0306-4522/91 $3.00 + 0.00 Pergamon Press pie fJ 1991 IBRO
h%?w~scii?nce voi. 45, No. 1, pp. 73-80, 1991 Printed in Great Britain
REPEATED ELECTROCONVULSIVE SHOCK INCREASES TACHYKMIN AND CHOLECYSTOKININ mRNA EXPRESSION IN VENTRAL PERIAQUEDUCTAL GRAY N.
Lrrmnnons,*t
K. BRODIN,$ C.-O. STILLER,* H. PEBSSON§ and E. BRODIN*
*Department of Pharmacology,
Karol&&a Institutet, Box 60400, S-IO401 Stockholm, Sweden IDepartment of Clinicai Pharmacology, Karolinska Hospital, Stockhohn, Sweden §Department of Medical Chemistry, Laboratory of Molecular Neurobiology, Karohnska Institutet, Stockholm, Sweden
Akatraet-The effect of repeated electroconvulsive shock (five shocks during 10 days) on preprocholecystokinin and preprotachykinin-A messenger RNA expression was studied in the mesencephalic periaqueductal gray and adjacent areas of rat using in situ hybridization histochemistry with specific oligonucleoti& probes. An increased number of p~pr~hol~sto~~n and p~~~~y~~n-A messenger RNA hybridization positive neurons (+30% and +47%, respectively) in the Edinger-Westphal nucleus was observed following repented electroconvulsive shock. In addition, both preprocholecystokinin and preprotachykinin-A messenger RNA expression, measured as grain density over single neurons, was si~ifi~n~y increased (+37% and +45%, respectively). The results indicate that cholccystoktmn- and substance Fcomarmng neurons m the lzdmger-Westphal nucleus are activated by repeated electroconvulsive shock, which may be related to the antidepressant and analgesic effects of electroconvulsive shock treatment.
El~tr~on~sive therapy (ECT) is widely used as treatment of severe affective disordersi and ECT has also been found to have an analgesic effect in patients suffering from chronic pain, especially in cases associated with dep~ssion.24,3’,4~~ The mechanisms of action of ECT in treatment of depression and pain are, however, still essentially unknown, even though involvement of several classical transmitter systems and peptidergic systems”‘~ has been implicated from animal experiments with electroconvulsive shock (ECS). Recently, it has been reported that repeated ECS increases tissue levels of substance P (SP) and cholecystokinin (CCK) in certain regions of rat cerebral cortex,6 an effect similar to that observed after subchronic treatment with antidepressant (monoamine up~ke-inhibiting) drugs.5” Treatment of rats with selective serotonin uptake inhibitors also increases the tissue level of SP in the mesencephalic periaqueductal gray (PAG),5 a region known to tTo whom correspondence should be addressed. Abbreviations: CCK. cholecvstokinin: cDNA, complementary deoxyribonucleic acid; dATP, deoxyadenosine ~phosp~te; DlT, di~iothr~to~ ECS, electrocouvulsive shock; ECT, electroconvulsive therapy; EWn, &linger-Westphal nucleus; mRNA, messenger ribonucleic acid; PAG, mesencephalic periaqueductal gray; PBS, phosphate-buffered saline; PPT-A, prepro~chyki~n-A; RIA, r~ioi~unoas~y; SSC, sahne sodium citrate buffer; SP, substance P; tRNA, transfer ribonucleic acid. 73
be involved in, for example, endogenous pain suppression.2*32 Immunohistochemical studies have shown that SPand CCK-like immunoreactivities are present in both terminals and cell bodies in rat and cat ventral PAG’s*‘6*M~37 and coexist in PAG neurons.37 Cells in ventral PAG, i.e. in the Edinger-Westphal nucleus (EWn), respond to noxious stimuli with an increased firing rateI and SP- and CCK-immunoreactive cells in this region have been found to project to the spinal cord and have been suggested to represent a descending system which could influence sensory transmission at the spinal leve1.30*38Recently, acute treatment with morphine, which is believed to exert part of its analgesic effect via an action on the PAG, was found to induce increased tissue levels of SP and CCK in this brain region3’ The recent development of the in situ hybridization technique4$ makes it possible to study expression of neuropeptide genes in defined neuron populations in the brain. Rat preprotachykinin A (PPT-A, a SP precursor) messenger RNA19 (mRNA) and rat CCK mRNAa have been sequenced and ~ornpl~en~~ DNA (cDNA) clones isolated. Hybridization probes from these cDNA clones as well as synthetic oligonucleotide probes, including those used in the present study, have with good agreement been used to document the expression of PPT-A mRNA3~20~22~45 and CCK mRNAi3~22~M*35~4’~43 in rat brain.
Fig. 1. Autoradiograms of PPT-A mRNA in situ hybridization histochemistry in rat mesencephalon. The upper panel shows an autoradiogram from a control animal and the lower panel shows an autoradiogram from an ECS-treated animal. The tissue was hybridized to an oligonucleotide PPT-A mRNA probe exposed to a photographic film. The hybridization signal considered to belong to the EWn was located 0.7 mm ventral to the ventral limit of the aaueduct at the coronal level shown in the figure. Arrows indicate cells shown in Fig. 3. CO, control
In the present study the effect of repeated ECS on expression of PPT-A and preproCCK mRNA was assessed in PAG and adjacent areas by in situ hybridization histochemistry combined with computerized image analysis.
EXPERIMENTAL PROCEDURES
Animals and tissue preparation
Male Sprague-Dawley rats (ALAB Laboratorietjanst. Sollentuna, Sweden; body weight 150-2oOg) were used. For ECS, the animals were anaesthetized with halothane-nitrous oxide-oxygen and stimulation was applied through ear-clip electrodes (1 s, 150 mA, 50
Sachs cycles, x s sinusoidal) using rodent shocker Elektronik Frieburg, F.R.G.). stimulations gave to generalized lasting for 5 s. group of animals received shocks over days (days and 10) five control were treated the same except that current was The rats killed by 24 h the last The brains immediately dissected frozen on ice and at -80°C. sections (16pm) the level EWn were on a (Leitz, Wetzlar, Sections of and control were paired thaw-mounted on (50 pg/ml)-coated In situ The tissues buffered saline
fixed in for 30
formalin in rinsed twice
4 min
ECS increases SP and CCK mRNA in PAG
75
Fig. 2. Autoradiograms of CCK mRNA in situ hybridization his&hem&y in rat mesencephalon. The upper panel shows an autoradiogram from a control animal and the lower panel shows an autoradiogram from an ECS-treated animal. The tissue was hybridized to an oligonucleotide CCK mRNA probe exposed to a photographic Iilm. The hybridization signal considered to belong to the EWn was located 0.7 mm ventral to the ventral limit of the aqueduct at the coronal level shown in the figure. Arrows indicate cells shown in Fig. 4. CO, control.
