Restoration of afferent arteriolar autoregulatory behavior in ischemia-reperfusion injury in rat kidneys

Abstract

Renal autoregulation is critical in maintaining stable renal blood flow (RBF) and glomerular filtration rate (GFR). Renal ischemia-reperfusion (IR)-induced kidney injury is characterized by reduced RBF and GFR. The mechanisms contributing to renal microvascular dysfunction in IR have not been fully determined. We hypothesized that increased reactive oxygen species (ROS) contributed to impaired renal autoregulatory capability in IR rats. Afferent arteriolar autoregulatory behavior was assessed using the blood-perfused juxtamedullary nephron preparation. IR was induced by 60 min of bilateral renal artery occlusion followed by 24 h of reperfusion. Afferent arterioles from sham rats exhibited normal autoregulatory behavior. Stepwise increases in perfusion pressure caused pressure-dependent vasoconstriction to 65 ± 3% of baseline diameter (13.2 ± 0.4 μm) at 170 mmHg. In contrast, pressure-mediated vasoconstriction was markedly attenuated in IR rats. Baseline diameter averaged 11.7 ± 0.5 µm and remained between 90% and 101% of baseline over 65–170 mmHg, indicating impaired autoregulatory function. Acute antioxidant administration (tempol or apocynin) to IR kidneys for 20 min increased baseline diameter and improved autoregulatory capability, such that the pressure-diameter profiles were indistinguishable from those of sham kidneys. Furthermore, the addition of polyethylene glycol superoxide dismutase or polyethylene glycol-catalase to the perfusate blood also restored afferent arteriolar autoregulatory responsiveness in IR rats, indicating the involvement of superoxide and/or hydrogen peroxide. IR elevated mRNA expression of NADPH oxidase subunits and monocyte chemoattractant protein-1 in renal tissue homogenates, and this was prevented by tempol pretreatment. These results suggest that ROS accumulation, likely involving superoxide and/or hydrogen peroxide, impairs renal autoregulation in IR rats in a reversible fashion.

NEW & NOTEWORTHY Renal ischemia-reperfusion (IR) leads to renal microvascular dysfunction manifested by impaired afferent arteriolar autoregulatory efficiency. Acute administration of scavengers of reactive oxygen species, polyethylene glycol-superoxide dismutase, or polyethylene glycol-catalase following renal IR restored afferent arteriolar autoregulatory capability in IR rats, indicating that renal IR led to reversible impairment of afferent arteriolar autoregulatory capability. Intervention with antioxidant treatment following IR may improve outcomes in patients by preserving renovascular autoregulatory function and potentially preventing the progression to chronic kidney disease after acute kidney injury.

INTRODUCTION

Acute kidney injury (AKI) is a major clinical complication occurring in ∼14% of hospitalized patients, with even modest increases in the serum creatinine concentration increasing morbidity, mortality, and costs (1, 2). Recent clinical studies indicated that >25% of AKI survivors developed chronic kidney disease (CKD) and end-stage kidney failure (3, 4). Ischemia-reperfusion (IR)-induced AKI (IR-AKI) is one of the major causes of AKI that occurs after cardiac arrest, shock, and renal transplantation (4, 5). IR-AKI is characterized by increased renal vascular resistance (RVR) and decreased renal blood flow (RBF) and glomerular filtration rate (GFR), as well as tubular injury and necrosis (6–8). Despite intensive research, specific therapies to treat or prevent renal IR injury do not exist.

Efficient renal function depends on stable RBF and GFR. Kidneys achieve these stable hemodynamic conditions through precise autoregulatory adjustments of afferent arteriolar resistance in response to fluctuations in arterial pressure (9, 10). This protective action of afferent arterioles is achieved by the combination of two distinct mechanisms: the intrinsic myogenic response and tubuloglomerular feedback (TGF). The myogenic response is inherent in preglomerular microvessels (11). TGF is a negative feedback system that coordinates NaCl delivery past the macula densa with fine resistance adjustments on the terminal aspect of afferent arterioles (12). Therefore, afferent arterioles are the principal resistance vessels determining renal autoregulatory efficiency. Reduction or loss of autoregulatory efficiency is a crucial contributor to development of pathological glomerular injury and the progression of CKD (9, 13–15). Although early studies have indicated that renal microcirculatory dysfunction appears to be more important after ischemia followed by reperfusion (16, 17), the renal autoregulatory capability after IR is still uncertain (18–22). Autoregulatory impairment after IR may be central to the reduction of GFR and to hypoxia-induced kidney injury, thus making it important to determine the mechanisms that underlie autoregulatory dysfunction post-IR.

The mechanisms contributing to renal microvascular dysfunction in IR have not been fully determined, but renal IR triggers a pathological increase in reactive oxygen species (ROS) accumulation and inflammation in kidneys (23). ROS play vital roles in physiological signaling including cellular growth and proliferation and vascular reactivity (23). IR produces significant amounts of ROS that further recruit inflammatory cells to the kidneys (24–26). Renal microvessels and tubular epithelial cells strongly express NADPH oxidase (NOX), which is the major source of ROS production (9). Li et al. (27) showed that ROS participate in regulating myogenic contractions via superoxide but not hydrogen peroxide (H₂O₂) (23). In addition, superoxide reportedly enhanced TGF by scavenging nitric oxide at the macula densa (28). Alternatively, excessive ROS production contributes to several harmful effects under pathological conditions such as hypertension, AKI, CKD, or diseases with elevated inflammation (9, 14, 23). Animal studies have revealed that increased ROS accumulation and inflammatory cytokine release are linked to renal autoregulatory impairment in hypertensive or high-salt-fed animals (9, 14, 29, 30). Thus, we hypothesized that acute ROS scavenging or reducing ROS accumulation may reduce the pathophysiological levels of ROS and restore renal microvascular autoregulatory capability, culminating in reduced kidney injury in an IR rat model. In the current study, we used the in vitro blood-perfused juxtamedullary nephron (JMN) preparation to directly assess autoregulatory responses of afferent arterioles from rats subjected to 60 min of bilateral renal ischemia followed by 24 h of reperfusion.

