Recent advances in central cardiovascular control: sex, ROS, gas and inflammation

The central nervous system (CNS) in concert with the heart and vasculature is essential to maintaining cardiovascular (CV) homeostasis. In recent years, our understanding of CNS control of blood pressure regulation (and dysregulation leading to hypertension) has evolved substantially to include (i) the actions of signaling molecules that are not classically viewed as CV signaling molecules, some of which exert effects at CNS targets in a non-traditional manner, and (ii) CNS locations not traditionally viewed as central autonomic cardiovascular centers. This review summarizes recent work implicating immune signals and reproductive hormones, as well as gasotransmitters and reactive oxygen species in the pathogenesis of hypertension at traditional CV control centers. Additionally, recent work implicating non-conventional CNS structures in CV regulation is discussed.


Introduction
According to the World Health Organization (WHO), cardiovascular (CV) disease accounts for approximately 17 million deaths a year worldwide 1 , of which more than half (9.4 million) are attributable to complications of hypertension 2 . In 2008, a staggering 40% of adults over the age of 25 had been diagnosed with hypertension 3 . The central nervous system (CNS) is essential to maintaining CV homeostasis. Traditional central autonomic CV control centers include the nucleus tractus solitarius (NTS), the rostral ventral lateral medulla (RVLM), and the caudal ventral lateral medulla in the brainstem; the parabrachial nucleus in the pons; and the paraventricular nucleus (PVN) in the hypothalamus. In addition, the area postrema (AP) in the hindbrain, and the organum vasculosum of the lamina terminalis (OVLT) and subfornical organ (SFO) in the forebrain, are sensory circumventricular organs (CVOs) characterized by the presence of a wide variety of receptors and the lack of the normal blood-brain barrier, which have also been implicated in central CV regulation. The renin-angiotensin aldosterone system (RAAS) has also been extensively implicated as a critical signaling system, components of which play central roles both as circulating hormones and as CNS neurotransmitters in the regulation of blood pressure (BP). There is growing evidence that the development and progression of hypertension involves dysregulation of the sympathetic nervous system (SNS) (SNS over-activity) (for review, see 4-6) and activation of the RAAS 7,8 .
Over the past 20 years, our understanding of CNS control of BP regulation (and dysregulation leading to hypertension) has evolved substantially. This review will summarize some of these paradigm shifts, focussing primarily on signaling molecules that either (i) are not classically viewed as CV signaling molecules (i.e. immune signals and reproductive hormones) or (ii) exert effects at CNS targets in a non-traditional manner, acting via membrane receptor-independent signaling mechanisms (i.e. gasotransmitters and reactive oxygen species [ROS]), all of which have been shown to have profound effects on the central control of BP. CNS structures, not conventionally thought of as CV control centers but that more recently have been shown to influence CV regulation, are also discussed.

