The bidirectional interaction between the sympathetic nervous system and immune mechanisms in the pathogenesis of hypertension

Abstract

Over the last few years, evidence has accumulated to suggest that hypertension is, at least in part, an immune‐mediated inflammatory disorder. Many links between immunity and hypertension have been established and provide a complex framework of mechanistic interactions contributing to the rise in BP. These include immune‐mediated inflammatory processes affecting regulatory brain nuclei and interactions with other mediators of cardiovascular regulation such as the sympathetic nervous system. Sympathoexcitation differentially regulates T‐cells based upon activation status of the immune cell as well as the resident organ. Exogenous and endogenous triggers activate signalling pathways in innate and adaptive immune cells resulting in pro‐inflammatory cytokine production and activation of T‐lymphocytes in the cardiovascular and renal regions, now considered major factors in the development of essential hypertension. The inflammatory cascade is sustained and exacerbated by the immune flow via the brain–bone marrow–spleen–gastrointestinal axis and thereby further aggravating immune‐mediated pathways resulting in a vicious cycle of established hypertension and target organ damage. This review summarizes the evidence and recent advances in linking immune‐mediated inflammation, sympathetic activation and their bidirectional interactions with the development of hypertension.

Linked Articles

This article is part of a themed section on Immune Targets in Hypertension. To view the other articles in this section visit http://onlinelibrary.wiley.com/doi/10.1111/bph.v176.12/issuetoc

Abbreviations

Ang II

angiotensin II

APCs

antigen‐presenting cells

CKD

chronic kidney disease

CVO

circumventricular organ

SHR

spontaneously hypertensive rat

TLRs

toll‐like receptors

TMAO

trimethylamine N‐oxide

Introduction

Hypertension is a major cause of mortality and morbidity worldwide affecting more than one billion adults. It is the most common modifiable risk factor for cardiovascular events and death. With an underlying cause established for only ~10% of cases (i.e. secondary hypertension), the majority of the hypertensive population suffers from essential hypertension, which is frequently characterized by sustained sympathetic activation. More recently, immune mechanisms have been proposed to contribute to the development of hypertension. Immune reactivity in experimental and human hypertension was initially believed to be a consequence, rather than a cause, of BP elevation. It has now become clear that immune‐mediated oxidative stress can initiate renal interstitial inflammation, loss of peritubular capillaries and medullary hypoxia, which in turn causes impaired pressure natriuresis and increased BP (Guyton et al., 1972; Franco et al., 2013). These findings laid the foundations for further investigations into the role of immunity‐induced pathways as a causal mechanism in the development of hypertension. Indeed, Franco et al. (2013) demonstrated a close relationship between renal immune cell infiltration and impaired pressure natriuresis, which was preventable by mycophenolate mofetil‐dependent reduction of immune infiltration in a salt‐sensitive hypertension rat model.

There has since been a shift from the perception of hypertension and hypertension‐mediated end‐organ damage merely being the result of elevated pressure on the vessel wall to a much more complex scenario involving immune mechanisms. Experimentally, immunological approaches for the treatment of BP have yielded promising results and provided evidence for a concerted interplay of innate and adaptive immunity. The objective of this review is to summarize the evidence and recent advances in linking immune‐mediated inflammation, sympathetic activation and their bidirectional interactions, with the development of hypertension.

Immunity and hypertension: compelling evidence for immune‐mediated hypertension

Experimental suppression of immune mechanisms is an elegant approach to explore their role in BP regulation. Indeed, immunosuppression or thymectomy blunted the development and maintenance of hypertension (White and Grollman, 1964; Okuda and Grollman, 1967; Harrison, 2014). BP and urinary excretion of inflammatory cytokines were decreased in hypertensive patients treated with mycophenolate mofetil for rheumatoid arthritis or psoriasis over a 3 month period, despite unchanged antihypertensive therapy or salt intake, but reversed when treatment was discontinued (Herrera et al., 2006). Human immunodeficiency virus‐positive patients with low CD4⁺ T‐cell counts were less prone to hypertension than those on antiretroviral therapy with normal CD4⁺ T‐cell counts, who developed hypertension to a similar degree as the normal population (Seaberg et al., 2005). Thymic transplantation from normotensive Wistar Kyoto rats reduced hypertension in spontaneously hypertensive rats (SHRs) (Ba et al., 1982). The requirement of a functional thymus for the maintenance of hypertension and the inability of thymectomized or athymic nude mice with renal infarction to maintain hypertension (Svendsen, 1976; Svendsen, 1977; Bataillard et al., 1986) support a pathogenic role of the immune system in hypertension. Moreover, mice with severe combined immunodeficiency were protected from developing hypertension (Crowley et al., 2010). The passive transfer of lymphocytes from hypertensive rat models to normotensive rats induced hypertension (Okuda and Grollman, 1967). Furthermore, hypertension was suppressed by anti‐thymocyte serum or immunosuppressive cyclophosphamide in SHR (Bendich et al., 1981; Dzielak, 1991), suggesting that inflammation is a cause rather than merely a consequence of hypertension.