METHODS

Animals

A total of 169 male Sprague–Dawley rats (350–375 g, Charles River Laboratories, Raleigh, NC) were used. All rats had free access to water and standard chow (LabDiet, PMI Nutrition, Brentwood, MO) and were maintained according to the National Institutes of Health Guide for the Care and Use of Laboratory Animals. All procedures were approved by the Institutional Animal Care and Use Committee of the University of Alabama at Birmingham.

Rat Model of IR Injury

Rats were anesthetized with pentobarbital (50 mg/kg body wt ip). All rats studied also received single doses of buprenorphine (0.05 mg/kg body wt) and carprofen (5 mg/kg body wt) subcutaneously just before surgery to reduce pain. Surgical sites were shaved and sterilized with 7.5% povidone-iodine and 70% alcohol. Under anesthesia, both renal arteries were occluded for 60 min with atraumatic microvascular clamps, and ischemia was verified on the kidney color change to pale, as described previously (20). This model was chosen based on the previous report in isolated-perfused rat kidneys that IR led to impaired renal autoregulation (20). After 60 min of ischemia, clamps were removed and reperfusion was verified. The muscle layer and skin incisions were closed with 4-0 Vicryl sutures (Med-Vet, Mettawa, IL) and with AUTOCLIP wound clips (Becton Dickinson, Franklin Lakes, NJ), respectively. Control animals received sham surgery but renal arterial occlusion was omitted. Afferent arteriolar function was assessed 24 h after reperfusion. To determine the effectiveness of tempol treatment on preventing ischemic kidney injury and inflammation, we prepared another set of rats pretreated with tempol in drinking water as described in Metabolic Cage Experiments. Urine (24 h) and kidneys were collected 48 h after reperfusion to measure kidney injury markers and mRNA expression.

In Vitro Blood-Perfused JMN Preparation

The JMN preparation was used to directly assess afferent arteriolar responses to experimental manipulations by measuring arteriolar diameter (31, 32). Two identical rats (a kidney donor and a blood donor) were used for each experiment. Under anesthesia, the right kidney of the kidney donor was cannulated and perfused with Tyrode buffer (Sigma-Aldrich, St. Louis, MO) containing 5.2% BSA (Calbiochem, Billerica, MA). Blood collected from both rats via a carotid artery cannula was mixed and processed for kidney perfusion, as previously described (31, 33). The mixed blood was centrifuged at 2,700 g for 13 min, and the plasma was filtered through a 0.2-µm filter (Corning, NY). The buffy coat was discarded from the packed cells, and packed erythrocytes were washed with 0.9% saline and centrifuged at 320 g for 14 min and at 2,700 g for 13 min. The plasma and washed erythrocytes were mixed and filtered through a 5-µm nylon mesh to yield a hematocrit of ∼33%. After completion of dissection, the perfusate was switched to reconstituted blood gassed with 95% O₂-5% CO₂ from a pressurized reservoir while perfusion pressure was held at 100 mmHg. The inner cortical surface of the kidney was bathed with 1% BSA-Tyrode buffer (superfusate) at 37°C. Test drugs applied to the kidney surface were dissolved in the superfusate and delivered to the kidney using an eight-channel switching manifold, whereas polyethylene glycol-conjugated superoxide dismutase (PEG-SOD) and polyethylene glycol-conjugated catalase (PEG-catalase) were delivered directly into the reconstituted perfusate blood. Only one afferent arteriole was used in each JMN preparation. Vessel images were displayed on a video monitor via a video camera (NC-70 Newvicon Video Camera, DAGE-MTI, Michigan City, IN) and recorded on DVD for later analysis. The inner diameter of the afferent arteriole was measured at the same position every 12 s and analyzed by averaging the diameters obtained over the last 2 min of a 5-min treatment period.

Experimental Protocols

Experiment 1: effect of IR on the autoregulatory behavior of afferent arterioles.

After completion of the JMN preparation, the kidney was switched from the 5.2% BSA perfusate to reconstituted blood and the inner surface of the kidney was superfused with 1% BSA superfusate. Kidney perfusion pressure was set at 100 mmHg for at least 20 min to equilibrate and achieve a stable arteriole diameter. After a 5-min baseline diameter recording, afferent arteriolar autoregulatory responses were assessed as perfusion pressure was reduced to 65 mmHg and then increased to 170 mmHg in 15-mmHg increments at 5-min intervals. The sham-operated (sham) group (n = 7) served as a control for the IR group (n = 8).

Experiment 2: effect of acute tempol exposure on the autoregulatory behavior of afferent arterioles following IR.

To assess the role of ROS on autoregulatory behavior, tempol, a membrane-permeable and metal-independent SOD mimetic, was administered in the 1% BSA superfusate onto the inner cortical surface of the kidney. The protocol was similar to the autoregulatory protocol as described in experiment 1, except the inner surface of the IR kidney received 1 mM tempol in the superfusate for 20 min after a 5-min baseline recording (IR + tempol, n = 6). The autoregulatory protocol was executed with continuous tempol superfusion. The average diameter during the last 2 min of the tempol period served as a new baseline diameter for normalization of the autoregulatory data with tempol. The 20-min incubation time with the tempol administration protocol used in this study was based on our previous observation showing renal protection of tempol against autoregulatory dysfunction (34). Since the previous study showed that acute administration of tempol did not alter the pressure-dependent afferent arteriolar response profile in normal rats (34), sham control experiments with tempol alone were omitted.

Experiment 3: effect of acute apocynin exposure on the autoregulatory behavior of afferent arterioles following IR.

To assess the source of ROS on afferent arteriolar autoregulatory behavior in IR rats, we applied the NOX inhibitor apocynin in the 1% BSA superfusate. The autoregulatory protocol was identical to the tempol experiment, except apocynin (0.1 mM) was administered in the superfusate instead of tempol (IR + apocynin, n = 6). Similar to tempol treatment, acute administration of apocynin also did not alter the pressure-dependent afferent arteriolar response profile in normal rats, as demonstrated by the previous study (30); therefore, sham control experiments using apocynin were omitted.

Experiment 4: effect of acute PEG-SOD exposure on the autoregulatory behavior of afferent arterioles following IR.

A membrane-permeable superoxide scavenger, PEG-SOD, was also used to assess the contribution of superoxide in autoregulatory impairment. The autoregulatory protocol was identical to experiment 1, except that PEG-SOD (100 U/mL) was delivered in the reconstituted perfusate blood. Experiments were conducted in sham (n = 6) and IR (n = 6) groups, respectively.

