Does angiotensin II have a significant tonic action. on cardiovascular neurons in the rostral and caudal

Transcription

1 Am J Physiol Regulatory Integrative Comp Physiol 279: R1392 R1402, Does angiotensin II have a significant tonic action on cardiovascular neurons in the rostral and caudal VLM? P. D. POTTS, 1 A. M. ALLEN, 2 J. HORIUCHI, 1 AND R. A. L. DAMPNEY 1 1 Department of Physiology and Institute for Biomedical Research, The University of Sydney, New South Wales 2006; and 2 Howard Florey Institute of Experimental Physiology and Medicine, The University of Melbourne, Victoria 3101, Australia Received 10 February 2000; accepted in final form 26 May 2000 Potts, P. D., A. M. Allen, J. Horiuchi, and R. A. L. Dampney. Does angiotensin II have a significant tonic action on cardiovascular neurons in the rostral and caudal VLM? Am J Physiol Regulatory Integrative Comp Physiol 279: R1392 R1402, The peptidic ANG II receptor antagonists [Sar 1,Ile 8 ]ANG II (sarile) or [Sar 1,Thr 8 ]ANG II (sarthran) are known to decrease arterial pressure and sympathetic activity when injected into the rostral part of the ventrolateral medulla (VLM). In anesthetized rabbits and rats, the profound depressor and sympathoinhibitory response after bilateral microinjections of sarile or sarthran into the rostral VLM was unchanged after prior selective blockade of angiotensin type 1 (AT 1 ) and ANG-(1O7) receptors, although this abolished the effects of exogenous ANG II. Unlike the neuroinhibitory compounds muscimol or lignocaine, microinjections of sarile in the rostral VLM did not affect respiratory activity. Sarile or sarthran in the caudal VLM resulted in a large pressor and sympathoexcitatory response, which was also unaffected by prior blockade of AT 1 and ANG-(1O7) receptors. The results indicate that the peptidic ANG receptor antagonists profoundly inhibit the tonic activity of cardiovascular but not respiratory neurons in the VLM and that these effects are independent of ANG II or ANG-(1O7) receptors. sympathetic vasomotor tone; arterial pressure; medulla oblongata THE VENTROLATERAL MEDULLA (VLM) in the brain stem contains several populations of neurons that are essential in the regulation of the cardiovascular system. In particular, the rostral part of the VLM contains a group of tonically active neurons that project to and excite sympathetic vasomotor and cardiac preganglionic neurons in the spinal cord. The rostral VLM sympathoexcitatory neurons receive excitatory inputs from a variety of regions, including a GABAergic inhibitory input from a group of neurons in the caudal VLM. Both the rostral and caudal VLM neurons play a critical role in the tonic and phasic control of sympathetic vasomotor activity and arterial pressure (for reviews see Refs. 10, 14). Although it is well established that the tonic resting activity of rostral VLM sympathoexcitatory neurons is Address for reprint requests and other correspondence: R. A. L. Dampney, Dept. of Physiology, F13, The Univ. of Sydney, Sydney, NSW 2006, Australia ( rogerd@physiol.usyd.edu.au). of crucial importance in the maintenance of sympathetic vasomotor tone, there is considerable controversy concerning the mechanisms that generate this tonic activity (10, 14, 21). However, it has been demonstrated that this tonic activity is not dependent on excitatory amino acid receptors in the rostral VLM, because bilateral injections of excitatory amino acid receptor antagonists into this region have little effect on arterial pressure (19, 31). On the other hand, recent experiments suggest that ANG II may play a role in the tonic excitation of rostral VLM neurons. There is a high density of ANG II receptors in the rostral VLM of various species, including humans (2 4, 6, 23), and their location corresponds very closely to that of sympathoexcitatory neurons in this region (4, 7). Microinjection of ANG II into the rostral VLM elicits an increase in arterial pressure and sympathetic activity (4, 16, 24, 28). In contrast, unilateral microinjections of peptidic ANG II receptor antagonists, such as [Sar 1,Ile 8 ]ANG II (sarile) or [Sar 1,Thr 8 ]ANG II (sarthran), into the rostral VLM result in a moderate fall in arterial pressure and sympathetic activity in the rabbit (28) and rat (24). Interestingly, however, bilateral microinjections of these antagonists into the rostral VLM result in a profound fall in arterial pressure (18, 33), which is accompanied by a large fall in renal sympathetic nerve activity (33). The results of these experiments in different species therefore suggest that ANG peptides may play an important role in maintaining the tonic activity of sympathoexcitatory neurons in the rostral VLM. Sympathoinhibitory neurons in the caudal VLM are also tonically active (10, 20). In this case, the tonic activity has been shown to depend, at least in part, on a tonic excitatory input to these neurons mediated by excitatory amino acid receptors (20). It is possible, however, that ANG II may also play a role in the tonic excitation of caudal VLM neurons. Like the rostral VLM, the caudal VLM also has a high density of ANG II receptors, as demonstrated in different species (3, 6, 23). Microinjection of ANG II into the caudal VLM results in a depressor and sympathoinhibitory re- The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. R /00 $5.00 Copyright 2000 the American Physiological Society

