Periaqueductal gray area and cardiovascular function

PERIAQUEDUCTAL

FRANCESCO Institute

GRAY AREA AND CARDIOVASCULAR FUNCTION

ROSSI,

SABATINO

MAIONE

and LIBERATO

BERRINO

oj Pharmucology and Toxicology, Fuculty of Medicine and Surgery, I1 Uni\*ersity of Nuples, L>iaBroggia 3, 80138 Naples, Italy Ruc,ril,ed in firm/ fom 9 Septemher

1993

SUMMARY The periaqueductal gray (PAG) area seems to play an important role in modulating several biological functions such as the triggering of stereotyped defence and reproductive behaviour, pain, anxiety and cardiovascular and respiratory activities. Anatomically this midbrain area is made up of symmetric neuronal columns arranged along the long axis of the aqueduct. In this paper we review the most important findings of the last IO-15 years about the interaction between the PAG area and the cardiovascular function. It is shown that these neuronal columns within the PAG area exhibit a viscerotropic organization which elicits both hypertensive and hypotensive responses. In particular, the stimulation of the ventral neuronal column evokes a hypotensive response associated with a regional decrease in the vascular resistance. On the contrary, the stimulation of the dorsal and lateral neuronal columns evokes arterial hypertension associated with specific changes of the vascular resistance. Recently the authors demonstrated that the glutamergic system in the PAG area (prevalently through NMDA subtype receptor) may also be involved in the control of cardiovascular system. Moreover, the involvement of the arginine vasopressin neuropeptide in the hypertension induced by administration of excitatory amino acids into the PAG area has been demonstrated. KEYWOWS:periaqucductal

gray

area,

cardiovascular

function,

NMDA,

rat.

INTRODUCTION The periaqueductal gray (PAG) area, or as originally defined the central gray area, anatomically represents a relatively homogeneous midbrain neural density that wraps the aqueduct along its entire length. Currently the PAG area is defined as being made up of neural cell columns projecting along the aqueduct in a rostrocaudal direction (Fig. 1). On each side of the midbrain there are dorsomedial, dorsolateral, lateral and ventrolateral cell columns. Other neural nuclei found along the PAG area (dorsal raphe nucleus, the I o-13 66 I X/%/O IOO27- 10/$08.00/0

01994 The Italian Pharmacological

Society

28

Phamucdo~it~al

Research,

Vol. 29, No. I,

1994

nucleus belonging to the III cranial nerve, the troclear nerve nucleus and the Edinger-Westphal nucleus), however, are not regarded belonging to this area. Neurophysiological and neuroanatomical studies in relatively recent years have been of fundamental importance in demonstrating the existence of these PAG area cell columns and have shed light on the regional functional differences, often overlapping among each other. In the last lo-15 years there have been above all two features of this area that have solicited interest among scholars: nociception control [l-8] and modulation of behaviour modifications, currently termed defensive behaviour [9-161. In fact, both electrical and chemical stimulation of the PAG area, as shown in a number of studies, lead surprisingly to sustained reduction of pain sensitivity as well as to aggressive behaviour subjecting the animal to fight and/or flight with typical voice modifications [17, 181. It is currently believed that both the flight response and the nociception modulation are two aspects of the same phenomenon since it is assumed that the PAG area is an organ that integrates more than one biological function and, therefore, has a role in coordinating responses that assure survival [9-13, 191. More recent studies seem to show that besides the above mentioned functions, the PAG area integrates two other important vital functions: the cardiovascular and the respiratory [20-261. For these two neurovegetative functions a specific anatomo-functional organization along the neural cell columns is emerging. As for the cardiovascular function it has been demonstrated that neurons, along the lateral and latero-ventral columns, selectively and in a rostrocaudal direction control specific zones of the vascular bed. These studies have, therefore, documented a regional modulation of the vascular tone by the PAG area [23,24,27-301. The aim of this review is to illustrate the major contributions, although few, in the literature of the last lo-15 years relative to the modulation of the cardiovascular function by the PAG area. Reference is made to current knowledge of midbrain neuroanatomy as this forms the basis of our understanding of the effects that this area has on vascular tone.

