Tempol Prevents Altered K⁺ Channel Regulation of Afferent Arteriolar Tone in Diabetic Rat Kidney

Abstract

Experiments were performed to test the hypothesis that oxidative stress underlies the enhanced tonic dilator impact of inward-rectifier K⁺ (K_IR) channels on renal afferent arterioles of rats with streptozotocin-induced diabetes. Sham and diabetic rats were left untreated or provided tempol in their drinking water for 26±1 days, after which afferent arteriolar lumen diameter and its responsiveness to K⁺ channel blockade were measured using the in vitro blood-perfused juxtamedullary nephron technique. Afferent diameter averaged 19.4±0.8 μm in sham rats and 24.4±0.8 μm in diabetic rats (P<0.05). The decrease in diameter evoked by Ba²⁺ (K_IR channel blocker) was 3-times greater in diabetic rats than in sham rats. Glibenclamide (K_ATP channel blocker) and tertiapin-Q (Kir1.1/Kir3.x channel blocker) decreased afferent diameter in diabetic rats, but had no effect on arterioles from sham rats. Chronic tempol treatment prevented diabetes-induced increases in both renal vascular dihydroethidium staining and baseline afferent arteriolar diameter. Moreover, tempol prevented the exaggeration of afferent arteriolar responses to Ba²⁺, tertiapin-Q, and glibenclamide otherwise evident in diabetic rats. Preglomerular microvascular smooth muscle cells expressed mRNA encoding Kir1.1, Kir2.1 and Kir6.1. Neither diabetes nor tempol altered Kir1.1, Kir2.1, Kir6.1 or SUR2B protein levels in renal cortical microvessels. To the extent that the effects of tempol reflect its antioxidant actions, our observations indicate that oxidative stress contributes to the exaggerated impact of Kir1.1, Kir2.1 and K_ATP channels on afferent arteriolar tone during diabetes and that this phenomenon involves post-translational modulation of channel function.

Long-term renal microvascular dysfunction is one of many complications of type 1 diabetes (T1D) that are thought to contribute to nephropathy.¹ In the early stages of the disease, the main alteration that occurs in the renal microvasculature is afferent arteriolar dilation; however, the events underlying tonic dilation of this vessel are poorly understood. Many studies have demonstrated an increase in reactive oxygen species (ROS) generation during diabetes,^2–5 as well as altered activity of one or more of the major antioxidant enzymes responsible for maintaining redox balance (i.e. superoxide dismutase (SOD) and catalase).^3–5 Treatment of diabetic rats with the membrane-permeable antioxidant tempol has been shown to decrease ROS accumulation and to increase the activity of antioxidant enzymes.^3,4 Tempol has also been shown to recover NO-dependent vasodilator function in afferent arterioles from diabetic rabbits,⁶ likely reflecting the effect of reduced superoxide anion ( ${\text{O}}_2^{·-}$) levels to decrease peroxynitrite generation, thereby restoring NO bioavailability.

Preglomerular microvascular resistance is coupled to vascular smooth muscle membrane potential and K⁺ channels are important determinants of membrane potential. In vascular smooth muscle, opening of K⁺ channels allows efflux of K⁺ from the cell resulting in membrane hyperpolarization, which causes closure of voltage-gated Ca²⁺ channels (VGCCs) and reduced Ca²⁺ influx, thereby evoking vasodilation. We previously reported that members of the inward-rectifier K⁺ channel family (K_IR channels), including Kir2.1, Kir1.1/3.x and ATP-sensitive K⁺ (K_ATP) channels, exert an exaggerated tonic dilator influence on afferent arteriolar tone in rats with streptozotocin (STZ)-induced T1D.^7,8 Moreover, afferent arteriolar dilator responses to K_ATP activation (pinacidil) are accentuated during T1D,⁷ while responses to K_IR activation (20 mmol/L K⁺) are attenuated.⁸ The mechanism underlying altered K⁺ channel control of afferent arteriolar function during T1D remains unclear. As ROS can alter the function of several K⁺ channels in vascular smooth muscle,⁹ we hypothesized that oxidative stress underlies the changes in K⁺ channel function that arises in the afferent arteriole during T1D. The goal of the present study was to determine if chronic treatment with the membrane-permeable antioxidant, tempol, would prevent the exaggerated tonic impact of K_IR channel family members on lumen diameter of the afferent arteriole in rats with STZ-induced T1D. We also explored the postulate that the changes in K_IR channel regulation of afferent arteriolar function might reflect altered expression of the channel at the mRNA and/or protein levels.

