Angiotensin II stimulates superoxide production in the thick ascending limb by activating NOX4

Abstract

Angiotensin II (ANG II) stimulates production of superoxide (O₂⁻) by NADPH oxidase (NOX) in medullary thick ascending limbs (TALs). There are three isoforms of the catalytic subunit (NOX1, 2, and 4) known to be expressed in the kidney. We hypothesized that NOX2 mediates ANG II-induced O₂⁻ production by TALs. To test this, we measured NOX1, 2, and 4 mRNA and protein by RT-PCR and Western blot in TAL suspensions from rats and found three catalytic subunits expressed in the TAL. We measured O₂⁻ production using a lucigenin-based assay. To assess the contribution of NOX2, we measured ANG II-induced O₂⁻ production in wild-type and NOX2 knockout mice (KO). ANG II increased O₂⁻ production by 346 relative light units (RLU)/mg protein in the wild-type mice (n = 9; P < 0.0007 vs. control). In the knockout mice, ANG II increased O₂⁻ production by 290 RLU/mg protein (n = 9; P < 0.007 vs. control). This suggests that NOX2 does not contribute to ANG II-induced O₂⁻ production (P < 0.6 WT vs. KO). To test whether NOX4 mediates the effect of ANG II, we selectively decreased NOX4 expression in rats using an adenovirus that expresses NOX4 short hairpin (sh)RNA. Six to seven days after in vivo transduction of the kidney outer medulla, NOX4 mRNA was reduced by 77%, while NOX1 and NOX2 mRNA was unaffected. In control TALs, ANG II stimulated O₂⁻ production by 96%. In TALs transduced with NOX4 shRNA, ANG II-stimulated O₂⁻ production was not significantly different from the baseline. We concluded that NOX4 is the main catalytic isoform of NADPH oxidase that contributes to ANG II-stimulated O₂⁻ production by TALs.

superoxide (o₂⁻) is a reactive oxygen species that plays an important role in the regulation of kidney function (13, 19, 86). Superoxide reduces urinary sodium excretion by lowering glomerular filtration rate through vasoconstriction (58), enhancing tubular-glomerular feedback (57, 57) and increasing sodium reabsorption in the thick ascending limb (70). In the thick ascending limb of the loop of Henle, O₂⁻ increases sodium chloride absorption by enhanced Na-K-Cl cotransport activity (41) and also increases bicarbonate absorption through enhanced Na/H exchange (42). Elevated concentrations of renal O₂⁻ have been implicated in the development of hypertension (60) and renal damage (61, 73).

O₂⁻ can be generated by NADPH oxidase, cyclooxygenase, xanthine oxidase, uncoupled nitric oxide synthase (2), and mitochondria (91); however, NADPH oxidase is the primary source of O₂⁻ production in the thick ascending limb (37, 52, 86). The NADPH oxidases are a family of enzymes composed of multiple homologous subunits. The catalytic NOX subunit has five isoforms (NOX1, NOX2, NOX3, NOX4, and NOX5) and two related enzymes, DUOX1 and DUOX2, identified thus far (43). NOX5 is not expressed in rodents (64), and NOX3 appears to be expressed only in the inner ear of human adults (47). The DUOX (dual oxidase) members are substantially different in structure and function from the NOXs with an additional transmembrane helix and long NH₂-terminal domain that appears to be involved in exclusive H₂O₂ production in a Ca⁺²-dependent manner (50). They have been detected in the thyroid and salivary glands as well as the mucosal epithelium of certain gastrointestinal and respiratory tracts (24).

Chabrashvili et al. (7) found that kidneys of spontaneously hypertensive rats express NOX1, 2, and 4 but did not specifically study the isoforms of the catalytic subunit expressed in the thick ascending limb. Yang et al. (97) reported that thick ascending limbs express NOX2 and 4, but they did not study NOX1. Thus we know of no previous reports of NOX1 being expressed in the thick ascending limb. To our knowledge, there are no reports that the DUOXs are expressed in the kidney.

