Reactivity of renal and mesenteric resistance vessels to angiotensin II is mediated by NOXA1/NOX1 and superoxide signaling

Abstract

Activation of NADPH oxidase (NOX) enzymes and the generation of reactive oxygen species and oxidative stress regulate vascular and renal function and contribute to the pathogenesis of hypertension. The present study examined the role of NOXA1/NOX1 function in vascular reactivity of renal and mesenteric resistance arteries/arterioles of wild-type and Noxa1^−/− mice. A major finding was that renal blood flow is less sensitive to acute stimulation by angiotensin II (ANG II) in Noxa1^−/− mice compared with wild-type mice, with a direct action on resistance arterioles independent of nitric oxide (NO) bioavailability. These functional results were reinforced by immunofluorescence evidence of NOXA1/NOX1 protein presence in renal arteries, afferent arterioles, and glomeruli as well as their upregulation by ANG II. In contrast, the renal vascular response to the thromboxane mimetic U46619 was effectively blunted by NO and was similar in both mouse genotypes and thus independent of NOXA1/NOX1 signaling. However, phenylephrine- and ANG II-induced contraction of isolated mesenteric arteries was less pronounced and buffering of vasoconstriction after acetylcholine and nitroprusside stimulation was reduced in Noxa1^−/− mice, suggesting endothelial NO-dependent mechanisms. An involvement of NOXA1/NOX1/O₂^•− signaling in response to ANG II was demonstrated with the specific NOXA1/NOX1 assembly inhibitor C25 and the nonspecific NOX inhibitor diphenyleneiodonium chloride in cultured vascular smooth muscle cells and isolated mesenteric resistance arteries. Collectively, our data indicate that the NOX1/NOXA1/O₂^•− pathway contributes to acute vasoconstriction induced by ANG II in renal and mesenteric vascular beds and may contribute to ANG II-induced hypertension.

NEW & NOTEWORTHY Renal reactivity to angiotensin II (ANG II) is mediated by superoxide signaling produced by NADPH oxidase (NOX)A1/NOX1. Acute vasoconstriction of renal arteries by ANG was blunted in Noxa1^−/− compared with wild-type mice. NOXA1/NOX1/O₂^•− signaling was also observed in ANG II stimulation of vascular smooth muscle cells and isolated mesenteric resistance arteries, indicating that it contributes to ANG II-induced hypertension. A NOXA1/NOX1 assembly inhibitor (C25) has been characterized that inhibits superoxide production and ameliorates the effects of ANG II.

INTRODUCTION

In the vasculature, NADPH oxidase (NOX) enzymes are major sources of reactive oxygen species (ROS) such as superoxide anion (O₂^•−) and hydrogen peroxide (H₂O₂), causing contraction and relaxation of conduit and resistance vessels regulating arterial blood pressure and local organ perfusion. NOX1 and NOX4 are abundant in vascular smooth muscle cells (VSMCs) (1), whereas NOX2 and NOX4 are abundant in endothelial cells of rodents (2, 3). ROS produced by the NOX family are important regulators of vasomotor function during physiological and pathological conditions such as hypertension, diabetes mellitus, and atherosclerosis (4, 5). In general, O₂^•− is a vasoconstrictor, whereas H₂O₂ is a vasodilator. This holds true for systemic and renal cortical arteries and arterioles, which regulate arterial blood pressure and renal blood flow (RBF), respectively.

A chronic angiotensin II (ANG II) infusion can cause hypertension by shifting the pressure natriuresis relation to a higher arterial pressure through mechanisms that result in renal salt retention caused by excessive tubular Na⁺ reabsorption and/or renal vasoconstriction (6–9). Hypertension caused by ANG II also causes increased NOX-dependent renal production of O₂^•−, resulting in reduced nitric oxide (NO) bioavailability, which leads to exaggerated renal vascular reactivity to vasoconstrictor agents (10). Similar potentiated reactivity to vasoconstrictors has been observed in systemic resistance vessels like mesenteric arteries (11, 12).

ANG II activates ANG II type 1 (AT₁) receptors to stimulate NOX enzyme activity and O₂^•− production in VSMCs, which causes vasoconstriction and chronic upregulation of inflammation, leading to pathologies such as hypertension and atherosclerosis (13). Excessive renal and vascular NOX-mediated O₂^•− production contributes to the pathogenesis and maintenance of chronic ANG II/AT₁ receptor-induced hypertension in mice and rats. In spontaneously hypertensive rats (SHRs) and deoxycorticosterone acetate (DOCA)-salt rats, AT₁ receptor antagonists, the NOX inhibitor apocynin, extracellular superoxide dismutase (SOD), or the antioxidant SOD mimetic tempol decrease renal vasoconstriction and Na⁺ retention as well as arterial pressure (10, 14–19). Chronic ANG II upregulates NOX1, NOX2, and NOX4 mRNA in VSMCs (1, 20, 21). In the renal cortex, activation of AT₁ receptors by ANG II upregulates renal cortical O₂^•− production and NOX1, p22^phox, p67^phox, and Mn-SOD expression while downregulating NOX4 and extracellular SOD mRNA (22).

Global gene deletion studies have established that NOX1, NOX2, and O₂^•− production contribute to the development of ANG II-induced hypertension in C57BL/6 mice and that increased NOX4/H₂O₂ is involved in salt-sensitive hypertension in Dahl rats (7, 23–25). The mechanisms by which NOX enzymes cause hypertension are not well understood. ANG II-induced hypertension in mice may be caused by systemic vasoconstriction, as shown in studies where NOX1 or NOX2 overexpression in the vasculature increases hypertension (23, 26). Upregulation of the distal nephron epithelial Na⁺ channel (ENaC) has been implicated in enhanced renal Na⁺ retention by NOX1/NOXA1/O₂^•− signaling in ANG II-induced hypertension in mice and by NOX4/H₂O₂ in Dahl salt-sensitive rats (7, 27–29). Renal vasoconstriction can also lead to Na⁺ retention. ROS produced by NOX are known to participate in ANG II-induced renal vasoconstriction. Superoxide mediates acute renal vasoconstriction induced by ANG II, U46619, endothelin, and vasopressin in mice, with attenuation of reactivity by the antioxidant SOD mimetic tempol (30–32).

Currently, little is known about the role of specific NOX isoforms in mediating the contraction of resistance arteries/arterioles in the kidney. NOX2 and p47^phox mediate part of the acute ANG II-induced contractions of isolated mouse afferent arterioles, with a potentiated role in chronic ANG II-induced hypertension (33). An intravenous infusion of ANG II for 30 min produces less renal vasoconstriction in anesthetized mice lacking NOX2 than in control C57BL/6 mice (34). In diabetic SHRs, the responses to ANG II, endothelin-1, and norepinephrine in isolated renal arteries are exaggerated in association with elevated renal NOX4 expression (35). To the best of our knowledge, there is no evidence that NOX1/NOXA1/O₂^•− signaling may play a role in the reactivity of rodent renal vessels to common vasoactive agents.

Thromboxane A₂ (TxA₂), prostaglandin H₂ (PGH₂), and 8-isoprostane-PGF_2α activate thromboxane prostanoid (TP) receptors to cause vasoconstriction, mediated in part by vascular O₂^•− production (36–43). TxA₂ and TP receptors are also of interest because they are involved in the action of an endothelium-derived contracting factor, as identified in renal and mesenteric arteries (44). The specific NOX isozymes involved in TP receptor-induced vasoconstriction are poorly understood.

The objective of the present study was to determine whether NOXA1/NOX1-derived O₂^•− plays a role in the acute regulation of RBF and reactivity of renal and mesenteric resistance arteries/arterioles in vivo and in vitro, respectively. Therefore, we examined renal vascular reactivity to acute ANG II and U46619, an agonist of the TP receptor, in wild-type and Noxa1^−/− mice. Similarly, we wanted to determine to what extent NOXA1/NOX1-mediated vasoconstriction is due to direct action independent of NO bioavailability versus that mediated by NO bioavailability.

