Title:
Intrinsic chemosensitivity of RVLM sympathetic premotor neurons in the in situ arterially perfused preparation of rats.
Authors: Tadachika Koganezawa1and Julian F.R. Paton2 Addresses:
1. Department of Physiology, Division of Biomedical Science, Faculty of Medicine, University of Tsukuba, Tsukuba, Ibaraki, 305-8575, Japan 2. School of Physiology and Pharmacology, Bristol Heart Institute, School of Medical Sciences, University of Bristol, Bristol, BS8 1TD, United Kingdom
Running title: Chemosensitive responses of RVLM sympathetic premotor neurons Keywords:
hypoxia, rostral ventrolateral medulla, sympathetic nervous system
The total number of words (excluding references and figure legends): 5,815 words Corresponding author: Tadachika Koganezawa, PhD Department of Physiology Division of Biomedical Science Faculty of Medicine University of Tsukuba 1-1-1 Tennodai, Tsukuba Ibaraki 305-8575, Japan Phone: +81-(0)29-853-3499 FAX:
+81-(0)29-853-3495
E-mail: [email protected] Subject area: Autonomic Neuroscience This is an Accepted Article that has been peer-reviewed and approved for publication in the Experimental Physiology, but has yet to undergo copy-editing and proof correction. Please cite this article as an Accepted Article; doi: 10.1113/expphysiol.2014.080069.
This article is protected by copyright. All rights reserved.
1
NEW FINDINGS
What is the central question of this study? Brain hypoperfusion is a key factor triggering hypertension through activation of cardiovascular sympathetic vasomotor nerves.
However, mechanisms of detecting brain
hypoperfusion remain unclear. We hypothesized that the sympathetic premotor neurons in the rostral ventrolateral medulla (RVLM) can sense asphyxia and cause sympathoexcitation.
What is the main finding and its importance? Functionally identified RVLM sympathetic premotor neurons were excited by hypoxia but less so by hypercapnia, before and after blockade of synaptic transmission. RVLM sympathetic premotor neurons can act as an important oxygen sensor during brain hypoxia/hypoperfusion, which may be important in maintaining sympathetic nerve discharge to support blood pressure and hence maintain brain perfusion.
This article is protected by copyright. All rights reserved.
2
ABSTRACT
Brainstem hypoperfusion is a major excitant of sympathetic activity triggering hypertension but the exact mechanisms involved remain incompletely understood. A major source of excitatory drive to preganglionic sympathetic neurons originates from the ongoing activity of premotor neurons in the rostral ventrolateral medulla (RVLM sympathetic premotor neurons).
The chemosensitivity profile of physiologically characterized RVLM
sympathetic premotor neurons during hypoxia and hypercapnia remains unclear.
We
examined whether physiologically characterized RVLM sympathetic premotor neurons can sense brainstem ischemia intrinsically.
We addressed this issue in a unique in situ arterially
perfused preparation before and after a complete blockade of fast excitatory and inhibitory synaptic transmission.
During hypercapnic-hypoxia, respiratory modulation of RVLM
sympathetic premotor neurons was lost but tonic firing of most RVLM sympathetic premotor neurons was elevated.
After blockade of fast excitatory and inhibitory synaptic transmission,
RVLM sympathetic premotor neurons continued to fire and exhibited an excitatory firing response to hypoxia but not hypercapnia. This study suggests that RVLM sympathetic premotor neurons can sustain high levels of neuronal discharge when oxygen is scarce.
The
intrinsic ability of RVLM sympathetic premotor neurons to maintain responsivity to brainstem hypoxia is an important mechanism ensuring adequate arterial pressure essential for maintaining cerebral perfusion in face of depressed ventilation and/or high cerebral vascular resistance.
This article is protected by copyright. All rights reserved.
3
INTRODUCTION
Previously, we found that reducing blood flow into the brainstem by acute bilateral occlusion of the vertebral arteries produced a prompt increase in arterial pressure and sympathetic nerve activity (Cates et al., 2011).
