Angiotensin II–NAD(P)H Oxidase–Stimulated Superoxide Modifies Tubulovascular Nitric Oxide Cross-Talk in Renal Outer Medulla
The source of superoxide (O2·−) production and cell-to-cell interactions of O2·− and nitric oxide (NO) in response to angiotensin II (AngII) were studied by fluorescence microscopic techniques to image rat renal outer medullary microtissue strips. Changes in intracellular O2·− were determined by dihydroethidium-ethidium ratios, and NO was determined with 4,5-diaminofluorescein diacetate. AngII (1 μmol/L) significantly increased O2·− in the isolated, medullary thick ascending limb (mTAL). These responses were inhibited by the superoxide dismutase mimetic 4-hydroxytetramethylpiperidine-1-oxyl (TEMPOL) and by the NAD(P)H oxidase inhibitors diphenylene iodonium and apocynin. AngII did not increase O2·− in either pericytes of isolated, intact vasa recta (VR) or pericytes of VR with a disrupted endothelium, even when surrounded by mTAL. However, AngII did increase O2·− when the tissue strips were preincubated with the NO scavenger 2-(4-carboxyphenyl)-4,4,5,5-tetramethylimidazoline-1-oxyl-3-oxide (carboxy-PTIO), indicating that cross-talk of O2·− from mTAL to the VR occurred but was normally inhibited by NO. Also, tissue O2·− reduction by TEMPOL increased the diffusion of NO from mTAL to the pericytes, indicating that cross-talk of NO from the mTAL to the VR is also inhibited by O2·−. We conclude that AngII stimulates O2·− production in mTAL via the NAD(P)H oxidase pathway and that interactions of O2·− and NO ultimately determine the effectiveness of in situ free-radical cross-talk between the mTAL and the VR.
Oxidative stress in the kidney has been found to play an important role in the regulation of renal medullary blood flow (MBF), tubular sodium reabsorption, and the long-term control of arterial pressure.1,2 It has been shown in short-term, anesthetized-rat studies that medullary interstitial infusion of the superoxide dismutase (SOD) inhibitor diethyldithiocarbamic acid (DETC) reduces MBF and sodium excretion without a change in blood pressure.1 Opposite changes were observed with medullary interstitial infusion of the cell-permeable SOD mimetic 4-hydroxytetramethylpiperidine-1-oxyl (TEMPOL).1 Consistent with these observations, long-term, medullary interstitial infusion of DETC led to a reduction of MBF and produced hypertension in the absence of changes in cortical blood flow.2 Because the same dose of DETC when given intravenously neither reduced MBF nor changed arterial pressure, that study demonstrated that an increase in oxidative stress, specifically within the renal medulla, would result in hypertension.
Techniques for fluorescence imaging of both nitric oxide (NO) and Ca2+ in the vasa recta (VR) and tubules of microtissue strips obtained from the medulla were developed recently in our laboratory.3,4 We have demonstrated that NO serves as a paracrine substance that mediates cross-talk between the tubular epithelium of the medullary thick ascending limb (mTAL) and contractile VR pericytes.4 It was shown that angiotensin II (AngII) could not directly increase intracellular NO levels in either endothelial cells or pericytes of the outer medullary VR. Rather, AngII stimulated NO release from the mTAL that diffused to the adjacent pericytes of the VR.4 This “tubulovascular cross-talk” demonstrated functional coupling of the vascular and tubular units and explained how regional elevations of AngII could, by the release of NO, result in both buffering of AngII-mediated sodium reabsorption and VR constrictor effects.5
Because AngII can also stimulate production of superoxide (O2·−) in a variety of cell types,6–8 we hypothesized that if this occurred in the outer medulla, then the release of O2·− might be able to directly influence medullary tubular function and also modify the effects of NO cross-talk between tubular and vascular structures to regulate MBF. Therefore, intracellular O2·− and NO responses to agonists were measured by using the fluorescent dyes dihydroethidium (DHE) and 4,5-diaminofluorescein diacetate (DAF-2DA), respectively. The VR and tubular elements were studied within the context of normal morphological relationships and in isolation to determine the source of the observed changes in these free radicals.
