66

Brain Research, 533 (1990) 66-72 Elsevier

BRES 16032

Cardiac chronotropic organization of the rat insular cortex Stephen M. Oppenheimer 1"3 and David F. Cechetto 1-3 1Department of Stroke and Aging, Robarts Research Institute, London, Ont. (Canada), Departments of 2physiology and 3Clinical Neurological Sciences, University of Western Ontario, London, Ont. (Canada) (Accepted 29 May 1990)

Key words: Insular cortex; Heart rate; Electrocardiogram; Cardiac arrhythmia

Clinical evidence implicates the cerebral cortex in the genesis of ECG changes and cardiac arrhythmias. Such findings are not infrequent following acute cortical stroke and during partial seizures. Electrical stimulation of the cerebral cortex, however, only rarely and inconsistently results in cardiac changes. When encountered, attendant alterations in blood pressure and respiration occur; consequently, it is unclear whether the cardiac effects are primary or secondary to these. Phasic insular cortex microstimulation linked to the ECG cycle, a new technique, elicits only heart rate effects, eliminating confounding variables. The insular cortex was chosen for study because of its profuse autonomic and limbic connectivity. Cardiac chronotropic sites were demonstrated in 37 chloralose-anesthetized rats, with tachycardia represented in the rostral posterior insula, and bradycardia in the caudal posterior insula. Both effects were abolished by atenolol but not by atropine, implying their mediation by respective increases or decreases in sympathetic activity. This is the first report of the demonstration of a cortical region wherein stimulation affects heart rate and no other parameter.

INTRODUCTION Clinical observations implicate the cerebral cortex in cardiac control: E C G repolarization changes and cardiac arrhythmias have been noted following acute cortical strokes 2'7A7 and during epileptic seizures 1'23. Similar alterations have been demonstrated in experimental models of epilepsy: a temporal relationship occurs between E C G changes, cardiac sympathetic nerve activity, and ictal and interictal spikes l ° ' n . Such arrhythmias may play an important role in sudden death following acute neurological lesions, and in unexpected death during epileptic seizures. Despite the wealth of clinical evidence, previous studies involving electrical stimulation of the mammalian cerebral cortex have rarely been successful in producing either E C G changes or cardiac arrhythmias 6'8"2L22. This contrasts with the relative ease of their elicitation from subcortical regions, including the hypothalamus and brainstem 9'~4'24. In all of the studies of cortical chronotropicity, the stimulus was applied as a large constant voltage, and not as a constant current, producing uncertainty as to the extent of current spread and making precise localization of the response difficult. Likewise, stimulation-induced blood pressure and respiratory changes in unventilated preparations, introduced uncer-

tainty as to whether the observed cardiac effects were primary, or secondary. Using a new technique of phasic cortical microstimulation synchronized with the R wave of the E C G , we have been successful in producing cardiac effects without associated autonomic or respiratory changes. Recent research has implicated the insular cortex in autonomic contro14'ls'19; consequently this area was chosen for investigation, as being likely to contain a cardioregulatory center. A preliminary communication of this work has been published in abstract form 16. MATERIALS AND METHODS There were 3 series of experiments. In the first of these, 16 male Wistar rats were initially anesthetized with 7% chloral hydrate (350 mg/kg); anesthesia was maintained throughout the experiment with a-chloralose. The animals were frequently tested to determine the need for supplementation, and to maintain a constant level of anesthesia. The femoral artery and vein were cannulated, and the arterial cannula connected to a Statham P23D volumetric pressure transducer for measurements of the blood pressure. Heart rate was obtained from the pulse pressure, using a Grass 7P44 tachograph. An endotracheal tube was inserted into each animal, which inspired 100% oxygen. Respiration was measured with a Fleisch pneumotachograph. Blood pressure, heart rate and respiration were recorded by a Grass RPS 7C8 polygraph. The animals' core temperatures were maintained at 37 °C with a rectal thermistor and a heating pad. The animals were placed in a stereotactic frame, and the parietal

Correspondence: D.F. Cechetto, Department of Stroke and Aging, Robarts Research Institute, 100 Perth Drive, London, Ont. N6A 5K8, Canada.

