Bram Research, 582 (1992) 196-207 © 1992 Elsevier Science Pubhshers B.V All rights reserved 0006-8993/92/$05.00

196

BRES 17788

Characteristics of caudal ventrolateral medullary neurons antidromically activated from rostral ventrolateral medulla in the rabbit Z.J. Gieroba, Y.-W. Li* and W.W. Blessing Department of Medtcme and Physiology, Centre for Neuroscience, Fhnders Umverslty of South Austraha, Bedford Park, SA (Austraha) (Accepted 21 January 1992)

Key words. Caudal ventrolateral medulla, Rostral ventrolateral medulla, Vasodepressor neuron, Al-catecholamine synthesizing neuron. Electrophyslology

We made extracellular recordings from 107 spontaneously active neurons m the caudal ventrolateral medulla, after identifying the cells by antidromlcally activating them from the rostral ventrolateral medulla, in urethane-anesthetlzed rabbits. We tested the response of these neurons to inputs from baroreceptors and chemoreceptors The median conduction velocity for antidromically activated neurons was 0 84 m/s. Raising blood pressure with intravenous noradrenahne excited 22% of 96 neurons tested, inhibited 61%, and had no effect on the remaining 17% The spontaneous discharge rate of neurons excited by an increase in blood pressure was 1.6 _+ 0 3 spikes/s, lower than the discharge rate of neurons inhibited by this procedure (4 9 + 0.5 spikes/s) Excitation of chemoreceptors by hypoxla Increased the discharge rate of 14/16 neurons tested m the group exoted by baroreceptor inputs In the group inhibited by baroreceptor inputs 21/35 neurons tested were excited and 12/35 neurons were inhibited by chemoreceptor inputs. Neurons excited by an increase m blood pressure were located m the previously defined caudal vasodepressor region and in a region just rostral to the obex, intermediate between the vasodepressor region and the rostral sympathoexotatory region. These neurons may form part of the central inhibitory link m the baroreceptor-vasomotor pathway Other antldromlcally activated neurons in the vasodepressor region may be inhibitory vasomotor cells with a function relatively independent of baroreceptor inputs, or they may be A1 catecholamlne neurons, with axons passing through the rostral medulla en route to the forebram INTRODUCTION

r e c e n t l y identified in rabbit 38 and rat 1. T h e n e u r o n s w e r e f o u n d to be excited by activation of p e r i p h e r a l b a r o r e -

A c t i v a t i o n of n e u r o n s in the caudal v e n t r o l a t e r a l medulla ( C V L M ) r e d u c e s s y m p a t h e t i c v a s o m o t o r t o n e 9"41

c e p t o r s , and by electrical s t i m u l a t i o n of the aortic de-

by an action largely d e p e n d e n t on the functional integ-

ceptor-related

rity o f s y m p a t h o e x c i t a t o r y n e u r o n s in the rostral v e n t r o lateral m e d u l l a ( R V L M ) 2'6'27"31'42. T h e C V L M v a s o d e -

sympathoexcitatory

rabbit, was u n d e r t a k e n to study a larger p o p u l a t i o n of

p r e s s o r n e u r o n s a p p e a r to be tonically active, since their

a n t i d r o m i c a l l y a c t i v a t e d C V L M n e u r o n s , and to test their

inhibition is itself associated with an increase in b l o o d

r e s p o n s e s to p e r i p h e r a l activation o f b a r o r e c e p t o r s and

pressure. T h e v a s o d e p r e s s o r n e u r o n s w e r e t h o u g h t to b e l o n g to

chemoreceptors.

the p o p u l a t i o n of A 1 c a t e c h o l a m i n e - s y n t h e s i z i n g n e u rons, also situated in the C V L M 9A8'19. H o w e v e r neu-

MATERIALS AND METHODS

r o a n a t o m i c a l studies indicate that axons of A1 n e u r o n s

Successful experiments were performecl on 41 male, New Zealand White rabbits (2.5-3 0 kg), anesthetized with urethane (1 5 g/kg, Sigma Chemical Co) infused into a marginal ear vein over 30 min. Scopolamine metbylbromide (50 /~g/kg, Sigma Chemical Co) was gwen to minimize airway secretions. The trachea was cannulated, the animal was paralysed with pancuromum bromide (0.5 mg/kg, i.v initially, with supplemental doses as necessary) and artificially ventilated with oxygen enriched air, using a Harvard model 681 rodent ventilator. The end-expiratory CO 2 was monitored (Datex Normocap CO 2 monitor, Helslnkl, Finland) and maintained at 35-40 mmHg A bilateral pneumothorax was Induced Body tern-

do not synapse in the R V L M 7'36, a l t h o u g h t h e i r axons t r a v e r s e this a r e a as t h e y ascend to the h y p o t h a l a m u s 34' 3s. O t h e r n e u r o n s in the C V L M , including 7 - a m i n o b u tyric acid ( G A B A ) - s y n t h e s i z i n g cells, do h a v e axons term i n a t i n g in the R V L M 5. To identify the v a s o d e p r e s s o r n e u r o n s e l e c t r o p h y s i o l o g i c a l l y we a n t i d r o m i c a l l y activ a t e d t h e m f r o m the R V L M . S a m p l e s of t h e s e cells w e r e

p r e s s o r n e r v e , in a g r e e m e n t with t h e i r p o s t u l a t e d b a r o r e inhibitory

action

on

the

rostral

cells. T h e p r e s e n t p r o j e c t , in the

General procedures

* Present address: Department of Physiology, F13, University of Sydney, NSW 2006, Australia Correspondence Z J Gleroba, Department of Physiology, Hinders Medical Centre, Bedford Park, SA 5042, Australia. Fax (61) (8) 204 5450.