-6.0

-7.6

-8.0

Fig. 1. Schematic representation of four coronal sections of the rat PAG area showing the position of the neuronal columns indicated as dorso-lateral column (D), lateral column (L), ventrolateral column (V) and dorso-medial column (DM). Other indications are: (III) oculomotor nucleus and (IV) trochlear nucleus. The coordinates of the atlas of Paxinos and Watson [54] (measured from the bregma: -6.0, -7.6, -8.0, -8.3) were applied.

pharmacological Research, Vol. 29, No. I( 1994

THE PAG AREA AS A MIDBRAIN

29

VASOMOTOR

CENTRE

Let us briefly review the events leading up to the identification of the PAG area as a centre involved in controlling cardiovascular function [29]. The first to demonstrate that electrical stimulation of the central gray area in cats was able to induce arterial hypertension were Kabat et al. in 1935 [3 11. It was not until the end of the 1950s that more detailed findings emerged relating to regional changes in blood flow due to stimulation of specific zones of the PAG area. The Swedish physiologists Eliasson and Lindgren were the first to measure regional blood flow [32-341 after central electrical stimulation. These authors showed that PAG areainduced hypertension is coupled with a specific change in blood flow; vasodilatation of the muscle vascular bed in the hindlimb and vasoconstriction in the skin and organs; although they did not explain why there was such a functional pattern. The first to develop a more precise hypothesis to explain the particular change in blood flow following PAG area stimulation was the group led by Abrahams in 1960 [35]. In fact, using freely moving cats Abrahams et al. demonstrated that such cardiovascular effects are also surprisingly coupled with behavioural changes of alertness and fighting and that the major blood flow is diverted to the striated muscles instead of the internal organs and skin. At this point it was not difficult for Abrahams et al. to hypothesize that such a pattern could only be explained by an increased metabolic requirement of the highest stressed areas during fight or flight. Between the end of the 1970s and early 1980s an important step forward was made thanks to two fundamental technical events: 1) substitution of electrical stimulation (aspecific because not only do local nerves get depolarized but also passing axons) in the brain with more precise and selective stimulation carried out with glass micropipettes charged with excitatory amino acid solutions [36]; 2) the surgical sectioning of the brain at the precolliculus which allowed use of decerebrated rather than anesthetized animals, thus making locomotive as well as cardiovascular findings possible. Afterwards, by the second half of the 1980s Carrive and Bandler [20-22, 27,281 demonstrated that the PAG area not only modulated the induction effects such as hypertension followed by tachycardia, but when the same area of the midbrain was stimulated more ventrally there was an immediate drop in blood pressure coupled with sustained and long-term bradycardia. As can be seen in Fig. 2, the PAG area section, more specifically involved in inducing the increased arterial blood pressure, is a small portion lateral to the aqueduct, while on the other hand, it is the ventrolateral part of the periaqueductal gray area that modulates hypotensive events. Figure 2 shows that such specific sites involved in the control of both hypertension and hypotension appear to be distributed along the axis parallel to the aqueduct. In other words, it appears that the lateral neural cell columns could modulate hypertension, whereas the ventrolateral ones could induce hypotension. These first observations suggest that neural cell columns, anatomically well defined, specifically modulate different cardiovascular functions and especially arterial blood pressure.

-6.0

0

HYPOTENSIVE sites i left) between 5% and 25% decrease in arterial pressure

? ?HYPERTENSI ,.^^.

E sites (right between ZO% and 504 increase in arterial pressure

1

Fig. 2. Schematic representation of four coronal sections of the rat PAG area showing the position of the injection and/or stimulation zones which generate hypertension or hypotension. The coordinates of the atlas of Paxinos and Watson [54] (measured from the bregma: -6.0, -7.6, -8.0, -8.3) were applied.

