Cardiovascular Brain Circuits

Abstract

The cardiovascular system is hardwired to the brain via multilayered afferent and efferent polysynaptic axonal connections. Two major anatomically and functionally distinct though closely interacting subcircuits within the cardiovascular system have recently been defined: The artery-brain circuit and the heart-brain circuit. However, how the nervous system impacts cardiovascular disease progression remains poorly understood. Here, we review recent findings on the anatomy, structures, and inner workings of the lesser-known artery-brain circuit and the better-established heart-brain circuit. We explore the evidence that signals from arteries or the heart form a systemic and finely tuned cardiovascular brain circuit: afferent inputs originating in the arterial tree or the heart are conveyed to distinct sensory neurons in the brain. There, primary integration centers act as hubs that receive and integrate artery-brain circuit-derived and heart-brain circuit-derived signals and process them together with axonal connections and humoral cues from distant brain regions. To conclude the cardiovascular brain circuit, integration centers transmit the constantly modified signals to efferent neurons which transfer them back to the cardiovascular system. Importantly, primary integration centers are wired to and receive information from secondary brain centers that control a wide variety of brain traits encoded in engrams including immune memory, stress regulating hormone release, pain, reward, emotions, and even motivated types of behavior. Finally, we explore the important possibility that brain effector neurons in the cardiovascular brain circuit network connect efferent signals to other peripheral organs including the immune system, the gut, the liver, and adipose tissue. The enormous recent progress vis-à-vis the cardiovascular brain circuit allows us to propose a novel neurobiology-centered cardiovascular disease hypothesis that we term the neuroimmune cardiovascular circuit hypothesis.

Neural brain circuits can be broadly defined as hardwired polysynaptic axonal connections from peripheral tissues to the brain and back to peripheral tissues. However, defining neural circuits at the structural and functional levels turns out to be one of the most humbling but also rewarding challenges in current medical research. Delineating circuits that involve the brain requires molecular, structural and functional information of the interactions and connections of peripheral and central nervous system (PNS, CNS) neurons with - in principle - all peripheral tissues and indeed most organ cells. The human brain as the center of both interoception (perception from bodily signals) and exteroception (perception from the environment) is certainly the most complex tissue that nature has ever coined^1–35. Indeed, if challenges can be overcome to make steps towards a better understanding of the brain, groundbreaking advances into human development and conceivably the pathogeneses of a large number of clinically important cardiovascular diseases (CVDs) may be within reach³⁶. Defining control of physiological responses such as stress-related circuits have already uncovered aberrant interactions between the brain and peripheral organ dysfunction in diseases as diverse as cancer³⁷, obesity²⁴, metabolism and diabetes mellitus³⁸, chronic lung diseases^{39, 40}, inflammatory bowel diseases^{20, 41}, autoimmune diseases⁴², psychiatric disorders^{43, 44}, and neural control of infectious diseases^{3, 28}. It is important to distinguish these types of circuits as hard-wired bona fide circuits from interactions mediated by soluble molecules including neurotransmitters and hormones produced by the immune, endocrine or the nervous systems, respectively⁴⁵. Progress in deciphering these circuits has recently expanded at an accelerating pace⁴⁶. How peripheral signals are being processed by distinct subgroups of sensory neurons in brain centers (also referred to as brain nuclei) such as the brainstem, the parabrachial nuclei, the hypothalamus and the insula have already provided mechanistic insight into multiple brain traits that had remained outside of any comprehension including the sickness syndrome⁴⁷, fear and empathy⁴⁸, jealousy, and autism and schizophrenia^{5, 8, 43, 44, 49}. For few circuits, neuron subtypes that integrate these signals and connect to other neuron subtypes projecting back to the periphery have been identified in mice⁵⁰. It is noteworthy that many circuits not only involve several constituents of the PNS and the CNS but that both interact with organ-specific relatively independently acting intrinsic nervous systems^{1, 26, 28, 51–53}.

Several recent lines of evidence point to a new - until recently elusive - circuit involving the arterial tree that we have termed the artery brain circuit (ABC)^{7, 8, 54}. The ABC appears to interact with the much better established heart brain circuit (HBC) to form a large dynamic cardiovascular brain circuit (CBC)^{26, 27, 35, 55}. In this review, we will discuss what is currently known about both the ABC and the HBC to form a global CBC network with its core constituents, i.e. the arterial tree and the heart. We will discuss that multiple types of CBC dysfunction directly affect a substantial number of CVD outcomes including myocardial infarcts (MIs), sudden death, acute and chronic heart failure, kidney dysfunction, stroke and dementia⁴⁶. Finally, we will propose a novel neurobiology-centered CVD hypothesis: The neuroimmune cardiovascular circuit hypothesis.

THE ARTERY BRAIN CIRCUIT (ABC)

The conventional target of atherosclerosis research has been the inner layer of the arterial wall, i.e. the intima, because atherosclerotic plaques arise there^56–66. However, it is well established that the disease involves all three arterial wall layers with the media and adventitia layers also being affected⁹. Unlike mainstream research with its focus on the intima plaque, we decided early to focus on the adventitia which forms the outer connective tissue coat of arteries: We found that the majority of immune cells during disease progression in mice accumulated in the adventitia rather than in plaques^{54, 67–74}. Since it became apparent that the adventitia is the arterial wall layer where the immune system senses atherosclerotic plaques in the intima by forming aggregates of innate and adaptive immune cells, interest in the adventitia gained momentum. Later, delineation of artery tertiary lymphoid organs (ATLOs) in the adventitia of diseased but not healthy arteries in hyperlipidemic mice as well structured leukocyte aggregates that resembled those found in cancer and autoimmune diseases and other chronic inflammatory diseases raised further interest^{68, 69}. Moreover, it became apparent that similar immune cell aggregates formed in atherosclerotic aortic aneurysms and coronary arteries in humans⁷⁵ and that disruption of them in mice altered the progression of atherosclerosis⁶⁹. The adventitia has been the subject of recent reviews and will therefore not be considered further here^{73, 74, 76–79}. Owing to space limitations, we will not cover extensive discussions of risk factors in this review as multiple recent reviews have been reported^56–60.

