of recording positions from photographs of 100 pm transverse, frozen sections of the medulla. Action potentials were displayed on an oscilloscope and a trigger circuit was used to generate standard pulses from those action potentials crossing a voltage gate. These pulses were counted and displayed as an analogue rate on a multichannel chart recorder together with drug application event pulses. Noxious stimuli were applied to the hind limbs and. tail either by heating to between 45 and 55°C with a radiant heat lamp or by immersion in warm water, the temperature of the skin being monitored in both cases with a miniature thermocouple, or by pinching with haemostats. Innocuous stimuli used included tapping and gentle stroking of the skin with a camel hair brush. At the end of each electrophysiological experiment a lesion was made by passing a brief DC current through the PAG stimulating electrode. the animal was perfused through the aorta with 4% paraf~~rmaldeh~d~ solution and the brain was removed for subsequent histology. Immunohistochemical techniques used have recently been described [38,46-491 but briefly, an antiserum to P-endorphin was prepared using synthetic peptide (Serva) conjugated with thyroglobulin and injected into rabbits. Tissue fixed @-endorpbin was visualized by using antiserum at a dilution of 1 in 300 and sites of antiserum binding were located with the immunoperoxidase method. The specificity
253
of the antiserum used was tested by incubation with 50 pg/ml fi-endorphin or ACTH for 4 h prior to application to the tissue sections. Staining with the antiserum was completely abolished by pre-incubation with /3-endorphin but was unaffected when ACTH was used. Photographs of stained sections were taken and maps of P-endorphin containing structures were constructed using a camera lucida.
Results Behavioural experiments Forty-four animals were found to respond to PAG stimulation. Motor arousal, sometimes in the form of a co-ordinated running behaviour, was seen particularly with higher stimulus currents (up to 300 PA) and longer pulse trains (up to 30 set). Twenty-three of the 44 rats showed a clear analgesia to noxious thermal or mechanical stimulation of the tail or hind paws following electrical stimulation of the PAG. The analgesia usually outlasted the stimulus producing it, as described previously by others [41], but motor arousal terminated with the stimulus so that there was little difficulty in separating the phenomena in all 23 animals. In 11 rats the analgesia produced by PAG stimulation was tested after administration of either 2 or 10 mg/kg of naloxone i.p. Four of the 8 rats given 2 mg/kg naloxone showed reversal of the PAG evoked analgesia and all 3 rats given 10 mg/kg showed reversal. Lesions located in frozen sections of the brain stem showed that stimulating electrode tips in animals displaying analgesia were located within the confines of the central grey matter between A 1.6 and P 0.5, and this is illustrated in Fig. 1. Electrophysiological experiments Fourteen animals showing analgesia phase of the investigation. Micropipettes
after PAG stimulation were used in this were inserted perpendicular to the surface
PO.5 Fig. I. Transverse section of rat brain stem at intervals indicated by the stereotaxic co-ordinates beneath each section, showing stimulation sites capable of supporting behavioural analgesia in the 14 rats used in electrophysiological experiments. The symbols are placed at positions determined by lesions made through the implanted stimulating electrodes and indicate the stimulus current necessary for the production of analgesia, such that *, 25-100 PIA; 0, 100-200 pA; and A. 200-300 PA. AP zero was the intra-aural line.
