Angiotensin II-induced superoxide and decreased glutathione in proximal tubules: effect of dietary fructose

Abstract

Angiotensin II exacerbates oxidative stress in part by increasing superoxide ($O_2^{-}$) production by many renal tissues. However, whether it does so in proximal tubules and the source of $O_2^{-}$ in this segment are unknown. Dietary fructose enhances the stimulatory effect of angiotensin II on proximal tubule Na⁺ reabsorption, but whether this is true for oxidative stress is unknown. We hypothesized that angiotensin II causes proximal nephron oxidative stress in part by stimulating NADPH oxidase (NOX)4-dependent $O_2^{-}$ production and decreasing the amount of the antioxidant glutathione, and this is exacerbated by dietary fructose. We measured basal and angiotensin II-stimulated $O_2^{-}$ production with and without inhibitors, NOX1 and NOX4 expression, and total and reduced glutathione (GSH) in proximal tubules from rats drinking either tap water (control) or 20% fructose. Angiotensin II (10 nM) increased $O_2^{-}$ production by 113 ± 42 relative light units·mg protein⁻¹·s⁻¹ in controls and 401 ± 74 relative light units·mg protein⁻¹·s⁻¹ with 20% fructose (n = 11 for each group, P < 0.05 vs. control). Apocynin and the Nox1/4 inhibitor GKT136901 prevented angiotensin II-induced increases in both groups. NOX4 expression was not different between groups. NOX1 expression was undetectable. Angiotensin II decreased GSH by 1.8 ± 0.8 nmol/mg protein in controls and by 4.2 ± 0.9 nmol/mg protein with 20% fructose (n = 18 for each group, P < 0.047 vs. control). We conclude that 1) angiotensin II causes oxidative stress in proximal tubules by increasing $O_2^{-}$ production by NOX4 and decreasing GSH and 2) dietary fructose enhances the ability of angiotensin II to stimulate $O_2^{-}$ and diminish GSH, thereby exacerbating oxidative stress in this segment.

INTRODUCTION

The proximal nephron is critical in maintaining Na⁺ homeostasis because it reabsorbs ≈65% of the filtered load of water and Na⁺ (11, 19, 43, 69). Oxidative stress in this segment has been reported in renal injury (7, 32, 86, 87) and hypertension (48, 83). Many factors and hormones regulate proximal nephron function; chief among these is angiotensin II (ANG II).

ANG II stimulates the production of reactive oxygen species (ROS) such as superoxide ($O_2^{-}$) and H₂O₂ and thus oxidative stress in many tissues (34, 46, 73, 79, 84, 90), including those of the kidney (21, 55, 57, 84). A role for ANG II-induced oxidative stress has been implicated in proximal nephron injury (32). Oxidative stress has also been reported to both decrease and increase proximal nephron Na⁺ reabsorption in models of hypertension (61) and diabetes (14, 63), respectively. ANG II directly enhances $O_2^{-}$ production by collecting ducts (77), thick ascending limbs (33, 49), and cultured cells of proximal tubule origin (20); however, whether this is true for native proximal tubules is unclear.

Many enzymes are capable of producing $O_2^{-}$ and H₂O₂, and thus oxidative stress. These include the family of NADPH oxidases (NOXs). Three family members are expressed along the nephron, including NOX1 (31, 74), NOX2 (9, 44, 74), and NOX4 (44, 74, 75). NOX4 has been reported to be a source of ROS in collecting ducts (47) and thick ascending limbs (36, 71, 72), whereas NOX2 is the primary source of ROS in the macula densa (18). Although NOX generates ROS in proximal tubules (61, 63), the isoform responsible has not been identified. Uncoupled nitric oxide synthase (27) and xanthine oxidase (52) also produce $O_2^{-}$, and these enzymes are expressed in thick ascending limbs (62) and proximal tubules (29). ANG II has been reported to stimulate ROS production by both NOXs (33, 49, 68, 77) and nitric oxide synthase (27); however, the source(s) of ROS production in the proximal nephron and the direct actions of ANG II on them have not been studied in detail.