67 In a third series of experiments, 6 male Wistar rats were prepared surgically as detailed above, and underwent pharmacological manipulation to determine the nature of the effector neurotrammitter(s). Atropine (0.3 mg/kg) i.v. was injected after obtaining a chronotropic response to ascertain cholinergic components; likewise, atenolol (5 mg/kg) i.v. was injected to block sympathetic effects. At the end of each experiment, the animals were perfused under deep anesthesia using an intracardiac catheter. The perfusate comprised a solution of 0.9% normal saline, followed by 10% formalin. The brains were removed and stored in 10% formalin for at least 2 days, before being cut in the coronal plane on a microtome. The sections were then stained with thionin and examined using a Leitz Diaplan microscope. Camera lucida tracings of the stimulation sites were thus obtained.

bone drilled. A glass microelectrode filled with 3 M saline, of internal tip diameter 10-20 gm, was inserted into the left insula using co-ordinates previously determined from recording and stimulating experiments in this region4'5. The ECG was monitored in the lead II configuration, and relayed to a signal discriminator which differentiated the R wave from other parts of the ECG waveform. An event-related trigger pulse was therefore generated, and used to activate a stimulator. Paired pulses were delivered to the microelectrode, which could be arranged to occur at any point in the ECG cycle. The pulses were separated by 20 ms and were generated coincident with the R wave. Allowing for central conduction delay4, this was likely to result in stimuli reaching the heart during the P wave of the next ECG cycle. At this point, effects on the heart rate would be maximized. With this method, it was possible to arrange stimuli coincident with either each ECG cycle, or any multiple thereof. This abolished the effects of stimulation on blood pressure or respiration. The stimulus parameters were: 2 ms pulse duration, 500 gA or less current. The duration of the train of pulses varied from 45 to 90 s. ReSlXmses were ascertained at progressive 200-/~m steps of microelectrode advancement. In this fashion, a map of the chronotropic organization of the insular cortex was compiled. In another series of experiments, 15 male Wistar rats were anesthetized with ketamine-xylazine 0.1 ml/100 g i.m. (ketamine 100 mg/ml, xylazine 20 mg/ml), and maintained under anesthesia with a continuous ketamine infusion. Single sites were explored as above for their chronotropic effects. A comparison of their distribution was made with that obtained from animals under chloralose anesthesia.

RESULTS A c h r o n o t r o p i c r e s p o n s e to s t i m u l a t i o n was d e t e r m i n e d if a m i n i m u m s u s t a i n e d c h a n g e in h e a r t r a t e o f 5 beats/min

(bpm)

occurred,

t e m p o r a l l y r e l a t e d to the

stimulus, in t h e a b s e n c e of c o n c o m i t a n t a l t e r a t i o n s of b l o o d p r e s s u r e o r respiration. T h e s e effects w e r e elicited by p a i r e d stimuli s e p a r a t e d by 20 ms c o i n c i d e n t with e a c h

ECG

BP U

HR mu

i

Stim. CC

Resp.

I "--"--'--V

I

CPu

AC

I1

GI

mll

Pir

\

|

zsb

c: °.

, c7

HR 2min BP ~bpm~ (ram HgD Fig. 1. Microelectrode track through a typical tachycardia site showing development of the response. The first column shows the first and then each subsequent 20th ECG trace following the onset of insular microstimulation for each site. The arrows denote maximal shift in the R wave during the development of the response. Heart rate and blood pressure recordings are displayed in the second column. The final column depicts respiration during microstimulation. Horizontal bars represent the stimulus duration; different time scales apply to the latter two columns. AC, anterior commissure; AI, agranular insular cortex; CC, corpus callosum; DI, dysgranular insular cortex; CPu, caudate/putamen; EC, external capsule; En, endopiriform nucleus; GI, granular insular cortex; LV, lateral ventricle; Pir, piriform cortex; BP, blood pressure; HR, heart rate; Resp., respiration; Stim., stimulus marker. 0

68 ECG BP Resp. HR

Stim.