MATERIALS AND METHODS

Animals

All animal procedures were performed in compliance with protocols approved by the University of Michigan and the University of North Carolina Institutional Animal Care and Use Committees in accordance with National Institutes of Health policies. Noxa1^−/− mice were generated by deleting the entire coding region of the mouse Noxa1 gene as previously described (45). Wild-type C57BL/6J mice were purchased from Jackson Laboratory (Bar Harbor, ME) and bred with Noxa1^−/− mice to generate heterozygous mice. Experiments were performed with wild-type and Noxa1^−/− littermate mice bred in-house from heterozygous breeding pairs. Mice were housed in ventilated cages with a 12:12-h light-dark cycle, and food and water were provided ad libitum.

Preparation of Animals for RBF Experiments

Mice were prepared for renal hemodynamic experiments as previously described (32, 46). Animals were anesthetized with an intraperitoneal injection of pentobarbital sodium (80 mg/kg body wt) and placed on a heated operating table that was servo-controlled to maintain body temperature at 37°C. The trachea was intubated with a phenylephrine (PE)-90 cannula to facilitate spontaneous breathing. A jugular vein was cannulated with three PE-10 catheters for the infusion of maintenance solutions and supplemental doses of anesthetic. Isotonic saline was infused continuously at 7 μL/min with 2% BSA in saline at 3 μL/min (Fraction V, Sigma). The BSA line was connected to a 10-µL HPLC injection valve to allow switching to a bolus injection of vasoactive agents. Arterial pressure was monitored through a PE-10 cannula inserted in a carotid artery with a pressure transducer (p23dB, Statham) connected to a Hewlett-Packard carrier amplifier (Model 8805B). Urine was drained through a PE-90 cannula secured in the bladder at the dome. The right kidney was exposed through a retroperitoneal flank incision, and the renal artery was mobilized free of connective tissue to allow installation of an ultrasonic transit time flow probe (0.5PSB transducer connected to a TS420 flow monitor, Transonic) for measurement of RBF. Renal nerves were removed during this procedure to eliminate reflex sympathetic input to the kidney that could influence RBF during alterations in arterial pressure.

Animals were allowed to stabilize 30–45 min after the surgical preparation before data were recorded. Analog arterial pressure and RBF signals were digitized with a PC-based data-acquisition system (Data Translation DT9816 analog-to-digital board controlled by DtEz software) at 120 samples/s and averaged over 1-s intervals before being written to disk for offline analysis. During data collection periods for analysis of renal vascular reactivity to vasoactive agents, the incoming arterial pressure and RBF data streams were smoothed with a low-pass Butterworth fifth-order digital filter with the frequency cutoff set to 2 Hz and then downsampled to 10 Hz for storage on disk. Data were collected ∼1 min before and 2–3 min after administration of the vasoactive agent.

Renal Responses to Vasoactive Agents

Ang II and the TxA₂ mimetic U46619 were injected individually as a bolus into a jugular vein and administered using a 10-μL HPLC injection valve connected in series with the BSA infusion line. For each injection, the infusion rate was increased to 60 μL/min immediately before the injection valve was operated and then returned to 3 μL/min when the injection had been completed, 18 s later. This procedure provided consistent 10-μL bolus injections of ANG II or U46619 at a rate of 1 μL/s. Doses of ANG II (1 and 3 ng) or U46619 (25 and 50 ng) were administered in random order. All doses were administered in each period in duplicate. The recovery period after each injection was based on the time required for the arterial pressure to return to the preinjection level but was always at least 10 min. This procedure limits the activation of ANG II and TP receptors to a minimum and minimizes tachyphylaxis associated with prolonged receptor activation.

Experiments were divided into two periods. Control data were obtained in the first period, followed by the administration of the NO synthase (NOS) inhibitor N^G-nitro-l-arginine methyl ester (l-NAME; 25 mg/kg iv) in the second period. A 20-min stabilization period was allowed for stabilization after l-NAME administration before retesting renal vascular reactivity. Maximum effects of a given agonist were determined by two-way ANOVA with repeated measures for overall dose and mouse genotype effects, with post hoc Bonferroni testing of comparisons between individual groups (GraphPad). Experiments of l-NAME effects were analyzed within a given group by one-way ANOVA with repeated measures and post hoc individual comparisons via the Holm–Sidak method (SigmaStat).

ANG II-Treated Mice

Three- to four-month-old male wild-type and Noxa1^−/− mice were randomly assigned to control or treatment groups. Mice were implanted with subcutaneous micro-osmotic pumps (Model 1002, Alzet, Durect, Cupertino, CA) to continuously deliver a slow pressor dose of ANG II (500 ng/kg/min, A9525, Sigma-Aldrich, St. Louis, MO) (47, 48) or vehicle (0.9% NaCl) for 14 days (45).

Kidney Arteriole Isolation

Preglomerular resistance arterioles were isolated using a modification of the iron oxide method as previously described (46). Mice were anesthetized using pentobarbital sodium (80 mg/kg ip). A midline incision was made, and the aorta was clamped above the renal arteries and cannulated at the bifurcation of the left and right common iliac arteries with a blunted 25-gauge butterfly needle. The left renal vein was punctured, and the kidneys were perfused with PBS containing 5 mM glucose and 0.1 mM CaCl₂ until all blood had been removed from the kidneys as evidenced by the clear solution exiting the left renal vein (∼5 mL). The kidneys were then perfused with 5 mL of 1% iron oxide solution followed by 1–2 mL of PBS and immediately excised and placed in PBS. The kidneys were decapsulated, and cortical slices were minced for 3 min with a razor blade and homogenized for 10 s with a TissueLyser. The vessels were removed using a magnet and passed through a bent 22-gauge needle followed by three passes through straight 25-gauge needles and sieving through a 55-μm sieve to remove tubular fragments and glomeruli. Vessels were placed in PBS containing 312.5 digestion units of collagenase (type 1A, Sigma, 2.5 mg/25 mL) for 30 min, followed by magnetic removal and passage through two 25-gauge needles and one 27-gauge needle. Preglomerular resistance arterioles were obtained by centrifugation at 5,000 rpm for 30 min at 4°C. Isolated arterioles were treated without or with 1 µM ANG II in PBS for 30 s before being flash frozen for ROS measurement.

Histology, Immunostaining, and Confocal Microscopy

Mice were euthanized with an overdose of inhaled isoflurane, their systemic vasculature was perfused with HBSS, and their kidneys and renal arteries were dissected and embedded in optimal cutting temperature (OCT) compound (No. 23-730-571, Fisher Scientific, Pittsburgh, PA) and snap frozen in liquid nitrogen. Sagittal sections of the hilum and coronal sections of the medulla and cortex were cut at a thickness of 5 mm from dissected kidneys. Frozen kidney sections were fixed in acetone and permeabilized in 0.1% Triton X-100, and immunofluorescent staining was then performed using antibodies against NOXA1 (Ab199, a gift from Dr. Ralf Brandes, Institut für Kardiovaskuläre Physiologie, Goethe-Universität, Frankfurt am Main, Germany) (49) or NOX1 (bs-3682R, Bioss, Woburn, MA) followed by goat anti-rabbit secondary antibodies conjugated with Alexa Fluor 594 (A11072, Thermo Fisher) and Alexa Fluor 488-conjugated aquaporin-2 (AQP2; bs-4611R-A488, Bioss) or FITC-conjugated α-smooth muscle actin (ACTA2; F3777, Sigma-Aldrich). Primary antibodies were used at 1:100 dilution, and secondary or conjugated antibodies were used at 1:200 dilution. The NOXA1 antibody was validated with blocking peptides by Ambasta et al. (50); other primary antibodies were validated commercially by the vendor. Secondary antibodies were validated by the omission of the primary antibody to ensure a lack of background staining. To observe fluorescence with DAPI (H-1200, Vector Laboratories), coverslips were mounted in Vectashield mounting medium. A Nikon Microphot-FX microscope was used to acquire fluorescence images using the same magnification, exposure, gain, and offset settings. All images were analyzed using ImageJ 1.53t (National Institutes of Health, Bethesda, MD). Fluorescence was determined by measuring the integrated density. Colocalization was determined by applying a color threshold and measuring the fluorescence-integrated density of overlapping channels.

