Impairment of microvascular endothelial Kir2.1 channels contributes to endothelial dysfunction in human hypertension

Abstract

Hypertension is associated with decreased endothelial function through reduced contributions of nitric oxide (NO). We previously discovered that flow-induced NO production in resistance arteries of mice and humans critically depends on endothelial inwardly rectifying K⁺ (Kir2.1) channels. The goal of this study was to establish whether these channels contribute to the impairment of endothelial function, measured by flow-induced vasodilation (FIV) in peripheral resistance arteries of humans with hypertension. We measured FIV in vessels isolated from subcutaneous fat biopsies from 32 subjects: normotensive [n = 19; 30.6 ± 9.8 yr old; systolic blood pressure (SBP): 115.2 ± 7 mmHg; diastolic blood pressure (DBP): 75.3 ± 5.7 mmHg] and hypertensive (n = 13; 45.3 ± 15.3 yr old; SBP: 146.1 ± 15.2 mmHg; DBP: 94.4 ± 6.9 mmHg). Consistent with previous studies, we find that FIV is impaired in hypertensive adults as demonstrated by a significant reduction in FIV when compared with the normotensive adults. Furthermore, our data suggest that the impairment of FIV in hypertensive adults is partially attributed to a reduction in Kir2.1-dependent vasodilation. Specifically, we show that blocking Kir2.1 with ML133 or functionally downregulating Kir2.1 with endothelial-specific adenoviral vector containing dominant-negative Kir2.1 (dnKir2.1) result in a significant reduction in FIV in normotensive subjects but with a smaller effect in hypertensive adults. The Kir2.1-dependent vasodilation was negatively correlated to both SBP and DBP, indicating that the Kir2.1 contribution to FIV decreases as blood pressure increases. In addition, we show that exposing vessels from normotensive adults to acute high-pressure results in loss of Kir2.1 contribution, as high pressure impairs vasodilation. No effect is seen when these vessels were incubated with dnKir2.1. Overexpressing wtKir2.1 in the endothelium resulted in some improvement in vasodilation in arteries from all participants, with a greater recovery in hypertensive adults. Our data suggest that hypertension-induced suppression of Kir2.1 is an important mechanism underlying endothelial dysfunction in hypertension.

NEW & NOTEWORTHY Impairment of endothelial function under high blood pressure is linked to the loss of inwardly rectifying K⁺ (Kir2.1) channels activity in human resistance arteries, leading to a reduction in flow-induced vasodilation and possibly leading to a vicious cycle between elevation of blood pressure, and further impairment of Kir2.1 function and flow-induced vasodilation.

INTRODUCTION

Hypertension and elevated blood pressure are prevalent diseases that predisposes to heart disease, heart failure, stroke, and kidney failure (1). A major underlying cause of hypertension is dysfunction of the endothelium, an inner lining of the blood vessels that regulates vasodilatory responses through the synthesis of nitric oxide (NO) and other relaxing factors (2, 3). Impairment of endothelium to induce vasorelaxation leads to increased peripheral vascular resistance, enhanced vasoconstriction, and vascular remodeling, which are known to contribute to the development of hypertension (2–4). The hallmark of endothelial function, and an essential mechanism for the control of blood flow to the microcirculation, is flow-induced vasodilation (FIV). FIV depends on the ability of endothelial cells (ECs) to sense mechanical force generated by blood flow and translate it into a biological response of releasing vasorelaxant factors. It is also well known that one of the major mechanisms responsible for FIV is flow-induced activation of a signaling cascade leading to the activation of serine/threonine kinase Akt1, which in turn activates endothelial NO synthase (eNOS), to generate NO (5, 6).

Our earlier studies demonstrated that flow-induced activation of Akt1/eNOS cascade critically depends on flow-sensitive endothelial K⁺ channels, Kir2.1 (7). Kir2.1 belong to a family of inwardly rectifying K⁺ channels that are ubiquitously expressed and are responsible for the maintenance of the resting potential in a variety of cells (8, 9). These channels have been identified in endothelial cells to be one of the earliest responders to flow, which increases channel activity resulting in K⁺ efflux and membrane hyperpolarization (10–12). Previously, we have found that genetic deficiency of Kir2.1 results in severe impairment of FIV in murine and human resistance vessels in an NO-dependent way resulting in increased vascular resistance and blood pressure in mice (7). Furthermore, we also found that hypercholesterolemia (13–15) and obesity (16) result in the suppression of endothelial Kir2.1 and its ability to respond to a mechanical stimulus, which leads to impaired FIV, suggesting that a loss of Kir2.1 is a major factor contributing to endothelial dysfunction in these diseased conditions.

However, it is not known whether microvascular endothelial Kir2.1 function is depressed in adults with elevated blood pressure and hypertension who are not hypercholesterolemic or obese compared with those with normal blood pressure. Therefore, in the present study we test the effect of elevated blood pressure on the contribution of endothelial Kir2.1 to flow-induced vasodilation of human arterioles. Our results show that there is a significant loss in Kir2.1 contribution to FIV in 1) adults with elevated blood pressure and 2) following acute elevations of intraluminal pressure in isolated arterioles from normotensive patients, suggesting that impairment of endothelial Kir2.1 may be one of the key mechanisms that contributes to hypertension.

