Journal of the Autonomic Nervous System, 38 (1992) 159-166
159
© 1992 Elsevier Science Publishers B.V. All rights reserved 0165-1838/92/$05.00 JANS 01261
Short Communication
Interaction of putative neurotransmitters in rostral ventrolateral medullary cardiovascular neurons S.K. A g a r w a l a n d F.R. C a l a r e s u Department of Physiology, University of Western Ontario, London, Ontario, Canada (Received 29 September 1991) (Revision received 13 January 1992) (Accepted 21 January 1992)
Key words: Glutamate; Acetylcholine; Substance-P; Rostral ventrolateral medulla; Cardiovascular neuron Abstract As recent immunohistochemical evidence has shown the coexistence of putative neurotransmitters in the rostral ventrolateral medulla (RVLM), we have investigated the possibility that there may be an interaction of putative transmitters on the firing frequency of cardiovascular neurons in the RVLM. Extracellul~r activity was recorded from 37 spontaneously firing units in the right RVLM of urethane anaesthetized and artificially ventilated rats. Nine of these units were classified as cardiovascular neurons because: (i) they were silenced by baroreceptor activation (1-3 /xg phenylephrine i.v.); and (ii) they showed rhythmicity of their spontaneous activity in synchrony with the cardiac cycle. Microiontophoresis of combinations of near threshold amounts of L-glutamate (GLU; 10 nA), acetylcholine (Ach; 30 nA) and substance-P (SP; 60 nA) showed a synergistic interaction of these substances with one another in eliciting changes in firing frequency of cardiovascular neurons. These results show that GLU and Ach, GLU and SP and Ach and SP interact synergistically to influence the firing frequency of cardiovascular neurons in the RVLM and suggest that these substances play a physiological role in the neural control of the circulation.
The peripheral and central nervous systems store and release large numbers of substances which may act as transmitters. Many of these substances are colocalized and coreleased by the same nerve terminal [11]. In many cases a classical transmitter (e.g. acetylcholine or amino acid) is located in the same nerve terminal with one or more neuropeptides [3]. Despite the growing body of information (primarily from immunohistochemical studies) demonstrating the colocaliza-
Correspondence: S.K. Agarwal, Playfair Neuroscience Unit, Toronto Hospital, Mc 11-417, Toronto, Ontario, Canada, M5T 2S8.
tion of a variety of putative neurotransmitters, there is very little information regarding functional interactions between these colocalized substances [2,5]. When cotransmission does occur, it often involves synergistic actions of the cotransmitters, each transmitter having a facilitatory effect on the action of the other transmitter on the postjunctional neuron [6]. The importance of the rostral ventrolateral medulla (RVLM) in the regulation of the cardiovascular system has been amply demonstrated [7,8,21,28]. However, the role of putative neurotransmitters in this area is just beginning to be investigated. Recently it has been demonstrated immunohistochemically that fibers and cell bodies in the RVLM contain cholinergic
160
receptors [8,22] and that microinjection of carbachol or physostigmine (cholinergic agonists) into the RVLM evokes pressor responses [8,25]. In addition, iontophoresis of acetylcholine (Ach) has been reported to excite sympathoexcitatory neurons in the RVLM [23]. Similarly, recent studies demonstrate the presence of substance-P (SP) like immunoreactivity [12,13,16] as well as SP receptors [10] in the RVLM. Microinjection of SP agonists into the RVLM produces dose dependent increases in arterial pressure (AP) and heart rate (HR) [26]. Recent work from our laboratory has demonstrated a functional interaction between putative transmitters and peptides following combined microinjection of putative transmitters into two areas of the central nervous system involved in cardiovascular control: the nucleus tractus solitarius and the intermediolateral cell column [14,15]. The present study was designed to investigate the possibility that a synergistic interaction occurs between classical transmitters and peptides in eliciting changes in firing frequency of cardiovascular neurons in the RVLM of the rat. We therefore recorded extracellular single unit activity during microiontophoresis of GLU, Ach and SP either singly or in combination in the vicinity of RVLM cardiovascular neurons. Experiments were done in 12 adult male Wistar rats (250-325 g, Charles River, Montreal), anaesthetized with urethan (Sigma, St. Louis, MO, 1.4 g/kg, i.p.). The animals were paralyzed (decamethonium bromide, Sigma, 3.3 m g / k g i.v. initially, with 0.35 mg supplements every 15-30 min) and artificially ventilated (1.8 to 2.4 m l / m i n and 85 to 95 strokes/min) with room air using a small animal ventilator (Harvard Apparatus, model 683). Supplemental doses of the anaesthetic were administered when necessary. The femoral artery and vein were cannulated. The arterial cannula was connected to a pressure transducer (Statham P23 Db) which was connected to a Grass polygraph (model 7) for continuous recording of arterial pressure AP. A Grass tachograph (7P4C), triggered by the arterial pressure pulse was used to monitor heart rate (HR). The electrocardiogram (ECG, lead II) was recorded with subcutaneous electrodes. The ve-
