Functional Neural-Bone Marrow Pathways: Implications in Hypertension and Cardiovascular Disease

Resistant hypertension - current perspective

Treatment resistant hypertension (TRHT) is characterized by persistently high arterial blood pressure, partly as a result of a dysfunctional autonomic nervous system (ANS), wherein sympathetic drive/norepinephrine (NE) spillover are increased and parasympathetic drive is decreased.^1–3 The difficulty in treatment of TRHT arises precisely from this partly neurogenic component, as the available drug therapies do not target the central nervous system (CNS) directly. Due to this, and despite recent advances in techniques such as renal denervation and carotid baroreceptor activation,^{4, 5} successful treatment and long term control of TRHT remain challenging.⁶ In an attempt to develop more effective treatments, many groups are investigating specific causes of TRHT. A large body of experimental evidence implicates both genetic and environmental influences, such as salt sensitivity and elevated systemic renin-angiotensin system (RAS) activity^7–14 in the pathophysiology of this disease. Furthermore, a majority of studies point to dysregulations in the activity within the cardiorespiratory brain regions as a reason for increased sympathetic and decreased parasympathetic drive to the peripheral organs,^14–21 resulting in end organ damage,^21–25 vascular/endothelial dysfunction,^{26, 27} and hormonal imbalance,²⁸ which perpetuate the pathophysiology and complicate treatment strategies. Despite our increasing understanding of TRHT, the origins of this brain dysregulation remain largely unknown. Recently, the activity of the immune system^29–31 and neuro-immune pathways in hypertensive patients and animal models of hypertension has been highlighted.^32–36 Studies suggest that both the sympathetic and the parasympathetic arms of the ANS can exert their influence on the activity of the immune organs, tissues and cells, and that it is the dysfunctional ANS-immune communication that may lead to hypertension and CVD.^35–40 The aim of this review is to summarize the latest advances in this field and review the current understanding of connections between the autonomic and immune systems, specifically the connections between the brain and the bone marrow (BM), the largest source of hematopoietic cells in the body. In addition, we will highlight the importance of BM activity in CVD and hypertension, and propose novel bidirectional brain- BM communication hypothesis whose dysfunction may have important implications in the development of therapeutic strategies for neurogenic TRHT.

Role of bone marrow in cardiovascular health and disease

BM is central in the regulation of hematopoiesis.^{41, 42} It constitutes ~4% of the total body mass in humans, produces ~500 billion hematopoietic cells each day, and accounts for ~90% of total hematopoiesis in the body.^{41, 42} BM hematopoietic stem and progenitors cells (HSPCs) interact with cells of secondary lymphoid organs such as the spleen, which regulate the HSPC differentiation and maturation.^43–46 Readily available BM-derived cells are released into the circulation diurnally or in response to diverse pathophysiological cues such as vascular or tissue injury.^{37, 47–51} The involvement of BM and HSPCs in CVD and hypertension has attracted significant attention in the last decade for several reasons. First, CVD has been linked with an overactive adaptive immune system. Increased inflammation has been reported in heart disease, stroke, vascular diseases and hypertension in both humans and animal models.^{29–31, 36, 37, 52–59} In fact, a correlation exists between elevated inflammatory responses and disease progression,^{31, 57, 60, 61} and immunosupression can delay or even arrest the progression of CVD.^62–67 Second, following an acute injury, the BM cells with angiogenic and reparative properties such as endothelial progenitor cells (EPCs)^{17, 68–73} are immediately recruited to the site of injury, aiding in repair of the tissue damage.^74–78 Unfortunately, some patients with advancing CVD and other related diseases have significantly decreased pools of EPCs,^{79, 80} and this decreased vascular potential further exacerbates the dysfunctional vasculature already seen in CVD. Third, an overactive renin-angiotensin system (RAS) is implicated in exaggerated immune system responses in human patients as well as animal models of cardiovascular diseases.^{14, 17, 81–83} In angiotensin II (Ang II)-dependent hypertension, increased systemic and splenic inflammatory cells and factors have a central role in initiation and maintenance of hypertension,^{17, 35, 40, 82, 84} presumably by conferring damage on the vasculature as a result of decreases in vascular repair-relevant progenitor cells. Consistent with this are observations that antihypertensive effects of the RAS inhibitors have been shown to improve the number and function of EPCs in CVD.^{71, 85, 86} Collectively, this data suggest a correlation between BM HSPCs and inflammation in development of CVD.