MATERIALS AND METHODS

Human Subject Recruitment

The study protocol and procedures were approved by the University of Illinois at Chicago Institutional Review Board (IRB No. 2018-1505). The IRB approved the human experimental procedures based on the Declaration of Helsinki and good clinical practice.

Participants included in the study were aged between 18 and 65 yr old, body mass index was lower than 35 kg/m², and waist circumference was under 88 cm in women and 102 cm in men. Participants with a medical history of cardiovascular disease, diabetes, chronic inflammatory or autoimmune diseases, or an active infection were excluded. Participants with blood pressure over 160/99 mmHg and hypercholesterolemia were also excluded. To avoid variations in endothelial function due to the menstrual cycle, women were only included at the follicular phase.

After written informed consent and medical history were obtained, the anthropometric measurements and vitals were assessed. Blood pressure measurements were taken during rest at seated position with an appropriate cuff size on the right brachial artery, at a heart level. Three readings were obtained at 1-min intervals using an automatic oscillometer device (HEM-907XL, Omron Corporation, Japan). The average of the three readings was then calculated. The mean arterial pressure (MAP) was calculated from the measured systolic and diastolic blood pressure (SBP and DBP, respectively) values using the following equation: MAP = DBP + [1/3 × (SBP − DBP)].

Blood samples and subcutaneous gluteal fat biopsy were collected by trained research personnel at the Clinical Research Center at the University of Illinois Chicago. For gluteal fat biopsy, a small incision (∼1 cm) was made to expose the subcutaneous fat just underneath the skin of the gluteal region. Approximately 1 mL of fat tissue was harvested by sharp dissection. The biopsy was transferred to the HEPES buffering solution (pH = ∼7.4 and 4°C) and was used to extract resistance arteries for microvascular function studies.

Flow-Induced Vasodilation Measurement

Isolated microvessels were dissected from the subcutaneous gluteal fat biopsy and transferred to a microscopic apparatus for continuous measurements of vascular diameter in response to flow or pharmacological stimuli, as previously described (7, 13, 17, 18). In brief, human-isolated arterioles (50–200 μm) were cannulated with glass micropipettes in an organ perfusion chamber and perfused with physiological salt solution (pH = ∼7.4) using a peristaltic pump and bubbled with air (5% CO₂-21% O₂) at a temperature of 37°C. The chamber was attached to an inverted microscope (Olympus CKX41) equipped with a camera and attached to a video monitor. Arterial inner diameters were measured using a video-measuring system (Model VIA-100; Boekeler).

Following a 30-min pressurization at 60 cmH₂O, arterial diameter was measured at baseline, after a preconstriction with endothelin-1 (ET-1, Sigma; 120–200 pM) of up to 60% of baseline diameter, and during intraluminal flow corresponding to pressure gradients of Δ10–Δ100 cmH₂O (3 min each, maintaining intraluminal pressure at 60 cmH₂O). The starting concentration of endothelin-1 (ET-1) used in this study was 120 pM for 5 min to a preconstriction of 60% of baseline diameter. The dosage was repeated with increments of 20 pM when no satisfactory preconstriction was reached. The average final concentration for both groups was 140 pM to obtain a 60% constriction from baseline in both groups. Vessels that did not preconstrict at least 50% at maximum dosage of 220 pM/L ET-1 were discarded from the study. Endothelial-independent dilation was assessed at the end of each protocol using papaverine (PAP). Vessels with dilation to PAP lower than 80% were removed from analysis (Table 1). FIV at each pressure gradient was calculated as the percent change using the following equation: FIV (%) = (diameter at each pressure gradient − ET-1 preconstricted diameter)/(baseline diameter − ET-1 preconstricted diameter) × 100. At the end of each protocol, endothelium-independent vasodilation was induced by papaverine (PAP, Sigma; 10⁻⁴ M). To explore the role of Kir2.1 in FIV, all arteries were incubated with either empty, dnKir2.1, or wtKir2.1 expressing adenoviral vectors for 48 h, as described later, before measuring FIV. To further determine the contribution of Kir2.1 and NO in FIV, vessels were exposed for 30 min to ML133 (100 μM; Sigma), Kir2.1 inhibitor (19), and NO synthase inhibitor [N^G-nitro-l-arginine methyl ester (l-NAME), Sigma; 10⁻⁴ M], respectively. The aforementioned FIV protocol was then repeated. Vascular smooth muscle-dependent vasodilation was also assessed by a dose-response curve with sodium nitroprusside (SNP).

Table 1.

Vessel inclusion criteria

	Average Constriction, %
	ET-1	Papaverine
Normotensive
Empty	61.44 ± 3.08	92.08 ± 2.46
dnKir2.1	65.23 ± 3.05	93.37 ± 2.13
wtKir2.1	61.95 ± 2.2	98.14 ± 0.89
Hypertensive
Empty	67.09 ± 2.2	92.63 ± 3.17
dnKir2.1	60.84 ± 2.31	90.07 ± 2.31
wtKir2.1	61.26 ± 1.72	91.74 ± 2.53

Values are means ± SE. ET-1, endothelin-1; Kir2.1, inwardly rectifying K⁺ channels; dn, dominant negative; wt, wild type.

