Brain Research Bulletin, Vol. 24, pp. 275-280.
0361-9230/90 $3.00 + .OO
0 Pergamon Press plc, 1990. Printed in the U.S.A.
Angiotensin II and Angiotensin (l-7) Excite Neurons in the Canine Medulla In Vitro KAREN L. BARNES.’ W. DOUGLAS KNOWLES AND CARLOS M. FERRARIO Section of Neuroscience, Department of Brain and Vascular Research The Research Institute of the Cleveland Clinic Foundation, Cleveland, OH 44195 Received 6 September 1989
BARNES, K. L., W. D. KNOWLES AND C. M. FERRARIO. Angiotensin II and angiotensin (I-7) excite neurons in the canine medulla in vitro. BRAIN RES BULL 24(2) 275-280, 1990.-Our group showed previously that the heptapeptide angiotensin (l-7) [Ang-( l-7)] is a bioactive product of the renin-angiotensin system, and produces dose-dependent cardiovascular effects similar to those evoked by Ang II when microinjected into the nucleus tractus solitarii (nTS) of the rat. The effects of Ang II were compared with those of Ang-( l-7) on single neuron activity recorded from the medial nTS or dorsal motor nucleus of the vagus (dmnX) in perifused horizontal slices of the canine dorsomedial medulla. Ang II excited 48% of 31 medial nTS neurons, but only activated 14% of 22 dmnX cells. Ang-(l-7) also excited half of the medial nTS cells and 14% of the dmnX neurons. Although most medial nTS neurons excited by Ang II were also activated by Ang-(l-7), two cells were excited by Ang II but not by Ang-(l-7). and one cell was excited by Ang-(l-7) but not by Ang II. Because Ang-(l-7) lacks direct vasoconstrictor effects, neurons in the dorsomedial medulla may have different receptor characteristics than peripheral tissues. The observation of a few medial nTS neurons excited by only one Ang peptide suggests that there may be a separate Ang-(l-7) receptor that participates in the physiological effects of Ang peptides mediated by the brain. In vitro brain slice
Angiotensin (l-7) Neuronal discharges Angiotensin II Dorsal motor nucleus of the vagus Nucleus tractus solitarii
Canine dorsomedial
medulla
II is the biologically active angiotensin peptide is based on analyses of structure-activity relationships performed on peripheral tissues (18.23) and findings suggesting the absolute requirement for the presence of an amino acid in the eighth position (carboxyl terminal) of the angiotensin molecule. However, recent studies from this group have challenged this view. Following the discovery that the amino terminal heptapeptide angiotensin (l-7) [Ang-( l-7)] was a major product of the incubation of ‘2SI-labelled angiotensin I with homogenates of the dorsomedial medulla, Campagnole-Santos et al. (5) showed that microinjection of Ang-(l-7) into the nTS of the rat produced dose-dependent depressor effects similar to those obtained with Ang II. The potential significance of these observations for the neuronal actions of angiotensin peptides prompted us to compare the actions of Ang II with those of Ang-( l-7) on the activity of single neurons recorded from the medial nTS or dmnX of in vitro slices of the canine dorsomedial medulla. This study extends our previous report (3) of neuronal excitation produced by Ang II in the same in
THE major structures of the dorsomedial medulla oblongata, the nucleus tractus solitarii (nTS), dorsal motor nucleus of the vagus (dmnX) and area postrema, contain the neurons that are the primary targets of the afferent fibers of the glossopharyngeal and vagus nerves. This region is also a primary site of the vasomotor actions of angiotensin II (Ang II). In addition to the pressor actions of small amounts of circulating Ang II mediated via the area postmma (13, 15, 21), microinjection of very low doses of the peptide into the parenchyma of the nucleus tractus solitarii (nTS) or dorsal motor nucleus of the vagus (dmnX) can alter both arterial blood pressure and heart rate in the dog (1) and rat (8). Our discovery of a discrete pattern of high affinity Ang II binding sites in the canine dorsomedial medulla that depend on the vagal innervation of this region (9) underscores the observation that Ang II modifies both tonic and baroreflex regulation of arterial pressure and heart rate (4,14). We have also shown (3) that Ang II excites about half of the spontaneously firing neurons recorded from the medial nTS in an in vitro preparation of the canine dorsomedial medulla. To date, investigations of the actions of angiotensin peptides in the dorsomedial medulla have been limited to Ang II and its octapeptide analogues. The commonly held assumption that Ang
vitro preparation. MBTBOD In 23 dogs
‘Requests for reprints should be addressed to Karen L. Barnes, Ph.D., Department Euclid Avenue, Cleveland, OH 44195-5070.
275
(8 to 12 kg male
of Brain and Vascular Research,
mongrels)
anesthetized
with
Cleveland Clinic Foundation,
2% 9500
276
BARNES, KNOWLES A
I 0
&SF
;ng II / 4
8
I 16
12
Time (mid
,,
AND FERRARlO
600,
0
45 Time
(mm)
8’
t
aCSF
) 5 set
,
3 mV
I
1. Time course of the excitation following application of Ang II (8 ng/@) to a neuron in the medial nTS. (A) Instantaneous frequency plot before and after giving aCSF or Ang II (arrows). Ang II produced a substantial increase in firing rate, which peaked after about two minutes, then declined toward baseline over the next 10 minutes. Linked arrows under letters above plot indicate location of extracellular recording segments shown in B and C, below. (B) Discharge pattern before and after application of aCSF (arrow). (C) Discharge pattern at the peak of the excitation produced by Ang II.
