Hypercholesterolemia (HC), a major risk factor for onset and progression of renal disease, is associated with increased oxidative stress, potentially causing endothelial dysfunction. One of the sources of superoxide anion is xanthine oxidase (XO), but its contribution to renal endothelial function in HC remains unclear. We tested the hypothesis that XO modulates renal hemodynamics and endothelial function in HC pigs. Four groups (n = 23) of female domestic pigs were studied 12 wk after either normal (n = 11) or HC diet (n = 12). Oxidative stress was assessed by plasma isoprostanes and oxidized LDL, and the XO system by plasma uric acid, urinary xanthine, and renal XO expression (by immunoblotting and immunohistochemistry). Renal hemodynamics and function were studied with electron beam-computed tomography before and after endothelium-dependent (ACh) and -independent (sodium nitroprusside) challenge, during a concurrent intrarenal infusion of either oxypurinol or saline (n = 5–6 in each group). HC showed elevated oxidative stress, higher plasma uric acid (23.8 ± 3.8 vs. 6.2 ± 0.8 μM/mM creatinine, P = 0.001), lower urinary xanthine, and greater renal XO expression compared with normal. Inhibition of XO in HC significantly improved the blunted responses to ACh of cortical perfusion (13.5 ± 12.1 and 37.2 ± 10.6%, P = 0.01 and P = not significant vs. baseline, respectively), renal blood flow, and glomerular filtration rate; restored medullary perfusion; and improved the blunted cortical perfusion response to sodium nitroprusside. This study demonstrates that the endogenous XO system is activated in swine HC. Furthermore, it suggests an important role for XO in regulation of renal hemodynamics, function, and endothelial function in experimental HC.
- oxidative stress
- uric acid
hypercholesterolemia (hc) is a major risk factor for the development and progression of atherosclerosis (45) and is associated with an increase in the incidence of coronary artery disease and cardiac events (1). Even at an early stage, HC can alter vasomotor regulation in both large vessels and the microcirculation (26, 46) and is responsible for the impairment of both the function and the structure of various vascular beds.
Moreover, HC has been demonstrated to be an independent risk factor for the onset (10) and progression (22) of renal disease and can both induce and worsen renal glomerular, interstitial, and vascular damage (30, 35). We have previously shown (11, 20, 42) that even a short exposure to diet-induced HC is associated with increased formation of oxidized LDL (ox-LDL) and reactive oxygen species (ROS). Increased oxidative stress impairs endothelial function in both humans and animal models (13, 14, 37), partly by reducing bioavailability of nitric oxide (NO) via its reaction with ROS. Moreover, ROS can induce renal injury both by direct cellular toxicity (3) and by promoting production of ox-LDL, which, in turn, further inactivates NO (12, 33) and directly contributes to tubulointerstitial disease (2) and glomerulosclerosis (16).
Superoxide anions and other ROS may be generated by several different enzymatic and nonenzymatic mechanisms. In the vascular endothelium the main source for superoxide is NAD(P)H-oxidase, but additional enzymes can induce ROS production, e.g., cyclooxygenase, uncoupled endothelial nitric oxide synthase (eNOS), and xanthine oxidase (XO) (28). XO can lead to superoxide production during the purine degradation process, which involves metabolism of hypoxanthine and xanthine to uric acid (4). XO activity has been demonstrated to be elevated in the plasma of hypercholesterolemic subjects and to contribute to endothelial dysfunction in HC animals (52) and humans (6). In the kidney, XO is also involved in ischemic injury (23). However, the contribution of XO-derived ROS to endothelial dysfunction in the kidney in early atherosclerosis has not been determined.
The purpose of the present study was to assess the role of XO in the hemodynamics and endothelial function in the kidney of pigs with diet-induced HC. For this purpose, we used electron beam-computed tomography (EBCT), which provides accurate and noninvasive measurement of single-kidney regional hemodynamics and function in vivo, and allows detection of subtle alterations in renal hemodynamics and function (8, 10, 11, 20, 31, 32, 42).