In another important line of research, in mice with deletion of the recombinase activating gene 1 (Rag1^−/−), which lack T‐lymphocytes and B‐lymphocytes, hypertension failed to develop when they were chronically exposed to angiotensin II (Ang II) (Guzik et al., 2007). Rag1 deletion also attenuated hypertension in Dahl salt‐sensitive rats (Mattson et al., 2013) further supporting a role of immune cells in hypertension development. However, in a more recent publication, Ji et al. (2017) reported that these Rag1^−/− mice on a C57BL/6J background lost their resistance to Ang II‐induced hypertension, which was independent of T‐cells. Enhanced sensitivity and up‐regulation of the renal angiotensin AT₁ receptor activity was identified as the most likely explanation, and this was attributed to the universal drive for genetic variation in inbred animals. Clearly, this phenotypic change in the Rag1 knockout C57BL/6J mice model has implications for investigators using this strain to study mechanisms of T‐cell modulation of Ang II‐dependent BP control. However, it does not necessarily compromise the findings from studies investigating the role of T‐cell modulation in Ang II‐induced hypertension, prior to this occurrence. Furthermore, there is strong evidence that T‐cells modulate other important physiological processes mediated by Ang II such as endothelial dysfunction and microvascular injury (Mian et al., 2016).

T‐cells as critical immune players in the genesis of hypertension

Sympathoexcitation differentially regulates T‐lymphocytes, and it has been shown that sympathetic stimulation induces noradrenaline‐mediated T‐cell activation and vascular inflammation (Marvar et al., 2010). The activation of T‐lymphocytes involves two steps: (i) T‐cell receptor interaction with antigen presented by the major histocompatibility complex and (ii) co‐stimulation via CD28, bound by the B7 ligands, CD80 and CD86 of the antigen‐presenting cells (APCs) – the absence of which induces T‐cell apoptosis (Frauwirth and Thompson, 2002). Depletion of neutrophils and monocytes with diphtheria toxin in mice eliminates the hypertensive response to Ang II, and hypertension is restored with repletion of monocytes but not granulocytes, highlighting the important role of monocyte antigen presentation to T‐cells (Wenzel et al., 2011; Harrison, 2014). Inhibition of co‐stimulation by CTLA4‐Ig, a pharmacological agent that binds B7 ligands on APCs, attenuated the associated BP rise as well as T‐cell activation, vascular infiltration and T‐cell‐mediated TNF‐α and IFN‐γ production in both Ang II and deoxycorticosterone acetate (DOCA) hypertension models (Vinh et al., 2010). Hypertension as a T‐cell‐driven inflammatory process became evident when passive transfer of T‐cells and not B‐cells restored hypertension following Ang II infusion accompanied by elevated levels of memory and circulating T‐cell markers including CD69⁺, CCR5 ⁺ and CD44^high (Guzik et al., 2007). Furthermore, the role of T‐cells in the pathogenesis of hypertension has recently been confirmed in humanized mice, in which the murine immune system is replaced by the human immune system (Itani et al., 2016).

In all models of salt‐sensitive hypertension, the tubulointerstitium of rat kidney is intensively infiltrated with lymphocytes and macrophages, which correlates with the degree of hypertension (Rodríguez‐Iturbe et al., 2004; Heijnen et al., 2014). Lymphocyte infiltration was also seen in perivascular aortic tissue in Ang II‐induced hypertension, DOCA‐salt hypertension and aldosterone‐induced hypertension (Guzik et al., 2007; Kasal et al., 2012). Significant immune infiltration precedes hypertension and intensifies with increasing severity of hypertension, suggestive of a cause and effect relationship (Rodríguez‐Iturbe et al., 2004; Franco et al., 2013). Blocking of immune infiltration using mycophenolate mofetil prevents hypertension in various animal models, such as SHR (Rodríguez‐Iturbe et al., 2001), Dahl salt‐sensitive rats (De Miguel et al., 2011) and salt‐driven hypertension models induced by protein overload glomerular disease (Alvarez et al., 2002), and in the Page kidney model when the renal parenchyma is subjected to external compression (Vanegas et al., 2005).

A distinct infiltrating T‐cell phenotype, characterized by amplified production of the chemokine CCL2, promotes leukocyte recruitment and enhances ROS and generation of cytokines, thereby intensifying the local and advanced inflammation causing organ dysfunction and overt hypertension (Wei et al., 2014). Elevated CCL2 levels were found in organs regulating BP (Bush et al., 2000), and T‐cells were identified to be an important source of CCL2 contributing to inflammation associated with hypertension. Leukocytes express the CCL2 receptor, CCR2, which binds CCL2 and several other chemokines such as CCL8, CCL13 and CCL27. CCR2 inhibition reduces vascular macrophage accumulation and reverses the pressor responses in DOCA‐salt‐induced hypertension (Chan et al., 2012). CCR2‐deficient mice exhibit decreased macrophage and monocyte infiltration into the arterial wall during Ang II‐induced hypertension (Bush et al., 2000), suggesting a crucial role of CCR2–CCL2 signalling in immune‐mediated inflammation of hypertension and associated organ dysfunction, which may be initially driven by infiltrating T‐cells leading to a pro‐inflammatory cascade.

