Mineralocorticoid Receptor Antagonism Improves Transient Receptor Potential Vanilloid 4-dependent Dilation of Cerebral Parenchymal Arterioles and Cognition in a Genetic Model of Hypertension

Abstract

Objective:

In a model of secondary hypertension, mineralocorticoid receptor (MR) antagonism during the development of hypertension prevents the impairment of transient receptor potential vanilloid 4 (TRPV4) activation in parenchymal arterioles (PAs) and cognitive impairment. However, it is unknown whether MR antagonism can improve these impairments when treatment begins after the onset of essential hypertension. We tested the hypothesis that MR activation in stroke-prone spontaneously hypertensive rats (SHRSP) leads to impaired TRPV4-mediated dilation in PAs that is associated with cognitive dysfunction and neuroinflammation.

Methods:

20–22-week-old male SHRSP ± eplerenone (EPL; 100mg/kg daily for 4 weeks) were compared to normotensive Sprague Dawley (SD) rats. Pressure myography was used to assess PA function. Cognition was tested using Y-maze. Neuroinflammation was assessed using immunofluorescence and qRT-PCR.

Results:

Carbachol-mediated endothelium-dependent dilation was impaired in SHRSP, and MR antagonism improved this without affecting myogenic tone. Dilation to TRPV4 agonist GSK1016790A was impaired in SHRSP, and ELP treatment restored this. IK_Ca/SK_Ca-mediated dilation was impaired by hypertension and unaffected by EPL treatment. TRPV4 and IK_Ca/SK_Ca channel mRNA expression were reduced in PAs from hypertensive rats, and EPL did not improve this. Impairments in PA dilation in SHRSP were associated with cognitive decline, microglial activation, reactive astrogliosis, and neuroinflammation; cognitive and inflammatory changes were improved with MR blockade.

Conclusions:

These data advance our understanding of the effects of hypertension on cerebral arterioles using a clinically relevant model and treatment paradigm. Our studies suggest TRPV4 and the MR are potential therapeutic targets to improve cerebrovascular function and cognition during hypertension.

Condensed Abstract:

MR activation is associated with cerebrovascular damage and cognitive decline. Using a model of essential hypertension and a clinically relevant treatment paradigm, we found that MR antagonism improves TRPV4-mediated dilation in cerebral arterioles, improves cognitive function, and reduces neuroinflammation during hypertension. These data suggest treatment with MR antagonists may be a beneficial intervention to improve cerebrovascular and cognitive function in hypertensive patients.

Introduction

Mineralocorticoid receptor (MR) activation has been linked to vascular damage in hypertension [1–4] and stroke [5,6] and is associated with cognitive and memory impairment in rodents [7] and humans [8]. In a mouse model of angiotensin II (AngII)-hypertension, we showed that MR activation during the development of hypertension impairs endothelium-dependent dilation and causes inward remodeling of cerebral parenchymal arterioles (PAs); these changes were associated with cognitive decline [7,9].

The PAs regulate blood flow from pial arterioles to the capillary bed, where gas and nutrient exchange occurs. Unlike pial arteries, PAs lack collateral connections and are considered the weak link in cerebral perfusion. Strikingly, occlusion of a single PA results in a microinfarct that is sufficient to cause cognitive decline [10,11]. The brain is a metabolically demanding organ that does not have energy stores; therefore, tightly regulated blood flow is an absolute requirement for proper function. Modifiable vascular risk factors such as hypertension disrupt cerebral hemodynamics and lead to cerebral hypoperfusion [12,13]. Chronic cerebral hypoperfusion is associated with cognitive decline because it prevents neurons from receiving the necessary nutrients [14–17]. In fact, appropriate vascular function is so vital for brain health that the severity of hypoperfusion can predict the rate of cognitive decline in patients with Alzheimer’s disease [18]. Our studies in AngII-hypertensive mice confirm the importance of vascular function in cognition, as the impaired PA endothelium-dependent dilation observed is associated with cognitive dysfunction [7].

We have shown that PA endothelium-dependent dilation is primarily mediated by transient receptor potential vanilloid 4 (TRPV4) channels and that this dilator pathway is impaired in mice with AngII-hypertension [7,19,20]. Ca²⁺ influx through TRPV4 triggers the activation of intermediate and small conductance Ca²⁺-activated potassium channels (IK_Ca/SK_Ca) in endothelial cells [21,22]. IK_Ca/SK_Ca activation produces a hyperpolarizing current that travels to vascular smooth muscle cells through myoendothelial gap junctions to cause vasodilation. The importance of TRPV4 in regulating cerebrovascular health is highlighted in our recent studies using a global TRPV4 knockout rat model. These rats exhibit cerebral hypoperfusion and impaired cognitive function [19].

Our previous studies implicating the MR and TRPV4 in hypertension-associated PA dysfunction were conducted in a model of secondary hypertension using angiotensin II, and the MR antagonists were administered as hypertension was developing [7,9]. Administration of pharmacological therapy alongside hypertension development does not reflect the timing of treatments in humans, where therapy would only be administered after hypertension is detected. This is a significant limitation of our previous studies and prevents us from knowing if hypertension-associated dementia is treatable or merely preventable. Our goal was to determine if cerebrovascular and cognitive impairments associated with hypertension can be improved with MR antagonism after the onset of hypertension. The current study utilized a rat model of polygenic multifactorial hypertension that accurately reflects human essential hypertension. We also used a clinically relevant treatment regime in which MR antagonist administration began after the development of hypertension. We tested the hypothesis that MR activation impairs PA TRPV4-mediated dilation in SHRSP, that these impairments are associated with impaired cognitive function and neuroinflammation, and that MR antagonism improves these impairments.

