Adaptive responses of rat descending vasa recta to ischemia

Abstract

tested whether rat descending vasa recta (DVR) undergo regulatory adaptations after the kidney is exposed to ischemia. Left kidneys (LK) were subjected to 30-min renal artery cross clamp. After 48 h, the postischemic LK and contralateral right kidney (RK) were harvested for study. When compared with DVR isolated from either sham-operated LK or the contralateral RK, postischemic LK DVR markedly increased their NO generation. The selective inducible NOS (iNOS) inhibitor 1400W blocked the NO response. Immunoblots from outer medullary homogenates showed a parallel 2.6-fold increase in iNOS expression (P = 0.01). Microperfused postischemic LK DVR exposed to angiotensin II (ANG II, 10 nM), constricted less than those from the contralateral RK, and constricted more when exposed to 1400W (10 µM). Resting membrane potentials of pericytes from postischemic LK DVR pericytes were hyperpolarized relative to contralateral RK pericytes (62.0 ± 1.6 vs. 51.8 ± 2.2 mV, respectively, P < 0.05) or those from sham-operated LK (54.9 ± 2.1 mV, P < 0.05). Blockade of NO generation with 1400W did not repolarize postischemic pericytes (62.5 ± 1.4 vs. 61.1 ± 3.4 mV); however, control pericytes were hyperpolarized by exposure to NO donation from S-nitroso-N-acetyl-dl-penicillamine (51.5 ± 2.9 to 62.1 ± 1.4 mV, P < 0.05). We conclude that postischemic adaptations intrinsic to the DVR wall occur after ischemia. A rise in 1400W sensitive NO generation and iNOS expression occurs that is associated with diminished contractile responses to ANG II. Pericyte hyperpolarization occurs that is not explained by the rise in ambient NO generation within the DVR wall.

INTRODUCTION

Juxtamedullary efferent arterioles form descending vasa recta (DVR) that traverse renal outer medullary vascular bundles to distribute blood flow (2, 30, 36). DVR are capillary-sized microvessels, ~10 to 15 µm in diameter, with contractile properties imparted by abluminal pericytes (28, 37, 42, 45). Vascular bundles have a unique niche in the physiology of the kidney and the maintenance of its oxygenation. Countercurrent diffusion of oxygen from DVR to ascending vasa recta reduces medullary oxygen tensions to low values (3–5). The outer medullary interbundle region that surrounds vascular bundles harbors the thick ascending limbs of Henle, pars recta, and collecting ducts that transport sodium, consume O₂, and are presumably vulnerable to hypoxic insults (1, 4, 5, 25, 26). The release of paracrine vasodilatory agents such as NO, adenosine, and prostaglandins may play important roles to limit DVR contraction and provide a defense against medullary infarction (10, 31, 43).

As with other organs, exposure of the kidney to ischemia can invoke adaptive responses that may serve to mitigate injury from recurrent insults (1, 15, 39, 40). A transient episode of ischemia followed by reperfusion (ischemia-reperfusion) has been used as an experimental maneuver to invoke and study such adaptation. Adaptive processes include modification of NO generation, regulation of nitric oxide synthase (NOS) isoforms, and formation of endogenous antioxidants. Modification of immune responses may also occur (3, 4).

Despite a longstanding focus on renal medulla as a vulnerable site for injury (6, 22, 23, 27), the intrinsic capacity for adaptation within the medullary vasculature itself is unexplored. We performed initial experiments to test whether an episode of ischemia leads to adaptive modifications that favor blunting of further vasoconstriction. Specifically, we applied 30 min of left renal artery cross clamp, followed by reperfusion. Subsequently, the left and right kidneys were harvested to measure microvessel properties. DVR dramatically increased NO release that was fully inhibited by the iNOS blocker 1400W. DVR pericytes became hyperpolarized, and isolated DVR constriction to angiotensin II application was limited. These observations demonstrate an intrinsic capacity of DVR pericytes and endothelia to become modified in ways that are likely to limit pericyte contraction and defend the medulla from ischemic injury.

METHODS

Ischemia-reperfusion.