FIG. 2. Excitatory responses produced by application of either Ang II (8 ng/pJ) or Ang-(l-7) (80 rig/p..), plotted as spikes per minute. Arrows under trace indicate time of application of aCSF, L-glutamate, Ang II or Ang-(l-7). (A) Excitatory responses to both Ang II and Ang-(l-7) of a medial nTS neuron with infrequent spontaneous discharges. Excitability of this cell was tested with L-glutamate (0.5 FM) prior to applying Ang peptides. (B) Excitation of a neuron in the dmnX by both Ang II and Ang-( 1-7). This cell had an irregular pattern of spontaneous discharges.
silk loops were placed around both vetebral and carotid arteries. The surface of the dorsal medulla oblongata was exposed as described previously (2). After the region was cooled with ice-cold oxygenated artificial cerebrospinal fluid (aCSF), the carotid and vertebral arteries were ligated. The brainstem was transected at the pontomedullary and spinomedullary junctions. The tissue was removed and prepared as horizontal slices for in vitro recording (3). A block containing the dorsomedial medulla was cut under oxygenated aCSF (2°C) with a vibratome (Fredrick Haer OTS 3000, Brunswick, ME) into 300 to 400 pm thick slices oriented in the horizontal plane, parallel to the tractus solitarius (TS). The slices were placed in the recording chamber of an Oslo type interface slice chamber (custom made at IBM Watson Research Center, Yorktown Heights, NY) and left undisturbed for at least one hour before electrophysiological studies were begun. Electrodes and drug pipettes were positioned under stereomicroscopic visual guidance using precision micromanipulators. Angiotensin peptides and aCSF were applied to the surface of the slice by microdrops ejected from broken micropipettes (tip diameters 40-60 pm). A micromanipulator (Narishige HMT-3, Tokyo, Japan) held three independently moveable pipettes above the slice near the recording electrode. Air pressure was used to form a microdrop (0.2 to 0.5 pl) on the pipette tip. The pipette was
briefly lowered to the surface of the slice to apply the drop onto the region surrounding the recording electrode. The pipettes contained either Ang II (8 ng/p,l in aCSF, pH 7.6), Ang-(l-7) (80 ng/kl in aCSF, pH 7.4), or aCSF. To compensate for occasional linear trends in baseline firing rate, the order of application of Ang II and Ang-( l-7) was reversed for successive neurons recorded during an experiment. Whenever possible, each peptide was given twice to each cell in either counterbalanced or alternating order. In addition, aCSF was given before and after each peptide application. The aCSF solution contained 124 mM NaCl, 5 mM KCl, 1.25 mM Na,HPO,, 2 mM MgSO,, 2 mM CaCl,, 26 mM NaHCO,, and 10 mM D-glucose, and was bubbled with 95% 0,/5% CO, and warmed to 36°C before use. Extracellular recordings were made with pipettes containing 0.9% NaCl saturated with Fast Green dye (resistance 5-15 MR, tip diameters O.80; Z=O.81, p>O.42). In the dmnX the effects of Ang II and Ang-( 1-7 1were the same on all 22 neurons. Both angiotensin peptides excited three neurons, but had no effect on the activity of the other 19 cells. When the baseline activities of responsive and nonresponsive neurons were compared, no clear differences were seen. The three dmnX cells that were excited by the angiotensin peptides had spontaneous firing rates of 28, 96 and 376 spikes per minute. while the median firing rate of the neurons which did not respond to the peptides was 79 spikes per minute, with a range of 0 to 384 spikes per minute. The median 2-sample test (17) indicated no differences in baseline firing between responsive and unresponsive cells (Z = 0.61. p>O.54). Figure 2A shows the time course of the firing rate plotted as spikes per minute for a medial nTS neuron that was excited by both Ang II and Ang-( l-7). Because the spontaneous firing rate of this cell was less than 10 spikes per minute, its excitability was verified with L-glutamate prior to administration of the angiotensin peptides. Two successive applications of Ang II generated peak firing rates of more than 250 spikes per minute. A similar excitation was produced by a microdrop of Ang-( l-7). Figure 2B is a plot of the activity of a neuron in the dmnX that was excited by both Ang II and Ang-( l-7). First. a microdrop of Ang II produced a substantial increase in firing rate. Two successive applications of Ang-( l-7) produced similar excitations, and Ang II again activated the cell. Figure 3 documents the firing pattern of one of the two medial nTS neurons that was excited repeatedly by Ang II, but was not activated by three applications of Ang-(l-7). Because the Ang-( 1-7) solution that failed to activate this neuron excited two other cells during this experiment, the Ang-(l-7) peptide was documented to be active. Figure 4 shows an example of a cell in the dmnX that was not excited by either angiotensin peptide, but was activated by L-glutamate at the end of the recording period. DISCUSSION
This study provides direct evidence for an excitatory neuronal action of Ang-(1-7) in subnuclear regions of the dorsomedial medulla that contain specific, high affinity Ang II binding sites. Furthermore, the neuronal effects of Ang-( l-7) were comparable to those produced by application of Ang II to neurons in the medial nTS and dmnX. Although in this study we did not assess the specificity of the neuronal excitation generated by either angiotensin peptide, we re rted previously (3) that the angiotensin antagonist [Sar’ ,Thr !P] Ang II blocked neuronal activation by Ang
KNOWLES AND FERRARIO
I1 in the medial nTS. The present observations are in agreement with our previous report (3) that Ang II excited about half of the spontaneously active neurons recorded in the medial nTS. Combining our previous findings (3) with those of the present study. we have now shown that the medial nTS contains a higher proportion of Ang II-sensitive neurons than does the dmnX because the ratio of responsive to nonresponsive neurons was 0.5 1 in the medial nTS, compared to 0.14 in the dmnX. Histological analysis of the position of the neurons recorded from the dmnX revealed that they were located within the dorsal portion of the nucleus. This finding is in agreement with the observation that the dorsal aspect of the canine dmnX contains a relatively low density of Ang II binding sites. in contrast to the high density of Ang II binding in the medial nTS and ventral dmnX (9). Thus the difference in the proportion of neurons that were responsive to Ang II in these two areas of the dorsomedial medulla may be explained in part by the differences in the density of Ang II binding sites in these subnuclear regions. Comparison of the mean and median baseline firing rates of neurons activated by Ang II with these measures for cells not responsive to the peptide did not