Materials and Methods

Experimental Models and Treatment

All experimental protocols were approved by the Michigan State University Animal Care and Use Committee and performed according to the National Institutes of Health Guide for the Care and Use of Laboratory Animals. Male 20–22-week-old stroke-prone spontaneously hypertensive rats (SHRSP; n=15–16 total for each group) from the colony housed at Michigan State University were used. All rats were given standard rodent chow containing 0.3% sodium (Envigo, diet 8904). A group of SHRSP were treated with the MR antagonist eplerenone (EPL; 100mg/kg/day), given orally suspended in 2g peanut butter daily for 4 weeks. This amount of peanut butter does not significantly alter the amount of dietary sodium. Treatment began when rats were 16–18 weeks of age. Age-matched male Sprague Dawley (SD) rats (n=16) purchased from Envigo (Indianapolis, IN) served as control. We used SD rats as the control for the SHRSP because previous studies found biological variability in Wistar Kyoto (WKY) rats that results in variable blood pressures [23]. WKY rats also express depressive behavior that confounds behavior analyses [24,25]. Though they do not share the same genetic background, SDs are commonly used as a normotensive control for SHRSP [26–28]. All animals studied were male and housed on 12h:12h light/dark cycle with food and water ad libitum.

Tail-Cuff Plethysmography

Blood pressure was measured in conscious rats by tail-cuff plethysmography using a RTBP1001 tail-cuff blood pressure system (CODA-6, Kent Scientific, Torrington, CT). Rats (n = 8) were randomly selected for blood pressure analysis. All rats were acclimatized to both handling and the blood pressure measurement system ahead of the experimental measurements. Of the 25 cycles measured, the first 15 were used as acclimation, and the remaining 10 were averaged to provide the final data point for each animal.

Laser Speckle Contrast Imaging

Pial cerebral blood flow was measured by laser speckle contrast imaging in n=8 rats randomly selected per group; variable numbers are due to the removal of an outlier identified by Grubb’s test and the death of one SHRSP under isoflurane. Rats were anesthetized under 2% isoflurane before imaging, and a 2cm vertical incision was made through the skin and connective tissue to expose the skull. The closed skull was cleaned using a saturated ferric chloride solution followed by hydrogen peroxide. To improve the resolution of the image, clear nail polish was applied to the surface of the skull. Before beginning the measurements, isoflurane was reduced to 1%, and the rats were equilibrated for 5 minutes. Images were then acquired for 1 minute at a rate of 22 images/s. Images were analyzed using PIMSoft software (PeriMed, Las Vegas, NV). Mean flux values were measured in regions of interest defined in each brain hemisphere between the bregma and lambda sutures.

Pressure Myography

The endothelial function of PAs was assessed by pressure myography as described previously by our lab [2,9,29,30]. A 5 × 3 mm section of the brain containing the middle cerebral artery (MCA) was dissected to isolate the arterioles. The pia with the MCA was separated from the brain, and the PAs branching off the MCA were used for experiments. A total of n=15–16 rats were used to complete dilation and myogenic tone experiments; each arteriole was used for only one concentration-response curve. Isolated arterioles were cannulated using two glass micropipettes in a custom-made cannulation chamber. PAs were equilibrated in physiological salt solution (PSS) containing 140mmol/L NaCl, 5mmol/L KCl, 1.8mmol/L CaCl₂, 1mmol/L MgCl₂, 10mmol/L HEPES, and 10mmol/L glucose. A servo-null system was used to pressurize the arterioles, and a leak test was performed before each experiment. Arterioles were pressurized to 60mmHg [2,20,29] until the development of stable myogenic tone (% tone = [1-(active lumen diameter/passive lumen diameter)] × 100. Arterioles that generated less than 20% myogenic tone were discarded. The diameter of the arterioles was recorded using MyoView 2.0 software (Danish Myo Technology, Aarhus, Denmark).

Parenchymal Arteriole Vasodilation

Endothelium-dependent vasodilation was assessed after myogenic tone generation by incubating the arterioles with increasing concentrations of the muscarinic receptor agonist carbachol (CCh; 10⁻⁹-10⁻⁵ mol/L) added to the bath. To determine the role played by TRPV4 in PA dilation, arterioles were incubated with the TRPV4 antagonist GSK2193874 (10⁻⁷mol/L) for 10 minutes after the development of stable myogenic tone, then CCh-induced dilation (10⁻⁹-10⁻⁵ mol/L) was assessed. A separate group of PAs were incubated with increasing concentrations of the TRPV4 agonist GSK1016790A (10⁻⁹-10⁻⁵ mol/L). To evaluate the role of IK_Ca/SK_Ca channels in PA dilation, arterioles were incubated with increasing concentrations of the agonist NS309 (10⁻⁹-10⁻⁵mol/L) in the bath.