B‐cells as additional players in the genesis of hypertension

Whereas much evidence supports an important role for T‐cells, the role of B‐cells has been less well documented. Only quite recently did Chan et al. demonstrate that B‐cells and the associated IgG production contributed to Ang II‐induced hypertension. Interestingly, a clinically available anti‐CD20 antibody approach reduced hypertension, providing an interesting potential therapeutic perspective for hypertension (Chan et al., 2015).

The concerted interplay of adaptive and innate immunity in the development of hypertension

The T‐cell‐dominant adaptive immune response in hypertension, initiated and coordinated by the innate immune components such as phagocytes, dendritic cells, monocytes/macrophages, NK cells and the pattern recognition toll‐like receptors (TLRs), is critical to the induction and sustenance of hypertension. These events are mediated through the downstream immune components such as ROS and reactive nitrogen species (Ryan, 2013). The peripheral blood monocytes of hypertensive patients have elevated mRNA levels of TLR2 and TLR4, which are reduced by intensive antihypertensive treatment, indicative of a relevant role of the innate immune response (Hartupee et al., 2007).

Oxidative stress has been observed in all experimental models of hypertension. ROS act as second messengers in Ang II signalling and i.c.v. injections of adenovirus encoding for SOD‐ablated Ang II‐dependent changes in heart rate and BP (Zimmerman et al., 2002). Furthermore, SOD deletion in the circumventricular organ (CVO) increases baseline BP and augments the Ang II‐dependent hypertensive response and promotes T‐cell and leukocyte infiltration as well as vascular inflammation, partly by modulating sympathetic outflow. (Lob et al., 2010). With respect to ROS, a feedforward loop between NADPH oxidase (NOX) and mitochondria results in the induction and maintenance of diffuse inflammation in organs regulating BP (Datla and Griendling, 2010; Dikalov and Nazarewicz, 2013). This inflammatory milieu is characterized by elevated levels of Ang II, aldosterone, cytokines and altered mechanical forces, stimulated enzyme sources such as the NOXs, uncoupled NOS and the mitochondrial ROS, which contribute to the development of hypertension (Harrison and Gongora, 2009).

Several cytokines are implicated in the pathogenesis of hypertension. Inflammatory markers (Chae et al., 2001; Engstrom et al., 2002) and T‐cell‐specific cytokines such as IL‐4, IL‐6, IL‐7, IFN‐γ‐inducible protein‐10, TNF‐α and IL‐17 (Madhur et al., 2010; Stumpf et al., 2011) are elevated in patients with essential hypertension. IL‐17 infusion caused endothelial dysfunction and hypertension in mice (Nguyen et al., 2013) and also induced oxidative stress in the placenta of pregnant rats, thereby promoting hypertension (Dhillion et al., 2012). The TNF‐α antagonist etanercept effectively prevented hypertension in animal models (Tran et al., 2009; Venegas‐Pont et al., 2010), and IL‐6‐deficient mice appear to be protected in certain hypertension models (Lee et al., 2006; Schrader et al., 2007; Brands et al., 2010). Short‐term administration of pressor doses of Ang II resulted in renal infiltration and activation of immune cells with subsequent development of salt‐sensitive hypertension in male Sprague–Dawley rats (Rodríguez‐Iturbe et al., 2001). The Ang II‐induced vascular dysfunction and arterial hypertension were driven by IFN‐γ‐mediated immune recruitment and activation of NK cells and monocytes (Kossmann et al., 2013). Furthermore, IFN‐γ is up‐regulated in kidneys of hypertensive mice (Crowley et al., 2010), and deletion of IFN‐γ receptor (IFNGR) prevented Ang II‐dependent end‐organ damage (Marko et al., 2012).

The ongoing metabolic and oxidative cellular damage in hypertension can result in the oxidative modification of endogenous molecules, exposure of intramolecular sites by protein cleavage that are otherwise not exposed to the immune system and release reactive intracellular molecules further propagating the damage (Harrison et al., 2011). Highly reactive oxidatively modified isoketals accumulate in the dendritic cells and activate T‐cell‐mediated immuno‐inflammation and promote hypertension in mice (Kirabo et al., 2014). Certain molecules such as oxidized LDL, heat shock proteins (HSP) and platelet glycoproteins have been suggested to elicit immune responses in atherosclerosis (Gotsman et al., 2008). Abnormal HSP responses have been noted in patients with essential hypertension (Kunes et al., 1992; Pockley et al., 2002; Pockley et al., 2003; Li et al., 2009), and increased expression of HSP70 has been observed in circulating lymphocytes of hypertensive subjects (Kunes et al., 1992) that caused clonal expansion of CD4 ⁺ T‐cells and induced delayed hypersensitivity in models of salt‐sensitive hypertension (Pons et al., 2013). HSP70 gene polymorphisms, namely, HSPA1A, HSPA1B and HSPA1L, predispose to hypertension in individuals of Uygur ancestry (Li et al., 2009). Furthermore, auto‐antibodies against AT₁ receptors and α₁‐adrenoceptors have been detected in patients with hypertension (Liao et al., 2002) indicative of a role for protein modification. Agonistic antibodies directed to the second extracellular loop of AT₁ receptors, initially described in pre‐eclampsia, have also been detected in patients with essential hypertension (Fu et al., 2000; Wei et al., 2011). Circulating anti‐endothelial IgG and IgM has also been observed in hypertension (Frostegard et al., 1998; Papadopoulos et al., 2006). Taken together, the ‘first hit’ phase is likely to be initiated by cytokine‐dependent oxidative damage and protein modification resulting in the generation of neoantigens, which contribute to the initiation of the inflammatory cascade of hypertension (Figure 1).

