Role of Noradrenergic Brain Nuclei in the Regulation of Carotid Artery Blood Flow: Pharmacological Evidence from Anesthetized Pigs with Alpha-2 Adrenergic Receptor Modulator Drugs

Abstract

Background:

Cerebral hypoperfusion and degeneration of the noradrenergic locus coeruleus (LC) occur early in Alzheimer’s disease (AD). Cerebral blood vessels are densely innervated by noradrenergic projections from the LC suggesting a functional role for the regulation of cerebral blood flow (CBF). Experimental LC stimulation, however, has provided no clarity as decreases or increases in CBF were reported from different experimental settings and investigators.

Objective:

To find out with pharmacological methods whether endogenously released norepinephrine (NE) increases or decreases carotid artery blood flow (CABF) in anesthetized pigs by investigating the effect of centrally acting alpha-2 adrenergic drugs, which increase (atipamezole) or decrease (xylazine) NE in the brain.

Methods:

CABF was measured by a Doppler-flow probe placed around the left carotid artery in pentobarbital anesthetized young pigs.

Results:

Neither current antihypertensive drugs nor pharmacological stimulation of dopamine, histamine, serotonin or acetylcholine receptors changed CABF. The alpha-2 adrenergic-receptor agonist xylazine decreased, while the antagonist atipamezole raised CABF. This rise was abolished by a combined treatment with endothelial NO-synthase inhibitor Nω-Nitro-L-arginine methyl ester (L-NAME) and the non-selective β-receptor antagonist propranolol. Propranolol alone did not decrease CABF in contrast to L-NAME but decreased CABF after L-NAME, surprisingly.

Conclusion:

Pharmacological evidence suggests that NE released in the brain of anesthetized pigs raises CABF involving β-adrenergic mechanisms and nitric oxide. If in awake humans NE released from the LC had vasodilator effects early LC degeneration could be involved in early cerebral hypoperfusion of AD. Moreover, a cerebral adrenergic vascular innervation deficit, possibly resulting from LC degeneration, and systemic endothelial dysfunction together may act synergistically to reduce CBF.

Keywords

Alpha-2 adrenergic drugs Alzheimer’s disease atipamezole carotid artery blood flow cerebral blood flow cerebral hypoperfusion degeneration locus coeruleus pigs propranolol xylazine

Introduction

The locus coeruleus (LC) is the main source of norepinephrine (NE) in the brain providing this stimulating transmitter throughout the brain via widespread efferent projections [1]. LC degeneration and abnormalities of the noradrenergic system in Alzheimer’s disease (AD) patients had been long observed, confirmed, and more deeply investigated in more recent studies [2 –6]. LC is the first area to develop neurofibrillary changes in AD. Therefore, LC could be the predominant site where AD-related pathology begins. Accumulating evidence suggests that the loss of noradrenergic innervation exacerbates the pathogenesis of AD. Apart from its role as a stimulating neurotransmitter in the regulation of behavior, NE is believed to have a role in providing neuroprotection against oxidative stress, neurotrophic support, and in limiting neuroinflammatory activation [7, 8].

Another change in AD, which appears early and even in the prodromal phase, is cerebral hypoperfusion, as shown in a number of studies in the last two decades [9 –14]. Vascular risk factors are involved in the pathogenesis of AD and cerebrovascular dysfunction is detected prior to the onset of cognitive and histopathological changes [15 –18].

The fact that cerebral hypoperfusion occurs early in the development of AD suggests that it could be causally involved in the pathophysiology by causing a chronic energy deficit that could lead to neurodegeneration over a long time period [19 –21]. Since both LC degeneration and cerebral hypoperfusion occur early in the development of AD, we raised the question whether and how these changes could be related as the LC is clearly connected to vascular innervation in the brain: LC projections innervate the cerebral vasculature throughout the brain with extensive varicosities designed for volume release of NE (non-synaptic transmission) [1, 22]. It is estimated that each LC neuron innervates 20 m of capillaries on average in humans [23]. Noradrenergic fibers from the locus ceruleus were identified in close proximity to brain microvessels for volume transmission of NE, and the existence of adrenoceptors has been shown in isolated brain arterioles and capillaries [22, 24]. A functional relevance of perivascular noradrenergic innervation originating in the LC had been suggested by the demonstration that LC dysfunction induced chemically in animals leads to an upregulation of vascular adrenergic receptors which was referred to as “de-afferentiation hypersensitivity” [22]. The pathophysiological correlate to this experimental situation of induced noradrenergic dysfunction in animals could be the finding that vascular alpha-2 (α2) adrenergic and β2-adrenergic receptors were significantly increased in cerebral vessels of AD patients [24].

According to the mechanism of neurovascular coupling, which denotes an increase in CBF with increasing neuronal activity by metabolic coupling mechanisms, one would assume vasodilator effects and an increase in CBF from a stimulatory brain transmitter. The literature on the effects of experimental LC stimulation on CBF is contradictory, however, showing increases or decreases in CBF. In more recent studies, where the claim of the authors was that the way of stimulation is more physiological, LC activation indeed increased CBF while several earlier investigations showed decreases in CBF [25 –28]. The controversial findings may be explained by the dual role of NE: it exerts vasoconstrictor effects by stimulating α-adrenergic receptors on vascular smooth muscles cells (VSCMs) as well as vasodilator effects by the stimulation of β1-adrenergic receptors on VSCMs and β2- and β3- and α2-adrenergic receptors on endothelial cells (EC) via release of nitric oxide (NO) so that LC stimulation could cause increases or decreases in CBF depending on whether β- or α-adrenergic receptors predominate on the target vessel and also on the intensity of the stimulation potentially leading to different levels of NE concentrations [29].

Given the controversy on the role of NE released from noradrenergic brain nuclei after experimental stimulation for the regulation of CBF – causing increases or decreases in animal studies – the aim of this study was to find out the role of endogenously released NE in the regulation of CBF by using a pharmacological approach, namely by investigating the effect of drugs on carotid artery flow (CABF) in anesthetized pigs that either inhibit or stimulate the release of NE from brain noradrenergic neurons. An important mechanism in noradrenergic innervation is the self-control and modulation of NE-release by presynaptic α2-adrenergic receptors. Their stimulation by endogenous NE or by pharmacological agonists such as xylazine inhibits the release of NE while antagonists at this receptor such as the selective α2-receptor antagonist atipamezole enhance NE-release [30]. α2-adrenergic antagonists have been investigated in animal models of cognition and neuroinflammation and have shown modest beneficial effects (for review, see [7]) but a systematic investigation of their effects on CBF has not been made. In vivo pharmacological CBF studies using appropriate drugs with defined modes of actions might help to elucidate the physiological mechanisms of CBF regulation including the role of different neurotransmitter systems in the regulation of CBF.

