0306-4522/90$3.00+ 0.00 Pergamon Press plc 0 1990IBRO

NeuroscienceVol. 34, No. 2, pp. 331-339, 1990 Printed in Great Britain

EXCITOTOXIN-INDUCED DEGENERATION VAGAL AFFERENT NEURONS

OF RAT

S. J. LEwts,*t A. J. M. VERBERNE,*$C. J. LOUIS& B. JARROTT,*11P. M. BEART* and W. J. LOUIS* University of Melbourne, *Clinical Pharmacology and Therapeutics Unit, and §Departments of Medicine and Pathology, Austin Hospital, Heidelberg, Victoria 3084, Australia Abstract-The inferior vagal or nodose ganglion contains the pcrikarya of vagal afferent neurons that function as cardiopulmonary and abdominal visceral receptors as well as aortic arch baroreceptors. In this study we have sought to utilize the axon-sparing properties of the excitotoxins kainic acid, N-methyl-o-aspartic acid and a-amino-3-hydroxy-4-isoxazolepropionic acid to destroy the perikarya of these sensory neurons and thus selectively de-afferent the vagus in the rat. Kainic acid (0.5 nmol/pI, 2 x 2~1) was applied topically to both nodose ganglia and the rats were allowed to recover for 7-8 days. Baroreceptor heart rate reflex activity was assessed in these conscious rats. Baroreceptor heart rate reflex gain was reduced (-51%) in kainic acid-treated rats, as was the maximal reflex bradycardia induced by the pressor agent, phenylephrine. Kainic acid treatment did not alter resting mean arterial pressure or heart rate. Vagal efferent neurons were spared by kainic acid treatment since bradycardic responses to electrical stimulation of the peripheral end of a cut vagus were not impaired. Histological studies showed marked destruction of perikarya within the nodose ganglia of kainic acid-treated rats: inflammatory and degenerative changes were evident at 2 days, and at 10 days there was considerable loss of neuronal cell bodies, but sparing of axons. Topical application to the nodose ganglion of a-methyl-DL-aspartic acid (6.8 nmol/pI, 2 x 2 PI), a non-excitotoxic dicarboxylic acid, failed to alter baroreflex sensitivity or produce perikaryal degeneration in nodose ganglia. In other studies, kainic acid (0.5 nmol/pl, 2 x 2 ~1) N-methyl-D-aspartic acid (6.8 nmol/pl, 2 x 2 ~1) or a-amino-3-hydroxy-4isoxazolepropionic acid (0.54 nmol/pI, 2 x 2 PI), when applied unilaterally to the nodose ganglion, produced after 7-8 days significant reductions in the 5-hydroxytryptamine-induced Bezold-Jarisch reflex after contralateral vagotomy and appreciable losses of neuronal cell bodies. These findings suggest that the excitotoxins may be useful as selective agents for inducing chemical deafferentation of mixed peripheral nerves and, in particular, for studying the role of vagal afferent neurons in cardiovascular reflexes.

The vagus or Xth cranial nerve contains afferent sensory fibres innervating thoracic and abdominal visceral structures,6.2S.26 as well as parasympathetic preganglionic motor fibres which have their perikarya

in the dorsal motor nucleus of the vagus and the ambiguus nucleus. ‘6.25Furthermore, the cell bodies of the vagal afferent fibers are located in the inferior or nodose ganglia (NG) which lie within the cervical vagal trunks. 25,33Central processes of these pseudobipolar neurons terminate principally throughout the rostro-caudal extent of the nucleus of the solitary tract,“.‘6.25.36while the peripheral processes convey baroreceptor, chemoreceptor and viscerosensory information to the central nervous system. Selective destruction of sensory neurons in mixed peripheral nerves has long been a desirable, but tPresent address: Department of Pharmacology, University of Iowa, Iowa City, IA 52242, U.S.A. IPresent address: Department of Pharmacology, University of Limburg, Maastricht 6200 MD, The Netherlands. //To whom correspondence should be addressed. Abbreviations: AMPA, cc-amino-3-hydroxy-4-isoxazolepropionic acid; DABP, diastolic arterial blood pressure: HR, heart rate; 5-HT, 5-hydroxytryptamine; Kk, kainic acid; a-MA, cc-methyl-DL-aspartic acid; MAP, mean arterial pressure; NC?, nodose ganglia; NMDA, Nmethyl-o-aspartic acid; SE., standard error of the mean.

