Roles of Oxidative Stress and AT1 Receptors in Renal Hemodynamics and Oxygenation in the Postclipped 2K,1C Kidney
The spontaneously hypertensive rat (SHR) exhibits angiotensin II (Ang II)–dependent oxidative stress and reduced efficiency of renal oxygen usage (QO2) for tubular sodium transport (TNa). We tested the hypothesis that oxidative stress determines the reduced TNa:QO2 ratio in the clipped kidney of the early 2-kidney, 1-clip (2K,1C) Ang II–dependent model. One week after sham operation (Sham) or clip placement, 2K,1C rats received for 2 weeks either a vehicle, the superoxide dismutase mimetic tempol (Temp), or candesartan (Cand). Oxidative stress was assessed from excretion of 8-isoprostaglandin F2α (PGF2α) and malondialdehyde (MDA) and renal oxygenation from pO2 in the renal cortex and from the ratio of calculated TNa and QO2 values. The mean arterial pressure (MAP) of Sham (113±6 mm Hg) was increased in 2K,1C vehicle-treated rats (148±4 mm Hg), but both Temp and Cand restored MAP to Sham levels. The excretions of 8-iso-PGF2α and MDA were higher in 2K,1C vehicle-treated rats compared with Sham and were normalized by Temp. The pO2 of Sham (42±2 mm Hg) was lower in 2K,1C vehicle-treated animals (28±2 mm Hg). This was restored to Sham values by Temp (36±3 mm Hg) but not by Cand (28±2 mm Hg). The TNa:QO2 of Sham (12.9±1.6) was reduced in 2K,1C vehicle-treated rats (9.7±2.8) and was restored to Sham values by Temp (13.7±2.5) but not by Cand (7.5±1.6). We conclude that the correction of oxidative stress in the 2K,1C model partially corrects renal cortical hypoxia and inefficient utilization of O2 for Na+ transport, independent of the fall in blood pressure.
Renal oxygen consumption (QO2) is closely related to the energy required for sodium transport (TNa). Classic studies have established that variations in sodium delivery and hence, transport, over a broad range are matched by proportionate changes in QO2.1,2 Prevention of glomerular filtration reduced QO2 to a low but measurable value, identified as the O2 required for basal kidney metabolism. However, across a broad range of glomerular filtration rates (GFRs), QO2 rose linearly with the GFR above the basal level. This defines the normal rate at which O2 is consumed to satisfy the energy requirements for TNa (15 to 25 μmol of Na transported per μmol of O2 consumed).
Recent studies have reported that the TNa:QO2 ratio is variable. Laycock and associates3 showed that the TNa:QO2 in the dog kidney was reduced by ≈50% during inhibition of nitric oxide (NO) synthase with l-nitroarginine. We showed a similar reduction in TNa:QO2 in kidneys from spontaneously hypertensive rats (SHR).4 The SHR is a model of reduced renal NO bioactivity associated with increased superoxide radical (O2·−).5 There is a complex interrelation between NO, O2·−, and Po2 or O2 usage in the tissues. In pulmonary arteries and vascular smooth muscle cells, both chronic hypoxia and hyperoxia can enhance O2·− levels.6,7 Increased O2·− interacts with NO, which reduces its bioactivity and produces peroxynitrite. Nevertheless, we detected a reduced renal cortical pO2 in the SHR, which we attributed to inefficient utilization of O2 for TNa as a consequence of functional NO deficiency during oxidative stress. We found also that the renal cortical hypoxia and reduction in Po2 in the SHR could be corrected by prolonged administration of the angiotensin receptor antagonist candesartan (Cand) but not by equally effective antihypertensive therapy with agents that do not block the renin-angiotensin system. Moreover, we found that Cand also restored NO bioactivity in the SHR kidney.8 Angiotensin II (Ang II) stimulates the expression of NADPH oxidase.9,10 We concluded that the reduced NO bioactivity and inefficient utilization of O2 in the SHR kidney could have been secondary to oxidative stress. The present study was designed to test this hypothesis in the early phase of 2-kidney, 1-clip (2K,1C) Goldblatt hypertension. The 2K,1C is a pathophysiological model of Ang II–dependent hypertension, which suppresses function in the clipped kidney in response to the induced renal artery stenosis.11 One week after the renal artery clip was placed, rats were treated with the superoxide dismutase mimetic tempol (Temp) to reduce O2·− or with Cand to lower blood pressure equivalently.