In the kidney, angiotensin II (ANG II) reduces urinary sodium excretion by reducing renal blood flow (40), enhancing tubuloglomerular feedback and stimulating sodium reabsorption along the nephron (20, 48). At least some of these actions are due to elevated O₂⁻ levels resulting from the stimulation of O₂⁻ production by NADPH oxidase (34, 46, 74). In endothelial cells, ANG II acutely increases O₂⁻ production by activating both NOX2- and NOX4-based NADPH oxidases (8, 30, 66, 87, 98). In the macula densa, NOX2 mediates the acute effects of ANG II on O₂⁻ production (21). Chronically, ANG II increases O₂⁻ production and NOX1 expression in vascular smooth muscle cells (10, 23, 49, 94). In the thick ascending limb, ANG II acutely stimulates O₂⁻ production (65) via the ANG II type 1 (AT₁) receptor; however, the NOX isoforms(s) responsible has yet to be determined.

Given the macula densa shares many characteristics with the thick ascending limb, we tested the hypothesis that ANG II-induced O₂⁻ production is mediated by NOX2-based NADPH oxidase in the thick ascending limb.

METHODS

Animals

Male Sprague-Dawley rats obtained from Charles River Breeding Laboratories (Wilmington, MA) were maintained on chow containing 1.1% K and 0.22% Na (Purina, Richmond, IN) for ≥5 days before use. Wild-type and NOX2 knockout mice (Jackson Laboratories, Bar Harbor, ME) were fed regular chow. Food and water were provided ad libitum. All protocols involving animals were approved by the Henry Ford Hospital Animal Care and Use Committee (Institutional Animal Care and Use Committee) in accordance with the National Institutes of Health's Guidelines for the Care and Use of Laboratory Animals.

Medullary Thick Ascending Limb Suspensions

Animals were anesthetized with ketamine (100 mg/kg ip) and xylazine (20 mg/kg ip). The abdominal cavity was opened, and the kidneys were flushed with 40 ml ice-cold 0.1% collagenase in physiological saline containing the following (in mM): 130 NaCl, 2.5 NaH₂PO₄, 4 KCl, 1.2 MgSO₄, 6 l-alanine, 1 Na₂ citrate, 5.5 glucose, 2 Ca dilactate, and 10 HEPES pH 7.4 via retrograde perfusion of the aorta. Kidneys were then removed, and the inner strip of the outer medulla was isolated, minced, and incubated in collagenase at 37°C for 30 min, with gentle agitation of the tissue every 5 min and gassing it with 100% oxygen. The collagenase solution was removed by centrifugation at 93 g, and the tubules were resuspended in oxygenated, ice-cold physiological saline and gently stirred on ice for another 30 min. The tissue was filtered through a 250-μm nylon mesh, and the filtered material was rinsed with ice-cold physiological saline. Approximately 100 μg of total protein were used for each sample when O₂⁻ was measured.

Some minor changes were required for the mouse thick ascending limb suspensions to optimize yield. The osmolality of the physiological saline was increased to 320 mosmol/kgH₂O with mannitol, the collagenase concentration was increased to 0.15%, and hyaluronidase (0.05%) was added. The times for collagenase incubation and stirring were reduced to 20 min, and the volume was reduced to 8 ml.

