Altered inflammatory response is associated with an impaired autonomic input to the bone marrow in the SHR

Abstract

Autonomic nervous system (ANS) dysfunction, exaggerated inflammation and impaired vascular repair are all hallmarks of hypertension. Considering the bone marrow (BM) is a major source of the inflammatory cells (ICs) and endothelial progenitor cells (EPCs), we hypothesized that impaired BM-ANS interaction contributes to dysfunctional BM activity in hypertension. In the SHR, we observed a >30% increase in BM and blood ICs (CD4.8⁺), and a >50% decrease in EPCs (CD90⁺.CD4.5.8^-) compared to the normotensive Wistar-Kyoto (WKY) rat. Increased tyrosine hydroxylase (70%) and norepinephrine (NE, 160%), and decreased choline acetyl transferase (30%) and acetylcholine esterase (55%) indicated imbalanced ANS in SHR BM. In WKY, night time-associated elevation in SNA (50%) and BM NE (41%) was associated with increased ICs (50%) and decreased EPCs (350%), while BM sympathetic denervation decreased ICs (25%) and increased EPCs (40%). In contrast, these effects were blunted in SHR, possibly due to chronic downregulation of BM adrenergic receptor α2a (by 50-80%) and β2 (30-45%). Application of NE resulted in increased BM IC activation/release, which was prevented by pre-administration of Ach. Electrophysiological recordings of femoral SNA (fSNA) showed a more robust fSNA activity in SHR compared to WKY, peaking earlier in the respiratory cycle, indicative of increased sympathetic tone. Finally, manganese-enhanced magnetic resonance imaging (MEMRI) demonstrated that pre-sympathetic neuronal activation in SHR was associated with an accelerated retrograde transport of the GFP-labeled pseudorabies virus from the BM. These observations demonstrate that a dysfunctional BM ANS is associated with imbalanced EPCs and ICs in hypertension.

Introduction

Autonomic dysfunction, characterized by increased sympathetic and decreased parasympathetic activity, is a hallmark of neurogenic hypertension ^{1, 2}. Recent evidence has indicated a direct interaction of the autonomic nervous system (ANS) with the immune system (IS) to regulate normal cardiovascular homeostasis. Thus, a dysfunctional neural-immune communication has been implicated in the pathogenesis of cardiovascular diseases (CVD) and hypertension^{3, 4}. Evidence of increased sympathetic nervous system (SNS) activity to immune organs in hypertension supports this contention⁵. Moreover, atherosclerotic vasculature is further compromised by mobilization of the BM-derived ICs following myocardial infarction, characterized by increased sympathetic nerve activity (SNA)⁶. On the other hand, the anti-inflammatory effects of the vagus nerve (i.e. parasympathetic, PNS) stimulation are demonstrated by lowered levels of the inflammatory cytokines and suppressed activation of ICs^{7, 8}. Therefore, the imbalance of the parasympathetic/sympathetic influence in hypertension contributes to the hypertensive phenotype by perpetuating the inflammatory response⁹. In contrast to the ICs, decreased circulating levels of endothelial progenitor cells (EPCs) and their dysfunction are demonstrated in hypertension and CVDs^10-13, suggesting EPCs' impaired abilities in repairing vascular damage. The EPCs, like ICs, appear to be neuro-regulated, as suggested by the diurnal pattern of their release into the circulation¹³. As the EPC numbers and function are inversely correlated in patients and rat models of hypertension^{10, 11, 14}, the overactive sympathetic drive to the BM, and/or the imbalance in the overall parasympathetic/sympathetic tone in hypertension, may also directly contribute to the dampened EPC numbers and function, leading to the impairment of the endothelial reparative processes and accelerating the vascular dysfunction and hypertension-associated pathophysiology. Taken together, these observations led us to hypothesize that the autonomic imbalance influences the release of the progenitor cells from the BM, thereby affecting the circulating IC and EPC levels. Spontaneously hypertensive rat (SHR), a model of neurogenic hypertension that exhibits an early onset of autonomic and endothelial dysfunctions and increased inflammatory response, has been used in this study to evaluate this hypothesis. We present direct evidence that an altered autonomic regulation of BM is associated with an imbalance in the EPC and IC levels in hypertension.

Methods

All experimental protocols are presented in the Methods section and are available in the Online Supplementary Data. All animal procedures were approved by the University of Florida Institute Animal Care and Use Committee.

