Proc. Nati. Acad. Scf. USAVol. 74, No. 7, pp. 3081-3085, July 1977Neurobiology

Immunohistochemical analysis of peptide pathways possibly relatedto pain and analgesia: Enkephalin and substance P

(spinal interneurons/descending spinal systems/primary afferents)

TOMAS HOKFELT*, AKE LJUNGDAHL*, LARS TERENIUSt, ROBERT ELDE**, AND GORAN NILSSON§Departments of * Histology and § Pharmacology, Karolinska Institute, Stockholm, and t Department of Medical Pharmacology, Uppsala University, Uppsala,Sweden

Communicated by U. S. von Euler, April 25,1977

ABSTRACT The distribution of Met-enkephalin- and sub-stance P-immunoreactive neurons was studied by indirect im-munofluorescence in some areas related to pain and analgesia.Met-enkephalin- and substance P-positive cell bodies and nerveterminals were observed in the periaqueductal central gray, thenucleus raphe magnus, the marginal layer and substantia ge-latinosa ofthe spinal trigeminal nucleus, and the dorsal hornof the spinal cord. Lesion experiments suggest that Met-en-kephalin neurons in the dorsal horn and possibly in the spinaltrigeminal nucleus are interneurons or propriospinal neuronswith nerve terminals in the laminae I and II of tfe cord and inthe superficial layers of the spinal trigeminal nucleus, respec-tively. These areas are also very rich in substance P-positivenerve terminals, mainly representing central branches of pri-mary afferent neurons.The present immunohistochemical-anatomical findings

support the hypothesis that stimulation-produced analgesia isrelated to activation of spinal and spinal trigeminal enkephalininterneurons forming axo-axonic synapses with (substance P?)pain afferents in the superficial laminae of the dorsal horn andthe spinal trigeminal nucleus. These interneurons may be acti-vated by sensory fibers and by descending fibers from medullarystimulation sites. Transmitter substances in these descendingfibers may be 5-hydroxytryptamine and substance P.

Morphine is a classical agent for the relief of severe, chronicpain. Strong analgesia can also be produced by electric stimu-lation of mesencephalic and medullary structures (refs. 1-9, seeref. 10), by acupuncture, or by "electroacupuncture" (11, 12),a modification of the classical Chinese procedure. Several ob-servations suggest a common element in these approaches forpain relief. For instance, narcotic antagonists, such as naloxone(13), block analgesia induced by electrical stimulation (6-9, 14)and by morphine injections (15). Furthermore, tolerance de-velops to electrostimulation analgesia and to morphine analgesiaand cross tolerance can be observed (16).

Recently, Hughes and collaborators (17) have identified twopentapeptides, Leu-enkephalin and Met-enkephalin, that mayrepresent endogenous ligands for opiate receptors. Availableevidence suggests that the activity profile of endogenous opioidsis similar to that of the opiates. For instance, enkephalin hasanalgesic activity in vivo and the narcotic antagonist naloxonewill also counteract enkephalin-induced elevation of painthresholds (18, 19). It is therefore conceivable that these ligandsmay be involved in stimulation-produced analgesia.

Using the immunohistochemical approach, we have beenable to outline the gross distribution of enkephalin-positivestructures in the central nervous system (refs. 20 and 21, see alsoref. 22). Enkephalin neurons were found in many areas of thebrain and spinal cord, indicating involvement in many differenttypes of central nervous system functions. Here, however, wefocus attention strictly on areas assumed to be of importance

in morphine- and stimulation-produced analgesia: the peri-aqueductal central gray, the medullary raphe nuclei, the spinaltrigeminal nucleus, and the spinal cord. It may be significantthat another peptide, substance P, originally discovered by vonEuler and Gaddum (23) and structurally characterized byLeeman and collaborators (24), is present in the same or adja-cent areas and also in a population of primary sensory neurons(25). In the present paper, the immunohistochemical distri-bution of the enkephalin and substance P systems is comparedwith neural substrates involved in pain and analgesia (see refs.10 and 26).

MATERIAL AND METHODS

Male albino rats (Sprague-Dawley, body weight 150 g) weredivided into three groups: (i) untreated rats, (ii) rats receivingan injection of 10 or 25 pug of colchicine dissolved in 0.9%(wt/vol) saline (1 gug//l) into the lateral ventricle or the cisternamagna, and (iii) rats receiving an intraspinal injection of 20 ,ugof colchicine dissolved in 0.9% saline (10 ,ug!/ul) at the lumbo-sacral level. The colchicine-treated animals were sacrificed24-48 hr after the injection.We have previously observed widespread enkephalin-posi-

tive nerve terminals whereas no immunoreactive cell bodiescould be detected (20). Colchicine has been used in other studiesto increase cell body levels of catecholamines (27), probably dueto inhibition of axonal transport.

