Excitatory amino acids in the rostral ventrolateral medulla support blood pressure during. water deprivation in rats. By:

Transcription

1 Excitatory amino acids in the rostral ventrolateral medulla support blood pressure during water deprivation in rats. By: Virginia L. Brooks, Korrina L. Freeman and Kathy A. Clow Department of Physiology and Pharmacology Oregon Health & Science University Portland, OR Running title: EAA to RVLM in water deprived rats Correspondence: Virginia L. Brooks, Ph.D. Department of Physiology and Pharmacology, L-334 Oregon Health & Science University 3181 SW Sam Jackson Park Rd Portland, OR (503) FAX: (503)

2 ABSTRACT Water deprivation is associated with regional increases in sympathetic tone, but whether this is mediated by changes in brainstem regulation of sympathetic activity is unknown. Therefore, this study tested the hypothesis that water deprivation increases excitatory amino acid (EAA) drive of the rostral ventrolateral medulla, by determining if bilateral microinjection of kynurenate (KYN; 2.7 nmol) into the RVLM decreases arterial pressure more in water-deprived compared to water-replete rats. Plasma osmolality was increased in 48 hr water deprived rats (313"1 mosm/kg H 2 O; P<0.05) compared to 24 hr water deprived rats (306"2 mosm/kg H 2 O) and water replete animals (300"2 mosm/kg H 2 O). KYN decreased arterial pressure by 28.1"5.2 mmhg (P<0.01) in 48 hr water-deprived rats, but had no effect in water-replete rats (-5.9"1.3 mmhg). Variable depressor effects were observed in 24 hr water-deprived animals (-12.5"2.4 mmhg, ns); however, in all rats the KYN depressor response was strongly correlated to the osmolality level (P<0.01; r 2 =0.47). The pressor responses to unilateral microinjection of increasing doses (0.1, 0.5, 1.0, 5.0 nmol) of glutamate were enhanced (P<0.05) during water deprivation, but the pressor responses to iv phenylephrine injection were smaller (P<0.05). These data suggest that water deprivation increases EAA drive to the RVLM, in part by increasing responsiveness of the RVLM to EAA such as glutamate. Keywords: kynurenic acid, glutamate, brain, arterial blood pressure, osmolality

3 INTRODUCTION The rostral ventrolateral medulla (RVLM) plays a crucial role in the maintenance of basal sympathetic activity (for reviews, see (6, 12, 29). Indeed, acute bilateral blockade of the RVLM causes marked decreases in arterial pressure and the recorded activity of many sympathetic nerves to cease. Moreover, there is evidence that excitatory amino acid (EAA) input into the RVLM underlies at least part of this tonic activity (6, 29). For example, in the cat or rabbit, bilateral blockade of EAA receptors in the RVLM by microinjection of the EAA antagonist kynurenic acid (KYN) produces a significant decrease in arterial pressure (5). In rats, bilateral KYN injections into RVLM generally do not decrease arterial pressure, unless inhibitory input from the caudal ventrolateral medulla or nucleus tractus solatarius is first eliminated (6, 18, 27). It has been proposed that in normal rats the tonic excitatory EAA input into RVLM is balanced by EAA drive of inhibitory interneurons (21, 29). Thus, KYN blocks both the excitatory and inhibitory actions of EAA, and blood pressure does not change significantly. Sympathetic tone does not remain constant, but changes in a number of physiological and pathophysiological states. For example, in multiple models of hypertension, basal sympathetic activity appears to increase (4, 7, 28). Experiments have been performed to determine if the increased sympathetic activity is mediated by a relative increase in the excitatory EAA drive of the RVLM and have shown that bilateral KYN microinjections decrease arterial pressure in rats with renovascular hypertension (1) as well as genetic models of hypertension, such as SHR (16) 1