ROS Measurements

ROS levels in tissue sections were measured as previously described (51). Sagittal and coronal sections of fresh-frozen renal tissues were incubated with 10 mM dihydroethidium (DHE; D11347, Thermo Fisher) in the presence or absence of 200 U/mL polyethylene glycol-superoxide dismutase (SOD; S9549, Sigma) in HBSS for 15 min at 37°C. Using a Nikon Microphot-FX microscope and monochrome camera, fluorescent images were acquired using identical magnifications, exposure, gain, and offset values. ImageJ (National Institutes of Health) was used to analyze the images, and the mean gray value (integrated density) was determined.

VSMCs were isolated from wild-type mice as previously described (45, 51). VSMCs and isolated afferent arteriole superoxide levels were determined by HPLC detection of 2‐hydroxyethidium (51, 52). Cells were grown to 80% confluence and quiesced overnight in DMEM. Following incubation with 10 mM of test compounds for 30 min [or 10 μM diphenyliodonium (DPI) as a positive control], VSMCs were treated with 100 ng/mL of TNF-α and 50 μM DHE for an additional 30 min. VSMCs were harvested in acetonitrile, and supernatants were dried with a Savant SpeedVac ISS 100. Afferent arterioles were treated with 50 μM DHE for 30 min, washed three times with PBS and homogenized in acetonitrile using a TissueLyser II (Qiagen, Hilden, Germany), and supernatants were dried with a Savant SpeedVac ISS 100. Samples were dissolved in PBS and analyzed with an Agilent 1100 HPLC system equipped with a Partisil 5-µm ODS3 250 × 4.6-mm or Kinetex 2.6-µm C18 100 Å 100 × 4.6-mm column (Phenomenex, Torrance, CA). Quantification was carried out by comparison with a standard curve generated from a 2‐hydroxyethidium standard (Noxygen Science Transfer & Diagnostics, Elzach, Germany). The pellet from the cells or tissue after harvesting was dissolved in T-PER lysis buffer (No. 78510, Thermo Fisher), and protein was quantified to standardize measurements.

Mesenteric Artery Myography

After mice were euthanized by an isoflurane overdose, their mesenteric arteries were removed and placed in physiological saline solution (130 mM NaCl, 4.7 mM KCl, 1.17 mM MgSO₄, 1.18 mM KH₂PO₄, 14.9 mM NaHCO₃, and 5.5 mM dextrose) at 4°C. Second- or third-order arteries were cleaned of adventitial fat, immediately mounted on the DMT 620M four-chamber myography (DMT, Aarhus, Denmark), heated to 37°C, and supplied with oxygen by bubbling a 95% O₂-5% CO₂ gas mixture into the chamber. The physiological saline solution was refreshed every 20 min throughout the experiment once the temperature reached physiological levels. We tested two mesenteric arteries from each mouse and used one mouse of each genotype in each experiment.

As determined by the DMT normalization module, mesenteric arteries were prestressed at ∼95% of the arterial pressure of 13.3 kPa (100 mmHg), which is the typical mean arterial pressure in mouse mesenteric arteries. Vessels were allowed to stabilize for 30 min after achieving pretension before the reactivity was assessed. Constriction trials had at least 20 min of rest time between each one to allow vessels to normalize. Two constriction trials were carried out with high-K⁺ buffer (60 mM K⁺), followed by a contraction with 1 × 10⁻⁵ M phenylephrine for 2 min and a relaxation with 1 × 10⁻⁵ M acetylcholine. When the vessels did not relax by at least 50% to acetylcholine, it was assumed that the endothelium was damaged, and such vessels were excluded from the analysis. Half-log dose-response curves were performed to assess contractions to phenylephrine and ANG II and relaxation to sodium nitroprusside and acetylcholine. Contractile responses of vessels were assessed independently of one another using the high-K⁺ contraction as a baseline for comparison with ANG II contraction and the time-dependent stability of the phenylephrine contraction. Half-log dose-response curves were used to evaluate relaxation responses immediately following maximal phenylephrine contraction. To prevent tachyphylaxis after contracting vessels repeatedly with ANG II, a resting period of 2 h was implemented.

Peptide Preparation

A cDNA containing the SH3 domain of human NOXA1 isoform 1 (residues 401–483 for transcript variant 1 and residues 401–476 for transcript variant 2) was cloned into the pET-30 vector containing a NH₂-terminal His-tag (Sigma). The protein was expressed in Rosetta 2 (DE3) pLysS competent cells (Sigma). Bacterial cultures (6 L) were grown to OD₆₀₀ = 0.6 at 37°C, and the temperature was then lowered to 18°C, 0.5 mM isopropylthiogalactoside (IPTG) was added, and the culture was grown overnight. Cells were harvested and resuspended in 180 mL (30 mL/L culture) that contained 50 mM sodium phosphate (pH 7.2), 100 mM sodium chloride, 30 mM imidazole, and Roche EDTA-free complete protease inhibitors. Cells were lysed with two rounds of sonication for 20 s at 40% amplitude, followed by a 40-s rest period between each pulse for 12 cycles. Lysates were cleared by centrifugation at 12,000 rpm for 40 min.

Supernatants were loaded onto HisTrap FF 5-mL columns for NOXA1-SH3-His or NCF1-PRR-GST, respectively, at 1 mL/min using a sample pump on an ÄKTA FPLC (GE Healthcare). Before sample loading, the HisTrap FF column was equilibrated with 10 column volumes (CV) of binding/wash buffer [50 mM sodium phosphate (pH 7.2), 500 mM NaCl, and 30 mM imidazole]. After sample loading, the column was washed with 15 CV of binding/wash buffer. The protein was eluted in a linear gradient to 100% elution buffer [50 mM sodium phosphate (pH 7.2), 500 mM NaCl, and 500 mM imidazole] over 20 CV with 1-mL fractions collected during elution. GSTrap FF columns were washed with 15 CV of binding/wash buffer after sample loading. The protein was eluted using 100% elution buffer [25 mM Tris (pH 7.5), 150 mM NaCl, and 10 mM reduced glutathione] over 10 CV with 1-mL fractions collected during elution. An SDS-PAGE gel and Coomassie staining were used to identify peak fractions containing NOXA1-SH3-His or NCF1-PRR-GST. Fractions were concentrated using Amicon Ultra-15 centrifuges (3- or 10-kDa cutoff, Sigma). Concentrated, the affinity-purified protein was loaded at 1 mL/min onto a HiLoad 26/600 Superdex 75 prep grade column using a sample pump on an ÄKTA Purifier. Before samples were loaded, the column was equilibrated with 1.2 CV of sample buffer (1× PBS and 2 mM β-mercaptoethanol). Protein was eluted isocratically over 1.3 CV with a 0.3-CV elution before fractionation and collection of 3-mL fractions over the remaining 1 CV. Peak fractions containing NOXA1-SH3-His or NCF1-PRR-GST were detected by SDS-PAGE gel and Coomassie staining. Fractions were pooled and concentrated using Amicon Ultra-15 centrifugal concentrators. The isolated proteins were stored at −80°C in 200-µL aliquots.

AlphaScreen Assay Development

The AlphaScreen detection system (Perkin-Elmer, Waltham, MA) is made up of a histidine (Nickel Chelate) detection kit (Product No. 6760619) and glutathione donor beads (Product No. 6765300). NOXA1-SH3v2-His and NCF1-PRR-GST were titrated from 0 to 10,000 nM in Tris buffer [20 mM Tris base (pH 7.5), 75 mM NaCl, 2 mM DTT, 0.05% Tween 20, and 0.1% BSA (Thermo Fisher)] in the 384-well OptiPlate (Perkin-Elmer) in a final volume of 5 µL. After incubation for 30 min, 15 mg/µL of nickel-chelate and streptavidin-conjugated beads were added and incubated for another 30 min. The EnSpire Multimode Plate Reader (Perkin-Elmer) was used to read the plates using a 640-nm dichroic excitation mirror and a 570-nm cutoff emission filter.