Functional organization of dorsal crnd r’entral PAG urea In regard to the hypertensive effects due to the neurons in lateral cell columns, Carrive et ul. [20, 211 showed that iliac artery tone together with its branches and those from the renal artery are modulated differently depending on whether neurons from the rostra1 or caudal area of the PAG area are stimulated. In fact, D,L-homocysteic acid (DLH) administration into the 3rd anterior portion of the lateral cell column induces after a few seconds a significant constriction of the iliac artery and its branches and no significant renal vasoconstriction. The contrary occurs when the neurons to be stimulated are more caudal along the same neural cell column. Administering DLH in the middle third and caudal portions of this lateral cell column induces an increased blood flow within the iliac artery coupled with renal artery constriction. While the lateral neural cell column arrangement of the PAG area was being demonstrated, Carrive and Bandler [281 demonstrated that a similar functional organization was true also for the ventrolateral neural cell column that modulates hypotension. In fact, two types of ventrolateral PAG areamediated hypotension can be evoked: one featuring reduced vascular tone along the iliac artery after rostra1 neurons stimulation, the other featuring reduced renal vascular tone with caudal neuron stimulation.

the sub-retrofacial (SRF) nucleus (also identified as the Currently rostroventrolateral area of the bulb) is attributed with the primary role of sympathetic tone pacemaker [37401. In fact, provided that spinal column intermedial-lateral neurons (IML) make up the sympathetic centre which functionally and anatomically is more directly coupled with peripheral organs, one must, however. remember that the IML alone is unable to modulate and maintain sympathetic outflow. In contrast, the SRF nucleus is the switchboard for a large quantity of inputs arising both from other nuclei in the CNS directly connected to cardiovascular function (hypothalamus, PAG area, CVLM, NTS, area postrema. Raphe nuclei) as well from extravascular sources.

pharmawlo,~id

Rcwtrr-ch,

Vol. 29. No. I. 1094

31

A fundamental contribution to neuroanatomical understanding, demonstrating the real link between the PAG area and SRF nucleus with other important bulbar structures in modulating cardiovascular function, comes from almost contemporary studies by Holstege and Luiten [41, 421. These studies clearly show that there is an important descending pathway arising directly from the PAG area and leading to the SRF nucleus. Carrive et al. demonstrated that such fibres originate primarily in lateral and ventrolateral cell columns [21, 27,281. This group of fibres is without decussation and offers inputs not only to the SRF, but to other brainstem nuclei with cardiovascular involvement as well, such as NTS and CVLM. et al. [27] documented that PAG columns (lateral and Besides, Carrive ventrolateral) have a common rostrocaudal viscerotopic organization and that the projections from PAG go to SRF with opposite direction (the rostra1 part of PAG columns projects to the caudal part of SRF; the caudal part of PAG columns projects to the rostra1 part of SRF). This scheme of connections is in agreement with the viscerotopic organization of SRF demonstrated by Lovick [24]. She, in fact, demonstrated that chemical stimulation of the SRF caudal area with DLH was able to induce arterial hypertension coupled with a sustained vasoconstriction of the external iliac artery, while stimulation of the more rostra1 SRF zone generated arterial hypertension with a sustained vasoconstriction of the renal artery. In addition, other more recent studies by Bandler and Carrive in 1991 [ 12, 28, 301 demonstrated that these two vascular beds (renal and iliac) are not the only ones directly affected by the PAG area. The external carotid artery is also influenced by lateral and ventrolateral cell column neurons. DLH applied to the caudal part of the lateral PAG area determines a marked vasoconstriction of the extracranial arterial bed, while administrating this excitatory amino acid in the rostra1 area of the lateral PAG causes vasodilatation. According to these results Carrive et al. [29, 301 realised the existence of a vasodilator column in the lateral PAG area. The viscerotopic organization of this column was the opposite of that of the vasoconstrictor column: the more rostra1 part of this column would cause vasodilatation of head vessels, while neurons that cause dilatation of the posterior limbs vessels lie more caudally. The pathways of the extracranial vasodilatation probably start in the lateral PAG and go to the parasympathetic preganglionic neurons of the pterygopalatine ganglion [29,43]. A cholinergic component (postganglionic cholinergic vasodilator fibres) may be considered to explain the vasodilator effect occurring in the iliac artery since systemic pretreatment with atropine is able to significantly reduce the initial phase of vasodilatation. On the other hand, since propranolol has instead been shown to be very effective in reducing the late phase of the iliac’ artery vasodilatation it is clear that the sympathetic nervous system (mostly through /I receptor mediation) can also play a role in generating iliac vasodilatation mediated by the vasodilator column of the PAG area [23, 24, 29, 33, 341. PAG urea neurotransmitters

imwlvement

in cardiovascular

regulation

It has recently been demonstrated that besides the important autonomic nervous system, the glutamergic system modulating

role played nociception

by the in the

Pharmacological Research, Vol. 29, No. I,

32

7 q 40 z

-

-

I994

GAMS CGP 25838

M 30 2 E

20

2 4

10 0

0.05

0.1

1.0 0.5 L-glu (kg/rat icv)