Antiserum to Met-enkephalin was raised in rabbits as de-scribed elsewhere (20). Crossreaction was about 10% withLeu-enkephalin and <0.1% with a-, A-, and 'y-endorphins,substance P, and somatostatin (L. Terenius, unpublished re-suIts). Furthermore, because Met-enkephalin occurs in amounts3-4 times greater than those of Leu-enkephalin in rat brain (28),we ascribe most of the immunofluorescence to Met-enkephalin.Antiserum against substance P was raised in rabbits as describedpreviously (29).The rats were perfused with formalin. After being rinsed,

the brain and spinal cord were cut on a cryostat (Dittes,Heidelberg) and processed for the indirect immunofluorescencetechnique (see ref. 30). Series of four consecutive sections atdifferent levels of the mesencephalon, lower brain stem, andspinal cord were incubated for 30 min in a humid atmosphereat 370 with antiserum to Met-enkephalin (diluted 1:10 or 1:20),substance P (diluted 1:40), and the two control sera, respec-tively. The control sera, in which the Met-enkephalin andsubstance P antibodies were blocked with the respective peptide(50 jug/ml of serum diluted 1:10), were included to establishthe specificity of the staining. After a rinse in phosphate-buf-fered saline, the sections were incubated with fluorescein iso-thiocyanate-conjugated antibodies (SBL, Stockholm; diluted1:4) under the same conditions as described above, rinsed in

3081

* Present address: Department of Anatomy, University of Minnesota,Minneapolis, MN.

Dow

nloa

ded

by g

uest

on

May

17,

202

0

3082 Neurobiology: Hokfelt et al.

1^~~~i

FIG. 1. Immunofluorescence studies. Arrowhead always points dorsally. DF, dorsal funiculus; P, tractus corticospinalis; asterisk indicatescentral canal. (Bar indicates 50Mm.) Anatomic sites: dorsal horn of the spinal cord (A, B); nucleus raphe magnus (C); intermediate gray matterof the sacral spinal cord (D-F). Treatments: untreated (A, B); colchicine-treated (C-F). Sera: antiserum to substance P (A, C, E); antiserumto Met-enkephalin (B, D); enkephalin control serum (F). A and B and D, E, andF are consecutive sections, respectively. A high density of substanceP-positive nerve terminals was seen in laminae I and II of the dorsal horn (A). The rectangle in A indicates approximately the area shown inB. The enkephalin-positive nerve terminals had a parallel but less dense distribution as seen at the higher magnification (B). Note lack of en-kephalin-positive fibers in the zone of Lissauer (arrow in B) whereas numerous substance P-positive fibers are present in this area (arrow inA). Numerous substance P-immunoreactive cell bodies were seen in the nucleus raphe magnus (C) after colchicine treatment. At the sacrallevel both enkephalin- (arrows in D) and substance P-positive (arrows in E) cell bodies were seen after intraspinal colchicine injections. Notethe networks of enkephalin- (D) and substance P-positive (E) nerve terminals. No fluorescent neurons or nerve terminals were seen after incubationwith control serum (F).

phosphate-buffered saline, mounted in phosphate-bufferedsaline/glycerol, 1:3 (vol/vol), and examined in a Zeiss fluores-cence microscope. The anatomical nomenclature is accordingto Palkovits and Jacobowitz (31).

RESULTSIn untreated rats, extensive networks of Met-enkephalin- andsubstance P-positive nerve terminals were observed in theperiaqueductal central gray and in the marginal layer and

Proc. Natl. Acad. Sci. USA 74 (1977)

Dow

nloa

ded

by g

uest

on

May

17,

202

0

Proc. Natl. Acad. Sci. USA 74 (1977) 3083

I

16FIG. 2. Immunofluorescence studies of the dorsal horn of the spinal cord (lower lumbar and upper sacral level) of colchicine-treated rat

after incubation with antiserum to Met-enkephalin. DF, dorsal funiculus. (Bar indicates 50 Am.) Cell bodies (arrows) are observed in superficial(B) and deeper laminae (C). B and C represent higher magnifications ofA as indicated by rectangles. Note that the fluorescent nerve terminalsare less bright compared to Fig. 1B.