This response was originally described by
Cushing (1901) who increased cerebral vascular pressure by elevating intracranial pressure. As had been suggested previously (Cushing, 1901; Rodbard & Stone, 1955; Dickinson & Thomson, 1960), we emphasized the importance of an intra-cranial detection system that was sensitive to blood pressure and/or cerebral blood flow and oxygen delivery (Paton et al., 2009) that would raise sympathetic activity and arterial pressure robustly to ensure adequate brain perfusion. Many premotor neurons for cardiovascular sympathetic preganglionic neurons are located in the rostral ventrolateral medulla (RVLM) and termed RVLM sympathetic premotor neurons (Dampney et al., 1979; Guyenet et al., 1989; Kumada et al., 1990; Dampney, 1994).
In numerous previous studies, the consensus of opinion is that RVLM
sympathetic premotor neurons influencing vasomotor and cardiac sympathetic motor outflows have ongoing activity, which is inhibited by baroreceptors and excited by peripheral chemoreceptors (Kumada et al., 1990) and distinct types of neighboring respiratory neurons (Moraes et al., 2014). Some studies have reported the possibility that RVLM sympathetic premotor neurons have chemosensitivity, especially to hypoxia, in vivo and in vitro (Sun et al., 1992; Sun & Reis, 1994a, b; Mazza et al., 2000; D'Agostino et al., 2009). However, in the in vivo studies, synaptic inputs into RVLM sympathetic premotor neurons were either intact or only partially blocked (Sun et al., 1992; Sun & Reis, 1994b). Therefore, the effects of synaptic This article is protected by copyright. All rights reserved.
4
inputs from other neurons to RVLM sympathetic premotor neurons could not be ignored. On the other hand, in the in vitro studies, dissociated neurons were intermingled with non-cardiovascular neurons from the RVLM (Sun & Reis, 1994a; Mazza et al., 2000; D'Agostino et al., 2009). Therefore, it remains unclear whether the RVLM sympathetic premotor neurons themselves can actually sense hypoxia and/or hypercapnia. In the present study, we hypothesized that a locus for the detection system of brainstem hypoperfusion was the rostral ventrolateral medulla and that RVLM sympathetic premotor neurons will be intrinsically sensitive to ischemia.
Therefore, our aim was to
assess changes of the firing of single sympathetic premotor RVLM sympathetic premotor neurons during hypercapnic-hypoxia and hypoxia alone both before and after blocking fast excitatory and inhibitory synaptic transmission.
This is not possible in vivo as global
blockade of excitatory and inhibitory synaptic transmission would be lethal.
Thus, we have
employed the in situ arterially perfused preparation of rat (Paton, 1996) that allowed long term extracellular recording, repeated and reversible bouts of severe hypercapnic-hypoxia and extreme manipulation of the extracellular milieu without detrimental effect to brainstem viability.
We report that RVLM sympathetic premotor neuronal firing is elevated in
response to ischemia in the presence of fast excitatory and inhibitory synaptic transmission. This unique characteristic makes the RVLM a prime locus for maintaining sympathetic discharge in conditions of cerebral hypoperfusion.
This article is protected by copyright. All rights reserved.
5
MATERIALS AND METHODS
All procedure conformed with the UK Animals (Scientific Procedures) Act 1986 and was approved by our institutional ethical review committee.
In situ arterially perfused preparations Seventy-four experiments were performed in decerebrate, unanesthetized and in situ arterially perfused preparations of Wistar rats (male, 60 - 90 g, University of Bristol colony or Japan SLC, Inc.).
General methods were based on those described previously (Paton,
1996). Briefly, a rat was anesthetized deeply with 5% halothane until there was no sign of withdrawal reflexes after pinching the tail or a forepaw.
The rat was bisected
sub-diaphragmatically and its upper body immersed in ice-chilled mock cerebrospinal fluid. Rats were decerebrated at the pre-collicular level and the fourth ventricle exposed by cerebellectomy.
The lungs were removed.
After transferring to a recording chamber, a
double-lumen perfusion catheter (4 F, Braintree Scientific) was introduced into the left ventricle and advanced into the ascending aorta.
Perfusion solution (see below for
constituents) was at constant flow using a roller pump (14 - 16 ml.min-1; Watson Marlow, 505LA).
Perfusion pressure was recorded from the second lumen of the catheter.
Preparations
were
paralyzed
with
vecuronium
bromide
(4
mg∙l-1,
Organon).