5.0

Fig. 3. Arterial blood pressure (ABP, mmHgfs.E.) changes in freely moving rats in response to microinjections of L-glutamic acid (L-glu, from 0.05 to 5 pg per rat) into the PAG area, pretreated or not, 5 min before, with 2-APV (1 pg per rat), GAMS (1 pg per rat) or CGP 25838 (1 ,ug per rat). Each point represents the mean+s.e. of 5 observations. *P
PAG area, may also be involved in cardiovascular changes that have their origins in this area [44]. In particular, the findings show that at this level the NMDA receptors are more involved than the non-NMDA ones in cardiovascular homeostasis. In fact, arterial hypertension induced by L-glutamic acid in the PAG area has been significantly reduced by a pretreatment with 2-amino-5phosphono valeric acid (2-APV) (a selective antagonist of NMDA receptors), but not by a pretreatment with glutamilaminomethyl-sulphonic acid (GAMS) (a selective antagonist of non-NMDA receptors) (Fig. 3). In addition, since it is known that NMDA receptors are modulated, to different degrees, by Mg2+, Zn2+, H+, polyamine, glycine and neurosteroids [45-501, a further study evaluated whether glycine [46] modulated NMDA receptors in viva within the PAG area. In agreement with the literature, the findings confirm the positive modulating role of glycine on NMDA receptors. They also demonstrate that there are NMDA receptors in the PAG area that regulate cardiovascular function and are sensitive to glycine modulation (Fig. 4). Glycine alone does not significantly modify This indicates that such inhibitory amino acid cardiovascular function. neurotransmitter is of little importance in modulating cardiovascular function in the PAG area [46]. In contrast, GABAergic neurotransmission, mostly consisting of short interneurons, has a clearer and more decisive action in negative modulation of all functions controlled by the PAG area, and therefore also cardiovascular ones [ 19, 401. As for the vascular tone, GABAergic modulation lowers the degree of lateral and lateroventral cell column excitation in neurons that more directly influence firing of the bulbar SRF nucleus [ 19,401. Serotonin is another neurotransmitter that has been shown to inhibit cardiovascular function. This molecule is widely available within the midbrain. The hypotension following stimulation of specific sites within the ventrolateral

pharmacological Research, Vol. 29, No. 1, 1994

? ?NMDA+GLY ? ?NMDA+HA

33

10 10

NMDA (pg/rat) Fig. 4. Arterial blood pressure (ABP, mmHgfs.!z.) changes in freely moving rats in response to microinjections of NMDA (from 0.01 to 1 pug per rat) into the PAG area, pretreated or not, 5 min before, with glycine (GLY, 10 pg per rat). The pretreatment with HA-966 (HA, 10 ,ug per rat), a selective antagonist of the glycine site on NMDA receptors, significantly prevented the glycine-induced increase of NMDA hypertension. *P
columns of the PAG area could, in fact, be due to activation of inhibitory efferent serotoninergic fibres that directly synapse with bulb vasomotor centres (CVLM, area postrema, NTS, SRF) [51]. This could be the consequence of excitatory pathway activation with fibres specifically projected to raphe nuclei (serotoninergic and inhibitory) which in turn innervate the SRF nucleus [23]. A brief word also needs to be said about some neuropeptides such as pendorphin, enkephalin [52] and substance P [53]. Although information regarding their cardiovascular regulation within the PAG area is scarce, it is well known that these same neuropeptides are very important in this cerebral area because they induce specific stereotyped behaviour and modulate nociception. In addition, we recently demonstrated that arginine vasopressin (AVP) can modulate cardiovascular function centrally of the PAG area. Intracerebral pretreatment with l-(~-mercapto-~,~-cyclo-pentamethylenepropionic)-2-(O-methyl)tyrosine (CGP 25838), a selective antagonist of V, receptors of AVP, significantly reduces glutamic acid pressor response (Fig. 3). This shows that AVP could be an important neuropeptide involved in generating the hypertension that follows glutamergic receptor activation [44]. In agreement with the literature, our findings confirm that besides being a hormone involved in the regulation of the electrolyte balance through V, receptors along the renal tubule, AVP can also behave as a neurotransmitter and/or a neuromodulator in the PAG area. cell