197 perature was momtored by a rectal thermistor probe and maintained at 38-39°C by a heating pad. A polyethylene catheter was inserted m the marginal ear vein for administration of drugs and into the left femoral artery for recording of artenal pressure (AP) and samphng artenal blood for blood gas and pH analysis. For admlmstratmn of sodmm cyanide, a polyvmyl catheter (PV 45) was inserted through the superior thyrotd artery and the tip was posluoned in the common carotid artery, close to carotid bifurcation A Statham P23 ID strain gauge transducer connected to a Grass model 7 polygraph was used to record AP Mean AP was obtained by filtering the phasic signal Heart rate was computed with a Grass 7P4F tachograph triggered by the phasic arterial signal. The head of the rabbit was fixed m a Kopf stereotaxlc frame and the medulla was exposed by recision and retraction of the atlanto-occlpltal membrane, and removal of the edges of the occipital bone Neck flexion was adjusted so that the dorsal surface of the medulla was horizontal. The rostral edge of the area postrema tn the mldhne (obex) served as the stereotaxlc zero for the rostro-caudal and medlo-lateral coordinates.

which passed a colhslon test 3° after ant~drom~c actwat~on were used for further study.

Baroreceptor sttmulatton Baroreceptor actiwty was altered by ralsmg AP more than 50 mmHg with an intravenous bolus rejection of noradrenahne (8/ag/ kg, Sigma, USA), or by lowering AP (20-50 mmHg) with an intravenous bolus injection of sodmm nitroprusside (250/ag/kg, Sigma, USA).

Chemoreceptor sttmulatton Chemoreceptors were stimulated by ventilating rabbits for approximately 1 mln with the following gas mixtures' 10% O 2 m N 2 (hypoxla), 5% CO2 + 10% 02 in N 2 (hypoxia/hypercapma) or 7% C O 2 i n O 2 (hypercapma). The gas mixtures (supphed by Commonwealth Industrial Gases Limited, Australia) were connected to the input of a ventilator. Bolus injection of sodmm cyamde (Aldrich, USA, 0.5 ml of 50/ag/ml m Ringer solution) into the common carotid artery, close to the carotid bifurcation was used for sUmulat~on hmited to peripheral chemoreceptors.

Electrtcal sttmulatton A monopolar tungsten electrode, insulated with glass except for the final 100/am (tip dmmeter 30/am) was connected via a Grass PSIU-6D stimulus isolation umt to a Grass $88 stimulator. After the dorsal surface of the medulla was exposed, the tip of the electrode was posluoned in the RVLM, 2-3 mm rostral to obex, 2.53.5 mm lateral to the mldhne, and 4.5-5 mm below the dorsal surface of the medulla. Electrical stimulation (cathodal pulses, 10 Hz, 0 5 ms duration, 500/aA, for 5 s) increased AP by approximately 50 mmHg. The stimulus intensity for antldrom~c activation of neurons located in CVLM was then set at 600-800/aA, 0.5 ms duration, 0 5-1 0 Hz At the end of most experiments a small electrolytlc lesion was made to locahse the posmon of the electrode tip. The left aortic depressor nerve (ADN) was isolated from the dorsolateral approach, placed on a bipolar stainless steel hook electrode, identified electrophyslologlcally by its charactenstlc pulse related rhythm and protected m paraffin o11. Electrical stimulation of the ADN (10 s tram, 50 Hz, 0 5 ms, 300/aA, cathode proximal) reduced a fall m AP by approximately 20 mmHg, accompanied by a bradycar&a. In 10 rabbits, the left cervical vagus nerve was exposed from the dorsolateral approach, cut low m the neck, placed on a bipolar stainless steel hook electrode and immersed m paraffin o11. Stimulation of the vagus nerve, to ehc~t antidromlc potentials m neurons m the nucleus amblguus, was accomphshed with single pulses of 500-900/aA, 0.25 ms duration, 0.5-1 Hz (cathode proximal).

Smgle umt recordings Single-umt recordings in the CVLM were made with the recordmg electrode inclined 18° from the vertical (tip rostral), 0 2-2.8 mm caudal to the obex, 2.5-3.6 mm lateral from the midhne and 2.5-5 mm ventral from the dorsal surface of the medulla. Because of the short distance between recording and stimulating electrodes, we recorded most antidromically activated units on the side contralateral to the stimulating electrode. The recording electrodes were either single glass-coated tungsten electrodes, or single glass electrodes filled with 0.5 M sodium acetate and 2% Pontamine sky blue (BDH Chemicals Ltd., Poole, UK). Umts activities were recorded conventionally with a N T l l 4 A differential amphfier (Neomedlx) filtered (band pass 500-10,000 Hz), amplified using a NL106 amplifier (Digitimer), monitored on a Tektronix 7313 oscilloscope and photographed with a Polaroid camera. Single unit activity, digitized by a window &scnmmator (Spike Trigger, NL200), was counted during set intervals (0 2-10 s) and &splayed as integrated activity on a Grass Polygraph. The onginal unit activity was continuously recorded into a casette data recorder (TEAC, MR-30) together with AP and end-expiratory CO 2. Recording sites were marked either by iontophoretic depositing of Pontamine sky blue through the glass recording electrode or, when tungsten recording electrodes were used, by making an electrolytic lesion. Only CVLM units

Htstology A catheter was placed m the ascendmg aorta at the end of most experiments, and the brain was fixed by perfuslon with aldehyde/ glutaraldehyde solution. The medulla was removed and 50 /am transverse sections were cut on a Vlbratome. Secuons were covered with bicarbonate buffered glycerol and examined m a Leltz fluorescence microscope. The formaldehyde-glutaraldehyde (FAGLU) fluorescence histochemical procedure ~6 was used to relate recording sites, marked wzth a lesion or with Pontamme sky blue (red m the fluorescence microscope), to the A1 group of catecholamine-containing neurons. In some cases, the same sections were rehydrated, stained with neutral red and covershpped with DPX. The posmons of the recording sites were mapped using an Olympus BH2 microscope, the Magellan Image Analysis System 2I and a Mclntosh II CX computer.