These studies were performed under guidelines recommended by the National Institutes of Health and were approved by the Georgetown University Animal Care and Use Committee. Young male Sprague-Dawley rats (80 to 100 g) were anesthetized with isoflurane (0.5% to 1.5%). A silver clip (0.2 mm) was placed around the left renal artery (2K,1C). Sham animals (Sham) were prepared similarly without clip placement. One week thereafter, the 2K,1C rats received for 2 weeks vehicle (Veh, n= 8), Temp (200 nmol · kg−1 · min−1 via osmotic minipump, n=8), or Cand (10 mg · kg−1 · d−1 added to drinking water, n=9). On day 13 of drug administration, the rats were placed in metabolic cages for a 24-hour urine collection. Whole-body oxidative stress was assessed from the excretion of 8-isoprostaglandin F2α (8-iso-PGF2α) and malondialdehyde (MDA) from both kidneys. 8-Iso-PGF2α was measured by an ELISA (Cayman Chemical, Ann Arbor, Mich) on extracted 24-hour samples. MDA was measured by chemical concentration of thiobarbituric acid–reactive substances.
On the following day, rats were anesthetized with Inactin (100 mg/kg IP; Research Biochemicals Inc) and prepared for measurements of mean arterial pressure (MAP), renal function, and renal pO2, as described previously4. To assess renal function in each kidney, urine was collected from the left ureter by placement of a cannula and from the right kidney, by a bladder catheter. The GFR was assessed by the clearance of [14C]inulin and renal plasma flow (RPF), by the clearance of [3H]para-aminohippuric acid. Renal blood flow (RBF) was calculated as RPF, factored by 1 minus the hematocrit value. At completion of the clearance period, samples were drawn from the femoral artery and left renal vein for measurement of O2 content and pO2. Care was taken to sample only venous blood emerging from the kidney, as described previously.4 Blood samples were analyzed in a blood gas cooximeter (Instrumentation Labs, Inc).
QO2 was calculated from the product of RBF and the difference between renal arterial and vein oxygen content. The TNa was calculated from the difference in the filtered Na+ load and the excreted Na+. The efficacy of oxygen usage for sodium transport was calculated from the ratio TNa:QO2. The renal pO2 in the outer and inner cortex were measured in the clipped kidney in separate animals by using an ultramicroelectrode, as described previously.12 The platinum-iridium electrode was coated with an O2-permeable membrane and connected to an ultra-high-impedence picoammeter (pA6000; World Precision Instruments). The analog output was digitally converted and displayed on a computer with an acquisition system (MacLab 4A; AD Instruments). The electrode provides a rapid, stable, and linear output in response to pO2 from 5 to 500 mm Hg. Before use, the electrode was calibrated to 4 different levels of O2. The electrode was inserted into the clipped kidney only and placed either in the outer or inner (advanced 2.0 mm) cortex. Statistics were assessed by ANOVA, with a post hoc Dunnett’s test to determine differences. Significance was determined at P<0.05.
All rats grew at normal rates. The body weights at the end of treatment did not differ between groups (Table 1). MAP measured under anesthesia was higher (P<0.001) in 2K,1C, Veh compared with Sham and was normalized by both Temp and Cand. The clipped kidney weight was lower in 2K,1C Veh and Cand compared with Sham but was not different from Sham in 2K,1C Temp. The right (unclipped) kidney weight was not different among groups. However, the ratio of left to right kidney weight was decreased in 2K,1C Veh compared with Sham and normalized in 2K,1C Temp rats (Table 1). This ratio was further increased in 2K,1C Cand.
As shown in Table 1, the GFR, factored by kidney weight, was reduced in the postclip kidney of 2K,1C, Veh compared with Sham. The low GFR associated with clipping was increased significantly in 2K,1C by Temp but not by Cand. Compared with Sham, the RPF was lower in 2K,1C Veh and was not changed by Cand or Temp. The filtration fraction was similar to Sham in 2K,1C Veh and 2K,1C Temp. However, the filtration fraction was reduced by Cand.
As shown in Figure 1, the excretion of 8-iso-PGF2α was higher (P<0.05) in 2K,1C Veh compared with Sham. Treatment with Temp (P<0.05) or Cand (P<0.05) reduced 8-iso-PGF2α excretion to a level comparable to those in Sham. The excretion of MDA also was higher (P<0.05) in 2K,1C Veh compared with Sham. However, whereas Temp reduced this (P<0.05) to a level comparable to that in Sham, Cand had no effect. Because the urine samples represent excretion from both kidneys, these values may not reflect the true level of oxidative stress in the clipped kidney.
O2 extraction, as shown by the difference in renal arterial and venous oxygen content, was elevated significantly in 2K,1C Veh. This was not changed by Temp or Cand (Table 2). However, O2 extraction in the 2K,1C Temp group was not different from Sham. There were no differences in O2 usage between groups. Changes in TNa reflected the pattern of changes in GFR.