Tissue Fluorescence Imaging
Renal microtissue strips were dissected from the outer medulla of the left kidney of male, pentobarbital-anesthetized (60 mg/kg IP), Sprague-Dawley rats (170 to 230 g; Harlan, Madison, Wis), and fluorescence imaging techniques were applied as reported previously.3,4 Left kidneys were removed and cleared of blood by perfusion with Hank’s balanced salt solution (Life Technologies) with 20 mmol/L HEPES (HBSSH; adjusted to pH 7.4; Sigma) and 1 mg/mL bovine serum albumin. A latex-microsphere solution (2.7% wt/vol, 0.2-μm diameter; Polysciences) was perfused in some rats to disrupt the endothelial cells of the VR4 so that changes in pericyte free radicals could be measured. A single layer of VR with surrounding mTAL or each structure alone was then carefully removed by stripping off a thin tissue layer to preserve the natural morphology and then placed on round, glass coverslips coated with tissue adhesive (Cell-tak, BD Biosciences) within 60 minutes after removal of the kidney.
l-Arginine (Arg, 100 μmol/L; Sigma) was added to HBSSH for free-radical measurements, and coverslips were loaded with either DHE (50 μmol/L in HBSSH-Arg; Molecular Probes) or DAF-2DA (10 μmol/L in HBSSH; Calbiochem-Novabiochem) for 1 hour at room temperature for intracellular O2·− or NO measurements, respectively; the coverslips were then washed twice to remove excess dye. Tissues were incubated for 30 minutes under 1 of the following conditions: (1) HBSSH-Arg alone; (2) HBSSH-Arg with 1 mmol/L of the SOD mimetic TEMPOL (Sigma); (3) HBSSH-Arg with 10 μmol/L of the flavoprotein inhibitor diphenylene iodonium (DPI, Sigma); (4) HBSSH-Arg with 1 mmol/L of the NADH oxidase inhibitor apocynin, which inhibits translocation of the p47 and p67 subunits of NAD(P)H oxidase9 (Sigma); or (5) HBSSH-Arg with 10 μmol/L of the NO scavenger 2-(4-carboxyphenyl)-4,4,5,5-tetramethylimidazoline-1-oxyl-3-oxide (carboxy-PTIO, Dojindo).
DHE conversion to ethidium (Eth) was used to determine intracellular O2·− levels, with DHE detected separately10 by using an excitation wavelength of 380 nm and an emission wavelength of 445 nm; Eth was excited at a 480-nm wavelength, and signals were collected through a 605-nm filter. DAF-2DA was excited at a 480-nm wavelength, and signals were collected through a 535-nm filter at 3-second intervals. Representative Eth signals obtained from a microtissue strip from endothelial cell–disrupted VR with associated TAL are shown in Figure 1.
O2·− and NO signals of interest were imaged in response to superfusion of the tissue strips with the drug vehicle, AngII (1 μmol/L, Sigma), or TEMPOL (1 mmol/L). Measurements of agonist-stimulated responses were followed by addition of either 1 mmol/L DETC (Sigma) to inhibit SOD together with 500 μmol/L menadione sodium bisulfite (Sigma)10 to stimulate mitochondrial O2·− release or the NO donor, diethylenetriaamine NONOate (DETA-NONOate, 100 μmol/L, Cayman). Together, these stimuli served as positive controls for dye loading and cell viability.3,4 Background fluorescence was also measured simultaneously to confirm that responses were specific to intracellular O2·−. Changes in O2·− levels were expressed as the ratio of Eth-DHE fluorescence values. NO levels were expressed as changes in fluorescence units in responses to agonists. Tissue strips were analyzed with imaging software (MetaFluor, Universal Imaging).3,4
Values are expressed as mean±SE (P<0.05). Responses were evaluated with a 1-way ANOVA, with a Dunn multiple-range test for O2·− and a Dunnet test for NO responses. An unpaired t test was used to compare drug vehicle and agonist responses at 250 or 300 seconds. All protocols were approved by the Institutional Animal Care Committee.