The assay conditions were optimized as previously described (53). The addition of 0.1% BSA to the basic Tris buffer increased the AlphaScreen signal and decreased the background. The DMSO addition was titrated from 0% to 5% in the basic buffer showing a DMSO tolerance of up to 2%. The optimal concentrations for high binding signal and low background were 10 nM NOXA1-SH3v2-His and NCF1-PRR-GST, and 10 µg/mL acceptor and donor beads were incubated for 3 h. Loading peptides followed by adding premixed beads to the assay plate was the preferred sequence for adding assay reagents. The specificity of the AlphaScreen NOXA1-SH3v2-His and NCF1-PRR-GST peptide-binding signals was determined with a competition assay using human NOXA1-SH3v1-His peptide containing an inhibitory sequence (EPDVPLA) (54) and, alternatively, the untagged human NOXO1-PRR peptide (peptide A: TAPPPTVPTRPSPGAIQSRCCTVTRRALE) (55).

For loading peptides and beads, a MicroFlo Select Reagent Dispenser equipped with a 1-µL cassette (Agilent BioTek) was used, and a Multimek NSX-384 Automated Liquid Handler (NanoScreen) was used for loading screening compounds. The assay robustness was evaluated for interplate and interday variations. The assay signal remained stable for over 24 h, allowing high-throughput batch screening.

High-Throughput Small-Molecule Library Screening

An annotated library of 22,000 small molecules was screened using the AlphaScreen assay. Test compounds were diluted in 4% DMSO-Tris buffer at a concentration of 10 µM and dispensed to 384-well OptiPlates. NOXA1-SH3v2-His and NCF1-PRR-GST peptides diluted in 0.1% BSA-Tris buffer were added to the plates, followed by the AlphaScreen acceptor and donor bead mix. Plates were sealed with TopSeal microplate seal (Perkin-Elmer), incubated for 3 h, and read using the EnSpire Multimode Plate Reader. NOXA1-SH3v1-His was used as a negative control for each plate. To eliminate compounds that may interfere with assay chemistry, potential hits were counterscreened using Biotin-His6 peptide (Perkin-Elmer) and AlphaScreen nickel-chelate and streptavidin-conjugate bead mix.

Luciferase Protein Complementation Assay

Full-length human NOXA1 and NCF1 cDNAs were cloned into pUC19-NLuc and pUC19-CLuc vectors containing NH₂- and COOH-terminal fragments of firefly luciferase, respectively. NOXA1-NLuc cDNA was then subcloned into pLenti-NEO and NCF1-CLuc cDNA into pLenti-HYG lentiviral backbone vectors and cotransfected into human embryonic kidney (HEK)-293T cells with FuGENE 6 transfection reagent (Promega). A stable cell line of NOXA1-NLuc/NCF1-CLuc HEK-293T cells was maintained in DMEM-HG supplemented with 10% fetal bovine serum, 1× penicillin/streptomycin, 400 µg/mL of G418 and 200 of μg/mL hygromycin. In 96-well white-wall, white-bottom assay plates were coated with polyethylenimine (PEI) (25 µg/mL). Cells were plated at 50,000 cells/well in a growth medium (DMEM-HG supplemented with 10% FBS and 1× penicillin-streptomycin) the day before the assay. Cells were incubated overnight at 37°C and 5% CO₂. On the day of the assay, test compounds were prepared in a growth medium at a concentration of 20 times the desired final concentration and 5 µL were added to the plate. After 3 h of incubation at 37°C and 5% CO₂, the plate was equilibrated at room temperature for 15 min before 100 µL of Bright-Glo reagent (Promega) was added to each well in the dark. After being shaken for 30 s, the plate was incubated for 2 min at room temperature, and luminescence (1-s integration) was read using an EnSpire Multimode Plate Reader (Perkin-Elmer).

Chemicals Used

Pentobarbital, ANG II, and l-NAME were obtained from Sigma Chemical. U46619 was purchased from Cayman Chemical.

Statistical Analysis

Statistical analyses were performed with Prism 9 (GraphPad Software, La Jolla, CA). Using the Shapiro–Wilk test, all data were checked for normality. An unpaired t test, one-way ANOVA followed by a Tukey multiple comparisons test, or two-way ANOVA with Fisher’s least-significant-difference multiple comparisons test were used to analyze the data. Fitting of potential inhibitor data was performed using a nonlinear log compound versus response least squares fit. IC₅₀ for each inhibitor was calculated using best-fit values. Significant differences were defined as P < 0.05.

RESULTS

Effects of Agonists on Renal Blood Flow in Mice With or Without Noxa1

Noxa1^−/− mice have been shown to be protected against ANG II-induced hypertension by measuring mean arterial pressure using telemetry over a period of 14 days, in which constant ANG II infusions are administered using osmotic minipumps (7). Na⁺ excretion analysis showed an acute phase of increased excretion time, 1–2 days with ANG II infusion, that was observed in wild-type mice but absent in Noxa1^−/− mice. RBF changes in response to agonists can reveal an aberrant pressure-natriuresis relationship, leading to altered renal Na⁺ handling. In this study, we examined RBF in wild-type and Noxa1^−/− mice in response to ANG II and U46619, a thromboxane mimetic. Baseline measures of mean arterial blood pressure, RBF, and renal vascular resistance were not significantly different between wild-type control and global Noxa1^−/− mice (Table 1). The vasoactive agents took 60 s to reach the kidney, with maximum vasoconstriction occurring at 85 s, followed by RBF returning to the control level 180 s after injection (Fig. 1, A and B). In control mice, ANG II (1 ng) produced the maximum reduction of RBF of 22%, whereas ANG II (3 ng) produced the maximum reduction of 50% of resting RBF. In contrast, ANG II caused weaker vasoconstriction in mice lacking NOXA1: 15% with 1 ng ANG II and 29% with 3 mg ANG II (Table 2 and Fig. 1C). Therefore, ANG II produced ∼40% less acute renal vasoconstriction in NOXA1-deficient mice, a statistically significant difference (P < 0.01, two-way ANOVA). In contrast, there was no genotype difference in renal vascular reactivity to U46619. Injections of 25 and 50 ng U46619 caused similar ∼15% and 29% reductions in RBF in wild-type and Noxa1^−/− mice (Table 2 and Fig. 1D).

Table 1.

Basal arterial pressure and renal artery function in young male mice (3–4 mo)

	Mean Arterial Pressure, mmHg	Renal Blood Flow, mL/min/g Kidney wt	Renal Vascular Resistance, mmHg/(mL/min/g Kidney wt)
Wild type (baseline)	96.6 ± 6.0	7.5 ± 2.9	39.5 ± 16.0
Noxa1−/− (baseline)	83.3 ± 4.5	6.5 ± 1.5	15.9 ± 2.1

Values are means ± SE; n = 8. Noxa1, NADPH oxidase A1.

Figure 1.

Renal blood flow (RBF) measurements showing that renal artery contraction to angiotensin II (ANG II) is impaired in NADPH oxidase A1 (Noxa1)^−/− mice. RBF was measured in young male mice (3–4 mo) after acute injection of ANG II to determine renal artery function. A: representative traces of the RBF change over time with ANG II added at time (t) = 0. Data are means ± SE; n = 3. B: representative traces of the RBF change over time with the thromboxane mimetic U46619 added at t = 0. Data are means ± SE; n = 3. C: changes in RBF in response to acute injection of ANG II with or without N^G-nitro-l-arginine methyl ester (l-NAME; 25 mg/kg) pretreatment. n = 8. *P < 0.01 using an unpaired t test. D: RBF changes in response to acute injection of U46619 with or without l-NAME (25 mg/kg) pretreatment. n = 8. *P < 0.01 and **P < 0.001 using one-way ANOVA. KO, knockout; WT, wild type.