Experiment 5: effect of acute PEG-catalase exposure on the autoregulatory behavior of afferent arterioles following IR.

Because superoxide produced in the cells can be converted by SOD to H₂O₂, we assessed autoregulatory capability following delivery of the membrane-permeable H₂O₂ scavenger PEG-catalase (1,000 U/mL) in the perfusate blood similarly to the aforementioned experiment with PEG-SOD. Experiments were conducted in sham (n = 6) and IR (n = 7) groups, respectively.

Experiment 6: afferent arteriolar response to a P2X1 receptor agonist, β,γ-methylene ATP, following IR.

As renal autoregulation is linked to P2X1 receptor activation (10, 35), we performed concentration/response experiments using the P2X1 agonist β,γ-methylene ATP (β,γ-mATP) in sham (n = 7) and IR (n = 6) rats, respectively. Briefly, after a 5-min baseline diameter recording, increasing β,γ-mATP concentrations (0.01–100 μM) were superfused directly onto the inner cortical surface for 5 min at each concentration.

Experiment 7: afferent arteriolar response to KCl-mediated membrane depolarization following IR.

Because loss of renal autoregulation in IR could be caused by microvessel necrosis (36), we performed KCl concentration/response experiments in sham and IR rats (n = 5/group). Briefly, after a 5-min baseline diameter recording, increasing KCl concentrations (30, 60, and 90 mM) were superfused directly onto the inner cortical surface for 5 min at each concentration.

Metabolic Cage Experiments

To determine the effectiveness of tempol treatment in preventing ischemic kidney injury and inflammation, we prepared another set of rats pretreated with tempol in the drinking water and urine was collected in metabolic cages. The following three groups were studied (n = 7/group): sham, IR, and tempol-pretreated IR (IR + tempol). All rats were individually housed in metabolic cages for 24 h before sham or IR surgery for a baseline recording of body weight, food and water consumption, and urine output. IR + tempol rats were pretreated with tempol (2 mM in drinking water) for 3 days. On day 3, renal ischemia was induced, and rats continued drinking tempol-treated water for 48 h. Additional metabolic recordings were taken between 24 h and 48 h after IR or sham operation. Collected urine was assessed for the urinary excretion of protein, albumin, and kidney injury markers [neutrophil gelatinase-associated lipocalin (NGAL) and kidney injury molecule-1 (KIM-1)]. Rats were euthanized 48 h postsurgery, and kidneys were collected for further mRNA analysis. Blood was collected for plasma creatinine (Cr) measurements.

Real-Time Quantitative RT-PCR Analysis of NOX Subunits and Inflammatory Gene Expression in Kidney Tissue Homogenates

Total RNA was extracted from the kidney cortex and medulla with TRIzol (Invitrogen, Thermo Fisher Scientific, Waltham, MA), treated with DNAase I to remove genomic DNA, and then purified using an RNA purification kit (Cat. No. 12183025, Invitrogen, Thermo Fisher Scientific). DNA-free RNA was reverse transcribed to cDNA using the SuperScript IV RT Kit (Cat. No. 18091050, Invitrogen, Thermo Fisher Scientific). cDNA was amplified with SYBR Green PCR in the LightCycler 480 system (Roche Diagnostics, Indianapolis, IN) and specific primers (Supplemental Table S1; all Supplemental material is available at https://doi.org/10.6084/m9.figshare.12963050.v1) for 40 cycles. Steady-state mRNA levels were calculated according to the threshold cycle generated with the LightCycler 480 software. mRNA expression of p47^phox, p67^phox, glycoprotein (gp)91^phox, monocyte chemoattractant protein (MCP)-1, transforming growth factor (TGF)-β, and tumor necrosis factor (TNF)-α in renal cortical and medullary homogenates was measured (primers are shown in Supplemental Table S1). Relative mRNA expression was normalized to GAPDH.

Proteinuria and Albuminuria Excretion and Plasma Cr Measurements

Urine protein content was assessed using the Bio-Rad protein assay (Bio-Rad). Urine albumin content was assessed using a commercially available rat albumin ELISA kit (Bethyl Laboratories, Montgomery, TX). Urinary excretion of protein and albumin was calculated from urine output over a 24-h period. Plasma Cr was measured using the picric acid method adapted for microtiter plates, as previously described (37).

Urinary NGAL and KIM-1 Excretion

Urine NGAL concentration was measured using a commercially available rat lipocalin-2 ELISA kit (Abcam). Urine KIM-1 concentration was measured using a commercially available rat TIM-1/KIM-1/HAVCR Quantikine ELISA kit (R&D Systems, Minneapolis, MN). Urinary excretion of NGAL and KIM-1 was calculated from urine output over a 24-h period.

Drug Preparation

All drug solutions were prepared fresh on the day of experiments. Tempol and apocynin were dissolved directly into 1% BSA superfusate at concentrations of 1.0 and 0.1 mM, respectively. PEG-SOD and PEG-catalase were dissolved in the reconstituted blood at concentrations of 100 and 1,000 U/mL, respectively. Tempol, apocynin, β,γ-mATP, PEG-SOD, and PEG-catalase were purchased from Sigma-Aldrich.

Statistical Analysis

All data are presented as means ± SE. Arteriolar responses are normalized as a percentage of the baseline diameter. For within-group analysis, one-way ANOVA for repeated measures was used followed by post hoc analysis with Dunnett’s multiple-range test. For analyses among groups, all data from afferent arteriolar experiments were analyzed using two-way ANOVA with a Bonferroni’s post hoc test. All other group data were analyzed using one-way ANOVA followed by a Dunnett’s multiple-range post hoc test. P values of <0.05 were considered significant. n represents the numbers of experiments in afferent arterial studies and numbers of rats in metabolic cage studies.

RESULTS

Autoregulatory Reactivity of Afferent Arterioles Is Blunted in IR but Is Improved by Acute Scavenging of ROS with Tempol

Initial afferent arteriolar diameters averaged 13.2 ± 0.4, 11.7 ± 0.5, and 11.4 ± 0.6 μm in sham (n = 7), IR (n = 8), and IR + tempol (n = 6) groups, respectively. IR significantly reduced baseline diameters compared with the sham group (P < 0.05). In the IR + tempol group, superfusion of tempol slightly but significantly increased baseline diameter to 12.3 ± 0.9 μm, representing a 7 ± 3% increase (P < 0.05).