4 and Dahl salt-sensitive rats (17). Thus, these hypertensive states are maintained in part by increased sympathetic tone, which is driven by tonically active EAA inputs to the RVLM. While changes in brain regulation of sympathetic outflow have been well-documented in numerous hypertensive models, whether similar changes occur in nonhypertensive sympathoexcitatory states has not been previously investigated. Water deprivation is characterized by volume depletion, increases in the osmolality of the body fluids and a homeostatically appropriate increase in sympathetic activity, at least to some vascular beds. Moreover, there is indirect evidence that this increase in sympathetic tone may involve increased EAA drive of the RVLM. First, acute increases in systemic osmolality activate neurons that originate in the paraventricular nucleus and that project to the RVLM (31). Second, nonspecific activation of the paraventricular nucleus by injection of the EAA D,L-homocysteic acid increases the activity of RVLM presympathetic neurons, and this activation is blocked by prior injection of KYN into the RVLM (33). However, whether EAA input to RVLM is increased during water deprivation has not been previously investigated. Therefore, the present study was performed to test the hypothesis that water deprivation increases EAA drive of the RVLM. To test this hypothesis, it was determined if bilateral microinjection of KYN into RVLM lowers arterial pressure in urethane anesthetized water deprived, but not water replete, rats. METHODS Animals. Experiments were performed using male Sprague-Dawley rats (Sasco, Inc., Wilmington, MA) weighing approximately g. All rats were housed in the institutional 2

5 facility a minimum of 5 days before experimentation. Rats were grouped 1-3, in a room with a 12h:12h light/dark cycle. All rats had free access to food (Purina 5001). Water deprived rats were housed singly for either 24 or 48 hr without water. Surgery. Anesthesia was induced with 5% isoflurane in 100% oxygen. A trachea tube was inserted so that the animals could be artificially ventilated, and a surgical plane of anesthesia was maintained with 2% isoflurane in 100% oxygen. Body temperature was maintained throughout the surgery and experiment at 37"1EC using a rectal thermister and a heating pad. Femoral arterial and venous catheters were implanted for arterial pressure measurements and infusions, respectively. The rats were then positioned in the stereotaxic device (David Kopf, Tujunga, CA), and a midline incision was made on the back of the head to expose the dorsal surface of the medulla and remove the atlanto-occipital membrane by a limited craniotomy. After completion of all surgical manipulations, a blood sample (400 µl) was collected for measurement of plasma osmolality, plasma protein concentration, hematocrit and plasma concentrations of sodium and chloride. An iv infusion of urethane (1.2 g/kg in one ml saline) was then administered over ~30 min; ten min after beginning the urethane infusion, the gas anesthetic was slowly withdrawn, but the rats were continually artificially ventilated with 100% oxygen throughout the experiment. After completion of the urethane infusion, the rats were allowed to stabilize for about min before experimentation. The depth of anesthesia was periodically assessed by confirming the lack of response to tail or paw pinch. Additional urethane (0.2 g/kg) was occasionally administered iv when needed. RVLM microinjections. Functional identification of the RVLM was made by observing >20 mmhg pressor responses to L-glutamate [1 nmol/100 nl (18, 19, 30). Single-barreled glass 3

6 micropipettes (20-40 µm tip diameter) containing glutamate were positioned as described by Keily and Gordon (19). Briefly, the tip of the rat's nose was pushed down until the calamus scriptorius was 2.4 mm posterior to the interaural line. Using calamus scriptorius as zero, injections were made: mm anterior, mm (usually 1.9) lateral and mm ventral to the dorsal surface of the medulla. Injections (100 nl) were made over approximately 3-7 sec using either a 1µL Hamilton syringe held in a micromanipulator or a PicoPump (WPI); the successful microinjection of drugs was verified by watching, through a microscope reticule, the movement of a small bubble in the injection tubing a distance calibrated to be 50 or 100 nl. No more than 3 penetrations were made per side before the experiment commenced. All drugs [except KYN, which was first dissolved in one part 100 mm sodium bicarbonate and then diluted with 9 parts artificial cerebrospinal fluid (acsf)] were dissolved in acsf containing (in mm): 128 NaCl, 2.6 KCl, 1.3 CaCl 2, 0.9 MgCl 2, 20 NaCO 3 and 1.3 Na 2 HPO 4. The ph of all solutions was corrected to 7.4. Before the pipette was filled with a new drug, it was flushed with several volumes of acsf and then several volumes of the new drug. At the conclusion of the experiment, ~50 nl of 2.5% Alcian Blue in 0.5 M sodium acetate was injected into the RVLM using the same pipette and coordinates used for injections. The brain was removed and placed in 4% formaldehyde in phosphate-buffered saline for at least 48 hr. The brain stem was subsequently cut into 50 µm sections using a cryostat; sections were mounted on glass microscope slides and counter-stained with neutral red. RVLM injection sites were verified against those previously published, within an area approximately 500 µm caudal to the caudal end of the facial nucleus, and ventral to nucleus ambiguus (Figure 1). 4