2 R1393 sponse (5, 24, 28), whereas microinjection of sarthran or sarile into the caudal VLM results in an increase in resting arterial pressure and sympathetic nerve activity (5, 24, 28). The nonselective ANG receptor antagonists sarthran and sarile used in the studies mentioned above are likely to act on all ANG II receptors, of which the AT 1 and AT 2 receptors have been cloned (see Ref. 6 for review). Therefore it is not clear whether the sympathoinhibitory and sympathoexcitatory effects that result from the injection of sarile or sarthran into the rostral and caudal VLM, respectively, are due specifically to blockade of AT 1 receptors, AT 2 receptors, or to some other effect. In the VLM of most species, the ANG receptors appear to be primarily of the AT 1 subtype (1, 22, 38). Surprisingly, however, microinjection of the selective AT 1 receptor antagonist losartan has little effect on resting arterial pressure or sympathetic nerve activity while still blocking the action of exogenous ANG II (8, 12, 16). Similarly, microinjection of the selective AT 2 receptor antagonist PD also has little effect on resting arterial pressure or sympathetic nerve activity (16). The contrasting effects of selective and nonselective ANG receptor antagonists are difficult to resolve. It is possible that sarthran and sarile act on receptors other than the AT 1 or AT 2 subtype. For example, recent work in the rostral VLM of the rat has suggested that a specific receptor subtype for the heptapeptide ANG- (1O7) is involved in the tonic regulation of arterial pressure (12). However, bilateral blockade of ANG- (1O7) receptors in the rostral VLM results in a much more modest reduction in arterial pressure (12) compared with that resulting from injection of sarthran or sarile (18, 33). Alternatively, because the selective AT 1 receptor antagonist losartan has a lower affinity for the AT 1 receptor compared with nonselective antagonists such as sarthran or sarile (29, 30), it is possible that a higher concentration of losartan may be required to equal the potency of the peptidic receptor antagonists. However, it is difficult to test this, because losartan appears to have nonspecific effects at higher concentrations (8). A third possibility is that combined blockade of AT 1 and ANG-(1O7) receptors, as produced by sarthran or sarile, results in a much greater effect on VLM neurons than blockade of either AT 1 or ANG- (1O7) receptors alone. Finally, it is possible that the peptidic antagonists have effects that are independent of their interaction with ANG receptors. The aims of the present study were to determine whether the effects on arterial pressure and sympathetic nerve activity produced by microinjection of peptidic ANG II receptor antagonists into the caudal or rostral VLM are dependent on AT 1 and/or ANG-(1O7) receptors, and whether these effects are due to antagonism of the actions of ANG II. In particular, we have tested 1) whether selective blockade of AT 1 receptors in the rostral VLM with the antagonist candesartan [which has a high affinity for AT 1 receptors, similar to that of sarthran or sarile (25)] has effects on arterial pressure and sympathetic nerve activity similar to those of the peptidic antagonists; 2) whether the cardiovascular effects of microinjection of the peptidic antagonists in the rostral and caudal VLM are affected by prior blockade of AT 1 and ANG-(1O7) receptors in these regions; and 3) whether the actions of the peptidic antagonists are due to a general neuroinhibitory effect. MATERIALS AND METHODS All procedures were performed in accordance with the Australian National Health and Medical Research Council Code of Practice. General Procedures in Rabbits Experiments were performed on New Zealand white rabbits (Laboratory Animal Services, University of Sydney) or mixed strain rabbits (Baker Medical Research Institute) ( kg body wt) of either sex. A marginal ear vein was cannulated, and the animals were anesthetized by administration of either pentobarbital sodium (35 mg/kg iv initially, followed by 9 15 mg kg 1 h 1 ) or urethan (1.5 g/kg iv). The body temperature was maintained within the range of C by a thermoregulated lamp. The trachea was cannulated, and the rabbit was ventilated with oxygen-enriched room air at a rate of 30 breaths/min at a volume that maintained end-tidal carbon dioxide at %. Catheters were placed in a femoral artery and femoral vein. The carotid sinus baroreceptors were denervated, as previously described (28), and the aortic and vagal nerves were cut to eliminate baroreceptor reflex effects that may have been secondarily evoked as a consequence of the changes in arterial pressure evoked by injection of ANG receptor antagonists into the VLM. The effectiveness of the baroreceptor denervation was confirmed by the absence of reflex heart rate responses to pharmacologically induced changes in blood pressure. The dorsal medulla was then exposed according to the procedures described previously (28). In most experiments the renal nerve was then exposed, whereas in the remainder, either the phrenic nerve or diaphragm was exposed to allow recording of respiratory activity (see Measurement of respiratory activity). After all surgical procedures, neuromuscular blockade was induced by administration of alcuronium chloride (0.1 mg/kg iv every 1 2 h). The effects of alcuronium chloride were allowed to wear off before additional doses were administered. The adequacy of anesthesia without neuromuscular blockade was verified by the absence of a withdrawal response to nociceptive stimulation and during neuromuscular blockade by a stable arterial pressure, heart rate, and renal sympathetic nerve activity (RSNA). The arterial pressure was measured via the femoral arterial catheter, and the mean arterial pressure (MAP) and heart rate were derived from the pulsatile signal by means of a low-pass filter and ratemeter, respectively. All signals were displayed on a polygraph chart recorder. Renal nerve recording. After exposure of the renal nerve, the distal end was crushed to eliminate afferent discharge, and the proximal end was placed on bipolar silver recording electrodes and covered with mineral oil to prevent drying. The signal from the electrodes was amplified, filtered (100 1,000 Hz), displayed using a MacLab system (ADInstruments, NSW, Australia), and monitored by means of an audio amplifier. The filtered nerve activity signal was rectified, integrated (resetting every 5 s), and recorded on a polygraph chart recorder and on videotape. At the end of the experiment, the baseline noise level was established by applying

