- Email: [email protected]
Plasma catecholamines associated with hypothalamically-elicited defense behavior
Physiology&Behavior,Vol. 36, pp. 867-873. Copyright©Pergamon Press Ltd., 1986. Printed in the U.S.A.
0031-9384/86 $3.00 + .00
Plasma Catecholamines Associated With Hypothalamically-Elicited Defense Behavior SUSAN L. STODDARD, .1 VALERIE K. BERGDALL,* DOUGLAS W. TOWNSENDt AND BARRY E. LEVIN¢
*Department of Anatomy, Indiana University School of Medicine, Fort Wayne, IN 46805 tDepartment of Mathematical Sciences, Indiana University-Purdue University at Fort Wayne Fort Wayne, IN 46805 and ~Department of Neurosciences, University of Medicine and Dentistry of New Jersey Newark, NJ 07103 R e c e i v e d 8 J u l y 1985 STODDARD, S. L., V. K. BERGDALL, D. W. TOWNSEND AND B. E. LEVIN. Plasma catecholamines associated with hypothalamically-elicited defense behavior. PHYSIOL BEHAV 36(5) 867-873, 1986.--Sympatho-adrenal (SA) activation was determined by measuring levels of norepinephrine (NE) and epinephrine (E) in bilateral adrenal venous and peripheral venous plasma of 20 anesthetized cats following stimulation of medial hypothalamic sites. Hypothalamic sites were selected that elicited affective defense behavior in the freely moving cat. Fifty-eight percent of these hypothalamic sites elicited a bilateral increase ~ 10 ng/min in the output of both adrenal catecholamines (CAs); these increases were greater from the gland ipsilateral to the side of stimulation. Other SA responses included both preferential increases or decreases in either NE or E. Under baseline conditions, an average of 67% of the NE in the peripheral venous plasma was contributed by the sympathetic noradrenergic nerves; hypothalamic stimulation at "defense" sites increased the contribution to 75%. The data suggest that hypothalamic regions that elicit defense behavior may overlap with regions that activate the adrenal medullary and sympathetic nervous systems. Cat Defense behavior Hypothalamic stimulation Norepinephrine Epinephrine Sympathetic nerves
Adrenal vein
A F F E C T I V E defense behavior (pupillary dilatation, piloerection, growling and hissing) is one of several aggressive behaviors that can be elicited by stimulation of specific regions in the hypothalamus of the cat [ 12,44]. Other aggressive behaviors include flight [14,36] and predatory, or quiet-biting, attack [38,44]. Since hypothalamically-elicited aggression is empirically similar to the naturally-occurring behavior [25], it has been used as a model to investigate neural structures that modulate emotional behavior [1, 27, 29, 38, 40]. Little, however, has been reported concerning the peripheral physiological correlates of these centrallyelicited behaviors. Cannon's report [5] that the "fight or flight" response is accompanied by sympatho-adrenal activation suggests that measurement of sympatho-adrenal (SA) parameters, (activation of the adrenal medulla and the sympathetic noradrenergic nerves) would be a meaningful approach to evaluating some of the physiological responses that are concomitants of centrally-elicited aggressive behavior. Previous investigations have used hypothalamic stimulation in cats and dogs to study output from the adrenal medulla [3, 10, 11, 15, 16, 30, 34, 35] or sympathetic nervous
Plasma catecholamines
activity [21]. In general, these studies used an exclusively anatomical approach, mapping the hypothalamus for areas which activated the SA system. Although this work has contributed much to our understanding of the central anatomy of the autonomic nervous system, it was not the purpose of these studies to correlate behaviors related to particular hypothalamic regions with coincident SA activation. Additionally, a number of these studies were done before the advent of sensitive assays for catecholamines (CAs). Thus bioassay techniques were used to quantify the output of adrenal CAs, and adrenal venous blood was collected over several minutes. Since the half-lives of adrenal CAs are quite short (1-2 min for E and 2.5--4 min for NE) [2, 23, 45], this extended sampling period, combined with an assay procedure that is less sensitive than current methodologies, has been unable to yield information about the hypothalamic control of minute-to-minute changes in adrenal medullary activity. The present study was undertaken as an initial step in correlating hypothalamically-mediated behavior with concomitant SA activation. Hypothalamic sites were identified which elicited affective defense behavior in the awake, freely
1Requests for reprints should be addressed to Dr. S. L. Stoddard, Indiana Univ. School of Medicine, 2101 Coliseum Blvd. E., Fort Wayne, IN 46805.
867
STODI)ARI) Ii'7 A I
868 moving cat. These behaviorally identified sites were subsequently stimulated and blood samples were collected from both adrenal and peripheral veins. Measurement of CA levels in these samples was used to differentiate adrenal medullary from sympathetic neural components of the SA response. Heart rate and blood pressure were continuously monitored to permit comparison between the activity of the SA and cardiovascular (CV) systems. Preliminary results of this study have previously been reported in abstract form [39]. METHOD
Elicitation of Aggressive Behavior Twenty adult cats (16 females and 4 males; mean BW =2.5 kg) were used in this study. During aseptic surgery, electrode guides were stereotaxically mounted on the skull overlying the medial hypothalamus (anterior: 9.0--12.5; lateral: 1.0-2.5). Additionally, a series of bolts was anchored to the skull for the purpose of securing the cat in a headholder during electrode placement and brain stimulation [28,29]. At least one week after stereotaxic surgery, the awake, unanesthetized animal was restrained in the headholder; calibrated, monopolar, stainless steel electrodes, 0.6 mm in diameter and bared at the tip for 0.5 ram, were lowered through the guides in 0.25 mm steps until stimulation of the hypothalamus elicited a hiss [40]. Each hypothalamic site was stimulated (Grass $88) at 60 Hz with biphasic, square wave pulses of 1 msec half-cycle duration. To approximate constant current conditions the constant voltage output of the stimulator was led through two stimulus isolation units (Grass SIU5) and a pair of 40,000 ohm resistors placed in series with the leads to the cat, and was monitored with an oscilloscope. The following day, the behavior elicited by stimulation of each electrode was evaluated by placing the experimental cat and a " t a r g e t " cat in an observation box. Electrodes were selected that elicited affective defense behavior: hissing, pupillary dilatation, retraction of the ears, and piloerection. The stimulated cat fixed its gaze on the target cat and either backed into a corner or struck with its paw at the target cat. The intensity of stimulation which reliably elicited affective defense within 10 sec was defined as the "behavioral intensity," and ranged from 0.2-0.8 mA (peak-to-peak).
NE
E NE
E NE
A 150
~
AI2.0
E
A 12.5
J
A,.5 aHd
i FIG. 1. Composite map of stimulation sites in the hypothalamus which elicited defense in the freely moving cat. Sites are plotted on drawings of coronal sections from the atlas of Jasper and AjmoneMarsan [17]. Symbols indicate the type of adrenal medullary response elicited by stimulation of each site at the behavioral intensity. Changes in NE output are on the left of each drawing; changes in E are on the right. Shaded circle: increase in CA output />10 ng/min; clear circle: increase in CA output/>0 but < 10 ng/min; circle with dots: decrease in CA output. Each circle is divided into halves: the left half shows the response from the adrenal gland ipsilateral to the side of hypothalamic stimulation; the right half shows the response from the contralateral adrenal gland. Half circles show the adrenal medullary responses in the 5 cats that had only the ipsilateral adrenolumbar vein cannulated (see the Method section). Abbreviations used: AC, anterior commissure; arid, dorsal hypothalamic area; F, fornix; Ha, anterior hypothalamic area; Hdm, dorsomedial nucleus; Hp, posterior hypothalamus; Mm, medial mammiUary nucleus; mt, mammillothalamic tract; OC, optic chiasm; OT, optic tract; PVH, paraventricular nucleus; SCh, suprachiasmatic nucleus: SON, supraoptic nucleus; vm, ventromedial nucleus.
and maintained between 101°-102°F with heating pads. Following the surgical procedures the animal was heparinized and allowed to stabilize for at least 30 min before blood sampling commenced.
Cannulation Procedure The following day the animal was anesthetized with sodium pentobarbital (47 mg/kg). Cannulae were placed in the femoral artery for the continuous monitoring of heart rate (HR) and mean arterial pressure (MAP), and in the atrial appendage through the left external jugular vein for the withdrawal of peripheral blood samples. Additionally both adrenolumbar veins [26] were exposed through a mid-ventral abdominal incision and directly cannulated with PE 90. (In the cat the adrenal vein exits from the hilus of the adrenal gland and joins with the adrenolumbar vein, which passes over the adrenal gland, draining the posterior body wall and adrenal vein. The adrenolumbar vein joins the inferior vena cava on the right and either the renal vein or inferior vena cava on the left. There is no " a d r e n o l u m b a r " vein in the human; rather this vessel is termed the adrenal vein.) In five of the 20 cats only the adrenolumbar vein ipsilateral to the side of hypothalamic stimulation was cannulated; the contralateral adrenolumbar vein was ligated at the junction with the inferior vena cava. The body temperature was monitored
Collection of Venous Blood Samples In the anesthetized animal, each behaviorally identified hypothalamic site (N=31) was stimulated for 30 sec at the intensity which evoked behavior when awake. F o r each stimulation, continuous sequential blood samples were taken concurrently from the adrenolumbar vein(s) and from the atrium during four l-rain periods: two prior to stimulation and two following stimulus onset. Blood was allowed to drip from the adrenolumbar cannulae into heparinized tubes kept on ice; the volume of each l-rain sample was measured. Seventy-one percent of the sites (N=22) were stimulated twice: once at the behavioral intensity, and once a t a standard, higher intensity (0,8 mA). Following each stimulation period, the volume of blood collected was replaced with an equal volume of Macrodex (dextran 70; Pharmacia). At the completion of the experiment, a small lesion was made via each electrode (10/zA for 3 rain). The animal was then sacrificed and perfused; the location of each electrode tip was determined through standard histological procedures [41].