We have recently proposed that it is not only the increased inflammatory damage, but also the imbalance in the BM-derived immune cell-dependent vascular damage and endothelial cell-driven vascular repair, is central to cardiovascular pathophysiology.^{17, 32, 33, 36} For example, the hypertensive phenotype of the spontaneously hypertensive rat (SHR) is characterized by chronically elevated inflammation and down-regulated BM-derived EPC number and function, as evidenced by the decreased ability of the EPCs to proliferate and migrate to the site of the damage, as well as the loss of EPC angiogenic ability.³⁶ Additionally, data suggests that central effects of Ang II-induced hypertension involve an increase in the BM-derived inflammatory cells (ICs), as well as a decrease in BM-derived EPC number and function, which cannot be attributed to secondary effects of systemic hypertension.^{14, 17} Therefore, the decrease in the EPC count coupled with their decreased function may perpetuate vascular damage in hypertension. One mechanism through which overactive Ang II in hypertension may affect the BM HSPC activity is by having a direct effect on the HSPC ‘stemness’, through promotion of premature differentiation.^{41, 42} This would diminish the angiogenic and reparative abilities of HSPCs, thereby compromising vascular repair. Furthermore, our data have shown that Ang II undermines the ability of HSPCs to home to the BM niches⁸⁷ which could further compromise their angiogenic functions. The increase in inflammatory stress can also directly reduce the function and activity of EPCs,^{88, 89} as evidenced by clinical data demonstrating that anti-inflammatory drugs and antioxidants exert beneficial effects on the EPCs.⁸⁹ While a transient inflammatory response may in fact stimulate EPC mobilization, thereby promoting tissue repair,⁹⁰ evidence suggests that chronic systemic inflammation may lead to functional impairment of EPCs and their depletion.^91–93 In line with this is the observation that treatment with BM-derived mononuclear cells does not improve neurological recovery following stroke in the SHR which is characterized by both increased systemic inflammation and elevated RAS, perhaps due to the high inflammation which remained widespread in these animals.⁹⁴ Taken together, it is reasonable to suggest that systemic RAS may act distinctly from its well-established classic effects to both increase ICs and decrease EPCs.

It is evident from the above discussion that production, mobilization and release of cells from the BM is critical in maintaining regular vascular homeostasis, and that dysregulation at any stage in these processes leads to devastating cardiovascular consequences. However, it is important to point out that some discrepancies exist in the literature. For example, there are studies that did not find a difference in EPCs between healthy and hypertensive humans,⁹⁵ while others linked dysfunctional EPCs only with hypertension that was associated with other comorbidities.^{96, 97} In one such study Delva et al⁹⁵ found no significant differences in EPC numbers and function between a group of carefully selected patients exhibiting a hypertensive phenotype and control subjects. However, this patient group may have had less severe hypertension because their treatment employed no more than two antihypertensive medications, namely RAS and beta blockers, both of which may directly affect the function of EPCs.^{69, 71, 98, 99} The hypertensive patients also were treated with aspirin, known for its beneficial vascular effects^{65, 66} including those benefiting the EPC function directly.¹⁰⁰ These and other studies pinpoint the complexity of the problem of TRHT, and further highlight the need for novel therapeutic targets. Some of the most relevant questions arising from the evidence to date are whether the BM activity is regulated by the brain, and if the dysfunctional brain-BM communication could account for CVD and hypertension.