Arteries directly talk to the brain via polysynaptic axonal connections

The discovery of a structural ABC was made possible by several lines of evidence: In landmark observations made by the founder of modern anatomy, i.e. the Belgian anatomist Andreas Vesalius, who in 1543 described the striking macroanatomical proximity of the PNS and the arterial tree¹⁷. Research beginning in the late 20^th century identified numerous interactions between the PNS, blood vessels and immune cells leading to two relatively independently studied areas: the area of the cardiovascular link^{13, 17, 31} and the area of neuroimmunology^{16, 20–23, 80}: imaging studies revealed that the PNS not only uses the adventitia as its main conduit to reach distant targets but that its axon endings directly gain access to the adventitia. Moreover, restructuring of the innervation of the adventitia was particularly pronounced in diseased segments of arteries: It involves newly formed axon networks⁵⁴. Focusing on the abdominal aorta of aged ApoE-deficient mice, two components of the PNS were identified in the formation of neuroimmune cardiovascular interfaces in the adventitia: The sympathetic nervous system and the sensory nervous system but - in this segment - not the parasympathetic nervous system. Interestingly, this type and location of aorta adventitia innervation resembled that of normal lymph nodes⁸¹. However, segments of other arterial tree segments showed a different type of innervation pattern with the vagus nerve innervating the aortic arch^{26, 52, 53}. In view of this apparent heterogeneity of artery innervation, we expect that future studies will identify further nervous system components in other artery segments depending on their respective functions and anatomical microdomains to regulate cardiovascular homeostasis.

The ABC Sensor

Two major PNS components were identified in the adventitia of diseased arteries of aged hyperlipidemic mice and human coronary arteries and abdominal aorta aneurysms: The sensory nervous system forms the ABC sensor and the sympathetic nervous system forms the ABC effector (Figures 1, 2). We interpret the restructuring events during atherosclerosis progression of both components of the ABC in the adventitia as a response of the PNS to the inflammatory environment afforded by the leukocytes that specifically infiltrate those adventitia segments that are burdened by plaques^{9, 69, 71}. Multiple molecules that are involved in the bidirectional communication between the sensory nervous system are likely to be produced by leukocytes but also by the axon endings including calcitonin gene-related peptide and multiple nociceptors: our immunohistochemical analyses of the adventitia showed marked expression of transient receptor potential V1^{12, 82–84} at axon endings as well as markers of axon growth cones. The transient receptor potential V1 is a classical nociceptor that responds to a variety of exogenous stimuli including heat, the chili pepper ingredient capsaicin (though these are unlikely mediators of the ABC) but - with relevance to atherosclerosis and other CVDs - to inflammatory mediators that may be induced in activated immune cells during atherosclerosis progression (Figure 1). Using virus tracing studies we showed that the upper abdominal aorta segment, is innervated by neurons in dorsal root ganglia (DRG) from where the sensory nervous system enters the spinal cord to reach the parabrachial nucleus and the central amygdala⁵⁴ (see for comparison below the distinct innervation patterns of the heart in the HBC in Figure 3). When taken together, this data established the sensory arm of the ABC.

Figure 1. The ABC sensor.

We propose that the ABC is initiated in the adventitia of diseased artery segments establishing a tridirectional communication network between innate immune cells, the arterial wall, and the peripheral nervous system. Inflammatory mediators may activate nociceptive receptors, for example, TRPV1 (transient receptor potential vanilloid 1), ion channel receptor, for example, Nav1.8 (voltage-gated sodium channel 1.8), P2X3 (P2X purinoceptor 3), cytokine receptors and GPCRs (G protein-coupled receptors) at axon endings of sensory neurons resulting in action potentials that are transduced to dorsal root ganglia to enter the spinal cord to be conveyed to brain nuclei (see below).

Figure 2. The ABC effector.

Sympathetic nervous system neurons are projected via the spinal cord to the periphery⁵⁴. Sympathetic nerves of celiac ganglion neurons directly innervate the adventitia.

Figure 3. The HBC.

The HBC is composed of a sensory arm through vagal and spinal afferents reaching the brain either via DRGs and the spinal cord or via the nodose ganglia; the motor arm contains sympathetic and parasympathetic efferents. The ICN communicates with the heart and the vagus and the sympathetic nervous system. The ICN is preferentially present in the atria (inset above the heart symbol). RA, right atrium; LA, left atrium; ICN, intrinsic cardiac nervous system.

The ABC Effector

The ABC effector of the sympathetic nervous system was identified using a combination of imaging and virus tracing experiments in hyperlipidemic mice⁵⁴. There was a marked increase of adventitia tissue epinephrine levels (and calcitonin gene-related peptide levels for the ABC sensor) in areas that are burdened by plaques (Figure 2). This data may be interpreted as a biologically relevant interaction network between the arterial wall, the immune system and the nervous system to form neuroimmune cardiovascular interfaces⁸⁵. Transcript analyses of these artery segments revealed the formation of multiple other neurotransmitters, cytokines, and growth factors as potential candidates for this interaction^{69, 71}.

THE HEART BRAIN CIRCUIT (HBC)

A better understanding of the multi-directional communication system between the heart and the brain in health and disease will fill important gaps. This will not only improve our understanding of the physiology of heart innervation but also open the way to delineate the basis for a series of heart diseases including atrial fibrillation, life-threatening arrhythmias such as ventricular tachycardia and ventricular fibrillation with their severe forms leading to sudden death. There are several major components in the HBC⁸⁶ (Figure 3): First, the sensory arm is composed of vagus afferents and spinal nerves from the heart; second, the CNS cardiovascular center in the brainstem integrates interoceptive signals and higher brain (examples are circadian clocks regulating heart rate) inputs that result in the generation of premotor outputs; third, extrinsic cardiac motor control through parasympathetic and sympathetic efferent nerves; and fourth, elaboration of local neural networks formed by neurons located within the heart which is also referred to as the intrinsic cardiac nervous system (ICN) (see below).

Sensing from the Heart

Signals from the heart and distinct artery segments such as the aortic arch are closely monitored by extrinsic sensory neurons in the vagus and spinal nerves, with their cell bodies located in the nodose ganglia or in DRGs (Figure 3); and for the arterial tree most likely all DRGs^{86, 87, 88}. More work is needed to delineate the relation between sensory neurons in the DRG versus those in the nodose ganglion. As a major neuronal subtype, vagal sensory neurons actively monitor numerous vital heart-derived cues to maintain cardiovascular homeostasis^{26, 89, 90}: pulsatile and tonic pressure changes are sensed at various locations such as atria, ventricles, and the aortic arch. In addition, vagal sensory neurons also monitor pathological changes like ischemia and inflammation, likely through secreted peptides and multiple signalling molecules^{89, 90}. Vagal aortic afferents with diverse sensory modalities, response patterns, and activation thresholds have been described. At least two specialized atrial receptors have been characterized, both of which are fast conducting A-fibers but may differentially sense atrial systole and filling. The majority of vagal ventricular afferents are capsaicin-sensitive C-fibers that are also sensitive to alkaloids such as veratridine as well as bradykinin⁹¹. Together, the results have shown that cardiac sensation via the vagus nerve is complex, demonstrating the important role of the HBC in maintaining cardiovascular homeostasis. Such heterogeneity and complexity present a considerable challenge for studies to better understand the general coding principles for cardiac signals and the underlying molecular and cellular mechanisms for the majority of cardiovascular responses. However, considerable efforts are underway to delineate vagal cardiovascular afferents from anatomical, molecular, and functional angels. A major question is if and how sensory endings are morphologically specialized for the input signals. Using a collection of anterograde tracing approaches, a variety of specialized vagal sensory endings have been characterized^{52, 55, 92, 93}. However, except for a clear link between aortic end-net endings and arterial blood pressure sensing⁵², signals sensed by these morphologically and genetically defined sensory axon endings largely remain unknown. Linking terminal morphologies with their electrophysiological properties and physiological roles for vagal cardiac afferents has been an extremely challenging task. However, delineation of them will undoubtedly provide new insights for better understanding the underlying sensory mechanisms implying major progress for future therapeutic interventions in diseases as varied as high blood pressure, arrhythmias and thrombus formation due to atrial fibrillation.