For the validation of our approach, we tested numerous drugs for comparison which we think is necessary because a systematic pharmacology of CBF does not exist yet and because experiments with a reasonable throughput may only be possible in anesthetized animals which is a limitation. In this intention, common vasodilator and antihypertensive drugs were investigated including an ACE-inhibitor, an angiotensin II-receptor blocker, an endothelin receptor antagonist and a L-type calcium channel blocker. As NE is a stimulating transmitter, for the purpose of comparison, we have also investigated the effect of several drugs on CABF in deeply pentobarbital anesthetized pigs, which raise concentrations of other stimulatory transmitters or mimic their actions including the dopamine-agonist apomorphine, fluoxetine, a histamine H3-antagonist and the cholinesterase-inhibitor donepezil. The latter increase the concentrations of serotonin, histamine or acetylcholine, respectively, and histamine is an important stimulating and arousal inducing transmitter in the brain.

Methods

All procedures involving experimentation on animal subjects have been performed according Directive 63/2010 of the European Commission implemented into German Animal Welfare Legislation and are registered by the Competent Authority (District Government in Darmstadt, Hesse) under No. HMR-4a/Anz.40.

Pigs of the German Landrace with an age of about 2-3 months weighing 20–35 kg were used. Pigs fed a standard pig diet were fasted for 20 h before the experiment but had free access to water.

Anesthesia

Pigs were premedicated with 2 mL Rompun^® 2% i.m. (xylazine HCl, 23.3 mg/mL) and 1 mL of Zoletil100^® (100 mg/mL; 50 mg/mL Tiletamine and 50 mg/mL Zolazepam) and anesthetized with an i.v.-bolus of 3 mL Narcoren^® (pentobarbital, 160 mg/mL) followed by a continuous intravenous infusion of 12–17 mg/kg/h pentobarbital. In addition to general anesthesia, local anesthesia was performed on all incisions with the long acting local anesthetic bupivacaine (Bupivacain^® 0,5%, Jenapharm). The animals were placed in a recumbent position, tracheotomized, and ventilated with room air and oxygen by a respirator (ABV-Intensiv, Fa. Stephan, D 56412 Gackenbach, Germany). Blood gas analysis (pO₂; pCO₂) was performed to control the oxygen supply via the respirator in order to maintain pO₂ >100 mm Hg and pCO₂ <35–40 mm Hg. Blood gas analysis: before start of infusion and during infusion at time points 30 and 60 min.

Measurement of CABF and blood pressure

To monitor blood pressure, a tip catheter (Millar PC 350) was inserted into the left femoral artery after an incision. CABF was measured via an ultrasonic flow probe (3 mm diameter; transit time flowmeter module and flow probe from HSE Electronics). The carotid flow probe was placed around the common carotid artery after enlarging the incision made for the tracheotomy. The pig carotid artery unlike human carotid artery does not branch into an internal and external carotid artery where the internal branch would feed the brain (reflecting a large portion of cerebral blood flow) but the common carotid artery trunk releases two major branches to extracranial structures on its way along the neck, a major branch to the chewing muscles quite distally and another branch cranially just before the carotid artery penetrates the skull. This narrow anatomical situation precludes the possibility of placing a flow probe around the carotid artery just after the last extracranial branch (which in the pig is then the correlate of the internal carotid artery). These extracranial arterial branches of the common carotid artery in the pig are quite large compared with the human anatomy reflecting the much larger size of head and tongue in the pig. To measure exclusively flow to the brain on the common carotid artery the two carotid branches to extracranial structures mentioned above were ligated (after having studied pig carotid anatomy via vascular casts).

Table 1

Effect of α2-adrenergic receptor modulator drugs on carotid artery blood flow and carotid artery resistance, mean blood pressure (BP) and heart rate in anesthetized pigs in comparison with vasoactive and centrally acting drugs

Drug		Carotid artery flow			Carotid artery resistance			Mean BP			Heart rate
		(ml/min)		% change	(mmHg*min/ml)		% change	(mmHg)		(bpm)
		baseline	drug	p<	baseline	drug	p<	baseline	drug	p<	baseline	drug	p<
Vehicle	Mean	133.5	134.3	0.6%	0.77	0.79	2%	89.4	86.3		93.0	89.7
n = 12	SEM	13.2	13.9		0.1	0.1		7.1	8.7		4.3	4.2
Atipamezol	Mean	123.1	170.8	39%	0.80	0.53	−34%	89.9	83.9		83.7	84.2
n = 13	SEM	7.7	10.6	0.000001	0.07	0.04	0.00001	6.2	5.4	0.001	3.1	3.0
Xylazine	Mean	123.9	77.6	−37%	0.81	1.39	71%	84.1	109.9		79.4	81.2
n = 10	SEM	20.7	7.9	0.01	0.08	0.14	0.0001	4.4	8.8	0.01	3.1	3.2
Cilostazol	Mean	138.6	176.6	27%	0.55	0.45	−19%	72.8	75.9		86.2	93.6
n = 5	SEM	7.4	11.3	0.05	0.04	0.02	0.05	6.1	7.8	0.01	5.3	5.2	0.01
Amlodipine	Mean	127.7	95.7	−25%	0.84	0.85	2%	94.0	75.4		87.1	90.1
n = 7	SEM	16.9	9.8	0.01	0.11	0.12		7.0	6.6	0.001	7.7	6.0
Irbesartan	Mean	142.8	144.0	1%	0.67	0.70	4%	86.3	87.7		90.4	86.6
n = 5	SEM	24.1	26.5		0.08	0.09		7.8	10.1		6.8	7.0
Ramipril	Mean	142.7	144.3	1%	0.67	0.71	6%	87.4	94.1		84.7	82.7
n = 3	SEM	28.3	27.6		0.09	0.11		6.1	3.8		12.7	11.6
Bosentan	Mean	121.3	119.0	−2%	0.69	0.71	2%	74.2	74.2		95.5	90.0
n = 4	SEM	13.9	13.8		0.14	0.15		5.7	5.7		5.2	4.2
Fluoxetine	Mean	115.7	108.7	−6%	0.83	0.97	17%	78.8	91.3		67.0	74.3
n = 3	SEM	35.0	35.2		0.19	0.23		4.4	6.2		9.5	6.4
Apomorphine	Mean	119.2	109.7	−8%	0.77	0.73	−5%	86.3	75.2		82.3	69.5
n = 6	SEM	11.2	8.6		0.05	0.05		6.1	4.7		6.8	5.3
A503	Mean	129.2	131.5	2%	0.71	0.74	3%	84.2	88.8		83.3	79.7
n = 6	SEM	9.4	11.4		0.07	0.05		4.7	4.1		4.1	3.7
Donepezil	Mean	124.3	130.0	5%	0.71	0.75	6%	78.9	83.2		82.8	73.5
n = 4	SEM	17.8	23.1		0.11	0.15		5.3	7.3		9.2	7.1

When stabile hemodynamic conditions and blood gas values were reached for at least 1.5 h the control values for the hemodynamic parameters were taken.