Stansttcal analysts Baseline mean AP, heart rate, unit &scharge rate, and changes of these values in response to baroreceptor and chemoreceptor stlmuh, were recorded. Data were expressed as mean + S.E.M. and changes m values were examined using a paired Student t-test

RESULTS Single u n i t a c t i v i t y w a s r e c o r d e d f r o m 107 s p o n t a n e ously active neurons. Antodromic RVLM

activation from the

w a s v e r i f i e d b y c o n s t a n t l a t e n c y , ability to fol-

low high stimulation frequencies,

characteristic IS-SD

components, and by all-or-nothing action potential, conf i r m e d in e v e r y c a s e b y c o l l i s i o n (Fig. 1 A ) o f t h e a n t i dromically evoked action potential with the orthodromic a c t i o n p o t e n t i a l r e s u l t i n g f r o m a s p o n t a n e o u s d i s c h a r g e 3°. A s t h e r e c o r d i n g e l e c t r o d e w a s a d v a n c e d i n t o t h e vasodepressor region, especially between

2.5 a n d 3 m m

below the dorsal surface of the medulla, we recorded from

units

with

obvious,

characteristic,

respiratory

rhythms. High amplitude spikes from neurons of the precerebellar

lateral

reticular

nucleus,

located

approxi-

m a t e l y b e t w e e n 4 a n d 4.5 m m b e l o w t h e d o r s a l s u r f a c e o f t h e m e d u l l a , c a u d a l t o t h e o b e x , w e r e also c h a r a c t e r istic. N e u r o n s w i t h t h e s e c h a r a c t e r i s t i c s w e r e n e v e r a n tidromically activated from the RVLM.

198

A

7.

~ ~ l ~

C total umts (n = 98) (median 4 0 spikes/s)

Z

0

1

2

3

4

5

6

.~ 1~

ted by

•

0

(median 1 spike/s)

1

2

3

4

12] "~ 104

il;' il :1 d

"°

"8" 9 "'1"5" 1"6"i9" _b

units excited (n = 21) by increase in blood pressure

"F. a I m l e 6t H

12.

7

5

6

7

8

9

10

units inhibited (n = 59) by increase in blood pressure (median 4.8 spikes/s)

t~t

:: .

0

.25

5

75

1 1 25 1.5 1.75 2 2.25

3.25

425

Conduction velocity (m/s)

0

1

2

3

4

5

6

7

8

9

15

.r'2

16 1~ 20

Discharge rate (spikes/s)

Ftg. 1. Electrophyslologtcal characteristics of antldromlcally activated, collision test-positive units recorded In rabbit caudal ventrolateral medulla (CVLM). A: collision test. Oscilloscope traces showing antidromlc activation of unit in CVLM by electrical activation from rostral ventrolateral medulla (RVLM) Each trace is taken from 5 superimposed oscilloscope sweeps, triggered by spontaneous spikes Top" RVLM stimulus (at arrowhead) was timed to occur just after the critical delay and produced a constant latency spike response (marked by star) on every occasion. Bottom. RVLM stimulus was timed lUSt less than the critical delay, and the antidromlc spike response was always canceled. Bars = 2 ms and 0.2 mV; B" frequency histogram of axonal conduction velocity antidromically activated units, C" frequency histogram of spontaneous discharge rate of 98 umts antldromlcally activated from RVLM Spontaneous discharge rate of 9 antldrommcally activated collision test-posmve units was not recorded,

W h e n C V L M neurons were antidromically activated from the R V L M the threshold stimulus intensity ranged between 100 and 750 # A . The amplitude of the action potential ranged between 70 and 1000 ktV (mean 272 _+ 19/~V). The distance between stimulating and recording electrodes was approximately 3 m m for ipsilateral and 6.7 m m for contralateral stimulation. Latencies ranged between 1.5 and 30 ms, and conduction velocity thus ranged between 0.22 and 4.46 m/s (median 0.84 m/s, Fig. 1B). Resting discharge rate ranged between 0.1 and 20 splkes/s (median 4 spxkes/s, Fig. 1C). Many antidromically activated C V L M units were not spontaneously active and in these cases a collision test was not possible. We r e c o r d e d an additional 101 such units, with median conduction velocity 0.84 m/s, threshold for antidromlc

activation 411 + 24/xA and amplitude 244 + 14 AtV. We excluded these units from further analysis.

Responses to alteration m baroreceptor inputs The response to a rise in AP, p r o d u c e d by intravenous noradrenaline, was assessed for 96 of the 107 units. The discharge rate of 21 units (22%) was increased (1.6 _+ 0.3 spikes/s to 6.8 + 0.8 spikes/s, P < 0.01, 4 ipsilateral, 17 contralateral); the discharge rate of 16 units (17%) was unaffected; the discharge rate of 59 units (61%) was decreased (4.9 _+ 0.5 to 0.7 _+ 0.2 spikes/s, P < 0.01, 9 ipsilateral, 50 contralateral) (Figs. 2, 3 and 4). Both the increase and decrease in unit discharge rate was observed within approximately 1 s of the change in AP. The time course reflected the changes in A P (Figs. 3 and

199

C

A ARTERIAL 150 ] PRESSURE 100 (ram Hg) 50

Units in CVLM arttidmmlcally activated from RVLM coUisimt test positive (107}

¢ 96 umts tested with 1.v NA J

UNIT

DISCHARGE

$

B ARTERIAL 150] PRESSURE 100 (mm Hg) 5O

UNIT DISCHARGE

I umt excated (11%)

21 umts extnted

16 umts unaffected

59 umts mtubtted

(22%)

(17%)

(61%)

wath Lv. SNP

wath LV SNP

I umt unaffected (11%)

7 umts mhlbtted (78%)

4 umts unaffected

smth t v SNP

20 umts excated (80%)

4 umts l umt unaffected mh~b~ted (16%) (4%)

¢ Fig. 2. Responses of antidromically activated units to intravenous injection of noradrenaline (NA' 8/~g/kg; marked by arrows m A and B). A' oscilloscope trace of unit excited by increase m blood pressure. Bars = 2 s, 0.2 mV; B: oscilloscope trace of unit inhibited by increase in blood pressure. Bars = 2 s, 0 1 mV; C: summary of results showing proportion of antidromically activated CVLM neurons which were tested to intravenous administration of noradrenahne and sodium mtroprusslde (SNP: 250/~g/kg).

4), except that mean AP remained slightly changed for some time after the neuron returned to its original discharge rate. The spontaneous discharge rate of units excited by an increase in AP was 1.6 + 0.3 spikes/s (median 1 spike/s), less than the discharge rate of units inhibited by injection of noradrenaline (4.9 + 0.5 spikes/s, P < 0.01; median 4.8 spikes/s; Fig. 1C). We tested the response of the 9 units, previously excited by an increase in AP, to hypotension induced by injection of sodium nitroprusside. This procedure decreased the discharge rate of 7 of the 9 units, from 1.5 + 0.4 to 0.4 + 0.2 spike/s (P < 0.01). Conversely, hypotension induced by sodium nitroprusside increased (5.8 + 1 to 7.3 + 1.3 spikes/s, P < 0.01) the discharge rate of 20 of 25 units (80%) previously inhibited by intravenous injection of noradrenaline.