suggest that differences in the level of spontaneous activity accounted for responsiveness to the peptide. Although in most of the cells examined in the present study the neuronal actions of Ang II and of Ang- 1( I-7) were qualitatively similar, we do not know whether these angiotensin peptides are equipotent because a 1O-fold higher concentration of Ang-( I-7) was used. In initial experiments comparing the neuronal effects of Ang II and Ang-( l-7). both peptides were applied at a concentration of 8 ng/kl. However, Ang-( l-7) produced peak excitations that were considerably smaller than those generated by Ang II at the same dose. In addition, our group (5) reported that a nanogram dose of Ang-(1-7) microinjected into the medial nTS of the rat evoked significantly smaller depressor responses than the same amount of Ang II. Thus in the present study we used Ang-( l-7) at a IO-fold greater concentration than Ang II, and obtained similar peak excitations with both peptides (see Fig. 2). Furthermore, the microdrop technique does not allow precise quantitation of the amount of peptide delivered to the cell, since the concentration is diluted while diffusing through the tissue to the recorded neuron. Nevertheless, these data are in agreement with recent studies by our group (5) showing that the monophasic depressor and bradycardic effects of Ang II and Ang-( l-7) microinjected into the nTS or dmnX of the rat were generally comparable. The present study could not determine whether the receptors that mediate the neuronal excitations produced by either Ang II or Ang-( l-7) are actually located on the recorded neuron. The neuronal excitation following application of either angiotensin peptide may have been produced by transmitter release from presynaptic terminals containing angiotensin receptors. The finding that Ang-( l-7) excited most of the neurons that responded to Ang II may suggest that both peptides act on a common receptor. In further support of this possibility, the cardiovascular effects of both Ang II and Ang-(l-7) microinjected into the nTS were blocked by prior administration of the Ang II antagonist [Sar’, Thr”] Ang II into the nTS (Campagnole-Santos ef al.. personal communication). Alternatively. there may be a separate receptor for each of these angiotensin peptides. The observation that three neurons in the medial nTS were activated by only one of these angiotensin peptides is consistent with this possibility. Recent studies by our group (10) have used in vitro receptor autoradiography to uncover the presence of binding sites for Ang II and Ang-(1-7) with similar affinities (~50 nM) in the rostra1 medial nTS and dmnX of the dog. A lower affinity (= 1 km) binding site for Ang-(l-7) was also found in these regions. Other subnuclei of the dorsomedial medulla contained only the low affinity Ang( I-7) binding site and high affinity (-= 1 nM) Ang II binding. This
NEURONAL
219
EFFECTS OF ANG II AND ANG-( l-7)
microregional heterogeneity of angiotensin receptors in the dorsomedial medulla may explain the present observation of some medial nTS neurons that responded to only one of the angiotensin peptides, in contrast to the majority of the cells that were excited by both Ang II and Ang-(l-7). Thus there may be different subgroups of neurons that mediate the various physiological effects of angiotensin peptides. Although the present study is the first report of direct neuronal actions of Ang-(l-7), we have shown previously (25) that this amino terminal heptapeptide fragment of Ang II is generated directly from radiolabeled Ang I in homogenates of the canine dorsomedial medulla. The endogenous presence of Ang-( l-7) as a major component of the profile of Ang peptides in the rat medulla has been documented by radioimmunoassay following HPLC separation (7). In situ hybridization has revealed the presence of mRNA for angiotensinogen in the dorsomedial medulla (19). Our group has also reported that Ang-( l-7) is equipotent to Ang II in stimulating the release of vasopressin from explants of the hypothalamus and neurohypophysis in the rat (26). The combination of these data led Ferrario et al. (20) to suggest that Ang-(l-7) is a major active angiotensin peptide in the brain. Other investigators have reported activation of neurons by the carboxyterminal heptapeptide Ang-(2-8) [Ang III]. In the feline subfomical organ (12) and the rat paraventricular nucleus (11,16) Ang III was a potent neuronal excitant. Ang III also activated neurons in the in vitro hippocampal slice (22). When given intraventricularly, Ang II and Ang III produced equivalent increases in blood pressure (27). Iontophoretically administered Ang III has been reported to be more potent than Ang II in exciting forebrain neurons (11, 12, 16). Furthermore, these investigators proposed that Ang III rather than Ang II is the centrally active angiotensin peptide. However, because their rationale depends on the blockade of the neuronal actions of Ang II by aminopeptidase
inhibitors that modify other peptide systems (24), other mechanisms may be involved in their findings. To our knowledge, the effects of Ang III on either neuronal activity medulla or cardiovascular function mediated not been reported.
in the dorsomedial by this region have
Our discovery that Ang-( l-7) excites the same neurons that are responsive to Ang II in the dorsomedial medulla raises new questions about which endogenous form of angiotensin may be responsible for the cardiovascular actions of this peptide family. In the rat we have shown that blockade of the actions of endogenous Ang II with [Sar’ ,Thr*] Ang II enhances the cardiac component of the baroreceptor reflex (4). Others have reported that Ang II microinjections attenuate the heart rate component of the baroreflex (6). Although the dorsomedial medulla is clearly a region where Ang II modulates the activity of neurons in pathways that subserve cardiovascular function, we now propose that the fragment Ang-(l-7) may mediate some of these effects. In summary, the present study has established neuronal actions for a member of the angiotensin peptide family that lacks vasoconstrictor activity. Moreover, this investigation has shown the existence of neuronal groups in the medial nTS and dmnX that are likely to mediate the cardiovascular effects of Ang II and Ang(l-7). However, the observation that a few neurons in the medial nTS are responsive to only one of these peptides suggests that there may also be a separate Ang-( l-7) receptor that participates in the physiological actions of the brain renin-angiotensin system.
ACKNOWLEDGEMENTS
Supported in part by the National Institutes of Health-Heart, Lung and Blood Institute Program Project I-R-6835, the Reinberger Foundation (Cleveland OH) and George Storer Foundation (Miami, FL), and the Federal Mogul Corporation (Detroit, MI).