substantia gelatinosa of the spinal trigeminal nucleus. It is im-portant to note that both enkephalin- and substance P-positivefibers were present only in the caudal parts of the spinal tri-geminal nucleus and that there was a striking overlap betweenthe distribution patterns of the nerve terminals containing thesetwo peptides. In the medullary raphe nuclei (nucleus raphemagnus and nucleus raphe pallidus), low to moderate concen-trations of substance P-positive and Met-enkephalin-positivenerve terminals were observed. In the spinal cord, the highestconcentrations were observed in laminae I and 11 (32) (Fig. 1A and B) with moderately dense networks in laminae IV-VII,the ventral horn, and the area around the central canal. In thefirst two laminae, there were considerably more substanceSP-positive than Met-enkephalin-positive fibers. Whereas thezone of Lissauer contained numerous cross-sectioned substanceP-positive fibers, no immunoreaction with Met-enkephalinantiserum could be observed. No Met-enkephalin- or substanceP-immunoreactive cell bodies could be observed in any of theseareas. 4

In colchicine-treated rats, numerous Met-enkephalin- andsubstance P-positive cells were seen in addition to nerve ter-minals. Generally, the immunofluorescence of nerve terminals,especially the enkephalin-positive ones, appeared less strongand distinct in colchicine-treated rats compared to untreatedrats.

Met-enkephalin- and substance P-immunoreactive cell bodieswere observed in the ventrolateral parts of the periaqueductalcentral gray and in the raphe magnus (Fig. IC) and raphepallidus nuclei. In the spinal trigeminal nucleus, many Met-enkephalin-positive cell bodies were observed in the marginallayer of its caudal part whereas substance P-positive cells wereobserved only occasionally. Numerous Met-enkephalin- andsubstance P-immunoreactive cell bodies were observed in thespinal cord (Figs. 1 D and E and 2). They were present mainlyin laminae I-V. There were more Met-enkephalin-positive thansubstance P-positive cells in the dorsal horn of the spinal cord.Strongly fluorescent substance P-positive axons were presentin the caudal medulla oblongata ventromedially to the decus-satio pyramidis. In this area, substance P-positive cells were also

observed. The results described above are summarized in Fig.3.

Met-enkephalin- and substance P-positive cell bodies, axons,and nerve terminals were observed also in many other areas ofthe central nervous system. A full account of the localizationof these neurons will be given elsewhere.None of the immunoreactive structures described above were

observed after incubation with control sera (Fig. IF).

DISCUSSIONThere is evidence that morphine and electrical stimulations ofsome brain loci and of peripheral nerves produce analgesia viathe same type of receptors and that the enkephalins representimportant endogenous ligands for these receptors. The presentfindings demonstrate enkephalin-positive nerve terminals inall areas where morphine is known to produce behavioral an-algesia, including the periaqueductal gray, the medullary raphenuclei, the caudal trigeminal nucleus, and the spinal cord. Ithas been inferred from several studies that the ultimate site ofaction of midbrain stimulation would in fact be the spinal tri-geminal nucleus or the spinal cord (refs. 5 and 33-36, see ref.10). It is possible that, in electroacupuncture, activation of en-dorphin systems also occurs at the segmental level because thisincreases spinal cerebrospinal fluid levels of endorphins (37).Our findings indicate that at the spinal level the enkephalinsystem is either interneuronal or propriospinal. Both en-kephalin-positive terminals and cell bodies were found inthe dorsal horn. Neither transection of the cord at the thoraciclevel nor unilateral dorsal rhizotomy at the lumbar and sacrallevel produced any marked changes in the concentration ofenkephalin-positive nerve terminals in the lumbar cord (38).The presence of enkephalin nerve terminals and enkephalincell bodies in the marginal layers and substantia gelatinosa ofthe spinal trigeminal nucleus may indicate the existence ofenkephalin interneurons also in this nucleus. It is suggested thatactivation of enkephalin interneurons in the spinal trigeminalnucleus and in the dorsal horn is the last and critical step instimulation-produced analgesia. This finding supports con-

Neurobiology: Hakfelt et al.