Activation of vagal cardiac afferents triggers physiological reflexes to regulate heart rhythm, blood pressure, cardiac output, respiration and renal functions to maintain cardiovascular homeostasis⁹⁰. Sensations of blood pressure changes and oxygen/CO₂ levels from the aorta via vagal aortic baroreceptor and chemoreceptor neurons regulate heart rate/blood pressure and respiratory patterns, respectively, similar to their carotid counterparts^{52, 53}. The Bainbridge reflex, i.e. a tachycardia response to an increase of blood volume, is a regulatory reflex initiated by atrial stretch or distension detected by vagal sensory endings in heart atria. Like the baroreflex, the Bainbridge reflex is bi-directional in the sense that a decrease in venous return, such as during hemorrhage and hypotension, will decrease heart rate⁹⁴. Thus understanding the Bainbridge reflex holds therapeutic potential to treat acute life-threatening conditions such as those experienced in the intensive care unit including shock or a low ventricular ejection fraction following a MI. Moreover, the Bezold-Jarisch reflex is an inhibitory cardiac reflex attributed to activation of vagal sensory endings on heart ventricles that causes severe bradycardia and hypotension^{95, 96}. When taken together, despite their highly important roles in cardiovascular regulation in both healthy and disease conditions, our basic knowledge about most vagal cardiac afferents remains limited. However, recent advances in mouse genetics are rapidly improving our understanding of vagal cardiac afferents. Several genes have been proposed to be critically involved in the baroreflex^97–99, i.e. the mechanosensitive ion channels PIEZO1 and PIEZO2 fulfil most criteria of being the primary baroreceptor that directly senses blood pressure changes⁵³. Knocking out both Piezo1 and Piezo2 from vagal sensory neurons in paired-like homeobox 2b-Cre mice resulted in an almost complete loss of the baroreflex. The same phenotype was also observed in mice with ablated Piezo2⁺ vagal sensory neurons. Moreover, optogenetic activation of Piezo2⁺ vagal sensory neurons induced pronounced bradycardia. Anatomically, Piezo2⁺ vagal afferents form macroscopic claws around the aorta, which are decorated with mechanosensory ‘end-net’ endings⁵². Surprisingly, it seems that ‘flower-spray’ endings are not involved in blood pressure sensing as ablating aortic ‘flower-spray’ endings did not impact the baroreflex. Therefore, it is apparent that aortic blood pressure is sensed by PIEZO1/PIEZO2 mechanosensitive ion channels through ‘end-net’ endings^{52, 53}.

Similar to vagal afferents, cardiac DRG neurons also sense mechanical changes and a variety of neurochemicals, many of which are involved in cardiac pathological conditions such as myocardial ischemia and inflammation¹⁰⁰. Stimulation of cardiac DRG afferents results in an increase of sympathetic tone and a decrease of parasympathetic tone. For example, electrical or chemical stimulation of cardiac sympathetic afferents blunted the arterial baroreflex¹⁰¹. Elevated cardiac DRG afferent activity has been described in CVDs including chronic heart failure and hypertension, and their roles have been extensively studied using various animal models.

Cardiovascular Control from the Brain

The brain directly communicates with the heart through two major routes: the parasympathetic pathway involving vagal motor neurons in the brainstem and postganglionic neurons in the heart; and the sympathetic pathway involving preganglionic neurons^{86, 102}. In general, parasympathetic outputs mediate negative chronotropic effects through cholinergic signalling while sympathetic outputs positively impact both chronotropy and inotropy via adrenergic signalling. Here we will focus on some recent results on circuit mapping.

Vagal motor neurons are primarily located in two brainstem nuclei: the dorsal motor nucleus of the vagus and the nucleus ambiguus. However, the molecular mechanisms for axon targeting for vagal cardiac motor neurons remain to be elucidated. Single cell studies using scRNAseq approaches have identified 8 dorsal motor nuclei of the vagus neuron subtypes, with distinct molecular markers and spatial distribution patterns¹⁰³.

Sympathetic postganglionic neurons densely innervate the sinoatrial and atrioventricular nodes and directly modulate pacemaker activity via norepinephrine release that acts primarily through the β1 adrenergic receptor (ADR) and downstream cAMP signalling¹⁰⁴, while other ADR types also contribute¹⁰⁵. In addition to norepinephrine, sympathetic neurons also co-release neuropeptide Y which binds to the neuropeptide Y2 receptor expressed on the presynaptic terminals of cholinergic neurons to inhibit parasympathetic inputs to pacemakers¹⁰⁴. Sympathetic nerves also extensively innervate atrial and ventricular cardiomyocytes and modulate cardiac contractility. Under disease conditions such as MI, sympathetic nerves undergo extensive remodelling, and such abnormal sympathetic activity may lead to arrhythmia and sudden cardiac death¹⁰⁶.

The Intrinsic Cardiac Nervous System (ICN)

Extensive local neural networks in the heart suggests that neurocardiac control is far more complex than previously recognized. Neurons of the heart, so-called intrinsic cardiac neurons with cell bodies residing in dispersed clusters called ganglionic plexuses, form the ICN^{30, 86, 104, 107}. This local neural network interacts with extrinsic afferent and efferent fibers to coordinate the heart-to-brain axis. Moreover, the ICN is able to act independently of extrinsic cardiac neurons as local actors. Both anatomical and electrophysiological evidence suggests that, like the enteric nervous system, a single ICN cluster might be composed of sensory, motor, and local circuit neurons that form a microcircuit to sense and process surrounding cardiac changes and control local cardiac functions. Moreover, ICN clusters among different cardiac ganglionic plexuses communicate with each other through interganglionic neurons to coordinate cardiac functions across the entire heart (Figure 3 inset).