O2·− Responses in Epithelial Cells of Isolated mTAL
Application of AngII (1 μmol/L) to isolated mTAL increased the Eth-DHE fluorescence ratio within epithelial cells, which reached a level of significance (P<0.05, n=6) at 230 seconds after stimulation (Figure 2A). The response was completely inhibited by preincubation with 1 mmol/L TEMPOL. As shown in Figure 2B, pretreatment with either of the NAD(P)H oxidase inhibitors DPI (10 μmol/L) or apocynin (1 mmol/L) inhibited AngII responses. Values averaged 0.01±0.02 (P<0.01, n=5) with DPI pretreatment and 0.09±0.07 (P<0.01, n=5) with apocynin at 250 seconds. These values contrasted with average levels of 0.50±0.08 (n=6) in the absence of the inhibitors, indicating that NAD(P)H oxidase was responsible for AngII-induced O2·− production in mTAL. Drug vehicle and positive control responses (1 mmol/L DETC with 500 μmol/L menadione) did not differ significantly between groups. This suggests that the NAD(P)H oxidase inhibitors did not produce nonspecific Eth/DHE fluorescence inhibition. The increase of O2·− levels in mTAL with AngII stimulation tended to be less than the drug vehicle responses when tissues were preincubated with TEMPOL, DPI, or apocynin. Although these results could indicate that there is a baseline production of O2·− in mTAL by NAD(P)H oxidase, they could also suggest that NO was coproduced in mTAL4 and could have interacted with O2·− to reduce O2·− levels further than the vehicle responses.
Evidence of O2·− Cross-Talk Between mTAL and VR Pericytes
To determine whether O2·− was produced in the pericytes of VR and whether O2·− from surrounding endothelial cells of the VR or epithelial cells of mTAL could reach the pericytes when stimulated with AngII, O2·− responses to AngII were studied in the microtissue strip under the following 3 conditions: (1) isolated VR, with the endothelium disrupted by microsphere perfusion to measure O2·− levels in pericytes alone; (2) isolated VR with intact endothelium to measure O2·− levels in pericytes attached to intact VR endothelium; and (3) VR with disrupted endothelium in the presence of surrounding mTAL to measure O2·− levels in pericytes attached to the mTAL. As shown in Figure 3, under none of these conditions was a response to AngII measured. Pericytes alone averaged 0.08±0.00 (n=6), pericytes attached to intact VR endothelium averaged 0.10±0.01 (n=5), and pericytes with mTAL attached averaged 0.08±0.03 (n=6) at 250 seconds. Drug vehicle O2·− responses of pericytes alone averaged 0.09±0.02 (n=6), pericytes attached to intact VR endothelium averaged 0.11±0.04 (n=5), and pericytes with surrounding mTAL averaged 0.11±0.03 (n=6). In contrast, the positive control of DETC (1 mmol/L) with 500 μmol/L menadione significantly increased O2·− in pericytes under all of these conditions. Pericytes alone averaged 0.38±0.09 (P<0.01, n=6), pericytes attached to intact VR endothelium averaged 0.50±0.06 (P<0.01, n=5), and pericytes with mTAL attached averaged 0.56±0.12 (P<0.01, n=6), indicating that those pericytes were capable of producing O2·−. Although pericytes attached to the intact VR endothelium or tubular epithelial cells of mTAL tended to exhibit higher O2·− levels in response to DETC with menadione than did isolated VR pericytes, these responses did not reach a level of significance. These data therefore indicate that the O2·− produced in mTAL in response to AngII does not normally diffuse into pericytes.
Despite these results, evidence was found to indicate that O2·− cross-talk can occur between the mTAL and VR pericytes under certain circumstances. In microtissue strips preincubated with the NO scavenger carboxy-PTIO (10 μmol/L), AngII (1 μmol/L) increased pericyte O2·− levels an average of 0.19±0.03 (n=6 P<0.05) compared with drug vehicle (0.11±0.01, n=6), as shown in Figure 3. This increase was not observed with AngII in the absence of surrounding mTAL (AngII: 0.06±0.01, n=6 vs vehicle: 0.08±0.03, n=6), whereas significant increases in pericyte O2·− levels were observed in response to treatment with DETC with menadione (0.49±0.07; P<0.01, n=6). O2·− responses stimulated by DETC with menadione tended to be higher in pericytes with mTAL (0.56±0.12, n=6) compared with pericytes alone (0.38±0.09, n=6), indicating that when large amounts of O2·− are produced in mTAL, O2·− can diffuse into pericytes. This was also dramatically observed when the tissue strips with mTAL were preincubated with carboxy-PTIO (10 μmol/L) to scavenge NO. When O2·− was stimulated by DETC with menadione in the reduced tissue NO, Eth/DHE levels in pericyte with mTAL attached were significantly (P<0.05) higher, averaging 0.72±0.02 (n=6), whereas the signal in pericytes alone averaged 0.49±0.07 (n=6). Taken together, these responses indicate that O2·− cross-talk between mTAL and pericytes can indeed occur, especially in situations when tissue NO produced from the mTAL fails to serve as a barrier to this movement.