Investigations were performed according to protocols approved by the Institutional Animal Care and Use Committee of the University of Maryland. Transient renal ischemia or sham surgery was performed by accessing kidneys of male Sprague-Dawley rats (150–200 g) via flank incision under ketamine/xylazine (80 mg/kg and 10 mg/kg, respectively) anesthesia. The left renal artery was occluded with a microclamp for 30 min. Sham treatments were performed in contemporaneous controls by exposing and manipulating the renal artery but not closing the clamp. After surgery, wounds were sutured, and the rats were returned to warmed cages for recovery. Buprenorphine was administered to provide analgesia. After 48 h, the postischemic and sham-treated rats were anesthetized to harvest the right and left kidneys.

Isolation of DVR.

Kidneys were harvested, sliced, and stored at 4°C in a physiological saline solution (PSS, in mM) NaCl 145, KCl 5, MgCl₂ 1, CaCl₂ 1, N-2-hydroxyethylpiperazine-N′-2-ethanesulfonic acid (HEPES) 10, glucose 10, pH 7.4 at room temperature. For experiments involving electrophysiological recording, wedges of renal tissue were digested at 37°C for 22 min in PSS containing a cocktail of collagenase 1A (0.5 mg/ml), protease XIV (0.4 mg/ml), and bovine serum albumin (1.0 mg/ml). For microperfusion, enzymatic digestion was omitted. In all cases, tissue was stored in PSS in a petri dish chilled to ~4°C on ice. DVR were explanted by hand dissection from outer medullary vascular bundles. Vascular bundles were recognized as parallel accumulations of vessels with clear lumens, largely red blood cell free, lying between the red blood cell-congested interbundle capillary plexus of inner stripe tissue. Isolated DVR are generally 500 to 1,000 µm long and 12–15 µm outer diameter as previously described (32, 44). For microperfusion, DVR were transferred to the stage of an inverted microscope and cannulated so that contractile responses could be quantified by videomicroscopy. For measurement of NO generation, the rate of accumulation of 4,5-diaminofluorescein diacetate (DAF-2 DA) fluorescence was quantified (41). Finally, pericyte membrane potentials were recorded by whole cell perforated patch clamp recording (34, 51).

Whole cell patch clamp recording.

Patch clamp pipette electrodes were made from borosilicate glass capillary tubing (1.5-mm outer diameter, 1.0-mm inner diameter; PG52151-4, World Precision Instruments). The electrodes were formed using a two-stage vertical puller (Narshige PP-830) and then heat polished to a final orifice of ~1 µm. Whole cell electrical access to the cytoplasm was achieved with nystatin as the pore-forming agent in the electrode buffer (33, 34). The electrode buffer was (in mmol/l): Potassium aspartate 120, KCl 20, NaCl 10, HEPES 10, pH 7.2, with nystatin (100 μg/ml, 0.1% DMSO) in ultrapure water. The extracellular buffer was PSS, as defined above. Recordings were performed at room temperature with a Multiclamp 700B Amplifier (Molecular Devices). Membrane potential was recorded at 10 Hz by zero current clamp and corrected for junction potentials.

Vasoactivity measurements.

To measure vessel contraction, DVR were cannulated with concentric pipettes, and their lumens opened by microperfusion. As described, vessel diameter was recorded with videomicroscopy and the inner diameter quantified off-line (32, 35). During microperfusion, DVR were positioned with micromanipulators (Instruments Technology and Machinery, San Antonio, TX) near a thermocouple in the entrance region of a custom-built chamber where the local temperature was maintained at 37°C with a feedback controller (CN9000A, Omega Engineering). Images of microperfused vessels were captured with a digital camera (Spot Camera, Diagnostic Instruments) by using a ×40 objective to yield final magnification of ~1,300. The luminal constriction from baseline was quantified by image analysis (National Institutes of Health, ImageJ software) as % reduction of internal diameter. The latter was calculated as [1 – (D/D0)] × 100, where D and D0 are the experimental and baseline diameters, respectively.

Fluorescent detection of NO generation with DAF-2 DA.