The O2 efficiency for Na+ transport in the clipped kidney is shown in Figure 2. The TNa:QO2 of Sham was 12.9±1.6 (μmol of Na+ transported per μmol of O2 consumed). This was reduced in 2K,1C Veh 9.7 to 1.8 (P<0.01) because of a sharp fall in TNa with relative preservation of QO2. The ratio was restored to Sham values in 2K,1C given Temp (13.7±2.5) because the reduction in TNa was matched by an equivalent reduction in QO2. However, the TNa:QO2 of 2K,1C given Cand was reduced even further (7.5±1.6).
The pO2 in the inner and outer cortex of the clipped kidney is shown in Figure 3. In Sham animals, the pO2 in the outer and inner cortex averaged 42±2 and 33±2 mm Hg, respectively. The pO2 in the outer cortex was reduced by 31% and in the inner cortex by 60% in 2K,1C Veh. Whereas 2K,1C rats given Temp had pO2 values equivalent to those in Sham in the outer cortex, these were reduced in the inner cortex. The pO2 was not improved by Cand compared with Veh 2K,1C. Indeed, there was a further reduction in pO2 in the inner cortex to 11±3 mm Hg.
The main new findings from this study are that the clipped kidney of the early 2K,1C model has an increased use of O2 in relation to TNa associated with a sharp reduction of pO2 in the outer cortex and especially the inner cortex. These abnormalities are largely prevented by Temp, which also prevented the whole-body oxidative stress associated with the clipping. Although Temp treatment led to a normal blood pressure in the 2K,1C rats, the defects in oxygenation were not corrected by equivalent reduction in blood pressure by Cand. Indeed, Cand led to a further decline in TNa:QO2 and to a reduction in pO2 in the inner cortex of the clipped kidney to the remarkably low value of 11 mm Hg. Moreover, the reduction in the ratio of kidney weight (clipped to unclipped) of 78% in 2K,1C rats given Veh was improved by Temp but worsened by Cand.
The relation between NO and O2·− affects both vascular tone and O2 consumption.13,14 NO at physiological levels can compete with O2 for the respiratory chain in mitochondria. Thus, low levels of NO after inhibition of NO synthase enhance O2 usage and reduce the levels of oxygen in various tissues, including the kidney. Decreased bioactive NO is found in the kidney in the SHR, which is a model of oxidative stress. In this setting, the low NO in the juxtaglomerular apparatus (JGA) can be ascribed to interaction with O2·−, because NO bioactivity increased 6-fold after local microperfusion of Temp to metabolize O2·−.8 Thus, we propose that the reduced TNa:QO2 and the reduced pO2 of the SHR kidney might be secondary to NO deficiency owing to interaction with O2·−. However, we cannot rule out additional effects of Temp in the development of renovascular hypertension. Because kidney size was not reduced in this group, it is plausible that the beneficial effects are related to more complete perfusion of the kidney during this typically ischemic phase of this model. However, this effect may also be related to superoxide dismutase activity, because Temp promotes longer bioactivity of NO and would help to maintain stable blood flow.
Further evidence is derived from the results of prolonged blockade of Ang II type I receptors with Cand. Cand corrected oxidative stress, restored NO bioactivity in the JGA, and normalized renal TNa:QO2 and renal pO2 levels in SHR.4 Therefore, we elected to study the relation between oxidative stress and oxygenation more directly in the present series in a model of extreme Ang II action in the clipped kidney of 2K,1C rats. Previous studies in the 2K,1C pig15 and rat16 showed that increased oxidative stress was associated with high renin levels in the early phase. We have consistently observed low pO2 in the cortex of hypertensive kidneys, which also demonstrate high oxidative stress and deficient NO. However, direct supporting evidence of this relation is currently unavailable. The finding that Temp corrected oxidative stress, reduced renal cortical pO2, and reduced renal oxygen usage for TNa provides direct evidence linking defective renal oxygenation to oxidative stress. Because TNa:QO2 may be dependent on available O2, the effect on superoxide and NO production may suggest that this parameter is linked to redox capacity within the kidney.
Studies in patients with renovascular hypertension due to unilateral renal artery stenosis show enhanced markers for oxidative stress and decreased renal excretion of NO metabolites associated with endothelial dysfunction.17 All of these parameters are normalized after correction of the hypertension by renal angioplasty. Our study suggests that the beneficial effect of angioplasty is not due to the reduction in blood pressure, because our previous study detected no benefit from nonspecific antihypertensive treatment in the SHR model and no benefits in the present study in 2K,1C hypertension with Cand. In another study of patients with unilateral renal artery stenosis, O2 saturation was consistently lower in the blood draining from the contralateral compared with the poststenotic kidney.18 The authors related the enhanced renal O2 uptake to a reduced filtration fraction, leading to a lower level of Na+ reabsorption. However, neither TNa nor QO2 could be assessed in this study. We did not measure the O2 content of the contralateral kidney in our study. Nevertheless, the O2 content of the postclip kidney was already reduced sharply (Table 2). Therefore, if the contralateral kidney has an even lower renal venous O2 content, as anticipated from this clinical study, then it must likely also have profound defects of oxygenation.