PBS, and delipidated in graded series of ethanol, including a 5-min incubation with chloroform. The sections were then air-dried. The hybridization cocktail contained 50% formamide, 4 x saline sodium citrate buffer (SSC) (1 x SSC is 0.15 M NaCl. 0.015 M sodium citrate. DH 7.0). 1 x Denhardt’s solution, 1% Sarcosyl, 0.02M‘ Na,Pd; (pH 7.0), 10% dextransulphate, 0.5mg/ml yeast transfer RNA (tRNA). 0.06 M dithiothreitol IDD17 and 0.1 ma/ml sheared salmon sperm DNA. A 48-mer ‘oligonucle&de complementary to rat PPT-A mRNA coding for amino acids 49-64, and thus complementary to a, /I and y -PPT-A mRNA,i9 was used to detect PPT-A mRNA expression. The sequence of the PPT-A-specific oligonucleotide was 5’GAACTGCTGAGGCTTGGGTCTTCGGGCGA’I-ICTCTGAAGAAGATGCTC. A 44-mer oligonucleotide complementary to rat preproCCK mRNA coding for amino
acids 89-103 was used to detect preproCCK mRNA8 (in the following text refered to as CCK mRNA). The sequence of the prep;oCCK-specific oligonucleotide- was S’AAATCCATCCCAGCCCATGTAGTCCCGGTCACTIATCCTGTGGCT. The oligonucleotide was labelled at the 3’-end with a[Srs]dATP using terminal deoxyribonucleotidyl transferase (International Biotech Inc., New Haven, Cl) to a specific activity of approximately 8 x 108 c.p.m./mg. The labelled probes were purified on a Nensorb column (DuPont, Wilmington, DE) prior to use. Following hybridization for 16 h at 42”C, the sections were rinsed five times for 15 min each in 1 x SSC at 55°C. Finally, the sections were rinsed in autoclaved water for 5 min, and then dehydrated through a series of graded alcohol and air-dried. The sections were exposed to X-ray lilm (Amersham jI-max) for four to eight days and to photographic emulsion (Kodak
76
N. LINDEFOKS et al.
Fig. 3. High magnification bright-field emulsion photomicrographs of PPT-A mRNA-positive neurons 111 ventromedial PAG from a representative control brain (left panel) and the ECS brain (right panel) from the same slides also shown in Fig.
In situ hybridization with cDNA probes for PPT-A mRNA (a generous gift from Dr Shigetada Nakanishi, Kvoto. Janan’*) and CCK mRNA (a generous gift from Dr Jack dixon,‘West Lafayette, U.6A.p in adjacent tissue sections of rat midbrain have, in separate control experiments, shown a similar hybridization pattern to those found with the PPT-A and CCK oligonucleotide probes, respectively. For comparison of distribution of the hybridization signal, we also performed hybridization with a 48-mer oligonucleotide probe specific for the unrelated rat tyrosine hydroxylase mRNA.**
Neuron counting and computerized image analysis The numerical density of hybridization-positive neurons was analysed using a Lcitz Ortholux microscope at x 313
magnification. Hybridization-positive neurons were defined as those expressing more than 10 grains per neuron (well above background density of grains). For each individual brain all neurons in EWn with significant levels of PPT-A mRNA and CCK mRNA, respectively, were counted in three sections and a mean value per section was estimated. The number of grains per neuron and the size of PPT-A mRNA- and CCK mRNA-positive neurons was calculated using a Crystal image analysis processor (Quantel Ltd, U.K.) at x 500 magnification. For this purpose, the microscope was equipped with a Panasonic WV 1500 TV camera to provide a video signal for the image analysis. Grain density was analysed over a total of 10 neurons in each animal and in five animals of each of the two groups. Over some of the neurons analysed the high density of grains made a direct numerical count futile since the grains were fused in some cases. However, the grains never covered a major part of hybridization-positive neurons. Instead, the
ECS increases SP and CCK mRNA in PAG
Fig. 4. High magnification bright-field emulsion photomicrographs of CCK mRNA-positive neurons in ventromedial PAG from a representative control brain (left panel) and the ECS brain (right panel) from the same slides also shown in Fig. 2. Scale bars = 50 pm. CO, control.
total area of grains, including all fused grains was divided with a standard grain area obtained from measurements of unfused grams over neurons. The area of a standard grain was arbitrarily defbred as 0.3 pm.Z The used method of grain counting (i.e. gram area counting) underestimates changes in grain volume since only a two-dimensional aspect of the grains is considered. However, this conservative aspect of the method implies that sign&ant changes in grain area measured are reliable, although in a semiquantitative manner. Evaluating the results one must bear in mind that the real change in mRNA content in single cells is most probably higher than the measured increase in grain area.
REsJLTs
In situ hybridization of coronal tissue sections at the level of mesencephalon showed PPT-A mRNA-
expressing neurons in the colliculus superior, interpeduncular nucleus, peritectal area and most parts of the PAG with a high level in the EWn (Fig. 1). Scattered hybridization-positive neurons were also found in the neocortex. CCK mRNA-expressing neurons were seen in layers II and III, and V and VI of the neocortex, in the medial geniculate nucleus, ventral tegmental area, substantia nigra, interpeduncular nucleus and intensely labelled cells were observed in the EWn (Fig. 2). In control animals, the number of CCK mRNA-positive neurons in the EWn was nearly two times higher than the number of PPT-A mRNA-positive neurons (Figs 3, 4). For comparison, hybridization with a 48-mer oligonucleotide probe specific for rat tyrosine
78
Fig. 5. Autoradiogram of a coronal section of rat midbrain including the EWn, substantia nigra and ventral tegmental area, labelled by in situ hybridization with an oligonucleotide specific for tyrosine hydroxylase mRNA. Note the labelling of tyrosine hydroxylase mRNA in substantia nigra and ventral tegmental area and the lack of signal in the EWn (indicated in the midline by the arrow).
Table 1. Number of preprotachykinin-A messenger RNAand cholecystokinin messenger RNA-positive neurons in the Edinger-Westphal nucleus of control rats and rats treated with repeated electroconvulsive shock PPT-A mRNA CCK mRNA
Control
ECS
31.4 + 2.8 57.1 + 4.9
46.2 + 1.6’ 74.3 + 3.2’
‘Significant difference to control group (P < 0.05; Student’s two-tailed r-test for paired samples). Values indicate the number of hybridization-positive neurons per tissue section (mean &-S.E.M.; n = 5).