Results

Circadian regulation of BM ICs and EPCs

First, we investigated if increased sympathetic drive impacted the BM activity in the SHR, by comparing the levels of EPCs and ICs in the BM at the times of lowest (i.e. 11 am, ‘Day’) and highest (i.e. 8 pm, ‘Night’) sympathetic drive. We observed a 50% increase in the low frequency of SBP, LF:SBP, in the WKY, and a 130% increase in the LF:SBP in the SHR at Night vs. Day (Fig 1A). This was associated with increased BM norepinephrine (NE) (WKY: 41%; SHR: 38%) at Night vs. Day (Fig 1B). Furthermore, the overall sympathetic drive as measured by LF:SBP was higher in the SHR vs. the WKY rats at both Day and Night (Fig 1A-B). Furthermore, a 158% increase in the density of tyrosine hydroxylase (TH) immunoreactivity (Fig 4A), a 30% decrease in the density of choline acetyl transferase (ChAT) (Fig 4B), and a 55% decrease in the density of acetylcholine esterase (AchE) (Fig 4C) immunoreactivity around the femoral BM blood vessels were observed in the SHR compared to the WKY, suggesting impaired BM ANS input. Increased sympathetic drive at Night in the WKY rats was associated with a 50% and 33% increase in the IC levels in the BM and blood, respectively (Fig 1C-D, left panel), and a 350% decrease in the blood EPCs (Fig 1F, left panel). In comparison, the ICs were higher and the EPCs were lower in the BM of the SHR both at Day and Night (Fig 1C-D), and in the blood of the SHR at Day (Fig 1E-F, left panel) compared to the WKY rat. Furthermore, there was a lack of the Night time-associated increase in the ICs and decrease in the EPCs in the SHR (Fig 1C-F, right panels), suggesting a dysfunctional response of the SHR BM to the circadian-related sympathetic drive changes. To further investigate this, we performed BM sympathetic denervation by dissection of the superior cervical ganglion (SCGx) in the WKY and SHR as previously described¹⁵. Forty-eight hours following the SCGx, we observed a 50% and 30% decrease in the BM NE in the WKY and SHR, respectively (Fig 2A) compared to the naïve controls. This was associated with a 25% decrease in the BM IC levels, and a 40% increase in the BM EPC levels in the WKY (Fig 2B, left panel), similar to the decrease in ICs and increase in EPCs observed from Night to Day in the WKY (Fig 1C-D). However, BM ICs showed a trend toward an increase but produced no significant change following SCGx, while we observed ∼18% increase in the BM EPC levels in the SHR (Fig 2B, right panel), suggesting attenuated responsiveness of the BM HSPCs to the BM sympathetic changes in the SHR compared to the WKY. Quantitative PCR showed a significant decrease in both the α2a- (by ∼50 at Day and ∼80% at Night) and β2- adrenergic receptors (by ∼45% at Day and ∼30% at Night) in BM MNCs of SHR compared to the WKY. This suggested a possible mechanism for loss of circadian regulation of BM cells in the SHR (Fig 3).

Figure 1. Effects of elevated sympathetic drive at Night vs. Day on EPC and IC levels in BM and blood in SHR and WKY.

A: Spectral analysis of SBP reveals elevation in the overall sympathetic drive (LF:SBP) at Night vs. Day in both the SHR and WKY (P<0.05 vs. Day), and higher overall sympathetic drive at Night in the SHR vs. WKY (P<0.05 vs WKY ‘Night’), n=4 per strain. B: Significantly higher NE protein levels were observed in the BM cell supernatant at Night vs. Day in both the SHR and WKY (P<0.05 vs. Day, n=12 per strain), and higher NE protein levels at both Day and Night in the SHR vs. WKY (P<0.05 vs WKY, n=12 per strain). C-F: Significantly higher ICs in the BM (C) and blood (E), and lower EPCs in the blood (F) were observed at Night in the WKY but not the SHR (P<0.05 vs. Day, n=6 time of the day). Significantly lower ICs and higher EPCs were observed at Day in the BM (C-D) and blood (E-F) in the WKY vs. SHR (P<0.05 vs. SHR ‘Day), n=6 per strain.

Figure 4. Elevated Tyrosine Hydroxylase (TH) and decreased Choline acetyltransferase (ChAT) and Acetylecholinesterase (AchE) immunostaining in the BM of the SHR.