Functional Downregulation and Overexpression of Kir2.1

Functional downregulation and recovery of Kir2.1 in endothelium was performed as previously described (7, 13). Freshly isolated human microvessels, or mouse mesenteric arteries, were incubated with adenoviral vectors of dominant negative Kir2.1 (dnKir2.1) or wild-type Kir2.1 (wtKir2.1) driven by an endothelial-specific promoter Cdh5 or empty adenoviral vector (GFP-Cdh5-AV, Vector Biolabs) at ∼100 multiplicity of infection (MOI) for 48 h in DMEM (Gibco) supplemented with 10% FBS (Gibco) and 1% penicillin-streptomycin (Gibco) at 37°C and 5% CO₂. The endothelial specificity of the model has already been validated in our previous study using an HA tag-specific fluorescence in intact versus denuded en face arteries (16). We have also established previously that both human and mouse resistance arteries maintain their reactivity to acetylcholine and to flow after being incubated with an empty adenoviral vector for 48 h with no impairment of endothelial function (7, 13, 14).

Patch-Clamp Electrophysiology

To validate the efficiency of the functional downregulation and overexpression of Kir2.1 in the endothelium, we assessed Kir2.1 currents via electrophysiology patch clamp following exposure to the viral constructs in mouse mesenteric arteries. Mouse mesenteric arteries were harvested from 10-wk-old C57BL/6 mice, cleaned, and incubated with the different adenoviral vectors for 48 h as detailed earlier. ECs were then isolated from these arteries for patch-clamp recording as previously described (20). Freshly isolated cells in suspension were allowed to adhere directly onto the flow chamber for at least 45 min before the minimally invasive flow (MIF) device assembly. After whole cell access, currents were recorded in the bath until a stable baseline was achieved, steady flow was applied by gravity perfusion to induce shear-activated inward K⁺ currents. Shear stress (τ; 0.74 dyn/cm²) was calculated using the equation τ = 6μQh2w, where μ is the fluid viscosity (0.009 g/cm·s), Q is the flow rate (300 μL/s applied by gravity perfusion), h is the height (0.1 cm) of the MIF chamber, and w is the width (2.2 cm) of the MIF chamber.

Pipettes with resistance (2–4 MΩ) were pulled using borosilicate glass (BF150-11010, Sutter) and a vertical puller (Model PP-830). Currents were recorded using an EPC10 amplifier and the PatchMaster Next (HEKA Electronik), filtered at 2 kHz, and recordings were digitized at 10 kHz. ECs were held at −30 mV and a voltage ramp of −130 to +40 mV was applied over 400 ms. The perforated patch was obtained by adding amphotericin B (250 μg/mL) to the pipette solution within 2–5 min after formation of a GΩ seal, and accepted recordings maintained a series resistance between 10 and 30 MΩ. When necessary, leak subtraction was performed offline to collect the most accurate data points at −100 mV for group analysis.

The effect of NO donor SNP on Kir2.1 channel current was also evaluated. For that, cultured mouse mesenteric artery endothelial cells were subjected to whole cell patch-clamp recordings before and after a 3-min treatment of 1 μM sodium nitroprusside dihydrate.

Acute High-Pressure Exposure

In the experiments evaluating the effects of acute high pressure in Kir2.1 function and FIV, we exposed the vessels to a pressure (150 cmH₂O or 110 mmHg) for 45 min, returned to regular pressure for 15 min (RP), preconstricted with ET-1 and proceeded with the FIV measurements as described in Flow-Induced Vasodilation Measurement. Percentage of change was calculated using the following equation: FIV (%) = (diameter at each pressure gradient − ET-1 preconstricted diameter)/(diameter at RP − ET-1 preconstricted diameter) × 100 (21).

In some experiments, the effect of high pressure in FIV was also determined in mesenteric arteries of conditional Kir-deficient mouse (Kir2.1^fl/fl Cdh5.CreERT2). To achieve this goal, a floxed Kir2.1 mouse model was generated on C57BL/6 background and crossed with Cdh5.CreERT2 mouse, an inducible EC-specific Cre that is activated by tamoxifen (Sigma). Mice were injected with 200 µL of tamoxifen solution (10 mg tamoxifen in 1 mL corn oil/ethanol solution, 10:1) during 5 consecutive days followed by 21 days waiting period. These mice were compared with the control that were injected with vehicle (corn oil/ethanol, 10:1).

Solutions and Reagents

HEPES buffer contained (in mM) 140 NaCl, 4 KCl, 1 MgCl₂, 5 glucose, 10 HEPES, and 2 CaCl₂ at pH 7.4. Physiological salt solution (Krebs) contained (in mM) 123 NaCl, 4.7 KCl, 1.2 MgSO₄, 2.5 CaCl₂, 16 NaHCO₃, 0.026 EDTA, 11 glucose, and 1.2 KH₂PO₄ at pH 7.4. Electrophysiology bath solution contained (in mM) 80 NaCl, 60 KCl, 10 HEPES, 1 MgCl₂, 2 CaCl₂, and 10 glucose at pH 7.4. Pipette solution contained (in mM) 5 NaCl, 135 KCl, 5 EGTA, 1 MgCl₂, 5 glucose, 10 HEPES at pH 7.2. Reagents were purchased from Sigma or Fisher Bioreagents unless stated otherwise.

Statistics

Unpaired Student’s t test, one-way ANOVA, or a two-way with or without repeated-measures ANOVA were used. Significance, in all cases, was set to P < 0.05. Bonferroni post hoc tests were used to determine where differences existed after significance was detected with ANOVA. Statistical analyses were performed using IBM SPSS Statistics (v.27, Chicago, IL) and GraphPad Prism (v.6.0, San Diego, CA). Data are presented as means ± SE or n (%).

