Introduction

Sympathoexcitation is closely associated with the pathogenesis of hypertension and the related organ damage.1, 2, 3, 4, 5 The sympathetic nervous system is regulated by the brain, especially the rostral ventrolateral medulla (RVLM) in the brainstem.6, 7 There are inputs into the RVLM from baroreceptors, chemoreceptors and visceral receptors via the nucleus of the solitary tract (NTS)7, 8, 9, 10, 11, 12 and the paraventricular nucleus of the hypothalamus.8, 13, 14 Therefore, the RVLM is known as a vasomotor center, which determines the basal central sympathetic outflow.

The effect of nitric oxide (NO) and/or oxidative stress in the brain on the regulation of the sympathetic nervous system has been investigated in many previous studies.15, 16, 17, 18, 19, 20, 21, 22, 23, 24 In particular, we have investigated the role of NO and/or oxidative stress in the RVLM. In the present review of our previous studies, we describe the importance of NO and/or angiotensin II type 1 receptor (AT1R)-induced oxidative stress in the RVLM on the regulation of the sympathetic nervous system and the contribution of these mechanisms to the pathophysiology of hypertension.

NO derived from an overexpression of eNOS in the RVLM causes sympathoinhibition

NO is an important mediator of intracellular signaling in various tissues, including the central nervous system.25, 26, 27 There are three NO synthase (NOS) isoforms: the constitutive neuronal NOS (nNOS), endothelial NOS (eNOS) and inducible NOS (iNOS). A number of studies have demonstrated the localization of nNOS, eNOS and iNOS within the brain.28 Considerable evidence indicates that NOS acts on the cardiovascular system, including the receptors and effectors of the baroreflex pathway.15, 16, 17 We previously developed an in vivo technique for eNOS gene transfer into the NTS of rats to determine the effects of NO in the NTS on the sympathetic nervous system in a conscious state.20, 24, 29, 30, 31 Compared to studies of the NTS, studies of the RVLM in both acute and anesthetized models have resulted in both sympathoexcitation and sympathoinhibition.24, 32, 33, 34, 35, 36, 37

In an experiment using whole-cell patch-clamp recordings of visualized RVLM neurons from a brainstem slice, NO increased the amplitude of excitatory postsynaptic currents.37 Therefore, to investigate the role of NO in the RVLM in a conscious state, we applied the technique described above to study the RVLM.38, 39, 40 In those studies, prolonged sympathoinhibition was observed after the overexpression of eNOS in the RVLM.38, 39, 40 Microinjection of either a NOS inhibitor or bicuculline, a γ-amino butyric acid (GABA) receptor antagonist, into the RVLM increased blood pressure to greater levels in the group with overexpression of eNOS in the RVLM compared with the control group. In addition, GABA levels in the RVLM measured by in vivo microdialysis increased in the group with overexpression of eNOS in the RVLM.

These results indicate that NO derived from the overexpression of eNOS in the RVLM caused sympathoinhibition mediated by an increased release of GABA. Similar to previous studies,16, 41 we suggest that NO derived from the overexpression of eNOS would mainly mediate GABA release by the activation of N-methyl-D-aspartate (NMDA) receptors in presympathetic neurons in the RVLM. In addition, these studies provided convincing evidence that chronic changes in neurotransmitters/neuromodulators in the RVLM have a sustained impact on blood pressure in awake animals. In stroke-prone spontaneously hypertensive rats (SHRSP), overexpression of eNOS in the RVLM elicited greater depressor and sympathoinhibitory responses with the enhancement of GABA release than in normotensive rats.39 These results could indicate that NO dysfunction and the resulting disinhibition of the RVLM contribute to sympathoexcitation in SHRSP animals. Furthermore, overexpression of eNOS in the RVLM improved the impaired baroreflex control of heart rate in SHRSP animals.40 In summary, NO derived from an overexpression of eNOS in the RVLM caused sympathoinhibition via GABA release and facilitated the baroreflex function.