3 R1394 local anesthetic to the renal nerve and crushing the nerve proximal to the recording electrodes. Measurement of respiratory activity. Respiratory activity was measured by recording the efferent discharge of the phrenic nerve or by measuring the electromyographic (EMG) activity of the diaphragm. The right phrenic nerve was isolated in the neck via a ventral approach, and its activity was recorded using bipolar electrodes, as described above for the renal nerve. The diaphragm EMG was measured via electrodes inserted into the muscle, which was exposed by a retroperitoneal approach. The phrenic nerve activity or diaphragm EMG signal was then rectified, filtered, and recorded on a polygraph chart recorder and on videotape. General Procedures in Rats Male Sprague-Dawley rats ( g body wt) were anesthetized initially with sodium brietal (80 mg/kg ip), tracheotomized, and artificially ventilated with 100% oxygen. The ventilation rate and volume were adjusted to keep arterial blood PCO 2 in the range of mmhg. Anesthesia was maintained by inhalation of % isoflurane at a level that abolished pedal withdrawal and corneal reflexes. Body temperature was maintained at 37 1 C with a servocontrolled heating blanket. The left jugular vein and left common carotid artery were cannulated for the intravenous administration of drugs and measurement of arterial pressure, respectively. Pulsatile arterial pressure, MAP, and heart rate were displayed and recorded using a MacLab system. The rat was placed in a stereotaxic frame with the head ventroflexed at 45 from horizontal. The atlantooccipital membrane was exposed through a midline incision and opened to expose the dorsal surface of the brain stem. After completion of all surgical procedures and establishment of a stable anesthetic plane, neuromuscular blockade was induced with pancuronium bromide administered in a bolus dose of 1 mg/kg iv. The neuromuscular blockade wore off after min, at which time the level of anesthesia was assessed again before readministration of pancuronium. Intramedullary Microinjections Microinjections of various compounds were made into sites within the rostral or caudal VLM of rabbits or rats by use of a glass micropipette held in place by a micromanipulator (Kopf Instruments). In rabbits, the compounds injected were sodium glutamate (8 12 nl of 200 mm solution), ANG II (Sigma, 40 nl of 1 mm solution), the ANG receptor antagonists sarthran or sarile (Sigma, 100 nl of 1 or 10 mm solution), the AT 1 receptor antagonist losartan (100 nl of 10 mm solution, kind gift of Merck, Whitehouse Station, NJ), the ANG-(1O7) receptor antagonist [D-Ala 7 ]ANG-(1O7) (Auspep or Bachem, 100 nl of 10 mm solution), the GABA A receptor agonist muscimol (Sigma, 200 nl of 2 mm solution), or the local anesthetic lignocaine (100 nl of 2% solution). In rats, the compounds injected were sodium glutamate (25 50 nl of 50 mm solution), ANG II (Auspep, 100 nl of 1 mm solution), sarile (Auspep, 100 nl of 1 mm solution), or the AT 1 receptor antagonist candesartan (100 nl of 10 mm solution, kind gift of Astra Hassle, Molndal, Sweden). The vehicle solution was either phosphate-buffered saline (ph 7.4) or artificial cerebrospinal fluid (ph 7.4). Injections were made by pressure injection, as described previously (28). The volume injected was measured by observing the displacement of the meniscus in the pipette with respect to a horizontal grid viewed through an operating microscope. The pressor region in the rostral VLM was identified bilaterally as the site where microinjection of glutamate resulted in a rapid increase in arterial pressure (usually mmhg in rabbits and 20 mmhg in rats) with a short onset latency ( 5 s). The pipette containing glutamate was then replaced with a pipette containing an ANG receptor antagonist. The antagonist was then microinjected into the physiologically identified pressor site on one side, and the pipette was then left in place for 1 2 min to allow the drug to diffuse away from the pipette tip. The pipette was then removed and placed into the contralateral pressor site, and the antagonist was then injected into that site. In some experiments, further microinjections were made into the rostral VLM pressor region of ANG II and/or ANG receptor antagonists. The depressor region in the caudal VLM was identified in rabbits as the site at which microinjection of glutamate resulted in a rapid decrease in arterial pressure of 30 mmhg with a short onset latency ( 5 s). The subsequent procedure was then the same as described above for the rostral VLM pressor region. Statistical Analysis Comparisons between means were made using the paired or unpaired t-test. A P value of 0.05 was taken to indicate a statistically significant difference. All values are expressed as means SE. RESULTS Effects of Bilateral Microinjections of Selective AT 1 Receptor Antagonists Rostral VLM. In six rabbits, bilateral microinjection of the selective AT 1 receptor antagonist losartan (1 nmol) into the functionally identified pressor region in the rostral VLM had little effect on baseline arterial pressure, heart rate, and RSNA, except for small and transient increases in these variables that lasted 2 min (Fig. 1A, Table 1). After injections of losartan into the rostral VLM pressor region, microinjection of ANG II (40 pmol) into the same site had no effect on arterial pressure, heart rate, and RSNA, confirming previous observations from our laboratory that this dose of losartan is sufficient to block the actions of exogenous ANG II in the rostral VLM of the rabbit (16). In five rats, bilateral injections of the high-affinity AT 1 receptor antagonist candesartan (1 nmol) into the functionally identified pressor region in the rostral VLM also resulted in no significant changes in either MAP ( 1 5 mmhg) or heart rate (6 8 beats/min). Caudal VLM. Bilateral microinjections of losartan (n 4) into the functionally identified depressor region in the caudal VLM of barodenervated rabbits also had little effect on baseline arterial pressure, heart rate, and RSNA, except for small and transient decreases in these variables that lasted 2 min (Fig. 1B, Table 1). After injections of losartan into the caudal VLM depressor region, microinjection of ANG II (40 pmol) into the same site in three rabbits had little effect on arterial pressure, heart rate, and RSNA, in contrast to the prolonged and significant depressor and sympathoinhibitory effect produced by microinjection of similar or smaller doses of exogenous ANG II in the caudal VLM