In mammals, the number of ICNs is broadly correlated with the size of the heart (number of ICNs per heart: ~1,000 in mouse, ~6,500 in rat, ~12,000 in pig, and 43,000 – 94,000 in human)³⁰. ICN clusters are discretely located on the dorsal atrial surface in small animals, but more profusely distributed across broader atrial regions and even extending to ventricles in larger animals including rabbit and human^{30, 108, 109}. The majority of ICNs reside on the supra-ventricular epicardial surface^{30, 86, 104}. Recent development in mouse genetics and adeno-associated virus-guided anatomical tracing has provided important tools. For example, pacemakers in the sinoatrial and atrioventricular nodes represent an important downstream target of ICNs. Electrical stimulation of cardiac plexuses resulted in a significant drop in the heart rate. On the other hand, surgical ablation of certain ICN ganglia blocks the bradycardia induced by vagus nerve stimulation. This ICN-sinoatrial node projection has been demonstrated in mice using a combination of genetically engineered animals (choline acetyl transferase-Cre mice) and specifically designed adeno-associated viruses that selectively target peripheral neurons¹¹⁰, and high-resolution volumetric light-sheet imaging of tissue-cleared hearts¹¹¹. Meanwhile, activating specifically designed adeno-associated viruses targeting cholinergic ICNs located at the posterior region of the heart using optogenetics causes significant bradycardia. Together, this provides important evidence for the ICN-sinoatrial node connections from both anatomical and functional aspects at cellular resolution. In addition to the sinoatrial and artrioventricular nodes, ICNs likely cover a much larger heart area. Extensive cholinergic innervation of atrial appendages, interatrial septum, and ventricles has been characterized in various animal species^{108, 112}. However, complete mapping of ICN projections on the heart remains to be an important challenge of future studies. Multiple ICN neuron subtypes with distinct morphologies, response properties, and physiological roles have been characterized. Immunohistochemical studies have revealed that a variety of neuromodulators and neurotransmitters are present in the ICNs^{30, 104, 113–115}. Although it has been well documented that different neurochemicals can be co-released, such as norepinephrine and neuropeptide Y in sympathetic neurons¹¹⁶, it is unclear whether neuropeptide Y or cocaine and amphetamine-regulated transcript-derived protein are also co-released with acetylcholine, and if so, how they differentially regulate heart functions. In a separate study using laser capture and microdissection, 154 genes were identified in 151 captured single ICNs from spatially distinct ganglionic plexuses¹¹⁷. Therefore, a detailed molecular architecture of the ICN network has not been established yet.

ICNs have been classified into different subtypes based on their distinct electrophysiological properties such as firing patterns^{113, 118–120}. Interestingly, whether some isolated ICNs have spontaneous rhythmic activities resembling pacemaker cells remains controversial. Electrical or pharmacological stimulation of ICNs resulted in multifaceted, frequency and location-dependent cardiac hemodynamic changes, suggesting that cardiac regulation by the ICN is under precise spatial and temporal control¹⁰⁴. For example, stimulating selective ICN ganglia may elicit bradycardia, tachycardia, or more complicated changes in heart rhythms¹²¹. Local stimulation of ICNs within a ganglionic plexus may also modulate remote atrial and ventricular tissues, likely mediated by long-range projecting motor neurons or through neural circuits among multiple ganglionic plexuses. Although traditionally been thought as parasympathetic ganglia, it was estimated through electrophysiology that only ~40% of ICN receive direct parasympathetic inputs from the vagus nerve¹²². On the other hand, in vivo electrophysiological recordings of ICN in anesthetized dogs demonstrated that some ICNs may receive direct or indirect sympathetic inputs from the stellate ganglion^{109, 123}.

A variety of CVDs are associated with malfunction of the ICNs, including atrial and ventricular fibrillation, myocardial ischemia, and heart failure^{86, 124}. For example, abnormal activation of intrinsic cardiac ganglia around the pulmonary veins resulted in atrial arrhythmia, and ICN ablation is a viable option for intervention to treat atrial fibrillation in the clinic. Although the initiation of atrial fibrillation in humans is thought to be associated with an imbalance between sympathetic and parasympathetic tones, a subset of atrial fibrillation events can be triggered by abnormal ICN activity without extrinsic cardiac inputs¹²⁵. Nevertheless, the underlying molecular and cellular mechanisms between aberrant ICN activities and atrial fibrillation remain an important area to develop therapeutic strategies in the area of cardiac arrhythmias. However, since ICN ganglia plexuses are heavily interconnected with each other, and various ICN subtypes may differentially regulate distinct heart functions, whether surgical ablation of predicted ICN ganglia plexuses in severe bradycardia remains to be determined¹²⁶. Furthermore, mechanical or chemical disruption of cholinergic ICNs also increases the likelihood of ventricular arrhythmia¹²⁷, which is a leading cause of sudden cardiac death under a variety of disease conditions. However, as electrical or chemical stimulation of selective ICNs triggers ventricular arrhythmias¹²⁸ reveals that ICN activities need to be precisely regulated to maintain cardiac homeostasis. Other data further showed afferent and efferent remodelling of ICNs after MI, suggesting a functional role of the ICN in MI and heart failure. Thus, a better understanding of the biology of ICNs is likely to provide new therapeutic targets for treating CVDs including life-threatening arrhythmias under a variety of disease conditions.

THE IMMUNE SYSTEM IN ARTERY AND HEART HOMEOSTASIS AND DISEASE

The immune system emerged as a critical regulator of heart and artery physiology and CVDs. Studies combining flow cytometry, intravital microscopy and mouse genetic reporter models revealed that the heart harbours a broad repertoire of tissue-specific immune cells. The latter include macrophages, mast cells, monocytes, neutrophils, innate lymphoid cells (ILCs), dendritic cells, B cells and T cells. Cardiac macrophages are abundant, ontogenically diverse and heterogeneous. Four main cardiac macrophage subsets have been identified and each one occupies specific anatomical locations. For example, macrophages and dendritic cells are found in the aortic valves and coronary arteries; the atrioventricular node contains high numbers of macrophages; the pericardial fluid harbours macrophages, mast cells and B cells; and the pericardial adipose tissue contains a diverse set of leukocytes. Importantly, recent studies have shown that cardiac resident immune cells actively contribute to cardiac development and steady-state physiology throughout life, e.g., electrical conduction and cardiac rhythm¹²⁹. In the context of CVDs, innate immune cells have been described to initially drive a pro-inflammatory response to overcome tissue damage and to promote wound healing at later stages of inflammation. Recruited and tissue-resident innate and adaptive immune cells have been shown to be involved in inflammation, revascularization, cardiomyocyte dedifferentiation and fibrotic scar formation in the injured cardiac tissue. The heterogeneity and role of cardiac immune cells in homeostasis and heart pathology, including MI, conduction disorders, myocarditis and endocarditis, have been extensively reviewed elsewhere^{129, 130}.