TRPV4 Channels in Myogenic Tone

To assess the role of TRPV4 specifically in endothelial cells in the maintenance of myogenic tone, the endothelium of PAs from SHRSPs was removed by passing an air bubble through the lumen of the arteriole. The PAs were then pressurized, and after the development of stable myogenic tone, we confirmed endothelium denudation by incubating the arterioles with CCh (10⁻⁴ mol/L) for 10 min. Successfully denuded arterioles did not dilate in response to CCh. PAs without endothelium were incubated with 10⁻⁷ mol/L GSK2193874 for 10 minutes, and changes in myogenic tone maintenance were recorded. At the end of each experiment, Ca²⁺ free buffer containing ethylene glycol tetraacetic acid (EGTA; 29mmol/L) and sodium nitroprusside (SNP; 10⁻⁵mol/L) were added to the bath to maximally relax the smooth muscle. Dilation was calculated using the following formula: [lumen diameter at drug concentration – baseline lumen diameter) / (passive lumen diameter – baseline lumen diameter)] × 100.

qRT-PCR

Brain tissue from rats used in pressure myography experiments was saved for qRT-PCR. RNA was extracted from whole brain tissue anterior to the MCA, excluding the olfactory bulb, as well as from rat PAs for qRT-PCR analysis using Trizol. RNA was reverse transcribed using VILO reverse transcriptase (Invitrogen, Carlsbad, CA). TAQMAN-specific probes were used for the PCR to assess the mRNA expression of TRPV4, IK_Ca, and SK_Ca in pooled PAs (~10 PAs per rat) and tumor necrosis factor alpha (TNF-α), tumor necrosis factor receptor (TNFR1), interleukin 6 (IL-6), synaptophysin (SYP), sortilin-related receptor 1 (SORL1), and amyloid precursor protein (APP) in brain tissue. mRNA expression is expressed as the fold change from control using the 2^−ΔΔCt method. β2-microglobulin was used for normalization.

Y-Maze

Y-maze was used to assess spatial recognition memory. The Y-maze consists of three identical arms placed at a 120° angle to one another. Unique spatial cues made of colored laboratory tape were placed at the far end of the wall. Each rat received two trials. In the first trial, rats were placed in the maze with one arm of the maze blocked off and were allowed to explore for 5 minutes. The blocked arm was rotated within the testing groups to prevent arm placement bias. After exploration, the rats were placed back in their home cages for a 60-minute retention time. In the second trial, all three maze arms were accessible, and rats were allowed to explore for 5 minutes. The first 2 minutes of this trial were used to assess spatial recognition memory, and the full 5 minutes were used to assess any locomotor differences between groups. Spatial recognition memory was measured by the percentage of time spent exploring the novel arm and the number of visits to the novel arm. Recordings were tracked and analyzed using EthoVision XT software.

Immunofluorescence

Brains from the rats used for the pressure myography experiments were also used for microglia and astrocyte analysis. Because the PA studies used arterioles branching off the MCA, brains for immunofluorescence studies were sectioned beginning about 3mm posterior to the MCA, and images were taken from the cortex. Brains were post-fixed in 4% paraformaldehyde for 48hrs, washed in 1x PBS (24hrs each), and stored in 20% sucrose-PBS until sectioned. The brains were sliced into 40μm sections for analysis of the cortex. For microglia quantification, free-floating sections were blocked and permeabilized in 0.1% Triton X-100 with 10% normal horse serum-PBS for 30min at room temperature, then incubated in 1:200 rabbit anti-ionized Ca²⁺-binding adapter molecule-1 (IBA-1, PA5–27436, Invitrogen, Rockford, IL) in blocking buffer overnight at 4°C. Sections were washed three times in 1x PBS and then incubated in secondary AlexaFluor 568 donkey anti-rabbit (A10042, Invitrogen, Rockford, IL). For astrocyte analysis, free-floating brain sections were blocked and permeabilized in 0.1% Triton X-100 with 10% normal horse serum-PBS for 30min at room temperature, then incubated in 1:1000 rabbit anti-glial fibrillary acidic protein (GFAP; ab7260, Abcam, Cambridge, MA) in blocking buffer overnight at 4°C. The sections were washed three times in 1x PBS and then incubated for 1hr in secondary AlexaFluor 568 donkey anti-rabbit (A10042, Invitrogen, Rockford, IL). Sections were washed 3 × 5min in 0.1% Tween-20 in 1x PBS, then mounted with Vectashield mounting medium. For each subject, two z-stack images were acquired in each hemisphere of the cortex immediately dorsal to the CA1 region of the hippocampus using a 20x objective coupled to a Zeiss LSM 880 confocal microscope. Iba-1- and GFAP-stained sections were analyzed by FIJI software (ImageJ, NIH). Representative astrocytes from GFAP-stained sections were reconstructed then analyzed using the Simple Neurite Tracer plugin described by others [31]. Briefly, z-stack images were loaded into FIJI, and the Simple Neurite Tracer was used to reconstruct individual astrocytes. Morphological analyses include process length and astrocyte volume using the volume filler application. All quantifications and analyses were conducted by an investigator blinded to study groups.

Plasma Aldosterone

Blood was collected by cardiac puncture prior to euthanasia in anesthetized rats. Plasma aldosterone levels were measured by ELISA (ADI-900–173, Enzo Life Sciences).

Drugs and Chemicals

GSK1016790A and GSK2193874 were purchased from Cayman Chemicals (Ann Arbor, MI). All other drugs and chemicals are obtained from Sigma Aldrich unless otherwise specified.