4 R1395 Fig. 1. Chart recordings showing the lack of effect of bilateral injections (L, left; R, right) of losartan into the pressor region in the rostral ventrolateral medulla (VLM) (A) or the depressor region in the caudal VLM (B) of the anesthetized rabbit, apart from small and transient changes in arterial pressure and renal sympathetic nerve activity (RSNA). The dose of losartan was sufficient to abolish the effects of subsequent microinjection of ANG II into the same sites. bpm, beats/ min. of rabbits not pretreated with an AT 1 receptor antagonist (28). Effects of Bilateral Microinjections of Nonselective ANG Receptor Antagonists Rostral VLM. Bilateral microinjections of the nonselective ANG receptor antagonist sarthran into the rostral VLM of barodenervated rabbits caused a dosedependent decrease in arterial pressure, heart rate, and RSNA (Table 1, Fig. 2). Injection of the drug into one side typically caused a modest fall in arterial pressure, heart rate, and RSNA, but subsequent microinjection into the contralateral side caused much larger falls in these variables (Fig. 2). The duration of the depressor response was also dose dependent and was 20 min after microinjection of the larger dose (1 nmol) into the rostral VLM (Table 1). In rats, bilateral microinjections of 0.1 nmol of sarile into the rostral VLM pressor region caused a decrease inmapof25 4 mmhg (n 9), which was very similar in magnitude to that evoked by the same dose of sarthran in the rabbit (Table 1). In contrast to the rabbit, however, the change in heart rate (2 12 beats/min) was not significant. Caudal VLM. Microinjection of sarthran into the depressor region in the caudal VLM of barodenervated rabbits caused a dose-dependent increase in arterial pressure, heart rate, and RSNA (Table 1, Fig. 3). A small increase in these variables occurred after microinjection of sarthran into one side, but there was a much greater increase after microinjection into the contralateral side (Fig. 3). The duration of the pressor response was also dose dependent and was 30 min after microinjection of the larger dose (1 nmol) into the caudal VLM (Table 1). Effects of Bilateral Microinjection of Nonselective ANG Receptor Antagonists After AT 1 Receptor Antagonists Rostral VLM. In six rabbits, bilateral microinjections of 1 nmol of sarthran were made into the rostral VLM pressor region 5 20 min after bilateral microinjections of losartan into the same sites. As described above, microinjections of losartan (1 nmol) had little effect on resting arterial pressure and RSNA. In four of these experiments, the effectiveness of AT 1 receptor blockade was confirmed by the lack of response to microinjections of ANG II (40 pmol) into the same sites (Fig. 4).

5 R1396 Table 1. Effects on cardiovascular variables of bilateral microinjections of sarthran or losartan into the rostral and caudal VLM of the anesthetized rabbit MAP, mmhg Heart Rate, beats/min RSNA, % baseline Drug Baseline Change Baseline Change Baseline Change Duration, min Rostral VLM Sarthran 0.1 nmol (n 4) * * * nmol (n 6) * * * 22 4 Losartan 1 nmol (n 6) * Caudal VLM Sarthran 0.1 nmol (n 3) * * * nmol (n 5) * * * 32 3 Losartan 1 nmol (n 4) * Values are means SE. VLM, ventrolateral medulla; MAP, mean arterial pressure; RSNA, renal sympathetic nerve activity. *P 0.05 vs. baseline levels. After blockade of AT 1 receptors, microinjections of sarthran still resulted in large decreases in arterial pressure, heart rate, and RSNA, which were of similar magnitude to those evoked by sarthran without pretreatment with losartan (Table 2). In five rats, microinjections of a mixture of sarile (100 pmol) and ANG II (100 pmol) were made bilaterally into the rostral VLM pressor region 5 15 min after microinjections of candesartan (1 nmol) into the same sites. Bilateral microinjections of the sarile/ang II mixture resulted in a decrease in MAP of 23 2 mmhg, which was not significantly different from the decrease in MAP evoked by this dose of sarile in the group of nine rats referred to above that were not pretreated with candesartan (25 4 mmhg). As before, however, the change in heart rate (12 6 beats/ min) after bilateral injections of sarile was not significant. Caudal VLM. After bilateral microinjections of losartan into the caudal VLM depressor region, bilateral microinjections of sarthran in four rabbits resulted in increases in MAP, heart rate, and RSNA that were of similar magnitude to those evoked by sarthran without pretreatment with losartan (Fig. 5, Table 2). Effects of Bilateral Microinjection of Nonselective ANG Receptor Antagonists After AT 1 and ANG-(1O7) Receptor Antagonists In five experiments in rabbits, 1 nmol of losartan and 1 nmol of the selective ANG-(1O7) receptor antagonist [D-Ala 7 ]ANG-(1O7) (27) were both injected bilaterally into the rostral VLM pressor region before subsequent bilateral injections of 1 nmol of sarthran (in 2 experiments) or sarile (in the remaining 3 experiments). We have shown previously (Potts and Dampney, unpublished observations) that 1 nmol of [D-Ala 7 ]ANG-(1O7) is sufficient to abolish the effects of ANG-(1O7) in the rostral and caudal VLM. Even after injections of both AT 1 and ANG-(1O7) receptor antagonists, bilateral microinjections of sarthran or sarile evoked decreases in MAP, heart rate, and RSNA that were very similar in magnitude to those evoked without pretreatment with these selective ANG receptor antagonists (Table 2). Fig. 2. Chart recording showing the profound fall in arterial pressure and RSNA after bilateral injections of sarthran into the pressor region in the rostral VLM of an anesthetized rabbit.