Immunohistochemistry reveals higher TH density, and lower ChAT and AchE densities around the blood vessels (brown staining) in the BM of SHR compared to WKY. The quantification was a result of extensive image analysis using Image J to analyze forty to fifty images per strain (P<0.05 vs. WKY; n=6 per strain).

Figure 2. Effects of superior cervical ganglion denervation (SCGx) on the BM EPC and IC levels in SHR and WKY.

A: A significant reduction in NE protein levels was observed in the BM cell supernatant in both the SHR and WKY following SCGx (P<0.05; n=3). B-C: Examples of raw FACS data for WKY (B) and SHR (C), showing the changes in CD4.8⁺ (top panel, highlighted in red box) and CD90⁺.CD4.5.8^- BM cells (bottom panel, highlighted in red box) in Control and SCGx rats. D: A significant decrease in the ICs in the WKY, and an increase in EPCs in both WKY and SHR were observed in the BM following SCGx (P<0.05 vs Naïve Control, n=3).

Figure 3. Chronic downregulation of adrenergic receptors β2 and α2a in the BM of SHR.

A-B: Quantitative RT-PCR in the whole BM MNCs shows significantly lower relative expression levels of adrenergic receptors β2 (A) and α2a (B) at both Day (white bars) and Night time (grey bars) (*P<0.05 vs Day; #P<0.05 vs WKY, n=6 per strain).

To investigate the mechanism of increased mobilization of the BM ICs at high sympathetic drive, we used in vivo real-time imaging of the tibial BM (Fig 5), to determine if increase in the local NE in the BM would influence mobilization of inflammatory BM cells by studying the mobilization of GFP-labeled ICs in response to administration of 6 μg/kg NE, in the absence and presence of 80 mg/kg Ach. A significant increase in the movement of ICs in response to NE was observed (Fig 5; online video link). This movement was significantly attenuated by pre-administration of Ach (Fig 5).

Figure 5. Direct effect of norepinephrine (NE) on activation and migration of BM ICs.

A: GFP-labeled CD4.8⁺ T cells were injected into a recipient mouse with an exposed tibial BM, and imaged in vivo under the fluorescence microscope. B-C: GFP-labeled ICs' movement was tracked in vivo before the NE injection. The panel on the left represents a still image of the bone marrow niche with GFP-labeled cells showing as bright gray, and their trajectory labeled by colored lines. The panel on the right represents the summation of each of the cells' trajectories plotted as the distance and velocity travelled. D-E: Representative movement of each individual GFP-labeled IC was plotted after the NE injection. F-G: Representative movement of each individual GFP-labeled IC was plotted after the NE injection in the presence of pre-administered Ach. H-I: NE significantly increased the distance (H) and the velocity of travel (I) of GFP-labeled BM ICs, which was attenuated by pre-administration of Ach (P<0.05 vs Control, n=7-15).

Loss of EPC function in the SHR

The overall decrease in the EPC numbers in the SHR compared to the WKY was accompanied with the loss of cell function: the angiogenic ability of the BM-derived cells was also reduced in the SHR compared to the age-matched WKY, as evidenced by a 65% decrease in the formed tube length, a 50% reduction in tube width formation ex vivo (Fig S1A-C), and a 35% decrease in the SHR BM EPC's ability to proliferate in response to SDF (Fig S1D).

Sympathetic nerve activity to the BM is altered in the SHR

Next, we characterized the activity of the sympathetic nerve innervating the femoral bone. Respiratory cycle triggered averages of simultaneously recorded phrenic (PNA) and femoral sympathetic nerve activities (fSNA) were made in the decerebrate artificially perfused rat (DAPR, Fig 6A) as established previously¹⁶. This in situ DAPR revealed a classic PNA pattern (Fig 6B, second and third panel) and a robust fSNA (Fig 6B, fourth and fifth panel). The influence of respiration on fSNA was enhanced by 9% CO₂ (Fig 6D, third panel, red arrows) and was completely blocked by hexamethonium (HEX, Fig 6D, fourth panel). Next, we compared the respiratory modulation of fSNA between the hypertensive and normotensive rat. The peak firing of fSNA in the SHR occurred earlier in the PNA cycle i.e. at the end of the inspiration (I) phase (Fig 6E, third panel, red arrow) and was approximately 25% more robust when compared to the normotensive control, the peak of which occurred in the post-inspiration (P-I) phase (Fig 6E, second panel, red arrow), typical of respiratory-sympathetic patterning^17,18. These responses were consistent and repeatable over several preparations (n=4 per strain).