elusive, goal, in neurobiological research. To that end, a number of different approaches have been attempted to induce chemical deafferentation of the vagus. These have included treatment with the neurotoxin capsaicin’4,20 or injection of various neurotoxic lectins38 into the vagal trunk. Capsaicin has been used with limited success to destroy vagal afferent neurons, but may selectively destroy only subpopulations of neurons within the NG,4,9 since capsaicin treatment did not affect baroreceptor reflexes in guinea-pigs’ or rats20 suggesting that those neurons conveying cardiovascular information were left intact. On the other hand, Bond and coworkers reported an impairment of the carotid sinus reflex in adult rats treated with capsaicin as neonates.2 The lectins appear to be unselective in that both afferent and efferent neurons are affected.38 Transection of mixed nerves unavoidably interrupts both motor and sensory fibres coursing through the nerve bundle. However, surgical denervation of arterial baroreceptors (sino-aortic denervation) by stripping away sensory fibres has been employed successfully in many laboratories,‘2,“~27 but has the disadvantage that some aortic arch baroreceptor fibres travelling within the vagal and sympathetic trunks’8,2’ are not interrupted by section of the aortic 331

S. J. LEWISet al.

332

depressor or superior laryngeal nerves.7 Furthermore, the afferent fibres arising from cardiopulmonary baroreceptors and other chemosensitive receptors within the cardiopulmonary region also travel within the vagal trunks33 and may not be readily sectioned in the rat’* without affecting vagal efferent function. In the larger animals, such as the cat and the sheep, surgical removal of vagal afferents at the level of the NG has been reported.22.23 Excitotoxins, which probably act at multiple glutamate receptors,29,30induce a selective pattern of neuronal degeneration when microinjected into the brain parenchyma. Specifically, these excitotoxins (e.g. kainic, N-methyl-D-aspartic and quisquahc acids) destroy neuronal perikarya while sparing axons of passage and nerve terminals.3.3’ In the present study, we have sought to determine whether the excitotoxins, kainic acid (KA), N-methyl-D-aspartic acid (NMDA) or a-amino-3-hydroxy-5-methyl-4isoxazolepropionic acid (AMPA), may induce selective destruction of vagal afferent neurons when applied directly to the NG with a concomitant attenuation of the baroreceptor heart rate (HR) reflex and the vagal cardiopulmonary reflex activated by 5-hydroxytryptamine (5-HT) (often referred to as the Bezold-Jarisch reflex33). A preliminary report of some of this work has been published elsewhere.34 EXPERIMENTAL PROCEDURES

All experiments described here were performed according to the ethical guidelines defined by the National Health and Medical Research Council of Australia. Surgical procedures Female Sprague-Dawley rats (18-20 weeks old) (Austin Hospital Animal House) were anaesthetized with a methohexitone/amylobarbitone mixture (16.7 and 30 mg/kg, i.p., respectively; Eli Lilly, Indianapolis, U.S.A.) and the left and right nodose ganglia were exposed via a mid-cervical incision and carefully desheathed. Desheathing was performed with the aid of needle-pointed forceps under a dissecting microscope. Preliminary experiments indicated that desheathing the ganglia enhanced the effectiveness of KA application. During this procedure care was taken not to damage the main vagal trunks or the superior laryngeal nerve which enter the ganglion. In eight rats the ganglia were superfused bilaterally with 2 ~1 of a KA (Sigma, St. Louis, U.S.A.) solution (0.47nmol/nl in 0.9% w/v NaCI) via a glass microsyringe. After 3 min the excess solution was absorbed into the tip of a sterile cotton bud. This procedure was repeated once more. Care was taken not to allow the KA solution to spread to the underlying petrosal ganglia (which contain the somata of the glossopharyngeal afferents emanating from the carotid body and carotid sinus). Control rats (n = 8) received similar applications (2 x 2 ~1) to the NG of the non excitotoxic amino acid x-methyl-LXaspartic acid (a-MA, 6.8 nmol/pl in saline; Sigma, St Louis, U.S.A.). The incisions were then closed and the rats were allowed 7-8 days to recover. After this time, each of the KAand r-MA-treated rats were reanaesthetized for surgical implantation of arterial and two venous cannulae into the abdominal aorta and jugular vein, respectively. After cannulation, the animals were allowed to recover for 48 h, during which the arterial cannulae were flushed twice daily with 0.2 ml volumes of heparinized saline (50 U/ml).