Figure 1 shows the afferent arteriolar autoregulatory response to step-wise changes in renal perfusion pressure. As previously reported (38), afferent arterioles from sham rats exhibited normal autoregulatory behavior evidenced by appropriate pressure-dependent vasodilation or vasoconstriction (Fig. 1, A and B). In contrast, IR rats showed blunted autoregulatory reactivity. Diameter remained steady at 101 ± 2% of baseline compared with 114 ± 4% in shams as perfusion pressure was reduced from 100 to 65 mmHg (P < 0.05; Fig. 1B). Increasing perfusion pressure to 170 mmHg decreased sham arteriolar diameter to 65 ± 3% of baseline, whereas diameters in IR rats remained within 93 ± 3% of baseline (P < 0.05). In contrast, superfusion of tempol improved pressure-dependent vasoreactivity in IR rats. Arteriolar diameter increased to 107 ± 3% of baseline when perfusion pressure was decreased to 65 mmHg and then decreased to 78 ± 1% of baseline at 170 mmHg (P < 0.05).

Figure 1.

Acute tempol treatment improved autoregulatory behavior of afferent arterioles in ischemia-reperfused (IR) rats. A: afferent arteriolar response to alterations in renal perfusion pressure was measured in male sham (circles, n = 7), IR (squares, n = 8), or IR + tempol (diamonds, n = 6) rats. Tempol (1 mM) was superfused onto the inner cortical surface of the IR kidney for 20 min before the autoregulatory protocol was started and throughout the whole experiment. B: data were normalized as a percentage of the baseline diameter at 100 mmHg. Values are means ± SE. For within-group analysis, one-way repeated-measures ANOVA with a Dunnett’s post hoc test was performed. *P < 0.05 vs. baseline diameter at 100 mmHg in the same group. For analyses among groups, two-way ANOVA with a Bonferroni’s post hoc test was performed. †P < 0.05 vs. sham rats at the same perfusion pressure; #P < 0.05 vs. IR + tempol at the same perfusion pressure. n represents the numbers of experiments.

Acute NOX Inhibition Restores Afferent Arteriolar Autoregulatory Behavior following IR

Similar to the impact of tempol, superfusion of the NOX inhibitor apocynin increased baseline afferent arteriolar diameter from 9.6 ± 0.4 μm to 10.6 ± 0.5 μm (n = 6, P < 0.05), representing a 10 ± 2% increment. Apocynin restored the autoregulatory response in IR rats (Fig. 2, A and B). When perfusion pressure was decreased from 100 to 65 mmHg, arteriolar diameter increased significantly to 112 ± 2% of baseline (Fig. 2B), whereas increasing perfusion pressure to 170 mmHg resulted in a pressure-dependent vasoconstriction to 73 ± 3% of baseline, which was significantly different from IR rats (P < 0.05) but indistinguishable from sham (P > 0.05).

Figure 2.

Acute treatment with apocynin improved autoregulatory behavior of afferent arterioles in ischemia-reperfused (IR) rats. A: the afferent arteriolar response to alterations in renal perfusion pressure was measured in male sham (circles, n = 7), IR (squares, n = 8), or IR + apocynin (diamonds, n = 6) rats. Apocynin (0.1 mM) was superfused onto the inner cortical surface of the IR kidney for 20 min before the autoregulatory protocol was started and throughout the whole experiment. B: data were normalized as a percentage of the baseline diameter at 100 mmHg. Values are means ± SE. For within-group analysis, one-way repeated-measures ANOVA with a Dunnett’s post hoc test was performed. *P < 0.05 vs. baseline diameter at 100 mmHg in the same group. For analyses among groups, two-way ANOVA with a Bonferroni’s post hoc test was performed. †P < 0.05 vs. sham rats at the same perfusion pressure; #P < 0.05 vs. IR + apocynin at the same perfusion pressure. n represents the numbers of experiments.

Acute PEG-SOD Perfusion Restores Afferent Arteriolar Autoregulatory Behavior following IR

Figure 3 shows the effect of PEG-SOD on the pressure-mediated afferent arteriolar responses in IR rats. When perfusion pressure was reduced from 100 to 65 mmHg, arteriolar diameter increased from 11.4 ± 0.8 μm to 13.2 ± 0.9 μm in the IR + PEG-SOD group (n = 6; Fig. 3A), representing 116 ± 2% of the baseline diameter (Fig. 3B). These arterioles exhibited pressure-dependent vasoconstriction to 65 ± 5% of baseline at 170 mmHg, almost identical to responses from sham kidneys.

Figure 3.

Acute polyethylene glycol-conjugated superoxide dismutase (PEG-SOD) treatment improved autoregulatory behavior of afferent arterioles in ischemia-reperfused (IR) rats. A: the afferent arteriolar response to alterations in renal perfusion pressure was measured in male sham (circles, n = 7), IR (squares, n = 8), or IR + PEG-SOD (diamonds, n = 6) rats. PEG-SOD (100 U/mL) was delivered in the reconstituted perfusate blood, and the kidney was perfused for 20–25 min before the autoregulatory protocol was started. B: data were normalized as a percentage of the baseline diameter at 100 mmHg. Values are means ± SE. For within-group analysis, one-way repeated-measures ANOVA with a Dunnett’s post hoc test was performed. *P < 0.05 vs. baseline diameter at 100 mmHg in the same group. For analyses among groups, two-way ANOVA with a Bonferroni’s post hoc test was performed. †P < 0.05 vs. sham rats at the same perfusion pressure; #P < 0.05 vs. IR+ PEG-SOD at the same perfusion pressure. n represents the numbers of experiments.

Acute PEG-Catalase Perfusion Restores Afferent Arteriolar Autoregulatory Behavior following IR

Figure 4 shows the effect of PEG-catalase on pressure-mediated afferent arteriolar responses in IR rats. Similar to PEG-SOD, perfusion with PEG-catalase restored autoregulatory behavior in IR rats. Compared with IR rats, afferent arterioles from the IR + PEG-catalase group (n = 7) vasodilated from 11.1 ± 0.3 μm to 12.6 ± 0.3 μm (114 ± 2% of baseline; Fig. 4B) when pressure was decreased from 100 to 65 mmHg and vasoconstricted to 74 ± 3% of baseline at 170 mmHg, similar to the sham group (P > 0.05).