nous cannula was used to inject 1-3 # g of phenylephrine (PE, Sigma, St. Louis, 0.1-0.3 ml of a 10 /xg/ml solution) for baroreceptor activation. The animal was placed in a stereotaxic apparatus with the bite bar 20 mm below the interaural line. The medulla was exposed by retracting the dorsal neck muscles, incising the atlanto-occipital membrane, and removing part of the occipital bone and the dura. Rectal temperature was maintained at 37.5 _+ 0.5°C with a thermostatically controlled heating blanket. Action potentials from spontaneously firing units in the RVLM were recorded extracellularly through one barrel of a five barrel glass micropipette assembly containing 2 M NaCI (2-5 MOhm impedance measured at 1 kHz). The electrode was inclined 20 ° with respect to the vertical in the sagittal plane with the tip pointing caudally and was advanced through the cerebellum into the RVLM (stereotaxic coordinates: 3.3-3.7 mm rostral to the obex, 1.6-2.1 mm lateral and 7.7-8.3 mm below the dorsal surface of the brain) by a hydraulic microdrive (Narishige, model M08). Electrical activity was amplified through a preamplifier (Dagan 2400; bandpass 0.3-10 kHz), displayed on an oscilloscope (Tektronix R5103N) and discriminated by a Neurolog NL200 spike trigger. The frequency of firing of single units was recorded on a polygraph along with AP and HR. Digitized unit activity (derived from the spike trigger) along with stimulation markers, the AP wave and ECG were also fed to an IBM-AT computer using a Data Translation board (DT 2801A). Data were analy-o~d by the use of a custom-written program for spike train analysis and Macmillan's ASYSTANT-plus software for averaging ECG and AP signals, together with neural activity. For microiontophoresis, 3 drugcontaining barrels of the micropipette assembly were each connected with silver wire to a microiontophoresis unit (Dagan 6400), and one barrel containing 0.5 M NaCl and Fast green was connected to a current compensation circuit. The following drugs were dissolved in distilled water and adjusted to the required pH using either 0.1 M HCI or 0.1 M NaOH: L-glutamate 0.2 M (Sigma, pH 8.0), acetylcholine hydrochloride 0.2 M (Sigma, pH 5.5), substance-P 3 mM (Sigma,
pH 6.5). Retaining currents (5-10 nA) were applied to the drug containing barrels between ejection periods. Compounds were ejected using positive currents (5-60 nA) except glutamate which was ejected with negative current. The right RVLM was explored for spontaneously active units. Units with a signal-to-noise ratio of less than three were not studied further. After a stable recording from a single unit in the RVLM was obtained, each unit was characterized by means of two tests. First, its barosensitivity was tested by recording the change in discharge rate in response to an increase in mean AP elicited by an i.v. bolus injection of 1-3 /zg phenylephrine (PE) and each unit was also tested for the presence of rhythmicity of spontaneous unit activity in relation to the cardiac cycle. A Schmitt trigger circuit was used to derive standardized pulses coincident with a point halfway up the rising slope of the R wave of the ECG. These pulses were used to construct peri-R-wave histograms of RVLM unit activity. Units which were abruptly and completely silenced by baroreceptor activation and which displayed cardiac cycle synchronous rhythmicity were regarded as cardiovascular neurons. GLU, Ach and SP were iontophoresed on these neurons with a current which elicited a small change in firing frequency. These changes in firing frequency were defined as threshold responses. G L U and Ach, G L U and SP, and Ach and SP were then delivered at the same time with the same current used when each of the substances was delivered alone. For control the same current used for the combined iontophoresis was passed through the current balancing barrel. At the end of each experiment the recording sites in the RVLM were marked with iontophoretic deposits of Fast green (15 /xA of cathodal current for 8 min). Animals were perfused via the femoral artery with 50 ml of PBS followed by 50 ml of a 10% formalin solution in PBS. The brains were removed and stored in formalin for 3 - 4 days. Frozen transverse sections (50 /xm) were cut and stained with thionine. Recording sites were mapped on diagrams of transverse sections of the rat brain from an atlas [19].