Vehicle

Drugs were dissolved in 1 ml of dimethylsulfoxide (DMSO) to which 5 ml of polyethylenegyclol 400 (PEG400) was added. This solution was further diluted with 9 ml of water containing 20% glucose. Infusion rate was set to deliver 20 ml/h. This vehicle was used for all drugs and the control vehicle infusion.

Dose selection for hemodynamic drugs

As oral administration is difficult and not reliable during anesthesia the drugs were given as intravenous infusions over 5 to 30 min (infusion pump Unita, B. Braun) into the right internal jugular vein in order to avoid the effect of bolus doses. Doses for the pig experiments were derived from clinically recommended human oral doses (medium dose). The oral human dose recommended for the respective drug for the treatment of human hypertension was multiplied by the fraction of human oral bioavailability and divided by an assumed body weight of 75 kg. The pig dose was calculated according to the body weight (mg/kg) of our pigs. The effect of the drug on CABF and hemodynamics was followed during the whole infusion period. Reported are the values at the end of the infusion period if not indicated otherwise. Each drug reported in Table 1 was tested in a separate group of pigs.

The hemodynamic data were recorded by an electronic device (Plugsys from Hugo-Sachs Electronics, Harvard) and continuously stored on a computer hard disk by an online data acquisition and analysis system (Notocord HEM Software Evolution 4.2, Croissy-sur-Seine, France). Carotid artery resistance was continuously calculated from blood pressure and CABF by the software (second tracing in Fig. 1).

Fig.1

Effect of atipamezole (A) and xylazine (B) on carotid artery blood flow (abbreviated here as CBF), carotid artery resistance (abbreviated as CBF-resistance) and systolic and diastolic blood pressure (BP) and heart rate (HR) in anesthetized pigs (individual examples). Effect of atipamezole (A) and xylazine (B) on carotid artery blood flow (ml/min), carotid artery resistance (in mmHg*min/ml) and systolic and diastolic blood pressure (BP in mmHg) and heart rate (HR as bpm) in anesthetized pigs. Upper curve in the tracing “BP” shows systolic blood pressure, the lower curve shows diastolic blood pressure. Intravenous infusion starts with the first vertical line and ends with the second one. Atipamezole (20 μg/kg) was intravenously infused in 15 min, xylazine (1.86 mg/kg) in 30 min.

Euthanasia

At the end of the experiments, pigs were euthanized by an overdose of pentobarbital followed by a lethal dose of potassium chloride.

The following drugs were tested on CABF: Ramipril, irbesartan, bosentan, amlodipine, donepezil, H3-receptor antagonist A503, apomorphine, fluoxetine, xylazine, Nω-Nitro-L-arginine methyl ester (L-NAME), propranolol and cilostazol. Drugs were purchased from Sigma Aldrich, Germany, with the exception of the H3-receptor antagonist A503, ramipril and irbesartan, which were synthetized by Sanofi.

Statistical analysis

All averaged data are presented as the mean±SEM. The Student’s t-test was used to determine the significance of paired observations. For the statistical analysis of the effects of more than one drug administered in a group ANOVA for repeated measures was used followed by the Newman-Keuls test for a comparison between the different drugs or baseline and drugs. Differences were considered significant at p <0.05.

Results

Vehicle infused in 15–30 min had no effect on CABF or hemodynamic parameters in a separate vehicle group (Table 1). Common antihypertensive drugs including the ACE-inhibitor ramipril (0.3 mg/kg), the angiotensin receptor blocker irbesartan (2 mg/kg), the calcium channel blocker amlodipine (0.3 mg/kg), the endothelin receptor blocker bosentan (1 mg/kg) showed no changes in CABF or carotid artery resistance in deeply pentobarbital anesthetized pigs (Table 1). Likewise, dopamine-agonist apomorphine (0.1 mg/kg), fluoxetine (1 mg/kg), a histamine H3-receptor antagonist (0.3 mg/kg) and the cholinesterase-inhibitor donepezil (0.15 mg/kg) had no effect on CABF. Surprisingly, amlodipine lowered CABF because carotid artery resistance did not decrease with the fall in blood pressure (BP). The PDE-inhibitors cilostazol (0.3 mg/kg) increased CABF and decreased carotid artery resistance. BP and HR showed a slight rise after cilostazol.

Table 2

Effect of atipamezole, propranol, L-NAME and combinations on carotid artery blood flow and carotid artery resistance in anesthetized pigs

Experiment	Carotid artery flow (ml/min)			Carotid artery resistance (mmHg*min/ml)
		% change versus preceding drug	p-value versus preceding drug		% change versus preceding drug	p-value versus preceding drug
Baseline (n = 8)	114.0±10.0			0.78±0.11
Propranolol	116.0±11.5	2%		0.79±0.13	1%
Baseline (n = 6)	121.0±10.1			0.68±0.06
L-NAME	87.5±7.3	−28%	p <0.0001	1.54±0.15	128%	p <0.0001
Propranolol	66.3±4.6	−24%	p <0.001	1.85±0.17	20%	p <0.01
Baseline (n = 7)	118.6±11.2			0.72±0.08
Atipamezol	154.9±12.9	31%	p <0.0001	0.50±0.05	−30%	p <0.0001
Propranolol	126.,9±11.1	−18%	p <0.0001	0.58±0.08	17%	p <0.05
Baseline (n = 7)	127.9±10.5			0.79±0.11
L-NAME	96.3±8.3	−25%	p <0.01	1.39±0.12	76%	p <0.001
Atipamezol	124.1±13.4	29%	p <0.01	1.12±0.14	−19%	p <0.05
Propranolol	89.9±10.5	−28%	p <0.01	1.60±0.21	42%	p <0.01

Hemodynamic data –mean blood pressure and heart rate of these experiments - are shown in Table 3.