Cells have several defense mechanisms to prevent or repair the deleterious effects of oxidative stress (16, 50). Chief among these is glutathione (16, 22, 50, 80). The reduced form of glutathione (GSH) is a three-amino acid peptide that undergoes cyclic oxidation and reduction of the sulfhydryl group on the terminal cysteine. During oxidation, two GSH molecules donate their electrons to reduce the target product, and a disulfhydryl linkage is formed between them, generating oxidized glutathione (GSSG) (40, 64). Oxidative stress can decrease both total glutathione and GSH. The effects of ANG II on the glutathione system in the kidney in general, and the proximal nephron specifically, have not been thoroughly investigated.

Since its introduction in the early 1970s, the use of low-cost high-fructose corn syrup has become widespread. This has increased consumption of fructose in the United States from <2 to >40 lb·yr⁻¹·person⁻¹ (41, 82). The average American consumes >11% of their daily caloric intake as fructose (53, 82), and the caloric composition of the diets of >17 million people in the United States is 20% or more fructose (82). Fructose-rich diets are also present in other industrialized countries (37, 60, 91). Dietary fructose increases renal oxidative stress (8, 12), particularly when combined with a high-salt diet (12, 26). Fructose-induced oxidative stress has been linked to both renal injury (2) and hypertension (12, 78). Increased oxidative stress can result from increased production of ROS such as $O_2^{-}$ and/or H₂O₂, decreased degradation of ROS, and/or reductions in other defense mechanisms (12). However, the mechanisms by which dietary fructose increases renal oxidative stress and the renal structures involved are poorly understood.

The proximal nephron is thought to be key in the hypertension (6, 25, 59) and renal injury (9, 42) caused by dietary fructose. Proximal tubules also reabsorb (23) and metabolize fructose (9a). Genetic deletion of fructokinase C from proximal tubules protects against renal injury (1), likely by reducing oxidative stress. Fructose alone increases Na⁺ reabsorption by this segment (6), and dietary fructose enhances the stimulatory effect of ANG II on transport (25, 26). Whether this is also true for $O_2^{-}$ production and oxidative stress as indicated by changes in GSH is unclear.

We hypothesized that ANG II induces proximal nephron oxidative stress in part by stimulating NOX4-dependent $O_2^{-}$ production and decreasing the amount of the antioxidant glutathione and that this is exacerbated by dietary fructose.

METHODS

Reagents.

Unless specified, all drugs and reagents were obtained from Sigma-Aldrich (St. Louis, MO). A Pierce Coomassie (Bradford) Protein Assay Kit was obtained from Thermo Scientific (Rockford, IL). Bicarbonate-buffered physiological saline contained (in mmol/L) 25 NaHCO₃, 114 NaCl, 4 KCl, 0.4 NaH₂PO₄, 2.5 Na₂HPO₄, 1.2 MgSO₄, 5.5 glucose, 6.0 dl-alanine, 2.0 Ca(lactate)₂, and 1.0 Na₃ citrate, pH 7.4 when gassed with 95% O₂-5% CO₂. HEPES-buffered physiological saline contained (in mmol/L) 10 HEPES (pH 7.4), 130 NaCl, 4 KCl, 0.4 NaH₂PO₄, 2.1 Na₂HPO₄, 1.2 MgSO₄, 5.5 glucose, 6.0 dl-alanine, 2.0 Ca(lactate)₂, and 1.0 Na₃ citrate. The osmolality of bicarbonate-buffered and HEPES-buffered physiological saline was adjusted to 300 ± 5 mosmol/L with mannitol. The glutathione kit (catalog no. 703002) was obtained from Cayman Chemical (Ann Arbor, MI).

Animals.

Male Sprague-Dawley rats (Charles River Breeding Laboratories, Wilmington, MA) weighing between 101 and 125 g were randomly assigned to one of two experimental groups. One group received a 20% fructose solution as their source of fluid (Fruct group), whereas the other group received tap water (control group). Twenty percent fructose was prepared fresh every other day. Both groups were on a purified diet (no. 5876, TestDiet, St. Louis, MO) containing 18.6% protein, 4.3% fiber, 59.3% carbohydrates, and 10% fat, providing 4.02 kcal/g and ~100 meq/kg Na⁺. The metabolic characteristics of this model have been previously published (24).