Responses to electrical stimulatton of A D N We assessed the effect of A D N stimulation on 5 units previously excited by intravenous noradrenaline. The discharge rate of these units increased from 2.7 + 1.3 spikes/s to 5 + 1 spikes/s (P < 0.05), the increase being apparent within approximately 1 s of the onset of stimulation, before AP started to fall (Fig. 3). Conversely, A D N stimulation decreased the discharge rate of 18 of

19 units previously inhibited by intravenous noradrenaline (5.3 + 0.7 to 1.4 + 0.3, P < 0.01, Fig. 4).

Responses to alteration in chemoreceptor inputs One minute of ventilation with normocapnic hypoxic gas mixture decreased p O 2 to less than 50 mmHg, with no change in p C O 2. This was associated with an increase in AP (from 90 _+ 2 m m H g to 111 + 2 mmHg, P < 0.01, n = 51) and a decrease in heart rate (from 259 + 4 to 172 _+ 9 beats/min, P < 0.01, n = 51). We tested the response to hypoxia of 16 units previously excited by intravenous noradrenaline. Fourteen (87%) of these units were excited by this procedure, their discharge rate increasing from 1.6 + 0.4 to 6.3 _+ 0.9 spikes/s (P < 0.01) (Fig. 3). We also tested the response to hypoxia of 35 units previously inhibited by intravenous noradrenaline (Fig. 4). Twenty one (60%) of these units were excited, the discharge rate increasing from 5.6 _+ 0.8 to 8.8 _+ 1.2 spikes/s (P < 0.01), 2 (6%) units were unaffected and 12 (34%) units were inhibited, their discharge rate decreasing from 6.9 + 1.3 to 0.8 + 0.3 spikes/s (P < 0.05) (Fig.

5). Ventilation of animals with hypoxic/hypercapnic gas mixture increased pCO2 to more than 60 m m H g and decreased pO2 to less than 50 mmHg. This was associated

200

B

150-

C

D

ARTERIAL PRESSURE 1130

(nun Hg~ 50

I 15sl

Coeatslmin)

r

UNIT DISCHARGE (spikes/0.4s) 0

noradrennline (8 Ixg/kg i.v.)

ADN stimulation

nitropru~ide (250 ~tg/kg i.v.)

hypoxia (10 % O2in N2)

F~g. 3. Polygraphtraces dlustratmg responses of arterial pressure, heart rate and discharge rate of CVLM umts antldromlcally activated from RVLM, to intravenous injection of noradrenaline (A), stimulation of aortic depressor nerve (ADN) (B), intravenous rejection of sodmm mtroprusside (C), ventilation of rabbit with hypomc gas mixture (D).

with an increase in AP from 90 + 2 to 107 + 1 mmHg (P < 0.01, n = 38) and a decrease in heart rate from 250 _ 4 to 184 + 8 beats/min (P < 0.01, n = 38). We tested the effect of hypoxia/hypercapnia of 10 units whose discharge rate was previously increased by intravenous noradrenaline. Nine of these 10 units were excited by hypoxia/hypercapnia, their discharge rate increasing from 1.8 + 0.6 to 7 + 0.8 (P < 0.01). Conversely, of 28 units previously inhibited by intravenous noradrenaline, 17 units were excited (5 + 0.8 to 7.6 + 1 spikes/s, P < 0.01) and 11 were inhibited (7,5 + 0.4 to 1.2 ___ 0.3 spikes/s, P < 0.01) by hypoxia/hypercapnia. Ventilation of animals with 7% CO2 in oxygen mcreased pCO 2 to more than 60 mmHg. There was no associated change in AP or heart rate. We assessed the response to hypercapnia of 8 units previously excited by intravenous noradrenaline. None of these units was affected. Of 25 units whose discharge rate was decreased by intravenous noradrenaline by increase in AP, 14 were unaffected by hypercapnia, 5 were excited (5.8 + 1.4 to 7.5 + 1.5 spikes/s, P < 0.01), 6 were inhibited (5.8 + 0.9 to 2 + 0.7 spikes/s, P < 0.01). Of 17 units excited by hypoxia and hypoxia/hypercapnia, 5 were excited by hypercapnia alone, 11 were unaffected and 1 was inhibited. Of 8 units inhibited by hypoxia and hypoxia/hypercapnia, 5 were inhibited by hypercapnia alone and 3

were unaffected (Fig. 5). We tested the response of 14 CVLM units to rejection of sodium cyanide into the common carotid artery. Injection of sodium cyanide caused bradycardia and increase in AP. We tested 3 antidromically activated units whose discharge rate had previously been increased by intravenous noradrenaline. The response of tested units was similar to the response to hypoxia, with 2 units being excited by the cyanide and 1 being unaffected. Cyanide injection inhibited the discharge rate of 8 of 11 units previously inhibited by intravenous noradrenaline, similar to the response to hypoxia. The remaining 3 units were excited by hypoxia and by cyanide injection.

Location of sttmulatmg and recording sites Histological determination of the stimulating sites revealed that the tip of the stimulating electrode had been m the region ventral, ventrolateral or ventromedial to the nucleus ambiguus (Fig. 6), in the region containing the bulbospinal sympathoexcitatory neurons. Histological examination of recording sites revealed that units were located from 0.5 mm rostral to 2 mm caudal to the obex, and from 2.8 mm to 3.6 mm lateral from the midline (Fig. 6), in and dorsomedial to the region containing the A1 catecholamine cells (Fig. 6). Units excited by an increase in AP were mainly in the vicinity of nucleus ambiguus, generally more dorsal and medial than units

201 1TM

150

ARTERIAL PRESSURE (mm Hg)

100

50

., 1 5 s ,

--T TEi] (beats/mm)

/ !

f

UNIT 4 ] DISCHARGE (spikes/0.4s) 0 m

noradrenaline (8 Ixg/kg i.v.)