[Arg8]-vasopressin acetate salt (0.8 nM, Sigma) was added into the perfusate to increase peripheral vascular resistance.
The composition of the perfusate was: NaCl 125 mM (Fisher
Scientific); NaHCO3 24 mM (Fisher Scientific); KCl 5 mM (The British Drug House, Ltd.); CaCl2 2.5 mM (Sigma); MgSO4 1.25 mM (The British Drug House); KH2PO4 1.25 mM This article is protected by copyright. All rights reserved.
6
(Sigma); dextrose 10 mM (The British Drug House).
The composition of the low
concentration of Ca2+ and high concentration of Mg2+ containing perfusate was: NaCl 117 mM ; NaHCO3 24 mM; KCl 5 mM; CaCl2 0.2 mM; MgSO4 1.25 mM; KH2PO4 1.25 mM; MgCl2 4 mM; dextrose 10 mM (The British Drug House).
The perfusate volume was 200
ml and contained 1.25% Ficoll®70 (Sigma), which is a synthetic polymer of sucrose used to increase osmotic pressure, and pre-warmed to 31°C.
Under control conditions, the perfusate
was equilibrated with 95% oxygen and 5% carbon dioxide.
ECG & Peripheral Nerve Recording The electrocardiogram was recorded and the R-wave discriminated and used to generate TTL pulses from which heart rate (HR) was derived.
Left phrenic nerve (PN) was
isolated at the level of thorax, cut close to the diaphragm, and its activity recorded via a glass suction electrode.
The left central vagus nerve (VN) at the cervical level and the left
thoracic sympathetic chain (SC) at T8 - T9 were isolated and recorded using glass suction electrodes.
Simultaneous activities of the PN, VN and SC were amplified (10k), filtered
(100-5000 Hz), and integrated (0.1 s time constant).
The inspiratory motor pattern consisted
of an incrementing discharge indicative of an eupneic-like pattern (St -John & Paton, 2003), which was used to gauge the viability of the preparation.
The SC displayed both respiratory
and non-respiratory modulated activity as described before (Pickering et al., 2003)
Recording activity of single RVLM sympathetic premotor neurons Single unit activity of neurons located in the left rostral ventrolateral medulla was recorded by micropipettes filled with 0.5 M sodium acetate and 2% pontamine sky-blue for marking recording sites.
Micropipettes were held in a 3-D micromanipulator and driven
This article is protected by copyright. All rights reserved.
7
into the RVLM in 2 m steps using a custom built stepper motor. microelectrodes was between 8 and 12 MΩ.
The tip resistance of the
Single unit activity was amplified (Axon
Instrument, Axoprobe-1A) with reference to a ground attached to a neck muscle; signals were filtered (100 – 5000 Hz, Digitimer, Neurolog).
The site of unit recording was marked by
iontophoretic deposition of dye (-100 V, 10 min).
Recording procedures were as follows:
on encountering a spontaneously active neuron in the RVLM, we tested baroreceptor sensitivity.
Aortic baroreceptors were stimulated by distension of the aortic arch using a
balloon tipped catheter (2 F, Edwards Lifesciences), which was advanced towards the arch via the descending aorta. The final position of the balloon was optimized to produce a reflex bradycardia when distended.
If the neuron was inhibited by baroreceptor stimulation,
we proceeded to test its reaction to peripheral chemoreceptor stimulation.
Peripheral
chemoreceptors were activated by an intra-aortic injection of NaCN (0.03%, 50 μl, The British Drug House) via a side port of the aortic perfusion cannula. bradycardia and sympathoexcitation recorded in the SC.
This produced reflex
If the unit was excited by this
stimulus, it was deemed a RVLM sympathetic premotor unit.
Finally, we tested for its
bulbospinal projections by using conventional antidromic activation from the spinal cord. Following a cervical laminectomy bipolar cashew-coated tungsten wire electrodes (custom made; tip diameter, less than 10 m) were placed into the dorsolateral funiculi under visual control at the C7 segmental level.
The stimulus site was determined by producing powerful
excitation of the SC activity evoked using 3 pulse stimulation (0.2 ms width, 100 Hz, 1 V). For antidromic activation of medullary neurons, single pulse stimulation and an intensity of 1.5 times threshold for SC activation was employed (typically
Intrinsic chemosensitivity of RVLM sympathetic premotor neurons in the in situ arterially perfused preparation of rats.