Animals in both experimental groups were housed in pairs under normal rat housing conditions with a 12:12-h light-dark cycle and ad libitum provision of food and fluids. After 6–8 days of dietary treatment, the animals underwent terminal surgery. Rats were anesthetized with ketamine (100 mg/kg body wt ip) and xylazine (20 mg/kg body wt ip) and given 100 units heparin (ip).

Male Dahl salt-sensitive rats in which NOX4 was knocked out were treated in a manner similar to control rats. These rats were only used to validate the NOX4 antibody.

This study was approved by the Case Western Reserve University Institutional Animal Care and Use Committee. All experiments were conducted in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

Proximal tubule suspension.

Proximal tubule suspensions were generated using methods similar to those we have previously used (24). Briefly, rats were anesthetized and an abdominal U-shaped incision was made. The kidneys were retroperfused from the abdominal aorta with 80 mL bicarbonate-buffered physiological saline at 37°C. Physiological saline contained 1 mg/mL collagenase type I and 2 U/mL heparin and was infused at 8 mL/min for 10 min. Digested kidneys were excised before flow was stopped and immediately cooled by immersion in physiological saline at 4°C. They were then transferred to a cold Lucite plate, and their cortexes were gently scraped with a razor blade to collect proximal tubules. This tissue was minced and transferred to a 5-mL conical tube. Tissue was disrupted by passing it through a pipette tip and stirring on ice for 5 min. The resulting suspension was filtered through a 390-µm mesh, and tubules were recovered by centrifugation at 100 g for 2 min at 4°C. The tubules were then rinsed, filtered through a 250-µm mesh, and recovered by centrifugation. The final pellet was resuspended in 3–5 mL physiological saline.

$O_2^{-}$ production.

$O_2^{-}$ was measured using methods similar to those we have previously used (27). Fruct and control suspensions were diluted to 200 µg protein/mL, and 200-µL aliquots were added to two plastic tubes containing bis-N-methylacridinium nitrate (Lucigenin) at a final concentration of 5 µmol/L and kept on ice. One tube was designated “basal.” The other tube was designated “ANG II stimulated.” $O_2^{-}$ production by each sample was measured sequentially. The order of the tubes was randomized to account for effects of time. The process was as follows. Eight hundred microliters of warm, oxygenated physiological saline alone or containing 10⁻⁸ mol/l ANG II was added to the tube and incubated at 37°C for 6 min. The tube was transferred to a luminometer (FB12/Sirius, Zylux Oak Ridge, TN). Two luminometers were used in this study which required the application of a small correction factor to normalize the data. Total luminescence in relative light units (RLU)/s was recorded at 4.8-s intervals for 7 min. The $O_2^{-}$ scavenger 4,5-dihydroxy-1,3-benzenedisulfonic acid (Tiron) was then added to a final concentration of 10 mmol/L, and luminescence was measured for 5 additional min. $O_2^{-}$ production was calculated as the steady-state luminescence without Tiron minus the luminescence of the last 2 min after Tiron addition. Protein content of the sample was measured using the Pierce Coomassie (Bradford) Protein Assay, and $O_2^{-}$ production was normalized for protein content. Differences in luminescence between the samples with and without ANG II were considered ANG II-stimulated $O_2^{-}$ production.

For the experiments involving apocynin and GKT136901, nonselective and NOX1/4 inhibitors, respectively, the chemical was added during the initial dilution of the suspensions.

Western blot analysis.