Table 3

Effect of atipamezole, propranol, L-NAME and combinations on mean blood pressure (BP) and heart rate in anesthetized pigs

Experiment	Mean BP (mmHg) Mean	p-value versus baseline	Heart rate (bpm) Mean	p-value versus baseline
Baseline (n = 8)	78.2±6.2		87.4±4.1
Propranolol	78.5 ±7.7	n.s	82.4±3.5	p <0.05
Baseline (n = 6)	76.6±6.9		96.0±8.4
L-NAME	129.8±11.5	p <0.0001	77.3±4.5	p <0.01
Propranolol	119.1±9.3	n.s	78.3±4.3	n.s
Baseline (n = 7)	75.8±2.2		86.0±4.6
Atipamezol	70.2±2.3	p <0.01	85.7±4.9	n.s
Propranolol	65.3±3.4	p <0.05	79.4±3.8	p <0.01
Baseline (n = 7)	91.0±6.5		85.9±7.3
L-NAME	126.0±6.2	p <0.0001	81.3±6.5	n.s
Atipamezol	127.7±6.8	n.s	82.3±6.5	n.s
Propranolol	132.8±9.2	n.s	82.0±5.2	n.s

Table 3 shows hemodynamic data for experiments shown in Table 2.

The α2-adrenergic receptor antagonist atipamezole (6.5 μg/kg) significantly reduced carotid artery resistance by 34% and raised CABF by 39% while the α2-adrenergic agonist xylazine (0.31 mg/kg) increased carotid artery resistance by 71% and decreased CABF by 37% significantly. The changes started very early after the beginning of the drug infusion and steadily increased until they almost reached a plateau (Fig. 1 shows typical examples).

Investigation of the mechanisms involved in the cerebrovascular vasodilator effect of atipamezole

Effect of propranolol and L-NAME alone and in combination on CABF

The inhibitor of the endothelial nitric oxide synthase (eNOS) L-NAME and the non-selective β-adrenergic receptor blocker propranolol were used to elucidate the mechanism by which atipamezole caused its cerebrovascular dilator effect. To judge possible effects of these drugs on the vasodilator effect of atipamezole, it was important to first investigate the effect of propranolol and L-NAME on CABF alone and the combination of both pharmacological tools. Propranolol given alone in a group of pigs had no effect on CABF and carotid artery resistance (Tables 2 and 3). L-NAME alone in contrast reduced CABF and strongly increased carotid artery resistance and BP as expected. Here, it is particularly important to look at the effect on carotid artery resistance as looking at the effect on CABF alone could be misleading because of the strong rise in BP as the perfusion pressure after L-NAME. The rise in BP considerably reduces the fall in CABF that would be expected from an almost doubling of the carotid artery resistance by L-NAME. This rise in BP after L-NAME tends to keep CABF high despite the strong increase in carotid artery resistance. While propranolol alone had no effect on CABF and carotid artery resistance, it further decreased CABF and increased carotid artery resistance after pretreatment with L-NAME significantly. Hence, adrenergic dysfunction (by high dose of propranolol) together with endothelial dysfunction (caused by L-NAME) had a strong synergistic vasoconstrictor effect on cerebral blood vessels.

Effect of propranolol and L-NAME on the vasodilator effect of atipamezole

In a separate group of pigs atipamezole increased CABF significantly. A bolus dose of propranolol after atipamezole significantly lowered CABF by raising carotid artery resistance, implicating β-adrenergic mechanisms in the rise of the vasodilator effect of atipamezole. Thus, propranolol diminished the vasodilator effect of atipamezole on cerebral vessels but it did not completely abolish it leaving room for another mechanism. Therefore, we investigated the effect of the eNOS-inhibitor L-NAME in combination with propranolol. Atipamezole given to another group of pigs pretreated with L-NAME still raised CABF significantly and decreased carotid artery resistance. Subsequent addition of the bolus dose of propranolol, however, fully reversed the rise in CABF and the decrease in carotid artery resistance, which was induced by atipamezole despite the presence of L-NAME. Hence, only L-NAME and propranolol together were able to abolish the vasodilator effects of atipamezole in a strong synergistic manner. After propranolol CABF was even lower than before administration of L-NAME and carotid artery resistance was higher. A synergistic effect was already seen combining these two drugs in the absence of atipamezole.

Discussion

In this study in young, healthy, pentobarbital anesthetized pigs, we show that α2-adrenergic drugs—agonists and antagonists—which modulate the release of NE in the brain noradrenergic nuclei, have a strong effect on unilateral CABF. Atipamezole, which stimulates the release of NE from noradrenergic nuclei, increased CABF, while xylazine inhibiting NE release decreased CABF by respective changes in carotid artery resistance. The literature on the effects of experimental LC stimulation on CBF is controversial and contradictory showing increases or decreases in CABF [25 –28]. Our own results using this pharmacological approach to elucidate the function of LC activation on CABF now strongly suggest that endogenously released NE from noradrenergic neurons has physiological vasodilator effects in cerebral vessels to increase CABF. This is in line with the expectation that the stimulatory activity of the noradrenergic system in the brain is associated with an increase in CBF. Theoretically, the controversial findings from reported experimental LC stimulation might be explained by the dual role of NE having either vasoconstrictor or vasodilator effects. The net effect after high intensity stimulation, which may release large amounts of NE, could also be vasoconstriction from a theoretical point of view because NE has vasoconstrictor or vasodilator effects depending on the presence of vasodilator β-adrenergic versus vasoconstrictor α-adrenergic vascular receptors and it could act differently at low versus high concentrations. However, the rise in CABF in our pigs showed a plateau or a very small rise after a phase of a fast rise (Fig. 1). There was no biphasic pattern that could be interpreted in the sense of a vasoconstrictor effect. Similar observations were made with xylazine in the opposite direction. Hence, our findings don’t offer any clue as to why the results obtained with LC-stimulation on CBF in animals were contradictory. The amount of NE that can be released by atipamezole as a blocker of presynaptic α2-adrenergic receptor is limited to the amount that is inhibited by the action of NE on the presynaptic α2-adrenergic receptor. A plateau of CABF increase was reached soon after starting the infusion of atipamezole which may indicate that the maximum of NE release by this mechanism had been achieved early. Strong LC stimulation may release more NE than atipamezole which still has the potential to cause vasoconstriction. Our conclusion of a vasodilator function of NE released in the brain, most likely via projections from the LC reaching the blood vessels, does not automatically apply to NE released in the periphery and transported to the brain via the blood stream which theoretically could reach vascular structures endowed with more vasoconstrictor α-adrenergic receptors than vasodilator β-adrenergic receptors.