Evidence of autonomic nervous system regulation of BM activity

The first evidence of the importance of ANS for BM cell homeostasis is derived from the circadian studies which demonstrated the regulation of BM cell activity.³⁶ It is well established that the release of BM HSPCs is rhythmically regulated in a circadian manner, for which sympathetic drive is essential.^{36, 47–50, 101, 102} In rodents, the increase in sympathetic drive at night is associated with the release of the ‘surveillance’ immune cells from the BM, which circulate through the body in search of an infection or injury.^{36, 47} The return of these cells to the BM signals the release of the repair cells, aimed to heal and repair.^{47, 102} However, in cases of prolonged infections or otherwise compromised immune responses, the accumulation and aging of the surveillance immune cells in the circulation leads to the loss of the trigger for release of the repair cells into the circulation.⁴⁷ Clinically, this may translate into a diminished ability of the body to repair itself in times of sustained inflammatory damage. Furthermore, loss of circadian rhythmicity of the BM cell release appears to be closely tied to any dysfunction in the sympathetic drive.³⁶ For example, the loss of the circadian rhythmicity observed in diabetes-related peripheral neuropathy is attributed to the decreased sympathetic drive to the BM and results in dysfunctional BM EPCs.¹⁰² On the other hand, our studies have demonstrated that elevated sympathetic drive to the BM results in a perturbed BM adrenergic receptor signaling system that is associated with loss of circadian rhythmicity in BM cell release.³⁶ Therefore, we infer that the BM HSPC environment is tightly regulated, which allows it to respond very quickly to an environmental change or injury, but which can also result in adaptation and plasticity of the BM cell responses, ultimately contributing to a pathological state. In the context of hypertension, we propose that the chronic elevation in the BM sympathetic drive leads to the loss of circadian rhythmicity of the BM cell release, resulting in chronically elevated systemic inflammatory responses and decreased EPC availability and function.³⁶ Therefore, clinical targeting of the elevated sympathetic drive in neurogenic hypertension could have a new dimension.

The second body of evidence highlighting the importance of ANS in BM function comes from studies demonstrating increases in inflammatory cells in several disease states. Most recently, Dutta et al³⁷ described the role of increased sympathetic drive in recruitment of BM cells following post-myocardial infarction (MI). Their data showed that increased sympathetic drive to the BM following MI caused the liberation of the BM HSPCs, which seeded in the spleen and boosted the production of monocytes. The increase in monocyte production and their infiltration caused the development of larger, destabilized atherosclerotic plaques with a more advanced morphology, resulting in higher incidence of recurrent myocardial events as well as increased incidence of stroke.³⁷ In a similar fashion, inflammatory cells have been shown to infiltrate the brain and contribute to the neuropathology in several diseases, including CVD.^103–107 While the exact mechanism underlying the extravasation of inflammatory cells into the CNS is still an area of active research, several hypotheses are under investigation. First, an increase in chemo-attractant pro-inflammatory molecules, such as CCL2, have been shown to be associated with the increase in BM derived microglia and monocytes in the CNS.^{104, 106} Bone marrow monocytes will home in on areas of the brain where the concentration of CCL2 is the highest. Whether this process involves the breakdown of the blood brain barrier (BBB) is still debated. Hypoxic-ischemic brain injury is associated with BM-derived microglia infiltration into areas of the brain where the BBB had been compromised.¹⁰⁶ However, other groups have shown that the infiltration of BM monocytes into the brain is independent of the BBB integrity.¹⁰⁸ Therefore, it seems that while the breakdown of BBB would facilitate the infiltration of BM derived cells into the brain, it is not necessary for this process to occur. In diabetes, for example, infiltration of BM-derived monocyte progenitors in the brain pre-sympathetic areas contributes to the increased pool of activated inflammatory microglia and subsequent neuro-vascular-glial inflammatory status.¹⁰⁴ Our preliminary data support this contention in the hypertension paradigm as well.¹⁰⁹ Studies on the mechanism of the infiltration of BM cells into the brain of hypertensive animals are underway, but could be due to a combination of BBB breakdown as described by Stern¹¹⁰ and increased chemo-attractant, pro-inflammatory molecules.^{111, 112} While all the differences between BM-derived and resident microglia have yet to be elucidated, there is one fundamental difference. Resident microglia work to survey their environment, become activated and help to repair acute injuries.^113–115 BM-derived microglia have only been described to move to the CNS in chronic conditions, indicating that either their role is only important in the long-term pathophysiology of neuroinflammation, or there are age-related effects on the homing of BM cells to the brain.¹¹⁶ The role of activated microglia in several important neurodegenerative/neuroinflammatory diseases is well established;^{17, 103–107, 112, 117, 118} however, their influence on pre-sympathetic neurons in the context of established CVD with chronically high sympathetic drive is currently understudied.