Figure 4.

Acute polyethylene glycol-conjugated catalase (PEG-catalase) treatment improved autoregulatory behavior of afferent arterioles in ischemia-reperfused (IR) rats. A: the afferent arteriolar response to alterations in renal perfusion pressure was measured in male sham (circles, n = 7), IR (squares, n = 8), or IR + PEG-catalase (diamonds, n = 7) rats. PEG-catalase (1,000 U/mL) was delivered in the reconstituted perfusate blood, and the kidney was perfused for 20–25 min before the autoregulatory protocol was started. B: data were normalized as a percentage of the baseline diameter at 100 mmHg. Values are means ± SE. For within-group analysis, one-way repeated-measures ANOVA with Dunnett’s post hoc test was performed. *P < 0.05 vs. baseline diameter at 100 mmHg in the same group. For analyses among groups, two-way ANOVA with a Bonferroni’s post hoc test was performed. †P < 0.05 vs. sham rats at the same perfusion pressure; #P < 0.05 vs. IR + PEG-catalase at the same perfusion pressure. n represents the numbers of experiments.

Acute PEG-SOD or PEG-Catalase Perfusion Does Not Alter Afferent Arteriolar Autoregulatory Behavior in Sham Rats

Figure 5 shows the effects of PEG-SOD or PEG-catalase on pressure-mediated afferent arteriolar responses in sham rats. Baseline diameters of afferent arterioles were similar among the three groups (Fig. 5A). Neither PEG-SOD (n = 6) nor PEG-catalase (n = 5) altered the overall autoregulatory profiles of afferent arteriolar responses except at a perfusion pressure of 170 mmHg, where perfusion of PEG-SOD slightly but significantly attenuated vasoconstrictor responses compared with the sham group alone (84 ± 4% vs. 65 ± 3%, P < 0.05; Fig. 5B). However, afferent arterioles still maintained the pressure-dependent vasoconstriction in the presence of PEG-SOD.

Figure 5.

Acute polyethylene glycol-conjugated superoxide dismutase (PEG-SOD) or polyethylene glycol-conjugated catalase (PEG-catalase) treatment did not alter afferent arteriolar autoregulatory profile in sham rats. A: the afferent arteriolar response to alterations in renal perfusion pressure was measured in male sham (circles, n = 7), sham + PEG-SOD (squares, n = 6), or sham + PEG-catalase (diamonds, n = 5) rats. PEG-SOD (100 U/mL) or PEG-catalase (1,000 U/mL) was delivered in the reconstituted perfusate blood, and the kidney was perfused for 20–25 min before the autoregulatory protocol was started. B: data were normalized as a percentage of the baseline diameter at 100 mmHg. Values are means ± SE. For within-group analysis, one-way repeated-measures ANOVA with a Dunnett’s post hoc test was performed. *P < 0.05 vs. baseline diameter at 100 mmHg in the same group. For analyses among groups, two-way ANOVA with a Bonferroni’s post hoc test was performed. †P < 0.05 vs. sham rats at the same perfusion pressure. n represents the numbers of experiments.

Reduced Concentration Dependent Reactivity to β,γ-mATP following IR

Figure 6 illustrates the afferent arteriolar responses to superfusion of the specific P2X1 receptor agonist β,γ-mATP. Increasing β,γ-mATP concentrations caused concentration-dependent vasoconstriction in sham rats (n = 7). Starting from a baseline diameter of 14.1 ± 0.5 μm (Fig. 6A), sham arteriolar diameters decreased to 93 ± 1% of baseline at a concentration of 0.01 μM and continued to vasoconstrict to 67 ± 6% of baseline at a concentration of 100 μM (Fig. 6B). In contrast, IR rats (n = 6) showed attenuated vasoconstriction to β,γ-mATP. Afferent arterioles barely vasoconstricted at a concentration of 0.01 μM (100 ± 1% of baseline) and exhibited only a modest reduction in diameter to 88 ± 4% of baseline at a concentration of 100 μM.

Figure 6.

Afferent arteriolar responses to a P2X1 receptor agonist, β,γ-methylene ATP, were impaired in ischemia-reperfused (IR) rats. A: the afferent arteriolar response to superfusion of β,γ-methylene ATP (0.01 − 100 μM) was assessed in male sham (circles, n = 7) and IR (squares, n = 6) rats. B: data were normalized as a percentage of the baseline diameter [control (Con)] at 100 mmHg. Values are means ± SE. For within-group analysis, one-way repeated-measures ANOVA with a Dunnett’s post hoc test was performed. *P < 0.05 vs. baseline diameter in the same group. For analyses between groups, two-way ANOVA with a Bonferroni’s post hoc test was performed. †P < 0.05 vs. sham at the same concentration. n represents the numbers of experiments.

Normal Afferent Arteriolar Vasoconstrictor Responses to Increasing KCl Concentration following IR

Figure 7 shows afferent arteriolar responses to KCl. Mean baseline diameter was smaller in IR kidneys than in sham kidneys (11.2 ± 1.2 μm vs.16.5 ± 1.8 μm in sham, n = 5/group, P < 0.05; Fig. 7A). Increasing KCl concentration caused a similar concentration-dependent vasoconstriction in both sham and IR rats (P > 0.05; Fig. 7B).

Figure 7.

Afferent arteriolar responses of ischemia-reperfused (IR) rats to increasing KCl concentration ([KCl]) did not differ from sham controls. A: the afferent arteriolar response to superfusion of KCl (30, 60, and 90 mM) was assessed in male sham (circles, n = 5) and IR (squares, n = 5) rats. B: data were normalized as a percentage of the baseline diameter [control (Con)] at 100 mmHg. Values are means ± SE. For within-group analysis, one-way repeated-measures ANOVA with a Dunnett’s post hoc test was performed. *P < 0.05 vs. baseline diameter in the same group. For analyses between groups, two-way ANOVA with a Bonferroni’s post hoc test was performed. †P < 0.05 vs. sham at the same concentration. n represents the numbers of experiments.