MATERIALS AND METHODS
This study was approved by the Institutional Animal Care and Use Committee of the Mayo Clinic College of Medicine (Rochester, MN). Four groups of female domestic cross-bred pigs (n = 23, mean body weight 49.2 ± 10.9 kg) were studied with EBCT after 12 wk of either a normal (n = 11) laboratory chow diet (Land O Lakes Purina Feed, Shoreview, MN), or high-cholesterol diet (HC, n = 12), which contained 2% cholesterol and 15% lard by weight (TD 93296, Harlan Teklad, Madison, WI; see Table 1).
After completion of 12 wk on the diet, blood and urine samples were collected from all pigs for measurement of serum lipid profile (Roche, Nutley, NJ) and creatinine (spectrophotometry, Creatinine Analyze 2, Beckman Coulter, Fullerton, CA) (21). In addition, plasma levels of PGF2 alpha-isoprostanes (EIA, Cayman Chemical, Ann Arbor, MI) (32, 53) and ox-LDL (ELISA, Mercodia, Winston Salem, NC) (8) served to assess systemic oxidative stress. Plasma levels of uric acid and urinary levels of xanthine (spectrophotometry) served as measures of the activity of the XO system.
The pigs were then randomized to obtain constant intrarenal infusion of either oxypurinol (oxy) or vehicle during performance of the subsequent EBCT studies. In each acute EBCT study, cortical, medullary, and papillary perfusion were measured before and after infusion of ACh or sodium nitroprusside, representing endothelium-dependent and -independent challenges, respectively.
After completion of studies, the pigs were euthanized with intravenous (100 mg/kg) Sleepaway (pentobarbital sodium, Fort Dodge Laboratories, Fort Dodge, IA). The kidneys were immediately dissected, and sections shock-frozen in liquid nitrogen (and maintained at −80°C) or preserved in formalin. Renal XO expression was then assessed using Western blotting and immunohistochemistry.
Spectrophotometric measurements of urinary xanthine levels.
For measurement of xanthine level, diluted, filtered urine was mixed with an internal standard (8–13 Adenine) and analyzed by liquid chromatography tandem mass spectrometry (PE Sciex API 3000 LC/MS/MS, Applied Biosystems, Foster City, CA). LC/MS/MS was performed using a mobile phase composed of 50 mM ammonium formate, pH = 5, and 1:1 mixture of 50 mM ammonium formate, pH = 5: methanol, and ran using a gradient. An Xterra MS C18 column (2.1 × 150 mm) was used to separate xanthine and hypoxanthine from the bulk of the specimen matrix. The MS/MS was operated in the selected reaction monitoring scanning mode. The ratios of the extracted peak areas of xanthine and hypoxanthine to an internal standard was used to calculate the concentration of xanthine in the sample (27).
EBCT studies were performed, as was previously described (31, 42). On the day of the studies, each animal was anesthetized with intramuscular ketamine (20 mg/kg) and xylazine (2 mg/kg), intubated, and mechanically ventilated with room air. Anesthesia was maintained with a mixture of ketamine (15.7 mg·kg−1·h−1) and xylazine (2.3 mg·kg−1·h−1) in saline, administered via an ear-vein cannula (0.05 ml·kg−1·min−1).
Under sterile conditions and fluoroscopic guidance, a 7F arterial guide was advanced from the left carotid artery to the abdominal aorta; a tracker catheter was advanced within the guide into one renal artery to serve for intrarenal infusions, as we have previously shown (10, 31). The arterial guide was maintained at a level above the renal arteries and served for vasodilator infusion and for monitoring mean arterial pressure (MAP) throughout the experiment. A pigtail catheter advanced through a vascular sheath in the left jugular vein was positioned in the right atrium for contrast media injections. ECG leads served for monitoring heart rate.
Animals were then transferred to the EBCT (Imatron C-150, Imatron South San Francisco, CA) scanning gantry. In one normal group of pigs (Noxy, n = 5) and one HC (HCoxy, n = 6), after a 15-min recovery period, a constant infusion of oxypurinol was initiated into the renal artery catheter. In each group, baseline renal perfusion and function were measured after a 30-min intrarenal infusion of oxypurinol (300 mg·min−1·kg−1) (6). This dose has been shown to achieve more than 90% inhibition of XO activity (15). The other normal (n = 6) and HC (n = 6) groups were infused with saline (0.1 ml·kg−1·min−1).