Our recent observations in rodent models of Ang II-dependent hypertension lead us to suggest that activation of microglia in the pre-sympathetic cardioregulatory brain areas may precede both the activation of the sympathetic drive and the increase in BP. Therefore, activation of microglia may be a crucial step in generation of high sympathetic drive to the periphery including the BM.^{32, 33} This concept does not disregard the effects of direct Ang II action on the pre-sympathetic neurons, which has been firmly established in neurogenic hypertension.^{7, 14, 15, 119} It is plausible to suggest that the healthy cardioregulatory system is capable of regulating its own responses to sporadic Ang II challenges, until these are intensified and/or aided by other environmental pro-hypertensive insults. Consistent with this idea, an important characteristic of the microglial cell is that it exhibits characteristics of a memory immune cell of the brain, meaning that it can be primed for activation.^{120, 121} Therefore, several pro-hypertensive signals may be required for microglial cell differentiation into a fully active pro-inflammatory stage, characterized by increased proliferation, migration and release of cytokines/neurotransmitters that will subsequently affect the state of the surrounding pre-sympathetic neurons. This will lead to increased sympathetic drive to the periphery, including to the BM, causing dysfunction in BM cell activity.¹⁷ Dysfunctional BM cell activity will contribute to decreased repair and increased inflammation, and consequently result in the infiltration of the immune cells to the brain, thereby perpetuating cardiovascular pathophysiology in a feed-forward manner.³²

These observations support the role of increased sympathetic drive in recruitment and regulation of BM cells. Evidence also exists for a role of the parasympathetic nervous system in regulation of BM-induced inflammation. Stimulation of the vagus nerve protects from excessive cytokine production and ameliorates experimental inflammatory disease via the inflammatory reflex mechanism. This involves activation of nicotinic acetylcholine receptor alpha7 (nAchRα7), on macrophages, lymphocytes, neurons and other cells.¹²² This inflammatory reflex has been described in the spleen, gut and, recently, the BM, where it has been shown that deficiency of the BM nAchRα7 impaired vagus nerve-mediated regulation of the pro-inflammatory tumor necrosis factor (TNF), in a mechanism dependent on the BM-derived macrophages.^{38, 39, 51, 122–125} In support of this data, our recent study has demonstrated a decrease in both acetylcholine (ACh) esterase and cholinergic acetyl transferase in the BM of the SHR compared to the Wistar-Kyoto rat (WKY), in addition to increased BM sympathetic drive.³⁶ As ACh appears to block the acute and direct effects of NE on the release of BM ICs, this would suggest that the anti-inflammatory parasympathetic effects directly oppose the pro-inflammatory effects of the sympathetic drive on the BM.³⁶ Interestingly, the vagal influence is reduced in hypertension; potentially further exacerbating the pro-inflammatory state. There is no evidence of direct vagal projections to the BM, and the parasympathetic effects are most likely to be endocrine and delivered via the BM extensive vasculature (Fig 1). Alternatively, the vagus may be indirectly affecting the BM by modulating the activity of the BM postganglionic sympathetic nerves, as it has been suggested in other lymphoid tissues¹²⁶ (Fig 1). Future studies should elucidate the cellular and neuronal mechanisms of the reduced parasympathetic drive in the hypertension-related inflammatory responses in the BM, and whether the cholinergic receptor system is as tightly regulated as the adrenergic receptors in the BM.

Figure 1.