of the medulla, within 1.5 mm of the midline and approximately 1 mm caudal to the obex. While the electrode was advanced. continuous microiontophoretic application of glutamate was used as a search stimulus. When a neurone was located it was tested in order to determine whether it would respond to noxious and/or innocuous peripheral stimuli. When a response to peripheral stimulation was established this was then challenged with PAG stimulation. using stimulus intensities earlier found adequate to produce analgesia in the same animal. The responses of a total of 62 neurones were examined in this way. Six of these were identified as dorsal column nucleus (DCN) neurones as they were located superficially within the medulla and responded only to innocuous stimuli applied to a restricted receptive field on the limbs. None of these DCN neurones showed any attenuation of their sensory responses by PAG stimulation. The remaining 56 neurones were all located deeper in the reticular part of the medulla, in the vicinity of nucleus reticularis ventralis (NRV) and responded in a characteristic manner, as previously described [7.18.2 1.351. No responses to tapping or brushing were seen but noxious stimuli readily evoked excitation, or in some cases inhibition, of these neurones. Receptive fields were large encompassing at least one hind limb and the tail and sometimes included the majority of the body surface. PAG stimulation directly inhibited 32 of these neurones with a further 3 neurones being inhibited after a brief excitation. thus. 62% of the CRF neurones tested were inhibited by analgesia supporting PAG stimulation and, most interestingly. the time course of the inhibition resembled that of the behavioural analgesia being prolonged beyond the stimulus and depressing the firing of the neurone for several minutes. A further 13 neurones (23%) were excited by PAG stimulation, and although no assumptions were made about the possible significance of this observation this group received no further study at this time. The remaining 8 neurones (14%) were not affected by PAG stimulation in any way. On 21 neurones with clear and reproducible excitatory responses to peripheral noxious stimuli. the ability of analgesia supporting PAG stimulation to attenuate these responses was tested. Following a series of control responses to noxious stimuli, a period of PAG stimulation was given and immediately followed by a noxious stimulus, which was then repeated at regular intervals until the neuronal response had regained its control size. Fourteen (66%) of the neurones studied in this way displayed either complete abolition of the neuronal nociceptive response or a reduction in its size. The reduction in response size characteristically outlasted the stimulus by several minutes and had a similar time course to the effects of the same PAG stimulus on the animal’s nociceptive behaviour. Microiontophoretic application of GABA or met-enkephalin readily depressed the firing of CRF neurones and attenuated their nociceptive responses as previously described [ 18,221. The actions of ,f%endorphin. applied with currents in the range 10-60 nA, were examined on 13 neurones, of which 8 showed a depression of firing rate, 2 showed a brief excitation followed by depression. 1 was excited and 2 were unaffected. Met-enkephalin (20-100 nA) depressed 16 of 17 neurones tested. Naloxone (2-40 nA) reversed the depression produced by P-endorphin on 2 neurones and that produced by met-enkephalin in a further 3 neurones but had little effect on that
255 I NALOX 2
L
.
D
[3
CJ .
I
.
c3
cl
P .
cl
I
i
D B-END I
*
20
GABA 10
Fig. 2. An analogue rate meter record of the firing of an NRV neurone, maintained by the continuous microiontophoretic application of glutamate 20 nA. GABA 10 nA (solid bars) and P-endorphin 20 nA (open bars) were both found to depress the firing of this neurone, although during the concurrent application of naloxone 2 nA (long open bar) the response to /3-endorphin was abolished. Partial recovery was seen 10 min after the end of the naloxone application. Time calibration is 2 min.
produced by GABA as illustrated in Fig. 2. The direct inhibitory effects of PAG stimulation on the spontaneous or glutamate maintained firing of CRF neurones were tested in the presence of naloxone on 13 neurones and on 9 of these the inhibitory effects were reduced. On a further 3 neurones naloxone was tested against the PAG induced depression of nociceptive responses and reduced the inhibition in 2 cases. The remaining result was equivocal as the control nociceptive response was enhanced by naloxone. In Fig. 3 an experiment on a CRF neurone excited by a continuous application of glutamate is shown. PAG stimulation depressed the firing of this neurone and the inhibitory effect was prolonged. After several minutes iontophoretic application of
I
I
I
I
Naloxone
20na I
PAG stim
Fig. 3. An analogue rate meter record of the firing of an NRV neurone maintained by the continuous application of glutamate 20 nA. Electrical stimulation applied to the PAG (indicated by solid bars) produced a prolonged depression of firing and after several minutes microiontophoretic application of naloxone 20 nA this inhibition was blocked. Recovery of the depressant effect of PAG stimulation was seen when the naloxone application was discontinued. The rat used in this study had previously shown analgesia to noxious tail pinch following PAG stimulation. This analgesia was reversed by 10 mg/kg of naloxone.
Tail pinch . . LO 7
.
.
.
.
.
.
Naloxone
.
.
.
.
.
I
.
.
.
.