Proximal tubules from Fruct and control suspensions were dissolved in CelLytic M cell lysis reagent. Duplicates of 5 and 10 µg protein were loaded onto a 10% Criterion TGX Stain-Free polyacrylamide gels (Bio-Rad Laboratories, Hercules, CA). Samples from Fruct and control groups were run in pairs on the same gel. Electrophoresis was performed to separate proteins, which were transferred onto polyvinylidene difluoride membranes via the iBlot 2 Dry Blotting System (ThermoFisher Scientific, Waltham, MA). Membranes were exposed in an Azure c600 system (Azure Biosystems, Dublin, CA) at 507 nm for 30 s, and images of total protein fluorescence were captured to use as loading controls. Membranes were incubated in blocking buffer composed of 5% nonfat milk in PBS with Tween 20 (PBS-T; 137 mM NaCl, 2.7 mM KCl, 10 mM NaH₂PO₄, and 0.1% Tween 20) for 60 min. They were then incubated with a primary rabbit monoclonal anti-NOX4 antibody (ab133303, Abcam) at 1:4,000 dilution in 5% nonfat milk in PBS-T for 120 min at room temperature. Membranes were washed with PBS-T for 15 min and incubated at 1:4,000 dilution in 5% BSA with horseradish peroxidase (HRP)-conjugated anti-rabbit IgG (Amersham Pharmacia Biotech) for 60 min. Membranes were washed with PBS-T for 15 min and incubated with a luminol-based chemiluminescent HRP substrate (Pierce Biotechnology, Rockford, IL) and exposed in the Azure c600 system for 100 s. An image was captured, and densitometry was performed using ImageJ. The anti-NOX4 antibody was validated using renal tissue from NOX4 knockout mice.

The method to test for NOX1 expression was similar except that the primary antibody was directed against NOX1 and several dilutions were tested (ab131088, Abcam).

Glutathione.

Two aliquots of 1,000 µg protein from Fruct and control suspensions were diluted to 5 mL with HEPES-buffered physiological saline in separate incubation chambers at 37°C with stirring. ANG II was added to one of the chambers at a final concentration of 10⁻⁸ mol/L and incubated for 7 or 15 min. The 7- and 15-min incubations gave similar results, so the data were pooled. After incubation, the suspension was spun at 100 g for 2 min, and the supernatant was removed. Tubules were collected and resuspended in 2 mL MES buffer [0.2 M 2-(N-morpholino)ethanesulphonic acid, 0.05 M phosphate, and 1 mM EDTA (pH 6.0)] and deproteinated with the same volume of metaphosphoric acid (1 g/10 mL water). The mixed solution was incubated for 5 min at room temperature and then centrifuged at 2,000 g for 2 min. The supernatant was collected and frozen at −80°C. Standards were treated in the same manner. Before the GSH assay, samples and standards were thawed on ice. Total glutathione and GSSG were measured using a Glutathione Assay Kit (catalog no. 703002, Cayman Chemical). GSH was calculated by subtracting the amount of GSSG from the amount of total glutathione.

Statistical analysis.

Results are expressed as means ± SE. P values were calculated using an unpaired t test or two-way ANOVA as appropriate. When two-way ANOVA was used, post hoc testing was performed using paired or unpaired t tests as appropriate. Corrections for multiple testing were made using Hochberg’s method (35). Corrected P values are reported in the text.

RESULTS

ANG II stimulates ROS production in many tissues (34, 46, 73, 79, 84, 90), but whether it does so in the proximal nephron, and the effects of dietary fructose, are unknown. We first tested the effect of ANG II on $O_2^{-}$ production in control tubules. ANG II (10⁻⁸ mol/L) stimulated $O_2^{-}$ production from 772 ± 57 to 885 ± 49 RLU·mg protein⁻¹·s⁻¹ (n = 11, P < 0.022; Fig. 1A). Dietary fructose increases renal oxidative stress and augments the ability of ANG II to stimulate proximal nephron Na⁺ reabsorption. Therefore, we next studied the effect of ANG II on Fruct tubules. We found that ANG II augmented $O_2^{-}$ production from 682 ± 48 to 1,083 ± 86 RLU·mg protein⁻¹·s⁻¹ (n = 11, P < 0.0009; Fig. 1B). The effect of dietary fructose alone was not significant, whereas both the effects of ANG II alone (P < 0.002) and the interaction were significant (P < 0.025).

Fig. 1.

Effect of 10⁻⁸ mol/L angiotensin II (ANG II) on $O_2^{-}$ production by proximal tubule suspensions. A: control tubules. Error bars are covered by symbols. B: tubules from rats that consumed 20% fructose for 6–8 days. Error bars are covered by symbols. n = 11 for each group. The increase in both groups combined was significant by ANOVA (P < 0.0002). Post hoc testing using a paired a t test showed that the effect of ANG II was significant in each group, as indicated in the figure. C: ANG II-induced $O_2^{-}$ production by proximal tubule suspensions from control rats and rats that consumed 20% fructose for 6–8 days. ANOVA indicated a significant interaction between ANG II and fructose (P < 0.025). Post hoc testing using a paired t test indicated that the effect of ANG II was greater in rats on the fructose diet, as indicated in the figure. RLU, relative light units.