Table 2.

Changes in arterial pressure and renal vascular resistance in response to acute injections of ANG II or U46619 with or without l-NAME (25 mg/kg) pretreatment

	l-NAME	Arterial Pressure, kPa	Renal Blood Flow, % Max Change	Renal Vascular Resistance, % Max Change
Wild type [ANG II (1 ng)]	−	13.8 ± 2.3	−22.5 ± 5.6	50.0 ± 14.9
Noxa1−/− [ANG II (1 ng)]	−	11.6 ± 2.9	−15.2 ± 4.3	31.0 ± 11.6
Wild type [ANG II (1 ng)]	+	13.7 ± 1.2	−32.7 ± 7.5	92.3 ± 37.8
Noxa1−/− [ANG II (1 ng)]	+	11.8 ± 3.4	−18.7 ± 6.2	41.9 ± 18.4
Wild type [ANG II (3 ng)]	−	24.4 ± 2.9	−50.2 ± 5.4	175.2 ± 45.5
Noxa1−/− [ANG II (3 ng)]	−	21.4 ± 5.0	−29.4 ± 4.1**	74.4 ± 15.8
Wild type [ANG II (3 ng)]	+	24.9 ± 3.7	−49.3 ± 10.0	302 ± 155
Noxa1−/− [ANG II (3 ng)]	+	24.1 ± 23.6	−38.0 ± 10.0	148.2 ± 56.5
Wild type [U46619 (25 ng)]	−	18.6 ± 3.1	−13.6 ± 1.6	33.2 ± 3.6
Noxa1−/− [U46619 (25 ng)]	−	18.0 ± 2.1	−15.9 ± 2.1	33.6 ± 6.1
Wild type [U46619 (25 ng)]	+	22.6 ± 5.5	−24.2 ± 4.4†††	54.7 ± 17.4
Noxa1−/− [U46619 (25 ng)]	+	17.1 ± 2.5	−21.1 ± 3.6†††	42.8 ± 10.8
Wild type [U46619 (50 ng)]	−	25.9 ± 4.2	−28.0 ± 2.2	71.5 ± 5.2
Noxa1−/− [U46619 (50 ng)]	−	23.4 ± 2.0	−29.7 ± 3.2	77.0 ± 10.3
Wild type [U46619 (50 ng)]	+	33.8 ± 8.4	−52.6 ± 8.5†††	519 ± 317
Noxa1−/− [U46619 (50 ng)]	+	20.7 ± 2.3	−52.4 ± 7.2†††	176 ± 41.0

Values are means ± SE; n = 8. ANG II, angiotensin II; l-NAME, N^G-nitro-l-arginine methyl ester; Noxa1, NADPH oxidase A1. **P < 0.01 vs. wild-type; †††P < 0.001 vs. no L-NAME.

Another important genotype difference observed was the ability of endogenous NO to blunt acute renal vasoconstriction produced by U46619 but not by ANG II. Based on acute l-NAME effects, transient renal vasoconstriction caused by ANG II was very weakly buffered, if any, by endogenous NO in both wild-type and NOXA1-deficient mice (Table 2 and Fig. 1C). Conversely, NO more effectively blunted renal vasoconstriction induced by U46619 (P < 0.001, two-way ANOVA). In wild-type mice, the maximum RBF decrease caused by 25 ng U46619 increased from 14% to 24% of control RBF after l-NAME inhibition of NOS, and that by 50 ng U46619 increased from 28% to 53% of RBF before injection of the vasoconstrictor. l-NAME treatment produced similar results in NOXA1-deficient animals (Table 2 and Fig. 1D). There were no significant differences in arterial pressure or renal vascular resistance changes in wild-type or Noxa1^−/− mice with either treatment (Table 2).

ANG II Increases NOXA1, NOX1, and ROS

To gain further insight into the NOXA1/NOX1 signaling system, we investigated NOXA1/NOX1 protein changes in kidney compartments after 14 days of ANG II infusion. Chronic ANG II increased NOXA1 protein expression in VSMCs of the renal artery of wild-type mice, as evidenced by colocalization with ACTA2 (Fig. 2A). Moreover, NOX1 expression was induced by ANG II in wild-type mice but not in Noxa1^−/− mice (Fig. 2B). This change was accompanied by elevated superoxide (DHE fluorescence) production in renal arteries of wild-type mice but not those of Noxa1^−/− mice (P < 0.05; Fig. 2C).

Figure 2.

NADPH oxidase (NOX)1/NOXA1 interaction and reactive oxygen species (ROS) are enhanced in renal arteries from wild-type mice treated with angiotensin II (ANG II) but not from Noxa1^−/− mice. Young male (3–4 mo) mice were implanted with miniosmotic pumps containing saline or ANG II for 14 days. Sections of kidney tissue were prepared, and measurements of protein expression and ROS were taken at the level of the hilum. A: representative immunofluorescence images and quantification of increased NOXA1 (red) expression following ANG II treatment in medial smooth muscle cells of the renal artery of wild-type mice as shown by colocalization with α-smooth muscle actin (ACTA2; green). n = 3. B: representative immunofluorescence images and quantification of increased NOX1 (red) protein levels after ANG II treatment in medial smooth muscle cells of the renal artery in wild-type mice but not in Noxa1^−/− mice as shown by colocalization with ACTA2 (green). n = 3–7. C: quantification of superoxide by dihydroethidium (DHE) fluorescence showed increased ROS levels in wild-type mice following ANG II treatment but not in Noxa1^−/− mice. Data are means ± SE; n = 3–5. Scale bar = 100 µm. *P < 0.05 using an unpaired t test (A) or one-way ANOVA (B and C).

Afferent arterioles are major resistance vessels regulating RBF in health and hypertension (56). To the best of our knowledge, NOXA1/NOX1 expression has not been studied in afferent arterioles. We identified afferent arterioles as small-caliber ACTA2-positive arterioles in the renal cortex of mice. Chronic ANG II increased NOXA1 and NOX1 protein expression in VSMCs of afferent arterioles of wild-type mice (Fig. 3, A and B). ROS measurements were performed in isolated afferent arterioles with short-term ANG II stimulation (30 s) and were found to be significantly elevated in wild-type arterioles while remaining unchanged in Noxa1^−/− arterioles (P < 0.05; Fig. 3C). Chronic ANG II treatment significantly induced NOXA1 and NOX1 colocalization in wild-type mice (P < 0.001; Fig. 3D).

Figure 3.

Angiotensin II (ANG II) increased NADPH oxidase (NOX)1 and NOXA1 protein levels in afferent arterioles of wild type but not Noxa1^−/−mice. Young male (3–4 mo) mice were implanted with osmotic minipumps containing saline or ANG II for 14 days. Protein expression in afferent arterioles was examined in kidney cross sections. A: representative immunofluorescence images and quantification of increased NOXA1 (red) protein levels after ANG II treatment in medial smooth muscle cells of afferent arterioles of wild-type mice as shown by colocalization with α-smooth muscle actin (ACTA2; green). n = 3. B: representative immunofluorescence images and quantification of NOX1 (red) protein expression in medial smooth muscle cells of the renal artery, as shown by colocalization with ACTA2 (green), was increased with ANG II treatment in wild-type mice but not in Noxa1^−/− mice. n = 4–6. C: superoxide levels in renal afferent arterioles treated with 10 μM ANG II were determined by 2-OH-ethidium (2-OH-E⁺) HPLC. D: representative immunofluorescence images and quantification of NOXA1 (red) and NOX1 (green) colocalization induced by ANG II treatment. Data are means ± SE. Scale bar = 100 µm. *P < 0.05, **P < 0.01, and ***P < 0.001 using an unpaired t test (A and D) or one-way ANOVA (B and C).