The arterial wall also contains resident innate immune cells, including macrophages and dendritic cells, but their role in artery physiology only recently started to be acknowledged. In contrast, monocytes, macrophages and neutrophils, are well known for their roles in the initiation, progression, and destabilization of atherosclerotic lesions¹³¹. Over the past three decades our understanding of atherosclerosis has switched from a simple cholesterol-storage disease to a lipid-driven inflammatory disease. Notably, it is now well established that the chronic inflammatory disease of arteries depends on inflammatory leukocyte supply from the bone marrow and extramedullary sites. In addition to monocytes, macrophages and neutrophils, recent evidence suggests a potential involvement of previously underappreciated cell types in atherosclerosis, namely, dendritic cells and platelets¹³². Similarly, diverse adaptive immune cells, such as T cell subsets (CD8⁺ T cells and CD4⁺ regulatory T cells) have been shown to promote or attenuate arterial disease in rodent models¹³³. Whether atherosclerosis and other CVDs is ordinary autoinflammatory disease as previously thought or whether autoimmune B cells and autoimmune T cells participate in the progression of atherosclerosis is currently emerging as an important new research focus^{134, 135}. In murine atherosclerosis models and human diseased tissues, recent evidence is pointing to a bona fide autoimmune component for B cells¹³⁶ and more recently autoimmune T cells^{133, 137, 138}. This data has important implications for understanding the poorly understood relation between bona fide autoimmune diseases such as systemic lupus erythematosus or rheumatoid arthritis and atherosclerosis known to exacerbate each other^139–142. There is recent progress in the area of the role and a detailed description of the contribution of different immune cell subsets to atherosclerotic lesion progression which can be found elsewhere^{131, 132}.

Despite the enormous progress towards the characterisation of cardiovascular immune cells, the cellular and molecular circuitries that underlie their function remain unclear. Given that cardiovascular immune cells are strategically located in close proximity to non-immune cells, such as cardiomyocytes, fibroblasts, endothelial cells, coronary microvessels, lymphatic vessels and nerves, it is very likely that interactions between immune and non-immune populations contribute to cardiovascular health and disease⁵⁴. We will focus below on the fast-evolving field of neuroimmune interactions and their potential involvement in cardiovascular physiology and pathology.

Neuroimmune Cell Interactions in Peripheral Organs

Recent technological advances have provided new insights on cellular and molecular mechanisms that underly the communication between the nervous and the immune systems. Over the last decade, numerous studies have revealed that immune cells and nerve terminals colocalize and functionally interact at discrete anatomical locations in peripheral organs, forming neuroimmune cell units that orchestrate host defence, tissue homeostasis and organismal physiology¹⁴³. These bidirectional neuroimmune interactions have been described across several organs, including the bone marrow, spleen, skin, lung, intestine and adipose tissue^{1, 2, 144} and – more recently in the adventitia of aged atherosclerotic mice⁵⁴. In addition, such neuroimmune circuits are key orchestrators of multiple physiological processes, including haematopoiesis, organogenesis, host defence, inflammation, tissue repair, metabolism and thermogenesis^{1, 2, 143, 144}. Different branches of the PNS, including the sensory, sympathetic, parasympathetic, and enteric nervous systems have been shown to form neuroimmune hubs, mostly with innate immune cells. While peripheral neuroimmune cell interactions have been extensively reviewed elsewhere^{1, 2, 143, 144}, here our goal is to provide examples of neuroimmune interactions by macrophages and ILCs, which highlight the potential impact of neuroimmune regulation of cardiovascular health and disease.

Tissue-resident macrophages in the intestine and adipose tissue have been described to function as important hubs that integrate neuronal-derived cues. Under steady state conditions, enteric neurons maintain muscularis macrophages through secretion of colony-stimulating factor-1, which is required for macrophage development¹⁴⁵. Reciprocally, muscularis macrophages directly regulate neuronal function via secretion of bone morphogenetic protein-2, which activates bone morphogenetic protein receptor-expressing enteric neurons¹⁴⁵. This bidirectional neuronal-macrophage crosstalk regulates the peristaltic activity of the colon. Additionally, commensal microbe-derived signals can modulate bone morphogenetic protein 2 and colony stimulating factor 1 expression by macrophages and enteric neurons, respectively. This suggests that an intricate microbiota-driven neuronal-macrophage axis controls gastrointestinal motility, contributing to tissue homeostasis under steady-state conditions¹⁴⁵. In the context of enteric bacterial infection, activation of extrinsic sympathetic ganglia enhances the tissue-protective profile of muscularis macrophages, via norepinephrine signaling in ADR2-expressing macrophages¹⁴⁶. Interestingly, muscularis macrophages have also been shown to play a role in the maintenance of intrinsic enteric-associated neurons during enteric infection¹⁴⁷. Specifically, ADR signalling in muscularis macrophages was shown to induce neuronal protection through an arginase 1-polyamine axis¹⁴⁷.

In adipose tissue, a discrete population of macrophages tightly associates with sympathetic fibres. These neuron-associated macrophages take up and catabolize norepinephrine through expression of the norepinephrine transporter and the norepinephrine-degrading enzyme monoamine oxidase A, respectively¹⁴⁸. During aging, a similar macrophage-mediated norepinephrine-degrading activity was reported in the adipose tissue¹⁴⁹. Tissue-resident macrophages were also found to regulate sympathetic innervation of the brown adipose tissue, thus indirectly controlling local norepinephrine bioavailability, thermogenesis, and obesity¹⁵⁰.

Macrophages are among the most abundant, well studied and functionally relevant - though endowed with diverse and sometimes opposite functions - immune subsets in heart physiology and in atherosclerotic plaques, regulating plaque growth. Despite intensive research, the potential of neuroimmune crosstalk in cardiac homeostasis and in the pathogenesis of CVDs remains largely unexplored. In the context of atherosclerosis, it has been shown that a neuroimmune guidance cue, i.e. netrin-1, and its receptor, are expressed by macrophages of human and mouse atherosclerotic plaques¹⁵¹. Selective deletion of netrin-1 in bone marrow-derived cells reduced atherosclerotic plaque size and complexity in mice deficient in the receptor for low-density lipoprotein, and promoted macrophage egress from plaques, possibly by inactivating the migration of macrophages toward the chemokines CC-chemokine ligand 12 and CC-chemokine ligand 19, which drive macrophage egress from plaques¹⁵¹. Still, evidence for direct and functional neuronal-macrophage crosstalk and its implications for heart and artery physiology and CVD is essentially lacking.