Evidence That O2·− Can Modify NO Cross-Talk From mTAL to Adjacent VR Pericytes
NO responses of the VR pericytes were measured after reduction of tissue O2·− levels with 1 mmol/L TEMPOL. These VR were obtained from kidneys perfused with microspheres to disrupt the endothelium. Superfusion of the microtissue strips with TEMPOL increased NO levels in VR pericytes, reaching a level of significance at 276 seconds after stimulation when compared with the vehicle response (Figure 4A). As shown in Figure 4B, when these pericytes were studied in the presence of surrounding mTAL, TEMPOL significantly increased DAF-2 fluorescence by 16.3±4.1 units (P<0.05, n=5) at 300 seconds compared with pericytes without mTAL (−5.6±7.9 units, n=5). Similarly, epithelial cells of the mTAL within these tissue strips increased by 23.5±4.1 units (P<0.01, n=5) in response to TEMPOL at 300 seconds compared with drug vehicle (−2.3±0.5 units, n=5; Figure 4B). Responses of drug vehicle and positive control stimuli (100 μmol/L DETA-NONOate) did not differ significantly between the 3 conditions. These results indicate that the increases in NO levels of pericytes could be attributed to the NO released from the mTAL. Together with the results summarized in Figure 3, these data indicate not only that NO modifies O2·− tubulovascular cross-talk but also conversely, that O2·− modifies NO cross-talk between the mTAL and the VR.
The results of this study indicate that within the renal outer medulla, the release of O2·− from tubular epithelial cells of the mTAL can be stimulated by AngII (Figure 2). Ortiz and Garvin11 observed in cortical TAL that scavenging O2·− with the cell-permeable SOD mimetic TEMPOL increased Arg-induced NO release and decreased Cl− absorption, indicating that O2·− might counterregulate the NO-induced reduction of NaCl absorption in this tubular segment.11 Previous studies in our laboratory have demonstrated that an increase in medullary O2·− levels by medullary infusion of the SOD inhibitor DETC reduced both MBF and sodium excretion.1 Taken together, AngII therefore appears to increase O2·− in mTAL that in turn scavenges NO and contributes to an increase of NaCl absorption in the mTAL. Greater levels of oxidative stress in the outer medulla would thereby be expected to enhance the net retention of filtered sodium in the intact kidney. Although the concentrations of AngII used in these studies exceeded levels in the systemic circulation and even reported intrarenal levels (10−9 to 10−10mol/L),12 mTAL O2·− responses to AngII in our medullary tissue strips showed linear increases from 10−8 to 10−6 mol/L AngII, with a plateau reached at >10−6mol/L. Because the microtissue strip thicknesses ranged from 50 to 80 μm and a time of 250 seconds was required to achieve significant O2·− responses, it is also likely that the AngII concentrations that actually reached receptor sites were lower than those delivered.