Methods

For an expanded Methods section, see the online Data Supplement (http://hyper.ahajournals.org).

Animals

Animals in this study (male Sprague-Dawley rats weighing 275–300 g; n=135) were treated according to the National Institutes of Health Guide for the Care and Use of Laboratory Animals using procedures approved by the University of Nebraska Medical Center Institutional Animal Care and Use Committee. T1D was induced by injection of streptozotocin (65 mg/kg IV; STZ rats) and moderate hyperglycemia was maintained by partial insulin replacement, as described previously.⁸ Sham rats received vehicle treatments. Half of the rats in each group received chronic antioxidant therapy, achieved by addition of tempol to the drinking water. As STZ rats are polydipsic, the concentration of tempol provided in the water was lower in the STZ+Tempol group (0.4 mmol/L) than in the Sham+ Tempol group (1 mmol/L). In preliminary experiments, these concentrations resulted in similar daily tempol doses in the two groups of rats. Terminal experiments were performed 26±1 days after STZ (or vehicle) injection. Each rat was housed in a metabolic cage for the three days preceding the terminal experiment, with urine output and water intake measured during the final 24 hours. An aliquot of the urine sample was retained and frozen until analysis for indices of oxidative stress.

In Vitro Blood-Perfused Juxtamedullary Nephron Technique

Under pentobarbital anesthesia, the left kidney was removed and flash-frozen, and the right kidney was perfused and harvested for study of afferent arteriolar function using the rat in vitro blood-perfused juxtamedullary nephron technique, as detailed previously.^10,11 Perfusion pressure at the cannula tip in the renal artery was set and maintained at 110 mmHg. The tissue surface was bathed continuously with Tyrode’s solution containing 10 g/L bovine serum albumin, approximating the composition of renal interstitial fluid. Videometric techniques were used to measure arteriolar lumen diameter, as previously described,⁸ during completion of the following protocols. Protocol 1: After a stabilization period, during which a single arteriole was selected for study, images of the arteriole were recorded during a 10 min baseline period (exposure to Tyrode’s bath alone). Subsequently, the impact of K_IR channels on afferent arteriolar diameter was evaluated based on the effects of sequential exposure to Tyrode’s bath containing 10, 30 and 100 μmol/L BaCl₂ (5 min each), followed by a 10 min recovery period (Tyrode’s bath alone). This procedure was typically followed by an assessment of the tonic impact of Kir1.1/Kir3.x channels, based on the effects of sequential 5 min exposure to 1, 10 and 100 nmol/L tertiapin-Q (TPQ).¹² Protocol 2: In other kidneys (not exposed to Ba²⁺ or TPQ), the baseline period was followed by an evaluation of the impact of K_ATP channels on afferent arteriolar function based on responses to sequential exposure to 30, 100 and 300 μmol/L glibenclamide (5 min each). The efficacy and specificity of the concentrations of Ba²⁺, TPQ and glibenclamide used in this study have been addressed previously.^7,8

Indices of Oxidative Stress

Dihydroethidium (DHE) staining was utilized to detect superoxide in kidneys from each group of rats, according to the method described by Zalba et al.¹³ Flash-frozen kidneys were sectioned and images of the renal cortex were obtained by laser confocal microscopy. DHE staining intensity was quantified from these images using Adobe^® Photoshop^® CS4 Extended.

Plasma and urine samples from each group of rats were assayed for thiobarbituric acid reactive substances (TBARS) using the OxiSelect™ TBARS Assay Kit (Cell Biolabs). The Amplex^® Red Hydrogen Peroxide/Peroxidase Assay Kit (Molecular Probes) was used to quantify H₂O₂ levels. Both assays were performed according to kit manufacturer instructions.