hydroxylase mRNA was performed which showed a lack of hybridization signal in PAG but a strong hybridization signal in neurons in the substantia nigra compacta and ventral tegmental area (Fig. 5). Repeated ECS increased the number of neurons expressing significant levels of PPT-A mRNA (+47%; P < 0.05;Student’s t-test for paired samples) as well as the number of grains (+45%; P < 0.05)over individual neurons in the EWn (Figs 1, 3, Tables 1, 2). A similar increase was seen in the number of CCK mRNA-expressing neurons (+30%; P < 0.05)and in the number of grains (+37%; P < 0.05)over individual CCK mRNApositive neurons (Figs 2, 4, Tables 1, 2). DISCUSSION
The present results demonstrate that repeated ECS increases the PPT-A and CCK mRNA expression in a group of neurons in the EWn in the ventromedial aspect of the rat mesencephalic PAG. PPT-A and CCK mRNA-positive neurons have previously been demonstrated in EWn using oligonucleotide probes and a similar hybridization protocol as in the present
study. 20~35,43 These cells are most probably identical to the periaqueductal SP/CCK neurons previously demonstrated to project to the spinal cord using immunohistochemical and retrograde tracing techniques. y),37,38The role of the SPjCCK neurons in the EWn is at present unclear, but, as previously suggested, these neurons may represent a descending system which influences pain transmission.30x3* Electrophysiological recordings from the EWn have revealed a group of neurons that responds to noxious stimuli like toe-pinch with an increased firing rate, an effect that may be blocked by systemic morphine in a naloxone-reversible manner.16 Although the role of SP and CCK in pain perception seems complex, both SP and CCK have been reported to have an antinociceptive effect upon peripheral administration,39.46 intrathecal injection9,‘7,27 as well as when injected intracerebroventricularly or directly in the PAG.‘.‘7,23,‘9 From the present results, it may be suggested that descending periaqueductal SPjCCK neurons may take part in the mediation of ECS-induced antinociception demonstrated in animals,14,” and possibly also in the pain-relieving effect of ECT in man (see Introduction). Interestingly, electrical stimulation in PAG for the treatment of chronic pain Table 2. Grain density over hybridization-positive neurons expressing preprotachykinin-A messenger RNA in the Edinger-Westphal nucleus of control rats and rats treated with repeated electroconvulsive shock PPT-A mRNA CCK mRNA
Control
ECS
64.7 + 5.2 86.4 & 4.4
93.8 + 6.1’ 118.2 k 3.7’
‘Significant difference to control group (P < 0.05; Student’s two-tailed r-test for paired samples). Mean values + S.E.M.; n = 5.
ECS increases SP and CCK mRNA in PAG in man, has been found to increase the level of SP-like
immunoreactivity in lumbar cerebrospinal fluid,* suggesting increased utilization of SP during pain relief. Chronic pain in man may be associated with depression and it has been suggested that the painrelieving effect of ECT, as well as that of antidepressant drugs, is not a true analgesic effect but is due to the antidepressive action of the treatment. However, it has been reported that following treatment with antidepressant drugs, significant pain relief is seen much earlier than the mo~~levating effect (i.e. within two to three days after onset of treatment) which may suggest that the pain relief is not merely a consequence of the antidepressant action of the drug.12 It should be emphasized that an altered activity in EWn neurons may be related to functions other than pain modulation. The PAG receives neuronal input from several brain regions including hypothalamus, limbic regions and frontal cortex.2s.BJ6 PAG neurons project to forebrain regions and SFcontaining neurons in the ventrom~ial PAG have, for example, been
79
reported to project to the nucleus accumbens.2’ The dorsal PAG has been considered to be the prototypical element of the so-called brain-aversive system which integrates and controls aversive/defensive behaviour in the response to threatening and stressful stimuliB Recently also, selective activation of cell bodies in the ventral PAG has been reported to induce defensive behaviour.26 Since stress is believed to be an important factor in the depressive syndrome,’ it may he suggested that the changes in peptide gene expression in the PAG encountered in the present study may also have relevance for the antidepressant effect of ECT. In conclusion, the present results indicate that SP- and CCK-containing neurons in the EWn are activated by repeated ECS. The importance of these findings for the antidepressant and analgesic effects of ECT remains to be investigated. Acknowledgements-This
study was supported by The Swedish Medical Research Council (nos 8653, 6836), Ake W&ergs Stiftelse, Loo and Hans Gstermans fond, and The Swedish Society of Medicine and Maanus Berawalls Stiftelse. _
REFERENCES
Ml,
4.
5.
6. 7.
664(x%42. Brodin E. and Meyerson B. A. (1983) Substance P in human cerebrospinal fluid effect of central 8rey or dorsal column stimulation for suppression of chronic pain. In Substance P-Dublin 1983 (eds Scabanek P. and Powell D.), pp. 190-191. Boole Press, Dublin. Brodin E., &ren S.-G. and Theodorsson-Norheim E. (1987) Effects of subchronic treatment with imipramin, zimelidine and alaproclat on regional tissue levels of substance P- and neurokinin A-/neurokinin B-like immunoreactivity in the brain and spinal cord of the rat. N~rop~r~cology 26, 581-590. Brodin K., Rosen A., Iwarson K., Ggren 0. and Brodin E. (1989) Increased levels of substance P and choi~ystokinin in rat cerebral cortex following repeated electroconvulsive shock and subchronic treatment with a serotonin untake inhibitor. Acta physiol. stand. i36: 613-614. Danysz W., Kostowski W. and Archer T. (1988) Some aspects of stress and depression. Prog. Neuropsychopharmac. biol. Psychiat. 12, 405-419.
8. Deschenes R. J., Lorenz L. J., Haun R. S., Roos B. A., Collier K. J. and Dixon J. E. (1984) Cloning and sequence analysis of a cDNA encoding rat p~p~chol~yst~kin~n. Proc. natn. Acad, Sci. U.S.A. 81,726-730. 9. Doi T. and Juma I. (1981) Intrathecal substance P depresses the tail-flick response-antagonism by naloxone. Naunyn-Schmiedeberg’s
Arch. Pharmac. 317, 135-139.
10. Fink M. (1987) Convulsive treatment in affective disorders. In Psychopharmacology: The Third Generation of Progress (ed. Meltzer H. Y.), pp. 1071-1076. Raven Press, New York. 11. Geiter C. H. and Nutt D. J. (1989) Chronic electroconvulsive shock and neurotransmitter receptors. Life Std. 44, 985-1006. 12. Gourlay G. K., Cherry D. A., Cousins M. J., Love B. L., Graham J. R. and McLachlan M. 0. (1986) A controlled study of a serotonin muptake blocker, zimelidine, in the treatment of chronic pain. Pain 25, 35-52. 13. Hasegawa M., Usui H., Araki K., Kuwano R. and Takahashi Y. (1986) Developmental and regional changes of cholecystokinin mRNA in rat brain. Fedn Eur. biochem. Sots Lett. 194, 224-226. 14. Holaday J. W., Tortella F. C., Long J. B., Belenky G. L. and Hitzemann R. J. (1986) Endogenous opioids and their rcepltoo evidence for involvement in the postictal effects of electroconvulsive shock. Ann. N.Y. Read. Sci. 462, 15. Innis R. B., Correa F. M. A., Uhl G. R., Schneider B. and Snyder S. H. (1979) Cholecystokinin octa-peptide-like immunoreactivity: histochemical localization in rat brain. Proc. natn. Acad. Sci. U.S.A. 76, 521-525. 16. Innis R. B. and Aghajanian G. K. (1986) Cholecystokinin-containing and nociceptive neurons in rat Edinger-Westphal nucleus. Brain Res. 363. 230-238. 17. Juma I. and Zetler G. (1981) Antinociceptive effect of centrally administered caerulein and cholecystokinin octapeptide (CCK-8). Eur. J. Pharmac. 73, 323-331.