Dow

nloa

ded

by g

uest

on

May

17,

202

0

3084 Neurobiology: Hbkfelt et al.

D .FIG. 3. Distribution of enkephalin-positive (left side) and sub-

stance P-positive (right side) cell bodies (asterisks) and nerve ter-minals (dots) at three levels of the lower brain stem (A = P 0.5 mm;B = P 4.5 mm; C = P 7.4 mm) and spinal cord. Note that enkephalin-and substance P-positive cell bodies and nerve terminals are presentin many more areas than indicated on this drawing. AF, anteriorfuniculus; cod, nucleus cochlearis dorsalis; DF, dorsal funiculus; FLM,fasciculus longitudinalis medialis; gr, nucleus gracilis; io, nucleus ol-ivaris inferior; LF, lateral funiculus; LL, lemniscus lateralis; LM,lemniscus medialis; nrv, nucleus reticularis medullae oblongatae parsventralis; nts, nucleus tractus solitarii; ntV, nucleus tractus spinalisnervi trigemini (spinal trigeminal nucleus); ntVd, nucleus tractusspinalis nervi trigemini pars dorsomedialis; nVII, nucleus originisnervi facialis; P, tractus corticospinalis; PCI, pedunculus cerebellarisinferior; PCS, pedunculus cerebellaris superior; rd, nucleus raphedorsalis; rgi, nucleus reticularis gigantocellularis; rl, nucleus reticularislateralis; rm, nucleus raphe magnus; rp, nucleus raphe pallidus; SGC,substantia grisea centralis; TSV, tractus spinalis nervi trigemini; vm,nucleus vestibularis medialis. The nomenclature as well as the crosssections of the lower brain stem are according to Palkovits and Ja-cobowitz (31).

clusions drawn by LaMotte et al. (39). They demonstrated highopiate receptor binding in the superficial laminae of the dor-sal horn and a marked decrease in the binding after dorsalrhizotomy. They suggested presynaptic localization of the op-

iate receptors in primary sensory afferents Enkephalin neuronsmay therefore form axo-axonic synapses (see refs. 26 and 40)on primary afferents, possibly on substance P afferents. Theremarkable overlap of enkephalin and substance P nerve ter-minal fields in the superficial layers of the spinal trigeminalnucleus and in laminae I and II of the dorsal horn is striking.Which central pathways may be involved in the activation

of spinal enkephalinergic neurons? Both the periaqueductalcentral gray and medullary raphe nuclei are effective stimu-lation sites for producing analgesia, suggesting that descendingsystems (41) must be involved. From the periaqueductal centralgray, descending projections to the spinal cord have been de-scribed in the cat (42) but so far not in the rat. From the me-dullary raphe nucleus of the rat, on the other hand, Brodal etal. (43) have demonstrated direct projections to the cord. In acombined neuroanatomical and electrophysiological study,Basbaum et al. (35, 36) analyzed the descending system in-volved in stimulation-produced analgesia (see also ref. 13). Thecell bodies are located in the nucleus raphe magnus and theaxons appear to run in the dorsolateral part of the lateral funi-culus, terminating in the dorsal horn. Basbaum et al. (35, 36)suggested, on the basis of preliminary experimental evidence,that there is a circuitry from the periaqueductal central grayto the spinal cord via the nucleus raphe magnus.With regard to the neurohumoral substrate of the medullary

descending system, both morphological evidence (44) andpharmacological analysis (45, 46) suggest that a 5-hydroxy-tryptamine projection may represent one link between thestimulation site and the spinal cord. The present study dem-onstrates both enkephalin- and substance P-positive cell bodiesin these nuclei. Whereas no evidence for descending enkephalinneurons exist, descending substance P pathways have in factbeen demonstrated (38) and it may be that substance P path-ways represent part of the system related to stimulation-pro-duced analgesia described by Basbaum et al. (35, 36). Becausesubstance P appears to exert an excitatory action (47), it couldactivate the enkephalin interneurons of the dorsal horn. In-terestingly, intraventricular or systemic administration ofsubstance P has in fact been found to produce naloxone-re-versible analgesia (48, 49). 5-Hydroxytryptamine, on the otherhand, appears mainly to be inhibitory and the analgesic effectmediated by descending 5-hydroxytryptamine systems has beenascribed to an axo-axonic influence on primary afferent nerveterminals in the dorsal horn (50).

Primary afferent neurons may also be involved in activationof enkephalin interneurons. However, the type of primary af-ferent neurons activated in classical acupuncture or electro-acupuncture is not known, and we can only speculate aboutthe primary afferent fibers that might activate the enkephalinneurons. There is now strong biochemical (51), neurophysio-logical (47), and immunohistochemical (25) evidence thatsubstance P is present in some primary sensory neurons. Otherprimary afferent neurons may contain a somatostatin-likepeptide (52). The substance P neurons have small cell bodiesand fine-caliber axons with peripheral branches in most tissues.Their central branches terminate mainly in laminae I and IIbut probably also in deeper laminae of the dorsal horn. Thisdistribution and these morphological characteristics are inter-esting in view of the results of Christensen and Perl (53) whofound that cells responding to noxious stimuli are located mainlyin lamina I. Furthermore, they may be considered to add fur-ther support for the view expressed by Henry (54) that sub-stance P "may have an excitatory role in central transmissionin spinal pain pathways acting ... possibly at the first afferentsynapse."