Inflammation and immune regulators as modulators of cardiovascular regulation and contributors to hypertension
Although it had been speculated decades ago that there was a relationship between the immune system and hypertension 9 , the demonstration of systemic markers of inflammation in patients with essential hypertension in the early 2000s 10,11 was a catalyst for renewed interest in the relationship between hypertension and the immune system. Emerging evidence suggests that both the innate and acquired immune systems are activated in hypertension, as inflammations in the kidney, vasculature (arteries), and CNS have all been shown to be involved in the pathogenesis of hypertension.
As an immediate first-line defence mechanism to infections or tissue injury, the innate immune system initiates a generalized inflammatory response involving dendritic cells, macrophages, natural killer (NK) T cells, and Toll-like receptors (TLRs), all of which have been shown to be activated in hypertension.
Dendritic cell activation has been shown to promote hypertension by stimulating T-cell proliferation which infiltrates both the kidney and arterial walls 12,13 . Similarly, macrophage infiltration of the kidney and arteries has been documented in experimental models of hypertension, and a decrease in macrophage infiltration is associated with an improvement of hypertension in these models of hypertension [14][15][16][17] . Recently, NK T-cell activation and TLRs (TLR4, in particular) have been suggested to play a role in hypertensionrelated inflammation 18,19 . The adaptive immune system responds to specific antigens and involves antigen presentation, lymphocyte activation, and antibody production. T cells have been shown to play a role in angiotensin II (ANG II)-induced hypertension 20 whereas endogenously produced ANG II increases T-cell activation 21 . Pro-inflammatory T-cell activation and the subsequent release of pro-inflammatory cytokines are associated with hypertension 22-24 whereas inhibition or genetic ablation of the B7/CD28 T cell costimulatory pathway has been shown to prevent experimental hypertension 12 . RAG-1 −/− mice and SCID mice, which lack both T and B cells, exhibited a blunted hypertensive response to ANG II infusion 20,25 , a response that returned when T cells were transferred into RAG-1 −/− mice 20 . T cell-produced cytokines (such as tumor necrosis factor alpha, or TNFα) and many of the interleukins (such as IL-6) have been shown to play a role in hypertension. TNFα antagonism 20,26 or genetic knockout of IL-6 27 has been shown to blunt ANG II-induced hypertension. The presence of agonist antibodies to ANG II receptors has been identified in a number of conditions that are characterized by elevated BP, such as preeclampsia 28,29 , refractory hypertension 30,31 , and malignant hypertension 32 .
Many studies have suggested that arterial inflammation within specific CNS locations is involved in the pathogenesis of hypertension. A role for inflammation in the NTS, a pivotal region for regulating arterial pressure baroreceptor reflex sensitivity, has been suggested in the development of hypertension, as studies have shown not only leukocyte accumulation within the NTS microvasculature 33 but also changes in gene expression of a variety of inflammatory molecules 34,35 and neurotrophic factors 36 in the NTS of spontaneously hypertensive rats (SHRs).
In addition, many of the cytokines, released as a consequence of immune system activation, have been shown to directly influence cardiovascular control centers in the CNS. Microinjection of IL-6 into the NTS attenuates baroreceptor function 37 and leads to speculation that abnormal gene expression of IL-6 in the NTS may be associated with hypertension. Augmentation of IL-1β, IL-6, or TNF-α expression and increased ROS observed in the RVLM following chronic intraperitoneal lipopolysaccharide administration have been suggested to be contributing factors to neurogenic hypertension induced by systemic inflammation 38 .
Early studies identified the anteroventral third ventricle (AV3V), a broad-based region located along the wall of the third ventricle which includes the OVLT, as a critical CNS structure in the pathogenesis of hypertension 39 . A more recent study not only confirmed that lesions of the AV3V region attenuate ANG II-induced hypertension but also implicated immune system involvement as AV3V lesions eliminated circulating T-cell activation and vascular infiltration normally observed in response to ANG II administration 40 . IL-1β has been shown to influence the excitability of SFO neurons 41 , and recent studies have demonstrated that microinjection of IL-1β (and of TNFα) into SFO increases BP and renal sympathetic nerve activity (SNA) 42 .
The PVN, a hypothalamic autonomic control center with welldocumented roles in CV regulation, has been implicated as a CNS structure in which immune signals may act to cause hypertension. Chronic ANG II infusion causes the expression of proinflammatory cytokines and markers of oxidative stress in the PVN, effects blocked by central administration of TNFα blocker 26 . Angiotensin-converting enzyme 2 (ACE2) overexpression in the PVN has also been shown to attenuate both ANG II-induced hypertension and expression of the pro-inflammatory cytokines TNFα, IL-1β, and IL-6 in the PVN 43 . Blockade of nuclear factorkappa-B (NFκB), a prominent transcription factor that governs inflammatory responses, in the PVN of rats resulted in decreased BP, pro-inflammatory cytokines, and ROS, as well as upregulation of key protective anti-hypertensive RAAS components, suggesting an important role for NFκB in PVN in the hypertensive response 44 . Finally, rats fed a high-salt diet demonstrated increased expression of IL-1β and decreased expression of the anti-inflammatory cytokine IL-10, in the PVN. These expression levels were augmented by stimulation of ROS production within the PVN 45 .