Since ANOVA showed that the effect of ANG II was significant when the groups were combined and there was a significant interaction between ANG II and fructose, we next performed post hoc testing to study whether the stimulation caused by ANG II was significant in each group separately and whether there was a difference in the effect of ANG II between groups. In control tubules, 10⁻⁸ mol/L ANG II enhanced $O_2^{-}$ production by 112 ± 42 RLU·mg protein⁻¹·s⁻¹ (n = 11, P < 0.022; Fig. 1C), while in Fruct tubules, ANG II increased $O_2^{-}$ production by 401 ± 74 RLU·mg protein⁻¹·s⁻¹ (n = 11, P < 0.0009; Fig. 1C). These data also show that ANG II stimulated $O_2^{-}$ production to a greater extent in Fruct tubules than in control tubules (P < 0.005; Fig. 1C). This difference was not due to dietary fructose affecting basal $O_2^{-}$ synthesis. Basal $O_2^{-}$ production by proximal tubules was not different between groups.

NOXs are a major source of $O_2^{-}$ in the kidney (84, 92). Therefore, we next tested whether this family of enzymes was the source of the ANG II-stimulated $O_2^{-}$. In the presence of apocynin, $O_2^{-}$ production was 204 ± 55 RLU·mg protein⁻¹·s⁻¹ in the absence of ANG II and 233 ± 85 RLU·mg protein⁻¹·s⁻¹ in its presence in control tubules (n = 5; Fig. 2A). Similarly, in Fruct tubules, $O_2^{-}$ production was 189 ± 27 RLU·mg protein⁻¹·s⁻¹ in the absence of ANG II and 216 ± 36 RLU·mg protein⁻¹·s⁻¹ in its presence (n = 5; Fig. 2A). These data show that in the presence of apocynin, dietary fructose alone did not alter basal $O_2^{-}$ production. Additionally, apocynin blunted the ability of ANG II to stimulate $O_2^{-}$ production in both groups and eliminated the difference between them (Fig. 2B).

Fig. 2.

Effect of 10⁻⁸ mol/L angiotensin II (ANG II) on $O_2^{-}$ production by proximal tubule suspensions from control rats and rats that consumed 20% fructose (FRUC) for 6–8 days in the presence of apocynin, a nonselective NADPH oxidase inhibitor. A: basal and ANG II-stimulated $O_2^{-}$ production. B: ANG II-induced $O_2^{-}$ production. RLU, relative light units. n = 5 for each group.

NOX4 is thought to be a primary contributor to oxidative stress in the kidney. Recently, a NOX1/4 inhibitor, GKT136901, has been reported (74, 81). Consequently, we tested the effect of GKT136901 on $O_2^{-}$ production (Fig. 3). In the presence of GKT136901, $O_2^{-}$ production was 447 ± 98 RLU·mg protein⁻¹·s⁻¹ in the absence of ANG II and 529 ± 58 RLU·mg protein⁻¹·s⁻¹ in its presence in control tubules (n = 5; Fig. 3A). Similarly, $O_2^{-}$ production was 433 ± 60 RLU·mg protein⁻¹·s⁻¹ in the absence of ANG II and 469 ± 54 RLU·mg protein⁻¹·s⁻¹ in its presence in Fruct tubules (n = 5; Fig. 3A). These data show that in the presence of GKT136901, dietary fructose alone did not alter basal $O_2^{-}$ production. Additionally, GKT136901 blunted the ability of ANG II to stimulate $O_2^{-}$ production in both groups and eliminated the difference between them (Fig. 3B).

Fig. 3.

Effect of 10⁻⁸ mol/L angiotensin II (ANG II) on $O_2^{-}$ production by proximal tubule suspensions from control rats and rats that consumed 20% fructose (FRUC) for 6–8 days in the presence of GKT136901, a NADPH oxidase 1/4 inhibitor. A: basal and ANG II-stimulated $O_2^{-}$ production. B: ANG II-induced $O_2^{-}$ production. RLU, relative light units. n = 5 for each group.