Bidirectional communication between the brain and the bone marrow in cardiovascular homeostasis. Pro-hypertensive signals such as elevated Ang II lead to neuro-vascular-glial inflammation in the brain cardioregulatory areas. Dysfunctional ANS output is characterized by elevation in the sympathetic and a decrease in the parasympathetic drive to the periphery, including the BM. This process is coupled with elevated peripheral Ang II and causes a persistent increase in the ICs and decrease in EPCs. The elevated ICs contribute to the vascular and tissue damage, while decreased EPC count and function contribute to the decreased repair of this damage, leading to cardio-renal pathology. The combination of extravasation of the ICs to the pre-sympathetic brain areas and the increased somatic afferent input from the BM to the brain, via activation of the putative TRPV1 channels adds to the neuro-vascular-glial inflammation and drives the dysfunctional ANS output to the periphery. These events perpetuate the resulting cardiovascular and renal pathophysiology and resistant hypertension.

The evidence discussed above suggests that increased sympathetic drive to the BM, in conjunction with decreased parasympathetic influence, may initiate a cascade of signaling events culminating in dysfunctional BM cell activity and leading to hypertension.³⁶ The increase in sympathetic nerve activity (SNA) in pre-hypertensive animal models and humans supports this contention.^{1–3, 127} However, direct measurements of BM SNA in the pre-hypertensive state must be performed in order to confirm this hypothesis. Nonetheless, cumulative evidence supports the view that the ANS has profound influence on the BM that could be detrimental in hypertension and other CVD.

Afferent input from bone marrow to the brain

In the previous section, we have discussed the efferent brain-BM pathway and the role of the ANS in regulation of the BM cell activity in health and CVD. Our data, discussed below, support the concept that an afferent neuronal input from the BM to the brain may also exist and may be involved in modification of the ANS efferent pathway to the periphery, including the BM. This bidirectional brain-BM communication hypothesis is based on the following evidence; first, we used an anaesthetized and non-paralyzed cat model to investigate the effects of a localized, bolus BM injection of capsaicin (0.1 ml) on cardio-respiratory variables. We observed that capsaicin injection into the BM produced an immediate dose-dependent BP elevation, accompanied by apneusis, as evidence by the diaphragm and thyropharyngeus muscle EMG recordings and measurement of end tidal CO₂ levels (Fig 2A). These responses were enhanced by microinjection of kynurenic acid in the nucleus of the solitary tract (NTS; Fig 2B), suggesting that neural processing of the capsaicin-sensitive BM afferent projection may involve the NTS. Second, we used a decerebrated artificially-perfused rat preparation (DAPR)³⁶ to measure direct electrophysiological effects of a similar capsaicin injections in the femoral BM of a rat on the vagal and sympathetic efferent drive in situ. We found that, in addition to the similar capsaicin-dependent effects produced on breathing, as evidenced by the changes in the phrenic and hypoglossal nerve activity, capsaicin also immediately elevated the perfusion pressure, and affected both the efferent vagal and the thoracic sympathetic nerve activity (Fig 3B). This observation suggests that capsaicin-sensitive BM afferents may modulate the ANS efferent drive. Capsaicin also produced delayed, longer-lasting cardiorespiratory responses in both the cat and the rat (not shown), but these were attributed to the systemic capsaicin effects. Last, we employed functional MRI (fMRI) in an anaesthetized rat to image the brainstem regions, which may directly be stimulated by the capsaicin-sensitive BM afferent activation. We observed that a single femoral BM capsaicin injection in the anaesthetized rat greatly enhanced neuronal activity in the brainstem nuclei including the trigeminal sensory nucleus (Sp5), locus coeruleus (LC), as well as the cardioregulatory brainstem nuclei such as the NTS and the nuclei of ventrolateral medulla, as measured by the fMRI (Fig 3A). These are persuasive evidence for the existence of an afferent BM-brain neural connection, which when activated, may centrally modify the ANS efferent signaling to the periphery.

Figure 2.