CABF was measured after ligation of the external branches of the carotid artery after having studied pig carotid anatomy using vascular casts. Flow into a vascular bed is only determined by its resistance and the perfusion pressure which is BP in this case. Therefore, ligation of the external branches does not have a direct effect (contrary to what one might intuitively assume) on the internal carotid artery except for a small rise in total vascular resistance resulting from the occlusion. This would lead to a small rise in BP as the perfusion pressure for the internal carotid artery which would be considered in the calculation of the vascular resistance of the internal cerebral carotid artery. A limitation of the study is that measuring CABF does not show how the blood is distributed into the different regions of the brain and the question is whether it represents CBF. We used the principle of gate control of influx or efflux into a system. The four “gates” of blood flow to the brain are the two carotid arteries and the vertebral arteries. One carotid artery represents the other. There is also no doubt that flow through the carotid artery is much higher than through the vertebral artery, already because of the different size of both arteries being exposed to the same perfusion pressure. Therefore, we think that the changes in CABF which are induced by the drugs used in this investigation represent the net changes in CBF. This investigation is about the regulation of blood flow to the brain by different drugs with no claim to distinguish between effects in different vascular sections along the vessel. In this paper we do not investigate either how an increased or decreased CABF is regionally distributed.

Common vasodilator and antihypertensive drugs including inhibitors of ACE by ramipril, of angiotensin II receptors by irbesartan, of endothelin receptors by the antagonist bosentan and of L-type calcium channels by amlodipine did not increase CABF in our anesthetized healthy pigs in line with scarce human data on CBF showing either no or little or inconsistent effects on CBF with these drugs. The lack of effects of ACE-, angiotensin-receptor-, and endothelin-inhibition in healthy young animals on acute hemodynamic parameters is a well-known phenomenon in cardiovascular pharmacology as these physiological systems are rather in an inactive state in young, healthy, and well-hydrated animals. In dementia, cardiovascular diseases and old age these systems might be activated either locally or systemically so that antagonism of their vasoconstrictor effects then might show an acute change in CBF. The PDE-inhibitor cilostazol, however, showed a decrease in carotid artery resistance indicating vasodilation and an increase in CABF in line with experimental and human data [31 –33]. cAMP is raised by the PDE-inhibitor cilostazol by inhibiting its degradation. cAMP is physiologically elevated by β-adrenergic stimulation by NE (among other stimulants) which fits to our reasoning on second messengers involved in the vasodilator mechanisms of NE that will be elaborated below [29].

Effect of other stimulants on CABF

The change in CABF induced by the α2-adrenergic modulator drugs is unique compared with four other drugs raising or mimicking other stimulatory transmitters. The dopamine-agonist apomorphine, the serotonin uptake inhibitor fluoxetine, a histamine H3-receptor antagonist, and the cholinesterase-inhibitor donepezil had no effect on CABF in our anesthetized pigs. Fluoxetine raises levels of serotonin while H3-receptor antagonism facilitates the release of histamine and other transmitters such as acetylcholine, NE, and serotonin. The isolated and unique effect of the noradrenergic system in the regulation of CABF in this model compared with other stimulating transmitters can be well explained. The LC innervates the cerebral blood vessels by noradrenergic projections in the brain by volume transmission. Functional importance of this vascular innervation by the LC is suggested by the observation that vascular adrenergic receptors are upregulated after experimental LC-destruction and in AD [22, 24]. No such direct innervation of cerebral blood vessels exists for the other stimulating transmitters nor are those receptors present on the vessels. In the non-anesthetized state these other stimulating transmitters could possibly indirectly increase CBF by activating neurons via the mechanisms involved in neurovascular coupling. In deep pentobarbital anesthesia neuronal activity is reduced and so probably also its possibility of raising CBF indirectly by the metabolic path [34]. We saw no signs of flattening or deepening of anesthesia by atipamezole or xylazine by watching the animals and controlling for reflexes. Obviously, even in deep pentobarbital anesthesia, which even lack spontaneous breathing, noradrenergic neurons are still operative so that the release of NE can be modulated by α2-adrenergic drugs.

Elucidation of the mechanisms involved in noradrenergic cerebrovascular dilatation

α2-adrenergic agonist xylazine, which inhibits NE release, decreased CABF in deep pentobarbital anesthesia while stimulation of NE release by the α2-adrenergic antagonist atipamezole raised CABF leading to the conclusion that endogenously released NE causes vasodilation. Current physiological knowledge is that stimulation of β-adrenergic mechanisms has vasodilator effects via direct effects in VSCMs and indirect effects via the endothelium while stimulation of α-adrenergic receptors on VSCMs is vasoconstrictor. By contrast, stimulation of α2-adrenergic receptors on endothelial cells releases the vasodilator NO. With regard to β-adrenergic mechanisms potentially explaining the vasodilator effect stimulation of β1-adrenergic receptors on VSCMs causes vasodilation via a rise in cAMP [29]. It is important to remind that the PDE-inhibitor cilostazol raising cAMP levels increased CABF in our pigs; cAMP can be seen as one of the downstream effectors of noradrenergic vascular vasodilator innervation. Stimulation of β2- and β3-adrenergic receptors on endothelial cells stimulates eNOS and COX to produce nitric oxide (NO) and prostacyclin by raising cAMP and by stimulating the Pi3/Akt pathway [29, 35].

We decided to use the β-adrenergic receptor blocker propranolol, which inhibits both β1- and β2-adrenergic receptors, and the eNOS inhibitor L-NAME to find out a possible role of β-adrenergic mechanisms in cerebrovascular dilatation induced by the α2-adrenergic antagonist atipamezole. Unfortunately, we could not find an appropriate β3-adrenergic receptor antagonist, which is effective in the pig as β3-adrenergic endothelial receptors are involved in the release of NO, to block all three known β-adrenergic receptor subtypes. As the main presumed vascular effect of β2- and β 3-adrenergic activation on endothelial cells is the release of NO via stimulation of eNOS we used the eNOS-inhibitor L-NAME.

To judge possible effects of these drugs on the vasodilator effect of atipamezole, it was important to investigate the effect of propranolol and L-NAME on CABF alone and the combination of both pharmacological tools, leading to a surprising discovery. As expected, L-NAME alone strongly reduced CABF and increased carotid artery resistance and BP. Propranolol given alone had no effect on CABF and carotid artery resistance. However, surprisingly, propranolol given in the L-NAME pretreated pigs further decreased CABF and increased carotid artery resistance in a significant manner revealing a role of β-adrenergic activation for cerebrovascular vasodilation that seemed to be masked by an intact, uninhibited eNOS. The use of L-NAME highlights the strong and important role of NO for brain perfusion. L-NAME also inhibits the neuronal nitric oxide synthase, which may contribute to its vasoconstrictor effects further reducing the availability of NO in vascular smooth muscle cells for vasodilation.