6 R1397 Fig. 3. Chart recording showing the large increase in arterial pressure and RSNA after bilateral injections of sarthran into the depressor region in the caudal VLM of an anesthetized rabbit. Effects of Microinjection of Nonselective ANG Antagonists on Respiratory Activity In four experiments in spontaneously breathing rabbits, unilateral microinjection of 1 nmol of sarile into the rostral VLM pressor region resulted in a significant fall in arterial pressure of 35 3 mmhg but had no detectable effect on the peak amplitude or frequency of phrenic nerve discharge or diaphragm EMG (Fig. 6A). Unilateral microinjection of muscimol (0.4 nmol in 200 nl, n 3) or local anesthetic (lignocaine, 2% in 100 nl, n 2) in the rostral VLM pressor region resulted in a similar decrease in MAP (29 2 mmhg), but in contrast to the effects of sarile, it also resulted in a clearcut decrease (24 2%) in the peak amplitude of phrenic nerve discharge (Fig. 6B) or diaphragm EMG activity. DISCUSSION The results of this study demonstrate that bilateral microinjections of the peptidic ANG receptor antagonists sarthran and sarile into the rostral VLM pressor region of the rabbit lead to a profound decrease in arterial pressure and RSNA, in support of previous observations in the rat (18, 33), and show for the first time that bilateral microinjections of these compounds into the caudal VLM result in a large increase in arterial pressure and RSNA. Our results also show for the first time that the decrease in arterial pressure evoked by sarile in the rostral VLM was not accompanied by any effect on respiratory activity, in contrast to muscimol or lignocaine, which reduced both arterial pressure and respiratory activity. The major new finding of this study, however, is that the profound depressor and sympathoinhibitory effects evoked by sarile and sarthran in the rostral VLM of two species (rabbit and rat) are not mediated by receptors for ANG II, because these effects were unaffected after microinjections of selective ANG receptor antagonists that completely blocked the actions of exogenous ANG II. In a previous study, Ito and Sved (18) found that the depressor effect evoked by sarile in the rostral VLM was blocked when ANG II was co-injected with Fig. 4. Chart recording showing effects of bilateral injections of sarthran into the pressor region in the rostral VLM in an anesthetized rabbit, subsequent to bilateral injections of losartan and ANG II into the same sites. Note that sarthran induces a profound fall in arterial pressure and RSNA even under conditions in which the normal pressor and sympathoexcitatory actions of ANG II are completely blocked.

7 R1398 Table 2. Effects on cardiovascular variables of bilateral microinjections of sarthran or sarile into the rostral or caudal VLM after blockade of AT 1 or ANG-(1O7) receptors in the rostral or caudal VLM MAP, mmhg Heart Rate, beats/min RSNA, % baseline Rostral VLM Sarthran (1 nmol) alone (n 6) Sarthran (1 nmol) after blockade of AT 1 receptors (n 6) Sarthran or sarile (1 nmol) after blockade of AT 1 and ANG- (1O7) receptors (n 6) Caudal VLM Sarthran (1 nmol) alone (n 5) Sarthran (1 nmol) after blockade of AT 1 receptors (n 4) Values are means SE., change; AT, receptor, ANG type 1 receptor. this compound. This finding was interpreted to indicate that sarile and ANG II compete for the same receptors. An alternative explanation for this finding, however, is that ANG II and sarile have excitatory and inhibitory effects that are mediated by different receptors, which tend to balance each other when the two compounds are co-administered. Such an interpretation would be consistent with the findings of the present study, as well as the previous study of Ito and Sved (18). The results also confirm previous findings (8, 12) that bilateral microinjections into the rostral VLM pressor region of the AT 1 receptor antagonist losartan do not reduce resting arterial pressure or sympathetic activity. Losartan has a lower affinity for AT 1 receptors than the peptidic antagonists sarile and sarthran (29, 30), so it is therefore conceivable that the lack of effect of losartan is due to incomplete blockade of AT 1 receptors. This seems unlikely, however, because 1) the dose of losartan injected (1 nmol) was sufficient to completely block the pressor effect of exogenous ANG II in the rostral VLM, and 2) another AT 1 receptor antagonist, candesartan, which has a similar affinity to that of sarthran and sarile for these receptors (25), also had no effect on resting arterial pressure when injected in the same dose as the latter compounds. A previous study found that bilateral microinjection of [D-Ala 7 ]ANG-(1O7), an antagonist of ANG-(1O7) receptors, into the rostral VLM of the rat leads to a decrease in arterial pressure (12), suggesting the possibility that the effects of sarile and sarthran are due to blockade of ANG-(1O7) receptors. The results of our experiments in the rabbit, however, show that bilateral microinjection of [D-Ala 7 ]ANG-(1O7) into the rostral VLM pressor region, even in combination with losartan, did not reduce the depressor and sympathoinhibitory response to subsequent bilateral injections of sarile or sarthran. The dose of [D-Ala 7 ]ANG-(1O7) injected (1 nmol) has been shown to completely block the pressor effects of ANG-(1O7) in the rostral VLM of the rabbit (Potts and Dampney, unpublished observations). In addition, bilateral injections of [D-Ala 7 ]ANG- (1O7), like losartan, had no detectable effect on the resting level of arterial pressure or RSNA in the rabbit. Furthermore, even in the rat, the decrease in resting arterial pressure resulting from bilateral injections of [D-Ala 7 ]ANG-(1O7) in the rostral VLM is rather modest ( 15 mmhg) (12), much less than that evoked by bilateral injections of sarthran or sarile ( 50 mmhg) (18). Apart from AT 1 and ANG-(1O7) receptors, sarthran and sarile also block the AT 2 and AT 4 receptor subtypes (37). It is unlikely that the effects of sarthran and sarile are due to blockade of AT 2 receptors, because a previous study from our laboratory has shown that selective blockade of these receptors in the rostral VLM of the rabbit has little effect on resting arterial pres- Fig. 5. Chart recording showing effects of bilateral injections of sarthran into the depressor region in the caudal VLM in an anesthetized rabbit, subsequent to bilateral injections of losartan into the same sites. Note that sarthran induces a large rise in arterial pressure and RSNA even after administration of a high dose of losartan into the same region.