REFERENCES 1 Averill, D. B.; Diz, D. I.; Barnes, K. L.; Ferrario, C. M. Pressor responses of angiotensin II microinjected into the dorsomedial medulla of the dog. Brain Res. 414:294-300; 1987. 2 Barnes, K. L.; Ferrario, C. M.; Conomy, J. P. Comparison of the hemodynamic changes produced by electrical stimulation of the area postrema and nucleus tractus solitarii in the dog. Circ. Res. 45: 136-143; 1979. 3. Barnes, K. L.; Knowles, W. D.; Ferrario, C. M. Neuronal responses to angiotensin II in the in vitro slice from the canine medulla. Hypertension 11:680-684; 1988. 4. Campagnole-Santos, M. J.; Diz, D. I.; Ferrario, C. M. Baroreceptor reflex modulation by angiotensin II at the nucleus tractus solitarii. Hypertension ll(Supp1. 1):167-171; 1988. 5. Campagnole-Santos, M. J.; Diz, D. I.; Santos, R. A. S.; Khosla, M. C.; Brosnihan, K. B.; Ferrario, C. M. Cardiovascular effects of angiotensin-(l-7) injected into the dorsal medulla of rats. Am. J. Physiol. 257:H324-H329; 1989. 6. Casto, R.; Phillips, M. I. Angiotensin II attenuates baroreflexes at nucleus tractus solitarius of rats. Am. J. Physiol. 250:193-198; 1986. 7. Chappell, M. C.; Brosnihan, K. B.; Diz, D. I.; Santos, R. A. S.; Khosla, M. C.; Ferrario, C. M. Angiotensin-(l-7) is an endogenous peptide in the rat brain. Sot. Neurosci. Abstr. 14:21; 1988. 8. Diz, D. I.; Barnes, K. L.; Ferrario, C. M. Hypotensive actions of angiotensin II microinjected into the dorsal motor nucleus of the vagus. J. Hyp-ertens. 2(Suppl. 3):53-56; 1984. 9 Diz, D. I.; Barnes, K. L.; Ferrario, C. M. Contribution of the vagus nerve to angiotensin II binding in the canine medulla. Brain Res. Bull. 17:497-505; 1986. 10 Diz, D. I.; Ferrario, C. M. Angiotensin (Ang) receptors in the rostra1 dorsomedial medulla recognize Ang-( l-7) with an affinity similar to that of Ang II. Hypertension 14:343; 1989. 11. Felix, D.; Harding, J. W. Manipulation of aminopeptidase activities: differential effects on iontophoretically applied angiotensins in rat
brain. J. Hypertens. 4(Suppl. 6):398-l; 1986. 12. Felix, D.; Schlegel, W. Angiotensin receptive neurones in the subfomical organ. Structure-activity relations. Brain Res. 149: 107116; 1978. 13. Ferrario, C. M.; Barnes, K. L.; Block, C. H.; Brosnihan, K. B.; Diz, D. I.; Khosla, M. C.; Santos, R. A. S. Pathways of angiotensin formation and function in the brain. Hypertension lS(Supp1. 1):in press; 1990. 14. Ferrario, C. M.; Gildenberg, P. L.; McCubbin, J. W. Cardiovascular effects of angiotensin mediated by the central nervous system. Circ. Res. 30:257-262; 1972. 15. Ferrario, C. M.; Ueno, Y.; Diz, D. I.; Barnes, K. L. The reninangiotensin system: physiological actions on the central nervous system. In: Birkenhager, W. B.; Reid, J. L., eds. Handbook of hypertension. Amsterdam: ElsevierNorth Holland Biomedical Press B.V.; 1986:431-454. 16. Fink, G. D.; Bruner, C. A.; Mangiapane, M. L. Area oostrema is critical for angiotensin-induced hypertension in rats. Hypertension 9:355-361; 1987. 17. Harding, J. W.; Felix, D. Angiotensin-sensitive neurons in the rat paraventricular nucleus: relative potencies of angiotensin II and angiotensin III. Brain Res. 410:130-134; 1987. 18. Hays, W. L. Statistics for psychologists. New York: Holt, Rinehart and Winston: 1963. 19. Khosla, M. C. Biological effects of angiotensin analogs and antagonists. In: Menard, J., ed. The renin-anaiotensin svstem. vol. 2. London: Gower Medical Publishing Ltd.; i985:2.2&2.29. 20. Lynch, K. R.; Hawelu-Johnson, C. L.; Guyenet, P. G. Localization of brain angiotensinogen mRNA by hybridization histochemistry. Mol. Brain Res. 2:149-158; 1987. 21. Otsuka, A.; Barnes, K. L.; Ferrario, C. M. Contribution of the area postrema to the pressor effects of angiotensin II. Am. J. Physiol. 251:H538-H546; 1986.
280
22. Palovcik, R. A.; Phillips, M. I. A comparison of the effects of angiotensin II and III on hippocampal slice neurons. Sot. Neurosci. Abstr. 12:151; 1986. 23. Peach, M. J.; Levens, N. R. Molecular approaches to the study of angiotensin receptors. Proc. 14th Midwest Conference of Endocrinology and Metabolism. 1980:171-194. 24. Quirk, W. S.; Harding, J. W.; Wright, J. W. Amastatin and bestatin-induced dipsogenicity in the Sprague-Dawley rat. Brain Res. Bull. 19:145-147; 1987. 25. Santos, R. A. S.; Brosnihan, K. B.; Chappell, M. C.; Pesquero, J.;
BARNES,
KNOWLES
AND FERRARIO
Chemicky, C. L.; Greene, L. J.; Ferrario, C. M. Converting enzyme activity and angiotensin metabolism in the dog brainstem. Hypertension ll(Supp1. 1):1-153-I-157; 1988. 26. Schiavone, M. T.; Santos, R. A. S.; Brosnihan, K. B.; Khosla, M. C.; Ferrario, C. M. Release of vasopressin from the rat hypothalamoneurohypophysial system by angiotensin-(l-7) heptapeptide. Proc Natl. Acad. Sci. USA 85:409511098; 1988. 27. Wright, J. W.; Morseth, S. L.; Abhold, R. H.; Harding, J. W. Pressor action and dipsogenicity induced by angiotensin II and III in rats. Am. J. Physiol. 249:R514-R521; 1985.