Baroreceptor heart rate reflex testing Baroreceptor HR reflex activity was assessed in these conscious, unrestrained rats as previously described.‘5 Briefly, reflex HR responses were produced by i.v. injections of the pressor agent phenylephrine (I-25pg/kg; KochLight, Colnbrook, U.K.) and the depressor agent, sodium nitroprusside (I-50 pg/kg; Roche, Sydney, Australia), dissolved in normal saline (0.9% NaCl w/v). Reflex HR changes to mean arterial pressure (MAP) manipulations were analysed by exponential sigmoidal curve analysis.‘3,35 This analysis yielded the baroreceptor HR reflex parameters (see Fig. 1C): I?, average gain (or sensitivity); P, and P,, maximum and minimum HR values; Range, P, - P, ; BP%, the MAP value at Range/Z; T,, and T,_, upper and lower thresholds. Cardiopulmonary reJlex testing Cardiopulmonary reflex testing was assessed in rats which had previously received unilateral applications to the NC of KA (as described above), or the other excitotoxins NMDA (6.8 nmol/nl in saline, 2 x 2 n; Sigma, St. Louis, U.S.A.), or AMPA (0.54 nmol/pl in saline, 2 x 2 ~1; Research Biochemicals Inc., Wayland, U.S.A.). Control rats received equal volumes of 0.9% NaCI w/v. After 778 days recovery, the rats were anaesthetized with urethane (1.25 g/kg, i.p.), tracheotomized and allowed to spontaneously breathe carbogen (95% 0,: 5% CO,)-enriched room air. Then, the right external jugular vein and the right femoral artery were cannulated for the intravenous injection of drugs and the measurement of arterial blood pressure and HR respectively. The Bezold-Jarisch reflex was elicited by bolus intravenous injection of 5-HT (0.5, 1, 2, 4, 8 and 16 pg/kg; as the creatinine sulphate; Sigma, St. Louis, U.S.A.) given at 3-min intervals. The reflexly-mediated reductions in HR and diastolic arterial blood pressure (DABP) were measured. The responses to 5-HT are characterized by immediate falls in HR and DABP, which rapidly return to pre-injection values, followed by a more prolonged secondary depression in arterial blood pressure. Only the immediate HR and DABP responses are associated with activation of the Bezold-Jarisch reflex.” A second doseeresponse curve to 5-HT was constructed in both groups of rats subsequent to vagotomy contralateral to the saline- or excitotoxin-treated sides. Peripheral vagal .stimulation experiment In this study, a functional assessment of the vagal parasympathetic efferent fibres was performed to evaluate the selectivity of the KA-induced destruction of vagal afferent neurons. Additional groups of KA-treated (n = 4) and control (n = 4) rats were prepared as described above with the exception that only the left ganglion was treated. After 7-8 days, the rats were anaesthetized with urethane (I .25 g/kg, i.p.), tracheotomized and allowed to breathe 5% CO,:95% 0, enriched room air. MAP and HR were monitored after cannulation of the left femoral artery. The vagus nerve ipsilateral to the treated NG was sectioned distal to the ganglion and the peripheral end was placed in contact with bipolar platinum electrodes and immersed in mineral oil. HR changes in response to electrical stimulations of the vagus nerve (1 mA, 1 ms duration, I, 5, 10, 15, 20 and 25 Hz for 10 s) were recorded. Histology Immediately following each of the cardiovascular experiments, both NG from each rat were removed under barbiturate or urethane anaesthesia and placed in phosphatebuffered formalin (4% formaldehyde, pH 7.4). Serial paraffin-embedded sections (4pm) of these ganglia were stained with Haematoxylin-Eosin and examined by light microscopy. Two other groups of rats (each n = 4) were treated with KA or a-MA and allowed 48 h to recover before removal of the NG.

Excitotoxin-induced

333

degeneration of rat vagal afferent neurons

Statistics

Statistical differences between means of the baroreceptor reflex parameters (including resting MAP and HR) were determined by Student’s unpaired t-test (two-tailed; Minitab Statistical Package, Pennsylvania State University, University Park, PA, U.S.A.). Cardiopulmonary reflex data were subjected to an initial analysis of variance with repeated measures.32 Subsequent orthogonal partitioning of the sums of squares demonstrated that the relationships describing the DABP and HR-log dose S-HT response curves were best described by linear partitioning. The slopes of each group (divided into pre- and post-vagotomy) and the standard errors (S.E.) of each of the slopes were then determined. The significance of differences between slopes was determined by the equation t(evaluative statistic) = b, - b,,/(S.E.i + SE.:), where b, and b, represent any two slopes, with the Bonferroni adjustment for multiple comparisons.” The significance of differences between the pre- and post-vagotomy responses to each dose (e.g. between dose comparisons) of S-HT were determined by incorporation of the error mean square term for the overall analysis of variance with repeated measures (i.e. all data points included) into Student’s modified l-test formula with the Bonferroni adjustment for multiple comparisons.)’ The group standard errors (divided into pre- and post-vagotomy) were determined by taking the square root of the error mean square term divided by the number of animals in the group.

a

550* 5002 450. z 3; floe . z B 350 . 5 300 250 200I

40

0

*

80

m

3

120

’

’

160

B

4

200

MAP(mn Hg) 550-

b

500. ; 450 * z 3 400 z 2 350 .

Kalnlc Acid

5 300 250 .