Reproductive hormones and cardiovascular regulation
The interest in the role of sex hormones in hypertension has been driven by a number of observations regarding sexual dimorphism in BP regulation in humans and animals. Epidemiological findings that prior to menopause the prevalence of essential hypertension is lower in women than in men of the same age 46 and that young women have lower resting SNA than men 47 , differences that disappear after menopause, suggest that estradiol is important in BP regulation and, in fact, may protect against hypertension. Findings that estradiol administration attenuates increases in BP normally exhibited by intact males and ovariectomized females, and prevents development of hypertension in experimental models of hypertension 48,49 , suggest a role for estradiol in the regulation of BP.
Studies in humans and animals suggest that exogenous testosterone may also play a crucial role in BP regulation. In humans, low testosterone levels have been correlated with higher BP 50,51 whereas testosterone replacement has been shown to cause significant reductions in BP 52,53 , suggesting a role for testosterone in BP regulation. Moreover, in experimental models of hypertension high BP develops more rapidly and becomes more severe in the male than in the female, effects which were shown to be androgen-dependent 48,54,55 . Further support for a role of testosterone in the etiology of hypertension is derived from studies showing that castration prevents the development of hypertension in SHR rats 56 .
Evidence for a role for central actions of estradiol on BP regulation is derived from a variety of sources. Firstly, many of the CNS sites with well-documented roles in CV regulation have been shown to possess estrogen receptors (ERα and ERβ) 57-61 . Moreover, intracerebroventicular (icv) administration of estradiol in ovariectomized mice and in male mice attenuated the increase in BP normally elicited by ANG II 62 . In rats, aldosterone/salt-induced hypertension is exhibited by intact males and ovariectomized females, effects attenuated by activation of central ER receptors. Central ER blockade 63 or icv injections of small interfering RNA-ERα (siRNA-ERα) or siRNA-ERβ 64 , on the other hand, augmented aldosterone-induced hypertension in intact females.
Further to these findings, estradiol has been shown to act via ERα or ERβ (or both) at specific brain regions in both males and females to influence sympathetic outflow and baroreflex function. The AP and SFO predominantly express ERα 57-62 , and estradiol has been shown to decrease the activity of AP 65 and SFO neurons 66 , and inhibits ANG II activation of AP 67 and SFO neurons 66 , whereas genetic knockdown of ERα in the SFO enhances ANG II-induced hypertension in female mice 68 .
Estrogen actions at ERβ in PVN inhibit hypertensive effects of glutamate activation 69 . In the RVLM, estradiol actions at ERβ receptors have been shown to cause decreases in BP in normotensive rats 70 and to attenuate aldosterone-induced increases in SNA and BP 64 whereas ERβ knockdown in RVLM or PVN results in the augmentation of aldosterone-induced increases in SNA and BP 64 , effects that are not seen in intact females 64 .
Relaxin, a member of the insulin family best known for its role in pregnancy, has also been shown to influence BP. Early studies revealed that chronic intravenous (iv) administration of relaxin elicited a decrease in BP in SHRs 71 . Relaxin binding sites and relaxin receptors have been shown to be widely distributed throughout the brain, including the SFO, NTS, and PVN 72 , suggesting that relaxin may be involved in the central control of BP. Hypertensive effects of central administration of relaxin into the dorsal third ventricle are totally abolished by lesions of the SFO 73 , identifying this CVO as one central target mediating these cardiovascular effects. A recent study demonstrating that acute microinjection of relaxin-2 into the PVN increased sympathetic outflow and BP in SHR, whereas chronic PVN administration caused a profound increase in BP in normotensive rats 74 , supports the conclusion that there are multiple central targets for this reproductive hormone/neurotransmitter. Moreover, this same study revealed that neutralization of endogenous relaxin reduced BP in SHR but had no significant effect in WKY 74 , suggesting a role for relaxin in the pathogenesis of hypertension.
Another reproductive peptide that warrants further investigation into its potential contribution to the pathogenesis of hypertension is prolactin, a hormone best known for its involvement in lactation and reproduction. Very few studies have investigated the role of prolactin in the central control of CV regulation despite epidemiological evidence suggesting correlations between circulating prolactin levels and increased BP. Plasma prolactin has been shown to be elevated in patients with essential hypertension 75 and preeclampsia 76,77 . Furthermore, higher plasma prolactin levels have been shown to be associated with increased risk of hypertension in menopausal 78 and post-menopausal 79 women and in preeclampsia 54 . Prolactin receptors are widely distributed throughout the body 80 and brain 81 . mRNA for the prolactin receptor has been reported in the PVN 81,82 , and we have identified the presence of the prolactin receptor at levels similar to the AT1 receptor in the SFO 83 . However, to our knowledge, studies investigating the CV consequences of central administration of prolactin (icv or microinjection into discrete brain nuclei) on BP, or the effects of prolactin on neuronal excitability in central CV control centers, are lacking.