7 Protocols. After the RVLM was identified functionally, ~30 min were allowed for stabilization and collection of baseline data, and then one of the following protocols was performed: 1). To determine if EAA drive of RVLM is increased during water deprivation, bilateral microinjections of KYN (2.7 nmol in 100 nl) or acsf were given into the RVLM of water replete and water deprived (24 hr and 48 hr) rats. After the first microinjection was completed, the pipette was removed and positioned on the other side, and the second injection was made. Both injections were usually completed within 1 min. Some rats were pretreated with an iv injection of V1 vasopressin antagonist [Manning Compound; d(ch 2 ) 1 5, Tyr(Me) 2, Arg 8 - vasopressin; 5 µg in 100 µl] at least 15 min before injection of KYN. The responses of animals pretreated with the V1 antagonist were not significantly different from those of untreated rats, so the data were combined. 2). To determine if the increased EAA drive exhibited by water deprived rats is due to increased sensitivity of the RVLM to glutamate, varying doses of glutamate (0.1, 0.5, 1, and 5 nmol) were unilaterally microinjected into the RVLM of water replete and 48 hr-water deprived rats. Different doses were administered in random order, using the same pipette, but the pipette was removed and re-filled with a different concentration of glutamate for each injection. At least 5 min were allowed between injections. While the anterior-posterior and medial-lateral positions were fixed, a total of 3-4 repetitions of the dose were given as the pipette was lowered in 0.2 mm increments in the region bracketing the functionally identified site. The largest response in each case was chosen for data analysis. 3). To determine if changes in the glutamate pressor response were due to changes in vascular sensitivity to adrenergic agonists, varying doses of phenylephrine were administered iv as a bolus in water replete and 48 hr water deprived rats. Rats were studied either intact or after iv 5

8 administration of the V1 vasopressin antagonist followed by the ganglionic blocker hexamethonium (30 mg/kg). Doses of phenylephrine (1,3,6,10 and 20 µg/kg) were given in random order; different doses were given after a minimum 5 min interval or after baseline values were achieved. 4). To determine if water deprivation is also associated with decreased responsiveness of RVLM to inhibitory neurotransmitters, water replete and water deprived (24 hr and 48 hr) rats received unilateral microinjections of gamma-aminobutyric acid (GABA; 1 and 10 nmol); injections were administered randomly, in duplicate (results were averaged),15-20 min apart. 5). To determine if reduced inhibitory input from baroreceptors increases the sensitivity of the RVLM to glutamate, rats underwent cervical vagotomy (n=4) or vagotomy/sinoaortic denervation (SAD; n=4) as previously described (14). One rat received SAD alone. Results were similar to those following vagotomy/sad; therefore, they were combined. After catheterization, but before the deafferentation, the rats were given an iv bolus of the V1 antagonist to minimize the hypertension caused by the procedure. Rats were then prepared for RVLM microinjections as described above, and responses to glutamate (1 nmol in 100 nl) were determined. Chemicals. KYN, glutamate, hexamethonium and phenylephrine were all obtained from Sigma, and the V1 antagonist was obtained from Bachem. Data analysis. Between group (water replete, 24 hr water deprived, 48 hr water deprived) differences were determined using ANOVA and the post-hoc Bonferroni Correction. Differences in changes in arterial pressure with time or in dose-response relationships between phenylephrine and arterial pressure were assessed with 2-way ANOVA for repeated measures. 6

9 Finally, linear least-squares regression analysis was used to determine if a relationship exists between plasma osmolality and the depressor response to KYN. RESULTS Basal values. Water deprivation increased plasma sodium and chloride concentrations (Table 1) and plasma osmolality (Figure 2). The rats were also volume depleted, as indicated by increases in hematocrit and plasma protein concentration (Table 1). Nevertheless, mean arterial pressure was not different between groups (Table 1). Responses to kynurenate. Bilateral microinjections of KYN did not alter arterial pressure in water replete rats, but decreased arterial pressure in 48 hr water deprived rats (Figures 2,3). The depressor response to kynurenate in 24 hr water deprived rats was variable (range, -2 to -20 mmhg), was not significantly different from water replete animals (Figure 2), and was smaller than the response in 48 hr deprived rats (P<0.05). However, when the data from all groups were analyzed together, the decrease in pressure was highly correlated with the plasma osmolality (Figure 2, inset; P=0.003; r 2 =0.47) Responses to glutamate. Dose-dependent pressor responses to unilateral glutamate microinjection were observed in both water replete and water deprived rats (P<0.001, ANOVA), but the responses of water deprived rats were larger (Figure 4). However, the pressor responses to bolus injections of phenylephrine were smaller in water-deprived rats (Figure 5, top). To assess the pressor effect of phenylephrine in the absense of baroreflexes and vasopressin, a second group of animals was studied after treatment with a V1 antagonist and ganglionic 7