0361-9230/90 $3.00 + .OO
0 Pergamon Press plc, 1990. Printed in the U.S.A.
Angiotensin II and Angiotensin (l-7) Excite Neurons in the Canine Medulla In Vitro KAREN L. BARNES.’ W. DOUGLAS KNOWLES AND CARLOS M. FERRARIO Section of Neuroscience, Department of Brain and Vascular Research The Research Institute of the Cleveland Clinic Foundation, Cleveland, OH 44195 Received 6 September 1989
BARNES, K. L., W. D. KNOWLES AND C. M. FERRARIO. Angiotensin II and angiotensin (I-7) excite neurons in the canine medulla in vitro. BRAIN RES BULL 24(2) 275-280, 1990.-Our group showed previously that the heptapeptide angiotensin (l-7) [Ang-( l-7)] is a bioactive product of the renin-angiotensin system, and produces dose-dependent cardiovascular effects similar to those evoked by Ang II when microinjected into the nucleus tractus solitarii (nTS) of the rat. The effects of Ang II were compared with those of Ang-( l-7) on single neuron activity recorded from the medial nTS or dorsal motor nucleus of the vagus (dmnX) in perifused horizontal slices of the canine dorsomedial medulla. Ang II excited 48% of 31 medial nTS neurons, but only activated 14% of 22 dmnX cells. Ang-(l-7) also excited half of the medial nTS cells and 14% of the dmnX neurons. Although most medial nTS neurons excited by Ang II were also activated by Ang-(l-7), two cells were excited by Ang II but not by Ang-(l-7). and one cell was excited by Ang-(l-7) but not by Ang II. Because Ang-(l-7) lacks direct vasoconstrictor effects, neurons in the dorsomedial medulla may have different receptor characteristics than peripheral tissues. The observation of a few medial nTS neurons excited by only one Ang peptide suggests that there may be a separate Ang-(l-7) receptor that participates in the physiological effects of Ang peptides mediated by the brain. In vitro brain slice
Angiotensin (l-7) Neuronal discharges Angiotensin II Dorsal motor nucleus of the vagus Nucleus tractus solitarii
Canine dorsomedial
medulla
II is the biologically active angiotensin peptide is based on analyses of structure-activity relationships performed on peripheral tissues (18.23) and findings suggesting the absolute requirement for the presence of an amino acid in the eighth position (carboxyl terminal) of the angiotensin molecule. However, recent studies from this group have challenged this view. Following the discovery that the amino terminal heptapeptide angiotensin (l-7) [Ang-( l-7)] was a major product of the incubation of ‘2SI-labelled angiotensin I with homogenates of the dorsomedial medulla, Campagnole-Santos et al. (5) showed that microinjection of Ang-(l-7) into the nTS of the rat produced dose-dependent depressor effects similar to those obtained with Ang II. The potential significance of these observations for the neuronal actions of angiotensin peptides prompted us to compare the actions of Ang II with those of Ang-( l-7) on the activity of single neurons recorded from the medial nTS or dmnX of in vitro slices of the canine dorsomedial medulla. This study extends our previous report (3) of neuronal excitation produced by Ang II in the same in
THE major structures of the dorsomedial medulla oblongata, the nucleus tractus solitarii (nTS), dorsal motor nucleus of the vagus (dmnX) and area postrema, contain the neurons that are the primary targets of the afferent fibers of the glossopharyngeal and vagus nerves. This region is also a primary site of the vasomotor actions of angiotensin II (Ang II). In addition to the pressor actions of small amounts of circulating Ang II mediated via the area postmma (13, 15, 21), microinjection of very low doses of the peptide into the parenchyma of the nucleus tractus solitarii (nTS) or dorsal motor nucleus of the vagus (dmnX) can alter both arterial blood pressure and heart rate in the dog (1) and rat (8). Our discovery of a discrete pattern of high affinity Ang II binding sites in the canine dorsomedial medulla that depend on the vagal innervation of this region (9) underscores the observation that Ang II modifies both tonic and baroreflex regulation of arterial pressure and heart rate (4,14). We have also shown (3) that Ang II excites about half of the spontaneously firing neurons recorded from the medial nTS in an in vitro preparation of the canine dorsomedial medulla. To date, investigations of the actions of angiotensin peptides in the dorsomedial medulla have been limited to Ang II and its octapeptide analogues. The commonly held assumption that Ang
vitro preparation. MBTBOD In 23 dogs
‘Requests for reprints should be addressed to Karen L. Barnes, Ph.D., Department Euclid Avenue, Cleveland, OH 44195-5070.
275
(8 to 12 kg male
of Brain and Vascular Research,
mongrels)
anesthetized
with
Cleveland Clinic Foundation,
2% 9500
276
BARNES, KNOWLES A
I 0
&SF
;ng II / 4
8
I 16
12
Time (mid
,,
AND FERRARlO
600,
0
45 Time
(mm)
8’
t
aCSF
) 5 set
,
3 mV
I
1. Time course of the excitation following application of Ang II (8 ng/@) to a neuron in the medial nTS. (A) Instantaneous frequency plot before and after giving aCSF or Ang II (arrows). Ang II produced a substantial increase in firing rate, which peaked after about two minutes, then declined toward baseline over the next 10 minutes. Linked arrows under letters above plot indicate location of extracellular recording segments shown in B and C, below. (B) Discharge pattern before and after application of aCSF (arrow). (C) Discharge pattern at the peak of the excitation produced by Ang II.