150

ARTERIAL PRESSURE

ADN stimulation

nitroprusside (250 Itg/kg i.v.)

hypoxia (I0 % O2in N2)

F

E

100

(ram Hg) 50

H E A R T RATE

(beats/rain)

i]

UNIT 4 ] DISCHARGE (spikes/0.4s) 0

hylx~apnic hypoxia (5% CO2,10% O2 in N2)

hyptnor.~

(7% C02 in Oa)

Fig. 4. Polygraph traces illustrating responses of arterial pressure, heart rate and discharge rate of CVLM unit antidromically activated from RVLM to intravenous injection of noradrenaline (A), stimulation of aortic depressor nerve (ADN) (B), intravenous injection of sodium mtroprusslde (C), ventllauon of rabbit with hypoxlc gas mixture (D), ventilation of rabbit with hypercapnic/hypoxlc gas mixture (E), ventdatlon of rabbit with hypercapmc gas mixture (F)

inhibited by an increase in AP, with some overlap.

Identification and properties of neurons antidromically activated from vagus nerve Because vagal cardiac preganglionic neurons in the nucleus ambiguus, in the general region of the vasodepressor neurons, are also likely to be activated by in-

creases in AP, in 10 rabbits we identified these neurons by antidromic activation from the cervical vagus. Single unit activity was recorded from 13 spontaneously active, collision test-positive, neurons, which were also excited by electrical stimulation of the A D N (discharge increase from 1 _ 0.3 to 4.22 + 1.02, n = 13, P < 0.01). All these cells were ipsilateral to the side of the stimulated

202

Units excited by NA (21)

Units inhibited by NA (59)

+

+

1,6 onIts tested +,th hr_w_x+aI

14 units excaed (87%)

I" 5 units tested with LL, h~ poxla• !

2umts unaffected (13%)

21 umts excited (60%)

2 units

unaffected (6%)

12 units inhibited (34%) _

t

lO units tested with l h~_poxm/h~:~ercapniaJ

9 units exated

1 unit unaffected

°L

I 8 units tested w, th hypercapnia

8 units unaffected

j

i 18units tested w i t h - I [ h_ypo__xia/hypercapnia I

17 umts excited

10 units tested with I • hypoxla/hypercapnla

1 unit mh~b~ted

J 17umtstesiedwlth I

hypercapnla

5 units excited

I 1 umts unaffected

l I

1 umt Inhibited

10 umts Inhibited

8un,tstested with --II hypercap,_m_a A

3 umts unaffected

5 umts inhibited

Fig. 5 Summary of results showing proportion of CVLM umts antldromlcally activated from RVLM, which responded to rejection of noradrenahne (NA), to ventilation with hypox~c, hypercapmc/hypoxlcand hypercapmc gas mixtures

vagus nerve. Histological determination of recording sites revealed that these neurons were located in the nucleus ambiguus, from 0.5 mm caudal to 1 mm rostral to the obex. The spontaneous discharge rate ranged between less than 1 to 2.5 spikes/s, (mean 1 + 0.3 spikes/s), with no obvious respiratory or cardiac rhythm. Latency for antidromic activation from the cervical vagus was 10.2 + 1.1 ms. The distance between stimulating and recording electrodes was approximately 45 mm, and conduction velocity therefore varied between 2.6 and 9 m/s, with mean 4.7 + 0.6 m/s. The effect of increasing AP was assessed for 9 of these neurons. All were activated, their discharge rate increasing from 1 + 0.3 to 4.7 + 1.3 spikes/s n = 9, P < 0.05; Fig. 7), and lasting for approximately 1 min, similar to the duration of the increase in AP and the bradycardia produced by injection of NA. The vagal preganglionic neurons all increased their discharge rate in response to 1 min of ventilation with the hypoxic gas mixture (1.3 + 0.4 to 6.4 + 1.2 spikes/s, n = 7, P < 0.01). During the hypoxic stimulus these units clearly demonstrated a respiratory rhythm. Each identified vagal preganglionic cell was tested to see if it could be antidromically activated by electrical stimulation of the RVLM. None could be activated by this procedure. Similarly, no CVLM cells originally an-

tidromically activated from the RVLM could also be antidromically activated by ipsilateral vagal nerve stimulation. DISCUSSION Our results confirm preliminary reports by Terui et al. 38 in rabbit, and Agarwal and Calaresu I in rat, documenting that the vasodepressor region of the CVLM contains neurons which can be antidromically activated from the sympathoexcitatory region of the RVLM. Like Terui et al. 38, we found that the discharge rate of most of these neurons is promptly affected by alteration of baroreceptor inputs. However, all 4 of their antidromically identified units were excited by activation of baroreceptors, accomplished either by raising AP or by electrical stimulation of the ADN. In contrast, of the 96 antidromically identified units we tested for baroreceptor sensitivity, 22% increased their discharge rate m response to activation of peripheral baroreceptors, accomplished by raising AP, but many units (61%) were inhibited by baroreceptor inputs and others (17%) were unaffected. Similarly, of 32 barosensitive, but not all antidromically activated, CVLM neurons tested, Agarwal and Calaresu 1 reported increases in the discharge rate of 20, and decreases in the discharge of 12. The barorecep-

203

B

•

-

•

-

•

-

*

-

u n i t s a n t i d r o m i c a l l y a c t i v a t e d f r o m RVLM excited b y increase in b l o o d p r e s s u r e u n i t s a n t i d r o m i c a ! l y a c t i v a t e d f r o m RVLM i n h i b i t e d b y increase in b l o o d p r e s s u r e u n i t s a n t i d r o m i c a l l y a c t i v a t e d f r o m RVLM u n a f f e c t e d b y increase in b l o o d p r e s s u r e u n i t s a n t i d r o m i c a l l y a c t i v a t e d f r o m v a g u s n e r v e (not f r o m RVLM) excited b y i n c r e a s e in b l o o d p r e s s u r e