10 blockade. The V1 antagonist decreased arterial pressure more (p<0.005) in water-deprived (- 20"2 mmhg; n=6) compared to water replete (- 6"3 mmhg; n=7) rats. However, subsequent hexamethonium injection produced similar depressor responses (-54"9 mmhg, replete; -52"6 mmhg, deprived; p>0.5). As in intact rats, the pressor responses to phenylephrine were smaller in V1- and hexamethonium-pretreated water-deprived rats (Figure 5, bottom). Responses to GABA. GABA injections dose-dependently decreased arterial pressure (P<0.001, ANOVA; Figure 6). The depressor responses were generally larger in water-deprived rats, although there was no difference between 24-hr and 48-hr deprived animals. Responses to glutamate after baroreceptor deafferentation. Unilateral glutamate microinjection (1 nmol) into RVLM increased arterial pressure more (p<0.005) in rats after vagotomy (25"2 mmhg, n=4) or vagotomy/sad (27"1 mmhg, n=5), compared to intact rats (22"1 mmhg, n=10). Nonetheless, it should be noted that these values may underestimate the differences between intact and deafferented animals, because the former values were obtained upon the first injection (e.g. response greater than 20 mmhg) without a systematic search for the most responsive site. DISCUSSION The major new findings of the present study are that 1) blockade of EAA receptors in the RVLM by bilateral microinjection of KYN decreases arterial blood pressure in water deprived, but not water replete, rats; and 2) the pressor responses evoked by unilateral injection of increasing doses of glutamate are greater in water deprived, compared to water replete, rats. 8

11 Collectively, these data suggest that water deprivation is associated with increased EAA drive of RVLM sympathetic premotor neurons, possibly due in part to increased sensitivity of the RVLM to EAA such as glutamate. Water deprivation causes a decrease in blood volume, as indicated indirectly in the present study by increases in hematocrit and plasma protein concentration. The volume contraction does not lead to hypotension, however, because of homeostatic activation of neurohumoral mechanisms. For example, both vasopressin and angiotensin II levels increase (e.g. 2), and data in the present study confirm that the increased vasopressin levels contribute to pressure maintenance. While indirect measures of overall sympathetic tone, such as plasma catecholamines (9, 20, 32) or the depressor response to ganglionic blockade (11) (present study), are not significantly increased, there is evidence that regional increases in sympathetic activity occur. First, lumbar sympathetic nerve activity (as a function of reflex-induced maximum) is increased (25), and heart rate is either increased or unchanged (2, 11, 25, 26). Furthermore, the adrenal mrna levels for tyrosine hydroxylase, the rate-limiting enzyme for norepinephrine production known to be increased by sustained increases in sympathetic activity, are elevated by water-deprivation, and this effect is blocked by adrenal denervation (3). On the other hand, renal sympathetic activity (as a percent of baroreflex maximum) is unchanged (26). Thus, basal sympathetic activity to some vascular beds is increased, but the mechanisms of altered brain regulation of sympathetic tone have not been previously investigated. The present study tested the hypothesis that increased EAA input to RVLM may contribute to regional increases in sympathetic tone, and the finding that bilateral microinjection of KYN into the RVLM lowers arterial pressure in 48 hr water deprived, but not water replete, rats supports 9