FIG. 2. Excitatory responses produced by application of either Ang II (8 ng/pJ) or Ang-(l-7) (80 rig/p..), plotted as spikes per minute. Arrows under trace indicate time of application of aCSF, L-glutamate, Ang II or Ang-(l-7). (A) Excitatory responses to both Ang II and Ang-(l-7) of a medial nTS neuron with infrequent spontaneous discharges. Excitability of this cell was tested with L-glutamate (0.5 FM) prior to applying Ang peptides. (B) Excitation of a neuron in the dmnX by both Ang II and Ang-( 1-7). This cell had an irregular pattern of spontaneous discharges.
silk loops were placed around both vetebral and carotid arteries. The surface of the dorsal medulla oblongata was exposed as described previously (2). After the region was cooled with ice-cold oxygenated artificial cerebrospinal fluid (aCSF), the carotid and vertebral arteries were ligated. The brainstem was transected at the pontomedullary and spinomedullary junctions. The tissue was removed and prepared as horizontal slices for in vitro recording (3). A block containing the dorsomedial medulla was cut under oxygenated aCSF (2°C) with a vibratome (Fredrick Haer OTS 3000, Brunswick, ME) into 300 to 400 pm thick slices oriented in the horizontal plane, parallel to the tractus solitarius (TS). The slices were placed in the recording chamber of an Oslo type interface slice chamber (custom made at IBM Watson Research Center, Yorktown Heights, NY) and left undisturbed for at least one hour before electrophysiological studies were begun. Electrodes and drug pipettes were positioned under stereomicroscopic visual guidance using precision micromanipulators. Angiotensin peptides and aCSF were applied to the surface of the slice by microdrops ejected from broken micropipettes (tip diameters 40-60 pm). A micromanipulator (Narishige HMT-3, Tokyo, Japan) held three independently moveable pipettes above the slice near the recording electrode. Air pressure was used to form a microdrop (0.2 to 0.5 pl) on the pipette tip. The pipette was
briefly lowered to the surface of the slice to apply the drop onto the region surrounding the recording electrode. The pipettes contained either Ang II (8 ng/p,l in aCSF, pH 7.6), Ang-(l-7) (80 ng/kl in aCSF, pH 7.4), or aCSF. To compensate for occasional linear trends in baseline firing rate, the order of application of Ang II and Ang-( l-7) was reversed for successive neurons recorded during an experiment. Whenever possible, each peptide was given twice to each cell in either counterbalanced or alternating order. In addition, aCSF was given before and after each peptide application. The aCSF solution contained 124 mM NaCl, 5 mM KCl, 1.25 mM Na,HPO,, 2 mM MgSO,, 2 mM CaCl,, 26 mM NaHCO,, and 10 mM D-glucose, and was bubbled with 95% 0,/5% CO, and warmed to 36°C before use. Extracellular recordings were made with pipettes containing 0.9% NaCl saturated with Fast Green dye (resistance 5-15 MR, tip diameters O.80; Z=O.81, p>O.42). In the dmnX the effects of Ang II and Ang-( 1-7 1were the same on all 22 neurons. Both angiotensin peptides excited three neurons, but had no effect on the activity of the other 19 cells. When the baseline activities of responsive and nonresponsive neurons were compared, no clear differences were seen. The three dmnX cells that were excited by the angiotensin peptides had spontaneous firing rates of 28, 96 and 376 spikes per minute. while the median firing rate of the neurons which did not respond to the peptides was 79 spikes per minute, with a range of 0 to 384 spikes per minute. The median 2-sample test (17) indicated no differences in baseline firing between responsive and unresponsive cells (Z = 0.61. p>O.54). Figure 2A shows the time course of the firing rate plotted as spikes per minute for a medial nTS neuron that was excited by both Ang II and Ang-( l-7). Because the spontaneous firing rate of this cell was less than 10 spikes per minute, its excitability was verified with L-glutamate prior to administration of the angiotensin peptides. Two successive applications of Ang II generated peak firing rates of more than 250 spikes per minute. A similar excitation was produced by a microdrop of Ang-( l-7). Figure 2B is a plot of the activity of a neuron in the dmnX that was excited by both Ang II and Ang-( l-7). First. a microdrop of Ang II produced a substantial increase in firing rate. Two successive applications of Ang-( l-7) produced similar excitations, and Ang II again activated the cell. Figure 3 documents the firing pattern of one of the two medial nTS neurons that was excited repeatedly by Ang II, but was not activated by three applications of Ang-(l-7). Because the Ang-( 1-7) solution that failed to activate this neuron excited two other cells during this experiment, the Ang-(l-7) peptide was documented to be active. Figure 4 shows an example of a cell in the dmnX that was not excited by either angiotensin peptide, but was activated by L-glutamate at the end of the recording period. DISCUSSION
This study provides direct evidence for an excitatory neuronal action of Ang-(1-7) in subnuclear regions of the dorsomedial medulla that contain specific, high affinity Ang II binding sites. Furthermore, the neuronal effects of Ang-( l-7) were comparable to those produced by application of Ang II to neurons in the medial nTS and dmnX. Although in this study we did not assess the specificity of the neuronal excitation generated by either angiotensin peptide, we re rted previously (3) that the angiotensin antagonist [Sar’ ,Thr !P] Ang II blocked neuronal activation by Ang
KNOWLES AND FERRARIO
I1 in the medial nTS. The present observations are in agreement with our previous report (3) that Ang II excited about half of the spontaneously active neurons recorded in the medial nTS. Combining our previous findings (3) with those of the present study. we have now shown that the medial nTS contains a higher proportion of Ang II-sensitive neurons than does the dmnX because the ratio of responsive to nonresponsive neurons was 0.5 1 in the medial nTS, compared to 0.14 in the dmnX. Histological analysis of the position of the neurons recorded from the dmnX revealed that they were located within the dorsal portion of the nucleus. This finding is in agreement with the observation that the dorsal aspect of the canine dmnX contains a relatively low density of Ang II binding sites. in contrast to the high density of Ang II binding in the medial nTS and ventral dmnX (9). Thus the difference in the proportion of neurons that were responsive to Ang II in these two areas of the dorsomedial medulla may be explained in part by the differences in the density of Ang II binding sites in these subnuclear regions. Comparison of the mean and median baseline firing rates of neurons activated by Ang II with these measures for cells not responsive to the peptide did not