PLASMA CATECHOLAMINES AND DEFENSE
Catecholamine Assay Plasma was immediately separated after the blood samples were drawn, and the proteins precipitated by the addition of 20 tzl 5 N perchloric acid to each 480/zl plasma. The supernatants of a 20,000 g centrifugation for 15 rain were stored at -70°C until assayed for N E and E within 1-2 months. NE and E were assayed in 50 txl aliquots of the perchloric acid supernatant by a sensitive radioenzymatic technique utilizing catechoi-O-methyl transferase and S-adenosyl methionine [methyl-3H] as a methyl donor [24].
Data Analysis The adrenal medullary activation which accompanied hypothalamically-elicited defense behavior was determined by comparing the absolute changes in the output of NE and E (ng/min) during the first minute following stimulus onset to the baseline mean. CA output was quantified by multiplying the level in ng/ml by the blood flow rate (ml/min); this calculation was chosen to incorporate observed variations in adrenolumbar venous flow rate. Since the baseline levels of CAs were variable both across and within cats, we wished to select a parameter for the evaluation of adrenal medullary activity that was independent of baseline values. Thus, absolute changes in CA levels were used rather than percent changes since, with the exception of E levels in the ipsilateral adrenal gland, absolute changes in CAs were not significantly correlated with baseline values. In all cases, however, percent changes in CA levels were significantly correlated with baseline values. Cardiovascular activation was determined by comparing the maximum changes in MAP and HR during stimulation to the baseline mean. Correlations between sympatho-adrenal and cardiovascular variables were evaluated using Spearman's rank-order correlation coefficient; these correlation coefficients were judged to be significant if the associated p value was 0.05 or less.
Peripheral Venous NE To determine sympathetic nervous system activity, it was necessary to estimate the proportion of N E present in peripheral (atrial) blood contributed by the sympathetic noradrenergic nerves. E was also present in the peripheral plasma. As the primary source of E is the adrenal medulla, we assumed that a proportion o f peripheral NE, also, was contributed by the adrenal medulla. Since both adrenal veins were cannulated or tied off, the amount of E in the peripheral plasma relative to that in the bilateral adrenal venous plasma was considered to represent that proportion of adrenal CA output which had " l e a k e d " from the adrenal medulla into the peripheral circulation, perhaps via extra-hilar accessory veins, as have been described for humans [20], or via the lymphatic system [42]. Extra-adrenal chromaffin cells (paraganglia) [32], or accessory adrenal glandular tissue (unpublished observations) may also have contributed E to the peripheral plasma. A proportion of the adrenal N E output equal to the proportion of " l e a k e d " E was subtracted from the total peripheral NE to yield an estimate of peripheral NE which was contributed by the sympathetic noradrenergic nerves: Peripheral N E from nerve terminals = Total peripheral N E - ((Peripheral E/Adrenal venous E) × Adrenal venous NE). RESULTS Stimulation of 31 hypothalamic sites elicited affective de-
869 TABLE 1 NUMBER OF HYPOTHALAMIC DEFENSE SITES THAT EVOKED VARIED PATTERNS OF INCREASES IN BILATERAL ADRENAL CATECHOLAMINE OUTPUT AT BEHAVIORAL (0.2-0.8 mA) AND MAXIMAL(0.8 mA) INTENSITIES OF STIMULATION Intensity of Stimulation Increases in CA output* Bilateral '~NE and E Bilateral '~E and unilateral 1'NE Bilateral '~NE and unilateral I'E Bilateral I'E only Bilateral I'NE only Unilateral TNE and E Unilateral I'E only Unilateral I'NE only
Behavioral Maximal (N=24) (N=20) 9 5 3 1 2 1 1 1
11 5 2 1 0 0 1 0
*Increases were defined as changes in CA output/>10 ng/min.
fense behavior in the awake cat. The location of these sites is shown in Fig. 1. Effective sites were located within the stereotaxic coordinates A 11.0-13.5, lat 0.25-2.0, and included the anterior hypothalamic area, the pefifornical area, the dorsomedial nucleus, and the ventromedial nucleus. F o r analysis the data were separated into those responses elicited by the behavioral intensity of stimulation, which was variable, and the uniform maximum stimulation intensity.
Baseline Adrenal CA Output The mean resting secretion of CA from a single adrenal gland (N = 100) +_ the standard error of the mean (SEM) was 104.0_+12.3 ng/min for N E and 39£-+13.2 ng/min for E. Thus, under baseline conditions, N E formed an average (_+SEM) of 72.2+2.2% of the total CAs released from the adrenal gland (range 12.2-97.7%).
Activation of the Adrenal Medulla Seven hypothalamic sites were stimulated in five cats in which only the adrenolumbar vein ipsilateral to the side of stimulation was cannulated. The results obtained from stimulation of these sites (shown in Fig. 1) supported results obtained from bilateral adrenolumbar cannulations. Therefore, since the results from bilateral cannulations were more complete, only these data will be discussed below. Hypothalamic stimulation at sites that elicited defense behavior significantly changed the relative proportions of NE and E which comprised the CA output of the adrenal medulla. At the behavioral intensity of stimulation N E comprised 58.9+2.8% of the CAs in the plasma sample following stimulus onset; at the maximum intensity of stimulation, the percentage of N E decreased further to 47.3_+ 3.(1%. It should be noted that even though hypothalamic stimulation significantly altered the proportion of CAs released by the adrenal medulla, the absolute levels of NE were still greater than the levels of E following approximately half of all stimulations. Various patterns of increases in adrenal CA output were seen following stimulation at hypothalamic defense sites (Table 1 and Fig. I). As shown in Table 1, the most frequent adrenal response was a bilateral increase (/>10 ng/min) in
" ; , I ( ) I ) I ) A R I ) t,~ 4i
~70
Behavioral Intensity
Maximal Behovioral
Intensit),Intensif)~ A 15.0
Maximal Behavioral Intensity Itensit~/
A
Maximal Intensity
~ I5
<:SZ) vzsT>
J
AI 1 . 0 ~
cantly greater from the ipsilateral gland than flom the con tralateral gland (mean-+SEM=61.6+13.0 ng/min w, 42.3-+9.6 ng/min, respectively). Similarly, the increases in NE from the ipsilateral gland were greater than the contralateral increases, but the differences were not significant (mean_+SEM=61.4_ 15.4 ng/min vs. 36.9-+13.3 ng/min, respectively). This significant difference between ipsilateral and contralateral adrenomedullary CA output was also apparent at the maximum stimulation intensity only for E. For all stimulations, the changes in bilateral adrenal NE output resulting from hypothalamic stimulation were significantly, but weakly, correlated (r-0.5} with the changes in bilateral adrenal E output.
Activation o f the Peripheral Noradrem'r.eic Ncrvu~
FIG. 2. Composite map showing mcreases in the level of adjusted peripheral NE, suggesting activation of the sympathetic nervous component of the SA response, following stimulation of hypothalamic defense sites. Increases are expressed as percent of baseline mean: clear circle: 0-1 lff%; one quarter shaded circle: 111-120%; top half shaded circle: 121-130%; three quarter shaded circle: 131145%; shaded circle: > 145%. Responses at the behavioral intensity of stimulation (0.2-0.8 mA) are shown on the left of each drawing; responses at the maximal intensity of stimulation (0.8 mA) are shown on the right. Only stimulations are shown that permitted measurement of changes in NE of peripheral origin (see the Method section). Abbreviations are the same as for Fig. 1.
both NE and E. Within this response type, the mean increases ( - S E M ) in adrenal NE and E output were 94.1 _+15.5 ng/min and 101.9_+ 14.2 ng/min respectively. Adrenal responses were also evaluated by comparing the increases in N E output to the increases in E output. Nine sites were identified that elicited greater absolute increases in NE, while eight other sites elicited greater absolute increases in E. There was a general pattern to the location of these two types of sites. Within the dorsomedial nucleus of the hypothalamus twice as many sites (4 vs. 2) evoked increases in N E greater than the increases in E. Additionally, within the ventromedial nucleus of the hypothalamus, sites which elicited greater increases in NE were generally located ventral to those sites which evoked greater increases in E. The exceptions to this were two sites located between the dorsolateral border of the ventromedial nucleus and the medial border of the column of the fornix at A 11.5; these sites elicited greater increases in NE. Although stimulation of hypothalamic defense sites usually increased the output o f adrenal CAs, stimulation of some sites evoked a decrease in the output of one or both adrenal CAs (Fig. 1). At the behavioral intensity of stimulation, a decrease in N E occurred bilaterally at 2 sites (Fig. 1: A 11.5 and 11.0), ipsilaterally at 3 sites (Fig. 1: A 12.5, 11.5, 11.0) and contralaterally at 1 site (Fig. 1: A 13.0). A decrease in E output occurred only in the ipsilateral gland following stimulation of 3 hypothalamic sites (Fig. 1: A 12.5 and 11.5). Stimulation of hypothalamic defense sites resulted in differences between the CA output of the glands ipsilateral and contralateral to the side of stimulation. At behavioral intensities of stimulation, the increases in E were signifi-
Calculations for adjusted peripheral NE permitted comparisons between baseline and post-stimulation NE levels in approximately 60% of the cases (Fig. 2), i.e., trials in which the " l e a k a g e " of CAs from the adrenal glands was not so great as to obscure any changes in NE from noradrenergic nerves. Under baseline conditions the mean amount (-+SEM) of peripheral NE contributed by sympathetic nerves was 66,9_+4.4%. Stimulation of hypothalamic defense sites at all intensities increased the mean amount of peripheral N E contributed by the sympathetic nerves to 75.1 _+2.8%. Activation of the sympathetic component of the SA response was arbitrarily defined as levels of adjusted peripheral NE following hypothalamic stimulation greater than 145% of the baseline mean [35]. Using this criterion, activation of the sympathetic nervous (SN) system accompanied behavioral stimulation of 30% of the defense sites and maximal stimulation of 63% of the sites (Table 2). Generally, increasing the intensity of stimulation from behavioral to maximal also increased the degree of activation of the SN system (compare right and left sides of Fig. 2). SN system activation was most often a concomitant of hypothalamically-elicited defense when the behavior was elicited by stimulation of the ventromedial nucleus (Fig. 2: A 12.0, 11.5 and 11.0) and its rostral projections (Fig. 2: A 13.5, 13.0, and 12.5).