12 this hypothesis. Interestingly, KYN did not produce a consistent or statistically significant depressor response in 24 hr water deprived rats, despite the existence of elevated Na and Cl levels and volume contraction. Thus, it may be that a larger or more prolonged stimulus is required before brain regulation of sympathetic tone is significantly altered. In support of this, we found that the magnitude of the depressor response to KYN is correlated with the basal level of osmolality. Furthermore, a previous study investigating brain regions activated during water deprivation, as assessed by expression of the inducible transcription factor, c- fos, indicates that activation of some sites, such as the subfornical organ, is greater after 48 hr compared to 24 hr of water deprivation (22). The nature of the EAA input to RVLM is complex and incompletely understood. As observed in the present and previous studies, in normal rats, bilateral blockade of EAA receptors generally does not significantly alter arterial pressure (for reviews, see (6, 27). One interpretation of this finding is that EAA drive is zero in unstressed rats. However, Lipski et al (21) and Ito and Sved (18) have offered an alternative explanation, that EAA excitatory input is offset by EAA drive of inhibitory input. In support of this model, Ito and Sved (18) have shown that, following blockade of the inhibitory influence from CVLM or NTS, KYN into the RVLM produces a profound decrease in arterial pressure. Moreover, they propose that EAA drive of RVLM can increase if the normal balance of inhibition and excitation shifts toward excitation. Increased excitation could occur secondary to increased excitatory EAA input or increased sensitivity of the RVLM to EAA. Our test of the latter possibility indicated that water deprivation increases the responsiveness of RVLM to glutamate injections. Further experiments demonstrated that the increased pressor response was not due to enhanced vascular reactivity to 10

13 alpha-adrenergic agonists, since these responses were reduced. Decreased baroreflex sensitivity also cannot explain the larger glutamate pressor responses, since baroreflex gain appears to be increased during water-deprivation (2, 24), and reduced phenylephrine pressor responses were also observed in ganglionic-blocked water-deprived rats, which effectively removes the baroreflexes. Thus, we conclude that the greater falls in pressure in water deprived rats after KYN microinjection suggest an increased EAA drive of RVLM neurons, due in part to increased sensitivity of the RVLM to EAA. Increased net EAA input to RVLM induced by water deprivation could also be secondary to reduced inhibitory influences on RVLM neurons. However, the present results reveal that GABA elicits a larger decrease in blood pressure in water deprived rats, indicating that the increased EAA excitatory drive is not due to decreased responsiveness of the RVLM to GABA. The mechanism by which water deprivation increases net EAA excitatory input to RVLM or sensitivity of RVLM to glutamate was not directly investigated, but two possibilities can be considered. First, it may be secondary to the increased osmolality. Indeed, we observed a strong relationship between basal osmolality and the depressor response to KYN (Figure 2). Increased glutamate pressor responses have also been observed in normal Sprague-Dawley and Dahl salt-resistant rats placed on a high salt diet (15, 17, 23), and both water deprivation and a high salt diet (8, 13) increase body fluid osmolality. In conscious water-deprived rats, acute normalization of the elevated plasma osmolality decreases lumbar sympathetic nerve activity (25), suggesting that the hypertonicity contributes to sympathetic tone. The brain neurocircuitry by which increased osmolality activates the sympathetic nervous system has not been completely delineated (31). However, because acute increases in osmolality activate PVN neurons that 11

14 project to the RVLM (31), the hypertonicity may contribute to the increased EAA drive of RVLM observed in the present studies. Second, water deprivation decreases blood volume, and therefore presumably reduces inhibitory baroreceptor influences in RVLM. The present experiments demonstrate that RVLM responsiveness to glutamate increases following decreases in baroreceptor input after acute deafferentation, similar to the increased responses reported after blockade of the CVLM (18), and similar to the enhanced glutamate pressor responses observed in water deprived rats. The enhanced pressor responses could occur secondary to reduced convergent GABAergic inhibition, and suggest that the increased EAA drive exhibited by water deprived rats could be produced by without a change in EAA input. Alternatively, the volume depletion, when sensed by the renal baroreceptor, could lead to increases in renin secretion and elevated plasma levels of angiotensin II, which could act on the brain to ultimately increase EAA input to the RVLM (10). Thus, these findings are also consistent with the possibility that the volume depletion, via reduced baroreceptor input or increased circulating angiotensin II levels, also contributes to the increased responsiveness to glutamate and the increased EAA drive in the RVLM. In summary, the present data indicate that, during water deprivation, arterial pressure is supported by increased EAA drive of the RVLM. Because the pressor responsiveness of the RVLM to glutamate is also increased, the increased EAA drive may be due in part to enhanced sensitivity to EAA. 12