CONCLUSIONS It may be said that the PAG

are symmetrically

area is made

up of a

placed along the aqueduct.

series of neural cell columns that

In any discussion about the PAG area, speaking only of one of the biological functions that it modifies (such as cardiovascular function) is complicated and artificial. In fact, it is now evident that the PAG area has a specific role of integrating manifold biological functions and not just one. The PAG area is thought of as a primary centre of data processing and coordination of responses needed for survival. In fact, besides the cardiovascular functions heretofore discussed, the PAG area influences respiration and other specific stereotypical behaviour such as reproduction, nociception, speech, mood and flight response. The PAG area, like other parts of the CNS such as the hypothalamus, is specialized to receive and decode inputs from the CNS and other organs and it is, therefore, the switchboard for important vital functions needed to ensure survival of the individual and species.

REFERENCES 1. Fardin V, Oliveras JL, Besson JN. A reinvestigation of the analgesic effects induced by stimulation of the periaqueductal gray matter in the rat. II. Differential characteristics of the analgesia induced by ventral and dorsal PAG stimulation. Brain RES 1984; 306: 125-39. 2. Hosobuchi Y, Rossier J, Bloom FE, Guillemin R. Stimulation of human periaqueductal gray for pain relief increases immunoreactive P-endorphin in ventricular fluid. Science 1979; 203: 279-81. 3. Jensen TS, Yaksh TL. The antinociceptive activity of excitatory amino acids in the rat brainstem: an anatomical and pharmacological analysis. Brain Res 1992; 569: 255-67. 4. Liebeskind JC, Guilbaud G, Besson JM, Oliveras JL. Analgesia from electrical stimulation of the periaqueductal gray matter in the cats: behavioural observations and inhibitory effects on spinal cord interneurons. Brain Res 1973; 50: 441-6. 5. Oliveras JL, Besson JM, Guilbaud G, Liebeskind JC. Behavioural and electrophysiological evidence of pain inhibition from midbrain stimulation in the cat. Exp Brain Res 1974; 20: 32-44. 6. Oliveras JL, Hosobuchi Y, Bruxelle J, Passot C, Besson JM. Analgesic effects induced by electrical stimulation of the nucleus raphe magnus in the rat: interaction with morphin analgesia. Abstract 7th International Congress of Pharmacology Paris, Vol. 1, 1978, pp. 119. 7. Oliveras JL, Guilbaud G, Besson JM. A map of serotoninergic structures involved in stimulation producing analgesia in unrestrained, freely moving cats. Brain Res 1979; 164: 3 17-22. 8. Oliveras JL, Besson JM. Stimulation induced analgesia in animals: behavioural investigations. Prog Brain Res 1988; 77: 141-57. 9. Bandler R, Depaulis A, Vergnes M. Identification of midbrain neurons mediating defense behaviour in the rat by microinjections of excitatory amino acids. Bdmv Brain Res 1985; 15: 107-19. 10. Bandler R, Carrive P. Integrated defence reaction elicited by excitatory amino acids injections in the midbrain periaqueductal gray region of the unrestrained cat. Brain Res 1988; 439: 95-106. 1 I. Bandler R, Depaulis A. Elicitation of intraspecific defense reactions in the rat from midbrain periaqueductal gray by microinjection of kainic acid, without neurotoxic effects. Neurosci Lett 1988; 88: 291-6. 12. Bandler R, Carrive P, Zhang SP. Integration of somatic and autonomic reactions within the midbrain periaqueductal gray: viscerotopic, somatopic and functional organization. P rag Brain Res 199 1; 87: 269-305. 13. Depaulis A, Bandler R, Vergnes M. Characterization of pretentorial periaqueductal gray

14. IS.

16. 17. 18 19.

20.

21.

22.

23. 24. 2s. 26. 27.

28.

29.