I

[50 nl

HR

(bpm)

i

2 min

O

BP

(mmHg)

Fig. 2. Microelectrode track through a typical bradycardia site. Explanation of the columns is as for Fig. 1. HBO, nucleus of the horizontal limb of the diagonal band; M, medial preoptic nucleus; Thai, thalamus. Other abbreviations as for Fig. 1.

ECG cycle (53% of stimulations), or every second (7%) or fourth ECG cycle (40%). Tachycardia responses generated an average rise in heart rate for all such sites of 15 bpm (range 5-30 bpm). The requisite current strength was generally 500/~A, although at some sites a threshold of 200/~A was obtainable. The mean latency of onset of tachycardia was 8 s (range 0-30 s). The response attained a peak level after 33 s (range 6-132 s), and lasted 138 s (range 60-360 s). It outlasted the stimulus by 61 s (range 0-285 s). Fig. 1 demonstrates a typical track through a tachycardia site within the posterior granular insular cortex. The first row of panels shows the first and each subsequent 20th ECG cycle from the onset of stimulation. No change in ECG morphology occurred despite the increase in heart rate (denoted by arrows indicating the maximum temporal shift in the R wave from the onset of stimulation). Continuous monitoring of blood pressure and respiration revealed no change in these parameters during the period of heart rate increase. This figure also illustrates the discrete localization of the response; microstimulation at sites along the same track, 200 ~m above or below the region of peak tachycardia, elicited

little if any change in heart rate. At bradycardia sites, the mean decrement in heart rate was 16 bpm (range 5-40 bpm). The response developed within 9 s (range 0-60 s), and reached a peak within 77 s (range 12-300 s). The mean duration was 320 s (range 90-960 s) and the response outlasted the application of the stimulus by a mean of 230 s (range 0--870 s). Fig. 2 shows a track through a bradycardia site using the same format as before. No alteration in ECG morphology is apparent, despite the change in heart rate. In order to ascertain the specificity of the technique, rats were stimulated at 2, 5, 7, 10, 15 and 20 Hz frequencies, independent of the ECG cycle. The lower frequencies elicited no consistent response, and at the higher ranges (15 Hz and above) blood pressure changes accompanied alterations in heart rate. These latter were of shorter duration and lesser amplitude than those seen with insular phasic microstimulation. When a delay between the R wave and the stimulus trigger was imposed, such that the majority of sympathetic activity in the cardiac nerves was anticipated to coincide with the T wave, no heart rate changes were elicitable. Combining the results for 23 responsive tracks (16

69

+ 1.0

+ 0.5

0.0

-0.5

Fig. 3. Map of cardiac chronotropic sites within the insular cortex (chloralose anesthesia). The numbers beneath each section refer to its location with reference to the bregma (0.0). Q, tachycardia sites; A, bradycardia sites; O, unresponsive sites. Abbreviations are as for Figs. 1 and 4. tachycardia and 7 b r a d y c a r d i a ) , and 4 negative tracks in 22 animals (including the 6 rats used in the n e u r o p h a r -

macological experiments), a c h r o n o t r o p i c m a p of the rat insular cortex was o b t a i n e d and comprises Fig. 3;

BID

HR ---

8tim

A.

B.

~

Pir

+0-5

4"0-0

HR

2rain

BP

(bpm) (nlmHg 0 Fig. 4. Map of cardiac chronotropic sites within the insular cortex (ketamine anesthesia). Typical responses from a tachycardia site (A), and a bradycardia (B) site are shown. Cl, claustrum; F, fornix; 3V, third ventricle. Other abbreviations as for Fig. 1.