After a 30-min stabilization, hemodynamic measurements were recorded, and the EBCT studies were performed to determine baseline renal hemodynamics and volume in both kidneys. Forty consecutive scans (over 3 min) were obtained at variable time intervals after a bolus injection (0.5 ml/kg over 1 s) of the nonionic, low-osmolar contrast medium iopamidol (Isovue370, Squibb Diagnostics, Princeton, NJ) into the right atrial catheter. After 15 min, a 10-min infusion of ACh (4 μg·kg−1·min−1) or sodium nitroprusside (6 nmol·kg−1·min−1) in random order was performed and the EBCT scans repeated.
To measure renal XO expression, frozen renal tissue of five HC and five normal pigs (including both cortex and medulla) was pulverized and homogenized at 4°C in chilled protein extraction buffer. The homogenate was incubated in buffer for 1 h at 4°C, and the homogenized lysates were then centrifuged for 15 min at 14,000 rpm. The supernatant was removed, and the protein concentration was determined by spectrophotometry with a protein assay (Coomassie Plus, Pierce, Rockford, IL). The lysate was then diluted 1:4 in 1 × polyacrylamide gel electrophoresis sample buffer and heated at 95°C to denature the proteins. The lysate was then loaded into a gel and run for standard Western blotting protocols with the rabbit antixanthine oxidase polyclonal antibody (1:10,000, Chemicon International) as primary antibodies, and anti-rabbit IgG horseradish peroxidase linked whole antibodies from donkey (1:500, Amersham Biosciences) as secondary antibodies. The membrane was exposed for 5 min to a chemiluminescence developing system (SuperSignal West Pico Chemiluminescent Substrate, Pierce) and then finally exposed to X-ray film (Kodak), which was subsequently developed, and intensities of the protein bands were determined by densitometry. The specificity of the immunoblotting was confirmed with negative and positive controls obtained by parallel experiments performed in the absence of the the primary antibody or with a known concentration of XO (enzyme purified from buttermilk, 0.05 μg, Sigma, St. Louis, MO), respectively.
Immunohistochemistry for XO was performed on deparaffinized renal tissue of five HC and five normal pigs, using prediluted monoclonal primary antibodies (LabVision,CA). The secondary antibody, IgG Envision Plus (Dako), was followed by staining with the Vector NovaRED substrate kit (Vector Laboratories, Burlingame, CA), and slides were counterstained with hematoxylin. Kidney sections (2 sections for each pig) were examined, and the staining was quantified (as fraction of surface area) using a computer-aided image-analysis program (MetaMorph, Meta Imaging Series 4.6).
EBCT data analysis.
The methodology used for EBCT data analysis has been previously described in detail (8, 10, 11, 20, 31, 32, 42). Briefly, regions of interest were selected from the images by tracing the aorta and the bilateral renal cortex, medulla, and papilla, and their densities were sampled. Time-density curves were generated for each region, which described the change in tissue density consequent to transit of contrast in that region. The curves were then fitted using a modified gamma-variate fit (31). From each segment of the curve, the area enclosed under the curve and its mean transit time were calculated from the curve-fitting parameters. Renal regional perfusion (ml·min−1·ml tissue−1), normalized single-kidney glomerular filtration rate (GFR; ml·min−1·ml tissue−1), cortical, medullary volumes, and renal blood flow (RBF) were subsequently calculated as previously described (31).
Results are expressed as means ± SE. Comparisons between experimental periods within groups were performed using paired Student's t-test, and among groups using ANOVA, with the Bonferroni correction for multiple comparisons, and unpaired Student's t-test if applicable. Statistical significance was accepted if P < 0.05.
Systemic characteristics of the study groups are shown in Table 2. At the end of the diet period, the pigs fed with HC and normal diets had similar body weights, MAP, heart rate, and serum creatinine. Total and LDL cholesterol levels were significantly elevated in the HC group compared with normal (P < 0.0001, Table 2), as were plasma isoprostanes and oxidized LDL (P < 0.05).