Since GKT136901 inhibits both NOX1 and NOX4, we measured their expression and the effect of dietary fructose. We found that proximal tubules expressed NOX4 but we could not detect NOX1. Figure 4A shows a representative Western blot for NOX4. A major band corresponding to the molecular weight of NOX4 was detected at ≈65 kDa. Figure 4B shows mean data from six such blots normalized for total protein loaded as measured by fluorescence of each lane in the gel at 507 nm. Expression of NOX4 by proximal tubules was 229 ± 41 arbitrary units in control tubules, whereas it was 167 ± 32 arbitrary units in those from Fruct tubules. Figure 4C shows the major band observed at 65 kDa was NOX4 because it was not detectable in renal homogenates from NOX4 knockout rats.

Fig. 4.

NADPH oxidase expression measured by Western blot analysis. A: representative membrane showing the entire gel. A major band was observed at ~65 kDa. Lanes 1 and 2, 10 µg of proximal tubule protein from controls; lanes 3 and 4, 10 µg of proximal tubule protein from rats that consumed 20% fructose for 6–8 days. B: quantification of 6 blots normalized by the total protein in each lane of the gel. C: validation of antibody. Lanes 1 and 2, 10 µg of proximal tubule protein from control rats; lane 3, 10 µg of proximal tubule protein from a NADPH oxidase 4 knockout rat on a Dahl salt-sensitive background.

Reactions involving glutathione are major intracellular antioxidant defense mechanisms that neutralize ROS including $O_2^{-}$, thereby preventing tissue oxidative damage (16, 22, 80). Reductions in total glutathione and GSH thus reflect oxidative stress. Two-way ANOVA indicated that fructose alone did not significantly affect total glutathione but that 10⁻⁸ mol/L ANG II did (P < 0.0001). Thus, we next analyzed whether ANG II decreased total glutathione separately in tubules from control and Fruct groups. ANG II diminished total glutathione by 3.7 ± 1.1 ng/mg protein in control suspensions (n = 18, P < 0.001) and by 6.0 ± 0.8 ng/mg protein in Fruct suspensions (n = 18, P < 0.002). However, the difference between groups did not reach statistical significance (P < 0.10; Fig. 5).

Fig. 5.

Angiotensin II (ANG II)-induced changes in total glutathione in proximal tubules from control rats and rats that consumed 20% fructose for 6–8 days. n = 18 in each group. ANOVA indicated a significant effect of 10⁻⁸ mol/L ANG II on total glutathione (P < 0.0001). Post hoc testing using paired t tests showed the effect was significant in each group (a: P < 0.001; b: P < 0.002). The difference between groups was not significant, as indicated in the figure.

GSH is a more sensitive indicator of oxidative stress than total glutathione, so we next measured GSH. Two-way ANOVA indicated that 10⁻⁸ mol/L ANG II alone significantly altered GSH (P < 0.0001) and there was a significant interaction between ANG II and fructose (P < 0.030). Fructose alone did not significantly affect GSH. Consequently, we analyzed the effect of ANG II on GSH by group and whether fructose augmented its actions. ANG II (10⁻⁸ mol/L) decreased GSH by 1.8 ± 0.8 ng/mg protein in control tubules (n = 18, P < 0.032) while diminishing it by 4.2 ± 0.9 ng/mg protein in Fruct tubules (n = 18, P < 0.002). The decrease in GSH caused by ANG II was greater in the Fruct group than in the control group (P < 0.047; Fig. 6).

Fig. 6.

Angiotensin II (ANG II)-induced changes in GSH in proximal tubules from control rats and rats that consumed 20% fructose for 6–8 days. n = 18 in each group. ANOVA indicated a significant effect of 10⁻⁸ mol/L ANG II on GSH (P < 0.001). Post hoc testing using paired t tests showed the effect was significant in each group (a: P < 0.032; b: P < 0.002). The difference between groups was significant, as indicated in the figure.

DISCUSSION

The first part of our hypothesis was that ANG II causes proximal nephron oxidative stress in part by stimulating NOX4-dependent $O_2^{-}$ production and decreasing the amount of the antioxidant glutathione. We found that 1) acute ANG II treatment stimulated $O_2^{-}$ production in control tubules; 2) this could be blunted by apocynin and GKT136901, nonselective and selective NOX1/4 inhibitors, respectively; 3) proximal tubules express NOX4 but not NOX1; and 4) acute ANG II treatment decreased both total glutathione and GSH.