I
Hz
PAG “stir-n
Fig. 4. An analogue rate meter record of the firing of an NRV neurone induced by no~tous pinching of the tail (solid bars). Electrical stimulation of the PAG (hatched bar below the record) initially abolished the excitatory response to noxious pinch and produced attenuation outlasting the stimulus by several minutes. Iontophoretic application of naloxone 6 nA did not prevent the initial abolition of the response but considerably accelerated its recovery.
naloxone, the inhibitory effect of PAG stimulation was lost. and partial recovery of the inhibitory response was seen after termination of the naloxone application. In Fig. 4 an experiment of the second type is illustrated. Excitatory responses of the neurone were evoked by periodic noxious pinching of the tail and PAG stimulation was seen to produce an attenuation of this nociceptive response. When naloxone was applied iontophoretically it had no effect on the size of the control
Ftg. 5. A transverse section of the caudal medulla of the rat at a level corresponding to I mm caudal to the obex. Plotted on this section are the locations of I I CRF neurones (open triangles) which showed naloxone reversible inhibition following electrical stimulation of the PAG, and the location of fibres and terminals showing /I-endorphin immunoreactivity (dots). ntV. caudal trigeminal nucleus: gr. gracile nucleus; nts, nucleus of the solitary tract: nrv, nucleus reticularis ventralis: ap. area postrema: p, pyramidal tract
251
responses to noxious stimulation but did reduce the effectiveness of PAG stimulation. The most striking effect was on the time course of recovery of the nociceptive response as even during the application of naloxone, the first noxious stimulus after PAG stimulation was without effect. Immunohistochemical localization of /3-endorphin and reconstruction of recording and stimulation sites The location of P-endorphin containing fibres was plotted onto diagrammatic transverse sections of the rat brain at appropriate AP co-ordinates matching recording sites in the CRF and stimulation sites within the PAG. In Fig. 5 the distribution of P-endorphin in the caudal medulla is shown together with the
Fig. 6. Photomontage showing location of branching /3-endorphin and b correspond to the regions thus identified in Fig. 5.
containing
fibres within CRF. Panels a
Fig. 7. A transverse section of rat brain stem at A 1.6 showing the location (open diamonds) of the 5 PAG stimulation sites at this AP plane which produced analgesia with currents < 100 PA and the distribution of /3-endorphin immunoreactivity (dots). Note the correspondence of /3-endorphin fibres and stimulation sites. A further descending tract of &endorphin fibres runs ventrolateral to PAG. PAG. periaqueductal grey; LM, medial lemniscus; SNC, substantia nigra pars compacta.
locations of those 1 I neurones found to exhibit naloxone reversible inhibition following PAG stimulation. P-Endorphin fibres can be seen to be in a patchy distribution in most parts of the CRF but do not spread laterally into the caudal trigeminal nucleus to any significant extent. Although only scattered individual fibres were found, and no dense beds of terminals, these individual fibres seemed to give off a large number of short collateral or terminal branches and this is illustrated in Fig. 6. In Fig. 7 the distribution of ~-endorp~n fibres within PAG is shown together with the location of 5 stimuIation sites from animals with stimulus currents for analgesia that were clearly lower than the currents needed to evoke motor activity. It was thought that these experiments would provide reliable location of the actual structures stimulated as the effects produced would have the smallest chance of being attributable to stimulus spread. All of these stimulus sites were clearly within the PAG and overlapped with the main group of ,&endorphin containing fibres passing through this area.