Obviously, pharmacological induction of an adrenergic dysfunction by a high dose of propranolol in the presence of endothelial dysfunction (pharmacological eNOS inhibition) leads to a synergistic decrease in CABF.

The rise in CABF by atipamezole when given as the first drug (to another group) was only partially but significantly reduced by a bolus dose of propranolol revealing involvement of β1/ β2-adrenergic mechanisms; CABF was still clearly higher and carotid artery resistance lower than without atipamezole. When atipamezole was given after pretreatment with L-NAME (in a separate experimental group) there was still a marked rise in CABF and a decrease in carotid artery resistance by atipamezole. After subsequent administration of a bolus dose of propranolol to the pigs that had already received L-NAME plus atipamezole, however, CABF fell and carotid artery resistance rose markedly, clearly indicating involvement of β-adrenergic mechanisms. Propranolol in this sequence of drugs applied not only fully reversed the changes induced by atipamezole but even caused a stronger vasoconstrictor effect than L-NAME alone (a trend). Hence, the magnitude of the vasoconstrictor effect of propranolol on the vasodilator effect of atipamezole was much stronger in L-NAME pretreated pigs than in the absence of L-NAME. Only L-NAME and propranolol together abolished the vasodilator effect of atipamezole. To summarize, propranolol had no cerebrovascular constrictor effect when given alone, a moderate vasoconstrictor effect on the atipamezole stimulated rise in CABF (in the absence of L-NAME) and a stronger vasoconstrictor effect on the atipamezole effect in L-NAME pretreated pigs and a clear vasoconstrictor effect in L-NAME pretreated pigs (no atipamezole). The fact that propranolol had no effect when given alone may be explained that in our deeply anesthetized pigs basal NE release acting on the cerebral vessels is low or absent, and, thus, propranolol has no effect on its own (in the absence of an adrenergic agonist). However, in the L-NAME pretreated group (no atipamezole) propranolol showed a clear cerebral vasoconstrictor effect although there is also no obvious reason why NE release should be augmented. Thus, eNOS inhibition by L-NAME seems to unmask a vasoconstrictor effect of β-adrenergic blockade. In the discussion on possible mechanisms, one should consider that apart from NO prostacyclin plays an important role in vascular function; both transmitters are co-released by endothelial cells by the same stimuli to act in synergy [36], although we have no concrete mechanistic explanation why the vasoconstrictor effect of propranolol only becomes apparent or gets stronger after inhibition of eNOS. We did not investigate the role of prostanoids in the action of atipamezole.

Altogether, the effect of propranolol reveals a role of β-adrenergic mechanisms in the cerebrovascular dilator effect of atipamezole meaning that the physiological release of NE increases CABF by β-adrenergic mechanisms. Secondly, there is a strong pharmacological interaction of propranolol and L-NAME in reducing CABF—with and without atipamezole—suggesting a pathophysiological potential which will be discussed below.

Experiments were performed in pigs in pentobarbital anesthesia, which is a deep anesthesia, that at the doses used most likely alters CBF. Typically, autoregulatory plateaus of CBF are not seen in deep anesthesia [37] which we can confirm in our pigs as hemorrhagic blood loss did not show any plateau for CABF but a fall in CABF with a falling BP (data not shown). Moreover, most likely due to the strong inhibitory effects of anesthesia on neurons, their metabolic activity and hence their metabolic effect on CBF is low so that CBF is usually lower at least in deep anesthesia compared with the awake state [34]. Since a systematic pharmacology of CBF does not exist, and given the paucity of pharmacological CBF investigations and data the limitation that such experiments may only be possible in anesthetized animals with a reasonable throughput should be accepted and discussed later on in the light of the concrete results delivered. To validate the experimental approach among 10 other drugs tested here only cilostazol showed a rise in CABF in line with human data and L-NAME showed the expected decrease.

Despite these considerations or limitations arising from anesthesia, we cannot see any potential anesthesia-related mechanism that could change the conclusion of a vasodilator role of NE in cerebral blood vessels in our experiments. Assuming the true physiological role of NE released from noradrenergic brain nuclei were vasoconstriction in the awake state as could be inferred from publications showing a vasoconstrictor effect of experimental LC stimulation (while others showed the opposite), why and how should anesthesia reverse it into a vasodilator role? Cerebral vessels are innervated by noradrenergic projections from the LC which we assume is the path for the vascular effects of NE [22]. It then depends on the presence of the type of adrenergic receptors at the site of volume release of NE by the noradrenergic projections from the LC whether vasodilator or vasoconstrictor effects of NE predominate. Vasodilator β-adrenergic receptors and α2-adrenergic receptors, which stimulate the release of NO from EC, have been found on cerebral vessels and their expression was found increased in AD and after chemical LC destruction in animals [22, 24]. This upregulation could indicate a lack of the physiological transmitter potentially occurring as a consequence of LC degeneration (de-afferentiation supersensitivity).

Direct noradrenergic innervation of cerebral vessels is not the only mechanism of possible vasodilation by NE. NE is a stimulatory transmitter which could increase CBF indirectly by activating neurons by the metabolic path (neurovascular coupling). However, this mechanism is probably diminished by the inhibitory effect of anesthesia on neurons and their metabolism so that in the awake state vasodilator influences of NE could even be stronger compared with what we found in anesthesia here [34].

Our conclusion that NE released as a consequence of the inhibition of presynaptic α2-adrenergic receptors is a physiological vasodilator to increase CBF depends on the assumption that the effect of these α2-adrenergic receptor modulating drugs stimulating or inhibiting NE release is predominant on the presynaptic α2-adrenegic receptors over possible direct postsynaptic vascular (α-adrenergic) effects. The latter are complex as they include direct vasoconstrictor effects on VSCMs and possible vasodilator effects mediated by endothelial NO release by α2-adrenergic stimulation [35]. This is most likely the case as the therapeutic use of these drugs in anesthesia is entirely based on the modulation of the presynaptic α2-receptors increasing or decreasing NE release; postsynaptic α2-receptors are not involved in the anesthetic mechanism. In contrast to our results, Bekar et al. [38] report vasoconstrictor effects of atipamezole on regional CBF in the hyperemic response to forelimb and hindlimb stimulation in an anesthetized rat model. We used a low or moderate dose of atipamezole (6.5 μg/kg) in an attempt to elucidate the physiological vascular function of very moderate increases in NE to get an idea of what could happen if the noradrenergic system failed (LC degeneration) while those investigators used a dose as high as 2-3 mg/kg. High intensity LC stimulation and/or high doses of atipamezole may release high amounts of NE that can well have vasoconstrictor effects from a theoretical point of view. Speculatively, this may rather relate to pathological stress responses than reflect the basic physiological function.