Authors: Tadachika Koganezawa1and Julian F.R. Paton2 Addresses:
1. Department of Physiology, Division of Biomedical Science, Faculty of Medicine, University of Tsukuba, Tsukuba, Ibaraki, 305-8575, Japan 2. School of Physiology and Pharmacology, Bristol Heart Institute, School of Medical Sciences, University of Bristol, Bristol, BS8 1TD, United Kingdom
Running title: Chemosensitive responses of RVLM sympathetic premotor neurons Keywords:
hypoxia, rostral ventrolateral medulla, sympathetic nervous system
The total number of words (excluding references and figure legends): 5,815 words Corresponding author: Tadachika Koganezawa, PhD Department of Physiology Division of Biomedical Science Faculty of Medicine University of Tsukuba 1-1-1 Tennodai, Tsukuba Ibaraki 305-8575, Japan Phone: +81-(0)29-853-3499 FAX:
+81-(0)29-853-3495
E-mail: [email protected] Subject area: Autonomic Neuroscience This is an Accepted Article that has been peer-reviewed and approved for publication in the Experimental Physiology, but has yet to undergo copy-editing and proof correction. Please cite this article as an Accepted Article; doi: 10.1113/expphysiol.2014.080069.
This article is protected by copyright. All rights reserved.
1
NEW FINDINGS
What is the central question of this study? Brain hypoperfusion is a key factor triggering hypertension through activation of cardiovascular sympathetic vasomotor nerves.
However, mechanisms of detecting brain
hypoperfusion remain unclear. We hypothesized that the sympathetic premotor neurons in the rostral ventrolateral medulla (RVLM) can sense asphyxia and cause sympathoexcitation.
What is the main finding and its importance? Functionally identified RVLM sympathetic premotor neurons were excited by hypoxia but less so by hypercapnia, before and after blockade of synaptic transmission. RVLM sympathetic premotor neurons can act as an important oxygen sensor during brain hypoxia/hypoperfusion, which may be important in maintaining sympathetic nerve discharge to support blood pressure and hence maintain brain perfusion.
This article is protected by copyright. All rights reserved.
2
ABSTRACT
Brainstem hypoperfusion is a major excitant of sympathetic activity triggering hypertension but the exact mechanisms involved remain incompletely understood. A major source of excitatory drive to preganglionic sympathetic neurons originates from the ongoing activity of premotor neurons in the rostral ventrolateral medulla (RVLM sympathetic premotor neurons).
The chemosensitivity profile of physiologically characterized RVLM
sympathetic premotor neurons during hypoxia and hypercapnia remains unclear.
We
examined whether physiologically characterized RVLM sympathetic premotor neurons can sense brainstem ischemia intrinsically.
We addressed this issue in a unique in situ arterially
perfused preparation before and after a complete blockade of fast excitatory and inhibitory synaptic transmission.
During hypercapnic-hypoxia, respiratory modulation of RVLM
sympathetic premotor neurons was lost but tonic firing of most RVLM sympathetic premotor neurons was elevated.
After blockade of fast excitatory and inhibitory synaptic transmission,
RVLM sympathetic premotor neurons continued to fire and exhibited an excitatory firing response to hypoxia but not hypercapnia. This study suggests that RVLM sympathetic premotor neurons can sustain high levels of neuronal discharge when oxygen is scarce.
The
intrinsic ability of RVLM sympathetic premotor neurons to maintain responsivity to brainstem hypoxia is an important mechanism ensuring adequate arterial pressure essential for maintaining cerebral perfusion in face of depressed ventilation and/or high cerebral vascular resistance.
This article is protected by copyright. All rights reserved.
3
INTRODUCTION
Previously, we found that reducing blood flow into the brainstem by acute bilateral occlusion of the vertebral arteries produced a prompt increase in arterial pressure and sympathetic nerve activity (Cates et al., 2011).