Figure 1.

The ‘first hit pro‐inflammatory phase’ of pre‐ hypertension. Hypertensive (HTN) signals, damage‐associated molecular patterns (DAMPs) and cytokine‐dependent ROS signalling causes protein modification generating new neoantigens such as oxidized LDL, modified receptor proteins, HSPs and platelet glycoproteins that elicit an immunological response in the cardiovascular target tissues via the activation of T‐cells. This phase initiates the inflammatory cascade in organs associated with BP regulation. NAdr, noradrenaline.

Interaction of the sympathetic nervous system and immunity in the development of hypertension

Physiological integration and processing of the interaction between the autonomic nervous system and the immune system determines the balance of the sympathetic responses at the peripheral sites including the lymphoid targets. Neural and non‐neural communication pathways keep the central autonomic nervous system informed of the peripheral immune status. In addition, the micro‐environment created by diverse signalling mechanisms mediated through neurotransmitters and neuromodulators including adrenoceptors, immune cells, cytokines and pathogens maintains neural communication. Importantly, immune cells express adrenoceptors and nicotinic receptors (nAChRs), which bind the sympathetic neurotransmitters to initiate immunomodulatory responses. Of interest in this context is the observation that bilateral ablation of renal sympathetic nerves with phenol prevented immune activation and renal inflammation and attenuated Ang II‐induced hypertension in mice (Xiao et al., 2015). Furthermore, catheter‐based renal denervation applied to lower BP by reducing sympathetic nervous system activation has recently been demonstrated to reduce monocyte activation and inflammatory markers in patients with essential hypertension (Zaldivia et al., 2017).

Essential hypertension is a clinically heterogeneous entity, and current evidence supports the notion that immunomodulation is specifically relevant in regard to three major factors contributing to the BP rise including (i) increased sympathetic tone, (ii) enhanced sodium and fluid overload and (iii) activation of the renin–angiotensin–aldosterone (RAS) cascade. In cases involving high sympathetic activation, a low‐grade systemic inflammatory environment is established by the amplified release of inflammatory cells into the blood circulation (Heidt et al., 2014). The cells of the innate immune system, monocytes in particular, are implicated in the pathogenesis of hypertension. Acute and long‐term BP elevation is linked to pre‐activated monocytes in animals and patients with hypertension (Dörffel et al., 1994; Wenzel et al., 2011). In fact, genetic depletion of LysM⁺ monocytes resulted in blunted Ang II‐induced hypertension in mice, whereas adoptive transfer of pro‐inflammatory monocytes restored the BP elevation, supporting the notion of an important role of monocytes in hypertension development (Wenzel et al., 2011). Stimulation of α₁‐adrenoceptors has a pro‐inflammatory effect on immune system responses and positively regulates the expression of inflammatory cytokines in the innate immune cells during stress or injury‐dependent release of catecholamines (Grisanti et al., 2011). In addition, the release of pro‐inflammatory cytokines is dependent on activation of adrenoceptors in a number of chronic inflammatory disease states (Maestroni, 2000; Heijnen et al., 2002; Perez et al., 2009). For example, in chronic polyarticular juvenile rheumatoid arthritis, IL‐6 release from mononuclear cells in peripheral circulation is dependent on expression of α₁‐adrenoceptors (Heijnen et al., 1996), while activation of these receptors was important for enhanced inflammatory responses in an established model of multiple sclerosis (Brosnan et al., 1985).

Sympathetic activation of the bone marrow (BM) and the spleen and the associated release of haematopoietic cells into the systemic circulation has been identified to play an important role in the pathogenesis of hypertension (Heidt et al., 2014). Sympathetic nerves densely innervate the BM and upon activation can release progenitor cells seeding to the spleen with a subsequent increase in monocyte production (Felten et al., 1984; Felten et al., 1985; Dutta et al., 2012). Enhanced T‐cell infiltration and increased sympathetic ischarge have also been identified in the carotid body of hypertensive patients (McBryde et al., 2013). Exaggerated RAS activation (Rudemiller and Crowley, 2016), with T‐cells and monocytes/macrophages accumulation in the kidney and vasculature, mediates tissue injury and deleterious remodelling in hypertension (Madhur et al., 2010). Activation of the sympathetic nervous system differentially regulates the immune system based upon activation status of the immune cell as well as the resident organ (Case and Zimmerman, 2016). Release of noradrenaline from the sympathetic nerve terminal potentiates the effects of the activated T‐lymphocytes in these cardiovascular‐related organs. In noradrenaline‐infused mice, increased T‐lymphocyte numbers and activation have been observed in the aorta (Marvar et al., 2010), whereas T‐lymphocytes from the spleen in this model demonstrated decreased proliferation that was accompanied by a reduction in IFN‐γ and TNF‐α (Case and Zimmerman, 2015). While stimulation of Th1 cells generated from naïve CD4⁺ lymphocytes by noradrenaline enhanced IFN‐γ production by more than twofold (Swanson et al., 2001), noradrenaline also caused mitochondrial superoxide‐dependent modulation of the canonical activation of splenic T‐lymphocytes in a model of sympathetically‐driven hypertension (Case and Zimmerman, 2015; Case et al., 2016). Short‐term noradrenaline administration augmented the number of CD8⁺ circulating T‐lymphocytes, but long‐term administration resulted in decreased T‐lymphocyte numbers (Maisel and Michel, 1989). Taken together, the available evidence suggests that noradrenaline can elicit differential immunomodulatory effects on different subclasses and stages of differentiated T‐lymphocytes. Naïve unchallenged T‐lymphocytes are suppressed by noradrenaline stimulation, whereas the functionality of pre‐existing activated T‐lymphocytes is exacerbated under different conditions such as specific systemic infections, stress or targeted T‐lymphocyte differentiation (Alaniz et al., 1999).