Preglomerular Microvascular Smooth Muscle Cell Culture

Preglomerular microvascular smooth muscle cells (PVSMCs) from normal rat kidney were cultured by the explant method, as previously described.¹⁴ PVSMCs at passage 3 were maintained in 5 or 20 mmol/L glucose for 2–3 weeks, then growth-arrested by exposure to serum-free conditions for 1 day prior to harvesting.

RNA Extraction and Real-Time RT PCR

RNA was extracted from cultured PVSMCs using the Absolutely RNA Miniprep Kit (Stratagene). Total RNA was measured in a 2 μl sample using a Nanodrop (Thermo Scientific). Samples were stored at −80°C until real-time RT-PCR was performed using the Rotor Gene Real-Time Detection System, following the QuantiTect^® SYBR^® Green RT protocol (Qiagen).

Cortical Microvessel Isolation

Vessels were isolated from rat kidney by sieving, according to the following method. Rats were anesthetized with pentobarbital sodium and the kidney was perfused with cold Ca²⁺- and Mg²⁺-free HBSS containing 5 or 20 mmol/L glucose (for Sham and STZ rats, respectively). The kidney was removed, sliced longitudinally, and demedullated. The remaining cortex pieces were sequentially sieved over 250 and 150 μm nylon mesh, rinsed by dental water jet, placed in a cold collagenase solution, and incubated at 37°C while shaking for 20 min. After incubation the vessels were gently sieved again over the 150 μm nylon mesh to wash away the collagenase, and placed in cell lysis buffer for western blotting.

Western Blot

Proteins in cortical microvessel lysates were separated by SDS-PAGE (4–15% gradient gel), transferred onto PVDF membranes, dried, blocked and incubated overnight with one of the following primary antibodies: rabbit anti-Kir1.1 (Chemicon, AB5196; 1:200 dilution), rabbit anti-Kir2.1 (Alomone, APC-026; 1:100), rabbit anti-Kir6.1 (Santa Cruz, sc20808; 1:200), goat anti-SUR2B (Santa Cruz, sc5793; 1:200), mouse anti-β-actin (Abcam, ab627c; 1:200). The membranes were then washed five times prior to 1 hr incubation with secondary antibody conjugated to an infrared dye. Quantification was achieved using the Odyssey™ Infrared Imaging System.

Statistical Analyses

Simple between-group comparisons were made by ANOVA. Effects of pharmacological agents on arteriolar lumen diameter were evaluated based on the average diameter measured during the final 2 min of each treatment period, with statistical comparisons made by two-way repeated measures ANOVA and, when appropriate, the Holm-Sidak multiple comparisons test. P values <0.05 were considered statistically significant. All data are presented as means±SEM.

Results

During the 3–4 weeks after injection of STZ/vehicle, blood glucose concentration was measured twice weekly and found to average 23.5±0.7 mmol/L (n=33 rats) in STZ rats, a value that was elevated significantly compared to Sham rats (5.2±0.1 mmol/L, n=31 rats). Chronic tempol treatment had no effect on blood glucose levels in either group (Sham+Tempol, 5.2±0.1 mmol/L, n=34 rats; STZ+Tempol, 23.3±0.9 mmol/L, n=37 rats). The rats in each group gained weight during this period; however, STZ and STZ+Tempol rats gained less weight than Sham and Sham+Tempol rats. Accordingly, at the time of the terminal experiment, body weight was significantly lower in STZ rats (324±5 g) and STZ+Tempol rats (321±5 g) compared to Sham rats (364±4 g) and Sham+Tempol rats (367±4 g). As the degree of hyperglycemia achieved in the rats with T1D was somewhat greater than evident in our preliminary studies, the extent of the osmotic diuresis was also magnified. As a result, the STZ+Tempol rats drank more water than anticipated (STZ+Tempol, 155±10 ml/day; Sham+Tempol, 35±3 ml/day; P<0.05). Consequently, the dose of tempol received by STZ+Tempol rats drinking 0.4 mmol/L tempol (62±4 μmol/day) was greater than that received by Sham+Tempol rats drinking 1 mmol/L tempol (35±3 μmol/day, P<0.05).