HR
(bpm)
6°I_
161
--Lf---
20
AP (mmHg)
2°l O ~-
Unit activity (Spikes/s)
Fig. 1. Responses of a unit in the R V L M to baroreceptor activation induced by i.v. bolus injections of 1 and 3 p,g of phenylephrine delivered at the arrowheads. Note abrupt decrease in firing frequency and gradual return of activity as arterial pressure returns to basal levels. Smaller divisions of time scale are in seconds.
Values given in the text are means _ S.E., unless indicated otherwise. Mean changes in firing frequency to combined microiontophoresis of GLU and Ach, G L U and SP and Ach and SP were compared with the responses of each substance alone by an analysis of variance (ANOVA) followed by Neuman-Keuls test for pairwise comparisons as indicated by the ANOVA. In addition, sums of responses to two substances given alone were compared to the changes in firing frequency of combined iontophoresis of two substances and were tested with a paired t-test. The probability level taken to indicate a significant difference was P < 0.05 for all tests. Spontaneous activity was recorded from 37 RVLM units located ventral to a layer of neurons with large spikes and a respiration related discharge pattern. Of these 37 units only 9 fulfilled the criteria for cardiovascular neurons. Typical responses of a cardiovascular unit to bolus injections of PE are shown in Fig. 1. In peri-R-wave histograms of the spontaneous discharge of each of these neurons, a rhythmicity of their activity in synchrony with the cardiac cycle was apparent at
162
120
',,,'
.o - .
-
•
,,
GIA
/
;.-~ ,....'-f // ")-," ;
,, 7 , . , -
i' ~
_.._,_..~,- 7 +,
-..~
l~ x
159
© 1992 Elsevier Science Publishers B.V. All rights reserved 0165-1838/92/$05.00 JANS 01261
Short Communication
Interaction of putative neurotransmitters in rostral ventrolateral medullary cardiovascular neurons S.K. A g a r w a l a n d F.R. C a l a r e s u Department of Physiology, University of Western Ontario, London, Ontario, Canada (Received 29 September 1991) (Revision received 13 January 1992) (Accepted 21 January 1992)
Key words: Glutamate; Acetylcholine; Substance-P; Rostral ventrolateral medulla; Cardiovascular neuron Abstract As recent immunohistochemical evidence has shown the coexistence of putative neurotransmitters in the rostral ventrolateral medulla (RVLM), we have investigated the possibility that there may be an interaction of putative transmitters on the firing frequency of cardiovascular neurons in the RVLM. Extracellul~r activity was recorded from 37 spontaneously firing units in the right RVLM of urethane anaesthetized and artificially ventilated rats. Nine of these units were classified as cardiovascular neurons because: (i) they were silenced by baroreceptor activation (1-3 /xg phenylephrine i.v.); and (ii) they showed rhythmicity of their spontaneous activity in synchrony with the cardiac cycle. Microiontophoresis of combinations of near threshold amounts of L-glutamate (GLU; 10 nA), acetylcholine (Ach; 30 nA) and substance-P (SP; 60 nA) showed a synergistic interaction of these substances with one another in eliciting changes in firing frequency of cardiovascular neurons. These results show that GLU and Ach, GLU and SP and Ach and SP interact synergistically to influence the firing frequency of cardiovascular neurons in the RVLM and suggest that these substances play a physiological role in the neural control of the circulation.