Glomerular injury is commonly caused by NOX-derived superoxide production in hypertension models and is reversed by tempol (57), which mimics SOD. NOX1 has been implicated as a causative factor for glomeruli injury in diabetic mice (58). In this study, we investigated glomeruli, which are circular hypercellular structures adjacent to AQP2-positive distal tubules and found that NOXA1 and NOX1 protein expression was increased in the glomeruli of wild-type but not those of Noxa1^−/− mice infused with ANG II (Fig. 4, A and B). As expected, glomeruli from wild-type mice had increased superoxide production, whereas glomeruli from Noxa1^−/− mice did not (P < 0.0001; Fig. 4C). NOX4 and NOX2 were also evaluated by immunofluorescence staining and were found to be unchanged after ANG II treatment (Supplemental Figs. S1 and S2).

Figure 4.

NADPH oxidase (NOX)1/NOXA1 and reactive oxygen species (ROS) levels are increased in glomeruli of angiotensin II (ANG II)-treated wild type but not Noxa1^−/− mice. Young male mice (3–4 mo) mice were implanted with osmotic minipumps containing saline or ANG II for 14 days. Protein expression and ROS levels were measured in kidney tissues at the level of glomeruli adjacent to aquaporin 2 (AQP2)-positive distal tubules (arrowheads). A: representative immunofluorescence images and quantification of NOXA1 (red) protein in glomeruli showing an increase in wild-type mice treated with ANG II. n = 9. B: representative immunofluorescence images and quantification of NOX1 (red) protein in glomeruli showing increased expression with ANG II treatment in wild-type mice but not in Noxa1^−/− mice. n = 9. C: dihydroethidium (DHE) fluorescence quantification of superoxide showed an increase in ROS in wild-type mice following ANG II treatment but not in Noxa1^−/− mice. Data are means ± SE; n = 4–6. Scale bar = 100 µm. *P < 0.05, **P < 0.01, and ****P < 0.0001 using an unpaired t test (A) or one-way ANOVA (B and C).

Function of the Mesenteric Artery in Wild-Type and Noxa1^−/− Mice

ANG II-induced hypertension adversely affects the renal vasculature (11, 59) as well as other resistance vessels like mesenteric arteries (11, 12). Because we found increased NOXA1/NOX1 expression in resistance arterioles of the kidney, we investigated whether the reactivity of mesenteric resistance arteries differs between mouse genotypes. Noxa1^−/− arteries were less responsive to phenylephrine contraction (P < 0.001), as shown by the dose-response curves (Fig. 5A). In experiments with acute ANG II, mesenteric arteries from Noxa1^−/− mice showed decreased contraction than wild-type mice (P < 0.05; Fig. 5B). Also, a dose-response curve was generated for acetylcholine and sodium nitroprusside in precontracted vessels to measure endothelium-dependent and -independent relaxation, respectively (Fig. 5, C and D). Two-way ANOVA showed a significant difference between mouse strains in response to acetylcholine relaxation (P < 0.01), with a more striking difference (P < 0.0001) in response to sodium nitroprusside. There was a significant interaction between mouse genotype and sodium nitroprusside concentration (P < 0.05) in two-way ANOVA. Differences can also be seen from the comparison between wild-type and Noxa1^−/− mice with respect to the stability of phenylephrine contraction over time (P < 0.001; Fig. 5E).

Figure 5.

Mesenteric resistance arteries are impaired by NADPH oxidase (NOX)1/NOXA1-dependent oxidative stress. Vascular reactivity experiments were performed using second- and third-order mesenteric arteries obtained from young male (3–4 mo) mice. A: concentration-response curve of mesenteric arteries to phenylephrine (PE). B: concentration-response curve of angiotensin II (ANG II) expressed as the percent maximal response to high-K⁺ buffer (KPSS). C: concentration-response curve of endothelium-dependent relaxation to acetylcholine. D: sodium nitroprusside-induced endothelium-independent vasodilation of mesenteric arteries precontracted with PE. E: stability of PE (10⁻⁵ M)-induced contractions over time. Two-way ANOVA was used to examine differences between wild-type (n = 9) and Noxa1^−/− mice (n = 8) and at each concentration tested. The lines were generated using nonlinear sigmoidal curve fitting. Data are means ± SE and were analyzed by two-way ANOVA. *P < 0.05, **P < 0.01, and ****P < 0.0001.

NOXA1 Inhibitor Screening

We have previously reported that NOXA1 binds to NCF1 and regulates NOX1 activity in VSMCs (49). We developed a high-throughput screening strategy to identify potential inhibitors of NOX1 activity by selecting small molecules that interfere with the binding between the SH3 domain of NOXA1 and the PRR domain of NCF1 (Fig. 6A) (54). Using the optimal amounts of NOXA1-SH3v2-His and NCF1-PRR-GST peptides, we established an AlphaScreen assay with an average z-value of 0.833 ± 0.04 and signal-to-background ratio of 45. We then screened a small-molecule library consisting of 22,000 annotated compounds through high-throughput screening. Significant inhibition of the AlphaScreen binding signal was identified as a positive hit. Potential hits were rescreened to eliminate false positives associated with positional and plate bias and counterscreened to eliminate assay-interfering compounds. HEK-293T cells stably expressing NOXA1-NLuc and NCF1-CLuc proteins fused to complementary fragments of firefly luciferase were incubated with test compounds, and a significant reduction in binding signal was determined as a positive hit.

Figure 6.

High-throughput screening identified potential NADPH oxidase (NOX)A1/NOX1 oxidase inhibitors. A: screening strategy to identify potential small-molecule inhibitors of NOXA1-NCF1 binding. B: levels of superoxide were determined by 2-hydroxyethidium HPLC analysis in male mouse vascular smooth muscle cells (VSMCs) incubated for 30 min with 10 μM of test compounds or DPI and then treated with TNF-α. Data were normalized to protein concentration (n = 5). C: fitted concentration-response curves of AlphaScreen signals using test compounds to inhibit NOXA1-SH3v2-His and NCF1-PRR-GST peptide binding. D: concentration-response fitted curve of luciferase protein complementation assay luminescence signals using test compounds to inhibit NOXA1-NLuc and NCF1-CLuc protein binding. E: concentration-response fitted curves of 2-hydroxyethidium HPLC analysis using test compounds to inhibit superoxide production in mouse VSMCs. F: production of superoxide was measured by 2-hydroxyethidium HPLC analysis in male mouse Noxa1^−/− VSMCs treated with TNF-α after incubation with 10 μM of test compounds or DPI for 30 min. Data were normalized to protein concentration (n = 4). G: superoxide generation was quantified in male mouse Ncf1^−/− VSMCs that had been incubated with 10 μM of test compounds or DPI for 30 min and then treated with TNF-α by HPLC analysis (n = 4). H: molecular structure and properties of compound C25. Data are means ± SE and were analyzed by one-way ANOVA. *P < 0.05, **P < 0.01, and ***P < 0.001. ROS, reactive oxygen species.

A total of 30 potential hits were identified following rescreening and counterscreening in control HEK-293T cells. We then tested all hits in a cellular ROS inhibition assay with cultured VSMCs in which NOX1 activity was induced with TNF-α treatment (45, 49). We identified five potential leads that inhibited TNF-α-induced superoxide levels below 50% of diphenyliodonium-inhibitable levels using HPLC analysis of 2-hydroxyethidium (Fig. 6B). Rescreening of all leads was done with AlphaScreen and luciferase complementation assays (Fig. 6, C and D). IC₅₀ values for these compounds in ROS inhibition assays ranged between 8 and 17 μM (Fig. 6E). IC₅₀ values for these compounds in ROS inhibition assays ranged between 8 and 17 μM (Fig. 6E). ROS levels in either VSMC genotype were not affected by any of the five compounds, indicating specific inhibition of NOXA1-NCF1 interaction (Fig. 6, F and G). VSMC cytotoxicity for all five compounds was below 3%, indicating low toxicity. We chose compound C25 for structural analysis because it has many commercially available structural analogs with >75% similarity (Fig. 6H). The preliminary structural analysis of C25 also indicated potential drug likeness [MW 479.5, topological polar surface area 80.8 Å2, lipophilicity cLogP 3.5, SwissADME (60)].