Akin to macrophages, ILCs are innate immune effectors that actively engage in neuronal crosstalk in peripheral tissues. At mucosal barriers, such as the intestine and lung, ILCs reside in close proximity to neuronal cells. In the intestine, a neuroglia-ILC3-epithelial axis promotes intestinal tissue repair upon inflammatory and infection insults¹⁵². Enteric glial cells sense microbial and host-derived alarmin cues and produce neurotrophic factors that act on tyrosine kinase receptor-expressing ILC3s, promoting the production of tissue-protective interleukin-22¹⁵². Enteric and pulmonary ILC2s have been also shown to integrate and respond to neuronal cues derived from cholinergic, sympathetic and sensory nerve fibers. Cholinergic neurons control ILC2-mediated responses to allergens or helminthic infections via production of neuromedin U^153–155. Neuromedin U signals through neuromedin receptor 1 expressed in ILC2s leads to a rapid and potent production of the type 2 inflammatory cytokines, interleukin-5 and interleukin-13, and of the tissue-protective amphiregulin, conferring immediate tissue protection to helminthic infection^153–155. Pulmonary and intestinal ILC2s also express the ADR2 and colocalize with adrenergic neurons in the intestine¹⁵⁶. Abrogation of ADR2-mediated signalling results in increased ILC2 responses, type 2 inflammation, and lower helminth infection burden, effects that are reversed by ADR agonist treatment¹⁵⁶. In the lung, vasoactive intestinal peptide released by nodose afferent nociceptor sensory fibres potentiates airway inflammation by stimulating ILC2s to produce interleukin-5, which can act directly on nociceptors, further increasing vasoactive intestinal peptide expression and generating a type 2 inflammatory signalling loop that potentiates allergic inflammation¹⁵⁷. Interestingly, ILC2s have recently been shown to promote heart tissue repair, to alleviate cardiac fibrosis upon acute cardiac injury and to improve the recovery of heart function following MI^{158, 159}. Nevertheless, whether neurons and ILCs co-operate to orchestrate cardiovascular physiology in health and disease remains unknown.

Long-Distance Neuroimmune Interactions

Recent studies have started to unravel how neuroimmune interactions regulate health and disease via brain-body axes. These studies are providing new insight into the brain regions involved in these complex regulatory networks and are revealing the mechanistic pathways through which neuroimmune circuits operate at an organismal level. It has recently been shown that light-entrained and brain-tuned circadian circuits regulate enteric ILC3s, intestinal homeostasis, gut defence and host lipid metabolism¹⁶⁰. Surgical- and tissue-specific genetic approaches identified the suprachiasmatic nuclei in the hypothalamus as a key regulator of enteric ILC3 homeostasis, intestinal physiology and health¹⁶⁰. In another study, neuro-mesenchymal units were shown to control ILC2s and obesity via a brain-adipose circuit¹⁶¹. Combined retrograde tracing and functional chemogenetic manipulations identified a sympathetic aorticorenal circuit that modulates ILC2s in gonadal fat and connects to higher-order brain areas, including the paraventricular hypothalamic nucleus in the hypothalamus¹⁶¹. Thus, neuro-mesenchymal units translate cues from long-distance neuronal circuitry into adipose-resident ILC2s function, shaping host metabolism and obesity¹⁶¹. In the context of adaptive immune responses, a brain-spleen circuit was shown to directly impact the formation of immunoglobulin-producing plasma cells via a paraventricular hypothalamic nucleus / central amygdala-splenic nerve -B cell axis⁵⁰. Stress-responsive neurons in the central amygdala and the paraventricular hypothalamic nucleus that express corticotrophin-releasing hormone were shown to be connected to the splenic nerve and pharmacogenetic activation of these neurons increased plasma cell formation after immunization⁵⁰. Interestingly, in the context of psychological stress, distinct brain regions were shown to differentially and rapidly shape leukocyte distribution and function across peripheral organs, impacting on disease susceptibility¹⁶². Motor circuits induced rapid neutrophil egress from the bone marrow to peripheral organs via skeletal-muscle-derived neutrophil-attracting chemokines, while corticotrophin-releasing hormone neurons in the paraventricular hypothalamic nucleus promoted monocyte and lymphocyte mobilization from blood and secondary lymphoid organs to the bone marrow through direct, cell-intrinsic glucocorticoid receptor signalling, thus protecting against autoimmunity, but impairing immunity to viral infections¹⁶². Finally, intestinal inflammation activates neurons in a brain area called the insular cortex or the insula. Artificial reactivation of these ‘immune-imprinted’ neurons is sufficient to generate organ-specific recall of inflammatory responses that resemble the initial inflammatory episode¹⁶³. Altogether, these and other studies (reviewed in²⁵) are raising increased interest in mapping the brain regions and the communicating pathways involved in brain-peripheral immune system communication. Nevertheless, whether long-distance neuroimmune circuits are homeostatic cardiovascular players largely remains elusive.

Lifestyle Factors

Long-distance neuroimmune circuits allow for the integration of external challenges (exteroception) into internal body homeostasis (interoception) as introduced above. How lifestyle factors, such as sleep, psychological stress, diet and exercise, are integrated in cardiovascular physiology, possibly by neuroimmune circuits, also largely remains unexplored. Both epidemiological evidence and data from animal models indicate that stress and high-fat/high-cholesterol diet aggravate chronic inflammatory diseases, while regular exercise and healthy sleeping habits help to prevent CVDs¹⁶⁴. It is becoming increasingly clear that lifestyle-derived factors shape haematopoiesis, leukocyte trafficking, and innate immunity^165–167. In agreement with these views, several studies unravel the pathways through which sleep, stress, diet and exercise modulate haematopoiesis and promote inflammatory leukocyte supply to the atherosclerotic lesion, accelerating disease progression^168–170. Interestingly, some of these studies have revealed neuroimmune pathways as the underlying mechanism linking lifestyle factors to CVD progression. As an example, it has been shown that sleep regulates haematopoiesis and protects against atherosclerosis in mice via a neuroimmune axis. Mice subjected to sleep fragmentation display reduced levels of hypocretin in the lateral hypothalamus. This stimulatory and wake-promoting neuropeptide regulates myelopoiesis and protects against atherosclerosis by limiting the production of macrophage colony-stimulating factor-1 by hypocretin-receptor-expressing pre-neutrophils in the bone marrow^171–173. In humans, autoimmune destruction of hypocretin causes narcolepsy and low plasma levels of hypocretin are associated with increased risk of MI and heart failure. Uncovering how lifestyle factors and cardiovascular neuroimmune circuits intersect to shape cardiovascular health, disease onset and progression is certainly amongst the most challenging tasks ahead.