Because treatment trials of angioplasty of the stenotic renal artery in human renovascular hypertension have been quite disappointing, attention has focused on the use of angiotensin-converting enzyme inhibitors or angiotensin receptor blockers. However, although these drugs can reduce blood pressure quite effectively, they can cause a further loss of GFR and size in the poststenotic kidney. This has limited enthusiasm for their use in this condition. The results in this model of acute 2K,1C hypertension may not be fully applicable to chronic renovascular hypertension in humans. Nevertheless, blocking of the angiotensin AT1 receptor in the 2K,1C model reduced the GFR and size of the postclipped kidney. In contrast, we observed that scavenging superoxide with Temp not only corrected hypertension but also normalized the GFR and the relative size of the 2 kidneys. Importantly, the angiotensin receptor blocker led to a further deterioration in TNa:QO2 and to a profound fall in pO2 in the deep cortex to values as low as 11 mm Hg, whereas these parameters of oxygenation were improved by Temp. This suggests that treatment aimed at correction of oxidative stress could have advantages over current therapy for renovascular hypertension.
In summary, we found that in renovascular hypertension generated by clipping of the left renal artery, oxidative stress was increased both systemically and in the kidney. In the clipped kidney, GFR, RBF, and the efficient use of oxygen were reduced, concomitant with hypertension. Correction of hypertension by Cand did not improve renal function or oxygen efficiency in the clipped kidney. Reduction of hypertension by Temp led to normalized renal function and oxygen efficiency in the clipped kidney. The renal tissue pO2 in both the outer and inner cortex of the left kidney was also reduced by clipping. The tissue pO2 was unaffected by Cand but was normalized by Temp. We conclude that suppression of oxidative stress in this model partially corrects renal cortical hypoxia and inefficient usage of O2 for sodium transport.
This study extends the observations that many of the physiological consequences of elevated Ang II are mediated by superoxide. Renal function and the efficient use of oxygen in the kidney are both suppressed during high superoxide conditions of renovascular hypertension. This suggests that the levels of oxygen in the kidney are dependent on the utilization of oxygen for the work of sodium transport. Therefore, the relation of oxygen with the important metabolites NO and superoxide may be partially dependent on energy requirements. This represents novel regulation of a potentially potent vasoactive family.
This work was supported by grants from the National Institute for Diabetes, Digestive, and Kidney Diseases (DK-36079 and DK-49870); the National Heart, Lung, and Blood Institute (HL-HL68686 to 01); the National Kidney Fund of the Nations Capital; and from funds from the George E. Schreiner Chair of Nephrology. We are grateful to Sharon Clements for preparation of this manuscript.
- Received September 19, 2002.
- Revision received October 21, 2002.
- Accepted December 6, 2002.
Laycock SK, Vogel T, Forfia PR, Tuzman J, Xu X, Ochoa M, Thompson CI, Nasjletti A, Hintze TH. Role of nitric oxide in the control of renal oxygen consumption and the regulation of chemical work in the kidney. Circ Res. 1998; 82: 1263–1271.
Adler S, Huang H. Impaired regulation of renal oxygen consumption in SHR. J Am Soc Nephrol. 2002; 13: 1788–1794.
Terada LS, Guidot DM, Leff JA, Willingham IR, Hanley ME, Piermattei D, Repine JE. Hypoxia injures endothelial cells by increasing endogenous xanthine oxidase activity. Proc Natl Acad Sci U S A. 1992; 89: 3362–3366.
Freeman BA, Crapo JD. Hyperoxia increases oxygen radical production in rat lungs and lung mitochondria. J Biol Chem. 1981; 256: 10986–10992.
Griendling KK, Minieri CA, Ollerenshaw JD, Alexander RW. Angiotensin II stimulates NADH and NADPH oxidase activity in cultured vascular smooth muscle cells. Circ Res. 1994; 74: 1141–1148.
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.
Lerman LO, Nath KA, Rodriguez-Porcel M, Krier JD, Schwartz RS, Napoli C, Romero JC. Increased oxidative stress in experimental renovascular hypertension. Hypertension. 2001; 37 (part 2): 541–546.