4,5-Diaminofluorescein diacetate (DAF-2 DA, 10 µMol/l) was loaded by incubation of the acetoxymethyl ester for 20 min. The probe reacts with NO to form a fluorescent product that accumulates within cells. We have previously shown that global loading of the vessel wall produces a combined signal from pericytes and endothelia reflecting NO accumulation (41). DAF-2 DA emission was quantified at 530 nm during excitation at 485 nm by using a photon counting photomultiplier assembly (D104B, Photon Technology International).

Immunoblotting.

The renal tissue slices were divided into outer and inner medulla homogenized (Polytron) in chilled lysis buffer containing 10 mmol/l Tris·HCl (pH 7.5), 2 mmol/l EDTA, 1 mmol/l sodium vanodate, 1% SDS, and 20 mmol/l N-ethylmaleimide. Protein concentrations were measured using a BCA Protein Assay Reagent Kit (Bio-Rad). All samples were stored at a protein concentration of 1–3 mg/ml and solubilized at 95°C for 5 min in Laemmli sample buffer. Twenty micrograms of protein per sample were loaded onto individual lanes, separated by electrophoresis on 4–15% polyacrylamide gels and electrophoretically transferred to polyvinylidene difluoride membranes. After blocking with 5% milk for 1 h, membranes were probed overnight at 4°C with the appropriate primary antibody. The secondary antibody was conjugated to horseradish peroxidase (Abcam). Sites of antibody-antigen reaction were visualized by using luminol-based chemiluminescence (Amersham) and exposure to X-ray film. These methods have been described in detail (29).

Reagents.

DAF-2 DA was obtained from Calbiochem. Nystatin, angiotensin II, collagenase 1A, protease XIV, N-[[3-(aminomethyl) phenyl]methyl]-ethanimidamide (1400W), S-nitroso-N-acetyl-dl-penicillamine (SNAP), and other chemicals were purchased from Sigma (St. Louis, MO). SNAP was prepared fresh daily. Other reagents were thawed and diluted to final concentration on the day of the experiment. Any excesses were discarded daily.

Statistics.

Data in the text and figures are reported as means ± SE. When more than one vessel was isolated and tested from one animal, the results were averaged. The n reported with individual experiments represents the number of animals tested. The significance of differences was evaluated with SigmaStat 3.11 (Systat Software, Point Richmond, CA). Comparisons between groups were performed with unpaired Student’s t-test. Repeated-measures ANOVA was used to analyze testing of effects of addition and removal of iNOS inhibition or NO donation on membrane potential with 1400W and SNAP washout studies. Post hoc comparisons employed Tukey’s test, P < 0.05 to reject the null hypothesis.

RESULTS

Weight, kidney size, and DVR diameter.

The means ± SE of body weight, kidney weight, and resting DVR internal diameter are shown in Figs. 1, A–C, respectively. When compared with sham-operated controls, neither body weight nor the weight of the nonischemic right kidney was affected by prior ischemia to the left kidney. However, after 48 h, the ischemia-treated left kidney, when compared with the contralateral nonischemic right kidney from the same rat, increased in weight (209 ± 15 vs. 189 ± 10 g, P < 0.05) (Fig. 1B). The resting internal diameter of microperfused DVR from the ischemia-treated left kidney, however, did not change (Fig. 1C).

Fig. 1.

Kidney size and resting descending vasa recta (DVR) diameters. A: body weight of sham-operated (n = 5) and ischemia-treated (n = 13) rats at the time of kidney harvest (N.S., not significant). B: left kidney (LK) and right kidney (RK) weights from sham (n = 5) and ischemia-treated (n = 13) rats at the time of harvest (**P < 0.01, LK vs. RK of ischemia-treated rats). C: resting microperfused DVR internal diameters (I.D., micrometers) from the right and left kidneys of ischemia-treated rats (n = 6).

NO generation by DVR increases after ischemia.