Western Blot

Thick ascending limb suspensions were obtained as described above and placed on ice (see Fig. 3). Tubules were then centrifuged and lysed by being vortexed in 100 μl of buffer containing 20 mM HEPES (pH 7.4), 2 mM EDTA, 300 mM sucrose, 0.1% IGEPAL, 0.1% SDS, 5 μg/ml antipain, 10 μg/ml aprotinin, 5 μg/ml leupeptin, 4 mM benzamidine, 5 μg/ml chymostatin, 5 μg/ml pepstatin A, and 0.105 M 4-(2-aminoethyl)benzenesulfonyl fluoride (Sigma-Aldrich, St. Louis, MO). Samples were centrifuged at 6,000 g for 5 min at 4°C, and protein content in the supernatant was measured using Coomassie Plus reagent (Pierce). Fifty micrograms of of total protein were loaded into each lane of a SDS-polyacrylamide gel. Ten-percent acrylamide gels were used for NOX1 and 8% acrylamide for NOX2 and 4. Proteins were separated by electrophoresis and transferred to a polyvinylidene difluoride membrane (Millipore, Bedford, MA). The membrane was incubated in blocking buffer comprised of 5% nonfat milk in TBS-T (20 mM Tris pH 7.6, 137 mM NaCl, and 0.1% Tween 20) for 60 min and then with a 1:1,000 dilution of each NOX isoform-specific antibody in blocking buffer for 2 h at room temperature. The NOX1 antibody was a rabbit polyclonal antibody (Santa Cruz, Santa Cruz, CA). The NOX2 antibody was a mouse polyclonal antibody (Abcam, Cambridge, MA). The NOX4 antibody was a rabbit polyclonal antibody (Abcam). The membrane was washed with TBS-T and incubated with a 1:1,000 dilution in blocking buffer of a secondary antibody against the appropriate IgG conjugated to horseradish peroxidase (Amersham Biosciences) for 60 min at room temperature. The membrane was washed with TBS-T, and the reaction products were detected with an enhanced chemiluminescence kit (Pierce ECL Western Blotting Substrate). The membrane was exposed to Fuji RX film and band analysis was performed using an Epson Expression 1680 scanner and density analysis software. Thick ascending limb suspensions were obtained as described above (see Fig. 7). The suspensions were then separated into two, 200-μl aliquots: control and ANG II. Following the treatment with vehicle or ANG II for 5 min at 37°C supplied with oxygen, the samples were processed as just described for Western blots.

Fig. 3.

Protein expression of each NADPH oxidase isoform in 50 μg of total protein from isolated thick ascending limbs. Each Western blot was probed with isoform-specific antibodies. NOX, NADPH oxidase.

Fig. 7.

Western blot showing the effect of 1 nM ANG II for 5 min at 37° on NOX4 protein expression using an isoform-specific antibody.

Superoxide Production Using a Lucigenin-Based Assay

Two-hundred-microliter aliquots of thick ascending limb suspensions were placed in glass tubes, 700 μl of warm, oxygenated physiological saline was added, and samples were placed in a dry bath at 37° C. Lucigenin (N,N′-dimethyl-9,9′-biacridinium dinitrate; Sigma-Aldrich) was added to a final concentration of 5 μM, and the tubules were incubated for 10 min at 37°C with 100% oxygen supplied. ANG II (1 nM) or vehicle was added 5 min into the 10-min incubation with lucigenin. Samples were placed in a luminometer (model FB12/Sirius; Zylux Oak Ridge, TN) gassed with 100% oxygen and maintained at 37°C. Depending on the experimental conditions, isolated thick ascending limbs were also incubated with or without 100 μM apocynin. When studying the role of NOX4, the left kidneys were transduced with an adenovirus carrying an RNA silencing sequence against rat NOX4 as described below. Data were acquired at a rate of four measurements per second and the average luminescence per second was accumulated for 6 min. The O₂⁻ scavenger 4,5-dihydroxy-1,3-benzenedisulfonic acid (Tiron; Sigma-Aldrich) was added to a final concentration of 10 mM, and measurements were continued for 10 min. The average luminescence of the last 2 min following the addition of Tiron was then subtracted from every data point before the addition of Tiron as a means of subtracting background or nonsuperoxide luminescence. Measurements were normalized for protein content.

NOX4 Short Hairpin RNA

Recombinant replication-deficient adenoviruses encoding a short hairpin RNA (shRNA) sequence for rat NOX4 (AdNox4shRNA) under the control of the H1 mouse RNA polymerase promoter were constructed by ViraQuest (North Liberty, IA) as described previously (78). A 63-base annealed oligomer containing a shRNA sequence was produced (sense: C CGT TTG CAT CGA TAC TAA; bases 1358–1376 in GeneBank seq NM_053524; hairpin loop linker and antisense: TTA GTA TCG ATG CAA ACG G).

Viral injection.