A possible effect of the α2-adrenergic receptor antagonist atipamezole on the adrenal gland, which means the possibility of peripherally released catecholamines, should also be considered but can be excluded since atipamezole did not raise heart rate [39].

Our experiments were performed in anesthetized pigs. The limitation that could result from anesthesia has been discussed above. The question is whether the effects found in pigs could be species dependent. There is no doubt that NE exerts vasoconstrictor effects when applied to isolated vessels or systemically or directly onto cerebral vessels in situ. The precise question, however, should be whether NE released from noradrenergic projections from the LC innervating the cerebral blood vessels has vasoconstrictor or vasodilator effects in humans. These human vessels are endowed with β-adrenergic receptors [22, 24] whose physiological function is vasodilation. If NE release from LC projections to cerebral vessels had vasoconstrictor effects as claimed by other authors in this controversial discussion on the role of NE released from the LC on cerebrovascular regulation, one would rather expect cerebral hyperperfusion in AD due to a loss of a vasoconstrictor influence because of LC degeneration instead of the demonstrated hypoperfusion.

Possible pathophysiological implications

Altogether, our investigations with α2-adrenergic modulator drugs and the non-selective β-adrenergic receptor blocker propranolol strongly suggest that NE released in noradrenergic brain nuclei such as the LC has a net physiological vasodilator role for cerebral blood vessels to increase blood flow as expected for a stimulating transmitter.

Therefore, we believe that LC degeneration occurring as the earliest detectable regional degenerative change in AD by causing noradrenergic dysfunction or deficiency has the potential to be causally involved in early cerebral hypoperfusion of AD. LC degeneration may exacerbate the development of AD not only by a deficit in neuronal stimulation or by unleashing neuroinflammation [7, 8] but also by causing or at least by contributing to cerebral hypoperfusion.

Since stimulation of β2- and β3- and α2-adrenergic receptors on endothelial cells activates the eNOS to release NO from the endothelium noradrenergic dysfunction (LC degeneration) could lead to a chronic deficit in endothelial NO-release in cerebral vessels and cause or contribute to cerebrovascular endothelial dysfunction [29]. Endothelial dysfunction has been implicated in cerebrovascular disease and AD [40]. Vascular risk factors are associated with cerebral hypoperfusion in AD [14 –17]. Therefore, it is reasonable to hypothesize that LC degeneration in AD by leaving a noradrenergic vascular innervation disturbance may provide an early and basic cerebrovascular disorder onto which each of the known different cardiovascular risk factors for AD by inducing endothelial dysfunction might act highly synergistically to cause cerebral hypoperfusion. The hypothesis is strongly supported by the synergistic effect of the combination of the non-selective beta-adrenergic receptor blocker propranolol with L-NAME to reduce CABF in experiments originally made as control experiments (no atipamezole). Systemic endothelial dysfunction, which arises from vascular risk factors such as CV diseases and age and which also affects the brain vessels [40], together with an early brain or AD specific change—noradrenergic vascular innervation deficit by LC degeneration—could act highly synergistically to explain early cerebral hypoperfusion in AD. What causes early degeneration of the LC is an unanswered question that is outside the scope of this publication [7, 23].

In conclusion, our investigations in anesthetized pigs with α2-adrenergic modulator drugs and the nonselective β-adrenergic receptor blocker propranolol strongly suggest that NE released from noradrenergic brain regions has a net physiological vasodilator role for cerebral blood vessels to increase blood flow involving β-adrenergic vasodilator mechanisms. Therefore, if NE had the same role in non-anesthetized humans, early degeneration of the LC in AD and the ensuing noradrenergic dysfunction could be causally involved in cerebral hypoperfusion occurring early in AD. Cerebral noradrenergic dysfunction together with systemic endothelial dysfunction arising from vascular risk factors and age could act highly synergistically to reduce CBF.

Disclosure STATEMENT

The author’s disclosure is available online (https://www.j-alz.com/manuscript-disclosures/18-0340r2).

References

Berridge

, Waterhouse

(2003) The locus coeruleus noradrenergic system: Modulation of behavioral state and state-dependent cognitive processes. Brain Res Brain Res Rev 42, 33–84.

Tomlinson

, Irving

, Blessed

(1981) Cell loss in the locus coeruleus in senile dementia of Alzheimer type. J Neurol Sci 49, 419–428.

Grudzien

, Shaw

, Weintraub

, Bigio

, Mash

, Mesulam

(2007) Locus coeruleus neurofibrillary degeneration in aging, mild cognitive impairment and early Alzheimer’s disease. Neurobiol Aging 28, 327–335.

Braak

, Thal

, Ghebremedhin

, Del Tredici

(2011) Stages of the pathologic process in Alzheimer disease: Age categories from 1 to 100 years. J Neuropathol Exp Neurol 70, 960–969.

Stratmann

, Heinsen

, Korf

, Del Turco

, Ghebremedhin

, Seidel

, Bouzrou

, Grinberg

, Bohl

, Wharton

, den Dunnen

, Rüb

(2016) Precortical phase of Alzheimer’s disease (AD)-related tau cytoskeletal pathology. Brain Pathol 26, 371–386.

Theofilas

, Dunlop

, Heinsen

, Grinberg

(2015) Turning on the light within: Subcortical nuclei of the isodentritic core and their role in Alzheimer’s disease pathogenesis. J Alzheimers Dis 46, 17–34.

Feinstein

, Kalinin

, Braun

(2016) Causes, consequences, and cures for neuroinflammation mediated via the locus coeruleus: Noradrenergic signaling system. J Neurochem 139(Suppl 2), 154–178.

Gannon

, Che

, Chen

, Jiao

, Roberson

, Wang

(2015) Noradrenergic dysfunction in Alzheimer’s disease. Front Neurosci 9, 220.

Kogure

, Matsuda

, Ohnishi

, Asada

, Uno

, Kunihiro

, Nakano

, Takasaki

(2000) Longitudinal evaluation of early Alzheimer’s disease using brain perfusion. SPECT J Nucl Med 41, 1155–1162.