Figure 6. Elevated sympathetic drive in the BM of the SHR.

A: Schematic of the in situ DAPR [decerebrated artificially-perfused rat preparation]. B: An example of the raw and integrated tracings of phrenic (PNA), and femoral sympathetic nerve activity (fSNA) at constant perfusion pressure (PP). C: Photograph of the femoral nerve bundle (b) innervating the femur (d) via the bone nutrient foramen (c). D: Phrenic-triggered fSNA in baseline conditions (5% CO₂) peaks immediately after the PNA peak i.e. in the P-I phase, characteristic of the SNA (panel 2, red arrow). 9% CO₂ activates the fSNA (panel 3, red arrow), and administration of HEX (300 μM) abolishes the fSNA pattern (panel 4). E: Top two panels represent PNA and fSNA from a control normotensive rat, and the bottom two panels are from the SHR. Phrenic-triggered fSNA is more robust and occurs earlier in the PNA cycle in the SHR (panel 4, red arrow) compared to the control rat (panel 2, red arrow). I: Inspiration, P-I: post- inspiration; M-E: mid-expiration; L-E: late expiration; HEX-hexamethonium.

Retrograde viral tracing reveals dysfunctional autonomic-BM communication in the SHR

Finally, MRI and GFP-PRV retrograde labeling experiments were carried out to further investigate increased autonomic-BM communication in hypertension. Mn²⁺-enhanced MRI, a technique commonly used to visualize elevated cellular activity in vivo, showed a 20-25% higher neuronal activity in the hypothalamic paraventricular nucleus (PVN) of the SHR compared to the WKY (Fig 7). In addition, the rate of the PVN retrograde labeling by the GFP-PRV injected in the BM was significantly faster in the SHR compared to the WKY (Fig 7B). For example, seven days post- GFP-PRV administration, the PVN neurons from the SHR showed robust GFP fluorescence, while little fluorescence was seen in the PVN of the WKY rats (Fig 7B). A similar increase in the labeling of neurons in other SHR autonomic brain regions (such as the NTS, RVLM, SFO) was observed (Online Fig S2 and S3). These responses were consistent and repeatable over several preparations (n=3 per strain for MRI; n=3 per strain per time point for GFP-PRV retrograde labeling).

Figure 7. Higher PRV-GFP retrograde labeling from the femoral bone marrow to the PVN is associated with higher neuronal activity in the PVN of the SHR compared to WKY.

A: MEMRI reveals significantly higher neuronal signal intensity in the PVN of the SHR (left panel, red dashed labeled area) compared to the WKY (right panel, red dashed labeled area). B: GFP staining reveals robust retrograde labeling in the PVN of the SHR (left panel) with very little GFP stain present in the WKY (right panel) at day 7 following the BM PRV injection. The higher magnification image demonstrates neuronal labeling (red dashed box; scale bar = 10 μm). C: Paxinos-Watson stereotaxic coordinates of the PVN.

Discussion

Our study is novel in a number of ways: 1) We are the first to establish the electrophysiological recordings of the sympathetic nerve innervating the femoral bone marrow (i.e. the fSNA, Figure 6). The electrical properties of fSNA in the SHR are similar to those of the thoracic SNA in the SHR¹⁹, in that its peak activity occurs earlier in the phrenic cycle and it is more robust compared to the normotensive control (Figure 6). 2) NE and TH levels in the BM of the SHR are increased. In addition, night time NE levels were significantly higher than daytime in the BM of both the WKY and SHR, when the overall sympathetic drive was higher in these animals. 3) IC levels increased and EPC levels decreased in the SHR compared to the WKY rats, and correlated with the increased sympathetic drive. 4) Direct application of NE into the BM activated the BM ICs. 5) Regular circadian regulation of the BM ICs and EPCs, which is present in the WKY, is significantly compromised in the SHR. 6) α2a- and β2-adrenergic receptor levels are significantly decreased in the SHR BM, suggesting a possible mechanism behind the loss of circadian regulation of ICs/EPCs in the SHR BM. 7) A decreased BM parasympathetic tone, as demonstrated by decreased BM ChAT and AchE in the SHR, may contribute to the inflammatory activation of the BM in the SHR. Taken together, we suggest that hypertension in the SHR is associated with a persistent increase in the sympathetic drive to the periphery, including the BM. This is associated with an increase in the BM NE and TH levels, resulting in impaired BM cell activity, reflected in the increased ICs and decreased EPCs in the SHR. Chronically high fSNA eventually leads to downregulation of adrenergic receptors in the BM, leading to the loss of circadian regulation of the BM cells in the SHR. This loss of circadian regulation, coupled with persistent increase in ICs and decrease in EPCs, compromises the ability of vasculature to repair the damage induced by hypertension, which is accentuated by the residual EPCs that become dysfunctional in the SHR.