Rat medullary thick ascending limbs were transduced in vivo with recombinant replication-deficient adenoviruses carrying the shRNA sequence for rat NOX4 (AdNOX4shRNA) as reported previously (71, 72, 78). Rats were anesthetized with ketamine (60 mg/kg ip) and xylazine (20 mg/kg ip), and the left kidney was exposed via a left flank incision, the fatty tissue around the renal pole was opened, and the renal artery and vein was clamped. Then, four 20-μl virus injections (4 × 10¹⁰ plaque-forming units/ml) were made using a custom-made 30-gauge needle, specific for the depth of the outer medulla and attached to polyethylene-10 tubing connected to a syringe pump (Harvard Apparatus, Holliston, MA) set at 20 μl/min. The needle was inserted perpendicularly to the renal capsule, parallel to the medullary rays and directed toward the medulla. Injections were made along the longitudinal axis of the kidney at 2.5-mm intervals. To avoid bleeding and leakage of the virus, the needle remained in place for 30 s after each injection was completed. The renal artery was unclamped after 8 min, the kidney was returned to the abdominal cavity, and the incision was sutured. Previously, we (71) have shown that ≥80% of thick ascending limb cells can be transduced using this technique. Thick ascending limbs from each kidney were isolated, as described above, 6–7 days posttransduction for the studies involving the NOX4 shRNA, following the time frame we established when determining when peak reduction of message was achieved.

RT-PCR.

Separate TAL suspensions from the adenovirus-injected left kidney and noninjected right kidney were prepared as described previously (79). Total RNA was isolated from the suspensions using an RNeasy Plus mini kit (Qiagen, Valencia, CA) and subsequently used for RT-PCR analysis using a OneStep RT-PCR kit (Qiagen) according to the manufacturer's instructions. Previously published NOX isoform-specific PCR primers (84) were used to assay mRNA expression. Thermal cycler conditions for OneStep RT-PCR were as follows: 1) reverse transcription: 30 min at 50°C; 2) HotStarTaq DNA polymerase activation at 95°C for 15 min; 3) cycling at 94°C for 15 s, 60°C for 5 s, and 72°C for 1 min; and 4) final extension at 72°C for 5 min. The total number of cycles used for detection of NOX isoforms in rat thick ascending limbs (Table 1) was 35 cycles using 1 μg RNA. Because of the differences observed in mRNA expression among the three NOX isoforms in thick ascending limbs, the number of cycles and amount of template RNA was adjusted to evaluate knockdown of NOX mRNA by an adenovirally delivered NOX4-specific shRNA (Fig. 6). For NOX 1 mRNA knockdown analysis, the PCR primers were as follows: 5′-GGAGTCCTCATTTTGTGGGGCAACC-3′ (forward) and 5′-GGCACCCGTCTCTCTACAAATCCAGT-3′ (reverse). The total number of cycles used for NOX1, NOX2, and NOX4 was 35, 30, and 31, respectively, and the amount of RNA used was 400, 100, and 200 ng. PCR products were analyzed on 1.5% agarose gels and stained with ethidium bromide (0.5 μg/ml). Band density of the expected 560-, 605-, and 409-bp amplicons for NOX1, NOX2, and NOX 4 was analyzed. Aliquots from samples of suspensions used for the NOX4 knockdown experiments in which O₂⁻ was measured were taken before the O₂⁻ measurements and used for RT-PCR to assure there were consistent reductions in expression of NOX4.

Table 1.

mRNA abundance of NOX1, NOX2, and NOX4 in equal amounts of thick ascending limb from rats and mice as determined by RT-PCR

Catalytic Subunit, OD units	Rat	Mouse
NOX1	8.55	3.13
NOX2	462.42	223.16
NOX4	120.55	119.77

OD, optical density; NOX, NADPH oxidase.

Fig. 6.

Effect of adenovirus delivery of NOX4 shRNA on NOX4 mRNA measured 6 or 7 days following transduction. (*P < 0.0001; n = 7).

Statistical Analysis

Results are expressed as means ± SE. Statistical analyses were performed by the Henry Ford Hospital Department of Biostatistics and Epidemiology. Data were evaluated using Student's t-test for paired experiments, Student's t-test of paired differences or ANOVA using Hochberg's method to adjust for multiple testing, taking P < 0.05 as significant.