('omparisott of SA and ('V Parameters The mean (-+SEM) of the maximal increases in HR and MAP during the 30 sec period of hypothalamic stimulation at behavioral intensity were 11.4_+2.4 beats/rain and 9.3_+2.3 mm Hg, respectively. These mean changes in CV parameters were relatively small despite the considerable output of CAs. The maximal changes in HR and MAP during hypothalamic stimulation were tested for correlation with the changes in adrenal CA output and adjusted peripheral NE. Stimulation at behavioral intensities elicited increases in ipsilateral adrenomedullary E output which were significantly, but weakly, correlated (r=0.44) with changes in HR. DISC USSION
Electrical stimulation of hypothalamic sites which elicited affective defense behavior in the freely moving cat also evoked a variety of sympatho-adrenal responses. Such variation has been reported previously following central nervous system stimulation both in acute, anesthetized preparations [10, 16, 34, 35], and in the awake animal [41]. However, the most frequent concomitants of SA activation which accompanied hypothalamically-elicited defense behavior were
PLASMA CATECHOLAMINES AND DEFENSE
871
TABLE 2 NUMBER OF HYPOTHALAMICDEFENSE SITES WHICH INCREASED THE LEVEL OF ADJUSTED PERIPHERAL NE (SUGGESTING SYMPATHETICNERVOUS SYSTEM ACTIVITY) AT BEHAVIORAL(0.2-0.8 mA) AND MAXIMAL(0.8 mA) INTENSITIES OF STIMULATION Intensity of Stimulation Behavioral (N=20)
Maximal (N= 19)
100-110
1
0
111-120 121-130 131-145 >145
2 3 2 6
0 1 2 12
% Baseline*
*Level of adjusted peripheral NE in the first atrial blood sample following the onset of hypothalamic stimulation expressed as percent of the baseline mean.
(1) a bilateral increase in the output of adrenal medullary NE and E; (2) a significant change in the ratio of CAs released from the adrenal medulla; (3) preferential activation of the ipsilateral adrenal medulla; and (4) activation of the sympathetic noradrenergic nerves.
Adrenal Medullary Component of the SA Response Baseline CA output. Adrenal CA output data from the present study were converted to ng/kg/min to permit comparison with previous studies that have reported adrenal CA output in cats [6, 7, 9, 10, 16, 18, 35]. These studies, which employed a variety of anesthetics and CA assay techniques, reported CA output from only the left adrenal gland. The variations in methodology, anesthetics and assay techniques yielded a wide range of average values for baseline CA output (E: 3.1-63.0 ng/kg/min; NE: 1.4--200.0 ng/kg/min). The mean resting levels (-+SEM) of bilateral adrenal CA output of 41.3_+4.8 ng/kg/min for NE and 15.9_+5.5 ng/kg/min for E obtained in the current study were within the range of values reported by other authors. Further, our results were in accord with the majority of other investigators that NE is the primary CA released from the feline adrenal medulla at rest [4, 10, 16, 18, 31, 35]. Activation associated with hypothalamically-elicited defense. Stimulation of hypothalamic sites which elicited defense behavior also generally elicited increases in both NE and E. These results are in partial agreement with previous reports. Katsuki et al. [19] and Briicke et al. [4] found medial hypothalamic stimulation in cats to result in marked increases only in E secretion. More recently Robinson et al. [35] found that electrical stimulation in the dorsal hypothalamic area elicited a significant increase in both adrenal CAs. Francke et al. [11] reported that hypothalamic and mesencephalic stimulation in cats elicited increases in adrenal CAs that were usually greater in the gland ipsilateral to the side of stimulation. Our results concurred with this in that stimulation of hypothalamic defense sites elicited a mean increase in adrenal E which was significantly greater from the ipsilateral gland. The mean increase in ipsilateral NE was also greater, although the difference was not significant.
In the present study the relative increase (percent above baseline) in output of adrenal E was nearly always greater than the relative increase in NE. In spite of the differences in proportional change of the adrenal CAs, the absolute increase of N E was greater than that of adrenal E in approximately half the cases. This response, in addition to the fact that NE was the main CA released from the resting adrenal in our study, resulted in post-stimulation values of N E that were frequently higher than E. Such observations suggest that, in the cat, NE is the predominant neurotransmitter and neurohumor. The question of selective secretion of NE or E from the adrenal medulla remains. The difficulty in evaluating others' data arises from the interpretation of the term " s e l e c t i v e " relative to the release of either of the adrenal CAs. Previous reports agree that both peripheral and central stimulation, in addition to various physiological and pharmacological stimuli, can alter the ratio of CAs released from the cat's adrenal medulla [6, 9, 16, 18, 31, 34, 35, 37, 43]. Results from the current study concur with this; hypothalamic stimulation which elicited defense behavior significantly altered the ratio of CAs. The difficulty in evaluating a " s e l e c t i v e " increase results from the criteria for measuring changes in CA output. In the current study, the relative increase in E was nearly always greater than that of N E , in large part due to the low resting levels of E. However, evaluation of the data in terms of absolute increases in CA output revealed that a greater output of NE was elicited as frequently as a greater output of E. Thus, based on absolute increases in adrenal CAs, the present investigation revealed instances of the "selective" secretion of each E and NE. Our further observation that the increases in output of adrenal N E and E were weakly, but significantly, correlated also suggests that there may be partially independent mechanisms modulating the release of these catecholamines. Using a different criterion, the relative increase in CA output, Robinson et al. [35] also'found evidence for the "selective" increase in each CA following hypothalamic stimulation.
Sympathetic Nervous Component of the SA Response Baseline levels of peripheral NE. Under baseline conditions, approximately two-thirds of the NE in the peripheral plasma was contributed by the sympathetic noradrenergic nerves; this was similar to the results in rats that 70% of the peripheral NE is overflow from the sympathetic nerve terminals [22,33]. Activation associated with hypothalamically-elicited defense. Hypothalamic sites which elicited defense behavior also effected an overall increase in the activity of the sympathetic nervous component of the SA re.sponse, as determined by the adjusted levels of peripheral NE. Stimulation at behavioral intensity increased the activity of the SN system to 110-145% baseline levels at 37% of the sites, and to more than 145% baseline at 30% of the sites. Stimulation at maximum intensity caused not only a greater absolute increase in peripheral NE, but also elevated the activity of the SN system to more than 145% baseline at 60% of the sites. It was of interest that the sites where stimulation caused the greatest activation of the SN system were related to the ventromedial nucleus. Stimulation of this hypothalamic region reliably elicits affective defense behavior [12]. Further, the pathway from the ventromedial nucleus which mediates this behavior involves an initial rostral projection that synapses in the anteromedial hypothalamus [13]. Compari-
872
ST(II)I)ARI) ¢ i A!
son between the anatomical maps presented by Fuchs e t a / . [12,13] and our Fig. 2 suggests that the central system which activates the sympathetic nervous component of the SA response is coincident, at least within the hypothalamus, with efferent pathways from the ventromedial nucleus. C a r d i o v a s c u l a r P a r a m e t e r s A s s o c i a t e d With I t y p o t h a l a m i c a l l y - E l i c i t e d D<['ense
A significant correlation between CV and SA parameters was found only between HR and ipsilateral adrenomedullary E output. In fact, only relatively minor changes were noted in CV parameters compared to the degree of activation of the adrenal medullary and sympathetic neural components of the SA response. These observations suggest that hypothalamic sites which elicit defense behavior do not concomitantly activate the CV and SA systems. The current data support previous suggestions [8,41], that there appear to be regions in the hypothalamus which differentially activate the sympatho-adrenal and the cardiovascular systems. CONCLUSIONS In the present study, hypothalamically-elicited defense behavior was most frequently accompanied by activation of both the adrenal medullary and sympathetic nervous components of the SA response. Activation of the sympathetic nervous component of the SA response was reflected in in-
creased levels o1' NE in the peripheral plasma; Ihc concomkant cardiovascular measurements did not reliably reflect this sympathetic nervous activity. In spite of the anatomical spread of hypothalamic sites which elicited defense behavior, it appears that the central systems which mediate this behavior are largely coincident with pathways that activate the adrenal medulla, and frequently overlap with areas that activate the peripheral noradrenergic nerves. However, because defense behavior can be elicited from various hypothalamic regions, the concomitant SA responses ma~, vat'?,. Thus, use of this centrally-elicited behavior as a model lbr aggression or emotional behavior may no! be appropriate in all cases.