On the contrary, alterations in sympathetic nervous system activity may be involved in stroke‐induced immunosuppression. In experimental stroke, both activation of the hypothalamic–pituitary axis and the sympathetic nervous system and β‐adrenoceptor blockade have been shown to substantially affect stroke‐induced immunomodulation (Prass et al., 2003). It has been suggested that brain injury‐induced generation of inflammatory cytokines such as TNF‐α, IL‐1β and IL‐6 interferes with the autonomic‐immune circuits to result in immune dysfunction (Dirnagl et al., 2007). Direct measurement of sympathetic activation in humans following a stroke using gold standard techniques such as microneurography or noradrenaline spillover (Lambert et al., 2008; Schlaich et al., 2010) has not been applied but are likely to shed some light into the relevance of these findings in a clinical setting. Overall, these findings are indicative of relevant crosstalk between the sympathetic and the immune system to sustain the immune‐mediated inflammation in hypertension (Figure 2).

Figure 2.

The sympathetic–immune crosstalk results in established hypertension. The brain–BM–spleen–gastrointestinal (GI) axis perpetuates and further exaggerates the immune cascade by causing immune infiltration and immune‐mediated inflammatory damage in the organs regulating BP, ultimately resulting in established hypertension. MHC, major histocompatibility complex; TCR, T‐cell receptor.

The immunomodulatory brain–bone marrow–spleen–gut axis: neuroinflammation is a characteristic feature of hypertension

The sympathetic nervous system is an integrative interface between the brain and the immune system and modulates both haemodynamic and immune responses (Elenkov et al., 2000). Immune infiltration and enhanced inflammation of neural regulatory centres have been suggested to be key players in the interaction between the immune and sympathetic nervous system as demonstrated in the nucleus tractus solitarii of the SHR (Gouraud et al., 2011). Similarly, in Sprague–Dawley rats, chronic Ang II infusion has been demonstrated to activate microglia and increase pro‐inflammatory cytokines in the paraventricular nucleus causing neurogenic hypertension. This effect was attenuated by minocycline, an antibiotic that passes the blood–brain barrier and has been shown to act as an inhibitor of microglia (Shi et al., 2010). Furthermore, i.c.v. infusion of Ang II increased sympathetic discharge of the splenic nerve and increased the expression of IL‐1, IL‐2, IL‐6, IL‐16 and TGF‐β1 in splenocytes, an effect that can be abolished by splenic sympathectomy, thereby linking the CNS to peripheral immune reactivation (Ganta et al., 2005). Anteroventral third ventricle lesions impair the activation of lymphocytes and their homing to the arterial walls and thereby ameliorate Ang II‐induced hypertension (Marvar et al., 2010). Indeed, lesions in this region appear to prevent any form of experimental hypertension (Brody et al., 1978). Chronic i.c.v. infusion of neuropeptides such as substance P and Ang II enhances T‐cell populations in experimental rodents (Fannon and Phillips, 1991).

The CVO lacks a blood–brain barrier making it susceptible to the influence of circulating hormones and cytokines. The CVO has abundant AT₁ receptors and NOX in the subfornical organ that produces ROS to enhance sympathetic outflow. SOD knockdown induces oxidative stress in the subfornical organ of the hypothalamus and increases vascular infiltration of activated lymphocytes, which is associated with worsening hypertension (Lob et al., 2010). Administration of a dominant negative Ras‐related C3 botulinum toxin substrate 1 (Rac1) has been shown to prevent NOX activation, highlighting the requirement of Rac1‐dependent Ang II effects in the brain (Zimmerman et al., 2004). Similarly, Ang II receptor knockout in the subfornical organ of the brain ameliorates the BP in DOCA‐salt hypertension (Hilzendeger et al., 2013). Suppressing the inflammation of the subfornical organ with minocycline or induction of IL‐10 overexpression ameliorates hypertension (Lob et al., 2013). Intracerebroventricular infusion of etanercept inhibits brain TNF‐α and protects against Ang II‐induced hypertension in rats by restoring the anti‐inflammatory balance and alleviating oxidative stress in the paraventricular nucleus (Venegas‐Pont et al., 2010; Sriramula et al., 2013). The elevation of circulating inflammatory markers characteristic of hypertension (Chae et al., 2001; Engstrom et al., 2002) may well be relevant for inflammatory processes in the CNS, thereby sustaining and further exaggerating the overall systemic inflammation (Wu et al., 2012).