IR Increased mRNA Expression of NOX Subunits and Cytokines in Renal Tissue

Figure 8A shows mRNA expression for p47^phox, p67 ^phox, and gp91^phox NOX subunits in renal cortical homogenates in sham, IR, and tempol-pretreated IR (n = 6/group), respectively. All three subunits were markedly elevated in IR compared with sham rats (P < 0.05). Tempol pretreatment significantly attenuated mRNA expression of these subunits in IR rats to levels comparable with sham rats (P > 0.05). Similar results were also observed in renal medullary homogenates (Supplemental Fig. S1A).

Figure 8.

Tempol pretreatment decreased ischemia-reperfusion (IR)-induced mRNA expression of NADPH oxidase (NOX) subunits and cytokines in renal cortical homogenates. A: NOX subunit [p47^phox, p67^phox, and glycoprotein (gp)91^phox] mRNA expression in renal cortical homogenates of male sham (white columns, n = 6), IR (gray columns, n = 6), and IR + tempol (black columns, n = 6) rats. B: mRNA expression of monocyte chemoattractant protein (MCP)-1, transforming growth factor (TGF)-β, and tumor necrosis factor (TNF)-α in renal cortical homogenates of sham, IR, and IR + tempol rats. IR + tempol rats were pretreated with tempol (2 mmol/L in drinking water) for 3 days prior to IR and continued over 48 h following IR. Values are means ± SE. Statistical analysis was performed using one-way ANOVA followed by a Dunnett’s multiple range post hoc test. †P < 0.05 vs. sham rats; #P < 0.05 vs. IR + tempol rats. n represents the numbers of rats.

Figure 8B shows mRNA expression for MCP-1, TGF-β, and TNF-α in renal cortical homogenates in sham, IR, and IR + tempol (n = 6/group), respectively. IR markedly increased MCP-1, TGF-β, and TNF-α mRNA expression (P < 0.05 vs. sham), and tempol pretreatment suppressed MCP-1 mRNA expression in IR rats (P < 0.05 vs. IR). Similar results were also observed for renal medullary homogenates (Supplemental Fig. S1B). Both TGF-β and TNF-α tended to decline in IR + tempol rats but were not statistically different from IR rats.

Effect of Tempol Pretreatment on Kidney Function and Injury Markers of IR Rats

Mean plasma Cr concentration markedly increased in IR rats (3.7 ± 0.5 mg/dL vs. 0.7 ± 0.0 mg/dL in sham, n = 7/group, P < 0.05). Tempol pretreatment tended to lower plasma Cr in IR rats (2.7 ± 0.6 mg/dL, n = 7) but was still higher (P < 0.05) than mean plasma Cr of sham rats. Figure 9, A–E, shows the urine output and urinary excretion of protein, albumin, NGAL, and KIM-1. All baseline parameters were similar for sham, IR, and IR + tempol rats (data not shown). Although urine output increased significantly on day 3 for both IR and IR + tempol groups (P < 0.05; Fig. 9A), tempol pretreatment significantly attenuated the proteinuria (Fig. 9B) and albuminuria (Fig. 9C) observed in IR rats. Tempol treatment normalized the proteinuria compared with that observed in sham rats, but albuminuria remained elevated (P < 0.05).

Figure 9.

Impact of tempol pretreatment on urine output, proteinuria, albuminuria, and urinary neutrophil gelatinase-associated lipocalin (NGAL) and kidney injury molecule-1 (KIM-1) excretion in ischemia-reperfused (IR) rats. All male rats (n = 7/group) were individually housed in metabolic cages, and urine (24 h) was collected between 24 and 48 h postsurgery. A: urine output in sham (circles), IR (squares), and IR + tempol (diamonds) rats. B: urinary protein excretion. C: urinary albumin excretion. D: urinary NGAL excretion. E: urinary KIM-1 excretion. IR + tempol rats were pretreated with tempol (2 mmol/L in drinking water) for 3 days prior to IR and continued over 48 h following IR. Values are means ± SE. Statistical analysis was performed using one-way ANOVA followed by a Dunnett’s multiple range post hoc test. †P < 0.05 vs. sham rats; #P < 0.05 vs. IR + tempol rats. n represents the numbers of rats.

Baseline urinary NGAL and KIM-1 excretion were similar for sham, IR, and IR + tempol rats (n = 7/group). IR markedly elevated both NGAL (Fig. 9D) and KIM-1 (Fig. 9E) excretion (P < 0.05 vs. sham). Tempol pretreatment normalized the increase in NGAL excretion (P < 0.05 vs. IR) and reduced KIM-1 excretion in IR rats; however, total KIM-1 excretion remained elevated (P < 0.05 vs. sham).

DISCUSSION

The current study provides direct evidence that IR decreases baseline afferent arteriolar diameter and impairs the intrinsic property of afferent arteriolar autoregulatory capability. Afferent arterioles from IR rats exhibited reduced vasoreactivity to perfusion pressure manipulation. IR also attenuated afferent arteriolar vasoconstriction to β,γ-mATP, an agonist of P2X1 receptors that is postulated to be essential for renal autoregulatory responsiveness (39–42). Treatment with tempol, apocynin, PEG-SOD, and PEG-catalase restored afferent arteriolar autoregulatory behavior following IR injury. These results suggest that IR-induced accumulation of ROS (superoxide and/or H₂O₂) in kidneys leads to renal microvascular dysfunction manifested as autoregulatory impairment. The superfusion experiments in particular showed the production of ROS directly altered afferent arteriolar autoregulation in a reversible, localized fashion. Importantly, in additional experiments, pretreatment of rats with tempol mitigated IR-induced elevations of renal mRNA expression for p47^phox, p67^phox, gp91^phox, and MCP-1. The combined findings demonstrated that IR injury generated sufficient ROS to promote a reversible afferent arteriolar autoregulatory dysfunction. An associated renoprotective effect of tempol in IR was demonstrated by decreased urinary excretion of albumin, protein, and NGAL.