ACKNOWLEDGEMENTS The authors wish to thank Ms. M. Culpepper for her technical assistance, Mr. S. Stoddard-Apter for his assistance in programming and data analysis, and Ms. E. Wilson for her expert typing. The comments of Drs. J. A. Vilensky and S. W. Carmichael in critically evaluating earlier drafts of the manuscript are appreciate& This project was supported in part by Grant Number 5-S07-RR-5371, awarded by the Biomedical Research Support Grant Program, Division of Research Resources. National Institutes of Health (SLS), a grant from the Harry Frank Guggenheim Foundation ~SLS) and by the Research Service of the Veterans Administration (BELL
REFERENCES
1. Adamec, R. E. and C. 1. Stark-Adamec. Limbic control of aggression in the cat. Prog Neuropsychopharmacol Biol Psychiatry 7: 505-512, 1983. 2. Axelrod, J., H. Weil-Malherbe and R. Tomchick. The physiological disposition of Ha-epinephrine and its metabolite metanephrine. J Pharmacol Exp Ther 127: 251-256, 1959. 3. Boddin, M. and W. Dierick. Catecholamines in blood during direct stimulation of the hypothalamus in dogs. Arch Int Physiol Biochim 82: 895--906, 1974. 4. Briicke, F., F. Kaindl and H. Mayer. Ober die Ver~inderung in der Zusammensetzung des Nebennierenmarkinkretes bei elektrischer Reizung des Hypothalamus. Arch Int Pharmacodyn Ther g$: 407-412, 1952. 5. Cannon, W. B. The emergency function of the adrenal medulla in pain and the major emotions. Am J Physiol 23: 356--372, 1914. 6. Critchley, J. A. J. H., P. Ellis and A. Ungar. The reflex release of adrenaline and noradrenaline from the adrenal glands of cats and dogs. J Physiol (Lond) 298: 71-78, 1980. 7. Eade, N. R. and D. R. Wood. The release of adrenaline and noradrenaline from the adrenal medulla of the cat during splanchnic stimulation. Br J Pharmacol 13: 390--394, 1958. 8. Ferguson, R. W., B. Folkow, M. G. Mitts and E. C. Hoff. Effect of cortical stimulation upon epinephrine activity. J Neurophysiol 20: 32%339, 1957. 9. Feuerstein, G. and Y. Gutman. Preferential secretion of adrenaline or noradrenaline by the cat adrenal in vivo in response to different stimuli. Br J Pharmacol 43: 764-775, 1971. 10. Folkow, B. and U. S. von Euler. Selective activation of noradrenaline and adrenaline producing cells in the cat's adrenal gland by hypothalamic stimulation. Circ Res 2: 191-195, 1954. 11. Francke, P. F., J. L. Culberson, S. W. Carmichael and R. L. Robinson. Bilateral secretory responses of the adrenal medulla during stimulation of hypothalamic or mesencephalic sites. J Neurosci Res 8: 1--6, 1982. 12. Fuchs, S. A. G., H. M. Edinger and A. Siegel. The organization of the hypothalamic pathways mediating affective defense behavior in the cat. Brain Res 330: 77-92, 1985.
13. Fuchs, S. A. G., H. M. Edinger and A. Siegel. The role of the anterior hypothalamus in affective defense behavior elicited from the ventromedial hypothalamus of the cat. Brain Re.s 330: 93-107, 1985. 14. Fuchs, S. A. G. and A. Siegel. Neural pathways mediating hypothalamically elicited flight behavior in the cat. Brain Res 306: 263-281, 1984. 15. Goldfien, A. and W. F. Ganong. Adrenal medullary and adrenal cortical response to stimulation of diencephalon. Am J Physiol 202: 205-211, 1962. 16. Grant, R., P. Lindgren, A. Rosen and B. Uvnas. The release of catechols from the adrenal medulla on activation of the sympathetic vasodilator nerves to the skeletal muscles in the cat by hypothalamic stimulation. Acta Physiol S t a n d 43: 135-154, 1958. 17. Jasper, H. H. and C. A. Ajmone-Marsan. Stereotaxic Atlas" q f the Diencephahm o f the Cat. Ottawa: National Research Council of Canada, 1954. 18. Kaindl, F. and U. S. von Euler. Liberation of nor-adrenaline and adrenaline from the suprarenals of the cat during carotid occlusion. Am J Physiol 166: 284-288, 1951. 19. Katsuki, S., T. Ikemoto, H. Shimada, F. Hagiwara and J. Kanal. The functional relationship between the hypothalamus and the anterior pituitary-adrenal system. Endocrinot Jpn 2: 303-312, 1955. 20. Khalique, A. and W. S. Monkhouse. The adrenal veins of man. J Anat 135: 846, 1982. 21. Kollai, M. and K. Koizumi. The mechanisms of differential control in the sympathetic system studied by hypothalamic stimulation. J Auton Nerv Syst 2: 377-389, 1980. 22. Kvetnansky, R., V. K. Weise, N. B. Thoa and I . J. Kopin. Effects of chronic guanethidine treatment and adrenal medullectomy on plasma levels of catecholamines and corticosterone in forcibly immobilized rats. J Pharmacol Exp Ther 2tl9: 287-291. 1979.
PLASMA CATECHOLAMINES
AND DEFENSE
23. Levin, B. E. and O. R. Hubschmann. Dorsal column stimulation effect on human cerebrospinal fluid and plasma catecholamines. Neurology 30: 65-71, 1980. 24. Levin, B. E., J. Triscari and A. C. Sullivan. Abnormal sympatho-adrenal function and plasma catecholamines in obese Zucker rats. Pharmacol Biochem Behav 13: 107-113, 1980. 25. Leyhansen, P. Cat Behavior. New York: Garland STPM Press, 1979. 26. McDonald, L. E. Veterinary Endocrinology and Reproduction, 3rd edition. Philadelphia: Lea & Febiger, 1980. 27. MacDonnell, M. F., S. H. Brown and B. Davy. Hyperexcitability in the neural substrate of emotional behavior in cats after alcohol withdrawal. J Stud Alcohol 36: 1480-1492, 1975. 28. MacDonnell, M. F., L. Fessock and S. H. Brown. Ethanol and the neural substrate for affective defense in the cat. Q J Stud Alcohol 32: 406--419, 1971. 29. MacDonnell, M. F. and S. Stoddard-Apter. Effects of medial septal stimulation on hypothalamically-elicited intraspecific attack and associated hissing in cats. Physiol Behav 21: 679-683, 1978. 30. Magoun, H. W., S. W. Ranson and A. Hetherington. The liberation of adrenin and sympathin induced by stimulation of the hypothalamus. Am J Physiol 119: 615-622, 1937. 31. Malmejac, J. Activity of the adrenal medulla and its regulation. Physiol Rev 44: 186--218, 1964. 32. Mascorro, J. A. Anatomical and morphological observations on the juxtaposition between the inferior mesenteric ganglion and paraganglion chromaflin cells in cats. Anat Rec 202: 121A, 1982. 33. Micalizzi, E. R. and D. T. Pals. Evaluation of plasma norepinephrine as an index of sympathetic neuron function in the conscious, unrestrained rat. Life Sci 24: 2071-2076, 1979. 34. Redgate, E. S. and E. Gelihorn. Nature of sympathetico-adrenal discharge under conditions of excitation of central autonomic structures. Am J Physiol 174: 475--480, 1953. 35. Robinson, R. L., J. L. Culberson and S. W. Carrnichael. Influence of hypothalamic stimulation on the secretion of adrenal medullary catecholamines. J Auton Nerv Syst 8: 89-96, 1983.
873 36. Romaniuk, A. Representation of aggression and flight reactions in the hypothalamus of the cat. Acta Biol Exp (Warsaw) 25: 177-186, 1965. 37. Rubin, R. P. and E. Miele. A study of the differential secretion of epinephrine and norepinephrine from the laerfused cat adrenal gland. J Pharmacol Exp Ther 164:115-121, 1968. 38. Siegel, A. and H. Edinger. Neural control of aggression and rage behavior. In: Handbook of the Hypothalamus, Vol 3, Part B, Behavioral Studies of the Hypothalamus, edited by P. J. Morgane and J. Panksepp. New York: Marcel Dekker, 1981, pp. 203-240. 39. Stoddard-Apter, S. L., V. Bergdall and B. E. Levin. Catecholamine levels in adrenal venous and peripheral plasma following stimulation of behaviorally identified hypothalamic sites in the cat. Soc Neurosci Abstr 9:111, 1983. 40. Stoddard-Apter, S. L. and M. F. MacDonnell. Septal and amygdalar efferents to the hypothalamus which facilitate hypothalamically-elicited intraspeciflc aggression and associated hissing in the cat. An autoradiographic study. Brain Res 193: 19-32, 1980. 41. Stoddard-Apter, S. L., A. Siegel and B. E. Levin. Plasma catecholamine and cardiovascular responses following hypothalamic stimulation in the awake cat. J Auton Nerv Syst 8: 343360, 1983. 42. Trautmann, A. and J. Fiebiger. Fundamentals of the Histology of Domestic Animals. Ithaca, NY: Comstock Publishing Associates, 1952. 43. Ungar, A. and J. H. Phillips. Regulation of the adrenal medulla. Physiol Rev 63: 787-843, 1983. 44. Wasman, M. and J. P. Flynn. Directed attack elicited from hypothalamus. Arch Neuro/6: 220-227, 1962. 45. Whitby, L. G., J. Axelrod and H. WeiI-Malherbe. The fate of H3-norepinephrine in animals. J Pharmacol Exp Ther 132: 193201, 1961.