To determine the correlations of maximal dilation or Kir2.1-dependent FIV component to outcomes of interests, Pearson product-moment correlation coefficients and scatter plots were used. To determine whether SBP, DBP, and MAP predict maximal dilation or Kir2.1-dependent FIV component, multiple linear regression models were performed, adjusting for LDL, body mass index, and age. These analyses were performed using SPSS (v.28).

RESULTS

Contribution of Kir2.1 to FIV Decreases in Human Subject with High Blood Pressure

In this study, we recruited a total of 32 men and women, diagnosed or not with hypertension, overall healthy, and ages ranging from 18 to 65 yr old. As described in Table 1, we evaluated 13 subjects with hypertension (45.31 ± 15.34 yr old; mean BP, 146/94 mmHg) and 19 normotensive subjects (30.58 ± 9.75 yr old; mean BP, 115/75 mmHg). Since we previously showed that Kir-dependent FIV is compromised by dyslipidemia (14), patients with abnormal elevated LDL were excluded from the study.

Vascular reactivity measured in isolated vessels in response to flow was significantly decreased (∼25%) in hypertensive subjects as compared with normotensive subjects (Fig. 1, A and B). We also demonstrate here that flow-induced dilation correlates negatively with SBP, DBP, and MAP but not with age, indicating that although subjects with hypertension were older than the control group (Fig. 1, C–F, and Table 2), which was not the determinant of the decrease in vascular function observed.

Figure 1.

Analysis of flow-induced vasodilation (FIV) in subjects with hypertension. A: FIV was measured by increasing the pressure gradients from 10 to 100 cmH₂O in resistance arteries isolated from fat biopsies of normotensive subjects or subjects with hypertension. Endothelium-independent vasodilation was assessed with papaverine (PAP) as a control. B: comparison of maximum dilation at Δ100 cmH₂O. Correlation of FIV at Δ100 cmH₂O vs. systolic blood pressure (SBP; C), diastolic blood pressure (DBP; D), mean arterial pressure (MAP; E), and age (F). *P < 0.05.

Table 2.

Demographics and clinical characteristics

Demographics/Clinical Characteristics	Normotensive	Hypertensive	P Value
n	19	13
Sex, n (%)
Men	9 (47.37)	10 (76.92)
Women	10 (52.63)	3 (23.08)
Age, yr	30.58 ± 9.75	45.31 ± 15.34	0.004
SBP, mmHg	115.21 ± 6.98	146.15 ± 15.16	0.000
DBP, mmHg	75.32 ± 5.73	94.38 ± 6.89	0.000
MAP, mmHg	88.61 ± 5.6	111.64 ± 8.02	0.000
BMI, kg/m2	24.65 ± 4.26	28.9 ± 4.43	0.007
LDL, mg/dL	96.33 ± 29.74	104.25 ± 38.98	0.286
HDL, mg/dL	56.27 ± 9.06	58.75 ± 11.26	0.265

Values are n (%) or means ± SD along with results of independent t test when applicable; n = 13–19 participants. Normotensive ranges are systolic blood pressure (SBP) ≤ 129 mmHg and diastolic blood pressure (DBP) ≤ 89 mmHg. Hypertensive ranges are SBP ≥ 130–159 mmHg and DBP ≥ 90–99 mmHg. MAP, mean arterial pressure; BMI, body mass index; LDL, low-density lipoprotein; HDL, high-density lipoprotein.

To establish whether Kir2.1 channels contribute to the impairment of endothelial function in peripheral resistance arteries of humans with hypertension, we used two complementary strategies to suppress Kir2.1: transfecting the arteries with a construct of dominant negative Kir2.1 driven by an endothelial-specific Cdh5 promoter (dnKir2.1) or blocking the channels with a pharmacological inhibitor ML133. In brief, resistance arteries freshly isolated from the biopsies were incubated with dnKir2.1 or with an empty adenoviral construct (empty) as an assay control for 48 h and measured FIV, as described in our earlier studies (7, 16). The validation of the functional downregulation of endothelial Kir2.1 channels with dnKir2.1 was assessed electrophysiologically using patch clamp in endothelial cells freshly isolated from mouse mesenteric arteries. Our data show that transducing the arteries with dnKir2.1 resulted in a significant decrease in endothelial Kir currents under static conditions and the loss of any current increase in response to flow (Fig. 2, A and B). The inhibitory effect of the downregulation or blocking of Kir2.1 on the FIV in arteries isolated from the biopsies of patients with hypertension was significantly smaller than in arteries isolated from normotensive subjects (Fig. 2, C and D). Analysis of the Kir2.1 component involved in FIV, calculated as the difference between maximal FIV with and without suppression of Kir2.1 in arteries isolated from the same biopsy, corroborates the results aforementioned by showing that in subjects with hypertension, the Kir2.1-dependent component of FIV is significantly reduced (Fig. 2E).

Figure 2.