RESULTS

Baroreceptor

200-

heart rate reflex parameters

I

40

Figure la shows baroreceptor HR reflex curves obtained from an u-MA-treated and a KA-treated rat, while Fig. 1b shows mean baroreceptor HR reflex curves obtained from the analyses of MAP/HR data from experiments with the KA- and a-MA-treated groups of rats. The slopes of the curves obtained from

KA-treated rats were reduced, indicating that in these rats the gain of the baroreflex was reduced relative to control rats. In KA-treated rats, G was reduced by 51% (P < 0.001) when compared to the control (a-MA-treated) group (Table 1). KA treatment had therefore reduced the reflex HR changes in response to both the pressor and depressor agents. In addition, an appreciable diminution of the baroreceptor HR reflex Range (-43%, P < 0.001) was also found in the KA-treated rats. This change was reflected in an increase in P, (the predominantly vagal arm of the reflex13) (Table 1) from 227 f 8 bts/min to 319 f 11 bts/min (P < 0.001). This indicates that, in the KAtreated rats, reflex bradycardic responses to the pressor agent were reduced. Differences between the KA- and u-MA-treated groups were not found for the remaining baroreflex parameters, resting MAP, HR or body weight. Cardiopulmonary

reflex qctivity

The reflexly-mediated reductions in DABP and HR in response to S-HT (OS-16 pg/kg i.v.) in the KA, NMDA and AMPA experiments prior to, and after, vagotomy contralateral to the saline- and excitotoxintreated side, are shown in Figs 2, 3 and 4, respec-

tively. Slopes describing the dose-response

curves for

.

1

80

*

1

.

120

’

*

4

160

200

160

200

NAP(mn Hg) 550500* 2 450 z > 400* i 5

350 .

Fi 300 . 250 2004n

Rfl

120 MAP (n Hg)

Fig. 1. (a) Baroreceptor HR reflex curves obtained from an a-MA-treated rat (0) and a KA-treated rat (0). Points are actual responses. (b) Mean baroreceptor HR reflex curves of KA-treated (n = 8 rats) and control a-MA-treated (n = 8 rats) groups. Each rat received 15-20 bolus injections of both phenylephrine (l-25pg/kg, i.v.) and sodium nitroprusside (I-SOpg/kg, i.v.) in volumes of I-20~1 (total volume ~250 ~1). (c) Diagrammatic representation of sigmoidal mean arterial pressure-heart rate (MAP-HR) curve describing baroreflex parameters. G, average gain (or sensitivity); P, and P,, upper and lower HR plateaus; Range (P2 -PI); BP~~, MAP value at the midpoint of the range; T, and TL, upper and lower reflex thresholds. Reflex changes in HR were analysed by exponential sigmoidal curve analysis. Non-linear regression utilizing least squares techniques was used to obtain maximum likelihood estimates of parameter values. MAP and HR were related using: HR = P, + Range/(1 + eA(MAP-sPw),HR = P,, A = -4.56 G/Range. Tu=BPSO- 1.317 Rangel4.56 G, T,_= BP% + 1.317 Range/4.56 G.

334

S. J. LEWISCI al.

Table I. Baroreceptor heart rate reflex and resting parameters determined in control (cc-methyl-Dt.-aspartic acid) and kainic acid-treated

Parameter

-4.0 * 227 i 494 * 267 f 116+3 135k5 96 * 96k 109*3 389 k 256 +

G (bts/min mmHg) P, (bts/min) Pz (bts/min) Range (bts/min) npso (mmHg) TU (mmHg) rL (mmHg) Correlation (%) Resting MAP (mmHg) Resting HR (bts/min) Body weight (g) Each

rats

Control (a-MA) 0.2 8 12 17

4

I 7 4

KA -2.0 * 0.2* 319* 11* 412 k 5 153*7* 115+3 139_+2 91*4 94+ 1 109 * 2 417+9 261 +5

Peripheral

value

represents the mean f S.E.M. (n = 8 rats/ group). *P 0.05 for all

0.5 ,

A, Saline

-20 -

E a

-4o-

z d

The bradycardic responses to electrical stimulation of the peripheral end of the vagus in KA-treated and control groups of rats are shown in Fig. 5. Stimulus-response curves in these two groups were virtually identical. Histology

Nodose ganglia removed from rats treated with KA showed a dramatic loss of perikarya (Fig. 6A-C). Figure 6A shows a normal ganglion taken from an cc-MA-treated rat, while Fig. 6B shows a KA-treated ganglion 48 h after treatment. The normal ganglion contains sheets and groups of closely packed ganglion cells, while the KA-treated ganglion is swollen due to an acute inflammatory reaction consisting of oedema and an infiltrate of inflammatory ceils. There is a 5-HT

F .E >

*

2 s w

a

8.0 16.0 I

0.5

5-HT (pg/kg,i.v.) 1.0 2.0 4.0 8.0 16.0

-40 -80 -120 -

-2oo-240L

-80 -

1

2.0

4.0

1.0 1

[I: I -160 -

-60 -

0.5

(lg/kg.i.v.)