In this study we have obtained evidence that a proportion of the reticular formation neurones within the caudal medulla of the rat, which are exclusively responsive to noxious inputs, are strongly inhibited by trains of electrical stimuli applied within the PAG. Furthermore, the stimuli producing these inhibitory effects
259
are sufficient to induce analgesia in the same animals, exhibiting a similar time course to the neuronal inhibitions. In view of the importance of the reticular formation as a relay in the ascending nociceptive system [9,10,12] it is tempting to relate these behavioural and neuronal responses to PAG stimulation, although it is noteworthy that reticular neurones have complex projections both descending [23] and ascending [9] the neuroaxis and are likely to be polyfunctional. The inhibitory effects of PAG stimulation were clearly seen on neurones whose firing was sustained by microiontophoretic application of the excitatory amino acid, glutamate, suggesting that the inhibitory mechanism was, at least in part, postsynaptic. The confirmation of this by intracellular recording would thus be of great interest particularly in view of the prolonged time course of these inhibitions. The reversal of a proportion of the inhibitory effects of PAG stimulation by microiontophoretic naloxone raises the possibility of an endogenous opioid substance acting as a neurotransmitter, and this has been inferred from the action of naloxone in reversing behavioural analgesia evoked by PAG stimulation [ 1,4,19,20]. Application of naloxone by microiontophoresis restricts the action of the drug to the immediate vicinity of the micropipette. Thus, any endogenous opioid acting to inhibit CRF neurones must be released locally. It has been suggested that naloxone may antagonize the actions of GABA in addition to those of opioids [19] and this observation is likely to be important in explaining the convulsant effects seen with toxic doses of the drug. However, we have recently described a short lasting inhibition of NRV neurones, evoked by single stimuli applied to PAG, which can be blocked by the GABA antagonist bicuculline but not by naloxone [37, see also 321 and which is quite distinct from the prolonged inhibition, evoked by trains of PAG stimuli, that we describe here. Many hypothalamic fibres project to the PAG and to the caudal medulla [15,43,47,48] and some of these contain oxytocin, vasopressin and related peptides [48]. We have recently shown that oxytocin will depress the firing of CRF neurones 1381 and thus this substance remains a transmitter candidate for those descending inhibitions which are resistant to both bicuculline and naloxone. Descending fibres also passing close to PAG contain catecholamines [52] and Shydroxytryptamine [2,50] and these too are worthy of consideration as transmitter candidates in this system. It is perhaps relevant that depletion of Shydroxytryptamine with parachlorophenylalanine will reduce the analgesia produced by stimulation of ventral PAG sites, close to the dorsal raphe, but does not reduce the analgesia evoked from more dorsal sites, such as those used in the present study [2j. Enkephalin containing fibres are found on the dorsal and lateral borders of NRV [44,51] but it is difficult to explain how they might be stimulated by pathways activated by electrodes within the PAG. There is evidence for an excitatory connection from PAG to nucleus raphe magnus and neurones in this region have descending enkephalin containing fibres [23]. However, these do not seem to terminate in the CRF as we have shown that horseradish peroxidase injected to the CRF does not undergo retrograde transport to raphe magnus although transport is seen direct to PAG [unpublished observations and 61. Our immunohistochemical studies suggest that P-endorphin may be a better candidate for the transmitter of naloxone sensitive
inhibition in the (‘RF as demonstration of the presence of /I-endorphrn 1’1hres within this region has been made [thla paper and i5] and injection of horseradish peroxidase to the (‘RF results in the retrograde labelling of neuronal cell bodies in the arcuate nucleus [47]. believed to be the origin of most P-endorphin containing axons in the central nervous hystem [X,15.46]. The stimulus sites which ue ~‘OLIIK! to have the lowest current threshold for producing beha\,ioural analgesia Lvere clearly located over the /!I-endorphin containing fibres passing through PAG at the same AP co-ordinates and it is tempting to attribute the naloxone sensiti\,c a&g&c and neuronal inhibitorv effects of PACT stimulation to activation of these fibres. ‘This is also consistent with the observations that /%endorphin can be released into the cerebrospinal fluid by analgesia producing pericentricular or PAG .stimulation [24] and that microinjection of ,&endorphin into midbrain reticular formation produces profound analgesia [27]. It seems unlikely that a /3-endorphin operated inhibition of caudal reticular neurones constitutes an adequate neuronal substrate for analgesia produced by PAG stimulation. However. when this is combined with a similar inhibition of more rostra] reticular neurones [36] and the likelihood that significant amount of /I-endorphin would be released in the vicinity of PAG and periventricular thalamus neurones thought to have a role in nociception [9,10,12,29] then a powerful net analgesic effect may well result. It may be possible to activate such an opioid operated descending inhibitory system by intense peripheral stimuli. as has been suggested by Besson and his colleagues [28]. for we have found that the inhibition of caudal reticular neurones by noxious thermal or mechanical stimuli [1X.35] or by vaginal distention [1X.20] is also naloxone reversible.
Acknowledgements
We are grateful to Endo for the gift of naloxone. Skilled technical assistance was provided by C. Allen, Catherine Crozier and P. Darby. This work is supported by grants from the Wellcome Trust, Medical Research Council and Royal Society.