70

A

B

I

C

Pre-Drug

BP HR

t Atropine

1I Atenolol

I 0

1oo

i 250

r

50

0 2 min "-0 HR BP (bpm) (mmHg)

Fig. 5. Effect of anticholinergic drugs and E-blockers on cardiac chronotropic responses. The first two columns show the first and each subsequent 20th ECG during stimulation prior to the injection of drugs, and following atropine and atenolol, at a taehycardia site (A) and a bradycardia site (B). The arrows denote maximal R wave shift during the response. The final column (C) shows typical blood pressure and heart rate effects prior to, and during drug manipulation at a tachycardia site. tachycardia sites exist within the rostral posterior insular cortex, whereas a discrete region within the caudal posterior insula engenders bradycardia. In ketamine anesthetized rats, a more circumscribed, borderzone region between the tachycardia and bradycardia sites was explored. Fig. 4 demonstrates that in this particular area, as with chloralose anesthesia, bradycardia responses are obtained from superficial, cortical locations. The tachycardia responses derive from deeper sites, probably due to excitation of fibers coursing caudally from more rostral regions. In order to ascertain the possible effector neurotransmitter(s) involved in the tachycardia response, 3 rats were anesthetized with chloralose, and injected intravenously or intra-arterially with atenolol after location of an appropriate site. Atenolol decreased the heart rate by a mean of 70 bpm (range 50-90 bpm) and abolished the tachycardia response to insular stimulation in all 3 animals (Fig. 5A). In two of the 3 rats, atropine was injected prior to atenolol (which was administered when the anticholinergic effects had abated) producing a mean heart rate elevation of 15 bpm. No effect of this drug upon the tachycardia response to insular stimulation was

documented (Fig. 5A). At bradycardia sites in two rats, atropine produced a mean rise in heart rate of 35 bpm (range 30-40 bpm), and did not affect the bradycardia response (Fig. 5B). Atenolol decreased the heart rate by 110 bpm (range 90-130 bpm) in 3 rats. The bradycardia effect was abolished in all of these animals (fig. 5B). DISCUSSION Phasic stimulation of the rat insular cortex linked to the R wave of the E C G resulted in purely cardiac chronotropic effects. Tachycardia is represented in the rostral posterior insular region, and bradycardia in the caudal posterior insula. In these latter regions, tachycardia responses could also be obtained from medial sites within, or adjacent to, the white matter. Such stimulation likely excites fibers arising in the more rostral posterior cortical insular regions, which pass caudally as part of the effector pathway leaving the insular cortex 25. In the case of animals anesthetized with ketamine, the borderzone area between bradycardia and tachycardia zones was investigated. A similar picture emerged, with

71 bradycardia elicitable from the cortical regions at this site, and tachycardia from the deeper, white matter. Recently, pressor and depressor sites have been identified within the insular cortex, and their distribution mapped 25. Pressor areas are represented in the rostral posterior insular cortex, and the depressor zone is more caudally placed. There are similarities, therefore, to the distribution of the cardiac chronotropic sites. As yet, there is no indication whether chronotropic and pressor effects derive from stimulation of the same cell by different means, or if two overlapping populations of cells exist, subserving different modalities, and with a similar rostrocaudal distribution. The optimum current frequency for the elicitation of blood pressure effects is 80 Hz 19. Phasic stimulation coincident with either each ECG cycle or every fourth cycle, represents frequencies of approximately 7 of 2 Hz, respectively. These are subthreshold for the elicitation of blood pressure changes both in our experiments and elsewhere 19. In our experiments, such frequencies when not phase-related to the R wave of the ECG generally did not result in a response. Stimulation of the chronotropic sites with 50 Hz current, produces either no effect at all on heart rate, or accompanying changes in blood pressure. The frequency characteristics of the cardiac chronotropic response may be determined by the time constant of the mediating pathways, such that stimulation at a frequency related to the average heart rate is required to dissect out chronotropic responses from the blood pressure effects that would otherwise accompany them. Alternatively, an oscillator may exist within the insular cortex, which requires stimulation at heart raterelated frequencies in order to become entrained. It might also be that such stimulation mimics a central oscillator without the need for entrainment. The minimal current strength necessary for the production of either chronotropic effect was generally 500 /tA. This likely indicates the requirement to entrain a large number of ceils which are spatially dispersed within the respective tachycardia or bradycardia sites. Despite the comparative size of the current, persuasive argument for the discrete localisability of the responses comes from the observation of their development as the electrode was advanced in 200-/am steps through the sites. Although the amplitudes of both the bradycardia and tachycardia responses were similar, they differed in their temporal characteristics. The time for the peak bradycardia response to develop was twice that for the corresponding tachycardia parameter. Likewise, the duration of the bradycardia response was 2.5 times that of the tachycardia effect; it outlasted the application of the stimulus 4 times as long as did the tachycardia response. These differences may be explained by disparities in the