15 ACKNOWLEDGMENTS The authors wish to thank multiple colleagues for their help and advice on the brainstem microinjection technique, including Shaun Morrison, Alan Sved, Geoff Head, Julie Moffitt and Frank Gordon. In addition, we are grateful for the technical assistance of Yue Qi. This work was supported in part by NIH grants HL35872 and HL70962 and Grant-in-Aid N from the American Heart Association. 13

16 REFERENCES 1. Bergamaschi C, Campos RR, Schor N and Lopes OU. Role of the rostral ventrolateral medulla in maintenance of blood pressure in rats with Goldblatt hypertension. Hypertension 26: , Brooks VL. Vasopressin and ANG II in reflex regulation of heart rate: effect of water deprivation. Am J Physiol 263: R756-R761, Brooks VL, Huhtala TA, Silliman TL and Engeland WC. Water deprivation and rat adrenals mrnas for tyrosine hydroxylase and the norepinephrine transporter. Am J Physiol 272: R1897-R1903, Brooks VL and Osborn JW. Hormonal-sympathetic interactions in long-term regulation of arterial pressure: an hypothesis. Am J Physiol 268: R1343-R1358, Dampney RA, Horiuchi J, Tagawa T, Fontes MA, Potts PD and Polson JW. Medullary and supramedullary mechanisms regulating sympathetic vasomotor tone. Acta Physiol Scand 177: , Dampney RA, Tagawa T, Horiuchi J, Potts PD, Fontes M and Polson JW. What drives the tonic activity of presympathetic neurons in the rostral ventrolateral medulla? Clin Exp Pharmacol Physiol 27: ,

17 7. Esler M, Rumantir M, Kaye D and Lambert G. The sympathetic neurobiology of essential hypertension: disparate influences of obesity, stress, and noradrenaline transporter dysfunction? Am J Hypertens 14: 139S-146S, Fang Z, Carlson SH, Peng N and Wyss JM. Circadian rhythm of plasma sodium is disrupted in spontaneously hypertensive rats fed a high-nacl diet. Am J Physiol 278: R1490-R1495, Fejes-Toth G, Naray-Fejes-Toth A and Ratge D. Evidence against role of antidiuretic hormone in support of blood pressure during dehydration. Am J Physiol 249: H42- H48, Fink GD. Long-term sympatho-excitatory effects of angiotensin II: a mechanism of spontaneous and renovascular hypertension. Clinical and Experimental Pharmacology and Physiology 24: 91-95, Gardiner SM and Bennett T. Interactions between neural mechanisms, the reninangiotensin system and vasopressin in the maintenance of blood pressure during water deprivation: studies in Long Evans and Brattleboro rats. Clin Sci 68: , Guyenet PG. Role of the ventral medulla oblongata in blood pressure regulation. In: Central Regulation of Autonomic Functions, edited by Loewy AD and Spyer KM. 15

18 New York: Oxford University Press, 1990, p Habecker BA, Grygielko ET, Huhtala TA, Foote B and Brooks VL. Ganglionic tyrosine hydroxylase and norepinephrine transporter are decreased by increased sodium chloride in vivo and in vitro. Auton Neurosci 107: 85-98, Hatton DC, Brooks VL, Qi Y and McCarron DA. Cardiovascular response to stress: baroreflex resetting and hemodynamics. Am J Physiol 272: R1588-R1594, Ito S, Gordon FG and Sved AF. Dietary salt intake alters cardiovascular responses evoked from the rostral ventrolateral medulla. Am J Physiol 276: R1600-R1607, Ito S, Komatsu K, Tsukamoto K and Sved AF. Excitatory amino acids in the rostral ventrolateral medulla support blood pressure in spontaneously hypertensive rats. Hypertension 35: , Ito S, Komatsu K, Tsukamoto K and Sved AF. Tonic excitatory input to the rostral ventrolateral medulla in Dahl salt-sensitive rats. Hypertension 37: , Ito S and Sved AF. Tonic glutamate-mediated control of rostral ventrolateral medulla. Am J Physiol 273: R487-R494, Kiely JM and Gordon FJ. Non-NMDA receptors in the rostral ventrolateral medulla 16