The peripheral and central nervous systems store and release large numbers of substances which may act as transmitters. Many of these substances are colocalized and coreleased by the same nerve terminal [11]. In many cases a classical transmitter (e.g. acetylcholine or amino acid) is located in the same nerve terminal with one or more neuropeptides [3]. Despite the growing body of information (primarily from immunohistochemical studies) demonstrating the colocaliza-
Correspondence: S.K. Agarwal, Playfair Neuroscience Unit, Toronto Hospital, Mc 11-417, Toronto, Ontario, Canada, M5T 2S8.
tion of a variety of putative neurotransmitters, there is very little information regarding functional interactions between these colocalized substances [2,5]. When cotransmission does occur, it often involves synergistic actions of the cotransmitters, each transmitter having a facilitatory effect on the action of the other transmitter on the postjunctional neuron [6]. The importance of the rostral ventrolateral medulla (RVLM) in the regulation of the cardiovascular system has been amply demonstrated [7,8,21,28]. However, the role of putative neurotransmitters in this area is just beginning to be investigated. Recently it has been demonstrated immunohistochemically that fibers and cell bodies in the RVLM contain cholinergic
160
receptors [8,22] and that microinjection of carbachol or physostigmine (cholinergic agonists) into the RVLM evokes pressor responses [8,25]. In addition, iontophoresis of acetylcholine (Ach) has been reported to excite sympathoexcitatory neurons in the RVLM [23]. Similarly, recent studies demonstrate the presence of substance-P (SP) like immunoreactivity [12,13,16] as well as SP receptors [10] in the RVLM. Microinjection of SP agonists into the RVLM produces dose dependent increases in arterial pressure (AP) and heart rate (HR) [26]. Recent work from our laboratory has demonstrated a functional interaction between putative transmitters and peptides following combined microinjection of putative transmitters into two areas of the central nervous system involved in cardiovascular control: the nucleus tractus solitarius and the intermediolateral cell column [14,15]. The present study was designed to investigate the possibility that a synergistic interaction occurs between classical transmitters and peptides in eliciting changes in firing frequency of cardiovascular neurons in the RVLM of the rat. We therefore recorded extracellular single unit activity during microiontophoresis of GLU, Ach and SP either singly or in combination in the vicinity of RVLM cardiovascular neurons. Experiments were done in 12 adult male Wistar rats (250-325 g, Charles River, Montreal), anaesthetized with urethan (Sigma, St. Louis, MO, 1.4 g/kg, i.p.). The animals were paralyzed (decamethonium bromide, Sigma, 3.3 m g / k g i.v. initially, with 0.35 mg supplements every 15-30 min) and artificially ventilated (1.8 to 2.4 m l / m i n and 85 to 95 strokes/min) with room air using a small animal ventilator (Harvard Apparatus, model 683). Supplemental doses of the anaesthetic were administered when necessary. The femoral artery and vein were cannulated. The arterial cannula was connected to a pressure transducer (Statham P23 Db) which was connected to a Grass polygraph (model 7) for continuous recording of arterial pressure AP. A Grass tachograph (7P4C), triggered by the arterial pressure pulse was used to monitor heart rate (HR). The electrocardiogram (ECG, lead II) was recorded with subcutaneous electrodes. The ve-
nous cannula was used to inject 1-3 # g of phenylephrine (PE, Sigma, St. Louis, 0.1-0.3 ml of a 10 /xg/ml solution) for baroreceptor activation. The animal was placed in a stereotaxic apparatus with the bite bar 20 mm below the interaural line. The medulla was exposed by retracting the dorsal neck muscles, incising the atlanto-occipital membrane, and removing part of the occipital bone and the dura. Rectal temperature was maintained at 37.5 _+ 0.5°C with a thermostatically controlled heating blanket. Action potentials from spontaneously firing units in the RVLM were recorded extracellularly through one barrel of a five barrel glass micropipette assembly containing 2 M NaCI (2-5 MOhm impedance measured at 1 kHz). The electrode was inclined 20 ° with respect to the vertical in the sagittal plane with the tip pointing caudally and was advanced through the cerebellum into the RVLM (stereotaxic coordinates: 3.3-3.7 mm rostral to the obex, 1.6-2.1 mm lateral and 7.7-8.3 mm below the dorsal surface of the brain) by a hydraulic microdrive (Narishige, model M08). Electrical activity was amplified through a preamplifier (Dagan 2400; bandpass 0.3-10 kHz), displayed on an oscilloscope (Tektronix R5103N) and discriminated by