The reciprocal impact: bone marrow‐derived inflammatory cells in hypertension

The bone‐marrow (BM) is extensively innervated by the sympathetic nervous system (Katayama et al., 2006), which regulates immunomodulation (Scheiermann et al., 2013), stem cell niche homeostasis and haematopoiesis (Hanoun et al., 2015). BM from hypertensive subjects is characterized by altered adrenergic signalling with exaggerated sympathetic tone, elevated levels of noradrenaline, loss of circadian rhythmicity and impaired mobilization of haematopoietic stem and progenitor cell populations (Afan et al., 1997; Katayama et al., 2006). The sympathetic overdrive of hypertension shifts the BM equilibrium towards the generation of myeloid progenitor cells, which induce peripheral inflammation and migrate to the brain to differentiate into brain macrophage/microglia to induce and sustain neuroinflammation (Santisteban et al., 2016). Cytokines, chemokines and ROS generated as a consequence of inflammation further accentuate the sympathetic overdrive centrally as well as in the peripheral organs, thereby further perpetuating hypertension. Impaired autonomic input to the BM has been associated with altered inflammatory responses in SHR (Zubcevic et al., 2014). The extravasation of inflammatory progenitors from the BM into the cardioregulatory centres of the brain results in microglial/macrophage activation to cause and sustain hypertension (Santisteban et al., 2015). The inflammatory progenitors from the BM migrate to the brain parenchyma in a CCL2 and CCR2‐dependent manner to exacerbate neuroinflammation (Ataka et al., 2013; Lampron et al., 2013) and infiltrate critical peripheral organs such as the heart and vasculature, carotid body, spleen, kidneys and gut to induce systemic inflammation (Rodríguez‐Iturbe et al., 2004; Gonzalez et al., 2015). It is pertinent to note that the CCL2–CCR2 chemokine system is a crucial player in the migration of T‐lymphocytes, monocytes and NK cells in rheumatoid arthritis and Type 2 diabetes (Madrigal and Caso, 2014). Moreover, Ang II directly stimulates CCL2 production in vascular smooth muscle cells (Chen et al., 1998), monocytes (Tsou et al., 2007) and hypothalamic neurons (Santisteban et al., 2013) via an AT₁ receptor‐mediated pathway (Koh et al., 2004; Dai et al., 2007; Marketou et al., 2011). However, the resting microglia of the adult brain express AT₁ receptors only when primed by pro‐hypertensive signals (Miyoshi et al., 2008).

The spleen: a neuroimmune hub differentially involved in cardiovascular and metabolic diseases

The spleen contains half of the body's monocyte population and mapping of the genetic, molecular and neurophysiological basis of hypertension‐induced inflammatory reflexes converge in the spleen (Rosas‐Ballina et al., 2008; Rosas‐Ballina et al., 2011). The spleen acts as a reservoir of pathogenic T‐cells, and activation of adrenergic splenic neurons regulates a T‐cell subset in the white pulp that eventually migrates to organs regulating BP. Moreover, the pro‐hypertension signals such as Ang II regulate haematopoietic stem cell proliferation (HSPCs) in the BM and enhance CCR2⁺ HSPCs, myeloid progenitors and inflammatory monocytes in the spleen, thereby contributing to hypertension (Kim et al., 2016). Hypertensive challenges activate splenic sympathetic discharge to prime the immune response, specifically the vagus–splenic drive mediated by nAChRs (particularly the α7nAChR), which links the brain and spleen (Rosas‐Ballina et al., 2008). The deficiency of α7nAChRs in apolipoprotein E knockout mice or splenic parasympathectomy increases the inflammatory profile characterized by elevated leukocyte counts and increased pro‐inflammatory cytokine expression within the spleen (Kooijman et al., 2015). The neural circuits that modulate the lymphocytes reveal that the action potential generated in the vagus nerve is transmitted to the celiac ganglion (the origin of the splenic nerve) to inhibit macrophage production of TNF‐α and other cytokines (Andersson and Tracey, 2012). Indeed, TNF‐α production is not inhibited by vagus nerve stimulation in splenectomized animals during endotoxemia (Huston et al., 2006), and i.v. administration of LPS increased efferent activity in the splenic and splanchnic sympathetic nerves in rats to induce a strong systemic inflammatory response (Martelli et al., 2014; Martelli et al., 2016), indicating an essential role for the spleen in mediating inflammatory signalling.