Fig. 6. Location of stimulating and recording sites in rabbit ventrolateral medulla. A: photomicrograph of RVLM showing position of stimulating electrode marked by lesion (arrow), ventromedlal to nucleus amblguus. Bar = 0.5 ram; B: diagram of rostral medulla showmg position of lesion indicated in A; C. diagram of 3 representative hemmsections of medulla (from obex level to 1.5 mm caudal to obex) showing location of recording site of 29 units; D: photomicrograph, using dark filter, of a recording site of unit, antldromically activated from RVLM, excited by increase in blood pressure, marked by Pontamine sky blue (ventrolateral to nucleus ambiguus, 1 mm caudal to obex). Bar = 1.2 ram; E: high magnification (with light filter) of the subregion indicated by the rectangle in D showmg recording site in the vicinity of A1catecholamine synthesizing cells (marked by arrows); F: photomicrographs of same section using FAGLU-fluorescence method Bars for E and F = 100/~m AP, area postrema; IO, mferior olive; LRN, lateral reticular nucleus; MVN, medial vestibular nucleus; nA, nucleus amblguus; nPH, nucleus prepositus hypoglossl; nR, nucleus raphe; nTS, nucleus tractus solitarlus; py, pyramidal tract; Vsp, spmal nucleus of the tngemmal nerve, XII, hypoglossal nucleus

tor responses of o u r neurons were consistent in that units excited by increase in A P were also excited by electrical stimulation of the A D N , and inhibited by b a r o r e c e p t o r unloading using intravenous nitroprusside. Similarly, neurons initially inhibited by raising A P were excited

when A P was lowered with nitroprusside, and those unaffected by raising A P were also mostly unaffected by lowering AP. The v a s o d e p r e s s o r region contains m a n y different classes of cells, including the A1 cells with axons passing

204 150

A

C

B

ARTERIAL PRESSURE 100 (ram Hg)

50 400

HEART RATE (beats/rain) 200'

UNIT DISCHARGE (spikes/0.4s)

.':

,,~',

'.l "

'

i)

~ ~ ',

noradrenaline (8 ~tg/kg i.v.)

ADN stimulation

,~ 11, ,In t ~ l ~

, ~. l ~ l ~ . ~ O ~ , ! .

,,

hypoxia (10 % 02 m N2)

Fig. 7 Polygraph traces illustrating responses of arterial pressure, heart rate and discharge rate of unit located m nucleus amblguus, 0 5 mm caudal to obex, anudromically activated from ipsilateral cervical vagus nerve, to intravenous rejection of noradrenaline (A), stimulatton of aortic depressor nerve (ADN) (B), ventilation with hypoxic gas mixture (C)

through the R V L M en route to the hypothalamus 29'34'3s and GABA-synthesizing neurons with axons terminating in the R V L M 5. Our neurons were all antidromically activated from the sympathoexcitatory region of the rostral medulla and therefore had an axon projecting to or through the region affected by the current from the stimulating electrode. Terui et al. 38 used threshold techniques to try and ensure that only C V L M neurons with axonal processes terminating on R V L M sympathoexcitatory neurons were antidromically activated. However, this method can only be relatively reliable. In our study we recognized that axons passing through R V L M , including those of the A1 cells, could also be activated by our stimulating electrode. Evidence from a companion electrophysiological study 26 is relevant to this issue. In the study we antidromically activated C V L M neurons after positioning our stimulating electrode m the supraoptic nucleus of the hypothalamus The median conduction velocity of the neurons identified from the supraoptic nucleus was 0.7 m/s, not significantly different from the mean conduction velocity of the C V L M units antidromically activated m the present study. However, when the stimulating electrode was in the supraopt~c nucleus, no anUdromically activated C V L M unit was ex-

cited by an increase in AP. Indeed 93% of 101 antidromically activated C V L M units decreased their discharge rate in response to an increase in A P and the remaining units were unaffected by the change in AP. Similarly, after antidromic activation from the supraoptic nucleus no C V L M units decreased their discharge rate in response to hypotension induced by intravenous nitroprusside. This finding is consistent with other electrophysiological studies of A1 cells 23'32. These results contrast with those of the present study in which 22% of neurons antidromically activated from the R V L M increased their discharge when A P was elevated, and nearly all these units decreased their discharge when A P was reduced. In the present study we can therefore be relatively confident that those C V L M neurons which were excited by baroreceptor inputs belong to a population of neurons which is separate from the population of A1 neurons. The discovery of the vasodepressor neurons in the C V L M led naturally to the hypothesis that they may function as the inhibitory link known to be present in the central baroreceptor-vasomotor reflex pathway. Evidence that this is likely to be the case was obtained in r a t s 3'17'2°'24'39 but experiments in rabbits 4"1° demonstrated that inhibition of neuronal function in the region

205 containing the vasodepressor neurons did not abolish the baroreceptor-vasomotor reflex. The demonstration in the rabbit 38 that some CVLM neurons, antidromically activated from the RVLM, receive excitatory baroreceptordriven inputs suggested that some CVLM vasodepressor neurons might be part of the baroreceptor-vasomotor pathway in this species. Our present demonstration that some antidromically identified CVLM neurons are excited by baroreceptor inputs is consistent with this view. We noted, as did Terui et al. aS, that some of the baroreceptor-excited neurons are located 0.5 mm caudal to 0.5 mm rostral to the obex, in a region intermediate between the CVLM and the RVLM. In a previous mapping study from our laboratory 25 we were unable to elicit any vasodepressor responses from this more rostral intermediate region. In a recent paper, Masuda et al. 31 demonstrated that injection of excitatory amino acid into the intermediate region elicits depressor responses in rabbits with cut aortic and vagus nerves. Li et al. z8 have confirmed that depressor responses are obtainable from the intermediate area in rabbits with cut baroreceptor nerves. It must be that the process of cutting the baroreceptor nerves somehow unmasks the presence of vasodepressor neurons just rostral to the obex. Masuda et al. 31 have demonstrated that inhibition of neuronal function in both the CVLM and the intermediate vasodepressor areas does eliminate the baroreceptor-vasomotor reflex in the rabbit, an observation also confirmed by Li et al. 28. Similar findings have now been reported for the rat 12. It therefore seems likely, that the antidromically activated units which are excited by baroreceptor inputs are inhibitory links in the central baroreceptor-vasomotor pathway. However, our present paper demonstrates that not all antidromically activated units are excited by baroreceptor inputs, and we believe that the pool of vasodepressor neurons in the CVLM is likely to include sympathoinhibitory neurons whose activity is relatively independent of baroreceptor inputs. The powerful tonic GABA-mediated inhibition of the vasodepressor neurons 8'4°, for example, is difficult to explain in terms of baroreceptor activity. Some of the many units inhibited by baroreceptor inputs in our study may also be vasodepressor neurons with a complex function, rather than A1 cells with axons passing through the RVLM. The low resting discharge rate of the vasodepressor neurons excited by an increase in AP is somewhat surprising, given the vigorous inhibitory contribution of baroreceptors to the resting discharge rate of RVLM sympathoexcitatory neurons 2"2"z'31. The low discharge rate, together with the powerful action on AP, of the depressor neurons makes the situation similar to that described by Ellenberger et al.14 for vagal preganglionic neurons. These authors found that a few cells with low