19 mediate somatosympathetic pressor responses. J Auton Nerv Syst 43: , Kiss A, Jezova D and Aguilera G. Activity of the hypothalamic pituitary adrenal axis and sympathoadrenal system during food and water deprivation in the rat. Brain Res 663: 84-92, Lipski J, Kanjhan R, Kruszewska B and Rong WF. Properties of presympathetic neurones in the rostral ventrolateral medulla in the rat: An intracellular study 'in vivo'. J Physiol (Lond ) 490: , Morien A, Garrard L and Rowland NE. Expression of Fos immunoreactivity in rat brain during dehydration: effect of duration and timing of water deprivation. Brain Res 816: 1-7, Pawloski-Dahm CM and Gordon FJ. Increased dietary salt sensitizes vasomotor neurons of the rostral ventrolateral medulla. Hypertension 22: , Russ RD, Brizzee BL and Walker BR. Role of vasopressin in cardiovascular responses to acute and chronic hyperosmolality. Am J Physiol 262: R25-R32, Scrogin KE, Grygielko ET and Brooks VL. Osmolality: a physiological long-term regulator of lumbar sympathetic nerve activity and arterial pressure. Am J Physiol 276: R1579-R1586,

20 26. Scrogin KE, McKeogh DF and Brooks VL. Is osmolality a long-term regulator of renal sympathetic nerve activity in conscious water-deprived rats? Am J Physiol 282: R560-R568, Sved AF, Ito S, Madden CJ, Stocker SD and Yajima Y. Excitatory inputs to the RVLM in the context of the baroreceptor reflex. Ann N Y Acad Sci 940: , Sved AF, Ito S and Sved JC. Brainstem mechanisms of hypertension: role of the rostral ventrolateral medulla. Curr Hypertens Rep 5: , Sved AF, Ito S and Yajima Y. Role of excitatory amino acid inputs to the rostral ventrolateral medulla in cardiovascular regulation. Clin Exp Pharmacol Physiol 29: , Tagawa T and Dampney RAL. AT 1 receptors mediate excitatory inputs to rostral ventrolateral medulla pressor neurons from hypothalamus. Hypertension 34: , Toney GM, Chen QH, Cato MJ and Stocker SD. Central osmotic regulation of sympathetic nerve activity. Acta Physiol Scand 177: 43-55, Trapani AJ, Undesser KP, Kehrer P and Bishop VS. Neurohumoral interactions in conscious dehydrated rabbit. Am J Physiol 254(23): R338-R347,

21 33. Yang Z, Bertram D and Coote JH. The role of glutamate and vasopressin in the excitation of RVL neurones by paraventricular neurones. Brain Res 908: ,

22 Table: Basal values Water Replete 24 hour water 48 hour Water Deprived Deprived Basal MAP ± ± ± 3.1 (mmhg) n=28 n=7 n=27 Na ± ± 1.8 * ± 0.5 * (meq/l) n=28 n=5 n=25 Cl ± ± 1.9 * ± 0.3 * (meq/l) n=25 n=5 n=25 Hematocrit 44.5 ± ± 0.5 * 51.1 ± 0.4 * (%) n=27 n=11 n=24 Protein 6.2 ± ± ± 0.1 * (g/dl) n=28 n=11 n=23 * indicates P<0.05 compared to water replete by ANOVA post hoc Bonferroni Correction. n is the number of rats. 20

23 FIGURE LEGENDS. 1. Coronal section through rat medulla illustrating sites of drug microinjections into RVLM in water replete (open circles), 24 hr water deprived (closed triangles) and 48 hr water deprived (closed circles). Section is mm from bregma; all injections were within " 200 µm from this section. 2. Change in mean arterial pressure (MAP) in response to bilateral kynurenate microinjection into RVLM (top) and plasma osmolality (bottom) in water replete (WR; open bar; n=8; control MAP, 108"4 mmhg), 24 hr water deprived (24 WD; hatched bar; n=6; control MAP, 104"8 mmhg) and 48 hr water deprived (48 WD; solid bar; n=7; control MAP, 107"9 mmhg) rats. Water deprivation increased osmolality and the change in MAP (P<0.001, ANOVA for both). Microinjection of artificial CSF (shaded bars; n=5-6) did not significantly alter MAP. Insert shows the relationship between the change in MAP (ª MAP, mmhg) and the basal osmolality (Osm, mosmol/kg H 2 O) for individual WR (open circles), 24 WD (half closed circles) and 48 WD (closed circles) rats. *: P<0.05 compared to water replete, H: P<0.05 compared to 24 WD by Bonferroni Correction. 3. Top: Time course of changes in mean arterial pressure (MAP) following bilateral microinjection of kynurenate (KYN) into the RVLM of water deprived (WD; solid circles; n=7) and water replete (WR; open circles; n=8) rats. Control MAP as in Figure 2. The arrow indicates the initiation of the first injection; the second injection was usually completed within one min. 21