References I Adams, J.E.. Naloxone reversal of analgesia produced hv brain stimulation in the human. PaIn. 2 (1976) 161-166. 2 Akll. H. and Mayer. D.J.. Antagonism of stimulation-produced analgesia hy p-CPA. a \erotonin bynthesis inhibitor. Brain Res.. 44 (1972) 692-697. 3 Akil. H. and Liebeskind. J.C.. Monoaminergic mechanism5 of stimulation-produced analge%l, Brain Res.. 94 (1975) 279-296. 4 Akil. H.. Mayer. D.J. and Liebeskind. J.C.. Antagonism of stimulation-produced analgesia bv naloxone, a narcotic antagonist, Science. I91 (1976) 96 I --962. material &bated in 5 Akil, H., Richardson. D.E., Hughes. J. and Barchas. J.A.. Enkephalin-like ventricular 46X-465.
cerebrospinal
fluid of pain patients
after analgetic
focal stimulation.
Science.
201 (1978)
261
6 Benz, A.J., Ascending
and descending
Abstr.. 5 (1979) Abstr. 2053. 7 Benjamin, R.M., Single neurones
projections
in the rat medulla
to the periaqueductal responsive
gray of the rat, Neurosci.
to nociceptive
stimulation.
Brain Res.,
24 (1970) 525-529. 8 Bloom, F.. Battenberg, E., Rossier, J., Ling, N. and Guillemin, R.. Neurones containing /3-endorphin in rat brain exist separately from those containing enkephalin: immunocytochemical studies. Proc. nat. Acad. Sci. (Wash.). 75 (1978) 1591-1595. 9 Bowsher, D., Collateralization of reticulothalamic
projections
and its possible
significance
for pain
sensation, Adv. Pain Res. Ther., I (1976) 245-249. 10 Bowsher, D., Role of the reticular formation in responses to noxious stimulation, pain, 2 (1976) 361-378. I1 Carstens, E., Klumpp. D. and Zimmerman, M., The opiate antagonist, naloxone. does not affect descending inhibition from midbrain of nociceptive spinal neuronai discharges in the cat. Neurosci. Lett., 11 (1979) 323-327. 12 Casey, K.L., Reticular formation and pain: toward a unifying concept, Res. Publ. Ass. nerv. ment. Dis., 58 (1980) 93-105. 13 Duggan. A.W. and Griersmith, B.T., Inhibition of the spinal transmission of nociceptive information by supraspinal stimulation in the cat, Pain, 6 (1976) 149- 161. 14 Fields, H.L., Basbaum, A.I., Clanton, C.H. and Anderson, SD.. Nucleus raphe magnus inhibition of spinal cord dorsal horn neurons, Brain Res., 126 (1977) 441-453. 15 Finley, J.C.W., Lindstrom. P. and Petrusz. P. Immunocytochemicat localization of @-endorphin containing neurons in the rat brain, Neur~nd~rinoiogy, 33 (1981) 28-42. 16 Hayes. R.L., Newlon, P.G.. Rosecrans, J.A. and Mayer, D.J., Reduction of stimulation-produced analgesia by lysergic acid diethylamide. a depressor of serotonergic activity, Brain Res., 122 (1977) 367-372. 17 Hayes, R.L.. Price, D.D., Ruda, M. and Dubner, R., Suppression of nociceptive responses in the primate by electrical stimulation of the brain or morphine administration: behavioural and electrophysiological comparisons, Brain Res., 167 (1979) 418-421. 18 Hill, R.G., Opioid peptides as inhibitory neurotransmitters acting on sensory neurones in the rat medulla. In: C. Ajmone Marsan and W.Z. Traczyk (Eds.), Neuropeptides and Neural Transmission, Raven Press, New York, 1980, pp. 199-208. 19 Hill, R.G., The status of naloxone in the identification of pain control mechanisms operated by endogenous opioids, Neurosci. Lett., 21 (1981) 217-222. 20 Hill, R.G. and Ayfiffe, S.J.. The antin~i~ptive effect of vaginal stimulation in the rat is reduced by naloxone. Pharmacol. Biochem. Behav.. 14 (1981) 631-632. 