0031-9384/86 $3.00 + .00
Plasma Catecholamines Associated With Hypothalamically-Elicited Defense Behavior SUSAN L. STODDARD, .1 VALERIE K. BERGDALL,* DOUGLAS W. TOWNSENDt AND BARRY E. LEVIN¢
*Department of Anatomy, Indiana University School of Medicine, Fort Wayne, IN 46805 tDepartment of Mathematical Sciences, Indiana University-Purdue University at Fort Wayne Fort Wayne, IN 46805 and ~Department of Neurosciences, University of Medicine and Dentistry of New Jersey Newark, NJ 07103 R e c e i v e d 8 J u l y 1985 STODDARD, S. L., V. K. BERGDALL, D. W. TOWNSEND AND B. E. LEVIN. Plasma catecholamines associated with hypothalamically-elicited defense behavior. PHYSIOL BEHAV 36(5) 867-873, 1986.--Sympatho-adrenal (SA) activation was determined by measuring levels of norepinephrine (NE) and epinephrine (E) in bilateral adrenal venous and peripheral venous plasma of 20 anesthetized cats following stimulation of medial hypothalamic sites. Hypothalamic sites were selected that elicited affective defense behavior in the freely moving cat. Fifty-eight percent of these hypothalamic sites elicited a bilateral increase ~ 10 ng/min in the output of both adrenal catecholamines (CAs); these increases were greater from the gland ipsilateral to the side of stimulation. Other SA responses included both preferential increases or decreases in either NE or E. Under baseline conditions, an average of 67% of the NE in the peripheral venous plasma was contributed by the sympathetic noradrenergic nerves; hypothalamic stimulation at "defense" sites increased the contribution to 75%. The data suggest that hypothalamic regions that elicit defense behavior may overlap with regions that activate the adrenal medullary and sympathetic nervous systems. Cat Defense behavior Hypothalamic stimulation Norepinephrine Epinephrine Sympathetic nerves
Adrenal vein
A F F E C T I V E defense behavior (pupillary dilatation, piloerection, growling and hissing) is one of several aggressive behaviors that can be elicited by stimulation of specific regions in the hypothalamus of the cat [ 12,44]. Other aggressive behaviors include flight [14,36] and predatory, or quiet-biting, attack [38,44]. Since hypothalamically-elicited aggression is empirically similar to the naturally-occurring behavior [25], it has been used as a model to investigate neural structures that modulate emotional behavior [1, 27, 29, 38, 40]. Little, however, has been reported concerning the peripheral physiological correlates of these centrallyelicited behaviors. Cannon's report [5] that the "fight or flight" response is accompanied by sympatho-adrenal activation suggests that measurement of sympatho-adrenal (SA) parameters, (activation of the adrenal medulla and the sympathetic noradrenergic nerves) would be a meaningful approach to evaluating some of the physiological responses that are concomitants of centrally-elicited aggressive behavior. Previous investigations have used hypothalamic stimulation in cats and dogs to study output from the adrenal medulla [3, 10, 11, 15, 16, 30, 34, 35] or sympathetic nervous
Plasma catecholamines
activity [21]. In general, these studies used an exclusively anatomical approach, mapping the hypothalamus for areas which activated the SA system. Although this work has contributed much to our understanding of the central anatomy of the autonomic nervous system, it was not the purpose of these studies to correlate behaviors related to particular hypothalamic regions with coincident SA activation. Additionally, a number of these studies were done before the advent of sensitive assays for catecholamines (CAs). Thus bioassay techniques were used to quantify the output of adrenal CAs, and adrenal venous blood was collected over several minutes. Since the half-lives of adrenal CAs are quite short (1-2 min for E and 2.5--4 min for NE) [2, 23, 45], this extended sampling period, combined with an assay procedure that is less sensitive than current methodologies, has been unable to yield information about the hypothalamic control of minute-to-minute changes in adrenal medullary activity. The present study was undertaken as an initial step in correlating hypothalamically-mediated behavior with concomitant SA activation. Hypothalamic sites were identified which elicited affective defense behavior in the awake, freely
1Requests for reprints should be addressed to Dr. S. L. Stoddard, Indiana Univ. School of Medicine, 2101 Coliseum Blvd. E., Fort Wayne, IN 46805.
867
STODI)ARI) Ii'7 A I
868 moving cat. These behaviorally identified sites were subsequently stimulated and blood samples were collected from both adrenal and peripheral veins. Measurement of CA levels in these samples was used to differentiate adrenal medullary from sympathetic neural components of the SA response. Heart rate and blood pressure were continuously monitored to permit comparison between the activity of the SA and cardiovascular (CV) systems. Preliminary results of this study have previously been reported in abstract form [39]. METHOD
Elicitation of Aggressive Behavior Twenty adult cats (16 females and 4 males; mean BW =2.5 kg) were used in this study. During aseptic surgery, electrode guides were stereotaxically mounted on the skull overlying the medial hypothalamus (anterior: 9.0--12.5; lateral: 1.0-2.5). Additionally, a series of bolts was anchored to the skull for the purpose of securing the cat in a headholder during electrode placement and brain stimulation [28,29]. At least one week after stereotaxic surgery, the awake, unanesthetized animal was restrained in the headholder; calibrated, monopolar, stainless steel electrodes, 0.6 mm in diameter and bared at the tip for 0.5 ram, were lowered through the guides in 0.25 mm steps until stimulation of the hypothalamus elicited a hiss [40]. Each hypothalamic site was stimulated (Grass $88) at 60 Hz with biphasic, square wave pulses of 1 msec half-cycle duration. To approximate constant current conditions the constant voltage output of the stimulator was led through two stimulus isolation units (Grass SIU5) and a pair of 40,000 ohm resistors placed in series with the leads to the cat, and was monitored with an oscilloscope. The following day, the behavior elicited by stimulation of each electrode was evaluated by placing the experimental cat and a " t a r g e t " cat in an observation box. Electrodes were selected that elicited affective defense behavior: hissing, pupillary dilatation, retraction of the ears, and piloerection. The stimulated cat fixed its gaze on the target cat and either backed into a corner or struck with its paw at the target cat. The intensity of stimulation which reliably elicited affective defense within 10 sec was defined as the "behavioral intensity," and ranged from 0.2-0.8 mA (peak-to-peak).
NE
E NE
E NE
A 150
~
AI2.0
E
A 12.5
J
A,.5 aHd
i FIG. 1. Composite map of stimulation sites in the hypothalamus which elicited defense in the freely moving cat. Sites are plotted on drawings of coronal sections from the atlas of Jasper and AjmoneMarsan [17]. Symbols indicate the type of adrenal medullary response elicited by stimulation of each site at the behavioral intensity. Changes in NE output are on the left of each drawing; changes in E are on the right. Shaded circle: increase in CA output />10 ng/min; clear circle: increase in CA output/>0 but < 10 ng/min; circle with dots: decrease in CA output. Each circle is divided into halves: the left half shows the response from the adrenal gland ipsilateral to the side of hypothalamic stimulation; the right half shows the response from the contralateral adrenal gland. Half circles show the adrenal medullary responses in the 5 cats that had only the ipsilateral adrenolumbar vein cannulated (see the Method section). Abbreviations used: AC, anterior commissure; arid, dorsal hypothalamic area; F, fornix; Ha, anterior hypothalamic area; Hdm, dorsomedial nucleus; Hp, posterior hypothalamus; Mm, medial mammiUary nucleus; mt, mammillothalamic tract; OC, optic chiasm; OT, optic tract; PVH, paraventricular nucleus; SCh, suprachiasmatic nucleus: SON, supraoptic nucleus; vm, ventromedial nucleus.
and maintained between 101°-102°F with heating pads. Following the surgical procedures the animal was heparinized and allowed to stabilize for at least 30 min before blood sampling commenced.
Cannulation Procedure The following day the animal was anesthetized with sodium pentobarbital (47 mg/kg). Cannulae were placed in the femoral artery for the continuous monitoring of heart rate (HR) and mean arterial pressure (MAP), and in the atrial appendage through the left external jugular vein for the withdrawal of peripheral blood samples. Additionally both adrenolumbar veins [26] were exposed through a mid-ventral abdominal incision and directly cannulated with PE 90. (In the cat the adrenal vein exits from the hilus of the adrenal gland and joins with the adrenolumbar vein, which passes over the adrenal gland, draining the posterior body wall and adrenal vein. The adrenolumbar vein joins the inferior vena cava on the right and either the renal vein or inferior vena cava on the left. There is no " a d r e n o l u m b a r " vein in the human; rather this vessel is termed the adrenal vein.) In five of the 20 cats only the adrenolumbar vein ipsilateral to the side of hypothalamic stimulation was cannulated; the contralateral adrenolumbar vein was ligated at the junction with the inferior vena cava. The body temperature was monitored
Collection of Venous Blood Samples In the anesthetized animal, each behaviorally identified hypothalamic site (N=31) was stimulated for 30 sec at the intensity which evoked behavior when awake. F o r each stimulation, continuous sequential blood samples were taken concurrently from the adrenolumbar vein(s) and from the atrium during four l-rain periods: two prior to stimulation and two following stimulus onset. Blood was allowed to drip from the adrenolumbar cannulae into heparinized tubes kept on ice; the volume of each l-rain sample was measured. Seventy-one percent of the sites (N=22) were stimulated twice: once at the behavioral intensity, and once a t a standard, higher intensity (0,8 mA). Following each stimulation period, the volume of blood collected was replaced with an equal volume of Macrodex (dextran 70; Pharmacia). At the completion of the experiment, a small lesion was made via each electrode (10/zA for 3 rain). The animal was then sacrificed and perfused; the location of each electrode tip was determined through standard histological procedures [41].