We show here that the typical diurnal increases in the sympathetic vasomotor drive at night are accompanied by similar changes in the BM NE; however, both the overall sympathetic vasomotor drive, as indicated by LF:SBP, and the BM NE are significantly higher in the SHR, indicating an increased sympathetic drive to the BM of the SHR. This is corroborated by elevated BM TH protein levels in the SHR (Figure 4), as well as our electrophysiological recordings showing activity changes in the BM fSNA in the SHR, which are similar to the changes observed in the thoracic SNA of the SHR and are indicative of elevated SNA¹⁹. Other evidence supports this contention. For example, GFP-PRV retrograde labeling of the neurons in the PVN and other cardioregulatory brain regions from the BM is accelerated and more robust in the SHR compared to the WKY. This is not due to the genetic diversity between the two rat strains, but is rather associated with hypertension, as accelerated labeling of the PVN neurons by BM administration of GFP-PRV is also observed in chronic angiotensin II-infused rat model of hypertension (not shown), and is in agreement with the heightened neuronal activity in the PVN of the SHR, as demonstrated by elevated manganese-enhanced MRI signals in the SHR (Figure 7).

In the WKY, the night time-associated increase in the BM sympathetic drive was accompanied by elevated BM and blood CD4.8+ levels and decreased EPC levels, suggesting circadian regulation of the BM cells (Figure 1). This concept is not new, as both animal and human studies have shown that the release of BM cells follows a regular circadian pattern, and that in rodents, the immune cells (i.e. the surveillance cells) are released at night, while the repair cells (including the EPCs) are released during the day²⁰. This circadian regulation of BM activity appears to be dependent on the sympathetic innervation, interruption of which results in pathological situations, as evidenced in diabetes¹³. Our observations in the SHR are consistent with this, as the circadian regulation of the BM cells appears to be impaired in the SHR (Figure 1). Thus, in the SHR, the IC levels remain chronically high while the EPC levels remain low. One can, therefore, postulate that the presence of the functioning circadian rhythmic regulation of the highs and lows in ICs and EPCs at night and day, associated with diurnal sympathetic changes, is the reason that the WKY rat does not develop hypertension. In contrast, loss of circadian regulation in the SHR results in persistently higher ICs and lower EPCs throughout, which, combined with the reduction in the EPC function in the SHR (Online Figure S1), may contribute to increased inflammation and compromised repair of the vascular damage, thereby perpetuating the hypertension-related cardiovascular pathophysiology in the SHR. The loss of circadian regulation of the BM cell activity in the SHR is also reflected in the lack of BM cell response following the BM denervation. As the BM NE levels are reduced following the BM sympathetic denervation in both rat strains, this results in decreased ICs and increased EPCs but only in the WKY and not the SHR (Figure 2D). This may be due to the remaining high NE levels in the BM of SHR compared to the WKY rats (Figure 2A). However, the lack of the BM cell response following BM denervation in the SHR, as well as the loss of circadian regulation of the BM cell activity discussed above, is more likely due to chronic downregulation of specific adrenergic receptors in the BM of the SHR, perhaps occurring in compensation of the chronically high sympathetic drive²¹. Alternatively, it is pertinent to point out that the BM denervation was a relatively short-term experiment (<48 hrs), which may not be sufficient time to correct the chronic adrenergic receptor dysfunction in the BM of SHR.