Ang II infusion increases the expression of tyrosine hydroxylase , the rate‐limiting enzyme in noradrenaline biosynthesis in adrenergic neurons and noradrenaline levels in the spleen. Splenectomy or splenic nerve ablation prevents the recruitment, activation and egression of pathogenic T‐cells from the spleen and consequent target organ infiltration (Carnevale et al., 2016). Splenic placental growth factor (PlGF)–sirtuin 1 (SIRT1) signalling mediated hypertension development in mice. Furthermore, splenic reimplantation from wild‐type donors restored the hypertensive response to Ang II, whereas mice re‐implanted with spleen from PlGF‐deficient spleen donors continued to be protected against the hypertensive effects of Ang II (Carnevale et al., 2014). Chronic administration of the selective SIRT1 inhibitor, Ex‐527 (selisistat), and genetic silencing of SIRT1 in PlGF‐deficient splenocytes restored the hypertensive response to Ang II (Carnevale et al., 2014). Cumulatively, these data suggest that the spleen is a bidirectional communication hub between the nervous system and immune system, and neural signals culminate in increased PlGF production and increased noradrenaline release resulting in the release of T‐cells and its downstream effectors, thereby contributing to hypertension and tissue damage. Therapeutic strategies targeting the sympathetic innervation of the spleen may be of interest.

The gastrointestinal connection

Recent findings suggest that gut‐related mechanisms may also be of relevance to the development of hypertension. Indeed, hypertension in rodents is accompanied by gut dysbiosis (Yang et al., 2015; Durgan et al., 2016; Santisteban et al., 2017), and a high‐fibre diet induced changes in the gut microbiota and prevented the development of hypertension and heart failure in hypertensive mice (Marques et al., 2017). Furthermore, faecal microbial transplantation from hypertensive patients to germ‐free mice resulted in increased BP, suggesting that alterations in gut microbiota may precede the onset of hypertension (Li et al., 2017). Significant changes in the gut microbiome characterized by decreased butyrate and acetate‐producing bacterial populations and altered Firmicutes/Bacteroidetes ratio have been linked with high BP. Treatment with minocycline not only reduces the neuroinflammation (Shi et al., 2010) but also rectifies the gut dysbiosis in both animal and human hypertension and expanded both acetate‐producing and butyrate‐producing bacteria and lowered BP (Yang et al., 2015) indicative of its potential in the treatment of hypertension. Reductions in the Lactobacillaceae and Prevotellaceae populations were identified in patients with chronic kidney disease (CKD), an important and common consequence of hypertension (Vaziri et al., 2013). Interestingly, patients on maintenance haemodialysis have about 100‐fold higher enterobacteria and enterococci levels than controls (Hida et al., 1996).

Sympathetic overdrive is associated with chronic gut inflammation, which in turn enhances gut permeability, dysbiosis and altered microbial metabolites in the systemic circulation. The HSPCs from the BM also migrate to the gut causing local inflammation and immune responses (Santisteban et al., 2016). Intrinsic gut mechanisms regulate the activation of T‐helper 17 cells, which was elevated in hypertension‐inducing marked gut microbial dysbiosis and an increase in circulating inflammatory cells (Kim et al., 2015). The changes in gut microbiota in hypertension were associated with changes in gut inflammatory status (Santisteban et al., 2017). Germ‐free mice were protected from Ang II‐induced arterial hypertension, vascular dysfunction and hypertension‐induced end‐organ damage. This protection was mediated by inhibition of recruitment and activation of the inflammatory myelomonocytic cells in the vasculature and altered cytokine signalling (Karbach et al., 2016). Short‐chain fatty acids (SCFAs) are metabolites of gut microbiota primarily composed of acetate, propionate and butyrate and have been implicated in host–microbiota immune‐inflammatory responses (Xiong et al., 2004; Samuel et al., 2008; Maslowski et al., 2009).

SCFAs directly activate the sympathetic nervous system and are not detectable in the plasma of germ‐free mice (Perry et al., 2016), indicating that all circulating SCFAs are microbial in origin, and any changes in SCFAs indicate an involvement of the gut microbiota. Recent studies analysing different SCFA receptor‐null mice such as those lacking the olfactory receptor 78 (Olfr78) or GPR41 (FFA3 receptor), expressed in renal vasculature, suggest a role of SCFAs in BP regulation (Pluznick, 2017). Olfr78‐null mice have lowered plasma renin levels and lowered baseline BP consistent with its localization in the site of renin storage and secretion (renal afferent arteriole) and in vascular smooth muscle cells of large vessels. GPR41‐null animals display altered vascular tone, functionally stiffer vessels and isolated systolic hypertension (Natarajan et al., 2016). Surprisingly, transplantation of faecal contents from salt‐resistant (R) to salt‐sensitive (S) Dahl rats resulted in consistent and significant elevation in BP during the remaining lifespan of the recipient rat and also shortened lifespan. The genome of R rats was resistant to BP alterations when microbiota was transplanted from the S to the R type. These effects may be mediated by SCFAs as both acetate and heptanoate levels were higher in the R to S transfer group (Mell et al., 2015). Population‐based studies have found that interventions that alter production of SCFAs correlate with changes in BP (Khalesi et al., 2014) indicative of the potential roles of SCFAs in influencing phenotypic changes to regulate BP.