Increased RVR, reduced GFR, endothelial dysfunction, and tubular damage are major characteristics of IR-induced kidney injury (6–8). Recent evidence also indicates that reduced peritubular capillary density is linked to the transition from AKI to CKD (43, 44). Although there remains some debate on whether RBF is reduced after IR, our study supports the concept of increased RVR in ischemic kidneys, as afferent arteriolar diameter from IR rats was determined to be significantly smaller than that from sham rats. Superfusion of tempol or apocynin on IR kidneys for only 20 min slightly but significantly increased baseline arteriolar diameters but had no effect on baseline diameter of normal afferent arterioles, as previously demonstrated (30, 34). This observation is further supported by perfusion with PEG-SOD or PEG-catalase (Fig. 5). Both PEG-SOD and PEG-catalase had little influence on the basal afferent arteriolar diameter of sham rats. Together, these observations suggest that IR-induced ROS accumulation may contribute to the increased RVR in IR.

Renal autoregulation plays a critical role in controlling RBF and GFR. Both RBF and GFR are principally regulated by afferent arteriolar diameter and reactivity. Earlier studies of renal autoregulation in IR animal models mostly used in vivo animal models, in vitro nonblood-perfused kidneys, or micropuncture approaches (18, 20, 21, 45–47). However, it has not been fully determined if renal autoregulation is impaired or enhanced after IR (18–22). The current study using the in vitro blood-perfused JMN preparation provides direct evidence that afferent arterioles of IR rats exhibit impaired pressure-induced diameter adjustments that are essential for appropriate autoregulatory responses. This impairment is manifested as a flatter pressure-diameter relationship in IR rat kidneys compared with sham rat kidneys with normal autoregulatory control (Fig. 1). As normal renal autoregulation is largely dependent on P2X1 receptor activation (10, 35), we determined the afferent arteriolar response to β,γ-mATP, a selective P2X1/P2X3 receptor agonist. Compared with afferent arterioles from sham rats, afferent arterioles from IR rats exhibited significantly attenuated vasoconstrictor reactivity to β,γ-mATP while vasoconstriction to KCl-mediated membrane depolarization was intact. The findings indicate a reversible process that promotes autoregulatory impairment in IR but is not caused by microvessel necrosis as previously reported (36). Since P2X1 is a major purinoceptor expressed in preglomerular microvessels (48, 49), attenuation of afferent arteriolar vasoconstriction to β,γ-mATP is consistent with the impaired autoregulatory behavior of afferent arterioles after IR.

Increased ROS production has been considered an initial event contributing to the pathogenesis of renal IR injury (24, 25, 50). Several clinical and experimental studies, however, have concluded with mixed results regarding the benefit of antioxidant treatment in AKI. For example, earlier micropuncture studies using a rat IR model revealed an increase in nephron plasma flow and single nephron GFR with probucol treatment despite no improvement on tubular injury (51, 52). Conversely, Chatterjee et al. (53) reported that tempol treatment reduced lipid peroxidation in rats subjected to 45 min of ischemia followed by 6 h of reperfusion. In an IR rat model induced by aortic clamping, tempol was less renoprotective despite reduction of renal inducible nitric oxide synthase and IL-6 expression (54). Clinical studies have shown that antioxidant treatment did not provide significant benefit for patients with a variety of AKI and CKD (50). In the present study, we found that 20 min of exposure to topical tempol restored the pressure-dependent responsiveness of afferent arterioles from IR rats, supporting the postulate that excess ROS compromise afferent arteriolar autoregulatory ability in IR rats. ROS are generated by the univalent reduction of oxygen by a number of enzymatic reactions including NOX, xanthine oxidase, mitochondria, and uncoupled nitric oxide synthase (55). In the past decade, many studies showed renoprotective benefits of NOX inhibition with apocynin against ischemic injury (56–59), but those studies focused on the effects of apocynin on kidney morphological changes or markers of kidney function following IR. One study by Basile et al. (60) showed that apocynin normalized ANG II-enhanced renal vasoconstriction in IR rats, suggesting that excess NOX-induced ROS modify renal microvascular reactivity. However, it is unknown if NOX-induced ROS alter afferent arteriolar autoregulatory responses. As afferent arterioles mainly contain NOX (34, 61), we determined the impact of NOX inhibition on renal autoregulation and observed that apocynin also completely restored the autoregulatory response in IR. The autoregulatory profile is virtually identical to the sham, suggesting that excess NOX-derived ROS alter renal microvascular function by altering mechanosensitivity. Overall, our results indicate that acute ROS scavenging or reducing NOX-mediated ROS accumulation restores renal microvascular autoregulatory capability in an IR rat model.

Superoxide is a major form of ROS and is involved in the myogenic response in normal mouse afferent arterioles (23, 62, 63). Superoxide is further converted to H₂O₂ by SOD, whereas H₂O₂ is finally converted into H₂O and O₂ by catalase. Our current findings revealed that perfusion of either PEG-SOD or PEG-catalase normalized afferent arteriolar autoregulatory ability in IR rats, as the pressure-diameter profile was undistinguishable from responses in sham rats. In contrast, neither PEG-SOD nor PEG-catalase perfusion altered pressure-dependent vasoreactivity of afferent arterioles in our experimental setting. These findings suggest that both superoxide and H₂O₂ may not play a major role in renal autoregulation under normal physiological conditions but they do contribute to afferent arteriolar autoregulatory impairment in IR rats. H₂O₂ was recently reported to have either renal vasodilator or vasoconstrictor actions based on the redox status (23, 64, 65). For example, 10 μM H₂O₂ led to vasodilation of isolated rat intrarenal arteries but caused vasoconstriction in an antioxidant environment if the intrarenal arteries were preincubated with butylated hydroxytoluene (65). Moreover, studies using mouse isolated-perfused afferent arterioles revealed that the myogenic response of afferent arterioles was unaffected by the addition of PEG-catalase, suggesting that the myogenic response is H₂O₂ independent under physiological conditions (63). The myogenic response of afferent arterioles, however, can be blunted by preincubation with H₂O₂ (25 μM) (63), highlighting that excess H₂O₂ production can impair renal autoregulatory capability. This is consistent with our current findings that H₂O₂ accumulation contributes to IR-induced renal autoregulatory impairment.