spontaneous discharge rates can, when activated, exert a powerful influence on heart rate. Most antidromically identified units whose discharge increased when AP was increased also increased their discharge rate during activation of chemoreceptors by hypoxia. Recently Sun and Spyer 37 reported that activation of chemoreceptors produces an initial short excitatory response in RVLM bulbospinal neurons, followed by a fall in their discharge rate. This fall could be mediated by an action on the vasodepressor neurons. Hypoxia is known to increase sympathetic vasomotor activity 13'15'26, and the role of depressor neurons in reflex responses to hypoxia is therefore not clear. The situation is complicated because it is difficult to determine whether the effects of hypoxia on the CVLM neurons is a primary chemoreceptor-mediated event, or a secondary event reflecting baroreceptor activity initiated by the chemoreceptor-mediated rise in AP. We identified 13 spontaneously active neurons by stimulation of the cervical vagus nerve. All these neurons were located in the nucleus ambiguus. Their discharge rate, their activation by increases in AP, and by chemoreceptors, was similar to the responses recorded in the vasodepressor neurons antidromically activated from the RVLM. However none of the vagal units could be antidromically activated from the RVLM. They differed from the vasodepressor neurons in their response to baroreceptors activation by intravenous noradrenaline, being activated during the whole period that the increase in AP produced bradycardia. The conduction velocity of axons of these cells was higher than that of the vasodepressor neurons, in the range of myelinated axons (group B), similar to that reported in rabbit 14'22 and in cat 11'33. Depressor neurons antidromically activated from RVLM were therefore clearly distinct from vagal preganglionic neurons. Conclusion Our study provides evidence for the existence, in the vasodepressor region of the caudal ventrolateral medulla and in a more rostral intermediate zone, of neurons which project to the rostral sympathoexcitatory area, and increase their discharge rate in response to elevation of arterial pressure or electrical stimulation of the ADN. These neurons, possibly GABA-synthesizing, could be part of the inhibitory link known to exist in the central baroreceptor-vasomotor pathway. Other neurons in the vasodepressor region are either inhibited or unaffected by baroreceptor inputs. These neurons could be A1 cells, with axons passing through the rostral sympathoexcitatory area en route to the hypothalamus, or they could be vasodepressor neurons with more complex, nonbaroreceptor, functions.

206

Acknowledgements Our study was supported by National Heart Foundation of Austraha and National Health and Medical Re-

search Councd Mr Jamle Haupt provided techmcal assistance

REFERENCES 1 Agarwal, S.K. and Calaresu, E R , Monosynaptlc connection from caudal to rostral ventrolateral medulla m the baroreceptor reflex pathway, Brain Res., 555 (1991) 70-74. 2 Agarwal, S K., Gelsema, A.J. and Calaresu, E R . , Neurons m rostral VLM are inhibited by chemical stimulation of caudal VLM m rats, Am. J Phystol, 257 (1989) R265-R270 3 Agarwal, S.K., Gelsema, A.J. and Calaresu, ER., Inhibition of rostral VLM by baroreceptor actwaUon xs relayed through caudal VLM, A m J Phystol, 258 (1990) R1271-R1278. 4 Blessing, W.W., Baroreceptor-vasomotor reflex after N-methylo-aspartate receptor blockade m rabbit caudal ventrolateral medulla, J Physiol., 416 (1989) 67-78 5 Blessing, W.W., Distribution of glutamate decarboxylase-containlng neurons in rabbit medulla oblongata w~th attention to lntramedullary and spinal projections, Neurosczence, 37 (1990) 171-185 6 Blessing, W.W, Depressor neurons m rabbit caudal medulla act wa GABA receptors in rostra1 medulla, Am. J Phystol, 254 (1988) H686-H692 7 Blessing, W W., Hedger, S.C., Joh, T H. and Willoughby, J.O , Neurons m the area postrema are the only catecholamme-syntheslzmg cells m the medulla or pons with projections to the rostral ventrolateral medulla (Cl-area) in the rabbit, Brain Res, 419 (1987) 336-340 8 Blessing, W.W and Rels, D.J., Ewdence that GABA and glycme-like inputs mhlb~t vasodepressor neurons m the caudal ventrolateral medulla of the rabbit, Neurosct Lett, 37 (1983) 5762. 9 Blessing, W.W. and Rels, D . J , Inhibitory car&ovascular function of neurons m the caudal ventrolateral medulla of the rabbit: relationship to the area containing A1 noradrenerglc cells, Brain Res., 253 (1982) 161-171 10 Blessing, W.W and Wdloughby, J . O , Depressor neurons m rabbit caudal medulla do not transmit the baroreceptor-vasomotor reflex, A m J. Phystol, 253 (1987) H777-H786 11 Clnello, J. and Calaresu, F R., Distribution of vagal cardlomhibltory neurons m the medulla of the cat, A m J Physlol, 238 (1980) R57-R64. 12 Cravo, S L , Morrlson, S.F and Reis, D.J., Differentiation of two car&ovascular regions within caudal ventrolateral medulla, A m J Phystol Regul. Integr Comp Physzol, 261 (1991) R985-R994. 13 Dorward, P . K , Burke, S L., Jamg, W and Cassell, J , Reflex responses to baroreceptor chemoreceptor and noclceptor inputs m single renal sympathetic neurones m the rabb~t and the effects of anaesthesm on them, J Auton. Nerv. Syst., 18 (1987) 39-54 14 Ellenberger, H., Haselton, J.R., Lxskowsky, D.R and Schnelderman, N., The location of chronotrop~c cardlomfubltory vagal motorneurons m the medulla of the rabbit, J Auton Nerv Syst, 9 (1983) 513-529. 15 Fukuda, Y., Sato, A., Suzuki, A. and Trzebskl, A., Autonomic nerve and cardiovascular responses to changing blood oxygen and carbon dioxide levels m the rat, J Auton. Nerv Syst, 28 (1989) 61-74. 16 Furness, J.B , Costa, M and Blessing, W.W, Simultaneous fixat~on and production of catecholamme fluorescence m central nervous t~ssue by perfuslon with aldehydes, Htstochem J, 9 (1977) 745-750 17 Gordon, F J., Aortic baroreceptor reflexes are mediated by NMDA receptors in caudal ventrolateral medulla, A m J Physml, 252 (1987) R628-R633. 18 Granata, A.R., Kumada, M and Re~s, D . J . Sympatholnhxb~t~on by Al-noradrenerg~c neurons is medmted by neurons m the