HC animals showed a significant increase of plasma levels of uric acid compared with normal (P = 0.001, Table 2), while urinary levels of xanthine were significantly lower in HC (P < 0.01). Renal XO expression was significantly (P < 0.05) increased in HC compared with normal pigs (Fig. 1A), and immunostaining showed that it was expressed throughout the kidney, but it was particularly prominent in the cortical, proximal, and distal tubules (P < 0.05 vs. normal, Fig. 1B). No differences in XO expression or activity were observed between HC or normal pigs that were randomized to vehicle compared with oxypurinol infusion (data not shown).
Renal hemodynamics and function.
Under basal conditions, cortical and medullary perfusions, volumes, and blood flows, as well as single kidney blood flow and GFR, were similar among all of the groups [P = not significant (ns); Table 3]. Intrarenal infusion of oxypurinol did not modify heart rate or blood pressure. In addition, it did not alter basal kidney hemodynamics in either normal and HC groups compared with vehicle-infused (Table 3) or contralateral kidneys (data not shown).
In the normal groups, ACh induced a significant increase in the EBCT-derived perfusion of the renal cortex, medulla, and papilla, as well as in RBF and GFR (P < 0.01 vs. baseline for all). In HC the increase in cortical, medullary, and papillary perfusion in response to ACh was not significant compared with baseline and reached values significantly smaller than in the normal group (P < 0.05 for all regions), as was the increase in RBF (P < 0.01 vs. normal). GFR showed a modest but significant (P < 0.05) increase, which was significantly attenuated compared with normal group (P < 0.05).
On the other hand, in HCoxy, the perfusion of renal cortex significantly increased in response to ACh (P < 0.001 from baseline), although it did not reach normal values (P < 0.05), and RBF showed a similar pattern ( P < 0.05 vs. HC, and P = ns vs. normal). Furthermore, medullary and papillary perfusion response both normalized, increased significantly (P < 0.005) in response to ACh, and were not different from normal (Fig. 2A and 3, P = ns). However, oxypurinol did not modify GFR response to ACh, which remained significantly lower compared with normal (Table 3).
In normal animals, EBCT-derived cortical perfusion and RBF increased significantly in response to sodium nitroprusside (P < 0.01 and P < 0.05, respectively), while medullary and papillary perfusions remained unchanged (Fig. 2B). GFR was also not modified by sodium nitroprusside. In HC, both cortical perfusion and RBF failed to increase (P = ns) in response to the drug, with a significant attenuation compared with the normal group (P < 0.01 and P < 0.05, respectively). Similar to the normal group, neither medullary and papillary perfusion nor GFR changed in response to sodium nitroprusside. Oxypurinol improved cortical perfusion response to sodium nitroprusside in HC pigs (Fig. 2B).
This study suggests that the endogenous xanthine oxidase system plays an important role in regulation of renal hemodynamics in HC. We observed that our hypercholesterolemic porcine model was characterized by increased activity of the XO system, as indicated by the elevated levels of uric acid, decreased urinary levels of xanthine, and increased renal protein expression of XO. Furthermore, acute blockade of this system in HC significantly improved cortical and restored medullary perfusion responses to both endothelium-dependent and -independent challenges.
HC is an established cardiovascular risk factor in humans (1) and is associated with impaired coronary artery epicardial and microvascular endothelium-dependent vasodilation (44) and myocardial perfusion and permeability (43). Similarly, we have shown that this condition can induce functional vascular changes in the kidney, both in vivo and in vitro (50), which precede the onset of overt atherosclerotic lesions (10, 20, 42). Subsequently, HC subjects may develop glomerulosclerosis (24), ox-LDL deposition (35, 36), and chronic tubulointerstitial damage (25). Increased oxidative stress appears to be one of the main pathogenic mechanisms mediating lipid-induced nephropathy, and involves increased generation of ROS, decreased NO bioavailability, and consequently endothelial dysfunction, a condition characterized by a blunted endothelium-dependent vasodilator response.