Proc. Nati. Acad. Sci. USA 74 (1977)

Dow

nloa

ded

by g

uest

on

May

17,

202

0

Proc. Natl. Acad. Sci. USA 74 (1977) 3085

The skillful technical assistance of Miss A. Nygards is gratefullyacknowledged. The generous supply of Leu-enkephalin donated byDr. Larry Stein, Wyeth Laboratories, Philadelphia, PA is gratefullyacknowledged.

This work was supported by grants from the Swedish Medical Re-search Council (04X-2887, 04X-04495, 25X-5065, and 04X-2886),Magnus Bergvalls Stiftelse, Sven och Ebba-Christina Hagbergs Stiftelse,and Harald och Greta Jeanssons Stiftelse and by Grant IROI HI18994-01, 555 from the U.S. Department of Health, Education, andWelfare through the National Heart and Lung Institute.The costs of publication of this article were defrayed in part by the

payment of page charges from funds made available to support theresearch which is the subject of the article. This article must thereforebe hereby marked "advertisement" in accordance with 18 U. S. C.§1734 solely to indicate this fact.

1. Reynolds, D. V. (1969) Science 164,444-445.2. Mayer, D. J., Wolfle, T. L., Akil, H., Carder, B. & Liebeskind,

J. C. (1971) Science 174, 1351-1354.3. Schmidek, H., Fohanno D., Ervin, F. R. & Sweet, W. H. (1971)

J. Neurosurg. 35,715-722.4. Melzack, R. & Melinkoff, D. F. (1974) Exp. Neurol. 43, 369-

374.5. Oliveras, J. L., Besson, J. M., Guilbaud, G. & Liebeskind, J. C.

(1974) Exp. Brain Res. 20, 32-44.6. Akil, H., Mayer, D. J. & Liebeskind, J. C. (1975) C. R. Hebd.

Seance Acad. Sci. 274, 3603-3605.7. Oliveras, J. L., Redjemi, F., Guilbaud, G. & Besson, J. M. (1975)

Pain 1, 139-145.8. Adams, J. E. (1976) Pain 2, 161-166.9. Akil, H., Mayer, D. J. & Liebeskind, J. C. (1976) Science 191

961-962.10. Mayer, D. J. & Price, D. D. (1976) Pain 2,379-404.11. Mayer, G. A. & Fields, H. L. (1972) Brain 95, 163-168.12. Andersson, J. A., Eriksson T., Holmgren, E. & Lindqvist, G. (1973)

Brain Res. 63,393-396.13. Jasinski, D. R., Martin, W. R. & Haertzen, C. A. (1967) J. Phar-

macol. Exp. Ther. 157,420-426.14. Sjolund, B. & Eriksson, M. (1976) Lancet ii, 1085.15. Pert, A. & Yaksh, T. (1974) Brain Res. 80, 135-140.16. Mayer, D. J. & Heyes, R. L. (1975) Science 188, 941-943.17. Hughes, J., Smith, T. W., Kosterlitz, H. W., Fothergill, L. A.,

Morgan, B. A. Morris, H. R. (1975) Nature 258,577-579.18. Belluzzi, J. D., Grant, N., Garsky, V., Sarantakis, D., Wise, C. D.

& Stain, L. (1976) Nature 260,625-626.19. Buscher, H. H., Hill, R. C., Romer, D., Cardinauz, F., Closse, A.,

Hauser, D. & Pless, J. (1976) Nature 261, 423-425.20. Elde, R. P., Hokfelt, T., Johansson, 0. & Terenius, L. (1976)

Neuroscience 1, 349-351.21. Hokfelt, T., Elde, R. P., Johansson, O., Terenius, L. & Stein, L.

(1977) Neurosci. Lett., in press.22. Simantov, R., Kuhar, M. J., Uhl, G. R. & Snyder, S. H. (1977) Proc.