0.5 I

post

*

\

6, Kainic acid

vagal stimulation

5-HT (pg/kg,i.v.) 1.0 2.0 4.0 8.0 16.0 , I

O-

B

comparisons). In contrast, in the KA-treated rats the slopes and magnitudes of DABP and HR doseresponse curves prior to contralateral vagotomy were reduced compared to the respective saline-control group. Vagotomy contralateral to the excitotoxintreated side resulted in significant reductions in the slopes of the DABP and HR dose-response curves for each of the excitotoxins when compared to the pre-vagotomy values in these rats and also to the values of the post-vagotomy saline-treated group responses (Table 2).

5-HT (pg/kg,i.v.) 1.0 2.0 4.0 8.0 16.0

I

,

I

I

Fig. 2. The effect of the unilateral application of saline and KA (0.5 nmol/$, 2 x 2~1) to the rat nodose ganglion on the Bezold-Jarisch reflex elicited by 5-HT. Panels A and B show pre- and post-contralateral vagotomy HR and DABP responses for saline-treated and KA-treated rats, respectively. *P < 0.05 compared to pre-vagotomy value. Error bars represent groups standard errors.

,

Excitotoxin-induced 5-HT A, Saline

1.0

0.5

I

of rat vagal

afferent

neurons 5-HT

(pg/kg.i.v.) 2.0

I

degeneration

4.0

8.0 18.0

,

0.5

1.0

0.5

1.0

1

(pg/kg.i.v.)

2.0

I

4.0

8.0 18.0

,

-240-

-80 -

5-HT 0.5

B,NMDA

1.0

I

I

5-HT

(pg/kg,i.v.)

2.0

4.0

8.0 18.0

,

1

(pg/kg,i.v.)

2.0

I

4.0

8.0 16.0

I

0

B

-20

E

-40

s k

-80

a

-80 -100

-240

Fig. 3. The effect of the unilateral application of saline and N-methyl-o-aspartic acid (NMDA; 6.8 nmol/pl, 2 x 2 ~1) to the rat nodose ganglion on the Bezold-Jar&h reflex elicited by S-HT. Panels A and B show pre- and post-contralateral vagotomy HR and DABP responses for saline-treated and NMDA-treated rats, respectively. *P < 0.05compared to pre-vagotomy value. Error bars represent groups standard errors.

A, Saline

5-HT(pg/kg,i.v.)

5-HT (wikg,i.v.) 0.5

1.0

2.0

4.0

8.0

I

I

I

I

I

o-

0.5

16.0

4.0

8.0

0

IF E z g '1

post pre

I a

-6O-

\

post

-80 -120

Pre \

-160 -200

\

-240

-SOL 5HT(pg/kg.i.v.)

5-HT (pg/kg,i.v.)

0.5

1.0

2.0

4.0

8.0

I

I

I

I

I

16.0

1 0

-40I-80I-120I-160, -

0.5

1.0

2.0

4.0

8.0

1

I

I

I

I

16.0

I

post *

-??$

-200I-80L

18.0

-40

-20 -

B,AMPA

2.0

f

*

-40 -

1.0

-240I-

Fig. 4. The effect of the unilateral application of saline and c(-amino-3-hydroxy-5-methyl-4isoxazolepropionic acid (AMPA; 0.54 nmol/pl, 2 x 2 ~1) to the rat nodose ganglion on the Bezold-Jarisch reflex elicited by 5-HT. Panels A and B shown pre- and post-contralateral vagotomy HR and DABP responses for saline-treated and AMPA-treated rats, respectively. *P < 0.05 compared to pre-vagotomy value. Error bars represent group standard errors.

*

S. J. LEWIS et ul.

336

Table 2. Slopes (*SE.) describing the relationship between the reflexly-mediated falls in diastolic arterial blood pressure and heart rate and the log,, dose of .5-hydroxytryptamine (0.5-16 fig/kg i.v.) prior to (Pre) and after (Post) vagotomy contralateral to the saline-. kainic acid-, N-methyl-D-aspartic acid- or a-amino-3-hydroxy-4-isoxazole propionic acid-treated nodose ganglion

Group

Time

Saline I;:‘)

Pre Post Pre

(n = 6) Saline’ (n = 5) NMDA (n = 6) Saline (n = 5) AMPA (n = 6)

Post Pre Post Pre Post Pre Post Pre Post

Slope k S.E. DABP (mmHg/log,, dose 5-HT) (beats/min 46.4 + 2.6 42.3 35.0 k* 3.3 3.3: 8.6 + 2.8*t 48.9 7 2.7 46.1 k 3.3 49.3 k 3.6 26.5 + 3.6*t 43.7 * 2. I 38.8 i 3.5 39.5 i_ 2.6 21.4 i 2.2*t