A: Respiratory (diaphragm and thyropharyngeus muscle electromyographic activity), end-tidal CO₂, and blood pressure responses produced by a single capsaicin injection in the left femur of a freely-breathing anesthetized cat. Injections of 250 μM and 500 μM capsaicin in the bone marrow produced significant apneusis characterized by a prolonged inspiratory activity, and caused an immediate increase in the blood pressure in a dose-response manner. B: Bilateral microinjection of kynurenic acid (50 nl; 100 mM) in the rostral NTS prior to the capsaicin injection in the bone marrow produced a prolonged apneusis lasting >60 seconds, and an increase in the blood pressure by >40 mmHg. These data suggest existence of a direct afferent sensory input from the bone marrow to the brain cardio-respiratory regions.

Figure 3.

A: Increase in the blood oxygenation level dependent (BOLD) signal in the brainstem regions (locus coeruleus, LC; rostral ventrolateral medulla, RVLM; nucleus of the solitary tract, NTS; spinal trigeminal nucleus, sp5) following a single capsaicin injection (0.1ml, 500 μM; marked by arrow) in the left femur of an anaesthetized rat. Data are represented as 2D BOLD activation map highlighting the magnitude BOLD response (left column in A). Functional image is overlaid on T2 anatomical scan or an electronic atlas of the rat brain (middle column in A). 3D volumetric rendition shows the volume of activation in brainstem nuclei (top right panel in A). Below 3D maps are BOLD signal time courses for various nuclei. B: Respiratory (phrenic), vagal, hypoglossal and thoracic sympathetic nerve activities, and perfusion pressure (corresponding to arterial pressure) following a single capsaicin injection (0.1 ml; 500 μM) in the left femur of a decerebrated, artificially-perfused in situ rat preparation.

Capsaicin activates the transient receptor potential vanilloid type 1 (TRPV1) channels, expressed in primary sensory nerve fibers of both the somatic and autonomic afferent neurons.¹²⁸ The role of capsaicin-sensitive TPRV1 channels have been well described in relation with pain perception, and more specifically with the bone-related pain such as during different forms of cancer.¹²⁹ The bone itself is innervated with the A and C fibers,¹³⁰ and secretion of substances such as growth factors, cytokines and chemokines from tumor cells within the bone stimulate the primary afferent nociceptors and induce pain.^131–133 Pro-inflammatory cytokines suggested to be involved in cancer-related bone pain include IL-1beta, TNFalpha, IL-6, TGFbeta and MCP-1,^134–137 all of which have been implicated in hypertension-related inflammatory responses.³² TNFalpha is shown to sensitize the TRPV1 channels in the bone, causing deregulations of TRPV1 in dorsal root ganglion neurons and hyperalgesia,¹³⁶ while upregulation of MCP-1 in the tumor-burdened femur bone correlates with the severity of tumor progression and pain.¹³⁷ Furthermore, activation of microglial CX3CR1 receptors by the fractalkine-producing spinal cord neurons mediates pain during the carcinoma growth in rat tibia.¹³⁸ A parallel could be drawn between the inflammatory mediators causing activation of the pain sensory afferents within the cancerous bone, and those associated with inflammation in CVD. Activation of TRPV1 channels has profound effects on cardiovascular homeostasis, and TRPV1 dysfunction is implicated in increased salt sensitivity, water and sodium balance in hypertension, and in worsening of ischemia-reperfusion injury in the heart, brain and the liver.^139–141 In view of our findings and the data from the literature, we suggest that TRPV1 in the BM may be relevant in the following manner: (i) increased pro-inflammatory mediators may sensitize the TRPV1 in the BM; and (ii) activation of TRPV1 may stimulate inflammatory responses in the BM cells via stimulation of the release of calcitonin related gene peptide and substance P from sensory nerves. This may be relevant in the paradigm of hypertension, which is known to be characterized by both increased BM inflammatory¹⁷ and dysfunctional TRPV1 responses.^{140, 141} Thus, sustained activation of BM TRPV1 by inflammatory factors present in hypertension may exaggerate BM and systemic inflammation, contributing to cardiovascular pathology in a feed-forward manner, as we have hypothesized in Fig 1. Furthermore, based on our current findings, activation of TRPV1 on BM sensory afferents may send a direct neuronal signal to the brainstem nociceptive- and cardioregulatory areas, thereby modifying the ANS efferent responses (Fig 2 and 3). Further experiments are needed to dissect the afferent neuronal pathways from the BM to the brain in health and CVD.