suggest that differences in the level of spontaneous activity accounted for responsiveness to the peptide. Although in most of the cells examined in the present study the neuronal actions of Ang II and of Ang- 1( I-7) were qualitatively similar, we do not know whether these angiotensin peptides are equipotent because a 1O-fold higher concentration of Ang-( I-7) was used. In initial experiments comparing the neuronal effects of Ang II and Ang-( l-7). both peptides were applied at a concentration of 8 ng/kl. However, Ang-( l-7) produced peak excitations that were considerably smaller than those generated by Ang II at the same dose. In addition, our group (5) reported that a nanogram dose of Ang-(1-7) microinjected into the medial nTS of the rat evoked significantly smaller depressor responses than the same amount of Ang II. Thus in the present study we used Ang-( l-7) at a IO-fold greater concentration than Ang II, and obtained similar peak excitations with both peptides (see Fig. 2). Furthermore, the microdrop technique does not allow precise quantitation of the amount of peptide delivered to the cell, since the concentration is diluted while diffusing through the tissue to the recorded neuron. Nevertheless, these data are in agreement with recent studies by our group (5) showing that the monophasic depressor and bradycardic effects of Ang II and Ang-( l-7) microinjected into the nTS or dmnX of the rat were generally comparable. The present study could not determine whether the receptors that mediate the neuronal excitations produced by either Ang II or Ang-( l-7) are actually located on the recorded neuron. The neuronal excitation following application of either angiotensin peptide may have been produced by transmitter release from presynaptic terminals containing angiotensin receptors. The finding that Ang-( l-7) excited most of the neurons that responded to Ang II may suggest that both peptides act on a common receptor. In further support of this possibility, the cardiovascular effects of both Ang II and Ang-(l-7) microinjected into the nTS were blocked by prior administration of the Ang II antagonist [Sar’, Thr”] Ang II into the nTS (Campagnole-Santos ef al.. personal communication). Alternatively. there may be a separate receptor for each of these angiotensin peptides. The observation that three neurons in the medial nTS were activated by only one of these angiotensin peptides is consistent with this possibility. Recent studies by our group (10) have used in vitro receptor autoradiography to uncover the presence of binding sites for Ang II and Ang-(1-7) with similar affinities (~50 nM) in the rostra1 medial nTS and dmnX of the dog. A lower affinity (= 1 km) binding site for Ang-(l-7) was also found in these regions. Other subnuclei of the dorsomedial medulla contained only the low affinity Ang( I-7) binding site and high affinity (-= 1 nM) Ang II binding. This
NEURONAL
219
EFFECTS OF ANG II AND ANG-( l-7)
microregional heterogeneity of angiotensin receptors in the dorsomedial medulla may explain the present observation of some medial nTS neurons that responded to only one of the angiotensin peptides, in contrast to the majority of the cells that were excited by both Ang II and Ang-(l-7). Thus there may be different subgroups of neurons that mediate the various physiological effects of angiotensin peptides. Although the present study is the first report of direct neuronal actions of Ang-(l-7), we have shown previously (25) that this amino terminal heptapeptide fragment of Ang II is generated directly from radiolabeled Ang I in homogenates of the canine dorsomedial medulla. The endogenous presence of Ang-( l-7) as a major component of the profile of Ang peptides in the rat medulla has been documented by radioimmunoassay following HPLC separation (7). In situ hybridization has revealed the presence of mRNA for angiotensinogen in the dorsomedial medulla (19). Our group has also reported that Ang-( l-7) is equipotent to Ang II in stimulating the release of vasopressin from explants of the hypothalamus and neurohypophysis in the rat (26). The combination of these data led Ferrario et al. (20) to suggest that Ang-(l-7) is a major active angiotensin peptide in the brain. Other investigators have reported activation of neurons by the carboxyterminal heptapeptide Ang-(2-8) [Ang III]. In the feline subfomical organ (12) and the rat paraventricular nucleus (11,16) Ang III was a potent neuronal excitant. Ang III also activated neurons in the in vitro hippocampal slice (22). When given intraventricularly, Ang II and Ang III produced equivalent increases in blood pressure (27). Iontophoretically administered Ang III has been reported to be more potent than Ang II in exciting forebrain neurons (11, 12, 16). Furthermore, these investigators proposed that Ang III rather than Ang II is the centrally active angiotensin peptide. However, because their rationale depends on the blockade of the neuronal actions of Ang II by aminopeptidase
inhibitors that modify other peptide systems (24), other mechanisms may be involved in their findings. To our knowledge, the effects of Ang III on either neuronal activity medulla or cardiovascular function mediated not been reported.
in the dorsomedial by this region have
Our discovery that Ang-( l-7) excites the same neurons that are responsive to Ang II in the dorsomedial medulla raises new questions about which endogenous form of angiotensin may be responsible for the cardiovascular actions of this peptide family. In the rat we have shown that blockade of the actions of endogenous Ang II with [Sar’ ,Thr*] Ang II enhances the cardiac component of the baroreceptor reflex (4). Others have reported that Ang II microinjections attenuate the heart rate component of the baroreflex (6). Although the dorsomedial medulla is clearly a region where Ang II modulates the activity of neurons in pathways that subserve cardiovascular function, we now propose that the fragment Ang-(l-7) may mediate some of these effects. In summary, the present study has established neuronal actions for a member of the angiotensin peptide family that lacks vasoconstrictor activity. Moreover, this investigation has shown the existence of neuronal groups in the medial nTS and dmnX that are likely to mediate the cardiovascular effects of Ang II and Ang(l-7). However, the observation that a few neurons in the medial nTS are responsive to only one of these peptides suggests that there may also be a separate Ang-( l-7) receptor that participates in the physiological actions of the brain renin-angiotensin system.
ACKNOWLEDGEMENTS
Supported in part by the National Institutes of Health-Heart, Lung and Blood Institute Program Project I-R-6835, the Reinberger Foundation (Cleveland OH) and George Storer Foundation (Miami, FL), and the Federal Mogul Corporation (Detroit, MI).