8 R1399 Fig. 6. Chart recordings showing effects of injection of sarile (A) and the neuroinhibitory compound muscimol (B) into the pressor region in the rostral VLM of a spontaneously breathing anesthetized rabbit. Note that muscimol produces a large decrease in both arterial pressure and peak phrenic nerve activity (PNA), but that sarile evokes a decrease in arterial pressure without any change in peak PNA. sure or RSNA (16). It is conceivable, however, that the actions of sarthran and sarile in the VLM are due to blockade of AT 4 receptors, which are activated by ANG IV [ANG-(3O8)] (36). It has been shown, for example, that intracerebroventricular injection of an analog of ANG IV induces a pressor response in rats (35). However, this effect appears to be mediated via AT 1 receptors rather than AT 4 receptors (35). Furthermore, although AT 4 receptors have a high density in many different brain regions, the medulla appears to lack such receptors (36). Nevertheless, the possibility that AT 4 receptors may be involved in mediating the actions of sarthran and sarile in the VLM cannot be definitively ruled out at this stage. It could be argued that the depressor effects of sarile and sarthran are due to blockade of receptors that are accessible to endogenous ANG II, but not to exogenous ANG II. For example, it is conceivable that the critical receptors are located within the synaptic cleft and are tonically activated by endogenous ANG II released from presynaptic terminals, whereas exogenously applied ANG II activates other extrajunctional receptors. The possibility that exogenous sarile and sarthran but not ANG II would have access to such synaptic receptors seems highly unlikely, however, because the former compounds have very similar molecular size and structure to those of ANG II. It is also most unlikely that inaccessibility to the critical receptors could explain the lack of effect of losartan or candesartan in the rostral VLM, because these nonpeptide antagonists have a much smaller molecular size and are also more lipophilic than the peptidic antagonists (34). A final possibility regarding the actions of sarthran and sarile is that the effects observed after their mi-

9 R1400 croinjection into the rostral or caudal VLM are due not to the compounds themselves, but to fragments produced by their degradation. For example, it is possible that the sarcosine moiety is cleaved from these compounds. Sarcosine (N-methylglycine) is a competitive inhibitor of the glycine transporter 1, or Gly T-1, which mediates glycine uptake (13). Thus it is conceivable that sarcosine derived from sarthran or sarile may facilitate the inhibitory action of glycinergic inputs to VLM neurons. At least in the rostral VLM, however, bilateral injections of the glycine antagonist strychnine have very little or no effect on resting arterial blood pressure (15, 26), indicating that under normal conditions there is very little tonic release of glycine in this region. Thus, although this possibility cannot be ruled out, it seems unlikely that a reduction in the uptake rate of glycine could explain the profound depressor effects produced by sarthran or sarile. The caudal VLM contains sympathoinhibitory neurons that tonically inhibit sympathoexcitatory neurons in the rostral VLM (10, 14). Thus the increase in arterial pressure and sympathetic activity evoked by sarthran in the caudal VLM suggests that this compound inhibits the sympathoinhibitory neurons in this region. As in the rostral VLM, this effect was completely unaffected by blockade of AT 1 and ANG-(1O7) receptors, indicating that the mechanism of action of these compounds in the caudal VLM is similar or identical to that in the rostral VLM. The finding that microinjection of sarile into the rostral VLM reduced arterial pressure but not respiratory activity, whereas injection of muscimol or lignocaine reduced both, suggests that sarile has differential effects on different types of neurons. This finding is consistent with the observation by Chan et al. (9) that microinjection of sarile into the rostral VLM reduced the firing rate of some neurons in this region but had no effect on others. It is possible that the effects of sarile or sarthran are specific to cardiovascular neurons, but further studies are needed to test this hypothesis. The results show that a unilateral microinjection of sarile or sarthran into the rostral VLM of barodenervated rabbits produced effects that were less than one-half those produced by bilateral microinjections. In contrast, we have previously shown in the same experimental preparation that unilateral microinjection of muscimol into the rostral VLM produced a very large depressor and sympathoinhibitory effect, more than one-half of that produced by bilateral microinjections of muscimol (17). These findings indicate that, compared with muscimol, sarthran and sarile have a less potent inhibitory effect on rostral VLM sympathoexcitatory neurons. Furthermore, it is possible that there is a nonlinear relationship between the degree of inhibition of the activity of rostral VLM sympathoexcitatory neurons and the subsequent effects on arterial pressure and sympathetic vasomotor activity. Thus the degree of inhibition of the activity of rostral VLM sympathoexcitatory neurons produced by a unilateral microinjection of sarthran or sarile may be below the threshold required to produce