21 Hill, R.G. and Morris, R., Responses of nucleus reticularis ventralis neurones of the rat to noxious stimuli and to electrical stimulation of the periaqueductal grey and raphe dorsalis, J. Physiol. (Lond.), 301 (1980) 3%39P. 22 Hill, R.G. and Morris, R., Responses of neurones in medial caudal medulla to iontophoretically applied opioids, Brit. J. Pharmacol., 70 (1980) 147P. 23 Hokfelt, T., Terenius, L., Kuypers, H.G.J.M. and Dann, O., Evidence for enkephalin immunoreactive neurons in the medulla oblongata projecting to the spinal cord, Neurosci. Lett., 14 (1979) 55-60. 24 Hosobuchi, Y.. Rossier. J., Bloom, F.E. and Guillemin. R., Stimulation of human periaqueductal gray for pain rehef increases immunoreactive /3-endorphin in ventricular fluid, Science, 203 (1979) 279-281. 25 Jurna. I.. Effect of stimulation in the periaqueducta~ grey matter on activity in ascending axons of the rat spinal cord: selective inhibition of activity evoked by afferent AS ad C fibre stimulation and failure of naloxone to reduce inhibition, Brain Res., 196 (1980) 33-42. 26 Konig. J.F.R. and Klippel, R.A., The Rat Brain. a Stereotaxic Atlas of the Forebrain and Lower Parts of the Brain Stem, Williams and Wilkins, Baltimore, Md., 1962. 27 Kuraishi. Y., Satoh. M., Harada, Y.. Akaike, A.. Shibata, T. and Takagi, H., Analgesic action of intrathecal and intracerebral P-endorphin in rats: a comparison with morphine, Europ. J. Pharmacol.. 67 (1980) 43-146. 28 Le Bars. D., Chitour, D., Kraus, diffuse noxious inhibitory controls
E.. Dickenson, A.H. and Besson, J.M.. Effect of naloxone (DNIC) in the rat, Brain Res.. 204 (1981),387-402.
upon
29 Ltebeskind. J.C. and Mayer, DJ Sornato~enmr~ txorhd rcqxmm in (he mcsencrph~ltc ccn~r;il gr;t\ matter ~>fthe rat. ISraIn l&b.. 27 (lY71) 133 151. 30 Liebman. J.M.. Majer. D.J. and Lleheskind. J.C.. Mesencrphalic central gray lesions and fear-rn
J.H., Inhibitory effects of nucleus raphe magnus nucleus to nocicepti\r compared with non-nocicepti\c
on neuronal tnputs. Pain.
32 Lovick. T.A. and Wolstencroft. J.H.. Inhibitton from nucleus raphe magnua of tooth pulp responses in medial reticular neurnnes of the cat can be antagonized h! hicuculline. Neurosci. Lett.. 19 (1980) 3255330. 33 Mayer. D.J. and Liebeskind. J.C.. Pam reduction by focal electrical stimulatton of the hrain: dn anatomical and behavioural analysis. Brain Res.. 68 (1974) 73-93. 34 Mayet-. D.J. and Price. D.D.. Central nervous system mechanisms of analgesia. Pain, 2 (1976) 379-404. 35 Mayer. M.L. and Hill, R.G.. The effects of intravenous fentanyl. morphine and naloxone on nociceptive responses of neurones m the rat caudal medulla. Neuropharmacology, 17 (1978) 533-539. 36 Mohrlands. J.S. and Gebhart. G.F.. Effects of focal electrical stimulation and morphine mtcroinjecttons m the periaqueductal gray of the rat mesencephalon on neuronal activity in the medullary reticular formation. Brain Res.. 201 (1980) 23-37. 37 Morris. R. and Hill, R.G.. Inhibitory mechanisms in the nucleus reticular& ventralis (NRV) of the rat. Neurosci. Lett.. Suppl. 5 (1980) S389. 38 Morris. R.. Salt. T.E.. Sofroniew, M.V. and Hill, R.G.. Actions of microiontophoretically applied nxytocin and immunohistochemical localization of oxytocin. vasopressin and neurophyain in the rat caudal medulla, Neurosci. Lett.. IX (1980) 163- 168. 