80
N.
hlEFOKS et ul.
18. Kawagushi Y., Hosimura M., Nawa H. and Nakanishi S. (1986) Sequence analysis of cloned cDNA for rat substance P precursor: existence of a third substance P precursor. Eiochem. Biophys. Rex Commun. 139, 1040-1046. 19. Krause J. E., Chirgwin J. M., Carter M. S., Xu 2. S. and Hershey A. D. (1987) Three rat preprotachykinin mRNAs encode the neuropeptides substance P and neurokinin A. Proc. naa. Acad Sci. U.S.A. 84, 881-885.. 20. Lanaud P.. Ponovici T.. Normand E.. Lemoine C.. Bloch B. and Roaues B. P. (1989) Distribution of CCK mRNA in particular regions (hippocampus, periaqueductal gray and thalamus) of the rat by in siru hybridization. Neurosci. Leit. 104, 3842. 21. Li Y.-Q.. Rao Z.-R. and Shi J.-W. (1990) Midbrain periaqueductal gray neurons with substance P- or enkephalin-like immunoreactivity send projection fibers to the nucleus accumbens in the rat. Neurosci. Len. 119, 269-271. 22. Lindefors N., BrenC S., Herrera-Marschitz M. and Persson H. (1990) Neuropeptide gene expression in brain is differentially regulated by midbrain dopamine neurons. Expl Brain Res. 80, 489-500. 23. Malick J. B. and Goldstein J. M. (1978) Analgesic activitv of substance P followinn intracerebral administration in rats. Life Sci. 23, 835-844. 24. Mandell M. (1975) Electroconvulsive therapy for chronic pain associated with depression. Am. J. Psych&. 132, 599-60 1. 25. Marchand J. E. and Hagino N. (1983) Afferents to the periaqueductal gray in the rat. A horseradish peroxidase study. Neuroscience 9, 95-106. 26. Morgan M. M. (1990) Modulation of nociception from the PAG: distinct dorsal and ventral systems. In The Midbrain Periaqueducral Gray Matter, N.A.T.O. Advanced Research Workshop Abstracts, p. 17. Castera-Verduzan, France.
27. Naranjo J. R., Sanches-Franc0 F., Garzbn and de1 Rio J. (1982) Analgesic activity of substance P in rats: apparent mediation by met-enkephalin release. Life Sci. 30, 4446. 28. Neafsey E. J., Hurley-Gius and Arvanitis D. (1986) The topographical organization of neurons in the medial frontal, insular and olfactory cortex projecting to the solitary nucleus, olfactory bulb, periaqueductal gray and superior colliculus. Brain Res. 377, 261-270. 29. Olds M. E. and Olds J. (1963) Approach-avoidance analysis of rat diencephalon. J. camp. Neurol. 120, 259-295. 30. Phipps B. S., Maciewics R., Sandrew B. B., Poletti C. I. and Foote W. E. (19873) Edinger-Westphal neurons that project to spinal cord contain substance P. Neurosci. Left. 36, 125-131. 31. Pietsky J. E. (1946) Disappearance of painful phantom limbs after electric shock treatment. Am. J. Psych&. 462, 599-60 1. 32. Reynolds D. V. (1989) Surgery in the rat during electrical analgesia induced by focal brain stimulation. Science 162, 444445. 33. Rosen A. and Brodin E. (1989) Effect of acute morphine treatment on peptide levels in the peri-aqueductal grey. Acfa physiol. stand. 136, 493494. 34. Savasta M., Ruberte E., Palacios J. M. and Mengod G. (1989) The colocalization of cholecystokinin and tyrosine hydroxylase mRNAs in mesencephalic dopaminergic neurons in the rat brain examined by in situ hybridization. Neuroscience 29, 363-369.
35. Seroogy K., Schaling M., Brene S., Dagerlind A., Chi S. Y., Hiikfelt T., Persson H., Brownstein M., Huan R., Dixin J.. Filer D.. Schlessinaer D. and Goldstein M. (1989) Cholecvstokinin and tvrosine hvdroxvlase messenger RNAs in neurons of’ rat mese&ephalon: peptide/monoa&e’coexistence studies using in siri hybridization combined with immunohistochemistry. Expl Bruin Res. 74, 149-162. 36. Shipcly M. and Ennis M. (1991) Organization of some forebrain inputs to the PAG. In The Midbrain PAG-Functional, Anatomical and Immunohistochemicaf Organization (eds Depaulis A. and Bandler R.). Plenum Press, New York. (In press.) 37. Skirboll L., Hokfelt T., Rehfelt J., Cue110 A. C. and Dockray G. (1982) Coexistence of substance P- and cholecystokinin-like immunoreactivity in neurons of the mesencephalic periaqueductal central gray. Neurosci. Lerr. 28, 35-39. 38. Skirboll L., Hlikfelt T., Dockray G., Rehfelt J., Brownstein M. and Cue110A. C. (1983) Evidence for periaqueductal cholecystokinin-substance P neurons projecting to the spinal cord. J. Neurosci. 3, 1151-l 157. 39. Stewart J. M., Getto C. J., Neldner K., Reeve E. B., Krivoy W. A. and Zimmerman E. (1976) Substance P and analgesia. Science 209, 784-785. 40. Urea G. and Nof-Reshef A. (1985) Spinal mechanisms of the analgesic action of electroconvulsive shock. Brain Res. 4, 110-118. 41. Voigt M. M. and Uhl G. R. (1988) Preprocholecystokinin mRNA in rat brain: regional expression includes thalamus. Molec. Brain Res. 4, 247-253. 42. von Hagen K. 0. (1957) Chronic intolerable pain. J. Am. med. Ass. 165, 773-777.
43. Warden M. J. and Young W. S. III (1988) Distribution of cells containing mRNAs encoding substance P and neurokinin B in the rat central nervous system. J. camp. Neural. 272, 90-l 13. 44. Weiss D. M., Bloomberg W. and Drew J. H. (1957) The use of electroshock as therapy for atypical and intractable pain. Trans. Am. Neural. Ass. 82, 166-168. 45. Young W. S., Bonner T. I. and Brann M. R. (1986) Mesencephalic dopamine neurons regulate the expression of neuropeptide mRNAs in the rat forebrain. Proc. natn. Acad. Sci. U.S.A. 83, 9827-9831. 46. Zetler G. (1980) Analgesia and ptos caused by caerulein and cholecystokinin octapeptide (CCK-8). Nemopharmacology 19, 415422.