connectivity of the two sites as have been documented recently for the insular pressor and depressor regions 25. No gross change in ECG morphology occurred during stimulation at any of the sites, and no cardiac arrhythmias were encountered. This could reflect the relative stability of the preparation under chloralose anesthesia. Similarly, it might also reflect that, in order to maximize changes in heart rate, the cardiac stimulus was adjusted to coincide with the P wave. Earlier investigations have indicated a central conduction delay of between 90 and 120 ms from the onset of stimulation of the insular cortex to the onset of activity in the renal nerve 4. A similar delay was considered likely for the onset of cardiac sympathetic nerve activity after insular stimulation, and cortical stimulation arranged accordingly. Preliminary studies have now demonstrated that phasic insular stimulation generating cardiac activity coincident with the T wave, produces a lethal arrhythmia in urethane-anesthetized rats TM. Both bradycardia and tachycardia produced by insular stimulation were abolished by the fl-blocker, atenolol. This suggests that the sympathetic nervous system is involved in both responses. In the case of bradycardia, this appears anomalous. However, one explanation is that atenolol acts centrally at an inhibitory sympathetic synapse. The parent fibers would arise in the caudal posterior insula, and exert an inhibitory effect either on the more anterior tachycardia sites, or at a lower level such as the lateral hypothalamic area. Here they would converge on the excitatory pathway and inhibit it. A further explanation implies that atenolol acts peripherally; both the bradycardia and tachycardia responses are mediated by the sympathetic nervous system and are consequently blocked by the drug. Some evidence for this exists: tachycardia is associated with an elevation of plasma norepinephrine levels, whereas stimulation of bradycardia sites results in a decrease (Oppenheimer et al., in preparation). The inability of atropine to abolish the chronotropic effect, implies that cholinergic mechanisms are unlikely to be involved in either response. It could be argued that incomplete blockade occurred. This is unlikely in view of the increase in heart rate produced by the injection. Likewise, the dosage employed was 10 times that commonly recommended for the production of cholinergic blockade in rats undergoing surgical procedures 3. Cerebral mechanisms have been thought to play a role in sudden death and the sudden epileptic death syndrome 12'13 (for review see ref. 17). Previously, no evidence has been obtained for a cortical site involved exclusively in cardiac control without concomitant blood pressure or respiratory effects. The demonstration of such an area within a part of the brain which has profuse

72 connectivity with autonomic centers and a marked visceral afferent input 4'2°'25 could offer a means of explaining these clinical phenomena.

Acknowledgements. This work was supported by the Heart and Stroke Foundation of Ontario. S.M.O. is a Fellow of the Canadian Heart Foundation; D.EC. is a Canadian Heart Foundation Scholar.