Inwardly rectifying K⁺ channel (Kir2.1) contribution to flow-induced vasodilation (FIV) in subjects with hypertension. A: representative traces of Kir2.1 currents in freshly isolated endothelial cells from mouse mesenteric arteries exposed to empty (top) or dominant-negative (dn)Kir2.1 (bottom) viral constructs. B: average of Kir 2.1 current densities with empty (n = 8; no flow, −24.62 ± 4.17 pA/pF; with flow, −39.25 ± 3.74 pA/pF; ****P < 0.0001) or dominant negative (dn)Kir2.1 (n = 6; no flow, −13.92 ± 1.85 pA/pF; with flow, −15.92 ± 1.51 pA/pF). C: FIV was measured by increasing the pressure gradients from 20 to 100 cmH₂O in resistance arteries isolated from fat biopsies of normotensive subjects or subjects with hypertension (*normotensive vs. normotensive + ML133 or normotensive + dnKir2.1, P < 0.05; *hypertensive vs. hypertensive + dnKir2.1, P < 0.05; #normotensive vs. hypertensive, P < 0.05). D: comparison of maximum dilation at Δ100 cmH₂O (*P < 0.05). Endothelium-independent vasodilation was assessed with papaverine (PAP) as a control. E: comparison of the Kir2.1-dependent component in normotensive vs. hypertensive vessels (*P < 0.05). FIV data in C and D for untreated normotensive and hypertensive conditions are replotted from Fig. 1 to serve as controls for the other experimental conditions (treatments with ML133 or dnKir2.1). In all experiments, treated and untreated arteries from the same biopsies were used in parallel for a paired comparison.

Furthermore, inhibition of eNOS by l-NAME resulted in a significant loss of FIV in arteries of normotensive patients exposed to an empty virus but had no effect on arteries exposed to the dnKir2.1 virus, consistent with our previous studies showing that Kir2.1 is upstream of eNOS (Fig. 3A). No effect of l-NAME on FIV was observed in the arteries of patients with hypertension (Fig. 3B), indicating that NO generation is impaired, as was shown previously. No difference between normotensive subjects and subjects with hypertension was observed in endothelial-independent vasodilation induced by a NO donor, SNP (Fig. 3C). Application of SNP had no significant effects on Kir2.1 currents in mesenteric ECs (Fig. 3, D and E). This is slightly different from a previous report showing a positive effect of NO on Kir2.1 currents in cardiomyocytes (22), which could be due to different cell types or experimental conditions, such as using distinct NO donors.

Figure 3.

Vasodilation responses with N^G-nitro-l-arginine methyl ester (l-NAME) and sodium nitroprusside (SNP). Flow-induced vasodilation (FIV) was measured by increasing pressure gradients from 20 to 100 cmH₂O in resistance arteries isolated from fat biopsies of normotensive subjects [A; *empty vs. empty + l-NAME, P < 0.05; *empty vs. dominant-negative inwardly rectifying K⁺ channel (dnKir2.1), P < 0.05; #empty vs. dnKir2.1, P < 0.05] or subjects with hypertension (B). Endothelium-independent vasodilation was assessed with papaverine (PAP) as a control. C: Kir2.1 channels contribution to SNP dose-dependent vasodilation in vessels from normotensive subjects or subjects with hypertension exposed or not to dnKir2.1. D: representative traces of Kir2.1 currents in freshly isolated endothelial cells from mouse mesenteric arteries treated or not with SNP. E: average of Kir2.1 current densities in cells treated or not with SNP (n = 6; −15.5 ± 4.9 pA/pF and −14.67 ± 4.8 pA/pF, respectively).

Since the definitions of normotensive and hypertensive groups and assigning the subjects to these groups might differ from study to study, we analyzed the correlation between the Kir2.1-dependent and Kir2.1-independent components of FIV and blood pressure variables including SBP, DBP, and MAP for individual patients. This analysis shows that Kir2.1-dependent FIV has significant negative correlation with all three parameters in individual subjects (Fig. 4, A–C). No significant correlations were found in this cohort between Kir2.1-dependent FIV and age or body mass index (BMI) (Table 3) and the negative correlation between Kir2.1-dependent FIV and blood pressure was sustained when corrected for age, BMI, and LDL (Table 4).

Figure 4.

Inwardly rectifying K⁺ channel (Kir2.1)-dependent vasodilation correlates negatively with systolic blood pressure (SBP), diastolic blood pressure (DBP), and mean arterial pressure (MAP). Correlation of Kir-dependent flow-induced vasodilation (FIV) at Δ100 cmH₂O vs. SBP (A), DBP (B), and MAP (C) (*P < 0.05). Correlation of Kir-independent flow-induced vasodilation (FIV) at Δ100 cmH₂O vs. SBP (D), DBP (E), and MAP (F).

Table 3.

Pearson correlations between vasodilation and variables of interests

	Maximum Dilation	Kir2.1-Dependent Vasodilation
		ML133	dnKir2.1
SBP	−0.52**	−0.55**	−0.27
DBP	−0.53**	−0.50**	−0.37*
MAP	−0.55**	−0.55**	−0.34
Age	−0.21	−0.26	−0.16
BMI	−0.25	−0.04	0.002
LDL	−0.19	−0.19	−0.16
HDL	−0.08	0.07	−0.12

Values are correlation coefficients; n = 32 participants. SBP, systolic blood pressure; DBP, diastolic blood pressure; MAP, mean arterial pressure; BMI, body mass index; LDL, low-density lipoprotein; HDL, high-density lipoprotein; Kir2.1, inwardly rectifying K⁺ channels; dn, dominant negative.

^*P < 0.05

^**P < 0.01.

Table 4.