a major effect on sympathetic vasomotor activity, whereas the inhibitory effect of unilateral microinjection of muscimol is above that threshold. Although the findings of the present study, together with those of previous studies (8, 12), indicate that AT 1 receptors make no significant contribution to the maintenance of the resting tonic activity of sympathoexcitatory neurons in the rostral VLM or of sympathoinhibitory neurons in the caudal VLM, these receptors nevertheless may play an important role under other conditions. For example, DiBona (11) reported that microinjections of candesartan into the rostral VLM of salt-depleted rats produced a depressor response, and a recent study from our laboratory (32) found that the excitation of rostral VLM sympathoexcitatory neurons resulting from disinhibition of neurons in the hypothalamic paraventricular nucleus was abolished by microinjection of losartan or another selective AT 1 receptor antagonist, L Thus it seems likely that AT 1 receptors may play an important role in mediating inputs to sympathoexcitatory neurons in the rostral VLM, and possibly also to sympathoinhibitory neurons in the caudal VLM, that are activated only under certain physiological conditions. Perspectives The finding by Ito and Sved (18) that bilateral injections of peptidic ANG receptor antagonists into the rostral VLM caused a profound fall in resting arterial pressure, similar to that which occurs after blockade of the spinal sympathetic outflow, led to the hypothesis that ANG II receptors in the rostral VLM play a critical role in generating tonic sympathetic vasomotor activity. The present study confirmed this observation in both rats and rabbits and addressed the important question regarding the mechanisms that underlie the actions of these compounds. Our results indicate that the mechanism of action of sarile and sarthran in the VLM does not involve blockade of the tonic excitatory effects of endogenous ANG II, ANG-(1O7), or other compounds whose effects are mediated via AT 1 or ANG-(1O7) receptors, such as ANG III (16). We have also recently found that the depressor and sympathoinhibitory effects evoked by sarile or sarthran in the rostral VLM were unaffected by complete blockade of glutamate and GABA A receptors (33), indicating that the mechanism of action of these compounds also does not involve effects on glutamatergic or GABAergic neurotransmission. Thus there appear to be two remaining possibilities regarding the mechanism of action of sarile and sarthran in the VLM: these compounds 1) may either inhibit the tonic excitatory effects or facilitate the tonic inhibitory effects of an unknown endogenous compound acting on rostral VLM sympathoexcitatory neurons, or 2) are agonists that directly inhibit these neurons via unknown receptors. The results of the present study do not distinguish between these two possibilities. At the same time, our results indicate that peptidic ANG

10 R1401 receptor antagonists may have a specific action on cardiovascular neurons in the VLM. Further studies will be required to elucidate the precise mechanism of action of these compounds on neurons in the rostral and caudal VLM, which play a pivotal role in central cardiovascular regulation (10). We gratefully acknowledge the gift of losartan from Merck & Co., Inc., Whitehouse Station, NJ, and of candesartan from Astra Hassle AB, Molndal, Sweden. This study was supported by the National Health and Medical Research Council of Australia (Grants and ). REFERENCES 1. Aldred GP, Chai SY, Song K, Zhuo J, MacGregor DP, and Mendelsohn FAO. Distribution of angiotensin II receptor subtypes in the rabbit brain. Regul Pept 44: , Allen AM, Chai SY, Clevers J, McKinley MJ, Paxinos G, and Mendelsohn FAO. Localization and characterization of angiotensin II receptor binding and angiotensin converting enzyme in the human medulla oblongata. J Comp Neurol 269: , Allen AM, Chai SY, Sexton PM, Lewis SJ, Verberne AJM, Jarrott B, Louis WJ, Clevers J, McKinley MJ, Paxinos G, and Mendelsohn FAO. Angiotensin II receptors and angiotensin converting enzyme in the medulla oblongata. Hypertension 9, Suppl 3: , Allen AM, Dampney RAL, and Mendelsohn FAO. Angiotensin receptor binding and pressor effects in cat subretrofacial nucleus. Am J Physiol Heart Circ Physiol 255: H1011 H1017, Allen AM, Mendelsohn FAO, Gieroba ZJ, and Blessing WW. Vasopressin release following microinjection of angiotensin II into the caudal ventrolateral medulla in the anaesthetized rabbit. J Neuroendocrinol 2: , Allen AM, Moeller I, Jenkins TA, Zhuo J, Aldred GP, Chai SY, and Mendelsohn FAO. Angiotensin receptors in the nervous system. Brain Res Bull 47: 17 28, Allen AM, Sasaki S, Dampney RAL, Mendelsohn FAO, and Blessing WW. Actions of angiotensin II in the ventrolateral medulla. In: Central Neural Mechanisms in Cardiovascular Regulation, edited by Kunos G and Ciriello J. Basel: Birkhäuser, 1991, p Averill DB, Tsuchihashi T, Khosla MC, and Ferrario CM. Losartan, nonpeptide angiotensin II-type 1 (AT 1 ) receptor antagonist, attenuates pressor and sympathoexcitatory responses evoked by angiotensin II and L-glutamate in rostral ventrolateral medulla. Brain Res 665: , Chan RKW, Chan YS, and Wong TM. Effects of [Sar 1,Ile 8 ]- angiotensin II on rostral ventrolateral medulla neurons and blood pressure in spontaneously hypertensive rats. Neuroscience 63: , Dampney RAL. Functional organization