REFERENCES 1 Averill, D. B.; Diz, D. I.; Barnes, K. L.; Ferrario, C. M. Pressor responses of angiotensin II microinjected into the dorsomedial medulla of the dog. Brain Res. 414:294-300; 1987. 2 Barnes, K. L.; Ferrario, C. M.; Conomy, J. P. Comparison of the hemodynamic changes produced by electrical stimulation of the area postrema and nucleus tractus solitarii in the dog. Circ. Res. 45: 136-143; 1979. 3. Barnes, K. L.; Knowles, W. D.; Ferrario, C. M. Neuronal responses to angiotensin II in the in vitro slice from the canine medulla. Hypertension 11:680-684; 1988. 4. Campagnole-Santos, M. J.; Diz, D. I.; Ferrario, C. M. Baroreceptor reflex modulation by angiotensin II at the nucleus tractus solitarii. Hypertension ll(Supp1. 1):167-171; 1988. 5. Campagnole-Santos, M. J.; Diz, D. I.; Santos, R. A. S.; Khosla, M. C.; Brosnihan, K. B.; Ferrario, C. M. Cardiovascular effects of angiotensin-(l-7) injected into the dorsal medulla of rats. Am. J. Physiol. 257:H324-H329; 1989. 6. Casto, R.; Phillips, M. I. Angiotensin II attenuates baroreflexes at nucleus tractus solitarius of rats. Am. J. Physiol. 250:193-198; 1986. 7. Chappell, M. C.; Brosnihan, K. B.; Diz, D. I.; Santos, R. A. S.; Khosla, M. C.; Ferrario, C. M. Angiotensin-(l-7) is an endogenous peptide in the rat brain. Sot. Neurosci. Abstr. 14:21; 1988. 8. Diz, D. I.; Barnes, K. L.; Ferrario, C. M. Hypotensive actions of angiotensin II microinjected into the dorsal motor nucleus of the vagus. J. Hyp-ertens. 2(Suppl. 3):53-56; 1984. 9 Diz, D. I.; Barnes, K. L.; Ferrario, C. M. Contribution of the vagus nerve to angiotensin II binding in the canine medulla. Brain Res. Bull. 17:497-505; 1986. 10 Diz, D. I.; Ferrario, C. M. Angiotensin (Ang) receptors in the rostra1 dorsomedial medulla recognize Ang-( l-7) with an affinity similar to that of Ang II. Hypertension 14:343; 1989. 11. Felix, D.; Harding, J. W. Manipulation of aminopeptidase activities: differential effects on iontophoretically applied angiotensins in rat
brain. J. Hypertens. 4(Suppl. 6):398-l; 1986. 12. Felix, D.; Schlegel, W. Angiotensin receptive neurones in the subfomical organ. Structure-activity relations. Brain Res. 149: 107116; 1978. 13. Ferrario, C. M.; Barnes, K. L.; Block, C. H.; Brosnihan, K. B.; Diz, D. I.; Khosla, M. C.; Santos, R. A. S. Pathways of angiotensin formation and function in the brain. Hypertension lS(Supp1. 1):in press; 1990. 14. Ferrario, C. M.; Gildenberg, P. L.; McCubbin, J. W. Cardiovascular effects of angiotensin mediated by the central nervous system. Circ. Res. 30:257-262; 1972. 15. Ferrario, C. M.; Ueno, Y.; Diz, D. I.; Barnes, K. L. The reninangiotensin system: physiological actions on the central nervous system. In: Birkenhager, W. B.; Reid, J. L., eds. Handbook of hypertension. Amsterdam: ElsevierNorth Holland Biomedical Press B.V.; 1986:431-454. 16. Fink, G. D.; Bruner, C. A.; Mangiapane, M. L. Area oostrema is critical for angiotensin-induced hypertension in rats. Hypertension 9:355-361; 1987. 17. Harding, J. W.; Felix, D. Angiotensin-sensitive neurons in the rat paraventricular nucleus: relative potencies of angiotensin II and angiotensin III. Brain Res. 410:130-134; 1987. 18. Hays, W. L. Statistics for psychologists. New York: Holt, Rinehart and Winston: 1963. 19. Khosla, M. C. Biological effects of angiotensin analogs and antagonists. In: Menard, J., ed. The renin-anaiotensin svstem. vol. 2. London: Gower Medical Publishing Ltd.; i985:2.2&2.29. 20. Lynch, K. R.; Hawelu-Johnson, C. L.; Guyenet, P. G. Localization of brain angiotensinogen mRNA by hybridization histochemistry. Mol. Brain Res. 2:149-158; 1987. 21. Otsuka, A.; Barnes, K. L.; Ferrario, C. M. Contribution of the area postrema to the pressor effects of angiotensin II. Am. J. Physiol. 251:H538-H546; 1986.
280
22. Palovcik, R. A.; Phillips, M. I. A comparison of the effects of angiotensin II and III on hippocampal slice neurons. Sot. Neurosci. Abstr. 12:151; 1986. 23. Peach, M. J.; Levens, N. R. Molecular approaches to the study of angiotensin receptors. Proc. 14th Midwest Conference of Endocrinology and Metabolism. 1980:171-194. 24. Quirk, W. S.; Harding, J. W.; Wright, J. W. Amastatin and bestatin-induced dipsogenicity in the Sprague-Dawley rat. Brain Res. Bull. 19:145-147; 1987. 25. Santos, R. A. S.; Brosnihan, K. B.; Chappell, M. C.; Pesquero, J.;
BARNES,
KNOWLES
AND FERRARIO
Chemicky, C. L.; Greene, L. J.; Ferrario, C. M. Converting enzyme activity and angiotensin metabolism in the dog brainstem. Hypertension ll(Supp1. 1):1-153-I-157; 1988. 26. Schiavone, M. T.; Santos, R. A. S.; Brosnihan, K. B.; Khosla, M. C.; Ferrario, C. M. Release of vasopressin from the rat hypothalamoneurohypophysial system by angiotensin-(l-7) heptapeptide. Proc Natl. Acad. Sci. USA 85:409511098; 1988. 27. Wright, J. W.; Morseth, S. L.; Abhold, R. H.; Harding, J. W. Pressor action and dipsogenicity induced by angiotensin II and III in rats. Am. J. Physiol. 249:R514-R521; 1985.