(Accepfed 7 May 1991)
h%?w~scii?nce voi. 45, No. 1, pp. 73-80, 1991 Printed in Great Britain
REPEATED ELECTROCONVULSIVE SHOCK INCREASES TACHYKMIN AND CHOLECYSTOKININ mRNA EXPRESSION IN VENTRAL PERIAQUEDUCTAL GRAY N.
Lrrmnnons,*t
K. BRODIN,$ C.-O. STILLER,* H. PEBSSON§ and E. BRODIN*
*Department of Pharmacology,
Karol&&a Institutet, Box 60400, S-IO401 Stockholm, Sweden IDepartment of Clinicai Pharmacology, Karolinska Hospital, Stockhohn, Sweden §Department of Medical Chemistry, Laboratory of Molecular Neurobiology, Karohnska Institutet, Stockholm, Sweden
Akatraet-The effect of repeated electroconvulsive shock (five shocks during 10 days) on preprocholecystokinin and preprotachykinin-A messenger RNA expression was studied in the mesencephalic periaqueductal gray and adjacent areas of rat using in situ hybridization histochemistry with specific oligonucleoti& probes. An increased number of p~pr~hol~sto~~n and p~~~~y~~n-A messenger RNA hybridization positive neurons (+30% and +47%, respectively) in the Edinger-Westphal nucleus was observed following repented electroconvulsive shock. In addition, both preprocholecystokinin and preprotachykinin-A messenger RNA expression, measured as grain density over single neurons, was si~ifi~n~y increased (+37% and +45%, respectively). The results indicate that cholccystoktmn- and substance Fcomarmng neurons m the lzdmger-Westphal nucleus are activated by repeated electroconvulsive shock, which may be related to the antidepressant and analgesic effects of electroconvulsive shock treatment.
El~tr~on~sive therapy (ECT) is widely used as treatment of severe affective disordersi and ECT has also been found to have an analgesic effect in patients suffering from chronic pain, especially in cases associated with dep~ssion.24,3’,4~~ The mechanisms of action of ECT in treatment of depression and pain are, however, still essentially unknown, even though involvement of several classical transmitter systems and peptidergic systems”‘~ has been implicated from animal experiments with electroconvulsive shock (ECS). Recently, it has been reported that repeated ECS increases tissue levels of substance P (SP) and cholecystokinin (CCK) in certain regions of rat cerebral cortex,6 an effect similar to that observed after subchronic treatment with antidepressant (monoamine up~ke-inhibiting) drugs.5” Treatment of rats with selective serotonin uptake inhibitors also increases the tissue level of SP in the mesencephalic periaqueductal gray (PAG),5 a region known to tTo whom correspondence should be addressed. Abbreviations: CCK. cholecvstokinin: cDNA, complementary deoxyribonucleic acid; dATP, deoxyadenosine ~phosp~te; DlT, di~iothr~to~ ECS, electrocouvulsive shock; ECT, electroconvulsive therapy; EWn, &linger-Westphal nucleus; mRNA, messenger ribonucleic acid; PAG, mesencephalic periaqueductal gray; PBS, phosphate-buffered saline; PPT-A, prepro~chyki~n-A; RIA, r~ioi~unoas~y; SSC, sahne sodium citrate buffer; SP, substance P; tRNA, transfer ribonucleic acid. 73
be involved in, for example, endogenous pain suppression.2*32 Immunohistochemical studies have shown that SPand CCK-like immunoreactivities are present in both terminals and cell bodies in rat and cat ventral PAG’s*‘6*M~37 and coexist in PAG neurons.37 Cells in ventral PAG, i.e. in the Edinger-Westphal nucleus (EWn), respond to noxious stimuli with an increased firing rateI and SP- and CCK-immunoreactive cells in this region have been found to project to the spinal cord and have been suggested to represent a descending system which could influence sensory transmission at the spinal leve1.30*38Recently, acute treatment with morphine, which is believed to exert part of its analgesic effect via an action on the PAG, was found to induce increased tissue levels of SP and CCK in this brain region3’ The recent development of the in situ hybridization technique4$ makes it possible to study expression of neuropeptide genes in defined neuron populations in the brain. Rat preprotachykinin A (PPT-A, a SP precursor) messenger RNA19 (mRNA) and rat CCK mRNAa have been sequenced and ~ornpl~en~~ DNA (cDNA) clones isolated. Hybridization probes from these cDNA clones as well as synthetic oligonucleotide probes, including those used in the present study, have with good agreement been used to document the expression of PPT-A mRNA3~20~22~45 and CCK mRNAi3~22~M*35~4’~43 in rat brain.
Fig. 1. Autoradiograms of PPT-A mRNA in situ hybridization histochemistry in rat mesencephalon. The upper panel shows an autoradiogram from a control animal and the lower panel shows an autoradiogram from an ECS-treated animal. The tissue was hybridized to an oligonucleotide PPT-A mRNA probe exposed to a photographic film. The hybridization signal considered to belong to the EWn was located 0.7 mm ventral to the ventral limit of the aaueduct at the coronal level shown in the figure. Arrows indicate cells shown in Fig. 3. CO, control
In the present study the effect of repeated ECS on expression of PPT-A and preproCCK mRNA was assessed in PAG and adjacent areas by in situ hybridization histochemistry combined with computerized image analysis.
EXPERIMENTAL PROCEDURES
Animals and tissue preparation
Male Sprague-Dawley rats (ALAB Laboratorietjanst. Sollentuna, Sweden; body weight 150-2oOg) were used. For ECS, the animals were anaesthetized with halothane-nitrous oxide-oxygen and stimulation was applied through ear-clip electrodes (1 s, 150 mA, 50
Sachs cycles, x s sinusoidal) using rodent shocker Elektronik Frieburg, F.R.G.). stimulations gave to generalized lasting for 5 s. group of animals received shocks over days (days and 10) five control were treated the same except that current was The rats killed by 24 h the last The brains immediately dissected frozen on ice and at -80°C. sections (16pm) the level EWn were on a (Leitz, Wetzlar, Sections of and control were paired thaw-mounted on (50 pg/ml)-coated In situ The tissues buffered saline
fixed in for 30
formalin in rinsed twice
4 min
ECS increases SP and CCK mRNA in PAG
75
Fig. 2. Autoradiograms of CCK mRNA in situ hybridization his&hem&y in rat mesencephalon. The upper panel shows an autoradiogram from a control animal and the lower panel shows an autoradiogram from an ECS-treated animal. The tissue was hybridized to an oligonucleotide CCK mRNA probe exposed to a photographic Iilm. The hybridization signal considered to belong to the EWn was located 0.7 mm ventral to the ventral limit of the aqueduct at the coronal level shown in the figure. Arrows indicate cells shown in Fig. 4. CO, control.