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

C1 area of the rostral medulla, J Auton Nerv Syst. 14 (1985) 387-395 Granata, A R , Numao, Y., Kumada, M and Re~s, D J , A1 noradrenerglc neurons tomcally mhlblt sympathoexc~tatory neurons of C1 area m rat bramstem, Brain Res. 377 (1986) 127147 Guyenet, P . G , Ffltz, T M and Donaldson, S . R , Role of excitatory amino acids m rat vagal and sympathetic baroreflexes, Brain Res, 407 (1987) 272-284 Halasz, P and Martin, P , Magellan Program for Quantttattve Analysts of Histochemtcal Sectlons, Umverslty of New South Wales, Sydney, 1985, 62 pp. Jordan, D , Khahd, M E M., Schnelderman, N and Spyer, K.M , The location and properties of preganghonlc vagal cardlomotor neurones m the rabbit, Pflugers Arch, 395 (1982) 244-250 Kannan, H., Osaka, T , Kasal, M., Okuya, S. and Yamashlta, H., Electrophyslologlcal propemes of neurons m the caudal ventrolateral medulla projecting to the paraventncular nucleus of the hypothalamus m rats, Bram Res, 376 (1986) 342-350. Kubo, T and Klhara, M , N-Methyl-D-aspartate receptors mediate tomc vasodepressor control in the caudal ventrolateral medulla of the rat, Bram Res, 451 (1988) 366-370 L1, Y W. and Blessing, W W , Localization of vasodepressor neurons in the caudal ventrolateral medulla in the rabbit, Brain Res, 517 (1990) 57-63 L1, Y W , Gleroba, Z.J. and Blessing, W W., Chemoreceptor and baroreceptor responses of neurons projecting from ml area to supraoptic nucleus m rabbit, Am. J Physzol, submitted. L1, Y W , Gleroba, Z J , McAllen, R M and Blessing, W.W, Neurons m rabbit caudal ventrolateral medulla mhlblt bulbosplhal barosensitwe neurons in rostral medulla, Am J Physzol, 261 (1991) R44-R51 L1, Y W , Poison, J , Sved, P. and Dampney, R A L , Inhlblt~on of depressor neurons m the rabb~t mtermedmte ventrolateral medulla abohshes the vasomotor component of the baroreceptor reflex, Proceedings of the Australian Socwty of Neurosclence, Adelaide, February 1992. LI, Y.W, Wessehngh, S.L. and Blessing, WW., Projections from rabb~t caudal medulla to C1 and A5 sympathetic premotor neurons, demonstrated with phaseolus leucoagglutmm and herpes simplex virus, J Comp Neurol, in press Llpskl, J , AntldrOmlC actwatlon of neurons as an analytical tool m the study of the central nervous system, J Neurosct Methods, 4 (1981) 1-32 Masuda, N , Term, M , Koshlya, N. and Kumada, M , Neurons m the caudal ventrolateral medulla medmte the arterial baroreceptor reflex by mhlbltmg barosensmve retlculospmal neurons in the rostral ventrolateral medulla in rabbits, J Auton. Nerv Syst, 34 (1991) 103-118 McAllen, R.M. and Blessing, W.W, Neurons (presumably A1cells) projecting from the caudal ventrolateral medulla to the region of the supraopt~c nucleus respond to baroreceptor inputs in the rabbit, Neurosct Lett, 73 (1987) 247-252 McAllen, R.M and Spyer, K . M , Two types of vagal preganghomc motoneurones projecting to the heart and lungs, J Physlol, 282 (1978) 353-364. McKellar, S and Loewy, A.D., Efferent projections of A1 catecholamme cell group m the rat: an autoradlographlc study, Brain Res, 241 (1982) 11-29 Sawchenko, P.E and Swanson, L W , The orgamzatlon of noradrenergic pathways from the bramstem to the paraventncular and supraoptlc nuclei in the rat, Bram Res Rev, 4 (1982) 275325. Sun, M K and Guyenet, P G , Effect of clomdme and ~,-amlnobutyrlc acid on the &scharges of medullo-spmal sympathoex-

207 citatory neurons in the rat, Bram Res, 368 (1986) 1-17. 37 Sun, M.K. and Spyer, K.M., Responses of rostroventrolateral medulla spinal vasomotor neurones to chemoreceptor stimulation in rats, J. Auton Nerv. Syst., 33 (1991) 79-84 38 Term, N., Masuda, N., Saekl, Y and Kumada, M , Activity of barosenslt~ve neurons m the caudal ventrolateral medulla that send axonal projections to the rostral ventrolateral medulla in rabbits, Neurosct. Lett., 118 (1990) 211-214. 39 Urbanskl, R.W. and Sapru, H N., Evidence for a sympathoexcitatory pathway from the nucleus tractus sohtarn to the ventrolateral medullary pressor area, J Auton. Nerv Syst, 23 (1988) 161-174.

40 Wlllette, R.N., Barcas, PP., Kneger, A.J. and Sapru, H.N., Endogenous GABAergxc mechamsms m the medulla and the regulation of blood pressure, J Pharmacol Exp. Ther, 230 (1984) 34-39. 41 Willette, R.N., Barcas, P E , Kneger, A.J. and Sapru, H N , Vasopressor and depressor areas m the rat medulla. ~dentlficatlon by micromjectlon of L-glutamate, Neuropharmacology, 22 (1983) 1071-1079. 42 Willette, R . N , Punnen, 8., Kneger, A.J. and Sapru, H N., Interdependence of rostral and caudal ventrolateral medullary areas in the control of blood pressure, Bram Res, 321 (1984) 169-174.