Inputs from the ABC and HBC Merge in Distinct Brain Nuclei

Brain nuclei are clusters of neurons at a specific location in the brain containing a variety of neuron subtypes which receive afferent signals or target peripheral organs. When we compared larger brain areas involving the ABC or the HBC, we noticed that both subcircuits shared several brain nuclei (Figure 5)^{26, 36, 54, 55, 174–182}. This observation raised an important new question: Are identical neuron subtypes shared by ABC and HBC afferents and/or efferents within these nuclei ? If this were to be the case, it would indicate that specific neuron subtypes may be concomitantly activated by the ABC and the HBC depending on physiological and/or pathophysiological inputs implying direct ABC-HBC interactions at the level of single brain neurons. To answer this important question, a series of experimental approaches using new mouse models and an extensive set of cutting-edge methods need to be employed. These studies will define the connectivity networks of the ABC and the HBC and identify the nature of their interactions and places within the global CBC connectome. Understanding the CBC connectivity map in the brain and the periphery may contribute to mechanistic insights into the atherosclerotic brain before life-threatening types of brain injury such as strokes ensue.

Figure 5. The ABC and the HBC share large functionally diverse brain areas but whether identical neuron subtypes are shared remains unknwon.

Major multifunctional and markedly heterogeneously brain areas are shared by the ABC and the HBC. However, whether neuron subtypes connect to arteries and the heart remains to be defined. Here we list broad functions that have been assigned to these brain nuclei with the potential to regulate CVDs: Insula (emotions)^{5, 6, 36, 183, 184}, Amygdala (memory)^{175, 185}, Hypothalamus (stress)¹⁸⁶, Rostral Ventrolateral Medulla^{179, 187} (stress), Nucleus Tractus Solitarius (body weight)^{26, 188}, and Dorsal Motor Nucleus of the Vagus (metabolism)^{55, 188}. Many other impacts and connections of these nuclei are not depicted for ease of reading.

ARTERIES AND THE HEART FORM A GLOBAL CBC NETWORK AND ARE ORGANIZED IN ENGRAMS

Recent progress to define a structural ABC^{7, 8, 54} together with the better understood HBC^{26, 55, 189, 190} and rapid advances on the role of the immune system to direct both the ABC and the HBC^{1, 2, 9, 69, 71, 133, 143, 191} support the presence of a global CBC network in which multiple cardiovascular subcircuits interact (Figures 1–7). When the known components of this network are considered from the point of view of neuroimmune circuit physiology and pathophysiology, several points deserve attention: The interaction between arteries and the heart to form a CBC is only beginning to be understood as the two have largely been studied by relatively independent areas in the past. While atherosclerosis research has focused on the atherosclerotic plaque in the intima^{56–58, 60, 192} and on molecular mechanisms of atherosclerosis risk factors including hyperlipidemia and lipid transport, hypertension, diabetes mellitus, obesity, lifestyle, research of the heart has been dedicated to physiological responses, MI healing and treatments of arrhythmias and heart failure. Now, time has come to bring artery and heart innervation together and begin to appreciate the role of the brain to power the CBC network and its subcircuits^{1, 5, 25, 26, 36, 55, 191, 193}. Broadly, the CBC can be functionally divided into CBC sensors (Figure 6) and CBC effectors (Figure 7). It should be noted that each of the participants of the CBC consist of multiple genetically, functionally and territorialized distinct subcircuits to provide well organized neuronal control nodes depending of specific requirements of the various artery segments and heart compartments. What may be the connectivity structure of these subcircuits ? Each subcircuit may be defined what has been termed engrams (the term engram was originally coined for bona fide memory neuronal networks) which can be described as assemblies of neurons that share inputs from or outputs to peripheral stimuli¹⁹⁴. Engrams are known to exceed the borders of a single brain nucleus by forming a collection of neurons stretching across several nuclei in a network-like pattern¹⁹⁴. The identification of engrams in the CBC has not yet been accomplished directly but their existence is likely given the involvement of the immune system in forming engrams in other disease conditions and being involved in CVD progression: Thus, brain engrams have been shown to exist for physiological but also for pathophysiological responses involving the immune system^{5–8, 43, 44, 46, 49, 54, 163, 193, 195} forming inflammation-related engrams that are defined as the ability of the brain to remember past tissue injury of any given peripheral organ^{6, 163, 196}. Therefore, brain engrams can also be defined as regulatory units of neurons that store information of past physiological organ-to-brain connections or organ destruction. Of note, engrams endow important networking tasks to regulate longterm brain-organ responses. Furthermore, new engrams may form throughout life to adjust to the needs of peripheral organ physiological responses that are controlled by the brain. For all these reasons, we view the identification of CBC engrams as one of the major prerequisites to study the impact of activating or inactivating distinct neurons to define their roles in driving or preventing CVDs. Understanding potential cardiovascular engrams in the framework of the neuroimmune cardiovascular circuit hypothesis is expected to open unprecedented opportunities to test the role of distinct neuron subtypes to affect CVD progression. To delineate these engrams a powerful tool box is available in mice facilitating translational and clinical trials.

Figure 7. Brain to organ connections form a CBC effector and yield engram-derived cues back to the cardiovascular system.

Retrograde virus tracing from the adventitia of the abdominal aorta segment revealed direct polysynaptic efferent connections of autonomic brain nuclei via the brainstem and the spinal cord to the celiac ganglion⁵⁴, while similar studies of heart innervation showed vagus efferent neurons in the brain stem innervating the aortic arch and the heart including the atria (location of the ICN) and ventricles^{106, 114}. In addition, it is known that spinal cord sympathetic nervous system innervation involves the bone marrow, the adrenal gland, the spleen and other internal organs. These brain-to-organ efferents contain information from cardiovascular injury or other types of past tissue injuries what can be defined as engrams^{6, 36, 163, 196, 197}.

Figure 6. Organ to brain connections form a CBC sensor and program engrams.

Focusing on the upper abdominal aorta, virus tracing has shown that the adventitia is directly innervated by neurons of DRG segments T6–12⁵⁴, while sensory neurons in the nodose ganglia of the vagus nerve and different DRGs innervate the heart (Figure 3)^{87, 88, 26, 52, 55, 92, 93}. Both routes of cardiovascular innervation gain access to the brainstem to be projected to higher brain areas along the pain pathway⁵⁴. The chronic flow of organ-to-brain signals may form sensory engrams of the cardiovascular system.