PBS, and delipidated in graded series of ethanol, including a 5-min incubation with chloroform. The sections were then air-dried. The hybridization cocktail contained 50% formamide, 4 x saline sodium citrate buffer (SSC) (1 x SSC is 0.15 M NaCl. 0.015 M sodium citrate. DH 7.0). 1 x Denhardt’s solution, 1% Sarcosyl, 0.02M‘ Na,Pd; (pH 7.0), 10% dextransulphate, 0.5mg/ml yeast transfer RNA (tRNA). 0.06 M dithiothreitol IDD17 and 0.1 ma/ml sheared salmon sperm DNA. A 48-mer ‘oligonucle&de complementary to rat PPT-A mRNA coding for amino acids 49-64, and thus complementary to a, /I and y -PPT-A mRNA,i9 was used to detect PPT-A mRNA expression. The sequence of the PPT-A-specific oligonucleotide was 5’GAACTGCTGAGGCTTGGGTCTTCGGGCGA’I-ICTCTGAAGAAGATGCTC. A 44-mer oligonucleotide complementary to rat preproCCK mRNA coding for amino
acids 89-103 was used to detect preproCCK mRNA8 (in the following text refered to as CCK mRNA). The sequence of the prep;oCCK-specific oligonucleotide- was S’AAATCCATCCCAGCCCATGTAGTCCCGGTCACTIATCCTGTGGCT. The oligonucleotide was labelled at the 3’-end with a[Srs]dATP using terminal deoxyribonucleotidyl transferase (International Biotech Inc., New Haven, Cl) to a specific activity of approximately 8 x 108 c.p.m./mg. The labelled probes were purified on a Nensorb column (DuPont, Wilmington, DE) prior to use. Following hybridization for 16 h at 42”C, the sections were rinsed five times for 15 min each in 1 x SSC at 55°C. Finally, the sections were rinsed in autoclaved water for 5 min, and then dehydrated through a series of graded alcohol and air-dried. The sections were exposed to X-ray lilm (Amersham jI-max) for four to eight days and to photographic emulsion (Kodak
76
N. LINDEFOKS et al.
Fig. 3. High magnification bright-field emulsion photomicrographs of PPT-A mRNA-positive neurons 111 ventromedial PAG from a representative control brain (left panel) and the ECS brain (right panel) from the same slides also shown in Fig.
In situ hybridization with cDNA probes for PPT-A mRNA (a generous gift from Dr Shigetada Nakanishi, Kvoto. Janan’*) and CCK mRNA (a generous gift from Dr Jack dixon,‘West Lafayette, U.6A.p in adjacent tissue sections of rat midbrain have, in separate control experiments, shown a similar hybridization pattern to those found with the PPT-A and CCK oligonucleotide probes, respectively. For comparison of distribution of the hybridization signal, we also performed hybridization with a 48-mer oligonucleotide probe specific for the unrelated rat tyrosine hydroxylase mRNA.**
Neuron counting and computerized image analysis The numerical density of hybridization-positive neurons was analysed using a Lcitz Ortholux microscope at x 313
magnification. Hybridization-positive neurons were defined as those expressing more than 10 grains per neuron (well above background density of grains). For each individual brain all neurons in EWn with significant levels of PPT-A mRNA and CCK mRNA, respectively, were counted in three sections and a mean value per section was estimated. The number of grains per neuron and the size of PPT-A mRNA- and CCK mRNA-positive neurons was calculated using a Crystal image analysis processor (Quantel Ltd, U.K.) at x 500 magnification. For this purpose, the microscope was equipped with a Panasonic WV 1500 TV camera to provide a video signal for the image analysis. Grain density was analysed over a total of 10 neurons in each animal and in five animals of each of the two groups. Over some of the neurons analysed the high density of grains made a direct numerical count futile since the grains were fused in some cases. However, the grains never covered a major part of hybridization-positive neurons. Instead, the
ECS increases SP and CCK mRNA in PAG
Fig. 4. High magnification bright-field emulsion photomicrographs of CCK mRNA-positive neurons in ventromedial PAG from a representative control brain (left panel) and the ECS brain (right panel) from the same slides also shown in Fig. 2. Scale bars = 50 pm. CO, control.
total area of grains, including all fused grains was divided with a standard grain area obtained from measurements of unfused grams over neurons. The area of a standard grain was arbitrarily defbred as 0.3 pm.Z The used method of grain counting (i.e. gram area counting) underestimates changes in grain volume since only a two-dimensional aspect of the grains is considered. However, this conservative aspect of the method implies that sign&ant changes in grain area measured are reliable, although in a semiquantitative manner. Evaluating the results one must bear in mind that the real change in mRNA content in single cells is most probably higher than the measured increase in grain area.
REsJLTs
In situ hybridization of coronal tissue sections at the level of mesencephalon showed PPT-A mRNA-
expressing neurons in the colliculus superior, interpeduncular nucleus, peritectal area and most parts of the PAG with a high level in the EWn (Fig. 1). Scattered hybridization-positive neurons were also found in the neocortex. CCK mRNA-expressing neurons were seen in layers II and III, and V and VI of the neocortex, in the medial geniculate nucleus, ventral tegmental area, substantia nigra, interpeduncular nucleus and intensely labelled cells were observed in the EWn (Fig. 2). In control animals, the number of CCK mRNA-positive neurons in the EWn was nearly two times higher than the number of PPT-A mRNA-positive neurons (Figs 3, 4). For comparison, hybridization with a 48-mer oligonucleotide probe specific for rat tyrosine
78
Fig. 5. Autoradiogram of a coronal section of rat midbrain including the EWn, substantia nigra and ventral tegmental area, labelled by in situ hybridization with an oligonucleotide specific for tyrosine hydroxylase mRNA. Note the labelling of tyrosine hydroxylase mRNA in substantia nigra and ventral tegmental area and the lack of signal in the EWn (indicated in the midline by the arrow).
Table 1. Number of preprotachykinin-A messenger RNAand cholecystokinin messenger RNA-positive neurons in the Edinger-Westphal nucleus of control rats and rats treated with repeated electroconvulsive shock PPT-A mRNA CCK mRNA
Control
ECS
31.4 + 2.8 57.1 + 4.9
46.2 + 1.6’ 74.3 + 3.2’
‘Significant difference to control group (P < 0.05; Student’s two-tailed r-test for paired samples). Values indicate the number of hybridization-positive neurons per tissue section (mean &-S.E.M.; n = 5).