Plasma levels and urinary excretion of TBARS were elevated significantly after 3–4 weeks of hyperglycemia in STZ and STZ+Tempol rats (Table 1) when compared to Sham. Plasma H₂O₂ concentration did not differ among the four groups of rats. Urinary H₂O₂ excretion was similar in Sham and Sham+Tempol rats, but was significantly increased in STZ rats when compared to Sham (P<0.001). The increased H₂O₂ excretion evident in STZ rats was blunted significantly in STZ+Tempol rats. Thus, tempol reduced the hyperglycemia-induced increase in H₂O₂ excretion but had no effect on plasma H₂O₂ levels.

Table 1.

Indices of Oxidative Stress

Groups	Plasma TBARS, μmol/L	TBARS Excretion, μmol/d	Plasma H2O2, nmol/L	H2O2 Excretion, nmol/d
Sham	7.8±1.1 (n=5)	0.31±0.03 (n=12)	1295±54 (n=5)	33±8 (n=14)
Sham+Tempol	6.6±0.4 (n=6)	0.36±0.04 (n=10)	1041±109 (n=6)	19±9 (n=12)
STZ	19.0±2.3* (n=5)	17.9±2.5* (n=12)	1792±172 (n=6)	403±89* (n=13)
STZ+Tempol	18.3±1.5* (n=6)	19.2±2.5* (n=12)	1663±181 (n=6)	190±57*† (n=12)

Values are mean ± SEM. TBARS indicates thiobarbituric acid reactive substances (MDA equivalents). Excretory data were derived from 24 hour urine collection in metabolic cages during the 24 hr prior to the terminal experiment. Plasma samples were obtained under anesthesia prior to harvesting the kidney for study of arteriolar function.

^*P<0.05 vs Sham,

^†P<0.05 vs STZ.

DHE staining to was utilized to assess the effect of tempol on production of ROS in the kidneys of Sham and STZ rats. Figure 1A provides representative DHE fluorescence images of renal cortical sections from each treatment group. The quantitative analysis shown in Figure 1B reveals that renal cortical vessel DHE fluorescence in STZ rats was significantly greater than that evident in Sham rats. Chronic tempol treatment of STZ rats prevented the elevation in vascular DHE fluorescence. Although the tendency for increased DHE fluorescence intensity in renal cortical tubules of STZ rats did not achieve statistical significance, tubular DHE staining in STZ+Tempol rats was significantly lower than in untreated STZ rats. Tempol treatment of Sham rats did not alter DHE fluorescence intensity in vessels or tubules. These data are in agreement with tempol acting as an antioxidant in the renal cortex, especially in vascular structures.

Figure 1.

Effects of T1D and chronic tempol treatment on renal cortical DHE staining. A, Representative images showing DHE staining of renal cortical vessels (arrows) and tubules in the four groups of rats. Scale bar = 50 μm. B, Quantitative comparison of DHE staining intensity in renal cortical vessels and adjacent tubules. n=4–7 vessels from 2–4 rats/group; *P<0.05 vs Sham, †P<0.05 vs untreated.

Figure 2 shows afferent arteriolar lumen diameter under baseline conditions in the four animal groups. Baseline afferent arteriolar diameter in STZ rats was 25% greater than in Sham rats (P<0.05). In contrast, baseline lumen diameter did not differ between STZ+Tempol rats, Sham rats and Sham+Tempol rats. Thus, chronic tempol treatment prevented the afferent arteriolar dilation typically evident in the STZ-model of T1D, while not altering baseline afferent diameter in non-diabetic rats.

Figure 2.

Effects of T1D and chronic tempol treatment on afferent arteriolar baseline diameter. n=20–23 arterioles/group. †P<0.05 vs Sham.

Figure 3A illustrates the afferent arteriolar lumen diameter responses to increasing concentrations of Ba²⁺ in each group of rats. In Sham rats, Ba²⁺ evoked a vasoconstrictor response (11±3% decrease in diameter, P<0.05) only at the highest concentration employed (100 μmol/L). In STZ rats, significant constrictor responses were evident during exposure to 30 and 100 μmol/L Ba²⁺, and these responses were exaggerated compared with arterioles from Sham rats (30 μmol/L, Δ diameter = −15±2%; 100 μmol/L, Δ diameter = −35±6%; both P<0.05 vs Sham). Afferent arteriolar responses to Ba²⁺ did not differ between Sham and Sham+Tempol rats; however, the exaggerated afferent arteriolar Ba²⁺ responsiveness evident in STZ rats was not manifest in STZ+Tempol rats. Indeed, the response to 100 μmol/L Ba²⁺ was reduced in STZ+Tempol rats compared with STZ rats (P<0.05). As a result, afferent arteriolar diameter responses to Ba²⁺ did not differ significantly between STZ+Tempol, Sham and Sham+Tempol groups. Thus, chronic tempol treatment prevented the exaggerated afferent arteriolar contractile response to Ba²⁺ otherwise evident in STZ rats, suggesting that oxidative stress contributes to the augmented dilator impact of K_IR channels on afferent arteriolar tone during T1D.