148.7 & 9.1 151.8 98.1 f+ 8.9 8.5$ 28.7 + 8.5*1 1,46.1 7 8.6 58.4 + 9.1 58.8 + 8.0 70.7 f 10.5*t 52.6 + 10.0 152.1 + 10.1 1,42.2 + 8.8 83.2 k 8.9*1

The number of rats in each group is shown in parentheses. post-vagotomy. tP -c0.05, comparing excitotoxin to $P < 0.05. comparing excitotoxin to saline pre-vagotomy

decrease in the number of ganglion cells, many of which show degenerative change. Ten days after KA treatment ganglion cells are sparse, and the node is contracted and replaced by fibrous tissue (Fig. 6C). Nuclear counts revealed a 40-50% loss of neuronal cell bodies relative to sections from or-MA-treated ganglia, with degenerative changes including a loss of Nissl substance, while axons appeared histologically normal. Perikaryal loss was not seen in cl-MA-treated ganglia. Ganglia treated with either AMPA or NMDA (not shown) also showed loss of cells and the presence of degenerating neurons 7-8 days after application to the ganglion. However, the extent of cell loss

0

5

Frequency(Hz) 10 15 20

25

I

Fig. 5. Bradycardic responses resulting from electrical stimulation (1 ms. I mA, 1-25 Hz, 10 s) of the peripheral end of the vagus ipsilateral to KA- (0) or saline-treated (0) nodose ganglia (n = 4 rats per group).

HR log,,, dose 5-HT)

*P < 0.05, comparing saline values.

post-vagotomy

(3&40%) was not as marked KA treatment.

pre- to values.

as that observed

after

DISCUSSION

The excitotoxins, KA, NMDA and AMPA, when directly superfused on to the NG, produced degeneration of the somata of vagal afferent neurons, including those mediating cardiopulmonary and baroreceptor information. However, only a 30-50% loss in cells was observed, suggesting that some neurons may be resistant to the excitotoxins. This phenomenon has been described for some populations of central neurons.” and cell death/ vulnerability is extremely dependent on mi1ieu.29 Additional studies in KA-treated rats indicated that vagal afferent neurons conveying arterial baroreceptor information to the central nervous system are also affected, without apparent functional impairment of the vagal parasympathetic efferent fibres. This preferential, excitotoxin-induced loss of the sensory perikarya is consistent with the pattern of excitotoxin-induced neuronal degeneration in the brain.24 That is, the intracerebral microinjection of excitotoxins results in damage to neuronal perikarya, whereas the axons of passage and nerve terminals are spared. Although the precise mechanism(s) by which excitotoxins produced their degenerative effects has not been fully established, there is considerable evidence that stimulation of glutamate receptors is

Fig. 6. The effect of KA application (2 nmol/4 ~1 normal saline) onto the rat nodose ganglion. A. Normal ganglion showing sheets and groups of closely packed ganglion cells each with a round eccentric nucleus within which is a compact nucleolus. The cytoplasm contains small granules (Nissl bodies) and around each cell is a row of small supporting satellite cells. The small elongated nuclei of the intervening tissue are those of neurilemmal cells. B. Nodose ganglion 48 h following KA treatment. C. Nodose ganglion 10 days after treatment with KA. Scale bar = 20pm.

338

S. J. LEWISef at.

inherent to the neurotoxic effects of these compounds.29.30 Receptor autoradiography with [‘HIglutamate in our laboratory has recently demonstrated that glutamate receptors appear to be associated with the somata of vagal afferent neuronsi More specifically, ligand displacement experiments have indicated that the kainate, quisqualate and NMDA subtypes of excitatory amino acid receptor appear to be present within the NG19 and these may be the neuroanatomical substrates through which the excitotoxins could initiate degenerative changes within the somata. In the present study, the doses of the excitotoxins applied to the NG were based on those used when administered intracerebrally to produce approximately equal degrees to neurotoxicity.3’ Clearly, the degree of attenuation of Bezold-Jarisch reflex responses in the three groups of excitotoxintreated rats was comparable, although KA appeared to be the most potent of the three agents tested with respect to the degree of reflex attenuation and cell loss, consistent with a predominance of the KA subtype of glutamate receptor.19 These results and the inability of a 15-fold higher dose of the non-excitotoxic dicarboxylic amino acid E-MA, to produce degeneration, support a specific receptor-mediated action of the excitotoxins in the production of neuronal degeneration. There have been no reports, to our knowledge, describing the actions of excitatory amino acids on tat vagal afferent neurons, although these substances are known to excite dorsal root C-fibres.’ However, prolonged treatment of dorsal roots with KA was not found to induce excitotoxicity. In contrast, application of glutamate and aspartate to rat sciatic nerve trunk was found to result in cytotoxicity.‘5 Since the vagal afferents activated by S-HT to elicit the Bezold-Jar&h reflex are of the unmyelinated C-fibre type,” it may at least be concluded that this type of afferent neuron is sensitive to the excitotoxins. The KA-induced reduction in the sensitivity and