hydroxylase mRNA was performed which showed a lack of hybridization signal in PAG but a strong hybridization signal in neurons in the substantia nigra compacta and ventral tegmental area (Fig. 5). Repeated ECS increased the number of neurons expressing significant levels of PPT-A mRNA (+47%; P < 0.05;Student’s t-test for paired samples) as well as the number of grains (+45%; P < 0.05)over individual neurons in the EWn (Figs 1, 3, Tables 1, 2). A similar increase was seen in the number of CCK mRNA-expressing neurons (+30%; P < 0.05)and in the number of grains (+37%; P < 0.05)over individual CCK mRNApositive neurons (Figs 2, 4, Tables 1, 2). DISCUSSION
The present results demonstrate that repeated ECS increases the PPT-A and CCK mRNA expression in a group of neurons in the EWn in the ventromedial aspect of the rat mesencephalic PAG. PPT-A and CCK mRNA-positive neurons have previously been demonstrated in EWn using oligonucleotide probes and a similar hybridization protocol as in the present
study. 20~35,43 These cells are most probably identical to the periaqueductal SP/CCK neurons previously demonstrated to project to the spinal cord using immunohistochemical and retrograde tracing techniques. y),37,38The role of the SPjCCK neurons in the EWn is at present unclear, but, as previously suggested, these neurons may represent a descending system which influences pain transmission.30x3* Electrophysiological recordings from the EWn have revealed a group of neurons that responds to noxious stimuli like toe-pinch with an increased firing rate, an effect that may be blocked by systemic morphine in a naloxone-reversible manner.16 Although the role of SP and CCK in pain perception seems complex, both SP and CCK have been reported to have an antinociceptive effect upon peripheral administration,39.46 intrathecal injection9,‘7,27 as well as when injected intracerebroventricularly or directly in the PAG.‘.‘7,23,‘9 From the present results, it may be suggested that descending periaqueductal SPjCCK neurons may take part in the mediation of ECS-induced antinociception demonstrated in animals,14,” and possibly also in the pain-relieving effect of ECT in man (see Introduction). Interestingly, electrical stimulation in PAG for the treatment of chronic pain Table 2. Grain density over hybridization-positive neurons expressing preprotachykinin-A messenger RNA in the Edinger-Westphal nucleus of control rats and rats treated with repeated electroconvulsive shock PPT-A mRNA CCK mRNA
Control
ECS
64.7 + 5.2 86.4 & 4.4
93.8 + 6.1’ 118.2 k 3.7’
‘Significant difference to control group (P < 0.05; Student’s two-tailed r-test for paired samples). Mean values + S.E.M.; n = 5.
ECS increases SP and CCK mRNA in PAG in man, has been found to increase the level of SP-like
immunoreactivity in lumbar cerebrospinal fluid,* suggesting increased utilization of SP during pain relief. Chronic pain in man may be associated with depression and it has been suggested that the painrelieving effect of ECT, as well as that of antidepressant drugs, is not a true analgesic effect but is due to the antidepressive action of the treatment. However, it has been reported that following treatment with antidepressant drugs, significant pain relief is seen much earlier than the mo~~levating effect (i.e. within two to three days after onset of treatment) which may suggest that the pain relief is not merely a consequence of the antidepressant action of the drug.12 It should be emphasized that an altered activity in EWn neurons may be related to functions other than pain modulation. The PAG receives neuronal input from several brain regions including hypothalamus, limbic regions and frontal cortex.2s.BJ6 PAG neurons project to forebrain regions and SFcontaining neurons in the ventrom~ial PAG have, for example, been
79
reported to project to the nucleus accumbens.2’ The dorsal PAG has been considered to be the prototypical element of the so-called brain-aversive system which integrates and controls aversive/defensive behaviour in the response to threatening and stressful stimuliB Recently also, selective activation of cell bodies in the ventral PAG has been reported to induce defensive behaviour.26 Since stress is believed to be an important factor in the depressive syndrome,’ it may he suggested that the changes in peptide gene expression in the PAG encountered in the present study may also have relevance for the antidepressant effect of ECT. In conclusion, the present results indicate that SP- and CCK-containing neurons in the EWn are activated by repeated ECS. The importance of these findings for the antidepressant and analgesic effects of ECT remains to be investigated. Acknowledgements-This
study was supported by The Swedish Medical Research Council (nos 8653, 6836), Ake W&ergs Stiftelse, Loo and Hans Gstermans fond, and The Swedish Society of Medicine and Maanus Berawalls Stiftelse. _
REFERENCES
Ml,
4.
5.
6. 7.
664(x%42. Brodin E. and Meyerson B. A. (1983) Substance P in human cerebrospinal fluid effect of central 8rey or dorsal column stimulation for suppression of chronic pain. In Substance P-Dublin 1983 (eds Scabanek P. and Powell D.), pp. 190-191. Boole Press, Dublin. Brodin E., &ren S.-G. and Theodorsson-Norheim E. (1987) Effects of subchronic treatment with imipramin, zimelidine and alaproclat on regional tissue levels of substance P- and neurokinin A-/neurokinin B-like immunoreactivity in the brain and spinal cord of the rat. N~rop~r~cology 26, 581-590. Brodin K., Rosen A., Iwarson K., Ggren 0. and Brodin E. (1989) Increased levels of substance P and choi~ystokinin in rat cerebral cortex following repeated electroconvulsive shock and subchronic treatment with a serotonin untake inhibitor. Acta physiol. stand. i36: 613-614. Danysz W., Kostowski W. and Archer T. (1988) Some aspects of stress and depression. Prog. Neuropsychopharmac. biol. Psychiat. 12, 405-419.
8. Deschenes R. J., Lorenz L. J., Haun R. S., Roos B. A., Collier K. J. and Dixon J. E. (1984) Cloning and sequence analysis of a cDNA encoding rat p~p~chol~yst~kin~n. Proc. natn. Acad, Sci. U.S.A. 81,726-730. 9. Doi T. and Juma I. (1981) Intrathecal substance P depresses the tail-flick response-antagonism by naloxone. Naunyn-Schmiedeberg’s
Arch. Pharmac. 317, 135-139.
10. Fink M. (1987) Convulsive treatment in affective disorders. In Psychopharmacology: The Third Generation of Progress (ed. Meltzer H. Y.), pp. 1071-1076. Raven Press, New York. 11. Geiter C. H. and Nutt D. J. (1989) Chronic electroconvulsive shock and neurotransmitter receptors. Life Std. 44, 985-1006. 12. Gourlay G. K., Cherry D. A., Cousins M. J., Love B. L., Graham J. R. and McLachlan M. 0. (1986) A controlled study of a serotonin muptake blocker, zimelidine, in the treatment of chronic pain. Pain 25, 35-52. 13. Hasegawa M., Usui H., Araki K., Kuwano R. and Takahashi Y. (1986) Developmental and regional changes of cholecystokinin mRNA in rat brain. Fedn Eur. biochem. Sots Lett. 194, 224-226. 14. Holaday J. W., Tortella F. C., Long J. B., Belenky G. L. and Hitzemann R. J. (1986) Endogenous opioids and their rcepltoo evidence for involvement in the postictal effects of electroconvulsive shock. Ann. N.Y. Read. Sci. 462, 15. Innis R. B., Correa F. M. A., Uhl G. R., Schneider B. and Snyder S. H. (1979) Cholecystokinin octa-peptide-like immunoreactivity: histochemical localization in rat brain. Proc. natn. Acad. Sci. U.S.A. 76, 521-525. 16. Innis R. B. and Aghajanian G. K. (1986) Cholecystokinin-containing and nociceptive neurons in rat Edinger-Westphal nucleus. Brain Res. 363. 230-238. 17. Juma I. and Zetler G. (1981) Antinociceptive effect of centrally administered caerulein and cholecystokinin octapeptide (CCK-8). Eur. J. Pharmac. 73, 323-331.