Statistical Analysis

All data are presented as means ± SEM. For analysis of artery vasodilation, two-way analysis of variance with repeated measures in one factor (pressure) was utilized followed by Bonferroni-adjusted t-tests for post-hoc comparisons. All other statistical analyses were assessed by One-Way Analysis of Variance or Kruskal-Wallis if non-parametric testing was required. The Grubb’s test was used to identify outliers. All statistical analyses were performed using GraphPad Prism 7.0 software (GraphPad, San Diego, CA). In all cases, statistical significance was denoted by p<0.05.

Results

Mineralocorticoid receptor antagonism does not affect blood pressure or cerebral pial artery blood flow in SHRSP

SD rats were chosen as the normotensive control for SHRSP in this study. Plasma aldosterone levels were measured to confirm they fell within expected ranges (SD: 196.5 ± 28.4; SHRSP: 545.8 ± 135; SHRSP + EPL: 775.1 ± 116.5pg/mL; data not shown). These measurements reflect what has been observed by others [32–38]. To confirm the SHRSP were hypertensive compared to SD rats and that EPL treatment does not alter blood pressure, systolic blood pressure was measured using tail-cuff plethysmography. Systolic blood pressure in SHRSP was significantly higher than controls (Figure 1A), and MR antagonism did not lower blood pressure. Hypertension is associated with cerebral hypoperfusion, and MR antagonism rescued blood flow impairments in our previous study [9]. Laser speckle contrast imaging was used to measure pial blood flow. SHRSP had a trend toward reduced pial artery perfusion compared to controls (Figure 1B; p=0.0633 vs. control by one-way ANOVA). Treatment with EPL did not improve pial artery perfusion.

Figure 1. Systolic blood pressure is elevated and pial blood flow is reduced in SHRSP compared to control, and this is not changed by MR antagonism.

A) Blood pressure in each group was measured by tail-cuff plethysmography. Data are presented as means ± SEM. SHRSP rats had elevated systolic blood pressure compared to SD controls. MR blockade did not affect blood pressure. B) Pial blood flow was measured by laser speckle contrast imaging. SHRSP rats had a trend toward reduced pial blood flow compared to controls (ANOVA p=0.0497). EPL treatment did not improve pial blood flow. *=different from control by one-way ANOVA with Bonferroni correction for multiple comparisons.

Mineralocorticoid receptor antagonism improves impaired endothelium-dependent dilation during hypertension

The role of MR activation in the generation of myogenic tone in PAs was assessed. The PAs from SHRSPs had increased myogenic tone compared to control. MR antagonism did not alter the amount of tone generated by PAs from SHRSP (Figure 2A); this confirms our published studies [2]. We then assessed endothelium-dependent dilation in response to CCh. The CCh-mediated dilation was impaired in the PAs from SHRSP, which was improved after EPL treatment (Figure 2B). TRPV4 antagonism with GSK2193874 (10⁻⁷mol/L) inhibited the CCh-induced dilation of PAs from all groups (Figure 3A–C), indicating an important role of TRPV4 activation in PA endothelium-dependent dilation in SHRSP and SD rats.

Figure 2. MR blockade improves PA endothelium-dependent dilation in rats with essential hypertension.

Myogenic tone and endothelium-dependent dilation in isolated PAs were assessed by pressure myography. Data are presented as means ± SEM. A) The PAs from SHRSP had increased myogenic tone and this was not prevented by EPL treatment. B) The CCh-mediated dilation was impaired in SHRSP, and this was prevented by EPL treatment. In Figure 2A n= number of vessels from 7–8 rats. *=different from control. Comparisons for Figure 2A made by Kruskal-Wallis and for Figure 2B by two-way ANOVA.

Figure 3. TRPV4 channels are critical regulators of PA dilation.

The role of TRPV4 in PA endothelium-dependent dilation was assessed by pressure myography. PAs were incubated with the TRPV4 antagonist GSK2193874 for 10 mins after the generation of myogenic tone. Data are presented as means ± SEM. TRPV4 inhibition with GSK2193874 prevented the CCh-mediated dilation in A) Sprague Dawley, B) SHRSP, and C) SHRSP+EPL. *=different from control by two-way ANOVA.

Mineralocorticoid receptor antagonism improves impaired TRPV4-mediated dilation but does not impact IK_Ca/SK_Ca-mediated dilation or TRPV4, IK_Ca, and SK_Ca channel mRNA expression during hypertension

To explore the role of MR signaling in endothelium-derived hyperpolarization (EDH)-mediated dilation, we assessed TRPV4-dependent dilation in PAs. The TRPV4 agonist, GSK1016790A, caused a robust dilation in PAs from SD rats (Figure 4A). Hypertension impaired the TRPV4-mediated dilation, and EPL treatment improved this response. The influx of Ca²⁺ through TRPV4 triggers the activation of nearby IK_Ca/SK_Ca channels, producing a K⁺ efflux that hyperpolarizes the endothelial cell, ultimately causing vasodilation. To test IK_Ca/SK_Ca-dependent dilation, PAs were incubated with increasing concentrations of the IK_Ca/SK_Ca channel agonist NS309 in the bath. At lower concentrations of NS309, SHRSP had impaired IK_Ca/SK_Ca-mediated dilation compared to SD controls, demonstrated by a greater EC50 value in SHRSP +/− EPL compared to controls (control: −8.148±0.54, SHRSP: −6.508±0.10, SHRSP+EPL: −6.389±0.11 logEC50; p=0.009 by one-way ANOVA). This was not improved by MR antagonism (Figure 4B); however, the maximum dilation produced in response to NS309 was similar in all groups. We assessed the mRNA expression of TRPV4, IK_Ca, and SK_Ca channels in isolated PAs using qRT-PCR. Our data suggest that the mRNA expression for TRPV4, IK_Ca, and SK_Ca was reduced in PAs from SHRSP compared to SD rats, and EPL treatment had no statistically significant effect (Figure 4C–E).