range of the baroreceptor HR reflex indicates that this excitotoxin has produced functional changes in the baroreceptor reflex consistent with the histological findings. The partial loss of baroreceptor reflex activity was expected since the procedures used in this study did not remove the glossopharyngeal baroreceptors emanating principally from the carotid sinus regions and which pass through the petrosal ganglia. Indeed, loss of either the vagal or glossopharyngeal baroafferents results in a reduction in the sensitivity of the baroreceptor HR reflex in anaesthetized rabbits, and both vagal and glossopharyngeal baroreceptors are able to compensate fully for the loss of the other with respect to full activation (or withdrawal) of sympathetic, but not vagal outflow to the heart.” Furthermore, in rats, the level of the lower HR plateau is governed mainly by vagal parasympathetic activity, whereas the upper HR plateau is influenced by both sympathetic and parasympathetic activity.” Since the upper HR plateau was unaltered, it would appear that KA does not damage vagal efferent fibres. This conclusion is supported by the finding that the bradycardic responses following electrical stimulation of the peripheral end of the vagus nerve were not affected by prior treatment of the ipsilateral NG with KA.

CONCLUSION

In summa~, the technique of excitotoxin application to the NG represents a convenient and novel method of inducing primary sensory deafferentation in a rat model and as such represents the first demonstration of excitotoxin-induced destruction of peripheral sensory neurons. Acknowted~emePts-This work was supported by a Programme Grant from the National Health and Medical Research Council of Australia. The authors thank MS L. Newnham and Mr C. Bradley for their assistance.

REFERENCES

I. Agrawal S. G. and Evans R. H. (1986) The primary afferent depolarizing action of kainate in the rat. Br. J. Pharmac.

87, 3455355. 2. Bond S. M., Cervero F. and McQueen D. S. (1982) Influence of neonatally administered capsaicin on baroreceptor and chemoreceptor reflexes in the adult rat. Br. J. Pharmac. 77, 517-524. 3. Britt M. D. and Wise R. A. (1981) Kainic acid spares fibers of the dorsal noradrenergic bundle. Bruin Res. Buff. 7, 437440. 4. Coleridge H. M., Coleridge J. C. G. and Kidd C. (1964) Role of the pulmonary arterial baroreceptors in the effect produced by capsaicin in the dog. J. Physic& Lo-d. 170, 272-285. 2nd edn (ed. Lajtha A.), pp. 299-329. 5. Coyle J. T. (1985) Neuron-specific cytotoxins. In Handbook c$Neurochemistry, Plenum Press, New York. 6. Donaghue S., Garcia M., Jordan D. and Spyer, K. M. (1982) Identification and brain stem projections of aortic baroreceptor afferent neurones in nodose ganglia of cats and rabbits. J. Physiol., Lond. 322, 337.-352. 7. Faber J. E. and Brody M. J. (1983) Reflex hemodynamic responses to superior laryngeal nerve stimulation in the rat. 1. Auton. Nero. Sysi. 9, 607-622.

8. Furness J. B., Elliott J. M., Murphy R., Costa M. and Chalmers J. P. (1982) Baroreceptor reflexes in conscious guinea-pigs are unaffected by depletion of cardiovascular substance P nerves. Neurosci. Let?. 32, X35-290. 9. Gamse R., Leeman S., Holzer P. and Lembeck F. (1981) Differential effects of capsaicin on the content of somatostatin, substance P and neurotensin in the nervous system of the rat. Nuunyn-Schmiedeherg’s Arch. Pharmac. 317, 140-148. IO. Guo G. B., Thames M. D. and Abboud F. M. (1982) Differential baroreflex control of heart rate and vascular resistance in rabbits. Circulat. Res. 50. 554-565.

Excitotoxin-induced

degeneration of rat vagal a&rent neurons

339

11. Gwyn D. G., Leslie R. A. and Hopkins D. A. (1979) Gastric afferents to the nucleus of the solitary tract in the cat. Neurosci. L&t. 14, 13-17. 12. Haun C. K. and Alexander N. (1985) Pulmonary edema and death induced by sinoaortic denervation in fastigial nucleus-lesioned rats. Bruin Res. 343, 89-94. 13. Head G. A. and McCarty R. (1987) Vagal and sympathetic components of the heart rate range and gain of the baroreceptor heart rate reflex in conscious rats. J. A&on. New. Syst. 21, 203-213. 14. Holzer P. (1988) Local effector functions of capsaicin-sensitive sensory nerve endings: involvement of tachykinins, calcitonin gene-related peptide and other neuropeptides. Neuroscience 24, 739-768. 15. Hugon J., Vallat J. M. and Leboutet M. J. (1987) Cytotoxic properties of glutamate and aspartate in rat peripheral nerves: histological findings. Neurosci. Left. 81, Id. 16. Kalia M. and Sullivan J. M. (1982) Brainstem projections of sensory and motor components of the vagus nerve in the rat. J. camp. Neurol. 211, 248-264. 17. Krieger E. M. (1964) Neurogenic hypertension in the rat. Circular. Res. 15, 511-521. 18. Krieger E. M. and Marseillan R. F. (1963) Aortic depressor fibres in the rat: an electrophysiological study. Am. J. Physiol. 205, 771-774. 19. Lewis S. J., Cincotta M., Verberne A. J. M., Jarrott B., Lodge D. and Beart P. M. (1987) Receptor autoradiography