of central pathways regulating the cardiovascular system. Physiol Rev 74: , DiBona GF. Central sympathoexcitatory actions of angiotensin II: role of type 1 angiotensin II receptors. J Am Soc Nephrol 10, Suppl 11: S90 S94, Fontes MAP, Silva LCS, Campagnole-Santos MJ, Khosla MC, Guertzenstein PG, and Santos RAS. Evidence that angiotensin-(1 7) plays a role in the central control of blood pressure at the ventrolateral medulla acting through specific receptors. Brain Res 665: , Guastella J, Brecha N, Wegmann C, Lester HA, and Davidson N. Cloning, expression and localization of a rat brain high-affinity glycine transporter. Proc Natl Acad Sci USA 89: , 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. New York, NY: Oxford, 1990, p Guyenet PG, Darnall RA, and Riley TAS. Rostral ventrolateral medulla and sympathorespiratory integration in rats. Am J Physiol Regulatory Integrative Comp Physiol 259: R1063 R1074, Hirooka Y, Potts PD, and Dampney RAL. Role of angiotensin II receptor subtypes in mediating the sympathoexcitatory effects of exogenous and endogenous angiotensin peptides in the rostral ventrolateral medulla of the rabbit. Brain Res 772: , Horiuchi J and Dampney RAL. Dependence of sympathetic vasomotor tone on bilateral inputs from the rostral ventrolateral medulla in the rabbit role of baroreceptor reflexes. Neurosci Lett 248: , Ito S and Sved AF. Blockade of angiotensin receptors in rat rostral ventrolateral medulla removes excitatory vasomotor tone. Am J Physiol Regulatory Integrative Comp Physiol 270: R1317 R1323, Kiely JM and Gordon FJ. Role of rostral ventrolateral medulla in centrally mediated pressor responses. Am J Physiol Heart Circ Physiol 267: H1549 H1556, Kubo T and Kihara M. N-methyl-D-aspartate receptors mediate tonic vasodepressor control in the caudal ventrolateral medulla of the rat. Brain Res 451: , Lipski J, Kanjhan R, Kruszewska B, Rong WF, and Smith M. Pre-sympathetic neurones in the rostral ventrolateral medulla of the rat: electrophysiology, morphology and relationship to adjacent neuronal groups. Acta Neurobiol Exp 56: , MacGregor DP, Murone C, Song K, Allen AM, Paxinos G, and Mendelsohn FAO. Angiotensin II receptor subtypes in the human central nervous system. Brain Res 675: , Mendelsohn FAO, Allen AM, Clevers J, Denton DA, Tarjan E, and McKinley MJ. Localization of angiotensin II receptor binding in rabbit brain by in vitro autoradiography. J Comp Neurol 270: , Muratani H, Ferrario CM, and Averill DB. Ventrolateral medulla in spontaneously hypertensive rats: role of angiotensin II. Am J Physiol Regulatory Integrative Comp Physiol 264: R388 R395, Ojima M, Inada Y, Shibouta Y, Wada T, Sanada T, Kubo K, and Nishikawa K. Candesartan (CV-11974) dissociates slowly from the angiotensin AT 1 receptor. Eur J Pharmacol 319: , Ross CA, Ruggiero DA, Park DH, Joh TH, Sved AF, Fernandez-Pardal J, Saavedra JM, and Reis DJ. Tonic vasomotor control by the rostral ventrolateral medulla: effect of electrical or chemical stimulation of the area containing C1 adrenaline neurons on arterial pressure, heart rate and plasma catecholamines and vasopressin. J Neurosci 4: , Santos RAS, Campagnole-Santos MJ, Baracho NCV, Fontes MAP, Silva LCS, Neves LAA, Oliveira DR, Caligiorne SM, Rodrigues ARV, Gropen C, Carvalho WS, Silva ACSE, and Khosla MC. Characterization of a new angiotensin antagonist selective for angiotensin-(1 7): evidence that the actions of angiotensin-(1 7) are mediated by specific receptors. Brain Res Bull 35: , Sasaki S and Dampney RAL. Tonic cardiovascular effects of angiotensin II in the ventrolateral medulla. Hypertension 15: , Schambye HT, Hjorth SA, Bergsma DJ, Sathe G, and Schwartz TW. Differentiation between binding sites for angiotensin II and nonpeptide antagonists on the angiotensin II type 1 receptors. Proc Natl Acad Sci USA 91: , Song K, Allen AM, Paxinos G, and Mendelsohn FAO. Mapping of angiotensin II receptor subtype heterogeneity in rat brain. J Comp Neurol 316: , Sun MK and Guyenet PG. Arterial baroreceptor and vagal inputs to sympathoexcitatory neurons in rat medulla. Am J Physiol Regulatory Integrative Comp Physiol 252: R699 R709, Tagawa T and Dampney RAL. AT 1 receptors mediate excitatory inputs to RVLM pressor neurons from hypothalamus. Hypertension 34: , 1999.

11 R Tagawa T, Horiuchi J, Potts PD, and Dampney RAL. Sympathoinhibition after angiotensin receptor blockade in the rostral ventrolateral medulla is independent of glutamate and GABA receptors. J Auton Nerv Syst 77: 21 30, Timmermans PB, Wong PC, Chiu AT, Herblin WF, Benfield P, Carini DJ, Lee RJ, Wexler RR, Saye JA, and Smith RD. Angiotensin II receptors and angiotensin II receptor antagonists. Pharmacol Rev 45: , Wright JW, Bechtholt AJ, Chambers SL, and Harding JW. Angiotensin III and IV activation of the brain AT 1 receptor subtype in cardiovascular function. Peptides 17: , Wright JW and Harding JW. Important roles for angiotensin III and IV in the brain renin-angiotensin system. Brain Res Rev 25: , Wright JW, Jensen LL, Roberts KA, Sardinia MF, and Harding JW. Structure-function analyses of brain angiotensin control of pressor action in rats. Am J Physiol Regulatory Integrative Comp Physiol 257: R1551 R1557, Yang SN, Lippoldt A, Jansson A, Phillips MI, Ganten D, and Fuxe K. Localization of angiotensin II AT 1 receptor-like immunoreactivity in catecholaminergic neurons of the rat medulla oblongata. Neuroscience 81: , 1997.