80
N.
hlEFOKS et ul.
18. Kawagushi Y., Hosimura M., Nawa H. and Nakanishi S. (1986) Sequence analysis of cloned cDNA for rat substance P precursor: existence of a third substance P precursor. Eiochem. Biophys. Rex Commun. 139, 1040-1046. 19. Krause J. E., Chirgwin J. M., Carter M. S., Xu 2. S. and Hershey A. D. (1987) Three rat preprotachykinin mRNAs encode the neuropeptides substance P and neurokinin A. Proc. naa. Acad Sci. U.S.A. 84, 881-885.. 20. Lanaud P.. Ponovici T.. Normand E.. Lemoine C.. Bloch B. and Roaues B. P. (1989) Distribution of CCK mRNA in particular regions (hippocampus, periaqueductal gray and thalamus) of the rat by in siru hybridization. Neurosci. Leit. 104, 3842. 21. Li Y.-Q.. Rao Z.-R. and Shi J.-W. (1990) Midbrain periaqueductal gray neurons with substance P- or enkephalin-like immunoreactivity send projection fibers to the nucleus accumbens in the rat. Neurosci. Len. 119, 269-271. 22. Lindefors N., BrenC S., Herrera-Marschitz M. and Persson H. (1990) Neuropeptide gene expression in brain is differentially regulated by midbrain dopamine neurons. Expl Brain Res. 80, 489-500. 23. Malick J. B. and Goldstein J. M. (1978) Analgesic activitv of substance P followinn intracerebral administration in rats. Life Sci. 23, 835-844. 24. Mandell M. (1975) Electroconvulsive therapy for chronic pain associated with depression. Am. J. Psych&. 132, 599-60 1. 25. Marchand J. E. and Hagino N. (1983) Afferents to the periaqueductal gray in the rat. A horseradish peroxidase study. Neuroscience 9, 95-106. 26. Morgan M. M. (1990) Modulation of nociception from the PAG: distinct dorsal and ventral systems. In The Midbrain Periaqueducral Gray Matter, N.A.T.O. Advanced Research Workshop Abstracts, p. 17. Castera-Verduzan, France.
27. Naranjo J. R., Sanches-Franc0 F., Garzbn and de1 Rio J. (1982) Analgesic activity of substance P in rats: apparent mediation by met-enkephalin release. Life Sci. 30, 4446. 28. Neafsey E. J., Hurley-Gius and Arvanitis D. (1986) The topographical organization of neurons in the medial frontal, insular and olfactory cortex projecting to the solitary nucleus, olfactory bulb, periaqueductal gray and superior colliculus. Brain Res. 377, 261-270. 29. Olds M. E. and Olds J. (1963) Approach-avoidance analysis of rat diencephalon. J. camp. Neurol. 120, 259-295. 30. Phipps B. S., Maciewics R., Sandrew B. B., Poletti C. I. and Foote W. E. (19873) Edinger-Westphal neurons that project to spinal cord contain substance P. Neurosci. Left. 36, 125-131. 31. Pietsky J. E. (1946) Disappearance of painful phantom limbs after electric shock treatment. Am. J. Psych&. 462, 599-60 1. 32. Reynolds D. V. (1989) Surgery in the rat during electrical analgesia induced by focal brain stimulation. Science 162, 444445. 33. Rosen A. and Brodin E. (1989) Effect of acute morphine treatment on peptide levels in the peri-aqueductal grey. Acfa physiol. stand. 136, 493494. 34. Savasta M., Ruberte E., Palacios J. M. and Mengod G. (1989) The colocalization of cholecystokinin and tyrosine hydroxylase mRNAs in mesencephalic dopaminergic neurons in the rat brain examined by in situ hybridization. Neuroscience 29, 363-369.
35. Seroogy K., Schaling M., Brene S., Dagerlind A., Chi S. Y., Hiikfelt T., Persson H., Brownstein M., Huan R., Dixin J.. Filer D.. Schlessinaer D. and Goldstein M. (1989) Cholecvstokinin and tvrosine hvdroxvlase messenger RNAs in neurons of’ rat mese&ephalon: peptide/monoa&e’coexistence studies using in siri hybridization combined with immunohistochemistry. Expl Bruin Res. 74, 149-162. 36. Shipcly M. and Ennis M. (1991) Organization of some forebrain inputs to the PAG. In The Midbrain PAG-Functional, Anatomical and Immunohistochemicaf Organization (eds Depaulis A. and Bandler R.). Plenum Press, New York. (In press.) 37. Skirboll L., Hokfelt T., Rehfelt J., Cue110 A. C. and Dockray G. (1982) Coexistence of substance P- and cholecystokinin-like immunoreactivity in neurons of the mesencephalic periaqueductal central gray. Neurosci. Lerr. 28, 35-39. 38. Skirboll L., Hlikfelt T., Dockray G., Rehfelt J., Brownstein M. and Cue110A. C. (1983) Evidence for periaqueductal cholecystokinin-substance P neurons projecting to the spinal cord. J. Neurosci. 3, 1151-l 157. 39. Stewart J. M., Getto C. J., Neldner K., Reeve E. B., Krivoy W. A. and Zimmerman E. (1976) Substance P and analgesia. Science 209, 784-785. 40. Urea G. and Nof-Reshef A. (1985) Spinal mechanisms of the analgesic action of electroconvulsive shock. Brain Res. 4, 110-118. 41. Voigt M. M. and Uhl G. R. (1988) Preprocholecystokinin mRNA in rat brain: regional expression includes thalamus. Molec. Brain Res. 4, 247-253. 42. von Hagen K. 0. (1957) Chronic intolerable pain. J. Am. med. Ass. 165, 773-777.
43. Warden M. J. and Young W. S. III (1988) Distribution of cells containing mRNAs encoding substance P and neurokinin B in the rat central nervous system. J. camp. Neural. 272, 90-l 13. 44. Weiss D. M., Bloomberg W. and Drew J. H. (1957) The use of electroshock as therapy for atypical and intractable pain. Trans. Am. Neural. Ass. 82, 166-168. 45. Young W. S., Bonner T. I. and Brann M. R. (1986) Mesencephalic dopamine neurons regulate the expression of neuropeptide mRNAs in the rat forebrain. Proc. natn. Acad. Sci. U.S.A. 83, 9827-9831. 46. Zetler G. (1980) Analgesia and ptos caused by caerulein and cholecystokinin octapeptide (CCK-8). Nemopharmacology 19, 415422.
(Accepfed 7 May 1991)