Ischemia-reperfusion has been shown to increase intrarenal NOS isoform expressions that can persist for days to weeks (39). We tested the hypothesis that an increase in NO generation occurs within the DVR wall, per se, as part of the process. We measured tonic NO generation by DVR harvested from left kidneys (LK) of sham-operated or ischemia-treated rats. The rate of accumulation of DAF-2 DA fluorescence in the DVR wall from the ischemic kidneys increased to a remarkable degree (Fig. 2). The ability of transient ischemia at one site to confer remote protection at a distant site has been frequently described (4, 9, 24). In our hands, however, left renal ischemia did not confer a significant rise in the generation of NO by DVR of the contralateral right kidney (RK) of the same rat (Fig. 3).

Fig. 2.

NO generation by DVR from left ischemia-reperfused (I/R) kidney is elevated. NO generation of DVR from the left kidneys of rats previously subjected to 30-min cross clamp (ischemia, LK, n = 7) or sham manipulation (sham, LK, n = 6) of the left renal artery. NO was quantified by measuring the fluorescence of the NO probe 4,5-diaminofluorescein diacetate (DAF-2 DA). NO generation was greater in ischemia-treated vessels (P < 0.05, I/R vs. Sham for all time > 0.76 min and P < 0.01 for all time > 3.6 min).

Fig. 3.

NO generation by DVR from ischemia-treated left and nonischemic right kidneys. NO generation was quantified in DVR from the left (LK) and right (RK) kidneys of rats (n = 7) previously subjected to 30-min ischemia by cross clamp of the left renal artery. NO was quantified by measuring the rise in fluorescence of the NO probe DAF-2 DA. NO generation was higher in vessels from the LK than the RK (P < 0.05 for all time > 11.3 min, P < 0.01 for time > 12.7 min).

A role for enhanced NOS2 (iNOS) expression in adaptive preconditioning of the kidney has previously been identified (40). In keeping with that observation, the large generation of NO derived from previously ischemic DVR was fully inhibited by iNOS blockade with N-[[3-(aminomethyl)phenyl]methyl]-ethanimidamide (1400W, 10 μM, Fig. 4). This observation contrasts with our prior studies, wherein iNOS played a negligible role in basal tonic release of NO from the DVR wall in untreated rats (7). Hence, NOS2 appears to be upregulated and activated in postischemic DVR. We also confirmed the observations of others (39) that iNOS expression was globally upregulated in left kidney outer medullary homogenates (Fig. 5).

Fig. 4.

Inhibition NO generation by 1400W. NO generation of DVR from the left kidneys of rats previously subjected to 30-min cross clamp of the renal artery were exposed to either vehicle (n = 7) or the inducible NOS (iNOS) inhibitor 1400W (n = 7, 10 μM). NO was quantified by measuring the rise in fluorescence of the NO probe DAF-2 DA. NO generation was prevented by 1400W (P < 0.01 for all time > 3 min).

Fig. 5.

Postischemic enhancement of iNOS expression. Expressions of iNOS and β-actin were quantified from homogenates prepared from the outer medullas of right (RK) and left kidneys (LK) of rats (n = 5) that had been previously subjected to 30-min ischemia by cross clamp of the left renal artery. Expression of iNOS, normalized to β-actin, was markedly greater in the left kidney (*P < 0.01).

DVR vasoconstriction.

To determine if postischemic adaptations of DVR affected their responsiveness to agonist stimulation, we measured ANG II (10 nM)-induced contractions. After recording baseline internal diameter, ANG II was introduced into the extracellular buffer for 5 min and then removed. ANG II vasoconstriction was blunted in DVR from the left kidneys (Fig. 6). In a past study, we showed that acute, nonselective blockade of all NOS isoforms with nitro-l-arginine methyl ester yielded potent vasoconstriction of explanted DVR but that the selective blockade of iNOS with 1400W had little effect (7). In contrast, blockade of iNOS with 1400W induced significant contraction of DVR from ischemia-treated left kidney compared with the untreated right kidney of the same animal (Fig. 7).

Fig. 6.

Angiotensin II (ANG II) constriction of postischemic DVR is reduced. Microperfused DVR internal diameters were quantified by videomicroscopy. Percent reduction of luminal diameter vs. Time is shown at baseline, during ANG II (10 nM) exposure and after washout. DVR from the left postischemic left kidney (LK) constricted less than those from the right kidney (RK) (*P < 0.05; **P < 0.01; n = 6 each).