C25 Inhibits Superoxide Production by VSMCs

ANG II has been shown to increase superoxide production in cultured VSMCs (61, 62). We examined the response of VSMCs isolated from wild-type mice to ANG II (100 nM) stimulation without and with ROS inhibitors. Besides being an inhibitor of all NOX isoforms, DPI is also a general flavoprotein inhibitor that targets enzymes that use NAPDH and has been shown to inhibit ROS generation in response to ANG II stimulation. The compound C25 is designed to specifically inhibit the NOXA1/NOX1 complex in the Runge laboratory. As shown in Fig. 7A, ANG II induced ROS production in cultured VSMCs (P < 0.05), whereas inhibitors DPI and C25 inhibited enhanced ROS production (P < 0.05). ANG II-induced contraction of freshly isolated mesenteric arteries was then examined in the presence of C25. Waiting 2 h before subsequent ANG II dose-response curves in the presence of C25 vehicle control (DMSO) eliminated the effects of tachyphylaxis (Fig. 7B). When C25 was added 30 min before the second ANG II challenge, it significantly reduced contractions induced by ANG II (P < 0.05; Fig. 7C).

Figure 7.

C25 inhibits superoxide production in vascular smooth muscle cells (VSMCs) and angiotensin II (ANG II)-induced vascular contraction. VSMCs and intact mesenteric arteries were isolated from young male (3–4 mo) wild-type mice. A: production of superoxide by cultured VSMCs was measured by HPLC using dihydroethidium following stimulation with 100 nM ANG II and inhibitors (n = 4–5). B: ANG II-induced contraction of mesenteric arteries (first contraction) and further contraction 2 h (second contraction) later in the presence of DMSO (vehicle control) (n = 5–9). C: contraction of the mesentery artery in response to ANG II (first contraction) and a subsequent contraction 2 h later (second contraction) with incubation of C25 for 30 min (n = 5–9). Response curves were generated using nonlinear sigmoidal curve fitting. Data are means ± SE and were analyzed by one-way ANOVA (A) and two-way ANOVA (B and C). *P < 0.05.

DISCUSSION

Renal vasoconstriction may contribute to Na⁺ retention, which is a causal pathophysiological mechanism in ANG II-induced hypertension (8, 63, 64). The goal of the present study was to evaluate renal vascular reactivity to acute ANG II and the TP receptor agonist U46619 in wild-type and Noxa1^−/− mice. In addition, we examined whether endogenous NO or direct actions on VSMCs regulate NOXA1/NOX1-mediated vasoconstriction.

This is the first report to demonstrate the involvement of NOXA1/NOX1 signaling in ANG II-induced renal vasoconstriction in vivo. ANG II stimulates the expression of NOX1 in VSMCs and kidneys, where NOXA1/NOX1 produces primarily intracellular O₂^•− (1, 21, 65, 66). NOX1 plays a role in O₂^•−-mediated Ca²⁺ mobilization and the contraction response of ANG II of aortic and cerebral VSMCs (1, 67, 68) and the hypertrophic response to chronic ANG II and impaired endothelium-mediated dilation of aortic VSMCs (23). In addition, genetic deletion of NOX1 promotes translocation of AT₁ receptors from the plasma membrane to the cytoplasm with reduced ANG II stimulation of cytosolic Ca²⁺ concentration (67). In afferent arteriolar VSMCs, ANG II rapidly stimulates NOX-derived O₂^•− production and Ca²⁺ release from sarcoplasmic reticulum stores. whereas H₂O₂ inhibits Ca²⁺ increases (69). In addition, superoxide rapidly depolarizes the plasma membrane and increases cytosolic Ca²⁺ by activating L-type Ca²⁺ entry channels in VSMCs, independent of endothelial NO (70, 71). At the whole kidney level, renal vasoconstriction caused by ANG II activation of AT₁ receptors is mediated by a combination of Ca²⁺ entry into VSMCs through L-type Ca²⁺ channels and Ca²⁺ mobilization from internal sarcoplasmic reticular stores (72, 73). Accordingly, we found that inhibition of NOXA1 and NCF1 (p47^phox) interaction using C25 inhibited ANG II-induced O₂^•− generation and arterial contraction.

Our mouse RBF experiments reveal a marked difference between ANG II/AT₁ receptors and U46619/TP receptors in renal vascular signaling. In contrast to the lack of contribution to the renal hemodynamic response to the thromboxane mimetic U46619, NOXA1/NOX1 signaling plays an important role in acute renal vascular reactivity to ANG II. ANG II elicits vasoconstriction that is not buffered by local endogenous NO, suggesting a direct action of NOX1-derived O₂^•− on VSMCs rather than dominant O₂^•− scavenging of NO and decreasing NO bioavailability (74). Conversely, renal TP receptors induce NO-mediated buffering of renal vasoconstriction independently of NOXA1/NOX1 stimulation of ROS. VSMCs in renal arteries, afferent arterioles, and glomeruli express NOXA1, which is stimulated by ANG II to produce O₂^•−. Thus, such stimulation is selective for ANG II but not for U46619. The ANG II/AT₁ receptor and U46619/TP receptor signaling differences support our previous finding that acute ANG II-induced renal vasoconstriction is independent of TP receptor stimulation and TxA₂-isoprostane production (75). Earlier studies have shown that acute TxA₂/TP receptor activation is associated with NOX-dependent O₂^•− generation that contributes to contraction of the SHR aorta (38) and that acute U46619/TP receptor activation causes NOX-dependent O₂^•− generation that contributes to contraction of VSMCs (41) and the entire kidney (32). The mechanisms through which U46619 stimulates O₂^•− (32) are unknown, but our findings indicate that it might involve a NOX other than NOX1. Muzaffar et al. (76, 77) found that U46619 stimulated NOX2 expression and O₂^•− production in cultured pig pulmonary arterial VSMCs.

Oxidative stress in cardiovascular disease is associated with reduced NO bioavailability. ANG II increases O₂^•−, reduces NO action, and induces endothelial dysfunction in the systemic vasculature (78, 79). Stimulation of protein kinase C by ANG II results in upregulation of NOX1, NOX2, and endothelial NOS (eNOS) (80). Chronic upregulation of NOX1 in VSMCs under pathophysiological conditions can result in the uncoupling of eNOS and a decrease in NO bioavailability, leading to impaired vasorelaxation (81). eNOS uncoupling also reduces NO/cGMP/cGK-1 signaling, further increasing O₂^•− production (82). The extracellular SOD limits O₂^•− concentration and its quenching of NO to improve endothelium-dependent vasodilation (82, 83). Chronic NO inhibition causes an increase in O₂^•− production in VSMCs, resulting in an increase in cytosolic Ca²⁺ concentration (84). Acute inhibition of NOS with l-NAME causes systemic hypertension and renal vasoconstriction (85). Inhibition of NOS increases arterial pressure by ∼30 mmHg and reduces renal cortical blood flow by ∼20% in wild-type C56BL/6J mice without causing any hemodynamic changes in mice lacking eNOS (86). A 54% increase in arterial pressure in the rat from NOS inhibition is accompanied by a 275% increase in renal vascular resistance and a 54% decrease in RBF (87). Our results are consistent with the previous reports and provide new evidence that the tonic dilator action of NO is independent of NOX1 under basal conditions in both wild-type mice and mice deficient in NOX1/NOXA1 signaling.