Figure 3.

Effects of T1D and chronic tempol treatment on afferent arteriolar diameter responses to K⁺ channel blockade. Shown are changes in lumen diameter evoked by increasing bath concentrations of A, Ba²⁺ (K_IR channel blocker; n=7–8 arterioles/group), B, TPQ (Kir1.1/Kir3.x channel blocker; n=5–8 arterioles/group) and C, glibenclamide (K_ATP channel blocker; n=6–8 arterioles/group). Due to high affinity binding of glibenclamide to albumin,³³ the free glibenclamide concentration in these experiments is estimated to be 1–10% of the values shown in panel C. *P<0.05 vs baseline diameter, †P<0.05 vs Sham.

Figure 3B summarizes the effects of T1D and tempol on afferent arteriolar responses to TPQ. Afferent arteriolar lumen diameter was unaffected by TPQ in kidneys from Sham and Sham+Tempol rats. In contrast, a 100 nmol/L TPQ evoked a significant decrease in afferent arteriolar diameter in STZ rats (Δ diameter = −11±3%, P<0.05), but not in STZ+Tempol rats. Thus, chronic tempol treatment prevented the afferent arteriolar constrictor response to TPQ that otherwise occurs in STZ rats, suggesting that tonic activation of afferent arteriolar Kir1.1/Kir3.x channels arises via an oxidative stress-dependent mechanism during T1D.

The results of experiments performed to evaluate the effects of T1D and tempol on K_ATP channel-dependent regulation of afferent arteriole tone are illustrated in Figure 3C. Glibenclamide (30–300 μmol/L) had no effect on afferent arteriolar lumen diameter in kidneys from Sham or Sham+Tempol rats. However, significant contractile responses to glibenclamide were evident in kidneys from STZ rats, with afferent arteriolar diameter decreasing 11±2 and 21±4% during exposure to 100 and 300 μmol/L glibenclamide, respectively. Moreover, across the entire glibenclamide concentration range employed in this study, afferent arteriolar diameter responses in STZ rats significantly exceeded those observed in Sham rats. Compared with the responses observed in STZ rats, afferent diameter responses to 100 and 300 μmol/L glibenclamide were significantly attenuated in kidneys from STZ+Tempol rats. In fact, glibenclamide did not significantly alter afferent arteriolar diameter in STZ+Tempol rats, yielding responses that did not differ from those observed in Sham and Sham+Tempol rats. The ability of chronic tempol treatment to prevent the afferent arteriolar contractile response to glibenclamide that is typically evident in STZ rats suggests that oxidative stress contributes to the emergence of a tonic dilator impact of K_ATP channels on the afferent arteriole during T1D.

Real-time RT-PCR was used to measure mRNA expression of the pore-forming subunits of TPQ-sensitive channels (Kir1.1, Kir3.1, Kir3.4), Ba²⁺-sensitive K_IR channels (Kir2.1), and K_ATP channels (Kir6.1, Kir6.2, Kir6.3; characteristically expressed in the vasculature, pancreas and zebrafish, respectively). Figure 4 summarizes the real-time RT-PCR data for the seven channels after maintenance of PVSMCs for 2–3 wks in culture media containing either normal (5 mM) or high (20 mM) concentrations of glucose. Kir1.1, Kir2.1 and Kir6.1 mRNA were detected in cultured PVSMCs, while no mRNA was detected for Kir3.1, Kir3.4 or the K_ATP negative controls (Kir6.2 and Kir6.3). The threshold cycle did not differ between normal and high glucose-treated PVSMCs for any of the channels studied, indicating that the T1D-induced changes in afferent arteriolar lumen diameter responses to K⁺ channel blockers were not likely due to a glucose-induced increase in Kir gene transcription in PVSMCs.