However, we were unable to fully determine whether NO in the RVLM causes sympathoinhibition because we examined the role of NO derived from an overexpression of eNOS in the RVLM. eNOS, nNOS and iNOS are all located in the RVLM.42 Previous studies have demonstrated that the application of a selective nNOS inhibitor into the RVLM caused a depressor effect.43, 44, 45 Moreover, the sympathoexcitation elicited by epicardial application of bradykinin was significantly attenuated by the unilateral microinjection of non-selective NOS inhibitors or a specific nNOS inhibitor into the RVLM, although the sympathoexcitation was unchanged by the injection of an eNOS inhibitor into the RVLM.46 Another previous report indicated that NO derived from nNOS induced sympathoexcitation via both NMDA and non-NMDA receptors.47 In contrast, overexpression of nNOS in the RVLM normalized the impaired baroreflex function due to sympathoinhibition in rats with heart failure.48

With regard to iNOS, we demonstrated that the overexpression of iNOS in the RVLM caused sympathoexcitation in a conscious state of normotensive rats49 and that the iNOS expression in the RVLM was enhanced in spontaneously hypertensive rats (SHR).50 However, other studies have indicated that the NO derived from iNOS caused sympathoinhibition44, 45 via GABA receptors47 and that iNOS in the RVLM was reduced in SHR.42 Inconsistent results have also been shown within the same species,34, 35, 36 although the variable actions of NO in the RVLM observed in different species might be explained by interspecies variation in NOS density and distribution.32, 33 These conflicting results suggest that the effects of NO on the regulation of the sympathetic nervous system would be different among NOS isoforms.

In addition, we have not clarified the effect of NO on the release of the excitatory amino acid glutamate in the RVLM.51 Stimulation of the cardiac sympathetic pathways activates RVLM neurons associated with glutamate as well as nNOS,52 and NO facilitates the action of glutamate in the RVLM through cyclic GMP/protein.37, 47 Although we previously demonstrated that overexpression of eNOS in the RVLM increased the release of glutamate, microinjection of a glutamate receptor antagonist into the RVLM did not alter blood pressure.32 Further examinations to clarify the precise mechanisms for the effects of each NOS isoform and glutamate in the RVLM are necessary.

Oxidative stress in the RVLM causes sympathoexcitation in hypertensive rats

We focused on oxidative stress in the RVLM as a major sympathoexcitation factor because oxidative stress acts as a counterpart to NO. We measured oxidative stress in the RVLM using the electron-spin resonance method or the thiobarbituric acid-reactive substances method and demonstrated that oxidative stress was increased in the RVLM of SHRSP animals53, 54, 55, 56, 57, 58, 59, 60, 61 and SHR.62, 63 We also confirmed that a microinjection of tempol, a membrane-permeable superoxide dismutase (SOD) mimetic, into the RVLM decreased blood pressure and heart rate in SHRSP animals but not in normotensive rats. Furthermore, we transfected adenoviral vectors encoding the manganese SOD (MnSOD) gene into the RVLM in SHRSP animals. Overexpression of MnSOD in the RVLM decreased blood pressure, heart rate and urinary norepinephrine excretion in SHRSP animals but not in normotensive rats. We also found reduced SOD activity in the RVLM of SHRSP animals compared with normotensive rats, and this reduction led to a decreased capability for scavenging superoxide anions.

These findings indicate that oxidative stress in the RVLM causes sympathoexcitation, and this mechanism was involved in the neural pathophysiology of hypertension in SHRSP animals. We have also demonstrated that oxidative stress in the RVLM caused sympathoexcitation in several other hypertensive models, such as salt-induced hypertension,62 dietary-induced hypertension57, 64 and experimental jet lag.56 Moreover, these results are consistent with previous studies of other investigators.65, 66 In addition, Oliveira-Sales et al.67, 68, 69 suggested that sympathoexcitation in renovascular (two-kidney one-clip) hypertensive rats was associated with oxidative stress in the RVLM and paraventricular nucleus of the hypothalamus. These results strongly suggest that the increase in oxidative stress in the RVLM is an important cause, not a result, of the sympathoexcitation, which then leads to hypertension.