PLASMA CATECHOLAMINES AND DEFENSE
Catecholamine Assay Plasma was immediately separated after the blood samples were drawn, and the proteins precipitated by the addition of 20 tzl 5 N perchloric acid to each 480/zl plasma. The supernatants of a 20,000 g centrifugation for 15 rain were stored at -70°C until assayed for N E and E within 1-2 months. NE and E were assayed in 50 txl aliquots of the perchloric acid supernatant by a sensitive radioenzymatic technique utilizing catechoi-O-methyl transferase and S-adenosyl methionine [methyl-3H] as a methyl donor [24].
Data Analysis The adrenal medullary activation which accompanied hypothalamically-elicited defense behavior was determined by comparing the absolute changes in the output of NE and E (ng/min) during the first minute following stimulus onset to the baseline mean. CA output was quantified by multiplying the level in ng/ml by the blood flow rate (ml/min); this calculation was chosen to incorporate observed variations in adrenolumbar venous flow rate. Since the baseline levels of CAs were variable both across and within cats, we wished to select a parameter for the evaluation of adrenal medullary activity that was independent of baseline values. Thus, absolute changes in CA levels were used rather than percent changes since, with the exception of E levels in the ipsilateral adrenal gland, absolute changes in CAs were not significantly correlated with baseline values. In all cases, however, percent changes in CA levels were significantly correlated with baseline values. Cardiovascular activation was determined by comparing the maximum changes in MAP and HR during stimulation to the baseline mean. Correlations between sympatho-adrenal and cardiovascular variables were evaluated using Spearman's rank-order correlation coefficient; these correlation coefficients were judged to be significant if the associated p value was 0.05 or less.
Peripheral Venous NE To determine sympathetic nervous system activity, it was necessary to estimate the proportion of N E present in peripheral (atrial) blood contributed by the sympathetic noradrenergic nerves. E was also present in the peripheral plasma. As the primary source of E is the adrenal medulla, we assumed that a proportion o f peripheral NE, also, was contributed by the adrenal medulla. Since both adrenal veins were cannulated or tied off, the amount of E in the peripheral plasma relative to that in the bilateral adrenal venous plasma was considered to represent that proportion of adrenal CA output which had " l e a k e d " from the adrenal medulla into the peripheral circulation, perhaps via extra-hilar accessory veins, as have been described for humans [20], or via the lymphatic system [42]. Extra-adrenal chromaffin cells (paraganglia) [32], or accessory adrenal glandular tissue (unpublished observations) may also have contributed E to the peripheral plasma. A proportion of the adrenal N E output equal to the proportion of " l e a k e d " E was subtracted from the total peripheral NE to yield an estimate of peripheral NE which was contributed by the sympathetic noradrenergic nerves: Peripheral N E from nerve terminals = Total peripheral N E - ((Peripheral E/Adrenal venous E) × Adrenal venous NE). RESULTS Stimulation of 31 hypothalamic sites elicited affective de-
869 TABLE 1 NUMBER OF HYPOTHALAMIC DEFENSE SITES THAT EVOKED VARIED PATTERNS OF INCREASES IN BILATERAL ADRENAL CATECHOLAMINE OUTPUT AT BEHAVIORAL (0.2-0.8 mA) AND MAXIMAL(0.8 mA) INTENSITIES OF STIMULATION Intensity of Stimulation Increases in CA output* Bilateral '~NE and E Bilateral '~E and unilateral 1'NE Bilateral '~NE and unilateral I'E Bilateral I'E only Bilateral I'NE only Unilateral TNE and E Unilateral I'E only Unilateral I'NE only
Behavioral Maximal (N=24) (N=20) 9 5 3 1 2 1 1 1
11 5 2 1 0 0 1 0
*Increases were defined as changes in CA output/>10 ng/min.
fense behavior in the awake cat. The location of these sites is shown in Fig. 1. Effective sites were located within the stereotaxic coordinates A 11.0-13.5, lat 0.25-2.0, and included the anterior hypothalamic area, the pefifornical area, the dorsomedial nucleus, and the ventromedial nucleus. F o r analysis the data were separated into those responses elicited by the behavioral intensity of stimulation, which was variable, and the uniform maximum stimulation intensity.
Baseline Adrenal CA Output The mean resting secretion of CA from a single adrenal gland (N = 100) +_ the standard error of the mean (SEM) was 104.0_+12.3 ng/min for N E and 39£-+13.2 ng/min for E. Thus, under baseline conditions, N E formed an average (_+SEM) of 72.2+2.2% of the total CAs released from the adrenal gland (range 12.2-97.7%).
Activation of the Adrenal Medulla Seven hypothalamic sites were stimulated in five cats in which only the adrenolumbar vein ipsilateral to the side of stimulation was cannulated. The results obtained from stimulation of these sites (shown in Fig. 1) supported results obtained from bilateral adrenolumbar cannulations. Therefore, since the results from bilateral cannulations were more complete, only these data will be discussed below. Hypothalamic stimulation at sites that elicited defense behavior significantly changed the relative proportions of NE and E which comprised the CA output of the adrenal medulla. At the behavioral intensity of stimulation N E comprised 58.9+2.8% of the CAs in the plasma sample following stimulus onset; at the maximum intensity of stimulation, the percentage of N E decreased further to 47.3_+ 3.(1%. It should be noted that even though hypothalamic stimulation significantly altered the proportion of CAs released by the adrenal medulla, the absolute levels of NE were still greater than the levels of E following approximately half of all stimulations. Various patterns of increases in adrenal CA output were seen following stimulation at hypothalamic defense sites (Table 1 and Fig. I). As shown in Table 1, the most frequent adrenal response was a bilateral increase (/>10 ng/min) in
" ; , I ( ) I ) I ) A R I ) t,~ 4i
~70
Behavioral Intensity
Maximal Behovioral
Intensit),Intensif)~ A 15.0
Maximal Behavioral Intensity Itensit~/
A
Maximal Intensity
~ I5
<:SZ) vzsT>
J
AI 1 . 0 ~
cantly greater from the ipsilateral gland than flom the con tralateral gland (mean-+SEM=61.6+13.0 ng/min w, 42.3-+9.6 ng/min, respectively). Similarly, the increases in NE from the ipsilateral gland were greater than the contralateral increases, but the differences were not significant (mean_+SEM=61.4_ 15.4 ng/min vs. 36.9-+13.3 ng/min, respectively). This significant difference between ipsilateral and contralateral adrenomedullary CA output was also apparent at the maximum stimulation intensity only for E. For all stimulations, the changes in bilateral adrenal NE output resulting from hypothalamic stimulation were significantly, but weakly, correlated (r-0.5} with the changes in bilateral adrenal E output.
Activation o f the Peripheral Noradrem'r.eic Ncrvu~
FIG. 2. Composite map showing mcreases in the level of adjusted peripheral NE, suggesting activation of the sympathetic nervous component of the SA response, following stimulation of hypothalamic defense sites. Increases are expressed as percent of baseline mean: clear circle: 0-1 lff%; one quarter shaded circle: 111-120%; top half shaded circle: 121-130%; three quarter shaded circle: 131145%; shaded circle: > 145%. Responses at the behavioral intensity of stimulation (0.2-0.8 mA) are shown on the left of each drawing; responses at the maximal intensity of stimulation (0.8 mA) are shown on the right. Only stimulations are shown that permitted measurement of changes in NE of peripheral origin (see the Method section). Abbreviations are the same as for Fig. 1.
both NE and E. Within this response type, the mean increases ( - S E M ) in adrenal NE and E output were 94.1 _+15.5 ng/min and 101.9_+ 14.2 ng/min respectively. Adrenal responses were also evaluated by comparing the increases in N E output to the increases in E output. Nine sites were identified that elicited greater absolute increases in NE, while eight other sites elicited greater absolute increases in E. There was a general pattern to the location of these two types of sites. Within the dorsomedial nucleus of the hypothalamus twice as many sites (4 vs. 2) evoked increases in N E greater than the increases in E. Additionally, within the ventromedial nucleus of the hypothalamus, sites which elicited greater increases in NE were generally located ventral to those sites which evoked greater increases in E. The exceptions to this were two sites located between the dorsolateral border of the ventromedial nucleus and the medial border of the column of the fornix at A 11.5; these sites elicited greater increases in NE. Although stimulation of hypothalamic defense sites usually increased the output o f adrenal CAs, stimulation of some sites evoked a decrease in the output of one or both adrenal CAs (Fig. 1). At the behavioral intensity of stimulation, a decrease in N E occurred bilaterally at 2 sites (Fig. 1: A 11.5 and 11.0), ipsilaterally at 3 sites (Fig. 1: A 12.5, 11.5, 11.0) and contralaterally at 1 site (Fig. 1: A 13.0). A decrease in E output occurred only in the ipsilateral gland following stimulation of 3 hypothalamic sites (Fig. 1: A 12.5 and 11.5). Stimulation of hypothalamic defense sites resulted in differences between the CA output of the glands ipsilateral and contralateral to the side of stimulation. At behavioral intensities of stimulation, the increases in E were signifi-
Calculations for adjusted peripheral NE permitted comparisons between baseline and post-stimulation NE levels in approximately 60% of the cases (Fig. 2), i.e., trials in which the " l e a k a g e " of CAs from the adrenal glands was not so great as to obscure any changes in NE from noradrenergic nerves. Under baseline conditions the mean amount (-+SEM) of peripheral NE contributed by sympathetic nerves was 66,9_+4.4%. Stimulation of hypothalamic defense sites at all intensities increased the mean amount of peripheral N E contributed by the sympathetic nerves to 75.1 _+2.8%. Activation of the sympathetic component of the SA response was arbitrarily defined as levels of adjusted peripheral NE following hypothalamic stimulation greater than 145% of the baseline mean [35]. Using this criterion, activation of the sympathetic nervous (SN) system accompanied behavioral stimulation of 30% of the defense sites and maximal stimulation of 63% of the sites (Table 2). Generally, increasing the intensity of stimulation from behavioral to maximal also increased the degree of activation of the SN system (compare right and left sides of Fig. 2). SN system activation was most often a concomitant of hypothalamically-elicited defense when the behavior was elicited by stimulation of the ventromedial nucleus (Fig. 2: A 12.0, 11.5 and 11.0) and its rostral projections (Fig. 2: A 13.5, 13.0, and 12.5).