Our results showing that ANG II stimulates $O_2^{-}$ production by native proximal tubules are novel and have not been previously reported. They are consistent with data showing that ANG II augments ROS production by cultured cells of proximal tubule origin (20). They are also similar to data showing that ANG II enhances $O_2^{-}$ and/or H₂O₂ production by outer medullary tissue (54, 56), thick ascending limbs (49, 85), macula densa (89), collecting ducts (77), and mesangial cells (4, 21). We show here that the source of ANG II-induced $O_2^{-}$ was most likely NOX4 because a selective NOX1/4 inhibitor blocked the actions of ANG II and NOX4 but not NOX1 protein expression was detected in control suspensions. Again, these data are similar to those reported for thick ascending limbs (49) and mesangial cells (4, 21). However, in the macula densa, ANG II enhances ROS production by both NOX2 and NOX4 (89), although NOX2 is the primary source (18) and ANG II-induced ROS production by NOX1 has been reported in vascular smooth muscle cells (5).

The glutathione system is a cellular defense mechanism against oxidative stress. Consequently, we studied the effects of ANG II on total glutathione and GSH. We found that acute ANG II treatment decreased GSH levels in proximal tubules consistent with it causing oxidative stress in this segment. These are the first data showing such an effect although similar results have been reported for other tissues. ANG II has been previously reported to diminish GSH in endothelial progenitor cells (30). In addition, a previous study (3) has reported that ANG II reduced glutathione peroxidase activity in the rat hippocampus, which can also indicate the levels of reduced GSH.

The family of superoxide dismutases catalyze the degradation of $O_2^{-}$ to H₂O₂. As such, they also represent a defense mechanism against oxidative stress produced specifically by $O_2^{-}$. However, while they reduce $O_2^{-}$, they simultaneously increase H₂O₂, another ROS capable of causing oxidative stress. Consequently, we did not assess superoxide dismutase activity, favoring GSH instead.

Dietary fructose increases renal oxidative stress, particularly when combined with a high-salt diet (12); however, the mechanisms by which dietary fructose increases renal oxidative stress and the nephron segments involved are unclear. The second part of our hypothesis was that dietary fructose enhanced ANG II-induced ROS production by proximal tubules and diminished glutathione. We found that dietary fructose significantly enhanced the ability of ANG II to 1) stimulate $O_2^{-}$ production by NOX and 2) decrease GSH, thereby exacerbating oxidative stress in the proximal nephron. However, dietary fructose did not change basal, unstimulated $O_2^{-}$ production by rat proximal tubules or NOX4 expression, total glutathione, or GSH.

Our data showing that 20% fructose in the drinking water did not alter basal $O_2^{-}$ production by proximal tubules is novel, not having been reported before. This finding would seem to be at odds with reports that dietary fructose increases oxidative stress in the kidney (12), heart (10), skeletal muscle cells (38), and peripheral blood mononuclear cells (65). The explanation for the disparate results is likely straightforward. In the reports referenced above, preparations were used in which the effects of ANG II were likely still at least partially intact, reflecting the in vivo situation more closely. In contrast, generation of proximal tubule suspensions takes ~45 min. In this time, one would expect the effects of plasma ANG II to wane such that they were no longer observable. We chose to use this preparation precisely for this reason. This allowed us to study both potential basal effects of dietary fructose and whether the response to ANG II was altered. If we had used a preparation in which endogenous ANG II and its actions were intact, they would have confounded our ability to investigate whether dietary fructose altered these effects.