Partial correlations between vasodilation and blood pressure adjusted for age, BMI, and LDL

	Maximum Dilation	Kir2.1-Dependent Vasodilation
		ML133	dnKir2.1
SBP	−0.57**	−0.61**	−0.43*
DBP	−0.46*	−0.42*	−0.39*
MAP	−0.55**	−0.54**	−0.44*

Values are correlation coefficients; n = 32 participants. BMI, body mass index; LDL, low-density lipoprotein; SBP, systolic blood pressure; DBP, diastolic blood pressure; MAP, mean arterial pressure; Kir2.1, inwardly rectifying K⁺ channels; dn, dominant negative.

^*P < 0.05;

^**P < 0.01.

In contrast, the Kir2.1-independent component of the FIV, defined as the residual FIV portion after Kir2.1 inhibition, was not correlated with any of the parameters of the blood pressure (Fig. 4, D–F), suggesting that the impairment of the FIV in subjects with high blood pressure should be specifically attributed to the Kir2.1-dependent FIV.

Rescue of FIV in Subjects with Hypertension with EC-Specific wtKir2.1

To determine whether impaired FIV in patients with hypertension can be rescued by increased expression of Kir2.1 channels in ECs, a set of arteries from both normotensive and hypertensive groups were exposed to adenoviral constructs expressing WT Kir2.1 (wtKir2.1) driven by the endothelial promoter. The validation of the overexpression with wtKir2.1 was assessed electrophysiologically using patch clamp in endothelial cells freshly isolated from mouse mesenteric arteries. Our data show that transducing the arteries with wtKir2.1 resulted in an increase in endothelial Kir currents, as expected, and a robust response to flow (Fig. 5, A and B). Overexpression of wtKir2.1 in endothelium resulted in a significant improvement in vasodilation in arteries of patients with hypertension and a smaller improvement in normotensive patients (Fig. 5C). Furthermore, comparing the vasodilatory response in arteries expressing wtKir2.1 with control arteries (expressing an empty virus) isolated from the same patient from the same biopsy shows that the improvement of maximal dilation (at 100 cmH₂O) was observed in almost all subjects, 9 of 10 patients with hypertension and 11 of 15 normotensive participants (Fig. 5F). Notably, although in normotensive subjects a Kir2.1-driven rescue improvement was observed only at maximal dilation at 100 mmHg, in patients with hypertension, there was a greater separation of the response curves with and without Kir2.1 overexpression resulting in significant improvement at 60 cmH₂O (Fig. 5E). No statistically significant effects were observed at lower flow rates (Fig. 5D).

Figure 5.

Role of wild-type inwardly rectifying K⁺ channel (wtKir2.1) overexpression in vascular function. A: representative traces of Kir2.1 currents in freshly isolated endothelial cells from mouse mesenteric arteries exposed to empty (top) or wtKir2.1 (bottom) viral constructs. B: average of Kir 2.1 current densities with empty (n = 8; no flow, −24.62 ± 4.17 pA/pF; with flow, −39.25 ± 3.74 pA/pF) or wtKir2.1 (n = 7; no flow, −46.75 ± 4.33 pA/pF; with flow, −60.83 ± 4.72 pA/pF), ****P < 0.0001, **P ≤ 0.01. C: flow-induced vasodilation (FIV) of normotensive and hypertensive vessels exposed to wtKir2.1 (*P < 0.05; n = 13 and n = 11, respectively). Control vessels were exposed to an empty adenoviral vector. Endothelium-independent vasodilation was assessed with papaverine (PAP) as a control. D–F: dilation improvement ratio with wtKir2.1 in normotensive and hypertensive vessels at Δ40, Δ60, and Δ100 cmH₂O, respectively (*P < 0.05).

Loss of Kir2.1 in Human and Mice Resistance Arteries Exposed to Acute High Pressure

So far, we have demonstrated in this work that Kir2.1-dependent FIV is compromised in human resistance arteries of patients exhibiting elevated blood pressure. Next, we test whether this effect can be a result of acute elevation of the intraluminal pressure. To address this question, we applied an acute intraluminal high pressure by elevating the intraluminal reservoirs to 150 cmH₂O to approximate 110 mmHg for 45 min in vessels obtained from normotensive subjects and subjects with hypertension, either exposed to dnKir2.1 or empty adenoviral vectors, before assessing FIV.

In control vessels (exposed to empty adenoviral vector) from normotensive subjects, we observe a decrease of 35.5% in vasodilation after acute high pressure when compared with vessels under regular pressure (Fig. 6, A and B). This reduction is similar to the one observed in dnKir2.1 vessels under regular pressure. However, in arteries expressing dnKir2.1, acute high pressure had no further effect on FIV, indicating that acute intraluminal pressure applied for only 45 min can eliminate the Kir2.1-dependent component of FIV (Fig. 6, A and B).

Figure 6.

Acute high pressure (HP) blunts endothelial inwardly rectifying K⁺ channel (Kir2.1) contributions to flow-induced vasodilation (FIV). A: FIV of empty and dnKir2.1 normotensive vessels exposed to HP for 45 min (*empty vs. all other conditions, P < 0.05, n = 5). Endothelium-independent vasodilation was assessed with papaverine (PAP) as a control. B: comparison of maximum dilation at Δ100 cmH₂O in normotensive vessels under regular pressure (RP) or HP, respectively (*P < 0.05). C: FIV of empty and dominant-negative (dn)Kir2.1 hypertensive vessels exposed to HP for 45 min (n = 4). Endothelium-independent vasodilation was assessed with PAP as a control. D: comparison of maximum dilation at Δ100 cmH₂O in hypertensive vessels under RP or HP, respectively. E: FIV of control (vehicle) and tamoxifen-induced Kir2.1-deficient mouse isolated microvessels exposed to HP for 45 min (*vehicle vs. all other conditions, P < 0.05, n = 5). Endothelium-independent vasodilation was assessed with PAP as a control. F: comparison of maximum dilation at Δ100 cmH₂O in the microvasculature of control and Kir2.1-deficient mice under RP or HP, respectively (*P < 0.05, **P < 0.01).