Sources of oxidative stress in the RVLM

Among the sources of oxidative stress in the RVLM, such as nicotinamide adenine dinucleotide phosphate [NAD(P)H] oxidase, xanthine oxidase, uncoupled NOS and mitochondria, we demonstrated that mainly the activation of the AT1 receptor/NAD(P)H oxidase produced the oxidative stress in the RVLM of SHRSP animals.58, 70, 71 These results are consistent with a previous report.72 The renin angiotensin system in the brain is associated with enhanced central sympathetic outflow,73, 74, 75, 76, 77, 78 and mitochondria-derived oxidative stress has been shown to mediate sympathoexcitation induced by angiotensin II in the brain.79, 80 Angiotensin II exogenously administered into the RVLM elicited a pressor response via sympathoexcitation.81, 82 Moreover, the inhibition of the AT1 receptor in the RVLM by AT1 receptor blockers caused sympathoinhibition in hypertensive rats.57, 58, 59, 62, 64, 71, 81 We found that the overexpression of MnSOD attenuated the angiotensin II-induced pressor response associated with oxidative stress in the RVLM.70 All of these results indicate that the AT1 receptor/NAD(P)H oxidase may be the main source of oxidative stress in the RVLM of hypertensive rats. Moreover, we showed that Rac1 was associated with the activation of NAD(P)H oxidase.20, 22, 23, 24, 55, 79

Rac1 is a small G protein involved in integrating the intracellular transduction pathways toward NAD(P)H activation and requires lipid modifications to migrate from the cytosol to the cell membrane. The inhibition of Rac1 caused by the transfection of the adenovirus vectors encoding a dominant-negative Rac1 into the RVLM or NTS decreased blood pressure, heart rate and urinary norepinephrine excretion in SHRSP animals but not in normotensive rats.22, 79 A blockade of Rac1 translocation from the cytosol to the membrane in the RVLM of SHRSP animals caused sympathoinhibition via a decrease in NAD(P)H oxidase activity and oxidative stress.55 Thus, our findings indicate that activation of the AT1 receptor/NAD(P)H oxidase associated with Rac1 mainly produced oxidative stress in the RVLM of SHRSP animals.

Mechanisms of oxidative stress-induced sympathoexcitation in the RVLM

We have demonstrated several mechanisms by which oxidative stress in the RVLM causes sympathoexcitation. Superoxide interacts with NO. As we consider that NO in the RVLM causes sympathoinhibition,29, 38, 39 decreased NO bioavailability in the RVLM induced by superoxide might cause sympathoexcitation. Recently, we demonstrated that a reduction in NO-mediated GABA release in the RVLM was involved in superoxide-induced sympathoexcitation in SHRSP animals.83 Regarding the relationship between superoxide and NO, we should also focus on the formation of peroxynitrite because the kinetics of the formation of peroxynitrite from superoxide and NO are strong.84 In fact, peroxynitrite in the RVLM has an excitotoxic effect.85

A recent study suggested that reactive oxygen species and reactive nitrogen species, such as peroxynitrite, could dose dependently regulate iNOS function and that peroxynitrite reduces both NO and superoxide productions via enzymatic iNOS dysfunction.86 Another report indicated that an interaction between NO and superoxide in the RVLM via the formation of peroxynitrite contributed to the hypotensive effect of NO after overexpression of eNOS in SHR.87 The role of peroxynitrite in the RVLM on the regulation of the sympathetic nervous system should be further examined.

The excitatory amino acid glutamate in the RVLM is known to cause strong sympathoexcitation.6, 7 We recently demonstrated that oxidative stress modulates the balance between glutamate and GABA in the RVLM of hypertensive rats.63 The results are consistent with previous reports, which demonstrated that NAD(P)H oxidase-derived superoxide in the RVLM was involved in the angiotensin II-induced pressor response via an enhancement of the presynaptic release of glutamate.72 We hypothesize that glutamate in the RVLM might be involved in oxidative stress-evoked sympathoexcitation.

We have also focused on the signal transduction associated with oxidative stress. The AT1 receptor activates caspase-3 in the RVLM through the Ras/mitogen-activated protein kinase/extracellular signal-regulated kinase, which is involved in sympathoexcitation in SHRSP animals.71 The activities of Ras, p38 mitogen-activated protein kinase, extracellular signal-regulated kinase and caspase-3 in the RVLM were elevated in SHRSP animals compared with those in normotensive rats. The phosphorylation of the pro-apoptotic proteins Bax and Bad, which release cytochrome c in the mitochondria, was shown to activate caspase-3.71 In contrast, the phosphorylation of the anti-apoptotic protein Bcl-2 inhibited caspase-3 activation. Intracerebroventricular infusion of a caspase-3 inhibitor reduced blood pressure and heart rate with sympathoinhibition in SHRSP animals but not in normotensive rats. Intracerebroventricular infusion of an AT1 receptor blocker caused sympathoinhibition and also reduced the activities of Ras, p38 mitogen-activated protein kinase, extracellular signal-regulated kinase and caspase-3 in the RVLM of SHRSP animals. These results are consistent with a previous report, which indicated that NAD(P)H oxidase derived superoxide and functioned in the activation of p38 mitogen-activated protein kinase or extracellular signal-regulated kinase1/2 by angiotensin II in the RVLM.72 Although the relationship between oxidative stress and activation of these kinases is bidirectional,88 these pathways could exist downstream of the AT1 receptor in the RVLM of SHRSP animals and could be related to the blood pressure elevation and sympathoexcitation observed in SHRSP animals.