Fig. 7.

Constriction of postischemic DVR during iNOS inhibition. Microperfused DVR internal diameters were quantified by videomicroscopy. Percent reduction of luminal diameter vs. Time is shown at baseline, during 1400W (10 µM) exposure, and after washout. DVR from the left postischemic kidney constricted to a greater degree than those from the right kidney (*P < 0.05; **P < 0.01; n = 8 each).

Pericyte membrane potential.

We have previously shown that both calcium (CaV) and sodium (NaV) ion channels are present in DVR pericytes that are regulated by membrane potential (29, 47). The contractile machinery of vascular smooth muscle cells is regulated by CaV-mediated calcium entry and hence by membrane potential. Thus hyperpolarization favors vasodilation (11). In postischemic left kidneys, the DVR pericytes were significantly hyperpolarized relative to those from the contralateral right kidney (62.0 ± 1.6 vs. 51.8 ± 2.2 mV, P < 0.05, Fig. 8A). Sham manipulation of the left renal artery did not hyperpolarize membrane potentials (54.9 ± 2.1 mV, Fig. 8B).

Fig. 8.

Hyperpolarization of DVR pericyte membrane potential after ischemia. A: resting membrane potential was measured in DVR pericytes from rats previously subjected to 30-min ischemia of the left renal artery. Pericytes from postischemic left kidney (LK) were hyperpolarized relative to pericytes from the nonischemic right kidney (RK; **P < 0.01; n = 8 each). B: resting membrane potential was measured in DVR pericytes from right kidney (RK) and left kidney (LK) of rats subjected to sham surgery. Sham manipulation did not affect DVR pericyte resting potential (N.S., not significant; n = 6 each).

NO can inhibit vessel contraction by reducing the sensitivity of the contractile apparatus to calcium ions and by regulating the channel activities that govern membrane potential. To follow up the observation that 1400W eliminated the increase in NO generation in postischemic DVR (Fig. 4), we tested whether it reversed membrane potential changes of ischemia-treated pericytes. Despite the exquisite sensitivity of NO generation to 1400W, it did not repolarize pericytes to control levels (Fig. 9A). In separate experiments we showed that NO donation from SNAP (10 µM) does hyperpolarize control DVR pericytes (Fig. 9B), confirming that acute NO-dependent regulation of membrane potential is a property of untreated DVR pericytes. Thus the large tonic rise in NO generation after ischemia (Figs. 2 and 3) does not underlie the concomitant membrane potential hyperpolarization show in Fig. 8.

Fig. 9.

NO generation and ischemia-induced hyperpolarization. A: membrane potential of DVR pericytes from the ischemia-treated left kidney (LK) and nonischemic right kidney (RK) was measured at baseline, during iNOS blockade with 1400W, and after washout. At baseline, LK pericytes were hyperpolarized relative to RK pericytes (*P < 0.05: n = 6 each). Blockade of iNOS with 1400W did not alter membrane potential in either group (N.S., not significant). B: membrane potentials of DVR pericytes from left kidney DVR of sham-operated rats were measured at baseline, during exposure to the NO donor SNAP (10 µM), and after washout. SNAP consistently hyperpolarized the pericytes (**P < 0.01). C and D: example membrane potential recordings during exposure to 1400W or SNAP. Values at the beginning of the traces are the initial resting potential.