AngII did not induce O2·− production within VR pericytes under normal conditions. Pallone13 has shown that AngII vasoconstricts isolated, perfused, VR vessels of the rat outer medulla, an event that we have shown is related to an increase of pericyte intracellular Ca2+ concentrations.4 Because endothelial cells of the VR did not effect O2·− levels in pericytes (Figure 3), it is unlikely that O2·− produced in endothelial cells directly modified the vasoconstriction of the isolated, perfused VR. However, under conditions in which NO production was reduced, diffusion of O2·− produced in the mTAL did modulate pericyte O2·− levels. This was clearly demonstrated when the microtissue strips were pretreated with the NO scavenger carboxy-PTIO (Figure 3). Conversely, when O2·− produced in the mTAL was scavenged by TEMPOL (Figure 4), the pericytes of surrounding VR showed increases in NO levels. Because NO of the pericytes in the absence of mTAL was not increased by TEMPOL, these observations indicate that the cross-talk of NO between the tubules and the VR is modified by regional O2·− production. We conclude from these observations that AngII-stimulation of O2·− in mTAL reduces NO-mediated “tubulovascular cross-talk” and buffering of the VR responses to vasoconstrictor compounds. These interactions might increasingly dominate and reduce the buffering effects of NO at higher concentrations of AngII. Furthermore, as higher levels of O2·− begin to predominate, greater production of peroxynitrite and H2O2 could lead to even greater reductions of MBF and sodium excretion. We have recently reported that medullary interstitial infusion of H2O2 reduces MBF and sodium excretion14 and that long-term medullary infusion results in sustained hypertension in Sprague-Dawley rats.15 In the present study, we did not focus on O2·− production in VR endothelial cells. It is possible that O2·− produced in endothelial cells could interact with endothelial NO and also modify endothelial function. Because coronary microvascular endothelial cells are known to produce O2·− in response to AngII, endothelial cells of the VR might respond similarly. Although we have demonstrated that endothelial cells do not normally effect pericyte O2·− levels, it remains to be investigated whether NO or O2·− can cross-talk between endothelial cells and pericytes under certain conditions.
Microdialysis studies in our laboratory indicate that O2·− molecules can diffuse within the interstitial space, as shown by the increased conversion of DHE to Eth within the dialysate fluid when the SOD inhibitor DETC was administered into the medullary interstitial space.2 Although the source of this O2·− was not determined, it was evident that O2·− was capable of diffusing through the renal interstitial space and crossing the dialysis membrane. Because we have recently shown that AngII-induced NO from mTAL diffused into pericytes to buffer the vasoconstrictor effects of AngII,4,5, we hypothesized that tubular O2·− and NO, which are both stimulated by AngII, would interact within the renal medullary interstitial space. The estimated diffusion distance of free radicals in extracellular fluid is 200 to 250 μm for NO16 and 50 μm for O2·−.17 Although O2·− does not appear to pass through lipid membranes, Terada18 demonstrated that O2·− produced in vascular endothelial cells could efflux into the extracellular space through anion channels. A limitation of using microtissue strips for free-radical measurement is that one cannot show direct evidence of a correlation between the amount of free radical and the physiologic regulation of VR blood flow. Because we were unable to increase intracellular O2·− by a known amount, no attempts were made to calibrate intracellular fluorescence to absolute O2·− levels. However, we have shown previously, using in vivo microdialysis techniques, that a reduction of O2·− levels in the renal medulla by locally administered TEMPOL results in increases in MBF and sodium excretion.1 Conversely, increased medullary O2·− levels by an SOD inhibitor showed the opposite effect.1,2
Although our data indicate the importance of NAD(P)H oxidase in AngII-induced production of O2·− in mTAL, other pathways could also contribute to this response. Because endothelial NO synthase is present in mTAL,19 AngII-stimulated NO production in these cells could increase peroxynitrite formation and enhance O2·− production through the so-called “NO synthase uncoupling” mechanism.20 We believe, however, that peroxynitrite might not be a functionally important product of O2·− metabolism in the outer medullary region of Sprague-Dawley rats because we have found that AngII results in substantial elevation of NO levels within epithelial cells of the mTAL.4 However, another pathway that appears to contribute significantly to the production of O2·− in the renal medulla is the mitochondrial redox pathway.1 At pressor doses of AngII, associated reductions of MBF could reduce oxygen levels in the medulla and stimulate mitochondria. PO2 in this region of the kidney is relatively low (≈20 to 30 mm Hg),21 and oxygen delivery in the medulla is flow limited because of a high rate of metabolic O2 utilization relative to blood flow. Therefore, we conclude that AngII stimulates O2·− production in the mTAL via the NAD(P)H oxidase pathway, and interactions between O2·− and NO act as the free-radical cross-talk between mTAL and VR to regulate MBF.