RESULTS

We first investigated the effect of 1 nM ANG II on O₂⁻ production in rat thick ascending limb suspensions. Suspensions treated with vehicle produced 183 ± 36 relative light units (RLU/mg protein) O₂⁻ in 5 min. Tubules treated with 1 nM ANG II produced 358 ± 79 RLU/mg protein in the same time, an increase of 96% (Fig. 1). Controls showed no significant increase. These data show that ANG II increases O₂⁻ production in thick ascending limb suspensions.

Fig. 1.

Effect of 1 nM ANG II on net O₂⁻ production in isolated rat thick ascending limbs. RLU, relative light units. Tubules were exposed to ANG II for 5 min (*P < 0.03; n = 6).

To test whether the effect of ANG II on O₂⁻ production is due to activation of NADPH oxidase, we used apocynin, an NADPH oxidase inhibitor. Tubules treated with 1 nM ANG II produced 358 ± 79 RLU/mg protein O₂⁻, while those treated with ANG II plus 100 mM apocynin only produced 219 ± 62 RLU/mg protein (Fig. 2A). Apocynin had no significant effect on basal O₂⁻ production (Fig. 2B). Thus apocynin significantly reduced ANG II-stimulated O₂⁻ production.

Fig. 2.

A: effect of 100 mM apocynin on net ANG II-induced O₂⁻ production in isolated thick ascending limbs. Tubules were exposed to apocynin for 10 min and ANG II for 5 min before acquisition of data. (*P < 0.02; n = 6). B: effect of 100 mM apocynin on basal O₂⁻ production in isolated thick ascending limbs. Tubules were exposed to apocynin for 10 min and ANG II for 5 min before acquisition of data (NS; n = 6).

To begin to test which NOX isoform is responsible for ANG II-stimulated O₂⁻ production, we studied which of the catalytic subunits shown to be expressed in the kidney are present in thick ascending limbs. With the use of isoform-specific antibodies, NOX1, 2, and 4 expression was detected in thick ascending limbs. Figure 3, top, middle, and bottom, shows an aliquot from a rat thick ascending limb preparation. The three bands in Fig. 3, top, are splice variants of the NOX1 isoform, which depending on species can range from 55–60 kDa (5). The Fig. 3, middle and bottom, illustrates the presence of a single variant of NOX2 and NOX4. mRNA for NOX1, 2, and 4 was also found in thick ascending limbs from both rats and mice (Table 1).

To assess the contribution of NOX2-based NADPH oxidase to ANG II-stimulated O₂⁻ production, we used NOX2 knockout mice. This is possible because thick ascending limbs from both mice and rats express NOX1, 2, and 4 (Table 1). ANG II-induced O₂⁻ production was 346 ± 63 RLU/mg protein in the wild-type mice (n = 9; Fig. 4). Similarly, ANG II-stimulated O₂⁻ production in the NOX2 knockout mice was 290 ± 78 RLU/mg protein (n = 9) (Fig. 4). These data suggest that NOX2-based NADPH oxidase does not contribute to ANG II-stimulated O₂⁻ production in thick ascending limbs.

Fig. 4.

Effect of 1 nM ANG II on net O₂⁻ production in isolated thick ascending limbs from wild-type or NOX2 knockout mice. Tubules were exposed to ANG II for 5 min before acquisition of data. (n = 9).

To determine whether NOX4 mediates the effect of ANG II in thick ascending limbs, we used isoform-specific shRNA. ANG II-stimulated O₂⁻ production was 448 ± 54 RLU/mg protein in nontransduced thick ascending limbs, but only 186 ± 23 RLU/mg protein in thick ascending limbs transduced with NOX4 shRNA (Fig. 5). We next tested whether the NOX4 shRNA affected basal O₂⁻ production. In the absence of ANG II, baseline O₂⁻ production was 141 ± 32 RLU/mg protein in the nontransduced thick ascending limbs and 138 ± 34 RLU/mg protein in the thick ascending limbs transduced with NOX4 shRNA. These data indicate that the NOX4 shRNA does not alter baseline O₂⁻ production. NOX4 mRNA was reduced by 77 ± 4% within 6–7 days, as determined by RT-PCR (Fig. 6). Transducing the thick ascending limb with NOX4 shRNA had no effect on NOX1 and NOX2 expression (Fig. 6).