with [3H]L-glutamate reveals the presence and axonal transport of glutamate receptors in vagal afferent neurones of the rat. Eur. J. Pharmac. 144, 413415. 20. Maggi C. A. and Meli A. (1988) The sensory-efferent functions of capsaicin-sensitive sensory neurones. Gen. Pharmac. 19, l-43. 21. McCubbin J. W., Masson G. M. C. and Page J. H. (1958) Aortic depressor nerves of the rat. Archs. Int. Pharmucodyn. Th&. 114, 303-306. 22. Mei N. (1966) Existence d’une separation anatomique des fibres vagales efferentes et afferentes au niveau du ganglion plexiforme du chat. Etude morphologique et electrophysiologique. J. Physiol., Paris 58, 253-254. 23. Mei N. and Dussardier M. (1966) Etude des lesions pulmonaries produites par la section des fibres sensitives vagales. J. Physiol., Paris 58, 427431. 24. Olney J. W. (1978) Neurotoxicity of excitatory amino acids. In Kainic Acid us a Tool in Neurobiology (eds McGeer

G., Olney J. W. and McGeer P. L.), pp. 95-121. Raven Press, New York. 25. Palkovits M. and Zaborsky L. (1977) Neuroanatomy of central cardiovascular control. Nucleus tract solitarii: afferent and efferent neuronal connections in relation to the baroreceptor reflex arc. Prog. Bruin Res. 47, l-34. 26. Portalier P. and Vigier D. (1979) Localization of aortic cells in the nodose ganglion by HRP retrograde transport in the cat. Neurosci. Left. 11, 7-11. 27. Porter J. P. and Brody M. J. (1986) A comparison of the hemodynamic effects produced by electrical stimulation of subnuclei of the paraventricular nucleus. Bruin Res. 375, 20-29. 28. Randich A. and Maixner W. (1984) [D-Ala2]-methionine enkephalinamide reflexively induces antinociception by activating vagal afferents. Phurmuc. Biochem. Behuu. 21, 441448. 29. Rothman S. M. and Olney J. W. (1986) Glutamate and the pathophysiology of hypoxic-ischemic brain damage. Ann. Neural. 19, 105-l Il. 30. Rothman S. M. and Olney J. W. (1987) Excitotoxicity and the NMDA receptor. Trends Neurosci. 10, 299-302. 3 1. Schwarz R., Scholz D. and Coyle J. T. (1978) Structure-activity relations for the neurotoxicity of kainic acid derivatives and glutamate analogues. Neurophurmucology 17, 145-I 5 1. 32. Snedecor G. W. and Cochran W. G. (1980) Sturistical Methods, 7th edn, pp. 267-269. Iowa State University Press, Ames, IA. 33. Thoren P. (1979) Role of cardiac vagal C-fibres in cardiovascular control. Rev. Physiol. Biochem. Phurmuc. 86, l-94. 34. Verberne A. J. M.. Lewis S. J., Jarrott B. and Louis W. J. (1987) Bezold-Jarisch reflex is inhibited bv excitotoxininduced destruction of vagal primary afferents neurones. Eu;. J. iharmuc. 139, 365-367. 35. Verberne A. J. M., Lewis S. J., Worland P. J., Beart P. M., Jarrott B., Christie M. J. and Louis W. J. (1987) Medial prefrontal cortical lesions modulate baroreflex sensitivity in the rat. Bruin Res. 426, 243-249. 36. Von Euler C., Hayward J. N., Martilla I. and Wyman R. J. (1973) Respiratory neurons of the ventrolateral nucleus of the solitary tract of the cat: vagal input, spinal connections and morphological identification. Bruin Res. 61, l-22. 37. Wallenstein S.. Zucker C. L. and Fleiss J. L. (1980) Some statistical methods useful in circulatine research. Circulut Res. 47, 1-9. 38. Wiley R. G., Blessing W. W. and Reis D. J. (1982) Suicide transport: destruction of neurons by retrograde transport of ricin, abrin and modeccin. Science 216, 889-890. (Accepted

13 September

1989)