('omparisott of SA and ('V Parameters The mean (-+SEM) of the maximal increases in HR and MAP during the 30 sec period of hypothalamic stimulation at behavioral intensity were 11.4_+2.4 beats/rain and 9.3_+2.3 mm Hg, respectively. These mean changes in CV parameters were relatively small despite the considerable output of CAs. The maximal changes in HR and MAP during hypothalamic stimulation were tested for correlation with the changes in adrenal CA output and adjusted peripheral NE. Stimulation at behavioral intensities elicited increases in ipsilateral adrenomedullary E output which were significantly, but weakly, correlated (r=0.44) with changes in HR. DISC USSION
Electrical stimulation of hypothalamic sites which elicited affective defense behavior in the freely moving cat also evoked a variety of sympatho-adrenal responses. Such variation has been reported previously following central nervous system stimulation both in acute, anesthetized preparations [10, 16, 34, 35], and in the awake animal [41]. However, the most frequent concomitants of SA activation which accompanied hypothalamically-elicited defense behavior were
PLASMA CATECHOLAMINES AND DEFENSE
871
TABLE 2 NUMBER OF HYPOTHALAMICDEFENSE SITES WHICH INCREASED THE LEVEL OF ADJUSTED PERIPHERAL NE (SUGGESTING SYMPATHETICNERVOUS SYSTEM ACTIVITY) AT BEHAVIORAL(0.2-0.8 mA) AND MAXIMAL(0.8 mA) INTENSITIES OF STIMULATION Intensity of Stimulation Behavioral (N=20)
Maximal (N= 19)
100-110
1
0
111-120 121-130 131-145 >145
2 3 2 6
0 1 2 12
% Baseline*
*Level of adjusted peripheral NE in the first atrial blood sample following the onset of hypothalamic stimulation expressed as percent of the baseline mean.
(1) a bilateral increase in the output of adrenal medullary NE and E; (2) a significant change in the ratio of CAs released from the adrenal medulla; (3) preferential activation of the ipsilateral adrenal medulla; and (4) activation of the sympathetic noradrenergic nerves.
Adrenal Medullary Component of the SA Response Baseline CA output. Adrenal CA output data from the present study were converted to ng/kg/min to permit comparison with previous studies that have reported adrenal CA output in cats [6, 7, 9, 10, 16, 18, 35]. These studies, which employed a variety of anesthetics and CA assay techniques, reported CA output from only the left adrenal gland. The variations in methodology, anesthetics and assay techniques yielded a wide range of average values for baseline CA output (E: 3.1-63.0 ng/kg/min; NE: 1.4--200.0 ng/kg/min). The mean resting levels (-+SEM) of bilateral adrenal CA output of 41.3_+4.8 ng/kg/min for NE and 15.9_+5.5 ng/kg/min for E obtained in the current study were within the range of values reported by other authors. Further, our results were in accord with the majority of other investigators that NE is the primary CA released from the feline adrenal medulla at rest [4, 10, 16, 18, 31, 35]. Activation associated with hypothalamically-elicited defense. Stimulation of hypothalamic sites which elicited defense behavior also generally elicited increases in both NE and E. These results are in partial agreement with previous reports. Katsuki et al. [19] and Briicke et al. [4] found medial hypothalamic stimulation in cats to result in marked increases only in E secretion. More recently Robinson et al. [35] found that electrical stimulation in the dorsal hypothalamic area elicited a significant increase in both adrenal CAs. Francke et al. [11] reported that hypothalamic and mesencephalic stimulation in cats elicited increases in adrenal CAs that were usually greater in the gland ipsilateral to the side of stimulation. Our results concurred with this in that stimulation of hypothalamic defense sites elicited a mean increase in adrenal E which was significantly greater from the ipsilateral gland. The mean increase in ipsilateral NE was also greater, although the difference was not significant.
In the present study the relative increase (percent above baseline) in output of adrenal E was nearly always greater than the relative increase in NE. In spite of the differences in proportional change of the adrenal CAs, the absolute increase of N E was greater than that of adrenal E in approximately half the cases. This response, in addition to the fact that NE was the main CA released from the resting adrenal in our study, resulted in post-stimulation values of N E that were frequently higher than E. Such observations suggest that, in the cat, NE is the predominant neurotransmitter and neurohumor. The question of selective secretion of NE or E from the adrenal medulla remains. The difficulty in evaluating others' data arises from the interpretation of the term " s e l e c t i v e " relative to the release of either of the adrenal CAs. Previous reports agree that both peripheral and central stimulation, in addition to various physiological and pharmacological stimuli, can alter the ratio of CAs released from the cat's adrenal medulla [6, 9, 16, 18, 31, 34, 35, 37, 43]. Results from the current study concur with this; hypothalamic stimulation which elicited defense behavior significantly altered the ratio of CAs. The difficulty in evaluating a " s e l e c t i v e " increase results from the criteria for measuring changes in CA output. In the current study, the relative increase in E was nearly always greater than that of N E , in large part due to the low resting levels of E. However, evaluation of the data in terms of absolute increases in CA output revealed that a greater output of NE was elicited as frequently as a greater output of E. Thus, based on absolute increases in adrenal CAs, the present investigation revealed instances of the "selective" secretion of each E and NE. Our further observation that the increases in output of adrenal N E and E were weakly, but significantly, correlated also suggests that there may be partially independent mechanisms modulating the release of these catecholamines. Using a different criterion, the relative increase in CA output, Robinson et al. [35] also'found evidence for the "selective" increase in each CA following hypothalamic stimulation.
Sympathetic Nervous Component of the SA Response Baseline levels of peripheral NE. Under baseline conditions, approximately two-thirds of the NE in the peripheral plasma was contributed by the sympathetic noradrenergic nerves; this was similar to the results in rats that 70% of the peripheral NE is overflow from the sympathetic nerve terminals [22,33]. Activation associated with hypothalamically-elicited defense. Hypothalamic sites which elicited defense behavior also effected an overall increase in the activity of the sympathetic nervous component of the SA re.sponse, as determined by the adjusted levels of peripheral NE. Stimulation at behavioral intensity increased the activity of the SN system to 110-145% baseline levels at 37% of the sites, and to more than 145% baseline at 30% of the sites. Stimulation at maximum intensity caused not only a greater absolute increase in peripheral NE, but also elevated the activity of the SN system to more than 145% baseline at 60% of the sites. It was of interest that the sites where stimulation caused the greatest activation of the SN system were related to the ventromedial nucleus. Stimulation of this hypothalamic region reliably elicits affective defense behavior [12]. Further, the pathway from the ventromedial nucleus which mediates this behavior involves an initial rostral projection that synapses in the anteromedial hypothalamus [13]. Compari-
872
ST(II)I)ARI) ¢ i A!
son between the anatomical maps presented by Fuchs e t a / . [12,13] and our Fig. 2 suggests that the central system which activates the sympathetic nervous component of the SA response is coincident, at least within the hypothalamus, with efferent pathways from the ventromedial nucleus. C a r d i o v a s c u l a r P a r a m e t e r s A s s o c i a t e d With I t y p o t h a l a m i c a l l y - E l i c i t e d D<['ense
A significant correlation between CV and SA parameters was found only between HR and ipsilateral adrenomedullary E output. In fact, only relatively minor changes were noted in CV parameters compared to the degree of activation of the adrenal medullary and sympathetic neural components of the SA response. These observations suggest that hypothalamic sites which elicit defense behavior do not concomitantly activate the CV and SA systems. The current data support previous suggestions [8,41], that there appear to be regions in the hypothalamus which differentially activate the sympatho-adrenal and the cardiovascular systems. CONCLUSIONS In the present study, hypothalamically-elicited defense behavior was most frequently accompanied by activation of both the adrenal medullary and sympathetic nervous components of the SA response. Activation of the sympathetic nervous component of the SA response was reflected in in-
creased levels o1' NE in the peripheral plasma; Ihc concomkant cardiovascular measurements did not reliably reflect this sympathetic nervous activity. In spite of the anatomical spread of hypothalamic sites which elicited defense behavior, it appears that the central systems which mediate this behavior are largely coincident with pathways that activate the adrenal medulla, and frequently overlap with areas that activate the peripheral noradrenergic nerves. However, because defense behavior can be elicited from various hypothalamic regions, the concomitant SA responses ma~, vat'?,. Thus, use of this centrally-elicited behavior as a model lbr aggression or emotional behavior may no! be appropriate in all cases.