30. 31. 32. 33. 34. 35.

neurons mediating intraspecific defensive behaviour in the rat by microinjections of kainic acid. RI-crin Res 1989; 486: I2 l-32. Fernandez De Molina A, Hunsperger RW. Organization of the subcortical system governing defence and flight reactions in the cat. / Physiol (Land.) 1962; 160: 200-13. Keay KA, Depaulis A, Breakspear MJ, Bandler R. Pre- and subtentorial periaqueductal gray of the rat mediates different defense responses associated with hypertension. Nrut-osci Snc Ahst 1990; 16: 598. Krieger JE, Graeff FG. Defensive behaviour and hypertension induced by glutamate in the midbrain central gray of the rat. Brazil J Med Biol Res 1985; 18: 61-67. Larson CR, Kistler MK. Periaqueductal gray neuronal activity associated with laryngeal EMG and vocalization in the awake monkey. Neurosci Lett 1984; 46: 261-6. Larson CR, Kistler MK. The relationship of periaqueductal gray neurons to vocalization in laryngeal EMG in behaving monkey. Exp Bruin Res 1986; 63: 596-606. Depaulis A, Vergnes M. Elicitation of intraspecific defensive behaviours in the rat by microinjections of picrotoxin, a GABA antagonist, into the midbrain periaqueductal gray matter. Bruin Res 1986; 367: 87-95. Carrive P, Dampney RAL, Bandler R. Excitation of neurons in a restricted portion of the midbrain periaqueductal gray elicits both behavioural and cardiovascular components of the defence reaction in the unanaesthetized decerebrated cat. Neurosci Lett 1987; 81: 273-8. Carrive P, Bandler R, Dampney RAL. Anatomical evidence that hypertension associated with the defence reaction in the cat is mediated by direct projection from a restricted portion of the midbrain periaqueductal gray to the subretrofacial nucleus of the medulla. Brain Res 1988; 460: 339-45. Carrive P, Bandler R, Dampney RAL. Somatic and autonomic integration in the midbrain of the unanaesthetized decerebrated cat: a discriminative pattern evoked by excitation of neurons in the subtentorial portion of the midbrain periaqueductal gray. Bruin Res 1989; 483: 25 l-8. Lovick TA. Projections from the brainstem nuclei to nucleus paragigantocellularis lateralis in the cat. J Auton Nerv Syst 1986; 16: l-l I. Lovick TA. Differential control of cardiac and vasomotor activity by neurons in nucleus paragigantocellularis lateralis in the cat. J Physiol (Lond.) 1987; 389: 23-35. Lovick TA. Inhibitory modulation of the cardiovascular defence response by the ventrolateral periaqueductal gray matter in rats. Exp Bruin Res 1992; 89: 133-9. Lovick TA. Midbrain influences on ventrolateral medullo-spinal neurones in the rat. Ex~ Brain Res 1992; 90: 147-52. Carrive P, Bandler R, Dampney RA. Viscerotopic control of regional beds by discrete groups of neurons within the midbrain periaqueductal gray. Bruin Res 1989; 493: 385-90. Carrive P, Bandler R. Viscerotopic organization of neurons subserving hypotensive reactions within the midbrain periaqueductal gray: a correlative functional and anatomical study. Bruin Res 1991; 541: 206-15. Carrive P. Functional organization of PAG neurons controlling regional vascular beds. In: Depaulis A & Bandler R, eds. The midbrain periaqueductal gray mutter-functional, anatomical, and neurochemical orgukation. New York; Plenum Press, 1991; 67-100. Carrive P, Bandler R. Control of extracranial and hindlimb blood flow by the midbrain periaqueductal gray of the cat. Exp Brain Res 1991; 84: 599-606. Kabat H. Magoun HW, Ranson JW. Electrical stimulation of points in the forebrain and midbrain: the resultant alterations in blood pressure. Arch Neural 1935; 34: 931-55. Elliason S, Lindgren P, Uvnas B. The hypothalamus, a relay station of the sympathetic vasodilator tract. Actor P/1ysiol Stand 1954; 31: 290-300. Lindgren P. The mesencephalon and the vasomotor system. Acta Physiol Stand 1955; 35 (Suppl. 121): I-183. Lindgren P, Rosen A, Strandberg P, Uvnas B. The sympathetic vasodilator outflow: a cortico-spinal autonomic pathway. .I Camp Neural 1956; 105: 95-109. Abrahams VC. Hilton SM, Zbrozyna AW. Active muscle vasodilatation produced by

36

36.