39 Morris. R., Hill, R.G.. Cahuaac. P.M.B. and Salt, T.E.. The involvement of endogenous opioids in the descending inhibition of caudal trigeminal and reticular nociceptive neurones evoked by analgesia producing periaqueductal grey atnnulation m the rat. In: B. Mattheus and R.G. Hill (Eda.). Anatomical, Physiological and Pharmacological Aspects of Trigeminal Pam. Excerpta Medica. Amsterdam. 1982, pp. 195-206. 40 Morrou. T.J. and Casey, K.L.. Analgesia produced by mesencephalic stimulation: effect on hulhoreticular neurons. Adv. Pain Rea. Ther.. 1 (1976) 50% 5 10. 41 Reynolds. D.Y.. Surgery in the rat during electrical analgesia induced by focal brain stimulation. Science, 164 (1969) 444-445. 42 Richardson, D.E.. Central gray stimulation for control of cancer pain. Adv. Pain Res. Ther.. 2 (1979) 487 -492. 43 Saprr. C.B., Loewy, A.D.. Swanson, L.W. and Cowan. W.M.. Direct hypothalamus-autonomic connections, Brain Rea.. 117 (1976) 3055312. 44 Sar. M.. Stumpf. W.E., Miller. R.J.. Chang. R.-J. and Cuatrecasas. P.. Immunohistochemical localization of enkephalin m rat brain and spinal cord. J. camp. Neurol.. I83 (1978) 17-38. 45 Sherman. J.E. and Liebeskind. J.C.. An endorphinergic. centrifugal substrate of pam modulation: Res. Publ. Ass. nerb. ment. Dis.. 58 (1980) recent findings. current concepts and complexities. 191.-204. 46 Sofroniew. M.V.. Immunoreactive P-endorphin and ACTH in the same neurons of the hypothalamic arcuate nucleus in the rat. Amer. J. Anal., 154 (1979) 283.-289. M.V. and Schrell. U.. Hypothalamic neurons proJectmg to the rat caudal medulla 47 Sofroniew, oblongata, examined by immunoperoxidase staining of retrogradely transported horseradish peroxidase. Neurosci. Lett.. I9 (1980) 2577263. 48 Sofroniew, M.M. and Schrell. U.. Evidence for a direct proJection from oqtocm and vasopressm neurons in the hypothalamic paraventricular nucleus to the medulla oblongata: tmmunohistochemical visualization of both the horseradish peroxidase transported and the peptide produced by the same neurons, Neurosci. Lett., 22 (1981) 211-217. 49 Sofroniew, M.V. and Schrell, C.. Long term storage and regular repeated use of diluted antisera in glass staining jars for increased sensitivity, reproducibility and convenience of single- and tw-color light microscopic immunocytochemistry. J. Histochem. Cytochem.. 30 (I 982) 504-5 11.
263
50 Steinbusch, H.W.M., Distribution of serotonin-immunoreactivity in the central neurone system of the rat cell bodies and terminals, Neuroscience, 6 (1981) 557-618. 51 Uhl, G.R., Goodman, R.R., Kuhar, M.J., Childers, S.R. and Snyder, S.H., Immunohistochemical mapping of enkephalin-containing cell bodies, fibres and nerve terminals in the brain stem of the rat, Brain Res., 166 (1979) 75-94. 52 Watson, S.J., Akil, H. and Barchas, J.D., A possible role for the dorsal periventricular catecholamine bundle in stimulation produced analgesia: a behavioural and histochemical study, Brain Res., 130 (1977) 335-342. 53 Watson, S.J., Richard, V.W. and Barchas, J.D., Adreno-corticotrophin in rat brain, Immunocytochemical localization in cells and axons, Science, 200 (1978) 1180-1182. 54 Yaksh, T.L., Yeung, J.C. and Rudy, T.A., An inability to antagonize with naloxone the elevated nociceptive thresholds resulting from electrical stimulation of the mesencephalic central gray, Life Sci., 18 (1976) 1193-1198. 55 Yeung, J.C.. Yaksh, T.L. and Rudy, T.A., Concurrent mapping of brain sites for sensitivity to the direct application of morphine and focal electrical stimulation in the production of antinociception in the rat, Pain, 4 (1977) 23-40.