Figure 4. MR blockade prevents impaired TRPV4-mediated dilation of PAs, but not reduced mRNA expression of ion channels during hypertension.

PA endothelium-dependent dilation mediated by TRPV4 and IK_Ca/SK_Ca channels was assessed by pressure myography. mRNA expression of these channels was assessed by qRT-PCR. Data are presented as means ± SEM. A) The TRPV4 agonist caused dilation in the PAs from control rats. This dilation was impaired in SHRSP, and impairment was prevented by MR blockade with EPL. B) PA dilation to IK_Ca/SK_Ca agonist NS309 was impaired in SHRSP compared to controls, which was not improved by MR antagonism with EPL. C–E) The mRNA expression of TRPV4, IK_Ca and SK_Ca channels was reduced in the PAs from SHRSP; these changes were not prevented by EPL treatment. 2BM was used as the housekeeping gene. *=different from control and SHRSP+EPL by two-way ANOVA in Figure 4A–B. *=different from control by one-way ANOVA in Figure 4C–E.

Endothelial TRPV4 activity modulates parenchymal arteriole myogenic tone generation

Our previous studies in AngII-hypertensive mice showed that TRPV4 inhibition with GSK2193874 reduced myogenic tone in the PAs from AngII-treated mice but not normotensive mice [7]. The same loss of myogenic tone was observed in the PAs from SHRSP and SHRSP+EPL but not in the normotensive SD rats. To explore if endothelial TRPV4 was involved in the loss of myogenic tone, we removed the endothelium from a group of PAs isolated from SHRSP. When endothelium-denuded PAs were incubated with GSK2193874, there was no loss of myogenic tone, suggesting that endothelial TRPV4 is involved in regulating myogenic tone in SHRSP (Figure 5).

Figure 5. Endothelial TRPV4 channel modulate PA myogenic tone generation.

The role of TRPV4 channels in myogenic tone generation was assessed by pressure myography. Data are presented as means ± SEM. TRPV4 inhibition with GSK2193874 resulted in a significant loss of myogenic tone in the PAs from SHRSP and SHRSP+EPL, but not in the normotensive controls. Denuding the endothelium of PAs from SHRSPs prevented the loss in myogenic tone after TRPV4 inhibition. *= different from control by one-way ANOVA.

Mineralocorticoid receptor antagonism improves cognitive function in SHRSP

We used the Y-maze to assess the role of MR activation in cognition in SHRSP. SHRSP spent less time in the novel arm of the Y-maze and had fewer total novel arm visits compared to SD rats, indicating impairment of spatial memory (Figure 6A, B). Notably, SHRSP traveled less total distance during the testing period than controls. MR antagonism did not alter the total distance traveled by SHRSP, but it improved the total time spent in the novel arm and the number of visits to the novel arm, indicating improvement of spatial memory (Figure 6C).

Figure 6. SHRSP have impaired spatial memory that is rescued by MR antagonism.

The Y-maze was used to measure spatial memory. Data are presented as means ± SEM. (A) SHRSP explored the novel arm in the Y-maze significantly less than controls. EPL treatment improved novel arm exploration. (B) SHRSP had fewer total visits to the novel arm compared to control, which was improved with EPL treatment. *=different from control by one-way ANOVA.

SHRSP have elevated mRNA expression of proinflammatory markers and increased activated microglia, which are reduced by MR antagonism

We assessed changes in proinflammatory markers in whole brain tissue from the frontal region of the brain. The mRNA expression of the proinflammatory markers TNF-α and its receptor, TNFR1, were significantly increased in SHRSP compared to SD rats, and these differences were eliminated by EPL treatment (Figure 7A, B). IL-6 mRNA expression was also elevated in SHRSP, but this was not significantly reduced by EPL (Figure 7C). Next, we used immunofluorescence to quantify and assess the quantity and morphology of cortical microglia labeled with an Iba-1 antibody. SHRSP had more microglia compared to control (Figure 7G). The microglia in the SHRSP also had larger somas, suggesting these microglia were in their activated state (Figure 7H) [39]. EPL treatment did not change the number of microglia, but did reduce their soma size.

Figure 7. MR blockade prevents increased brain mRNA expression of genes associated with neuroinflammation and cognition in hypertension.

qRT-PCR was used to measure mRNA expression of genes associated with inflammation and cognitive function. Data are presented as means ± SEM. The brain mRNA expression of the inflammatory markers A) TNF-α and B) its receptor TNFR-1 were elevated in SHRSP compared to control and this was prevented by EPL treatment. C) The inflammatory marker IL-6 was also increased in SHRSP but EPL treatment did not prevent this increase. D) APP mRNA expression was elevated and E) SORL1, a protein involved in APP processing, trended toward an increase in SHRSP (p=0.0612 vs. control). These changes were prevented by MR antagonism. The number of ionized Ca²⁺-binding adapter molecule 1 (IBA-1)-positive microglia was quantified in the cortex. F) Representative images of IBA-1 staining. G) The number of IBA-1-positive cells was elevated in SHRSP compared to SD controls; this was unchanged by MR antagonism. H) IBA-1-positive cells in the cortex of SHRSP had greater soma sizes compared to SD controls and this was prevented by MR antagonism. Scale bar=100μm *=different from control by one-way ANOVA or Kruskall-Wallis. Abbreviations: TNF-α=tumor necrosis factor α, TNFR1=tumor necrosis factor receptor 1, IL-6=interleukin-6, APP=amyloid precursor protein, SORL1=sortilin-related receptor 1

Mineralocorticoid receptor antagonism corrects the elevated mRNA expression of markers associated with cognitive dysfunction in SHRSP

We also measured the mRNA expression of proteins involved in cognition. mRNA expression of amyloid precursor protein (APP) and sortilin-related receptor (SORL1), a protein involved in APP processing, were increased in SHRSP compared to control; these differences also were reduced by EPL treatment (Figure 7D, E).