This response was originally described by
Cushing (1901) who increased cerebral vascular pressure by elevating intracranial pressure. As had been suggested previously (Cushing, 1901; Rodbard & Stone, 1955; Dickinson & Thomson, 1960), we emphasized the importance of an intra-cranial detection system that was sensitive to blood pressure and/or cerebral blood flow and oxygen delivery (Paton et al., 2009) that would raise sympathetic activity and arterial pressure robustly to ensure adequate brain perfusion. Many premotor neurons for cardiovascular sympathetic preganglionic neurons are located in the rostral ventrolateral medulla (RVLM) and termed RVLM sympathetic premotor neurons (Dampney et al., 1979; Guyenet et al., 1989; Kumada et al., 1990; Dampney, 1994).
In numerous previous studies, the consensus of opinion is that RVLM
sympathetic premotor neurons influencing vasomotor and cardiac sympathetic motor outflows have ongoing activity, which is inhibited by baroreceptors and excited by peripheral chemoreceptors (Kumada et al., 1990) and distinct types of neighboring respiratory neurons (Moraes et al., 2014). Some studies have reported the possibility that RVLM sympathetic premotor neurons have chemosensitivity, especially to hypoxia, in vivo and in vitro (Sun et al., 1992; Sun & Reis, 1994a, b; Mazza et al., 2000; D'Agostino et al., 2009). However, in the in vivo studies, synaptic inputs into RVLM sympathetic premotor neurons were either intact or only partially blocked (Sun et al., 1992; Sun & Reis, 1994b). Therefore, the effects of synaptic This article is protected by copyright. All rights reserved.
4
inputs from other neurons to RVLM sympathetic premotor neurons could not be ignored. On the other hand, in the in vitro studies, dissociated neurons were intermingled with non-cardiovascular neurons from the RVLM (Sun & Reis, 1994a; Mazza et al., 2000; D'Agostino et al., 2009). Therefore, it remains unclear whether the RVLM sympathetic premotor neurons themselves can actually sense hypoxia and/or hypercapnia. In the present study, we hypothesized that a locus for the detection system of brainstem hypoperfusion was the rostral ventrolateral medulla and that RVLM sympathetic premotor neurons will be intrinsically sensitive to ischemia.
Therefore, our aim was to
assess changes of the firing of single sympathetic premotor RVLM sympathetic premotor neurons during hypercapnic-hypoxia and hypoxia alone both before and after blocking fast excitatory and inhibitory synaptic transmission.
This is not possible in vivo as global
blockade of excitatory and inhibitory synaptic transmission would be lethal.
Thus, we have
employed the in situ arterially perfused preparation of rat (Paton, 1996) that allowed long term extracellular recording, repeated and reversible bouts of severe hypercapnic-hypoxia and extreme manipulation of the extracellular milieu without detrimental effect to brainstem viability.
We report that RVLM sympathetic premotor neuronal firing is elevated in
response to ischemia in the presence of fast excitatory and inhibitory synaptic transmission. This unique characteristic makes the RVLM a prime locus for maintaining sympathetic discharge in conditions of cerebral hypoperfusion.
This article is protected by copyright. All rights reserved.
5
MATERIALS AND METHODS
All procedure conformed with the UK Animals (Scientific Procedures) Act 1986 and was approved by our institutional ethical review committee.
In situ arterially perfused preparations Seventy-four experiments were performed in decerebrate, unanesthetized and in situ arterially perfused preparations of Wistar rats (male, 60 - 90 g, University of Bristol colony or Japan SLC, Inc.).
General methods were based on those described previously (Paton,
1996). Briefly, a rat was anesthetized deeply with 5% halothane until there was no sign of withdrawal reflexes after pinching the tail or a forepaw.
The rat was bisected
sub-diaphragmatically and its upper body immersed in ice-chilled mock cerebrospinal fluid. Rats were decerebrated at the pre-collicular level and the fourth ventricle exposed by cerebellectomy.
The lungs were removed.
After transferring to a recording chamber, a
double-lumen perfusion catheter (4 F, Braintree Scientific) was introduced into the left ventricle and advanced into the ascending aorta.
Perfusion solution (see below for
constituents) was at constant flow using a roller pump (14 - 16 ml.min-1; Watson Marlow, 505LA).
Perfusion pressure was recorded from the second lumen of the catheter.
Preparations
were
paralyzed
with
vecuronium
bromide
(4
mg∙l-1,
Organon).