Physiological sample preparation, species, and agonist can all affect the effectiveness of NO in blunting agonist-induced renal vasoconstriction. We found that endogenous NO blunted renal vasoconstriction induced by U46629 in vivo but not ANG II in anesthetized mice. Similar to our observation, an earlier study found that acute ANG II decreased renal cortical perfusion in wild-type mice both before and after l-NAME treatment and also in mice lacking eNOS (86). As shown by our results, TP receptors, but not AT₁ receptors, stimulate cytosolic Ca²⁺ and eNOS to compensate for agonist-induced renal vasoconstriction in mice, rather than transfer Ca²⁺ from contracting VSMCs to endothelial cells (88, 89). Another study of isolated perfused mouse kidneys reported a higher EC₅₀ dose and maximum vasoconstriction induced by ANG II in kidneys lacking eNOS compared with mice lacking neuronal NOS or wild-type mice, suggesting that NO production by eNOS blunts or buffers ANG II-induced vasoconstriction (90). In the kidneys of rats, acute ANG II increases eNOS activity and NO production with NO/cGMP signaling, blunting ANG II-induced renal vasoconstriction (91–93). ANG II/AT₁ receptor stimulates NO production by rat afferent arterioles, and NOS/NO inhibition magnifies acute contractions caused by ANG II (94–96). In diabetes mellitus, however, excess O₂^•− impairs the buffering effect of NO on ANG II-induced vasodilation of afferent arterioles (97). To our knowledge, such a relationship has never been tested in vivo in mouse kidneys. In isolated afferent and efferent arterioles of wild-type mice, eNOS generated NO that counteracts acute constriction produced by ANG II (98, 99). Prostanoids, but not NO, blunt ANG II-induced vasoconstriction in some rat vascular beds. For example, cyclooxygenase (COX) inhibition potentiates renal vasoconstriction to ANG II, whereas l-NAME inhibition of NOS enhances constriction of mesenteric arteries but not of renal arteries (100). It has also been previously reported that vasoactive eicosanoids and TP receptors mediate acute norepinephrine-induced renal vasoconstriction and NO release in the preglomerular vasculature of rats (101). At the whole kidney level, we were surprised to find that endogenous NO does not buffer the magnitude of ANG II-induced renal vasoconstriction in the mouse, either wild-type or Noxa1^−/−. The previously published studies and our observations demonstrate the differences in contraction that can be seen between isolated vessel preparations, ex vivo whole kidney experiments, and in situ readings of renal arteries in live animals.

Our myograph data show that NOX/NOXA1 signaling contributes to the magnitude of contraction of mesenteric resistance arteries of wild-type mice to ANG II and phenylephrine, with less pronounced vasoconstriction of arteries of Noxa1^−/− mice. Pagano and coworkers (61) have previously demonstrated that isolated mesenteric arteries from Nox1^−/− mice had impaired contraction to ANG II while responding relatively normally to phenylephrine. There is no clear explanation for the difference in phenylephrine results. Consistent with our results, both ANG II and U46619 reduce mesenteric arterial blood flow by increasing VSMC cytosolic Ca²⁺ by a combination of increased Ca²⁺ entry and release from sarcoplasmic reticular Ca²⁺ stores (102), whereas relaxation is caused by a NO/cGMP-mediated reduction in cytosolic Ca²⁺ (103). The NOXA1/NOX1 signaling pathway contributes to vasoconstriction by increasing cytosolic Ca²⁺ concentration (84, 104). In the absence of NOXA1 and nonfunctional NOX1, Ca²⁺ handling may be altered enough to limit the contraction of VSMCs. Furthermore, sensitivity to NO relaxation is likely increased without NOX1/NOXA1 production of O₂^•−. This may explain our findings of attenuated agonist-induced contractions and enhanced reactivity to NO in arteries lacking NOXA1. Relevance to pathophysiology is shown by combined inhibition of NOX1 and NOX4 attenuating portal hypertensive syndrome via modulation of mesenteric arterial hyporeactivity and angiogenesis in rats (105). Interpretation of our myography results is unaffected by the stability of phenylephrine contraction due to the α-adrenergic-induced endothelial NO relaxation after phenylephrine-induced contractions (106). Assessment of vascular relaxation with acetylcholine and sodium nitroprusside after phenylephrine application began before the onset of α-adrenergic relaxation, progressed faster than its timeline, and was of greater magnitude, thereby indicating that the responses are due to the agonists tested and not by α-adrenergic relaxation. The trials were always run in parallel with vessels of wild-type and Noxa1^−/− mice tested concurrently to provide a direct comparison of genotypes.

Previous basal myography experiments in p47phox^−/− mice showed increased vascular reactivity to acetylcholine-induced relaxation in conjunction with a reduction in baseline blood pressure (107). In this study, Noxa1^−/− mice exhibited a similar relaxation response induced by acetylcholine that was exacerbated by endothelium-independent relaxation agents. This implies that the response of the resistance artery to NO is specifically impaired, although the acetylcholine-stimulated endothelium can cause vascular relaxation through multiple mechanisms (108, 109). Alternatively, acetylcholine may stimulate a vasoconstrictor agent that opposes the dilatory effect of NO, an agent not involved in the vascular response to the direct NO donor sodium nitroprussinde. The increased vascular relaxation of mesenteric arteries of Noxa1^−/− mice in response to sodium nitroprusside supports weaker renal vasoconstriction and lower RBF in Noxa1^−/− mice treated with ANG II. We have previously observed reduced baseline blood pressure in Noxa1^−/− mice based on time-course data in a hypertension experiment (7). A one-time blood pressure measurement in this study failed to achieve significance likely due to the fact that mean arterial pressure was investigated as opposed to systolic pressure in our previous study. There was a trend toward lower mean arterial pressure in Noxa1^−/− mice in the present study. Together with data from p47phox^−/− mice, these results strongly suggest that NOX1 signaling regulates vascular tone and blood pressure.

A limitation of our study was that we were limited to male mice since we and others have observed that female mice are not susceptible to ANG II-induced hypertension (7, 110). It is hypothesized that estrogen inhibits NOX activity and ROS generation, which results in resistance to ANG II-induced hypertension in female mice (110). We also observed less upregulation of ROS, NOXA1, and renal collecting duct ENaC levels in female mice than in male mice when exposed to ANG II (7).

In conclusion, our results indicate that NOX1/NOXA1/O₂^•− signaling contributes to acute ANG II-induced vasoconstriction in the renal and mesenteric beds of wild-type mice and thus in hypertension induced by ANG II. In contrast, U46619-induced renal vasoconstriction occurs independently of NOX1/NOXA1. A second distinction between the signaling of the two agonists in the mouse is that renal vasoconstriction induced by ANG II is not blunted by endogenous NO, indicating a direct action of ANG II signaling in renal resistance arteries/arterioles independent of NO bioavailability, whereas U46619 stimulates vasoconstriction that is effectively blunted by NO in both wild-type and Noxa1^−/− mice.

DATA AVAILABILITY

Data will be made available upon reasonable request.

SUPPLEMENTAL DATA

Supplemental Fig. S1: https://doi.org/10.6084/m9.figshare.20984104.v1.

Supplemental Fig. S2: https://doi.org/10.6084/m9.figshare.20984764.v1.

GRANTS

This work was supported by National Heart, Lung, and Blood Institute Grant HL139842.

DISCLOSURES

M.S.R. is a member of the Board of Directors at Eli Lilly and Company. None of the other authors has any conflicts of interest, financial or otherwise, to disclose.

AUTHOR CONTRIBUTIONS

M.D.S., M.S.R., W.J.A., and N.R.M. conceived and designed research; M.D.S., A.E.V., X.Y., Y.C., H.A.N., N.M., and M.S.R. performed experiments; M.D.S., A.E.V., and M.S.R. analyzed data; M.D.S., W.J.A., and N.R.M. interpreted results of experiments; M.D.S., A.E.V., and M.S.R. prepared figures; M.D.S., A.E.V., and N.R.M. drafted manuscript; M.D.S., W.J.A., and N.R.M. edited and revised manuscript; M.D.S., W.J.A., and N.R.M. approved final version of manuscript.