In subjects with hypertension, which already have a compromised Kir2.1-dependent vasodilation, we did not observe any further effects when exposed to acute high pressure in neither empty nor dnKir2.1 (Fig. 6, C and D). This again corroborates our previous findings that high intraluminal pressure, both chronic and acute, has a profound impact on endothelial Kir2.1 response to flow.

To confirm the loss of endothelial Kir2.1 function in response to acute intraluminal high pressure, we applied the same protocol on mesenteric arteries from Kir2.1^fl/fl Cdh5.CreERT2 mice. Here we see that, similarly to vessels of normotensive human patients, exposure to acute high pressure results in loss of Kir2.1 channel contribution to FIV, as high pressure impairs vasodilation in vehicle control vessels, but no effect is seen in EC-specific Kir2.1-deficient vessels (Fig. 6, E and F).

DISCUSSION

Several mechanisms contribute to the impairment of vasodilation induced by elevated blood pressure, and understanding them is crucial for developing strategies to mitigate these effects. Over the past decades, various contributing mechanisms for high intraluminal pressure have been described (1, 23–25) and it was established that endothelial dysfunction is one of the key factors (2). Here, we demonstrate a novel mechanism that contributes significantly to the endothelial dysfunction in patients with hypertension: the impairment of flow-sensitive endothelial K⁺ channels, Kir2.1 and the loss of Kir2.1-dependent vasodilation. We also show that acute elevation of the intraluminal pressure alone is sufficient to impair Kir2.1 function and Kir2.1-dependent vasodilation suggesting that it is a direct effect of elevated pressure.

Our previous work showed that FIV, a hallmark of the endothelial response to flow, critically depends on endothelial Kir2.1 channels (7). Specifically, we have demonstrated that genetic deficiency of Kir2.1 results in severe impairment of FIV in an NO-dependent way resulting in increased vascular resistance and blood pressure in mice (7). Here, we show for the first time that the functional loss of endothelial Kir2.1 channels contributes to the impairment of vascular function in peripheral resistance arteries of humans with hypertension, which can be partially restored by the overexpression of Kir2.1 channel. Moreover, we show that the loss of Kir2.1-dependent FIV in subjects with hypertension is associated with the increase in intraluminal pressure, independently of age, BMI, and other risk factors.

We propose that the mechanism by which the loss of Kir2.1 contributes to the development of hypertension is a decrease in the bioavailability of NO. Indeed, a decrease in the bioavailability of NO is known to be a key factor in the development of hypertension, as the impairment of the NO-induced vasodilation leads to increased vascular resistance and elevated blood pressure. Previously, it was shown that a decrease in NO production in patients with hypertension and in animal models of hypertension can be due to eNOS uncoupling, possibly related to increased oxidative stress (26–28). Here, we show that the impairment of NO production in hypertension can be due to the disruption of the major flow-induced mechanotransduction pathway that involve activation of flow-sensitive K⁺ channels leading to the activation of Akt1/eNOS phosphorylation cascade, a pathway demonstrated in our previous studies (7). Although it is possible that the loss of Kir2.1 contribution to vasodilation is secondary to eNOS, our data suggest that this is not the case. Specifically, the observation that overexpression of Kir2.1 in the endothelium restores FIV in patients with hypertension by 33.5% indicates that the loss of NO bioavailability previously reported in the literature is to a significant degree a result of the impairment on endothelial Kir2.1 channels. However, since the restoration is not complete, other factors are also contributing to the loss of NO.

Several Kir2.1/NO-independent pathways have been described to contribute to vasodilation, including prostaglandins (PGI₂ and PGE₂) and a family of endothelial-derived hyperpolarizing factors (EDHF), that in response to shear stress are released by the endothelium to provide vasodilation (29). It has been shown that EDHF-mediated responses can offer a vasodilator reserve in cases of hypertension and may temporarily compensate for endothelial dysfunction caused by reduced synthesis or availability of NO (30). One of these EDHF responses that can mediate endothelium-dependent relaxations involves a reactive oxygen species (ROS)-dependent mechanism, hydrogen peroxide (H₂O₂), by potentiating Ca²⁺ release from endothelial stores, opening of hyperpolarizing endothelial calcium-activated potassium (K_Ca) channels that initiates the dilator response by triggering hyperpolarization and K⁺ release (31). This compensatory shift in the vasodilator mechanism, where H₂O₂ replaces NO and prostaglandins as the primary vasoactive mediator of dilation, has also been observed in isolated microvessels exposed to acute intraluminal high pressure (32).