37. 38.

39.

40.

41. 42.

43.

44.

45. 46. 47. 48. 49.

50. 51. 52.

53.

54.

Pharmacological Research, Vol. 29. No. I. 1994

stimulation of the brain stem: its significance in the defence reaction. .I Physiol (Lond.) 1960; 154: 491-513. Goodchild AK, Dampney RAL, Bandler R. A method of evoking physiological responses by stimulation of cell bodies but not axons of passage within localized regions of the nervous system. J Neurosci Meth 1982; 6: 351-63. Ciriello J, Caverson MM, Polosa C. Function of the ventrolateral medulla in the control of the circulation. Brain Res Rev 1986; 11: 359-91. Dampney RAL, Czachurski J, Dembowsky K, Goodchild AK, Seller H. Afferent connections and spinal projections of the pressor region in the rostra1 ventrolateral medulla of the cat. J Auton Nerv Syst 1987; 20: 73-86. Ross CA, Ruggiero DA, Joh TH, Park DH, Reis DJ. Rostra1 ventrolateral medulla: selective projections to toracic autonomic cell column from the region containing Cl adrenalin neurons. J Comp Neural 1984; 228: 168-85. Strack AM, Sawyer WB, Hughes JH, Platt KB, Loewy AD. A general pattern of CNS innervation in the sympathetic outflow demonstrated by transneuronal pseudorabbits viral infection. Bruin Res 1989; 491: 156-62. Holstege G. Direct and indirect pathways to lamina I in the medulla oblongata and spinal cord of the cat. Prog Bruin Res 1988; 77: 47-94. Luiten PGM, ter Horst GJ, Steffens AB. The hypothalamus, intrinsic connections and outflow pathways to the endocrine system in relation to the control of feeding and metabolism. Prog Neurol 1987; 28: l-54. Spencer SE, Sawyer WB, Wada H, Platt KB, Loewy AD. CNS projections to the pterygopalatine parasympathetic preganglionic neurons in the rat: a retrograde transneuronal viral cell body labeling study. Bruin Res 1990; 534: 149-69. Maione S, Berrino L, Vitagliano S, Leyva J, Rossi F. Interactive role of L-glutamate at the level of the PAG area, for cardiovascular tone and stereotyped behaviour. Bruin Res 1992; 597: 166-9. Asher P, Nowak L. The role of divalent cations in the N-methyl-D-aspartate response of mouse central neurons in culture. J Physiol (Lond.) 1988; 399: 247-66. Berrino L, Vitagliano S, Maione S, Rossi F. Modulation by glycine on vascular effect of NMDA: in vivo experimental research. Amino Acids 1993; 4: 127-32. Johnson JW, Asher P. Glycine potentiates the NMDA responses in cultured mouse brain neurons. Nature (Lond.) 1987; 325: 529-31. Maione S, Berrino L, Vitagliano S, Leyva J, Rossi F. Pregnenolone sulfate increases the convulsant potency of NMDA in mice. Eur J Pharmacol 1992; 219: 477-9. Ransom RW, Stek NL. Cooperative modulation of [3H]MK801 binding to the N-methylD-aspartate receptor-ion channel complex by L-glutamate, glycine and polyamines. J Neurochem 1988; 51: 830-6. Traynelis SF, Cull-Candy SG. Proton inhibition of N-methyl-D-aspartate receptors in cerebellar neurons. Nature (Lond.) 1990; 345: 347-50. Nicholas AP, Hancock MB. Evidence for projections from the rostra1 medullary raphe onto medullary catecholamine neurons in the rat. Neurosci Lett 1990; 108: 22-8. Ma QP, Shi YS, Han JS. Naloxone blocks opioid peptide release in nucleus accumbens and amygdala elicited by morphine injected into periaqueductal gray. Bruin Res Bull 1992; 28: 351-4. Li YQ, Zeng SL, Rao ZR, Shi JW. Serotonin-, substance P- and tyrosine hydroxylaselike immunoreactive neurons projecting from the midbrain periaqueductal gray to the nucleus tractus solitarii in the rat. Neurosci Lett 1992; 134: 175-9. Paxinos G, Watson C. The rut bruin in the stereotuxis coordinates, 2nd edn. San Diego: Academic Press, 1986.