Reactive astrogliosis in SHRSP is reduced with MR antagonism

Astrocytes play a critical role in relaying information between the cerebral vasculature and the neurons and are therefore important in maintaining cognitive function. A GFAP antibody was used to label astrocytes in the cortex near the corpus callosum. SHRSP had more astrocytes compared to control (Figure 8B). Astrocytes from SHRSP also had greater soma and process thickness, referred to as astrocyte volume, than control (Figure 8C). A greater astrocyte volume indicates that the cells are reactive. MR antagonism reduced the overall number of astrocytes and improved their morphology.

Figure 8. MR antagonism prevents astrogliosis observed in SHRSP.

The number of glial fibrillary acidic protein (GFAP)-positive astrocytes were quantified in the cortex. Data are presented as means ± SEM A) Representative images of GFAP staining. B) The total number of astrocytes was elevated in the cortex of SHRSP compared to control; this was prevented by EPL treatment. C) Astrocytes in the cortex of SHRSP had increased volume, which was prevented by EPL treatment. D) The total process length of astrocytes in the cortex of SHRSP treated with EPL was elevated compared to control. Scale bar=100μm. *=different from control by one-way ANOVA.

Discussion

Our goal was to examine the role of MR activation in cerebral arteriolar function, cognition, and neuroinflammation in a hypertensive model using a clinically relevant treatment paradigm. We used SHRSP to show that: 1) MR activation is involved in the impairment of TRPV4-mediated PA endothelium-dependent dilation observed in hypertension; 2) Endothelial TRPV4 channels modulate myogenic tone during hypertension; 3) MR antagonists, delivered after the onset of hypertension, reduce hypertension-associated cognitive dysfunction, neuroinflammation, and reactive astrogliosis. These new data complement and expand on our previous findings in a mouse model of AngII-hypertension and improve our understanding of the role of vascular MR signaling in cerebral arteriolar and cognitive function. These studies suggest MR antagonists may be an effective pharmacological tool to mitigate the increased risk of vascular cognitive impairment and dementia development associated with hypertension. Hypertension-associated dementia is a significant public health concern. Midlife hypertension significantly increases the risk of dementia later in life, with a 20% increased dementia risk for every 10mmHg of systolic blood pressure elevation [40]. The current studies were designed to reflect this temporal development of hypertension with treatment beginning just before midlife in the SHRSP given their shortened life span [41].

MR activation plays a significant role in the development of hypertension; it is estimated that up to 60% of patients with essential hypertension have hyperaldosteronism [42]. Excessive MR activation is associated with endothelial dysfunction and heart failure [43,44]. Elevated MR activation leads to increased inflammation, and oxidative stress is associated with endothelial dysfunction via impaired nitric oxide (NO)-mediated dilation in peripheral arteries [45,46]. Less is known about the molecular mechanisms related to MR-induced cerebrovascular damage, particularly in vessels like the PAs that are not dependent on NO for dilation.

MR activation has been implicated in cerebrovascular injury in two stroke models [6,47]. MR activation also regulates cerebral artery remodeling during hypertension; we have shown that MR antagonism prevents and reverses the inward remodeling of middle cerebral arteries and PAs in SHRSP [2–4]. Using a mouse model of AngII-hypertension, we have shown that MR activation results in impaired cerebral PA dilation, inward hypotrophic remodeling, and reduced cerebral perfusion; these changes were associated with impaired cognitive function [7,9]. Data presented here support the hypothesis that MR antagonism protects against vascular and cognitive dysfunction during hypertension.

Using tail-cuff plethysmography, we confirmed that SHRSP have significantly elevated systolic blood pressure compared to SD rats and that EPL treatment did not affect blood pressure. This mimics our findings from previous studies showing that MR antagonism does not lower blood pressure in AngII-hypertensive mice [7,9] or SHRSP [2,4], including one utilizing radiotelemetry, the gold standard for blood pressure measurement [3]. Therefore, the effects of MR antagonism presented here are independent of blood pressure.

Cerebral hypoperfusion increases the risk of cognitive impairment and dementia [15,17]. Studies from our lab show that hypertension reduces cerebral pial artery perfusion in AngII-hypertensive mice, and MR antagonism, administered while the hypertension is developing, prevents this [9]. We used laser speckle contrast imaging to measure pial blood flow in the current study. We observed that SHRSP had a trend toward reduced pial blood flow compared to controls. The p-value for the one-way ANOVA was p=0.0497, but statistical significance was lost after application of Bonferroni correction for multiple comparisons. In contrast to our previous studies, EPL treatment did not improve pial flow. The difference in these findings could be due to the difference in species and the model of hypertension used. In our previous study using AngII hypertensive mice, the EPL and AngII administration occurred simultaneously. This contrasts with the current study, SHRSP develop hypertension by 6 weeks of age, and EPL treatment did not begin until the rats were 16–18 weeks of age [48]. Technical limitations may also have contributed to the lack of an observed effects of EPL. Measurements were made through a closed skull, which significantly reduces the image resolution. The laser penetration for the PeriMed system is 1mm; thus, we were only able to measure flow in the pial arterioles. It is possible that the changes we observed in PA function caused reduced flow in the deeper brain structures; MRI would be required to make these measurements, which is outside the scope of these studies.