Figure 4.

K channel mRNA levels in PVSMCs indicated by real-time RT-PCR. PVSMCs (passage 3) were maintained in media containing 5 or 20 mM glucose for 2–3 weeks. α-Tubulin was used as loading control. Kir6.2 and Kir6.3 were negative controls. n=5 experiments, with each experiment using PVSMCs from a different rat cultured under both low and high glucose conditions.

Figure 5 summarizes the Kir channel protein levels in renal cortical microvessels obtained from each group of rats. Western blotting confirmed the presence of Kir1.1, Kir2.1, Kir6.1 and SUR2B at the protein level in renal cortical microvessels. The Kir1.1 antibody also detected a strong signal in renal medullary thick ascending limb homogenate (positive control; data not shown). Kir1.1 and Kir2.1 protein expression did not differ between vessels from the Sham, Sham+Tempol, STZ and STZ+Tempol groups. Thus, neither T1D nor chronic antioxidant therapy altered Kir1.1 or Kir2.1 expression in renal cortical microvessels. Additionally, we probed for pore-forming and regulatory subunits comprising vascular K_ATP channels, Kir6.1 and SUR2B, respectively; however, there was no significant difference in protein levels of either subunit between Sham, Sham+Tempol, STZ and STZ+Tempol groups. Thus, neither T1D nor chronic antioxidant therapy altered Kir6.1 or SUR2B expression in renal cortical microvessels. These western blot data demonstrate that the effects of T1D and chronic tempol treatment on Kir regulation of arteriolar function are not due to change in channel protein level but, rather, likely reflect post-translational alterations in channel activity.

Figure 5.

Effects of T1D and chronic tempol treatment on Kir1.1, Kir2.1, Kir6.1, and SUR2B l protein levels in renal cortical microvessels. A, representative Western blots. B, intensometric analysis. n=5–7 rats/group.

Discussion

Accumulating evidence suggests that oxidative stress plays a role in a variety of vascular and renal complications of T1D. The present study explored the possibility that oxidative stress contributes to the dysregulation of K⁺ channel-dependent control of afferent arteriolar tone that occurs during the early stage of T1D in the rat. Animals were studied three weeks after onset of STZ-induced T1D, during which partial insulin replacement yielded a condition of moderate hyperglycemia. The STZ rats demonstrated the polydipsia and polyuria characteristic of this model, as well as an increase in basal afferent arteriolar lumen diameter that is generally considered to contribute to diabetic hyperfiltration.^8,15,16 We found that chronic tempol treatment prevented the T1D-induced increase in basal afferent arteriolar diameter, but did not affect this parameter in kidneys from non-diabetic rats. In addition, the effect of T1D to exaggerate afferent arteriolar constrictor responses to blockade of various K_IR channel family members was prevented by tempol treatment. Tempol also prevented increased DHE staining in renal cortical vessels from kidneys of STZ rats. To the extent that the effects of tempol represent its antioxidant actions, these observations implicate oxidative stress as a significant factor underlying the renal microvascular dysfunction that arises during the early stage of T1D in the rat, acting at least in part through effects on K⁺ channel regulation of arteriolar tone.