It is interesting that both PEG-SOD and PEG-catalase restored afferent arteriolar autoregulatory responsiveness in IR rats. Theoretically, application of PEG-SOD would reduce superoxide production, whereas PEG-catalase would reduce H₂O₂ and have little effect on superoxide. However, several studies have suggested that SOD could depend, at least in part, on the activity of the H₂O₂-scavenging system (66–69). Hink et al. (66) reported that H₂O₂ inactivated SOD activity in a dose-dependent manner in the apolipoprotein E^−/− mouse aorta, suggesting that H₂O₂ exacerbates oxidative stress and aortic injury of apolipoprotein E^−/− mice by inactivating SOD. Furthermore, Salo et al. (68, 69) found that bovine SOD is inactivated by exposure to H₂O₂ and that this inactivation was enhanced by incubation with erythrocytes. Following IR, both SOD and catalase activities decreased, whereas superoxide and H₂O₂ production increased in the afferent arterioles and renal cortex, as demonstrated in IR mice (70). As a consequence, one could speculate that removing H₂O₂ by perfusion of PEG-catalase might improve SOD activity and, therefore, reduce superoxide accumulation in afferent arterioles of IR rats. Currently, the effect of PEG-catalase on SOD activity has not been ascertained in the IR kidney. Certainly, more studies are required to define the underlying mechanisms by which IR-induced superoxide and/or H₂O₂ accumulation alters the ability of afferent arterioles to sense changes in transmural pressure.

Multiple studies have highlighted the strong association between ROS, inflammation, and impaired renal autoregulation (9, 14, 71). For example, Fellner et al. (30) reported that apocynin treatment prevented afferent arteriolar autoregulatory impairment in rats fed a high-salt diet through a reduction of kidney p67^phox expression. Treatment with the chemokine (C-C motif) receptor 2b antagonist abated renal monocyte and macrophage infiltration but also protected against renal autoregulatory impairment in ANG II-infused hypertensive rats (29). Those studies suggested important roles for ROS and MCP-1 in renal microvascular dysfunction. However, it is unknown if this also occurs in IR kidney injury. Thus, we prepared another set of rats pretreated with tempol and found that mRNA expression of the major NOX subunits (p47^phox, p67^phox, and gp91^phox), as well as other cytokines and MCP-1, was substantially upregulated in both cortical and medullary tissue of IR kidneys. Tempol pretreatment, however, prevented this enhanced mRNA expression such that they remained at the control level, suggesting enhanced NOX activity in IR kidneys. Tempol pretreatment also mitigated the rise in MCP-1 mRNA expression. Our current findings reveal a potential link between increased renal ROS and MCP-1 accumulation and IR-induced afferent arteriolar autoregulatory impairment. Further studies are needed to determine if NOX and MCP-1 upregulation is derived from the renal microvasculature or if it reflects cross talk from necrotic tubular cells following IR.

Tempol pretreatment did not alter kidney function but did reduce protein and albumin excretion in IR rats. We also measured urinary NGAL and KIM-1 excretion, which are considered early biomarkers of proximal tubular injury after IR. Consistent with prior reports (72, 73), both NGAL and KIM-1 excretion were markedly increased in IR rats. Remarkably, tempol pretreatment prevented the increase in NGAL excretion, whereas KIM-1 excretion remained elevated in IR rats. NGAL is released from immune cells and is elevated in IR kidneys and inflammatory conditions (74), but a recent study in mice also indicated that NGAL is significantly generated from hepatocytes after kidney IR (75), suggesting that NGAL can be released from remote organs (liver) after ischemic kidney injury. Overall, our study revealed that scavenging ROS restored afferent arteriolar autoregulatory reactivity and reduced kidney inflammation and injury following IR.

Perspectives and Significance

Renal autoregulation is critical in controlling renal blood flow and GFR via precise adjustment of afferent arteriolar diameter. This study provided direct evidence that ischemia-reperfusion led to renal microvascular dysfunction manifested by impaired afferent arteriolar autoregulatory efficiency and attenuated vasoconstriction to P2X1 receptor activation. Acute administration of scavengers of ROS following IR restored afferent arteriolar autoregulatory capability in IR rats, indicating that renal IR led to reversible impairment of afferent arteriolar autoregulatory capability. Thus, impaired afferent arteriolar autoregulatory capacity after renal IR was largely associated with alteration of mechanosensitivity rather than irreversible microvascular structural changes such as necrosis. Afferent arteriolar autoregulatory capability was also normalized by acute perfusion of PEG-SOD or PEG-catalase, suggesting that excessive superoxide and/or H₂O₂ contributed to IR-induced impairment of renal autoregulation. Furthermore, we demonstrated that scavenging ROS by pretreatment with tempol ameliorated IR-mediated increases in steady-state mRNA of NOX and MCP-1 in the kidney. Although pretreatment experiments using tempol demonstrated efficacy, the overall clinical relevance of these latter studies remains uncertain. However, of potential relevance to the clinical situation, intervention with antioxidant treatment following renal IR may improve outcomes in patients by preserving renovascular autoregulatory function and potentially preventing the progression to CKD after AKI. A second limitation of the present studies was the use of the in vitro blood-perfused JMN preparation in which both myogenic response and TGF are intact (76–78). Future studies will therefore be required to determine which autoregulatory components—the myogenic response, TGF, or both—contributed to the overall autoregulatory impairment after IR.

GRANTS

This study was supported by National Institutes of Health (NIH) Grant R01DK106500 and American Heart Association Grant-in-Aid 15GRNT25240015 (to Z.G.) and by NIH R01DK044628 (to E.W.I.). W. Feng was supported by American Heart Association Scientist Development Grant 15SDG25760063. P. W. Sanders was supported by Merit Award 2 I01 CX001326 from the United States Department of Veterans Affairs Clinical Sciences R&D Service and by the NIH George M. O'Brien Kidney and Urological Research Centers Program (Grant 2P30DK079337).

DISCLOSURES

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

AUTHOR CONTRIBUTIONS

W.F., E.W.I., and Z.G. conceived and designed research; W.F., C.E.R., I.E.O., S.R.A., S.I.M., and Z.G. performed experiments; W.F., C.E.R., I.E.O., S.R.A., S.I.M., and Z.G. analyzed data; W.F., P.W.S., E.W.I., and Z.G. interpreted results of experiments; W.F. and Z.G. prepared figures; W.F., C.E.R., and Z.G. drafted manuscript; W.F., C.E.R., I.E.O., S.R.A., S.I.M., P.W.S., E.W.I., and Z.G. edited and revised manuscript; W.F., C.E.R., I.E.O., S.R.A., S.I.M., P.W.S., E.W.I., and Z.G. approved final version of manuscript.