PAs play a critical role in regulating cerebrovascular resistance through their ability to generate significant myogenic tone. The myogenic tone in PAs from SHRSP was increased relative to SD rats, and this was not reversed with EPL treatment. We produced similar results in previous studies that utilized WKY rats as controls for SHRSP [2]. We acknowledge these data are at odds with an earlier study showing spironolactone treatment increased myogenic tone in SHRSP [3]. That study was conducted in middle cerebral arteries from twelve-week-old rats treated as hypertension developed. The differences in arteries, ages of rats, and treatment regime used are likely behind the observed differences.

Impaired endothelium-dependent dilation in cerebral arteries and arterioles from SHRSP has been reported by others [49–51]. In keeping with our hypothesis, PAs from SHRSP exhibited impaired endothelium-dependent CCh-mediated dilation, which was improved with EPL treatment. In all three experimental groups, PA dilation to CCh was prevented when TRPV4 was blocked with GSK2193874, supporting our previous findings that endothelium-dependent dilation in PAs is TRPV4 dependent [7,19,20]. It is important to note that other groups have proposed that NO also contributes to dilation in PAs [52]. While we cannot explain the differences in our findings, we acknowledge the possibility that EPL treatment improved dilation through a NO-mediated mechanism. Thus, we used a specific TRPV4 agonist (GSK1016790A) to further examine the impact of MR activation in TRPV4-mediated dilation. PAs from SHRSP had markedly impaired TRPV4-mediated dilation, which was reversed by EPL treatment.

TRPV4 activation in endothelial cells produces a Ca²⁺ influx that activates IK_Ca/SK_Ca channels, producing K⁺ efflux and hyperpolarization [53–55]. This EDH crosses to smooth muscle cells via myoendothelial gap junctions to hyperpolarize the smooth muscle cells and produce vasodilation. We used the IK_Ca/SK_Ca agonist NS309 to assess dilation via EDH, which is downstream of TRPV4 activation. While NS309 is an agonist for both IK_Ca and SK_Ca channels, IK_Ca channels are the predominate channels influencing dilation in PAs [21]. The EC50 values suggest the PAs from SHRSP were less sensitive to the effects of NS309. However, the maximum dilation in response to high concentrations of NS309 was similar across the groups. This finding is reflected in our qRT-PCR data from PAs showing that SHRSP +/− EPL had similarly reduced mRNA expression of these Ca²⁺-activated K⁺ channels compared to controls, suggesting the difference in IK_Ca/SK_Ca-mediated dilation is due to hypertension and is not an MR-mediated event.

We also found that TRPV4 mRNA expression is reduced in SHRSP +/− EPL compared to control, despite improvement in TRPV4-mediated dilation. While we did not measure protein expression in the present study, studies using mesenteric arteries from SHRSP have shown reduced protein levels of TRPV4 [56]. Taken together, these data present the possibility that in untreated SHRSP, TRPV4-mediated dilation is impaired at the level of functionality rather than protein expression. Endothelial TRPV4 channels are localized to myoendothelial projections by A-kinase anchoring protein 150 (AKAP150), ensuring their proximity to IK_Ca/SK_Ca channels. A previous study found that the interaction between TRPV4 and AKAP150 is dysregulated during hypertension [57]. EPL treatment may improve the cooperative gating between TRPV4 and AKAP150 to maintain channel function. The lack of an effect of EPL on TRPV4 mRNA expression is at odds with our previous studies in AngII-hypertensive mice that showed that EPL treatment prevented changes in TRPV4 mRNA expression. This could be explained by differences in the rodent and hypertension models used. However, it is also possible that the differences in treatment paradigms, as described previously, are also significant. It is also worth mentioning that a previous study from our group did not identify a role for the MR in the dilation of PAs from SHRSP [2]. However, we used nifedipine, an L-type calcium channel blocker, in that study to measure the dilatory response. L-type calcium channels are the primary source of calcium influx during myogenic tone generation [58]. Directly blocking these channels produces a robust dilation that is not endothelium-dependent.

We previously showed that TRPV4 activation modulates myogenic tone generation in hypertensive mice, and this effect is not modulated by MR activation [7]. The same was true in the current studies. TRPV4 inhibition resulted in a loss of myogenic tone in the PAs from SHRSP. The loss of tone was not observed when the endothelium of the arterioles was removed, suggesting endothelial TRPV4 channels are responsible for the effect. This presents the possibility that under pathological conditions, TRPV4 activates other signaling pathways such as cytosolic phospholipase A₂ (cPLA₂) to modulate vascular tone [59]. However, the mechanisms for the actions of endothelial TRPV4 in myogenic tone maintenance remain to be explored.