In contrast to the renal cortex, adequate delivery of oxygen for metabolic needs to the renal medulla is flow dependent. We propose that reactive oxygen species released in response to changes in sodium transport in the mTAL influence vascular tone of the surrounding VR and thereby communicate tubular metabolic needs to this vasculature. We have recently shown that AngII stimulation of mTAL increased the production of NO in mTAL epithelial cells and that this NO diffused to nearby VR pericytes. We proposed that this NO cross-talk mechanism would help buffer the direct vasoconstrictor actions of AngII on the VR and help maintain adequate blood flow for tubular metabolism. In the present study, we have shown that AngII can also stimulate the production of O2·− in mTAL, a response that would be counterproductive and would reduce MBF. These opposing events, however, appear to be dominated normally by the large production of NO that our present and previous studies have shown to prevail in the overall control of blood flow. Although we have demonstrated that O2·− production by mTAL can reduce the NO cross-talk between the mTAL and VR, we have also shown that normally, the release of NO is capable of protecting this region from the effects of stimulated O2·− production. In situations when NO production is reduced or O2·− production is excessive, we would anticipate that blood flow to this region would be compromised, with resulting ischemic damage.
This work was supported by the National Heart, Lung, and Blood Institute grant HL-29587. The authors thank Glenn Slocum for his expert assistance with the microscopy and Meredith M. Skelton for careful review of the manuscript.
- Received February 20, 2003.
- Revision received March 12, 2003.
- Accepted August 7, 2003.
Zou AP, Li N, Cowley AW Jr. Production and actions of superoxide in the renal medulla. Hypertension. 2001; 37: 547–553.
Makino A, Skelton MM, Zou AP, Roman RJ, Cowley AW Jr. Increased renal medullary oxidative stress produces hypertension. Hypertension. 2002; 39: 667–672.
Mori T, Dickhout JG, Cowley AW Jr. Vasopressin increases intracellular nitric oxide concentration via Ca2+ signaling in inner medullary collecting duct. Hypertension. 2002; 39: 465–469.
Dickhout JG, Mori T, Cowley AW Jr. Tubulovascular nitric oxide cross-talk buffers Ang II–induced medullary vasoconstriction. Circ Res. 2002; 91: 487–493.
Cowley AW Jr, Mori T, Mattson DL, Zou AP. Role of renal NO production in the regulation of medullary blood flow. Am J Physiol. 2003; 284: R1355–R1369.
Lang D, Mosfer SI, Shakesby A, Donaldson F, Lewis MJ. Coronary microvascular endothelial cell redox state in left ventricular hypertrophy: the role of angiotensin II. Circ Res. 2000; 86: 463–469.
Lassegue B, Sorescu D, Szocs K, Yin Q, Akers M, Zhang Y, Grant SL, Lambeth JD, Griendling KK. Novel gp91(phox) homologues in vascular smooth muscle cells: nox1 mediates angiotensin II–induced superoxide formation and redox-sensitive signaling pathways. Circ Res. 2001; 88: 888–894.
Li N, Yi FX, Spurrier JL, Bobrowitz CA, Zou AP. Production of superoxide through NADH oxidase in thick ascending limb of Henle’s loop in rat kidney. Am J Physiol. 2002; 282: F1111–F1119.
Ortiz PA, Garvin JL. Interaction of O2− and NO in the thick ascending limb. Hypertension. 2002; 39: 591–596.
Navar LG, Harrison-Bernard LM, Nishiyama A, Kobori H. Regulation of intrarenal angiotensin II in hypertension. Hypertension. 2002; 39: 316–322.
Chen YF, Cowley AW, Zou AP. Increased H2O2 counteracts the vasodilator and natriuretic effects of superoxide dismutation by TEMPOL in the renal medulla. Am J Physiol Regul Integr Comp Physiol. In press.
Makino A, Skelton MM, Zou AP, Cowley AW Jr. Increased renal medullary H2O2 leads to hypertension. Hypertension. 2003; 42: 25–30.
Lancaster JR Jr. The physical properties of nitric oxide: determinants of the dynamics of NO in tissue. In: Ignarro LJ, ed. Nitric Oxide: Biology and Pathobiology. New York, NY: Academic Press; 2000.
Laursen JB, Somers M, Kurz S, McCann L, Warnholtz A, Freeman BA, Tarpey M, Fukai T, Harrison DG. Endothelial regulation of vasomotion in apoE-deficient mice: implications for interactions between peroxynitrite and tetrahydrobiopterin. Circulation. 2001; 103: 1282–1288.