Fig. 5.

Effect of adenovirus delivery of NOX4 short hairpin (sh)RNA on ANG II-induced O₂⁻ production in isolated thick ascending limbs. Tubules were isolated 6 or 7 days after transduction with NOX4 shRNA and treated with 1 nM ANG II for 5 min before acquisition of data (*P < 0.005; n = 7).

Given that 1) the nature of our experimental design prevented us from testing all four treatments (basal, shRNA, ANG II, and ANG II plus shRNA) all at once; and 2) we did not achieve 100% knockdown of the NOX4 mRNA, it is conceivable that there is a residual acute effect of ANG II that is due to ANG II rapidly stimulating NOX4 expression as has been seen in mesangial cells (6). To test this, we performed Western blot to study whether a 5-min exposure to ANG II increases NOX4 expression. Figure 7 is a representative blot that shows that a 5-min exposure to 1 nM ANG II has no effect on the abundance of NOX4 protein (cumulative data: 0.82 vs. 0.88 arbitrary units control vs. ANG II, respectively; n = 3) in thick ascending limbs. This greatly reduces the likelihood that there is any residual acute effect of ANG II on O₂⁻ production after knocking down NOX4 that is due to a rapid stimulation of NOX4 expression.

Finally, we tested whether adding apocynin to suspensions transduced with NOX4 shRNA further reduced O₂⁻ production in the presence of ANG II. Adding apocynin had no further effect on ANG II-stimulated O₂⁻ production (Fig. 8). These data indicate that there is no apocynin-inhibitable NADPH oxidase activity in thick ascending limb suspensions transduced with NOX4 shRNA.

Fig. 8.

Effect of 100 mM apocynin on net ANG II-induced O₂⁻ production in isolated thick ascending limbs transduced with NOX4 shRNA. Tubules were exposed to apocynin for 10 min and ANG II for 5 min before acquisition of data (n = 8).

DISCUSSION

We hypothesized that ANG II stimulates O₂⁻ production in the thick ascending limb via activation of NOX2-based NADPH oxidase. To address this hypothesis, we first needed to confirm that NADPH oxidase is the source of O₂⁻ in ANG II-stimulated thick ascending limbs. To do this, we used the NADPH oxidase inhibitor apocynin. We found that apocynin significantly inhibited ANG II-stimulated O₂⁻ production but had no effect on basal production. The limits of the experimental design prevented us from testing all four treatments at once (basal, basal plus apocynin, ANG II, and ANG II plus apocynin). However, the measured O₂⁻ production in suspensions treated with ANG II plus apocynin was similar to the rates measured in untreated tubules and those treated with apocynin alone. These data demonstrate that essentially all ANG II-stimulated O₂⁻ is generated by apocynin-sensitive NADPH oxidase(s) in this segment. These results are similar to those reported by us (34) and other investigators (52).

We next set out to identify the isoform of NADPH oxidase that is involved in the observed response to ANG II. First, we studied which isoforms of the catalytic subunit of NADPH oxidase found in the kidney are expressed in the thick ascending limb by Western blot and RT-PCR. Three of the isoforms of the catalytic subunit of NADPH oxidase are expressed in the kidney; NOX1, 2, and 4 (25). Thus we only examined whether thick ascending limbs express these three isoforms; we were able to detect all three at both the mRNA and protein level (Fig. 3).

Our results regarding NOX 2 and 4 are consistent with those of Yang et al. (97) who previously detected these isoforms in the thick ascending limb. Other investigators have also shown that proximal tubules (15, 15, 51, 68, 76, 81) and macula densa cells (21) express either NOX2, 4, or both. Although NOX1 has been found in mesangial cells (55), podocytes (18), and vascular smooth muscle cells (25), we are unaware of any previous indications of its presence in renal tubules with the exception of a proximal tubule cell line (44).