Gasotransmitters and cardiovascular regulation: hydrogen sulfide
Gasotransmitters are endogenously produced membrane permeable gas molecules which act at specific, targeted cells via membrane receptor-independent signaling mechanisms to exert well-defined physiological effects. The action(s) of nitric oxide (NO) and carbon monoxide (CO) at peripheral tissues and in the CNS to influence cardiovascular regulation are well documented 84,85 . More recently, a third gasotransmitter, hydrogen sulfide (H 2 S), an environmental air pollutant with well-known deleterious health effects, has been identified and suggested to play a role in the pathogenesis of hypertension. H 2 S is endogenously produced from catalysis of L-cysteine by using four enzymes: cystathionine β-synthase (CBS), cystathionine γ-lysase (CSE), or 3-mercaptopyruvate sulfur transferase (3MST) in tandem with cysteine aminotransferase (CAT). CBS is highly expressed in the CNS where it produces H 2 S from L-cysteine 86 , whereas CSE is the predominant enzyme expressed in the myocardium and vasculature smooth muscle cells 87 . Though predominantly found in the mitochondria where they work in tandem to produce H 2 S, 3MST and CAT are also expressed in the brain and vascular endothelium 88 . In addition, H 2 S can be produced in red blood cells by the conversion of polysulfides which are obtained from dietary sources 89 .
Evidence for a role of H 2 S in the pathogenesis of hypertension is suggested by the observation that plasma H 2 S concentrations are lower in patients with grade 2 or grade 3 hypertension, portal hypertension, and pulmonary hypertension 90-92 and in preeclampsia where plasma H 2 S levels and placental CBS mRNA expression are decreased 93,94 . H 2 S has been shown to be endogenously produced in peripheral vascular tissues and has been demonstrated to be a potent vasodilator, causing vasorelaxation in mesenteric arteries 95 , aortic rings 96,97 , the ductus arteriosis 96 , and pulmonary arteries 98 via actions on vascular smooth muscle cells. Unlike its gasotransmitter counterparts, NO and CO, vascular smooth muscle relaxation occurs independently of cGMP pathway activation. Activations of Ca 2+ -activated potassium channels (BKCa) 99 , ATP-sensitive potassium channels (K ATP ) 100 , Kv7 voltage-gated potassium channels 97 , and cytochrome P-450 2C (Cyp2C) 99 have all been implicated as mechanisms of the H 2 S vasorelaxation.
A bolus iv injection of H 2 S elicited an immediate depressor response in normotensive rats 100 whereas chronic intraperitoneal administration of H 2 S decreases BP in hypertensive rats 101-104 . These findings, along with the fact that mice lacking CSE exhibit hypertension and reduced endothelium-dependent vasorelaxation 105 , provide evidence of a direct role for H 2 S in BP regulation.
A role for H 2 S in the central control of BP stems from studies demonstrating that icv administration of H 2 S has been shown to dosedependently decrease BP, effects which are followed by potent long-lasting hypertension actions attributed to modulation of H 2 S on K ATP channels and α adrenergic stimulation, respectively 106 . Furthermore, microinjection of H 2 S into discrete brain nuclei known for their involvement in CV regulation has also been shown to affect BP. H 2 S administration into the RVLM elicits decreases in BP, effects again mediated by K ATP channels 107 , whereas similar microinjections into the PVN 108 and SFO 109 have been shown to dose-dependently increase BP. Moreover, H 2 S has been shown to influence the excitability of neurons in the NTS 110 , PVN 111 , and SFO 109 , CNS areas involved in CV regulation.