Recently, additional central mechanisms of sympathoexcitation associated with oxidative stress were considered, such as perivascular macrophages in the brain,89, 90 transcription factor nuclear factor kappa-B91 and microglial cytokines92 in the brain. We also recently demonstrated that neuron-astrocyte uncoupling mediated by AT1 receptor-induced oxidative stress in the RVLM caused sympathoexcitation in SHRSP animals.93, 94

There are several unsolved issues. First, there may be multiple targets of oxidative stress in the RVLM. We have not clarified which types of cells, such as presympathetic neurons, interneurons or the axon terminals from NTS and paraventricular nucleus to the RVLM, are targets of oxidative stress in the RVLM. We demonstrated that oxidative stress triggered calcium influx in the neural cells; this mitochondrial calcium accumulation led to sympathoexcitation associated with mitochondrial oxidative stress production.70 These results were similar to a previous report.80 We also determined that oxidative stress in the RVLM enhanced glutamatergic excitatory inputs and attenuated GABAergic inhibitory inputs from the paraventricular nucleus into the RVLM.63 Interestingly, Chan et al.72 clearly demonstrated that NAD(P)H oxidase-derived superoxide in the RVLM was involved in the angiotensin II-induced pressor response via enhancement of the presynaptic release of glutamate to RVLM neurons. Our recent results indicate that oxidative stress-induced sympathoexcitation might be mediated by the apoptosis of astrocytes in the RVLM.71, 93 However, we have not conducted electrophysiological examinations to clarify the effect of oxidative stress on the electrophysiological characteristics of the RVLM neurons.

Several previous studies performed by Kumagai et al.95, 96, 97 indicated that the RVLM bulbospinal neurons in SHR depolarized and the firing rate increased by angiotensin II. Furthermore, angiotensin II produced a sustained inward current and increased the frequency and amplitude of the excitatory postsynaptic currents. These angiotensin II-evoked responses in the RVLM might be associated with oxidative stress, and angiotensin II-induced oxidative stress has the potential to reduce the expression of the voltage-gated potassium channel in the RVLM.98, 99 Moreover, the AT1 receptor enhanced the frequency of glutamate-sensitive spontaneous excitatory postsynaptic currents in the RVLM.72 Further studies to explore these unsolved questions are required.

Effects of AT1 receptor blockers on oxidative stress in the RVLM and sympathetic nervous system

The AT1 receptor in the brain has an important role in the pathophysiology of hypertension.7, 9, 73, 74, 79, 100, 101 Interestingly, peripherally administered AT1 receptor blockers can penetrate the blood–brain barrier and block the AT1 receptor within and outside the brain, although the extent of the blocking action within the brain varies among AT1 receptor blockers.58, 64, 102, 103, 104, 105 A high density of AT1 receptor is present in brain regions that are involved in the regulation of the sympathetic nervous system, such as the circumventricular organs outside of the blood–brain barrier that peripherally administered AT1 receptor blockers are able to access, and also regions inside the blood–brain barrier.101

Recent studies suggest that systemically administered AT1 receptor blockers also act on the AT1 receptor within the brain, thereby reducing blood pressure in hypertensive rats.22, 58, 64, 103, 106, 107, 108 The extent of the actions of AT1 receptor blockers within the brain might depend partly on the lipophilicity and pharmacokinetics of the blockers.102, 105 We have also examined the sympathoinhibitory effect via a reduction in oxidative stress through the inhibition of the AT1 receptor in the brain, and our results are consistent with many previous studies.64, 101, 102, 103, 104, 105, 106, 107 Orally administered telmisartan or olmesartan reduced blood pressure and urinary norepinephrine excretion in SHRSP animals and was associated with a reduction in oxidative stress production in the brainstem, including the RVLM.22, 108 Orally administered telmisartan has been shown to inhibit AT1 receptor-induced oxidative stress in the RVLM and cause sympathoinhibition in SHRSP animals to a greater extent than candesartan, in spite of similar depressor effects.58 Similar results were obtained in obesity-induced hypertensive rats treated with telmisartan or losartan.64 Thus, it is conceivable that orally administered AT1 receptor blockers could cause sympathoinhibition associated with the blockade of the AT1 receptor in the RVLM. Moreover, the sympathoinhibitory effects caused by orally administered AT1 receptor blockers may not be class effects.