The macula densa and thick ascending limb share many characteristics, and NOX2-based NADPH oxidase is responsible for ANG II-stimulated O₂⁻ production in the macula densa (21). Thus we first tested whether NOX2-based NADPH oxidase mediates the effects of ANG II in the thick ascending limb. To test this, we used thick ascending limb suspensions isolated from NOX2 knockout mice. ANG II stimulated O₂⁻ production in the knockout mice in a manner similar to that of the wild-type mice. There was also no significant difference in baseline O₂⁻ production between the knockout mice and the wild-type mice. From these data, we conclude that, contrary to our hypothesis, NOX2 does not mediate ANG II-stimulated O₂⁻ production in thick ascending limbs.

We next tested the hypothesis that NOX4 mediates ANG II-induced O₂⁻ production in the thick ascending limb using adenoviral delivery of NOX4 shRNA directly into the outer medulla of rat kidneys. We found that peak knockdown of NOX4 mRNA expression was ∼77% after 6–7 days. We measured mRNA rather than protein in these experiments so that we could assess the reduction of NOX4 in each suspension that was used for measurements of O₂⁻. We found that treating tubules with NOX4 shRNA significantly reduced ANG II-induced O₂⁻ production compared with ANG II alone. The shRNA had no effect on baseline O₂⁻ production.

As mentioned, the limits of the experimental design prevented us from testing all four treatments at once (basal, basal plus shRNA, ANG II, and ANG II treated with shRNA). Given this and the fact that NOX4 knockdown was not 100%, it is possible that there is some residual ANG II-stimulated O₂⁻ production due to activation of NOX4-based NADPH oxidase. To test for this, we first measured whether ANG II could acutely increase NOX4 expression. We found that ANG II had no effect on NOX4 expression. This result is different from mesangial cells where ANG II has been reported to increase NOX4 expression in as little as 5 min (6). These data and those showing that the measured O₂⁻ production in suspensions treated with ANG II and NOX4 shRNA were similar to rates measured in untreated tubules and those treated with shRNA alone reduce the likelihood that there is a significant residual NOX4-based NADPH oxidase activity after treatment with ANG II.

Finally, given that 1) we used mice to test the role of NOX2 and all other experiments were performed in rats; and 2) we did not specifically test for a role for NOX1, it is conceivable that there is a residual acute effect of ANG II on O₂⁻ production that is due to either 1) differences in the role of NOX2 in rats and mice; and/or 2) activation of NOX1. To test these possibilities, we examined the effect of apocynin on O₂⁻ production in the presence of ANG II after treatment with NOX4 shRNA. Adding apocynin to suspensions treated with NOX4 shRNA did not reduce the ANG II-stimulated O₂⁻ production. These data indicate that neither NOX1 or 2 or any apocynin-inhibitable NOX other than NOX4 is stimulated by ANG II in thick ascending limbs. Thus taken altogether the data indicate that NOX4-based NADPH oxidase is responsible for ANG II-stimulated O₂⁻ production in thick ascending limbs.

Our results showing that NOX4 mediates the effects of ANG II on O₂⁻ production by thick ascending limbs are novel. This is evidenced by the fact that there are few reports of ANG II specifically stimulating NOX4-based NADPH oxidase in the kidney. In renal mesangial cells, ANG II acutely increases O₂⁻ via NOX4-based NADPH oxidase (27, 28). However, ANG II has been shown to stimulate NOX4-based NADPH oxidase in a number of other cell types including vascular smooth muscle cells (36, 54, 54, 75, 88) and endothelial cells (3, 89, 89).