10.

Matsuda

(2001) Cerebral blood flow and metabolic abnormalities in Alzheimer’s disease. Ann Nucl Med 15, 85–92.

11.

Johnson

, Albert

(2000) Perfusion abnormalities in prodromal AD. Neurobiol Aging 21, 289–292.

12.

Okonkwo

, Xu

, Oh

, Dowling

, Carlsson

, Gallagher

, Birdsill

, Palotti

, Wharton

, Hermann

, LaRue

, Bendlin

, Rowley

, Asthana

, Sager

, Johnson

(2014) Cerebral blood flow is diminished in asymptomatic middle-aged adults with maternal history of Alzheimer’s disease. Cereb Cortex 24, 978–988.

13.

Hays

, Zlatar

, Wierenga

(2016) The utility of cerebral blood flow as a biomarker of preclinical Alzheimer’s disease. Cell Mol Neurobiol 36, 167–179.

14.

Wolters

, Zonneveld

, Hofman

, van der Lugt

, Koudstaal

, Vernooij

, Ikram

; Heart-Brain Connection Collaborative Research Group (2017) Cerebral perfusion and the risk of dementia: A population based study. Circulation 136, 719–728.

15.

de la Torre

(2012) Cerebral hemodynamics and vascular risk factors: Setting the stage for Alzheimer’s disease. J Alzheimers Dis 32, 553–567.

16.

De Silva

, Faraci

(2016) Microvascular dysfunction and cognitive impairment. Cell Mol Neurobiol 36, 241–258.

17.

Kisler

, Nelson

, Montagne

, Zlokovic

(2017) Cerebral blood flow regulation and neurovascular dysfunction in Alzheimer Disease. Nat Rev Neurosci 18, 419–434.

18.

Di Marco

, Venneri

, Farkas

, Evans

, Marzo

, Frangi

(2015) Vascular dysfunction in the pathogenesis of Alzheimer’s disease - A review of endothelium-mediated mechanisms and ensuing vicious circles. Neurobiol Dis 82, 593–606.

19.

de la Torre

, Mussivand

(1993) Can disturbed brain microcirculation cause Alzheimer’s disease? Neurol Res 15, 146–153.

20.

de la Torre

(2000) Impaired cerebromicrovascular perfusion. Summary of evidence in support of its causality in Alzheimer’s disease. Ann N Y Acad Sci 924, 136–152.

21.

Farkas

, Luiten

(2001) Cerebral microvascular pathology in aging and Alzheimer’s disease. Prog Neurobiol 64, 575–611.

22.

Kalaria

, Stockmeier

, Harik

(1989) Brain microvessels are innervated by Locus Ceruleus noradrenergic neurons. Neurosci Lett 97, 203–208.

23.

Satoh

, Iijima

(2017) Roles of tau pathology in the locus coeruleus (LC) in age-associated pathophysiology and Alzheimer’s disease pathogenesis: Potential strategies to protect the LC against aging. Brain Res. 10.1016/j.brainres.2017.12.027.

24.

Kalaria

, Andorn

, Harik

(1989) Alterations in adrenergic receptors of frontal cortex and cerebral microvessels in Alzheimer’s disease and aging. Prog Clin Biol Res 317, 367–374.

25.

Toussay

, Basu

, Lacoste

, Hamel

(2013) Locus coeruleus stimulation recruits a broad cortical neuronal network and increases cortical perfusion. J Neurosci 33, 3390–3401.

26.

de la Torre

(1976) Evidence for central innervation of intracerebral blood vessels: Local cerebral blood flow measurements and histofluorescence analysis by the sucrose-phosphate-glyoxylic acid (SPG) method. Neuroscience 1, 455–457.

27.

Goadsby

, Duckworth

(1989) Low frequency stimulation of the locus coeruleus reduces regional cerebral blood flow in the spinalized cat. Brain Res 476, 71–77.

28.

Raichle

, Hartman

, Eichling

, Sharpe

(1975) Central noradrenergic regulation of cerebral blood flow and vascular permeability. Proc Natl Acad Sci U S A 72, 3726–3730.

29.

Rath

, Balligand

, Dessy

(2012) Vasodilator mechanisms of beta receptor blockade. Curr Hypertens Rep 14, 310–317.

30.

Giovannitti

Jr , Thoms

, Crawford

(2015) Alpha-2 adrenergic receptor agonists: A review of current clinical applications. Anesth Prog 62, 31–39.

31.

Kawamura

, Watanabe

, Kimura

(1985) Effect of cilostazol, a new antithrombotic drug, on cerebral circulation. Arzneimittelforschung 35(7A), 1149–1154.

32.

Sakurai

, Hanyu

, Sato

, Kume

, Hirao

, Kanetaka

, Iwamoto

(2013) Effects of cilostazol on cognition and regional cerebral blood flow in patients with Alzheimer’s disease and cerebrovascular disease: A pilot study. Geriatr Gerontol Int 13, 90–97.

33.

Matsumoto

, Shimodozono

, Miyata

, Kawahira

(2011) Effect of cilostazol administration on cerebral hemodynamics and rehabilitation outcomes in poststroke patients. Int J Neurosci 121, 271–278.

34.

Van Aken

, Van Hemelrijck

(1993) An overview of the influence of anesthesia on cerebral blood flow and cerebral metabolism. Minerva Anestesiol 59, 615–620.

35.

Conti

, Russomanno

, Corbi

, Izzo

, Vecchione

, Filippelli

(2013) Adrenoreceptors and nitric oxide in the cardiovascular system. Front Physiol 4, 321, 1–11.

36.

Mitchell

, Ali

, Bailey

, Moreno

, Harrington

(2008) Role of nitric oxide and prostacyclin as vasoactive hormones released by the endothelium. Exp Physiol 93, 141–147.

37.

Dagal

, Lam

(2009) Cerebral autoregulation and anesthesia. Curr Opin Anaesthesiol 22, 547–552.

38.

Bekar

, Wei

, Nedergaard

(2012) The locus coeruleus-norepinephrine network optimizes coupling of cerebral blood volume with oxygen demand. J Cereb Blood Flow Metab 32, 2135–2145.

39.

Artalejo

, Olivos-Oré

(2018) Alpha2-adrenoceptors in adrenomedullary chromaffin cells: Functional role and pathophysiological implications. Pflugers Arch 470, 61–66.

40.

Katusic

, Austin

(2016) Neurovascular protective function of endothelial nitric oxide - recent advances. Circ J 80, 1499–1503.