DISCUSSION

DVR supply the renal medulla with blood flow. They traverse vascular bundles of the outer medulla and distribute perfusion to the outer and inner medulla (2, 30). It has been logically assumed that modification of DVR vasoconstriction could alter regional blood flow and serve as a control point for preservation of medullary integrity when oxygen delivery is compromised (6, 14, 15, 27). Several signaling mechanisms may be involved. Paracrine agents (e.g., adenosine, NO, CO, prostaglandins, H₂O₂) released by nephrons may supply the afferent components of feedback loops to vasodilate DVR and augment perfusion toward an optimal set point (10, 31, 38). Evidence has favored possible roles of NO, CO, reactive oxygen species (ROS), and cellular inflammation in that process (15, 17, 21, 24). In this study, we showed that DVR can adapt to ischemic stress, pointing to the likelihood that one component of tissue defense may be regulation of processes within the DVR wall. In our hands, transient ischemia enhanced 1400W-sensitive NO generation (Figs. 2 and 3). This might favor vasodilation to mitigate damage from repeat ischemic events. High levels of NO production might also serve to reduce sodium reabsorption by neighboring nephrons to blunt O₂ consumption (21). A beneficial role for iNOS-mediated NO generation, however, is not a foregone conclusion. At very high concentrations, NO itself may be injurious by acting as a radical that modifies cellular functions through nitrosylations (16, 18). It would be of great interest to test a model system wherein postischemic regulation in DVR can be prevented independent of modifications in other epithelial or vascular structures. Such an approach would enable testing of DVR modifications to analyze the role in survival and recovery. A means for independent control over iNOS expression in DVR is a technically elusive maneuver. It is of parallel interest that ischemia-reperfusion injury in iNOS-deficient mice can be lethal (39).

The relative roles of NO generation that occur in regions intrinsic or extrinsic to vascular bundles has been uncertain. Several observations bear on this issue. NO bioavailability can be altered through generation of ROS, particularly superoxide. Both NO and superoxide are released by thick ascending limbs of Henle that lie near vascular bundles. Those agents respectively inhibit and augment Na⁺ reabsorption so that modification of their production rates could defend against ischemia by lowering O₂ utilization. In addition, however, NO and ROS generations intrinsic to the DVR wall do modify vascular tone in isolated, explanted vessels (7). Hence the extent to which vascular bundle DVR NO formation adjusts NO within the interbundle region of the outer medulla is difficult to predict. It seems reasonable to assume that the steepness of laterally directed gradients between intra- and interbundle environments could be modified by generation within the DVR wall. Interbundle NO and superoxide levels are likely to be governed by interplay of several events: vascular synthesis, epithelial synthesis, diffusive flux, convective flux (via blood flow), and the chemical reactions between species. The result of perturbing one component is difficult to predict (12, 13).

The magnitude of the increase in DVR NO generation that occurred after ischemic conditioning (Figs. 2 and 3) is remarkable. It was entirely blocked by iNOS inhibition with 1400W. We have previously shown that basal NO generation in untreated DVR arises from endothelial constitutive NOS and neuronal NOS (7). These studies favor the conclusion that increased activity of iNOS alone accounts for postischemic changes in NO generation in DVR (Fig. 4). In our hands, the increase in NO after ischemia exceeds that resulting from any stimulus we have previously investigated, including vasoactive agonists or luminal shear (41, 49, 50). In vivo, the net direction of NO diffusive flux, toward or away from vascular bundles, might be altered or perhaps even reversed by such a large effect. Hemoglobin provides a sink for removal of NO so that orientation of NO gradients to favor diffusion toward DVR lumens seems likely under most conditions. Even in that case, however, a large rise in intrinsic DVR NO generation, such as observed herein, might increase the spatial average NO in environments surrounding vascular bundles thereby lowering intrabundle O₂ utilization by sodium reabsorbing epithelia.

It has long been frustrating that cogent in vivo experimental interrogation of the physiology of the outer medulla is nearly impossible because overlying cortex renders the region inaccessible. True luminal flows, shear, oxygen tension, and their interactions cannot be known or truly approximated to the in vivo state in explanted structures. The possibility that NO and ROS released by DVR could modify NO gradients in the vicinity of vascular bundles to alter salt transport and pressure natriuresis cannot be dismissed. Thus extrapolation of these findings to the in vivo setting needs to be cautious. Along the same lines, postischemic manipulation of ROS generation or free radial scavenging would be of interest in future studies to determine if it affects regulations of NOS isoforms or ion channel activities.