Reactive oxygen species and cardiovascular control
When produced at appropriate concentrations, ROS have been implicated in the regulation of many critical physiological processes, including cell signaling, maintenance of appropriate vascular tone, inflammation, and immune responses. ROS overproduction, on the other hand, is a feature common to a number of pathological conditions, including hypertension.
A role for ROS in hypertension is suggested in humans as a positive correlation between BP and biomarkers of oxidative stress in patients with essential hypertension has been reported 112,113 . Furthermore, mice lacking nicotinamide adenine dinucleotide phosphate (NADPH) oxidases, key enzymes in the production of ROS, are protected against experimental hypertension 114 , whereas overexpression potentiates ANG II-induced hypertension 115 .
ROS production in specific CNS cardiovascular control centers, including both brain stem (NTS, RVLM) and hypothalamic (PVN) nuclei, and within the CVOs (SFO) has been shown to play a role in neurogenic hypertension [116][117][118] . Superoxide dismutase (SOD), an enzyme that metabolizes superoxide, overexpression in the brain abolished the hypertensive response normally observed in response to icv ANG II administration 119 , whereas specific SOD3 deletion in the SFO increased baseline BP and potentiated ANG II-induced increases in BP 120 . Interestingly, this same study showed that ROS in the SFO leads to infiltration by activated lymphocytes in the peripheral vasculature 120 , linking oxidative stress in the CNS with immune activation in the periphery, which in concert would serve to intensify hypertension.
A high-salt diet increases NADPH oxidase (NOX-2 and NOX-4) expression in the PVN, whereas microinjection of amino-triazole (ATZ), a catalase inhibitor which increases ROS, into the PVN augments renovascular hypertension as well as increasing BP in normal rats 45 .

A role for 'other' central nervous system structures in the central control of blood pressure
This review has focussed on actions of non-traditional CV signaling molecules at CNS structures with well-documented roles in CV regulation. Another emerging area that warrants mention is the role of CNS regions not classically viewed as CV control centers that have been suggested to play a role in the pathogenesis of hypertension, secondarily or as a co-morbidity to other disease states. For example, the explosion of obesity research further to the discovery of leptin in the 1990s 121 has highlighted the involvement of a number of CNS autonomic control centers not typically viewed as CV control centers, such as the arcuate nucleus and the anterior hypothalamus, in the pathogenesis of hypertension as a consequence of direct actions of metabolic signals in these areas (for review, see 122,123). Furthermore, many metabolic signals associated with obesity have been demonstrated to influence BP regulation via actions at the 'classical' CNS CV control centers. Further study of the actions of traditional CV signals (such as ANG II) within these non-traditional CV CNS centers may elucidate previously unknown roles of these regions in normal CV regulation.

Conclusions
In this brief review, we have highlighted some emerging new perspectives which over the past 20 years contributed new and important information to the evolution of our understanding of CNS mechanisms involved in central CV control. The areas we have chosen to discuss are far from an exhaustive list of what is new and interesting, but do emphasize that this is a continually developing area of research with an inherent complexity associated with the requirement for integration of diverse autonomic systems. This points us in the direction of understanding that we perhaps should not expect to consider either single brain areas or single signalling molecules as "cardiovascular" at the expense of also describing their roles in other systems. Such conclusions point us to the broader perspective that all of these brain areas, signaling molecules, and autonomic systems contribute to the complex homeostatic regulation which maintains our "milieu interior" in a state of optimal health.

Open Peer Review Current Referee Status:
Editorial Note on the Review Process are commissioned from members of the prestigious and are edited as a F1000 Faculty Reviews F1000 Faculty service to readers. In order to make these reviews as comprehensive and accessible as possible, the referees provide input before publication and only the final, revised version is published. The referees who approved the final version are listed with their names and affiliations but without their reports on earlier versions (any comments will already have been addressed in the published version).