In normal rat kidney, voltage-gated K⁺ channels (K_V channels), large-conductance Ca²⁺-activated K⁺ channels (BK_Ca channels) and Ba²⁺-sensitive K_IR channels contribute a tonic dilator influence on afferent arteriolar tone.⁸ In kidneys from STZ rats, members of the K_IR channel family (but not K_V, BK_Ca or small-conductance Ca²⁺-activated K⁺ channels) exert an enhanced dilator impact on afferent arteriolar tone, and pharmacological agents that block various K_IR channels reduce afferent arteriolar diameter to normal values.^7,8 The multiple members of the K_IR channel family differ with regard to electrophysiological and pharmacological properties.¹⁷ Using real time RT-PCR, we detected mRNA encoding Kir2.1 (the most prominent vascular K_IR channel), Kir1.1 (the ROMK channel) and Kir6.1 (pore-forming subunit of the vascular K_ATP channel) in PVSMCs. Notably, we were unable to detect in PVSMCs mRNA encoding Kir3.1 (GIRK1) or Kir3.4 (GIRK4), both of which can be inhibited by TPQ. Western blotting confirmed the presence of Kir2.1, Kir1.1, Kir6.1 and SUR2B at the protein level in cortical microvessels. Thus, it is likely that the afferent arteriolar effects of Ba²⁺ reflect inhibition of Kir2.1 (and, perhaps, other K_IR family members, which exhibit varying Ba²⁺ sensitivities), those of TPQ reflect inhibition of Kir1.1, and those of glibenclamide reflect inhibition of the vascular K_ATP channel (Kir6.1/SUR2B) in PVSMCs. Similar to our previous observations,^7,8 exaggerated afferent arteriolar contractile responses to Ba²⁺, TPQ and glibenclamide were observed in kidneys from STZ rats, compared with Sham rats, thereby implicating Kir2.1, Kir1.1 and K_ATP channels in the afferent arteriolar dilation accompanying T1D. Importantly, chronic tempol treatment prevented this phenomenon, normalizing arteriolar diameter and contractile responsiveness to Ba²⁺, TPQ and glibenclamide. Evaluation of mRNA expression and K⁺ channel protein levels did not reveal any differences between tempol- and vehicle-treatments in either rat group. Therefore, the presented evidence indicates that neither T1D nor high levels of glucose per se alters channel expression in the preglomerular microvasculature. Rather, the altered K⁺ channel regulation of afferent arteriolar tone evident in these experiments more likely reflects post-translational modulation of channel function.

It is well documented that tempol can act as a superoxide scavenger as well as exerting other antioxidant effects, i.e. by functioning as a SOD-mimetic, increasing endogenous SOD activity, exerting a catalase-like action, and reducing NADPH-dependent superoxide generation.^18,19 Tempol is generally administered to rats as 1–2 mmol/L in drinking water, resulting in a daily dose of ~30–60 mmol/day.¹⁹ In our attempt to choose a concentration of tempol to provide for STZ+Tempol rats, we underestimated the extent of the polydipsia that these animals would experience; however, the resulting doses administered in Sham+Tempol and STZ+Tempol groups fall within the range (30–180 mmol/day) at which tempol is maximally effective in exerting its antioxidant effects without apparent dose-dependency.¹⁸ Within this dosage range, the effects of tempol are almost exclusively attributed to its antioxidant actions,^18,19 although a few studies have found that tempol has effects that are not mimicked by other antioxidants such as tiron.^20,21 Tempol has also been reported to exert beneficial effects without altering systemic indices of oxidative stress such as plasma F₂-isoprostane levels.²² Similarly, we found no effect of tempol on the T1D-induced increase in plasma TBARS or TBARS excretion, indices of systemic and renal lipid peroxidation. Some component of the sustained elevation in TBARS evident in STZ+Tempol rats may reflect the ability of H₂O₂ to promote lipid peroxidation via the Fenton reaction, as tempol did not decrease plasma H₂O₂ levels and only blunted (but did not completely prevent) the increase in H₂O₂ excretion accompanying T1D. Indeed, tempol exerts complex renal effects in STZ rats, decreasing renal cortical NAD(P)H oxidase activity^3,4 and reducing glomerular superoxide and H₂O₂ levels, while not altering the increased tubular H₂O₂ or the elevated renal HOCl levels.³ Importantly, in the present study that focused on renal vascular function, tempol prevented the increase in vascular DHE staining (an indicator of superoxide generation in situ) in kidneys from STZ rats. Thus, although we cannot rule out any off-target effects of tempol, the beneficial impact of this agent on K⁺ channel control of afferent arteriolar tone likely reflects its antioxidant action exerted in the T1D-induced state of renal cortical oxidative stress.^5,23,24