The shared membrane potential of pericytes and endothelia regulates voltage-gated calcium entry and DVR contraction (48, 50, 51). The importance of such pathways in the regulation of vascular smooth muscle of other renal arteriole segments has been well documented. Many classes of voltage-gated calcium channels (L-type, T-type, and P/Q type CaV) are present in afferent arterioles and juxtamedullary efferent arterioles throughout the cortex (20). Our studies have reinforced those findings. Electrophysiological signatures of nifedipine sensitive L-type CaV (and NaV1.3 sodium channels) are observed in DVR pericytes. L-type channel blockade blunts DVR contraction and modifies cytoplasmic calcium transients induced by vasoconstrictors and mechanical stimulation (47, 50, 52). Finally, L-type channel blockade can increase medullary blood flow, as quantified by laser-Doppler recordings performed in vivo (19, 46). Hence, pericyte membrane potential may be a pivotal regulator of calcium entry and medullary perfusion. The finding that postischemic pericytes become hyperpolarized (Figs. 8 and 9) and that DVR contraction to ANG II is blunted (Fig. 6) favors the interpretation that electrical adaptations, i.e., ion channel activities, are components of protective conditioning.

The mechanisms responsible for postischemic pericyte hyperpolarization are uncertain; however, knowledge of pericyte channel architecture points to the candidates. Pericyte membrane potential is set as a balance between depolarizing influence of calcium-dependent chloride channels (CaCC) (33, 34), hyperpolarizing influences of K⁺ channels (8, 33, 34), and hyperpolarization from electrogenic transport by Na,KATPase (8). Based on these data, we can only state that any changes in their activities are unlikely to occur downstream of NO signaling (Fig. 2) because iNOS inhibition with 1400W blocked NO generation (Fig. 4) but did not concomitantly repolarize the postischemic pericytes (Fig. 9). Determination of the responsible mechanisms that produce adaptation after ischemia will require individual measurements of channel activities. Other signaling molecules may be important. For example, products of arachidonic acid, CO, or ROS could be of prime importance. Hyperpolarizing effects of endothelia on pericytes via myo-endothelial gap junction coupling might also play a role (48, 51). Finally, the importance of electrophysiological regulation in the vasculature on renal injury and overall survival is an obvious question. We can only point out that postischemic hyperpolarization might blunt vasoconstriction to limit injury. Based on this study, however, it is overreaching to conclude that electrophysiological adaptation is a quantitatively important aspect of tissue defense.

These studies involved multiple comparisons of vascular properties between groups of rats because paired comparison with and without ischemia is not possible. We identified robust modification of iNOS expression, NO generation, contractility, and pericyte membrane potential. Concern over a type 1 statistical error seems minimal for two reasons. First, most P values <0.01 (Figs. 1–4 and 5–9). Second, results were individually reproducible in separate series of experiments (NO generation in Figs. 2, 3, 4; contractility in Figs. 6 and 7; and membrane potential in Figs. 8 and 9). This effort is a first study of postischemic adaptation within the DVR wall. As such, we focused on a single “dose” of ischemia (30-min cross clamp) and single postischemic time point (48 h). This provides a limited view of the plausible events that might occur in the vasculature after ischemia or toxic challenges. Notably, prior studies using ischemia-reperfusion showed a durable effect to upregulate NOS isoforms over periods of weeks (39). It will eventually be of interest to determine if the regulatory events in the DVR wall occur earlier and are similarly durable. Our identification of a large role for iNOS in the DVR wall parallels findings in mice which the consequences of its deletion on murine survival were severe (39).

GRANTS

These studies and personnel were supported by National Institute of Diabetes and Digestive and Kidney Diseases Grants R01 DK-042495 and R01 DK-067621, and the Baltimore Veterans Administration Medical Center.

DISCLOSURES

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

AUTHOR CONTRIBUTIONS

Z.Z. and T.L.P. conceived and designed research; Z.Z., K.P., and T.L.P. performed experiments; Z.Z., K.P., and T.L.P. analyzed data; Z.Z., K.P., and T.L.P. interpreted results of experiments; Z.Z., K.P., and T.L.P. prepared figures; Z.Z., K.P., and T.L.P. edited and revised manuscript; Z.Z., K.P., and T.L.P. approved final version of manuscript; T.L.P. drafted manuscript.