Yet another microbial metabolite that has drawn significant attention is trimethylamine derived from dietary phosphatidylcholine or l‐carnitine, metabolized in the liver following absorption to form trimethylamine N‐oxide (TMAO) and excreted by the kidneys (Wang et al., 2011; Tang et al., 2013). Accumulation of circulating TMAO results in adverse effects such as macrophage/foam cell activation, platelet hyperresponsiveness and vascular and endothelial cell dysfunction and leads to atherosclerosis, thrombosis and cardiorenal impairment (Wang et al., 2011) as well as prolonging the hypertensive effect of Ang II (Ufnal et al., 2014). Also, TMAO concentrations in plasma correlated positively with cardiovascular disease in humans (Koeth et al., 2013; Wang et al., 2015). Moreover, plasma TMAO levels (r = 0.43) were inversely related to vascular endothelial function as assessed by brachial artery flow‐mediated dilation, whereas they were positively correlated with systolic arterial levels (r = 0.47; R. A. Gioscia‐Ryan and D. R. Seals, unpubl. data, Raizada et al., 2017). TMAO levels were high in CKD patients and directly contributed to progressive renal fibrosis and dysfunction in animal models (Tang et al., 2015). Taken together, these studies indicate that (i) the gut microbiome may have a different interaction with the immune system in the prehypertensive and hypertensive states compared with the normotensive state and (ii) dysbiosis of the gut microbiota increases sympathetic activity, and the consequent immune responses are likely to contribute to the chronic systemic inflammation associated with hypertension.

Effect of sympathoinhibitory therapy on immune‐mediated inflammation in hypertension

In addition to its well‐described effects on cardiovascular control, data from laboratory and clinical studies provide substantial evidence for the role of sympathetic nervous system‐mediated immune mechanisms in the pathogenesis of hypertension. Inhibition of sympathetic overdrive in hypertension can be achieved by a number of strategies including weight reduction and exercise, pharmacotherapy and device‐based therapeutic approaches. Disorders such as obesity and the metabolic syndrome are associated with increased sympathetic nerve activity and circulating markers of oxidative stress and inflammation, which can be markedly improved with exercise and weight loss (Festa et al., 2000; Schlaich et al., 2015). Exercise training down‐regulates TLR signalling, attenuates oxidative stress and sympathetic activation and prevents the increase in weight, arterial pressure and white adipose tissue accumulation in rats fed with a high‐fat diet (Li et al., 2015). Pharmacological sympathetic inhibition can be achieved by inhibiting sympathetic outflow at the level of the rostral ventral medulla by means of imidazoline I₁ receptor agonists such as moxonidine, which has been demonstrated to exert anti‐inflammatory properties in postmenopausal hypertensive women (Pöyhönen‐Alho et al., 2008) in addition to its beneficial effects on BP and metabolic parameters (Chazova et al., 2006). Another relevant relationship in the current context is the sympathetic nervous system‐induced increase in expression of the sodium/glucose co‐transporter‐2 (SGLT2). Interestingly, SGLT2 inhibition with empagliflozin reduced glucotoxicity‐induced inflammation and fibrosis in human proximal tubular cells (Panchapakesan et al., 2013; Matthews et al., 2017). Interventional approaches to target sympathetic overactivity include device‐based therapeutic strategies such as renal denervation and carotid body ablation. Both carotid body deafferentation and renal denervation reduced T‐cell infiltration in brain tissue and aorta of SHRs (McBryde et al., 2013). The beneficial effects of catheter‐based renal sympathetic denervation applied in patients with resistant hypertension appear at least in part to be mediated by immunomodulation via inhibition of monocyte activation and inflammation (Zaldivia et al., 2017).

Overall, these findings suggest that targeting sympathetic overactivity as a therapeutic principle affects the immune system on various levels, which may at least in part account for the beneficial cardiovascular and metabolic effects observed with these approaches.

Conclusion

The sympathetic nervous system plays a vital role in both activating and limiting immune‐inflammatory responses. This, in turn, encourages thinking that fine tuning of the immune‐inflammatory mechanisms responsible for enhanced sympathetic activity may be useful in the management of hypertension. However, to develop successful therapeutic interventions, an even better understanding and more research focussing on this link is warranted. Another important issue is to determine how modulation of sympathetic nerve activity could be exploited therapeutically to beneficially influence immune‐inflammatory responses. Finally, understanding the immune flow via the brain–BM–spleen–gut axis that perpetuates the immune cascade including immune infiltration and immune‐mediated inflammatory damage in the organs regulating BP will be important. New technologies for manipulating the sympathetic nervous system are already being investigated in clinical trials. Investigating the full consequences of the new modalities used in the treatment of resistant hypertension, such as renal denervation and carotid body ablation, on the immune system is likely to yield relevant insights and is currently being explored.

Nomenclature of targets and ligands

Key protein targets and ligands in this article are hyperlinked to corresponding entries in http://www.guidetopharmacology.org, the common portal for data from the IUPHAR/BPS Guide to PHARMACOLOGY (Harding et al., 2018), and are permanently archived in the Concise Guide to PHARMACOLOGY 2017/18 (Alexander et al., 2017a, 2017b, 2017c, 2017d, 2017e).

Conflict of interest

The authors declare no conflicts of interest.

Carnagarin R., Matthews V., Zaldivia M. T. K., Peter K., and Schlaich M. P. (2019) The bidirectional interaction between the sympathetic nervous system and immune mechanisms in the pathogenesis of hypertension, British Journal of Pharmacology, 176, 1839–1852, doi: 10.1111/bph.14481.