This is an attractive concept in support of bidirectional brain-BM communication. However, our fMRI data indicate prominent activation of the sensory input-receiving nuclei such as the sp5 brainstem nucleus and the LC (Fig 3).^{142, 143} It is highly probable that the sensory afferents originating in the BM and ending in the brainstem are of somatic origin. The NTS, an important cardioregulatory region of the brainstem, is a site of convergence of somatic and autonomic regulatory mechanisms.^{144, 145} In line with this, our data show that pharmacological manipulation of the NTS modified the cardiovascular responses to BM capsaicin (Fig 2B), and that the NTS activity increases following BM capsaicin injection (Fig 3A). In hypertension, NTS activity is heavily influenced by the overactive RAS and especially Ang II, which increases the inhibitory GABAergic influence in the NTS and dampens the vagal baroreflex.^{20, 146} Therefore, in the context of hypertension, high inflammation may be causing the sustained activation of the BM sensory afferents, which together with Ang II-dependent dampened vagal responses characteristic of hypertension, could be part of a feed-forward loop resulting in even more inflammation in the periphery.

Based on this evidence, we propose the following bidirectional brain-BM communication hypothesis for cardiovascular homeostasis (Fig 1): Pro-hypertensive signals such as elevated brain and systemic RAS leads to neuro-vascular-glial inflammation in the brain cardioregulatory areas. The resultant dysfunctional ANS output, characterized by elevation in the sympathetic (and a decrease in the parasympathetic) drive to the periphery, including the BM, causes a persistent increase in the ICs and decrease in EPCs. The elevated ICs contribute to the vascular and tissue damage, while decreased EPC count and function contribute to the decreased repair of this damage, leading to cardio-renal pathology. Extravasation of the ICs to the pre-sympathetic brain areas may add to the neuro-vascular-glial inflammation and drives the dysfunctional ANS output to the periphery, thereby perpetuating the resulting cardiovascular and renal pathophysiology and resistant hypertension. The contribution of the novel somatic afferent input from the BM to the brain to our hypothesis remains to be fully investigated.

Future perspectives

In view of the evidence presented here, we believe that the autonomic control of BM plays a critical role in cardiovascular homeostasis and there appears to be a bidirectional communication between the brain and BM in this process. However, many questions remain unanswered. First, what are the exact neuroanatomical sympathetic and parasympathetic pathways innervating the BM? The loss of the BM sympathetic innervation following denervation of the superior cervical ganglion has been shown; however, exact pre-sympathetic and spinal neuronal pathways need to be elucidated to fully understand these pathways. Second, the exact nature of the BM sympathetic drive needs to be established. We have recently alluded to the respiratory coupling of the BM sympathetic nerve activity, but the barosensitivity of the BM sympathetic drive and whether this is changed in hypertension remains unknown. Third, the neuroanatomical pathways and the involvement of the afferent sensory input from the BM to the brain in hypertension need to be established. Fourth, the specific role of the vagus in regulation of the BM cells in hypertension needs to be expounded. We believe that these questions will yield important answers, which will move the field of neurogenic hypertension forward and provide novel therapeutic targets in the treatment of TRHT. One of the obvious potential therapeutic targets would be activated microglia that influence the activity of neurons within the brain cardioregulatory areas, some of which may control the BM and affect the function of the BM HSPCs. A recent study showed that treatment with minocycline, an orally available anti-inflammatory antibiotic, which can cross the blood brain barrier and inhibit microglial activation, reduced blood pressure and improved body weight and glucose levels in drug-resistant, morbidly obese/Type2 diabetic patients who were also hypertensive.¹⁰⁴ Therefore, our future investigations will focus on drug discovery of specific, orally available, blood brain barrier-crossing anti-inflammatory agents in order to alleviate neurogenic hypertension.