What is the mechanism of the impairment of endothelial Kir2.1 channels in hypertension? We have previously shown that an increased pressure in isolated vessels is enough to reduce vascular function that is unique from models of hypertension that have confounding effects of circulating angiotensin II and inflammatory mediators (21, 32). Here, we show that this phenomenon is observed only in normotensive patients and critically depends on the contribution of endothelial Kir2.1 channels, whereas in patients with hypertension, in which the vasodilation is already impaired, acute elevation of pressure has no further effect. This is evidence that elevation of the pressure alone impairs the function of endothelial Kir2.1 channels. Furthermore, although the exact mechanism by which high pressure affects endothelial Kir2.1 channels remains unclear, the acute effect of pressure on Kir2.1 function could be due to the internalization of Kir2.1 protein from the plasma membrane of the endothelial cells to intracellular membranes via endocytosis. An alternative possible explanation is that pressure impairs the sensitivity of the channels to flow, possibly because of the damage to endothelial glycocalyx layer that we have shown earlier to be required for Kir2.1 flow response (16). It is well established that high blood pressure is associated with increased oxidative stress (33) and that oxidative stress has a detrimental effect on the endothelial glycocalyx (34, 35). Previous work from our group has shown that flow sensitivity of Kir depends on the integrity and thickness of the endothelial glycocalyx and that obesity-induced alterations of the endothelial glycocalyx renders Kir2.1 insensitive to shear stress, triggering the endothelial dysfunction in the microvasculature (16). Thus, similarly, disruption of the glycocalyx by pressure-generated ROS would render endothelial Kir2.1 channels to become flow insensitive resulting in the loss of their contribution to FIV. It is also well known that glycocalyx is sensitive to hemodynamic environment (35–37) and it is possible that elevation of the intraluminal pressure by itself is disruptive. Alternatively, ROS may also impair endothelial Kir2.1 channels by activating phospholipase A₂ (PLA₂), which in turn degrades phosphatidylinositol 4,5-bisphosphate (PIP₂), a membrane phospholipid required for Kir2.1 channel activity (27). Oxidative modification of the protein residues can also lead to changes in channel properties, including reduced K⁺ current (38, 39). Further studies are needed to discriminate between these possibilities.

It is noteworthy that age plays an important role on endothelial dysfunction and on loss of vascular reactivity (40). Recent studies showed impairment of endothelial Kir2.1 function in cerebral arterioles in aged mice (41, 42). A critical point, however, is that in these and other aging studies in mice, the comparison is made over a very significant age differences, such as comparing aged mice over 24 mo old, equivalent to humans over 70 yr old, with young mice (4–6 mo old), akin to humans aged 26 to 34 yr old. In contrast, in our current research, although the age difference between normotensive patients and patients with hypertension is statistically significant, the age range is relatively narrow and comparable between the groups. Nonetheless, we show here that our responses are independent of age perhaps indicating that age-associated declines in endothelial health are better associated with increased BP. In addition, since Kir-dependent FIV can also be compromised in patients with dyslipidemia, to avoid any interference or mixed effects, and guarantee that the results observed here were fully independent from the ones caused by the excess of lipids, patients with abnormal elevated LDL were excluded from the study.

A limitation of the study is the scarcity of the biological material obtained from the human biopsies, which prevents obtaining data on the expression of the channel protein and determining whether the functional differences between normotensive patients and patients with hypertension are a result of differential Kir2.1 expression. Briefly, each biopsy contains only a small piece of tissue with two to three vessels of 100–150 µm in diameter and 1–2 mm in length. This small amount of material does not yield enough endothelial cells for assessing the expression of the channels that are also known to be functional at very low protein expression levels. Moreover, pulling materials from multiple patients may not be justified for the IRB protocol. The same limitation applies to single segments of arteries exposed to acute high intraluminal pressure, both human and mouse. Future studies should address this limitation by obtaining discarded materials or establishing an animal model to recapitulate these results.

Perspectives and Significance

This is a novel discovery showing a link between endothelial Kir2.1 channels and impairment of endothelial function under high pressure. We suggest that elevated pressure impairs Kir2.1 activity in the endothelial cells, leading to a reduction in FIV and possibly leading to a vicious cycle between elevation of blood pressure and further impairment of Kir2.1 function and FIV. Furthermore, there may be clinical and translational implications in the potential therapeutic rescue of Kir2.1 function to benefit NO-dependent vasodilation. Future research focusing on Kir2.1 channel modulation may offer new therapeutic strategies to improve endothelial function and vascular health in individuals with hypertension.

DATA AVAILABILITY

Data presented in this study are available on request from the corresponding authors.

GRANTS

This research was supported by National Institutes of Health Grants R01HL141120 (to I. Levitan and S. A. Phillips); R01HL073965, R01HL083298, and R56AG082099 (to I. Levitan); R01HL130513 (to S. A. Phillips); and UL1TR002003 (to S. A. Phillips).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

N.F.D.C., I.F., S.A.P., and I.L. conceived and designed research; N.F.D.C., I.F., S.T.G., J.C.-S., K.M.B., and S.J.A. performed experiments; N.F.D.C., S.T.G., J.C.-S., and C.-L.H. analyzed data; N.F.D.C., I.F., S.A.P., and I.L. interpreted results of experiments; N.F.D.C. prepared figures; N.F.D.C. drafted manuscript; N.F.D.C., I.F., C.-H.H., S.A.P., and I.L. edited and revised manuscript; N.F.D.C., I.F., S.T.G., J.C.-S., K.M.B., S.J.A., C.-L.H., S.A.P., and I.L. approved final version of manuscript.