The mechanism through which the pro-oxidant milieu accompanying T1D alters function of these K_IR channel family members is not known. Among the three channels targeted by this study, the literature provides the strongest evidence for ROS modulation of K_ATP channel function. H₂O₂ and peroxynitrite dilate cerebral arterioles by activating K_ATP channels,²⁵ and peroxynitrite hyperpolarizes and relaxes smooth muscle in rabbit internal carotid artery (but not in the common carotid artery) through K_ATP channel activation.²⁶ Superoxide has been reported to variably inhibit or activate K_ATP channels, depending on the vascular bed studied.⁹ ROS could influence K_ATP channel function by altering ATP levels in the cell or by acting directly on either the Kir6.1 or the SUR2B subunits of the channel. Recent studies of stably expressed Kir6.1 in HEK-293 cells (in the absence of any SUR subunit) revealed that co-application of hypoxanthine and xanthine oxidase produced a significant increase in Kir6.1 currents that could be abolished by tempol.²⁷ Thus, superoxide or other ROS might act directly on Kir6.1 to increase the contribution of K_ATP channels to afferent arteriolar tone during T1D. Such a scenario might also contribute to the exaggerated impact of Ba²⁺-sensitive Kir2.1 channels on afferent arteriolar tone during diabetes. These channels are generally activated by moderate increases in extracellular K⁺ concentration and, indeed, are responsible for K⁺-mediated vasodilation.²⁸ While we are unaware of any reports that ROS directly modulate function of this channel, a ROS-induced increase in K_ATP channel activity may create a microenvironment of high extracellular [K⁺] sufficient to provoke Kir2.1 activation. This scenario might explain our previous observation that exposure to 20 mmol/L K⁺ does not further dilate afferent arterioles from STZ rats.⁸ Thus, ROS-stimulated K_ATP channel activity and subsequent Kir6.1 activation could contribute to vascular smooth muscle membrane depolarization, reduced Ca²⁺ influx, and vasodilation. Obviously, more studies are needed to investigate the validity of this scenario.

The mechanism through which ROS might stimulate a Kir1.1-dependent influence on afferent arteriolar tone during T1D also remains speculative. On one hand, superoxide reduces Kir1.1 activity in renal epithelial cells via a Src kinase-dependent mechanism that results in channel endocytosis.²⁹ This scenario is directionally opposite of that necessary to underlie our observations. On the other hand, in pulmonary artery smooth muscle cells, H₂O₂ increases expression of the stress-responsive serum and glucocorticoid-inducible kinase Sgk1,³⁰ which is a well-documented activator of Kir1.1 in renal epithelial cells.²⁹ H₂O₂ excretion is markedly increased in our model of STZ-induced T1D;¹⁵ hence, a Sgk1-mediated process may contribute to the ROS-dependent increase in the Kir1.1 contribution to afferent arteriolar dilation during T1D. However, there is no information available regarding whether or not regulation of Kir1.1 channels in afferent arteriolar vascular smooth muscle and renal epithelial cells occurs via similar mechanisms.

Finally, it seems prudent to note the potential interplay between the phenomena unveiled by the present study (i.e. that oxidative stress promotes afferent arteriolar dilation during T1D via effects on K⁺ channels) and evidence that oxidative stress during T1D diminishes afferent arteriolar NO bioavailability, reducing agonist-induced NO-dependent dilation as well as the tonic dilator impact of endogenous NO (which should promote afferent arteriolar contraction).^6,31 Undoubtedly, control of afferent arteriolar tone involves both K⁺ channel-dependent regulation of membrane potential as well as the tonic influence of NO (which does not generally evoke hyperpolarization), with the net effect of these processes likely determining dilator tone. Thus, the decline in NO bioavailability during T1D may actually blunt the dilator impact of increased K_IR channel activation during this state of oxidative stress. Further study is required to unravel the interplay between these processes in control of afferent arteriolar function during T1D.

Perspectives

Diabetes is the leading cause of end-stage renal disease,³² making it increasingly important to unravel the complex mechanisms through which T1D impairs renal function. Evidence from this study and others^3,4,6 suggest that the antioxidant tempol may be useful for preventing the renal damage thought to contribute to the progression of diabetic nephropathy. However, the effects of tempol are complex and not uniformly exerted within the kidney, or systemically, and it has not yet been possible to discern the identity of the offending oxidant and its specific impact on various K_IR channel family members in the renal microvasculature. Continued investigation in this field may unveil targeted antioxidant approaches for preventing the renal complications of T1D.