ACKNOWLEDGEMENTS The authors wish to thank Ms. M. Culpepper for her technical assistance, Mr. S. Stoddard-Apter for his assistance in programming and data analysis, and Ms. E. Wilson for her expert typing. The comments of Drs. J. A. Vilensky and S. W. Carmichael in critically evaluating earlier drafts of the manuscript are appreciate& This project was supported in part by Grant Number 5-S07-RR-5371, awarded by the Biomedical Research Support Grant Program, Division of Research Resources. National Institutes of Health (SLS), a grant from the Harry Frank Guggenheim Foundation ~SLS) and by the Research Service of the Veterans Administration (BELL
REFERENCES
1. Adamec, R. E. and C. 1. Stark-Adamec. Limbic control of aggression in the cat. Prog Neuropsychopharmacol Biol Psychiatry 7: 505-512, 1983. 2. Axelrod, J., H. Weil-Malherbe and R. Tomchick. The physiological disposition of Ha-epinephrine and its metabolite metanephrine. J Pharmacol Exp Ther 127: 251-256, 1959. 3. Boddin, M. and W. Dierick. Catecholamines in blood during direct stimulation of the hypothalamus in dogs. Arch Int Physiol Biochim 82: 895--906, 1974. 4. Briicke, F., F. Kaindl and H. Mayer. Ober die Ver~inderung in der Zusammensetzung des Nebennierenmarkinkretes bei elektrischer Reizung des Hypothalamus. Arch Int Pharmacodyn Ther g$: 407-412, 1952. 5. Cannon, W. B. The emergency function of the adrenal medulla in pain and the major emotions. Am J Physiol 23: 356--372, 1914. 6. Critchley, J. A. J. H., P. Ellis and A. Ungar. The reflex release of adrenaline and noradrenaline from the adrenal glands of cats and dogs. J Physiol (Lond) 298: 71-78, 1980. 7. Eade, N. R. and D. R. Wood. The release of adrenaline and noradrenaline from the adrenal medulla of the cat during splanchnic stimulation. Br J Pharmacol 13: 390--394, 1958. 8. Ferguson, R. W., B. Folkow, M. G. Mitts and E. C. Hoff. Effect of cortical stimulation upon epinephrine activity. J Neurophysiol 20: 32%339, 1957. 9. Feuerstein, G. and Y. Gutman. Preferential secretion of adrenaline or noradrenaline by the cat adrenal in vivo in response to different stimuli. Br J Pharmacol 43: 764-775, 1971. 10. Folkow, B. and U. S. von Euler. Selective activation of noradrenaline and adrenaline producing cells in the cat's adrenal gland by hypothalamic stimulation. Circ Res 2: 191-195, 1954. 11. Francke, P. F., J. L. Culberson, S. W. Carmichael and R. L. Robinson. Bilateral secretory responses of the adrenal medulla during stimulation of hypothalamic or mesencephalic sites. J Neurosci Res 8: 1--6, 1982. 12. Fuchs, S. A. G., H. M. Edinger and A. Siegel. The organization of the hypothalamic pathways mediating affective defense behavior in the cat. Brain Res 330: 77-92, 1985.
13. Fuchs, S. A. G., H. M. Edinger and A. Siegel. The role of the anterior hypothalamus in affective defense behavior elicited from the ventromedial hypothalamus of the cat. Brain Re.s 330: 93-107, 1985. 14. Fuchs, S. A. G. and A. Siegel. Neural pathways mediating hypothalamically elicited flight behavior in the cat. Brain Res 306: 263-281, 1984. 15. Goldfien, A. and W. F. Ganong. Adrenal medullary and adrenal cortical response to stimulation of diencephalon. Am J Physiol 202: 205-211, 1962. 16. Grant, R., P. Lindgren, A. Rosen and B. Uvnas. The release of catechols from the adrenal medulla on activation of the sympathetic vasodilator nerves to the skeletal muscles in the cat by hypothalamic stimulation. Acta Physiol S t a n d 43: 135-154, 1958. 17. Jasper, H. H. and C. A. Ajmone-Marsan. Stereotaxic Atlas" q f the Diencephahm o f the Cat. Ottawa: National Research Council of Canada, 1954. 18. Kaindl, F. and U. S. von Euler. Liberation of nor-adrenaline and adrenaline from the suprarenals of the cat during carotid occlusion. Am J Physiol 166: 284-288, 1951. 19. Katsuki, S., T. Ikemoto, H. Shimada, F. Hagiwara and J. Kanal. The functional relationship between the hypothalamus and the anterior pituitary-adrenal system. Endocrinot Jpn 2: 303-312, 1955. 20. Khalique, A. and W. S. Monkhouse. The adrenal veins of man. J Anat 135: 846, 1982. 21. Kollai, M. and K. Koizumi. The mechanisms of differential control in the sympathetic system studied by hypothalamic stimulation. J Auton Nerv Syst 2: 377-389, 1980. 22. Kvetnansky, R., V. K. Weise, N. B. Thoa and I . J. Kopin. Effects of chronic guanethidine treatment and adrenal medullectomy on plasma levels of catecholamines and corticosterone in forcibly immobilized rats. J Pharmacol Exp Ther 2tl9: 287-291. 1979.
PLASMA CATECHOLAMINES
AND DEFENSE
23. Levin, B. E. and O. R. Hubschmann. Dorsal column stimulation effect on human cerebrospinal fluid and plasma catecholamines. Neurology 30: 65-71, 1980. 24. Levin, B. E., J. Triscari and A. C. Sullivan. Abnormal sympatho-adrenal function and plasma catecholamines in obese Zucker rats. Pharmacol Biochem Behav 13: 107-113, 1980. 25. Leyhansen, P. Cat Behavior. New York: Garland STPM Press, 1979. 26. McDonald, L. E. Veterinary Endocrinology and Reproduction, 3rd edition. Philadelphia: Lea & Febiger, 1980. 27. MacDonnell, M. F., S. H. Brown and B. Davy. Hyperexcitability in the neural substrate of emotional behavior in cats after alcohol withdrawal. J Stud Alcohol 36: 1480-1492, 1975. 28. MacDonnell, M. F., L. Fessock and S. H. Brown. Ethanol and the neural substrate for affective defense in the cat. Q J Stud Alcohol 32: 406--419, 1971. 29. MacDonnell, M. F. and S. Stoddard-Apter. Effects of medial septal stimulation on hypothalamically-elicited intraspecific attack and associated hissing in cats. Physiol Behav 21: 679-683, 1978. 30. Magoun, H. W., S. W. Ranson and A. Hetherington. The liberation of adrenin and sympathin induced by stimulation of the hypothalamus. Am J Physiol 119: 615-622, 1937. 31. Malmejac, J. Activity of the adrenal medulla and its regulation. Physiol Rev 44: 186--218, 1964. 32. Mascorro, J. A. Anatomical and morphological observations on the juxtaposition between the inferior mesenteric ganglion and paraganglion chromaflin cells in cats. Anat Rec 202: 121A, 1982. 33. Micalizzi, E. R. and D. T. Pals. Evaluation of plasma norepinephrine as an index of sympathetic neuron function in the conscious, unrestrained rat. Life Sci 24: 2071-2076, 1979. 34. Redgate, E. S. and E. Gelihorn. Nature of sympathetico-adrenal discharge under conditions of excitation of central autonomic structures. Am J Physiol 174: 475--480, 1953. 35. Robinson, R. L., J. L. Culberson and S. W. Carrnichael. Influence of hypothalamic stimulation on the secretion of adrenal medullary catecholamines. J Auton Nerv Syst 8: 89-96, 1983.
873 36. Romaniuk, A. Representation of aggression and flight reactions in the hypothalamus of the cat. Acta Biol Exp (Warsaw) 25: 177-186, 1965. 37. Rubin, R. P. and E. Miele. A study of the differential secretion of epinephrine and norepinephrine from the laerfused cat adrenal gland. J Pharmacol Exp Ther 164:115-121, 1968. 38. Siegel, A. and H. Edinger. Neural control of aggression and rage behavior. In: Handbook of the Hypothalamus, Vol 3, Part B, Behavioral Studies of the Hypothalamus, edited by P. J. Morgane and J. Panksepp. New York: Marcel Dekker, 1981, pp. 203-240. 39. Stoddard-Apter, S. L., V. Bergdall and B. E. Levin. Catecholamine levels in adrenal venous and peripheral plasma following stimulation of behaviorally identified hypothalamic sites in the cat. Soc Neurosci Abstr 9:111, 1983. 40. Stoddard-Apter, S. L. and M. F. MacDonnell. Septal and amygdalar efferents to the hypothalamus which facilitate hypothalamically-elicited intraspeciflc aggression and associated hissing in the cat. An autoradiographic study. Brain Res 193: 19-32, 1980. 41. Stoddard-Apter, S. L., A. Siegel and B. E. Levin. Plasma catecholamine and cardiovascular responses following hypothalamic stimulation in the awake cat. J Auton Nerv Syst 8: 343360, 1983. 42. Trautmann, A. and J. Fiebiger. Fundamentals of the Histology of Domestic Animals. Ithaca, NY: Comstock Publishing Associates, 1952. 43. Ungar, A. and J. H. Phillips. Regulation of the adrenal medulla. Physiol Rev 63: 787-843, 1983. 44. Wasman, M. and J. P. Flynn. Directed attack elicited from hypothalamus. Arch Neuro/6: 220-227, 1962. 45. Whitby, L. G., J. Axelrod and H. WeiI-Malherbe. The fate of H3-norepinephrine in animals. J Pharmacol Exp Ther 132: 193201, 1961.