REFERENCES

zures: analysis of 66 cases, Epilepsia, 25 (1984) 84-88. 13 Lown, B., DeSilva, R.A. and Lenson, R., Roles of psychologic stress and autonomic nervous system changes in provocation of ventricular premature complexes, Am. J. Cardiol., 41 (1978) 979-985. 14 Malmo, R.B. and Mundl, W.J., Cardiovascular and respiratory responses to electrical stimulation of the midbrain in the rat, Int. J. Psychophysiol., 1 (1983) 75-81. 15 Mesulam, M.-M. and Mufson, E.J., Insula of the old world monkey. III. Efferent cortical output and comments on function, J. Comp. Neurol., 212 (1982) 38-52. 16 Oppenheimer, S.M., Hachinski, V.C. and Cechetto, D.E, Cardiac chronotropic organisation of the rat insula, Soc. Neurosci. Abstr., 15 (1989) 595. 17 Oppenheimer, S.M., Cechetto, D.F. and Hachinski, V.C., Cerebrogenic cardiac arrhythmias: cerebral ECG influences and their role in sudden death, Arch. Neurol., 17 (1990) 513-519. 18 Oppenheimer, S.M., Hachinski, V.C., Wilson, J.X. and Cechetto, D.F., The insula and cardiac arrhythmias: implications for stroke, Stroke, 21 (1990) 174. 19 Ruggiero, D.A., Mraovitch, S., Granata, A.R., Anwar, M. and Reis, D.J., A role of insular cortex in cardiovascular function, J. Comp. Neurol., 257 (1987) 189-207. 20 Saper, C.B., Convergence of autonomic and iimbic connections in the insular cortex of the rat, J. Comp. Neurol., 210 (1982) 163-173. 21 Showers, M. and Crossby, E., Somatic and visceral responses from the cingulate gyrus, Neurology, 8 (1958) 561-565. 22 Ueda, H., Arrhythmias produced by cerebral stimulation, Jap. Circ. J., 26 (1962) 225-230. 23 VanBuren, J.M., Some autonomic concomitants of ictal automatism, Brain, 81 (1958) 505-528. 24 Weinberg, S.J. and Fuster, J.M., Electrocardiographic changes produced by localized hypothalamic stimulations, Ann. Int. Med., 53 (1960) 332-341. 25 Yasui, Y., Breder, C.D., Saper, C.B. and Cechetto, D.E, Autonomic responses and efferent pathways from the insular cortex in the rat, J. Comp. Neurol., in press.

1 Blumhardt, L.D., Smith, P.E.M. and Owen, L., Electrocardiographic accompaniments of temporal lobe epileptic seizures, Lancet, ii (1986) 1052-1055. 2 Burch, G.E., Meyers, R. and Abildskov, J.A., A new electrocardiographic pattern observed in cerebrovascular accidents, Circulation, 9 (1954) 719-723. 3 Guide to the care and use of experimental animals, Vol. 1, Canadian Council on Animal Care, 1980, p. 96. 4 Cechetto, D.E and Chen, S.J., Subcortical sites mediating sympathetic responses from insular cortex in rats, Am. J. Physiol., 258 (Regulatory Integrative Comp. Physiol.) (1990) R245-R255. 5 Cecheno, D.E, Saper, S.B., Evidence for a viscerotopic sensory representation in the cortex and thalamus in the rat, J. Comp. Neurol., 262 (1987) 27-45. 6 Deigado, J.M.R., Circulatory effects of cortical stimulaton, Physiol. Rev., 40 (1960) 146-170. 7 Dimant, J, and Grob, D., Electrocardiographic changes and myocardial damage in patients with acute cerebrovascular accidents, Stroke, 8 (1977) 448-455. 8 Kenedi, I. and Csanda, E., Electrocardiographic changes in response to electrical stimulation of the cerebral cortex, Acta Physiol. Acad. Sci. Hung., 16 (1959) 165-173. 9 Korteweg, G.C.J., Boeles, J.T.E and TenCate, J., Influence of stimulation of some subcorticai areas on electrocardiogram, J. Neurophysiol., 20 (1957) 100-107. 10 Lathers, C.M. and Schraeder, P.L., Autonomic dysfunction in epilepsy: characterization of autonomic cardiac neural discharge associated with pentylenetetrazol-induced epileptogenic activity, Epilepsia, 23 (1982) 633-647. 11 Lathers, C.M., Schraeder, P.L. and Weiner, F.L., Synchronization of cardiac autonomic neural discharge with epileptogenic activity: the lockstep phenomenon, Electroenceph. Clin. Neurophysiol., 67 (1987) 247-259. 12 Leestma, J.E., Kaleikar, M.B., Teas, S.S., Jay, G.W. and Hughes, J.R., Sudden unexpected death associated with sei-