We show here that dietary fructose augmented the ability of ANG II to stimulate $O_2^{-}$ production by proximal tubules. This has not been previously reported. Our results are consistent with those previously reported for other tissues. Ten percent fructose in the drinking water causes increased oxidative stress and nitric oxide synthase activity and expression in the kidney, which are reversed by dietary supplementation of (−)-epicatechin (66). In vitro, fructose added to the culture media of skeletal muscle cells provoked mitochondrial ROS formation, mitochondrial dysfunction, and progressive apoptosis (38, 39). In vivo, fructose ingestion before aerobic exercise in male athletes resulted in a significant rise in oxidative stress, not seen with glucose ingestion (15). Rats drinking 10% fructose present systemic oxidative stress (65). About 20 wk of 20% high fructose and 8% high sodium has been reported to elevate oxidative stress due to a decrease in superoxide dismutase activity and induce salt-sensitive hypertension. Both were blunted by tempol (12). The increase in systolic blood pressure caused by concomitant administration of salt and fructose (6, 28) is completely reversed when rats are chronically given tempol in the drinking water (88). In addition, dietary fructose causes tubular hyperplasia and proliferation in proximal tubules and renal injury (58, 70).

We next tested whether NOXs were the source of fructose-induced ANG II-stimulated $O_2^{-}$ production. Our data show that both apocynin and GKT136901 prevented the increase. Together, these data indicate that the source of $O_2^{-}$ was NOX1 or NOX4. Since NOX1 was undetectable in our samples (data not shown) and NOX4 expression levels were not changed by dietary fructose, it is most likely that the increased activation of NOX4 was the source of the increased production of $O_2^{-}$. These are the first data showing that dietary fructose augments the ability of ANG II to activate NOX4.

On the surface, our data appear to be at odds with those reporting that xanthine oxidase is likely the primary source of fructose-induced ROS in the proximal nephron (8). However, those authors used a diet that contained much more fructose than the diet used here. Additionally, it may be that ROS produced by xanthine oxidase stimulates NOX4 activity or the other way around. ROS-induced ROS production has been reported in the vasculature (17). NOX4-produced ROS has been reported to induce an increase in NOX2 mRNA (17), and, although it has not been reported, the reversed induction may also occur. Finally, NOX4 may not be the only source of ROS. NOX4-mediated oxidative stress has been reported to induce endothelial nitric oxide synthase uncoupling in the kidney, which leads to decreased nitric oxide production and increased $O_2^{-}$ production (13, 27); a similar mechanism has also been reported in the vasculature (45, 51). Our data do not eliminate such a possibility.

Finally, we studied the effect of dietary fructose and acute ANG II treatment on total glutathione and GSH. Our data show that acute ANG II treatment caused a greater reduction in GSH in Fruct tubules than in control tubules. These are the first data showing that dietary fructose enhances the ability of ANG II to diminish GSH in proximal tubules. These results are consistent with dietary fructose enhancing oxidative stress caused by ANG II. Although ANG II decreased total glutathione, this effect was not augmented by dietary fructose. Although our data showing that dietary fructose alone did not have a statistically significant effect on GSH seem to be at odds with studies reporting that dietary fructose causes systemic depletion of reduced GSH (15, 67), the likely explanation is the same as that described above, i.e., waning of the effects of ANG II in an in vitro preparation.

The physiological significance of increased oxidative stress caused by dietary fructose in the proximal nephron is likely to be profound. Previously, we reported that dietary fructose enhances the ability of ANG II to stimulate Na⁺ reabsorption in proximal tubules via increases in both Na⁺/H⁺ exchange activity (24, 26) and expression (24). We also have shown that dietary fructose stimulates fructose reabsorption by this segment (23). These physiological effects are likely to be at least partly dependent on ANG II-induced $O_2^{-}$ production.

In summary, we have shown that acute ANG II treatment stimulates $O_2^{-}$ production by NOX4 and reduces both total glutathione and GSH in the proximal nephron and that dietary fructose enhances the effects of ANG II on these parameters augmenting oxidative stress.

GRANTS

This work was supported in part by National Heart, Lung, and Blood Institute Grant HL-128053 (to J. L. Garvin).

DISCLOSURES

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

AUTHOR CONTRIBUTIONS

A.G.-V. and J.L.G. conceived and designed research; N.Y., A.G.-V., and J.L.G. performed experiments; N.Y. and J.L.G. analyzed data; N.Y., A.G.-V., and J.L.G. interpreted results of experiments; N.Y. and J.L.G. prepared figures; N.Y. drafted manuscript; A.G.-V. and J.L.G. edited and revised manuscript; N.Y., A.G.-V., and J.L.G. approved final version of manuscript.