Sympathoinhibition by targeting oxidative stress in the RVLM of hypertensive animals

3-Hydroxy-3-methylglutaryl coenzyme A reductase inhibitors (statins) are potent inhibitors of cholesterol biosynthesis. In addition, a previous study suggested that statins reduce blood pressure in patients with hypertension,109 and the potential sympathoinhibition of statins has been demonstrated.110, 111, 112 We have also demonstrated that orally administered atorvastatin causes sympathoinhibition and improves impaired baroreflex sensitivity via a reduction in oxidative stress through the inhibition of the AT1 receptor-NAD(P)H oxidase and the upregulation of MnSOD in the RVLM of SHRSP animals;54, 55, 111 these findings are consistent with a previous study of the vasculature.113 Moreover, we also determined that orally administered atorvastatin could increase NOS in the brainstem of SHRSP animals.114 From these results, we consider that statins have a potential to cause sympathoinhibition via a reduction in oxidative stress and/or an activation of NOS in the RVLM.

Several calcium channel blockers have been confirmed to cause sympathoinhibition via a reduction in oxidative stress in the RVLM of hypertensive rats. Orally administered amlodipine115 or azelnidipine61 caused sympathoinhibition via a reduction in oxidative stress in the RVLM of SHRSP animals. We confirmed that orally administered azelnidipine inhibited the NAD(P)H oxidase activity and activated MnSOD in the RVLM of SHRSP animals.61 Furthermore, a combination of atorvastatin and amlodipine60 or olmesartan and azelnidipine116 had additive effects of sympathoinhibition via a reduction in oxidative stress in the RVLM.

Interestingly, caloric restriction57 and exercise training59 have been shown to have a pivotal role in sympathoinhibition via a reduction in oxidative stress through the blockade of the AT1 receptor in hypertensive rats. These results suggest that adipocytokines and/or insulin resistance may affect the AT1 receptor in the RVLM and cause sympathoexcitation.

Perspectives and conclusions

The figure 1 shows our conceptualization of the regulation of the sympathetic nervous system via NO derived from overexpression of eNOS and/or AT1 receptor-induced oxidative stress in the RVLM. Regarding treatment for subnormal sympathoexcitation in cardiovascular diseases, our studies and other reports indicate that a blockade of the AT1 receptor in the RVLM may be necessary.64, 101, 102, 103, 104, 105, 106, 107, 108 AT1 receptor blockers are widely used in the treatments for hypertension.117 In addition, AT1 receptor blockers may have protective effects on neurons, thereby reducing the incidence of stroke and improving cognition.117, 118

Figure 1
figure 1

A schema showing our concept of the regulation of sympathetic nerve activity via NO derived from the overexpression of eNOS and AT1R-induced oxidative stress in the RVLM. GABA, γ-amino butyric acid; NAD(P)H oxidase, nicotinamide adenine dinucleotide phosphate.

Recent results from studies of renal nerve ablation in patients with resistant hypertension suggest that renal afferent nerves may also contribute to blood pressure elevation.119, 120, 121 Renal afferent nerves project directly into many areas in the central nervous system, such as the NTS and hypothalamus.122, 123 Oxidative stress was shown to mediate sympathoexcitation in the phenol renal injury model of hypertension in which the renal afferent nerves are stimulated.124 In this model, the AT1 receptor and NAD(P)H oxidase in the brain are activated. Previous studies have suggested that increased oxidative stress production and reduced nNOS expression may be involved in these mechanisms, and these changes lead to the alteration of cytokines in the brain.124, 125

We suggest that the AT1 receptor and related oxidative stress production in the RVLM should be therapeutic targets for the treatment of abnormal sympathoexcitation in hypertension. In conclusion, our studies suggest that the increase in AT1 receptor-induced oxidative stress and/or decrease in NO in the RVLM mainly cause sympathoexcitation in hypertension.