a Neurolog NL200 spike trigger. The frequency of firing of single units was recorded on a polygraph along with AP and HR. Digitized unit activity (derived from the spike trigger) along with stimulation markers, the AP wave and ECG were also fed to an IBM-AT computer using a Data Translation board (DT 2801A). Data were analy-o~d by the use of a custom-written program for spike train analysis and Macmillan's ASYSTANT-plus software for averaging ECG and AP signals, together with neural activity. For microiontophoresis, 3 drugcontaining barrels of the micropipette assembly were each connected with silver wire to a microiontophoresis unit (Dagan 6400), and one barrel containing 0.5 M NaCl and Fast green was connected to a current compensation circuit. The following drugs were dissolved in distilled water and adjusted to the required pH using either 0.1 M HCI or 0.1 M NaOH: L-glutamate 0.2 M (Sigma, pH 8.0), acetylcholine hydrochloride 0.2 M (Sigma, pH 5.5), substance-P 3 mM (Sigma,
pH 6.5). Retaining currents (5-10 nA) were applied to the drug containing barrels between ejection periods. Compounds were ejected using positive currents (5-60 nA) except glutamate which was ejected with negative current. The right RVLM was explored for spontaneously active units. Units with a signal-to-noise ratio of less than three were not studied further. After a stable recording from a single unit in the RVLM was obtained, each unit was characterized by means of two tests. First, its barosensitivity was tested by recording the change in discharge rate in response to an increase in mean AP elicited by an i.v. bolus injection of 1-3 /zg phenylephrine (PE) and each unit was also tested for the presence of rhythmicity of spontaneous unit activity in relation to the cardiac cycle. A Schmitt trigger circuit was used to derive standardized pulses coincident with a point halfway up the rising slope of the R wave of the ECG. These pulses were used to construct peri-R-wave histograms of RVLM unit activity. Units which were abruptly and completely silenced by baroreceptor activation and which displayed cardiac cycle synchronous rhythmicity were regarded as cardiovascular neurons. GLU, Ach and SP were iontophoresed on these neurons with a current which elicited a small change in firing frequency. These changes in firing frequency were defined as threshold responses. G L U and Ach, G L U and SP, and Ach and SP were then delivered at the same time with the same current used when each of the substances was delivered alone. For control the same current used for the combined iontophoresis was passed through the current balancing barrel. At the end of each experiment the recording sites in the RVLM were marked with iontophoretic deposits of Fast green (15 /xA of cathodal current for 8 min). Animals were perfused via the femoral artery with 50 ml of PBS followed by 50 ml of a 10% formalin solution in PBS. The brains were removed and stored in formalin for 3 - 4 days. Frozen transverse sections (50 /xm) were cut and stained with thionine. Recording sites were mapped on diagrams of transverse sections of the rat brain from an atlas [19].
HR
(bpm)
6°I_
161
--Lf---
20
AP (mmHg)
2°l O ~-
Unit activity (Spikes/s)
Fig. 1. Responses of a unit in the R V L M to baroreceptor activation induced by i.v. bolus injections of 1 and 3 p,g of phenylephrine delivered at the arrowheads. Note abrupt decrease in firing frequency and gradual return of activity as arterial pressure returns to basal levels. Smaller divisions of time scale are in seconds.
Values given in the text are means _ S.E., unless indicated otherwise. Mean changes in firing frequency to combined microiontophoresis of GLU and Ach, G L U and SP and Ach and SP were compared with the responses of each substance alone by an analysis of variance (ANOVA) followed by Neuman-Keuls test for pairwise comparisons as indicated by the ANOVA. In addition, sums of responses to two substances given alone were compared to the changes in firing frequency of combined iontophoresis of two substances and were tested with a paired t-test. The probability level taken to indicate a significant difference was P < 0.05 for all tests. Spontaneous activity was recorded from 37 RVLM units located ventral to a layer of neurons with large spikes and a respiration related discharge pattern. Of these 37 units only 9 fulfilled the criteria for cardiovascular neurons. Typical responses of a cardiovascular unit to bolus injections of PE are shown in Fig. 1. In peri-R-wave histograms of the spontaneous discharge of each of these neurons, a rhythmicity of their activity in synchrony with the cardiac cycle was apparent at
162
120
',,,'
.o - .
-
•
,,
GIA
/
;.-~ ,....'-f // ")-," ;
,, 7 , . , -
i' ~
_.._,_..~,- 7 +,
-..~
l~ x