[Arg8]-vasopressin acetate salt (0.8 nM, Sigma) was added into the perfusate to increase peripheral vascular resistance.
The composition of the perfusate was: NaCl 125 mM (Fisher
Scientific); NaHCO3 24 mM (Fisher Scientific); KCl 5 mM (The British Drug House, Ltd.); CaCl2 2.5 mM (Sigma); MgSO4 1.25 mM (The British Drug House); KH2PO4 1.25 mM This article is protected by copyright. All rights reserved.
6
(Sigma); dextrose 10 mM (The British Drug House).
The composition of the low
concentration of Ca2+ and high concentration of Mg2+ containing perfusate was: NaCl 117 mM ; NaHCO3 24 mM; KCl 5 mM; CaCl2 0.2 mM; MgSO4 1.25 mM; KH2PO4 1.25 mM; MgCl2 4 mM; dextrose 10 mM (The British Drug House).
The perfusate volume was 200
ml and contained 1.25% Ficoll®70 (Sigma), which is a synthetic polymer of sucrose used to increase osmotic pressure, and pre-warmed to 31°C.
Under control conditions, the perfusate
was equilibrated with 95% oxygen and 5% carbon dioxide.
ECG & Peripheral Nerve Recording The electrocardiogram was recorded and the R-wave discriminated and used to generate TTL pulses from which heart rate (HR) was derived.
Left phrenic nerve (PN) was
isolated at the level of thorax, cut close to the diaphragm, and its activity recorded via a glass suction electrode.
The left central vagus nerve (VN) at the cervical level and the left
thoracic sympathetic chain (SC) at T8 - T9 were isolated and recorded using glass suction electrodes.
Simultaneous activities of the PN, VN and SC were amplified (10k), filtered
(100-5000 Hz), and integrated (0.1 s time constant).
The inspiratory motor pattern consisted
of an incrementing discharge indicative of an eupneic-like pattern (St -John & Paton, 2003), which was used to gauge the viability of the preparation.
The SC displayed both respiratory
and non-respiratory modulated activity as described before (Pickering et al., 2003)
Recording activity of single RVLM sympathetic premotor neurons Single unit activity of neurons located in the left rostral ventrolateral medulla was recorded by micropipettes filled with 0.5 M sodium acetate and 2% pontamine sky-blue for marking recording sites.
Micropipettes were held in a 3-D micromanipulator and driven
This article is protected by copyright. All rights reserved.
7
into the RVLM in 2 m steps using a custom built stepper motor. microelectrodes was between 8 and 12 MΩ.
The tip resistance of the
Single unit activity was amplified (Axon
Instrument, Axoprobe-1A) with reference to a ground attached to a neck muscle; signals were filtered (100 – 5000 Hz, Digitimer, Neurolog).
The site of unit recording was marked by
iontophoretic deposition of dye (-100 V, 10 min).
Recording procedures were as follows:
on encountering a spontaneously active neuron in the RVLM, we tested baroreceptor sensitivity.
Aortic baroreceptors were stimulated by distension of the aortic arch using a
balloon tipped catheter (2 F, Edwards Lifesciences), which was advanced towards the arch via the descending aorta. The final position of the balloon was optimized to produce a reflex bradycardia when distended.
If the neuron was inhibited by baroreceptor stimulation,
we proceeded to test its reaction to peripheral chemoreceptor stimulation.
Peripheral
chemoreceptors were activated by an intra-aortic injection of NaCN (0.03%, 50 μl, The British Drug House) via a side port of the aortic perfusion cannula. bradycardia and sympathoexcitation recorded in the SC.
This produced reflex
If the unit was excited by this
stimulus, it was deemed a RVLM sympathetic premotor unit.
Finally, we tested for its
bulbospinal projections by using conventional antidromic activation from the spinal cord. Following a cervical laminectomy bipolar cashew-coated tungsten wire electrodes (custom made; tip diameter, less than 10 m) were placed into the dorsolateral funiculi under visual control at the C7 segmental level.
The stimulus site was determined by producing powerful
excitation of the SC activity evoked using 3 pulse stimulation (0.2 ms width, 100 Hz, 1 V). For antidromic activation of medullary neurons, single pulse stimulation and an intensity of 1.5 times threshold for SC activation was employed (typically