Natl. Acad. Sci. USA, 74,2167-2171.23. Euler, U. S. von, & Gaddum, J. H. (1931) J. Physiol. (London)

72,74-87.24. Chang, M. M., Leeman, S. E. & Niall, H. D. (1971) Nature New

Biol. 232, 86-8725. Hokfelt, T., Kellerth, J.-O., Nilsson, G. & Pernow, B. (1975)

Science 190, 889-890.

26. Kerr, F. W. (1975) Pain 1, 325-356.27. Dahlstr6m, A. (1968) in New Aspects of Storage and Release

Mechanisms of Catecholamines, Bayer Symposium II, ed.Schumann, H. V. & Kroneberg, G. (Springer, Berlin), pp. 20-26.

28. Smith, T. W., Hughes, J., Kosterlitz, H. W. & Sosa, R. P. (1976)in Opiates and Endogenous Oplold Peptides, ed. Kosterlitz, H.W. (North Holland, Amsterdam), pp. 57-62.

29. Nilsson, G., Pernow, B., Fischer, G. H. & Folkers, K. (1975) ActaPhysiol. Scand. 94,542-544.

30. Coons, A. H. (1958) in General Cytochemical Methods, ed.Danielli, J. F. (Academic Press, New York), pp. 399-422.

31. Palkovits, M. & Jacobowitz, D. M. (1974) J. Comp. Neurol. 157,29-42.

32. Rexed, B. (1952) J. Comp. Neurol. 96, 415-496.33. Liebeskind, J. C., Guilbaud, G., Besson, J. M. & Oliveras, J. L.

(1973) Brain Res. 50, 441-446.34. Proudfit, H. K. & Anderson, E. G. (1975) Brain Res. 98, 612-

618.35. Basbaum, A. I., Clanton, C. H. & Fields, H. L. (1976) Proc. Natl.

Acad. Sci. USA 73,4685-4688.36. Basbaum, A. I., Marley, N. J. E., O'Keefe, J. & Clanton, C. H.

(1977) Pain 3, 43-56.37. H6kfelt, T., Elde, R. P.,Johansson, O., Ljungdahl, A., Schultzberg,

M., Fuxe, K., Goldstein, M., Nilsson, G., Pernow, B., Terenius,L., Ganten, D., Jeffcoate, S. L., Rehfeld, J. & Said, S. (1977) inAmerican College of Neuropsychopharmacology, eds. Lipton,M., Aghajanian, G. & Bloom, F. (Raven Press, New York), inpress.

38. Sjolund, B., Terenius, L. & Eriksson, M. (1977) Acta Physiol.Scand., in press.

39. LaMotte, C., Pert, C. B. & Snyder, S. H. (1976) Brain Res. 112,407-412.

40. Rethelyi, M. & Szentagothai, J. (1973) in Handbook of SensoryPhysiology, ed. Iggo, A. (Springer, Berlin, Heidelberg, NewYork). Vol. II, pp. 207-252.

41. Holmqvist, B., Lundberg, A. & Oscarsson, 0. (1960) Arch. Ital.Biol. 98, 60-80.

42. Kuypers, H. G. J. M. & Maisky, V. (1975) Neurosci. Lett. 1,9-14.

43. Brodal, A., Taber, E. & Walberg, F. (1960) J. Comp. Neurol. 114,239-260.

44. Dahlstrom, A. & Fuxe, K. (1965) Acta Physiol. Scand. Suppl. 64,247,5-36.

45. Akil, H. & Mayer, D. J. (1972) Brain Res. 44,692-697.46. Guilbaud, G., Besson, J. M., Oliveras, J. L. & Liebeskind, J. C.

(1973) Brain Res. 61, 417-422.47. Konishi, S. & Otsuka, M. (1974) Brain Res. 65,397-410.48. Stewart, J. M., Getto, C. J., Neldner, K., Reeve, E. B., Krivoy, W.

A. & Zimmermann, E. (1976) Nature 262,784-785.49. Malick, J. B. & Goldstein, J. M. (1977) Fed. Proc. 36,994.50. Proudfit, H. K. & Anderson, E. G. (1973) Brain Res. 65, 542-

546.51. Takahashi, T. & Otsuka, M. (1975) Brain Res. 87, 1-11.52. Hokfelt, T., Elde, R. P., Johansson, O., Luft, R., Nilsson, G. &

Arimura, A. (1976) Neuroscience 1, 131-136.53. Christensen, B. N. & Perl, E. R. (1970) J. Neurophysiol. 33,

293-307.54. Henry, J. L. (1976) Brain Res. 114, 439-451.

Neurobiology: H6kfelt et al.

Dow

nloa

ded

by g

uest

on

May

17,

202

0