Cerebrovascular dysfunction in hypertension is associated with vascular cognitive impairment and dementia development [60–62]. Our previous studies in AngII-hypertensive mice demonstrated that hypertension is associated with cognitive decline that was prevented by MR antagonism [7]. In this study, we used the novel arm Y-maze to identify impairments in spatial recognition memory. This test uses the innate characteristic of rats to preferentially explore novel areas [63]. During the probe trial, SHRSP explored the novel arm significantly less than the control group, and this was improved with EPL treatment. Further, SHRSP had fewer total visits to the novel arm, which was enhanced with EPL. Notably, there was a difference in the total distance traveled between SD controls and SHRSP +/− EPL treatment, with SDs traveling more than SHRSP groups. While EPL treatment improved the exploration time and visitation frequency of SHRSP in the novel arm, it did not impact the total distance traveled. Therefore, it is unlikely that differences in distance traveled played a role in the cognitive improvement in the EPL group. These data indicate that MR antagonism improves PA function during hypertension and prevents the cognitive impairment that occurs alongside vascular dysfunction.

MR activation has been linked to increased inflammation, which, in turn, contributes to hypertension-associated cognitive decline and cerebral small vessel disease [64–66]. In the current study, we showed that the mRNA expression of TNF-α and its receptor, TNFR-1, were significantly increased in the brain from SHRSP, and these differences were reduced with EPL treatment. Others have shown that MR antagonism decreases plasma TNF-α in SDs with heart failure [67,68], but to the best of our knowledge, we are the first to show a role for the MR in regulating the expression of TNF-α and its receptor in brain tissue from SHRSP. IL-6 expression was also elevated in SHRSP; however, this was not improved with EPL treatment. Using immunofluorescence, we showed that SHRSP had more microglia in the cerebral cortex than control. Others have identified a role for the MR in regulating microglia polarity; microglia from mice with myeloid cell-specific MR deletion favored M2 polarity compared to control, indicating MR activation is associated with the inflammatory properties of microglia [69]. It should be noted that these studies were conducted in animals undergoing an experimental model of multiple sclerosis. Thus, the microglia were activated by inflammation associated with autoimmunity. Our data suggest increased inflammation in hypertension has a similar effect on cortical microglia. Microglia from SHRSP had an increased average soma size, indicating a preference for the inflammatory M1 polarity. EPL treatment reduced microglial activation, indicating less neuroinflammation in the EPL-treated group; however, EPL treatment did not impact microglia proliferation in SHRSP.

The presence of extracellular β-amyloid plaque deposits is a requirement for diagnosing Alzheimer’s disease [70]. Hypertension has been found to exacerbate cognitive dysfunction associated with β-amyloid plaque deposition in humans [71] and rodents [72–74] and is also associated with greater plaque accumulation [75]. β-amyloid is a product of its precursor protein, APP, which is processed by both amyloidogenic and non-amyloidogenic pathways. While deposition of β-amyloid plaques is uncommon in rats, especially those within the age range studied here [76], we observed an increase in the expression of the mRNA for APP, which was reversed by EPL treatment. Increased expression of SORL1, a protein involved in several pathways of APP proteolytic processing, was also observed in SHRSP; this also was corrected by EPL treatment. Mutation or dysfunction of SORL1 is a possible contributor to β-amyloid plaque deposition [77,78]. At present, it is unclear if the elevated mRNA expression of APP in SHRSP is directly associated with cognitive impairment in this group. Our data suggest that APP expression is modulated by MR activity, but additional experiments are necessary to determine which pathways of APP processing are active in SHRSP and if this plays a role in cognitive function.

Astrocyte numbers and morphology were assessed by GFAP staining in the cortex. Astrocytes have classically been considered to relay information from neurons to arterioles. However, a novel role for these cells has recently been established, showing bi-directional communication extending from arterioles to neurons [79]. Reactive astrogliosis is observed during increased neuroinflammation and hypertension [80,81]. SHRSP had more cortical astrocytes than control and had astrocytes with increased soma and process volume, which is indicative of astrogliosis. These differences were prevented by MR blockade. SHRSP did not have an elevated summed process length in these astrocytes, but EPL treatment in SHRSP did increase the process length of the astrocytes compared to control. Increased total process length of the astrocytes could enhance connections between PAs and neurons, potentially improving cognitive function during hypertension.

Some limitations in our study must be acknowledged. First, all our studies were conducted in male rats to expand our previous findings to a new model; thus, exploring possible sex differences is outside the scope of the current study, but we will consider this in future studies. In the future, we should also identify the cell-specific actions of MR signaling in PA myogenic tone and vasodilation. It is important to acknowledge the disagreement in the data presented in this paper; EPL treatment improves endothelium-dependent dilation yet does not improve cerebral pial blood flow in SHRSP. The conflict observed here is due to the difference in the areas of the vasculature examined. While PAs dive deeply into the brain, the laser speckle contrast imager can only measure blood flow through the surface arteries and arterioles. In the future, it would be beneficial to analyze blood flow alterations in the hippocampal region to accompany our PA studies; however, this is not possible at present.

In summary, our data confirm the key findings from a previous study in a model of human essential hypertension. PAs are important for the development of cerebral small vessel disease, and we show that MR activation impairs the TRPV4-mediated dilation of PAs during hypertension. TRPV4 may be a critical determinant of the hypertension-associated changes in cerebral arteriolar function. Endothelial TRPV4 channels are modulators of myogenic tone generation and maintenance in the PAs during hypertension. Impairment of vasodilation and alterations in myogenic tone may reduce cerebral blood flow during hypertension and increase the risk of cognitive decline and neuroinflammation in cerebrovascular disease.