24 The decrease in pressure was greater in WD rats (P<0.001, ANOVA, group and interaction). Bottom: representative tracing of bilateral KYN microinjection into RVLM of a WD rat. Arrows indicate time of injections of KYN into the RVLM of each side of the brain. 4. Effect of unilateral glutamate microinjection into the RVLM on mean arterial pressure (MAP) in water deprived (WD, closed circles; n=8) and water replete (WR, open circles; n=7) rats. MAP (in mmhg) just prior to injections for WR rats was (glutamate doses in nmol in parentheses): 106"3 (0.1), 106"3 (0.5), 108"4 (1.0) and 104"4 (5.0), and for WD rats was: 108"7 (0.1), 109"6 (0.5), 113"6 (1.0) and 105"7 (5.0). Pressor responses were greater in WD rats (P<0.05, ANOVA, dose, group and interaction). Unilateral artificial CSF injections did not alter MAP in either WD (change is -0.3"0.3 mmhg, n=9) or WR (change is -0.8"0.5, n=11) rats. 5. Effect of bolus iv phenylephrine injections on mean arterial pressure (MAP) in intact (top) and V1-pretreated/ hexamethonium-pretreated (bottom) water replete [WR, open circles; n=8 (intact), n=7 (pretreated)] and water deprived [WD, closed circles; n=7 (intact), n=5 (pretreated)] rats. MAP (in mmhg) just prior to injections in intact WR rats was (phenylephrine doses in µg/kg in parentheses): 103"3 (1), 108"2 (3), 110"3 (6), 104"2 (10) and 102"2 (20), and in intact WD rats was: 109"8 (1), 111"6 (3), 110"6 (6), 111"7 (10) and 107"6 (20). In rats pretreated with the V1-antagonist and hexamethonium, MAP (in mmhg) just prior to injections was (phenylephrine doses in parentheses): 51"3 (1), 47"4 (3), 50"3 (6), 50"3 (10) and 54"4 (20), and in WD rats was: 41"2 (1), 36"2 (3), 38"2 (6), 40"2 (10) and 41"2 (20). Pressor 22

25 responses were smaller in both untreated and pretreated WD rats (P<0.01, ANOVA, dose, group and interaction). 6. Effect of unilateral microinjection of gamma-aminobutyric acid (GABA) on mean arterial pressure (MAP) in water replete [open bars; n=9; control MAP, 104"7 mmhg (1 nmol) and 106"6 mmhg (10 nmol)], 24 hr water deprived [24 WD; n=9; 105"7 mmhg (1 nmol) and 103"5 mmhg (10 nmol)] and 48 hr water deprived [48 WD; n=6; control MAP, 88"7 mmhg (1nmol) and 92"6 mmhg (10nmol)] rats. *: P<0.05 compared to replete by Bonferroni Correction. 23

26 Replete 24 WD 48 WD Ο Ο Ο Ο Ο Ο Ο Ο Ο Ο Ο Ο Ο Ο Ο Ο Ο Ο Ο Ο ΟΟ Figure 1. 24

27 0 Change in MAP (mmhg) MAP Osm * * Osmolality (mosmol/kg H 2 0) WR 24 WD 48 WD Figure 2. 25

28 20 WR (n=8) WD (n=7) Change in MAP (mmhg) Begin KYN Time (min) AP (mmhg) MAP (mmhg) HR (bpm) Kyn: 2.7 nmol 1 min Figure 3. 26

29 Change in MAP (mmhg) WR (n=7) WD (n=8) Dose of Glutamate (nmoles) Figure 4. 27

30 50 WR (n=8) WD (n=7) Change in MAP (mmhg) Dose of Phenylephrine (µg/kg) Change in MAP (mmhg) WR (n=7) WD (n=5) Dose of Phenylephrine (µg/kg) Figure 5. 28

31 0 1 nmol GABA 10 nmol GABA Change in MAP (mmhg) * * Replete (n=9) 24 WD (n=9) 48 WD (n=6) * Figure 6. 29