Unprecedented Opportunities to Treat CVDs May Emerge from Understanding the Large CBC Network

As the CBC has only recently emerged as a hardwired polysynaptic nervous system axonal network, we can only imagine what is going on behind (not yet supported by data) closed doors. Progress in this area of neurobiology, however, is needed to develop preventive and therapeutic approaches to understand root causes of CVDs. Several principle aspects may come from preclinical models: Direct interference into the impact of neuron subsets in the brain to directly determine the longterm effect of the activity state of these neuron subtypes on a variety of CVD outcomes including atherosclerosis, mechanisms of arrhythmias, myocardial healing, ischemia perfusion injury are needed. Moreover, the PNS afferents and efferents within the CBC require analyses using electrophysiological approaches to identify nerve territories to apply localized surgical approaches as reported recently in a mouse model⁵⁴. Within the brain compartment of the CBC, defining primary and secondary integration centers including the brain endocrine system will yield further mechanistic insights. Mapping distinct neuroimmune cardiovascular interfaces by scRNAseq approaches such as those already applied to the adventitia or the ICN may lead to the identification of molecular cues such as neurotransmitters, inflammatory mediators or receptors expressed on neuronal cells or nonneuronal cells that participate in the CBC. Such advances will help to identify their functions in the cardiovascular system and guide translational studies in human brains and the human PNS. Therefore, we expect that the identification of new tangible experimental approaches and of new CBC-related biomarkers in the circulation from a combination of basic research and clinical studies in the coming years.

A State-of-the-Art Toolbox is Required to Uncover the Structure and Function of the Global CBC Network: Warm up for a Marathon

While there is considerable progress in defining neuronal circuits in experimental mice, studies of both the ABC and the HBC to form a global CBC is in its infant stages. While the toolbox (Figure 8) to delineate the CBC further will allow to answer urgent questions as related to the identification of neuron subtypes and address the impacts of neuron subtypes on CVD progression, multiple additional methods and approaches have to be employed including the generation of new mouse models that allow to assay both atherosclerosis progression and heart function in both short term and long term experiments. These mouse models may combine distinct heart-related transgenic and tissue-specific mouse models^{52, 53, 55} on a hyperlipidemic genetic background^{54, 198}. In particular, translational studies including clinical trials as they relate to the CBC are demanding tasks as the former are at present largely limited to morphological approaches using a variety of descriptive imaging tools. Yet, the combination of functional magnetic resonance imaging and machine learning algorithms provide new approaches to identify neuronal cells forming engrams¹⁹⁷. However translational studies will come into further focus if work in mice will yield testable hypotheses in CBC research. This work is predicted to come from postmortem studies of human neuronal tissues using single nuclear RNAseq analyses^{199, 200}, conventional immunohistochemical analyses and multiple brain imaging approaches^{201, 202}. We highlight here six major core methods that are all required to make progress to investigate the CBC (Figure 8).

Figure 8. Toolbox to investigate the CBC.

Progress in understanding the CBC requires state-of-the-art tools that exploit neurobiology approaches including tissue clearing²⁰³, multiomics^{133, 204}, optogenetics^{184, 205, 206}, chemogenetics^{207, 208}, electrophysiology^{54, 55, 209}, and multiple tracing methods^{54, 55, 161, 163}. We refer to excellent recent reviews covering each of the methods here for the interested reader.

THE NEUROIMMUNE CARDIOVASCULAR CIRCUIT HYPOTHESIS

Multiple hypotheses have been suggested in the past to address the multifactorial nature of the pathogenesis of atherosclerosis as the major pathology underlying CVDs, including the lipid storage/retention hypothesis, the response to injury hypothesis, the oxidized low density lipoprotein hypothesis, the immune and inflammation hypothesis among others yet none of them has provided therapeutic options in the clinic^{56, 58, 61, 129, 131, 210–213}. As reviewed above, recent research into the pathogenesis of atherosclerotic coronary artery disease and other segments of the arterial tree together with a better understanding of the innervation of the heart suggests the existence of an ABC and an HBC^{1, 6, 25, 26, 36, 46, 52–55, 163, 184, 214}. Moreover, the ABC and the HBC form a global CBC network regulating both physiological and pathophysiological responses of the cardiovascular system. The understanding of the workings of the CBC is still in an infant stage requiring application of state-of-the art toolboxes to progress (Figure 8). Here, we propose a new comprehensive hypothesis of the physiology and pathophysiology of the cardiovascular system with the nervous system as its core element that we term the Neuroimmune Cardiovascular Circuit Hypothesis: In this hypothesis, the brain is viewed as the principal hub in which both genes and interoceptive and exteroceptive cues including risk factors are received, processed and integrated¹⁹⁵. Furthermore, it integrates previous hypotheses related to inflammation and the immune system.

CONCLUSIONS

Recent progress in both basic and translational research into the anatomy and function of artery and heart physiology has established the contours of a CBC consisting of two major essential components: the ABC and the HBC. Within the CBC, the brain acts as a forceful instructor, master switch and control hub to regulate physiological responses. The complexity of the nervous system combined with the multilayered cardiovascular and immune systems makes the CBC highly vulnerable, however, to trigger CVDs when homeostatic regulation goes awry. A further disease susceptibility relates to the brain as an integration hub not only for interoceptive inputs from bodily signals but also for exteroceptive cues from the environment. In this regard we propose to view exteroception in the context of the common CVD risk factors such as unhealthy diets, sedentary lifestyles, and other harmful participants that manifest themselves as high blood pressure, type II diabetes mellitus, obesity and hyperlipidemia. When the wide variability of the genetic makeup of each individual human is taken into consideration with the above, a new holistic hypothesis regarding the mechanisms that control CVDs may be considered. We propose to tentatively name this hypothesis the neuroimmune cardiovascular circuit hypothesis (Figure 9).

Figure 9. The Neuroimmune Cardiovascular Circuit Hypothesis.

The neuroimmune cardiovascular circuit hypothesis emphasizes the central role of the brain in affecting CVD progression. It incorporates several components: The interoceptive impacts of arteries and the heart on the CBC in addition to the genetic background and lifestyle exteroceptive impacts on each individual human. Depending on the composition of these inputs and actions of - at times opposite - factors, cardiovascular homeostasis may be either maintained or disrupted to accelerate CVD progression.

Figure 4. Neuroimmune interactions regulate CVD progression via long-distance brain-body axes.

Long-distance neuroimmune circuits allow for the integration of exteroceptive and interoceptive cues in health and disease. Bidirectional connections between immune cells and the nervous system empowers neuroimmune axes with the unique capacity to efficiently perceive, integrate and respond to the ever changing external and internal environmental perturbations. We hypothesise that neuroimmune circuits operate in the heart and arteries to orchestrate cardiovascular physiology, health and disease.

Nonstandard Abbreviations and Acronyms:

ABC

artery-brain circuit

CBC

cardiovascular brain circuit

CNS

Central Nervous System

DRG

dorsal root ganglion

HBC

heart-brain circuit

ICN

intrinsic cardiac nervous system

ILC

innate lymphocyte cell

MI

myocardial infarct

NA

nucleus ambiguus

PNS

peripheral nervous system

scRNAseq

Single Cell RNA sequencing, method to identify the transcriptome of single cells