The interpretation of our results relies on the ability of apocynin to inhibit NOX1, 2, and 4-based NADPH oxidase activity. Although the ability of apocynin to inhibit NADPH oxidase is assumed to be due to its ability to prevent assembly of the enzyme's subunits (82), this has not been directly tested for all isoforms of the enzyme. Our data clearly indicate that it inhibits NOX4. Essentially all ANG II-induced O₂⁻ production in this study was inhibited by both NOX4 shRNA and apocynin. These results are consistent with the literature. Several studies have shown that NOX4-based NADPH oxidase activity is apocynin sensitive (4, 9, 12, 22, 62, 69, 83, 93). There is also an abundance of evidence that apocynin is effective for inhibiting NOX1 in a variety of tissues (11, 12, 17, 45, 85, 90, 93, 95). Finally, apocynin has been shown to inhibit NOX2-based NADPH oxidase (1, 1, 32, 99).

Our study shows that apocynin reduces O₂⁻ levels. We assume this is due to inhibition of production rather than simple scavenging. Although apocynin has been shown to scavenge reactive oxygen species, a much higher concentration than used in our studies is required (35). Furthermore, we (38, 80) have shown previously that the concentration we use does not scavenge O₂⁻ produced by hypoxanthine/xanthine oxidase, which is consistent with a previous study (82) that demonstrated that in cell-free systems apocynin is more effective at scavenging H₂O₂ at low concentrations (3–300 μM). Although we have not directly tested the mechanism of action of apocynin inhibition, it is thought that apocynin works by preventing assembly of the catalytic subunit with the p47 subunit. This would be consistent with our studies (34, 38) indicating that both ANG II-stimulated and luminal flow-stimulated O₂⁻ production is absent when p47 is absent. It is also consistent with data in mesangial cells showing that aldosterone increases O₂⁻ production via NOX1-/NOX4-based NADPH oxidase with concomitant translocation of p47 and p67 within the same time frame in an apocynin-sensitive manner (63).

Our results show that NOX4-based NADPH oxidase produces O₂⁻ rather than H₂O₂. This issue is somewhat contentious in the literature. There are studies indicating that NOX4 may produce H₂O₂ directly instead of superoxide (16, 77). However, a number of studies (26, 31, 59, 68) indicate that NOX4 first produces O₂⁻, which is then dismutated to H₂O₂. In this study, O₂⁻ was detected with lucigenin, which was validated as a method to detect O₂⁻ in 1998 (56). It has been used since then to measure O₂⁻ production when trying to discriminate it from H₂O₂ (35, 67). To show that in our hands lucigenin is in fact measuring O₂⁻, we added the O₂⁻ scavenger Tiron to the media at the end of each protocol and subtracted any luminescence not eliminated by Tiron. Tiron has also been used for several years to specifically scavenge O₂⁻ (33, 67). Finally, there are a number of recent studies (29, 29, 92) in a variety of tissues identifying NOX4 as the source of reactive oxygen species in which Tempol was used to reduce O₂⁻. Tempol is a superoxide dismutase mimetic. Thus it would not eliminate H₂O₂ if it, rather than O₂⁻, were the reactive oxygen species produced by NOX4-based NADPH oxidase.

It has also been suggested that NOX4-based NADPH oxidase is constitutively active and cannot be acutely stimulated by agonists (77). However, other than the work done in our laboratory indicating that ANG II rapidly stimulates O₂⁻ production in the thick ascending limb, acute regulation of NOX4 activity has also been demonstrated in other tissues such as the proximal tubule (68), mesangial cells (27, 28), and cardiac fibroblasts (14).

In summary, we have shown that ANG II stimulates NADPH oxidase-dependent O₂⁻ production in thick ascending limbs and this is due to activation of NOX4-based rather than NOX1- or NOX2-based NADPH oxidase.

GRANTS

This work was supported by National Heart, Lung, and Blood Institute Grants 5P01-HL-028982-30, 1P01-HL-090550-03, and 5R01-HL-070985-09 (to J. L. Garvin).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

Author contributions: K.J.M., N.J.H., and J.L.G. conception and design of research; K.J.M. and N.J.H. performed experiments; K.J.M. analyzed data; K.J.M. and J.L.G. interpreted results of experiments; K.J.M. prepared figures; K.J.M. drafted manuscript; K.J.M. and J.L.G. edited and revised manuscript; K.J.M. and J.L.G. approved final version of manuscript.