BM adrenergic receptors are crucial in the BM cell responses²⁰. In line with this, we observed that NE applied locally to the BM activated the BM ICs, confirming a direct effect of NE on the BM cells. Interestingly, this effect of NE was blocked by pre- application of Ach in the BM, suggesting that the parasympathetic influence in the BM may counteract/dampen the effects of the SNA. As PNA is generally reduced in the SHR^22,23, and, as it appears, in the BM too, as demonstrated by reduced AchE and ChAT protein levels in the BM of SHR compared to the WKY rats, it may be that this also contributes to dysfunctional BM cell activity in the SHR. This is consistent with previous data showing that stimulation of the vagus nerve ameliorates experimental inflammatory diseases^{7, 24}. However, further experiments, perhaps with the use of genetically modified animals where chimeric mice are generated by adoptive BM transfer from the nicotinic AchR KO mice, are needed to elaborate the involvement of parasympathetic influence in the BM in hypertension. We also recognize a limitation of this study in that it does not distinguish the effects of the BM NE, delivered directly via the increased BM fSNA, from the peripheral NE delivered via the BM blood vessels. Availability of adrenergic receptor KO mice may be useful in chimeric experiments to address this issue in the near future.

In summary, we propose that there is a loss of circadian regulation of the BM cells due to dysfunctional sympathetic/adrenergic mechanisms in the BM, resulting in chronically high ICs and low EPCs in the SHR. This study raises key questions: do changes in the BM sympathetic drive precede the development of high BP, or are they involved in the establishment of the hypertensive pathophysiology? Although both animal and human studies implicate pre-hypertensive elevation in the sympathetic drive^{3, 25}, further experiments to measure changes in fSNA prior to development of high BP are warranted. Nonetheless, the present study demonstrates a dysfunctional BM activity in hypertension, which is associated with changes in fSNA.

Perspectives

In this study, we present the first direct evidence for an impaired circadian regulation of the BM cell activity in the SHR. We present the hypothesis that an impaired sympathetic input to the BM promotes imbalance in the BM ICs and EPCs, which may result in an increased inflammatory-dependent vascular injury, and compromise the vascular repair in hypertension. This hypothesis is supported by the following: (1) enhanced functional neural connections between the pre-sympathetic brain regions and the BM in hypertension; (2) NE directly activates the BM ICs, which is attenuated by pre-application of Ach; (3) elevated TH and NE in the BM of SHR; (4) circadian control of the BM activity was impaired in the SHR, which exhibited increased ICs and decreased EPCs compared to the WKY, due to chronic downregulation of adrenergic receptor levels in the SHR BM. Thus, targeting the sympatho-adrenergic mechanisms the BM presents a novel strategy for consideration in neurogenic hypertension.

Novelty and Significance.

What is relevant?

• Increased inflammation and reduced vascular repair are hallmarks of hypertension and cardiovascular diseases.

• Increased sympathetic drive contributes to inflammation by mobilizing the inflammatory cells from the spleen in hypertension, and from the BM in myocardial infarction.

• The anti-inflammatory effects of the vagus nerve stimulation are demonstrated by lowered levels of the inflammatory cytokines and suppressed activation of inflammatory cells in inflammatory diseases.

What is new?

• This article presents direct evidence of altered sympathetic drive to the BM, which is associated with the dysfunctional BM-derived EPCs and ICs in the rat model of neurogenic hypertension.

• Retrograde labeling using GFP-labeled pseudorabies virus shows increased neuronal communication between the brain pre-sympathetic nuclei and the BM in the SHR.

• Increased sympathetic drive to the BM is associated with increased BM TH and NE. Decreased ChAT and AchE suggests impaired parasympathetic influence to the BM. Furthermore local delivery of NE to the BM increases the mobilization of the inflammatory cells in vivo which can be antagonized by similar localized delivery of acetylcholine.

Summary.

The study is the first direct evidence of elevated sympathetic drive to the BM in hypertension, which is associated with decreased BM-derived EPC counts and function, and increased BM-derived IC counts and mobilization. We suggest that repairing the balance in the EPCs and ICs presents a novel antihypertensive target in drug resistant hypertension.

Non-standard Abbreviations and Acronyms

DAPR

Decerebrated artificially-perfused rat

EPCs

endothelial progenitor cells

ICs

Inflammatory cells

BM

bone marrow

PVN

paraventricular nucleus

SFO

subfornical organ

IO

inferior olive

RVLM

rostral ventrolateral medulla

NTS

nucleus of the solitary tract

fSNA

femoral sympathetic nerve activity