As we have observed before (8, 32), in the present study, HC pigs showed augmented oxidative stress, as demonstrated by the increase of plasma isoprostanes and ox-LDL compared with normal animals. In the presence of increased abundance of ROS, LDL particles can become oxidized to form a particularly atherogenic lipoprotein species (48). The main source of superoxide production in the arterial wall in HC has been identified in the endothelium and has implicated various ROS-generating enzymatic pathways such as NAD(P)H oxidase, uncoupled eNOS, and XO (28). Circulating XO has been suggested to be specifically involved in the mechanism of peripheral endothelial dysfunction in HC (6, 52). XO can play a crucial role in the generation of ROS in the kidney in pathological conditions, such as renal ischemia-reperfusion injury (23) and hypertension-induced renal hypertrophy (34), and in the present study, we assessed its involvement in HC-induced changes in renal hemodynamics. Indeed, we observed increased expression of XO throughout the kidney of HC pigs, especially in the cortical proximal and distal tubules. The significant increase in plasma levels of uric acid and decrease in xanthine levels in the urine also suggested increased activity of this enzyme.
Our previous studies in a porcine model of HC showed blunted kidney perfusion responses to both ACh and sodium nitroprusside (20). The present study extends our previous observation and shows that acute XO inhibition with oxypurinol improves both ACh and sodium nitroprusside-induced renal vasodilation. Interestingly, XO was mainly expressed in renal tubules, yet its inhibition improved renal perfusion responses, suggesting cross-talk between renal tubules and blood vessels (40) or possibly involvement of tubuloglomerular feedback in the hemodynamic effect of XO. The improvement in endothelial function might have resulted from decreased generation of ROS and thereby increased avaibility of NO. In addition, the blunted response to sodium nitroprusside in HC may result from increased vasoconstrictor activity, and the improved response to sodium nitroprusside in HCoxy may implicate XO in regulation of renal vasoconstrictors.
The HC model also exhibited increased plasma levels of uric acid, which may constitute an independent cardiovascular risk factor (29, 38, 47). Interestingly, uric acid has antioxidant properties, because it can scavenge superoxide, hydroxyl radical, and peroxynitrite (29), so that an increased level of uric acid in HC could represent a compensatory effect to the increased oxidative stress. However, contrarily, uric acid is also a prooxidant, because urate and urate metabolites can amplify lipid peroxidation and inflammation (29). Furthermore, in addition to the ROS produced during the reaction catalyzed by XO, evidence suggests that uric acid generated during the same reaction might also directly and independently contribute to endothelial dysfunction (17, 39).
Notably, oxypurinol improved, but not completely restored, cortical perfusion in HC pigs. Although the relative response to ACh was increased in all of the regions of the kidney in HCoxy (Fig. 2, A and B), the absolute values that cortical perfusion reached after ACh infusion (Table 3) were improved but remained significantly lower than those reached by normal or Noxy pigs. In contrast, the response of medullary and papillary perfusion was completely normalized. Because the expression of XO was similarly increased in all renal regions, these results imply a greater sensitivity of the medulla and papilla to XO-derived radicals. The incomplete improvement in the cortical perfusion and GFR responses could reflect activation of other ROS-generating enzymes, such as NAD(P)H oxidase or uncoupled eNOS, which could account for the persistence of impaired renal hemodynamic response. Indeed, we have previously demonstrated increased expression of NAD(P)H oxidase in the kidney of HC pigs (7). Alternatively, other mechanisms could be involved in the lack of complete restoration of endothelial function by oxypurinol, including elevated levels of asymmetric dimethylarginine, an endogenous competitive inhibitor of eNOS (5), decreased availability of the eNOS substrate l-arginine (18), or reduced availability of the eNOS cofactor tetrahydrobiopterin (49).
Previous studies have shown that a single dose of allopurinol or oxypurinol increased urinary xanthine clearance and decreased plasma uric acid (51) and that the dose that we used resulted in more than 90% inhibition of XO (15). However, it is possible that acute blockade of ROS formation by either XO or other mechanisms may not suffice to improve renal hemodynamics, because of chronic renal tissue injury that resulted from the HC diet (7, 8, 10, 42). This is supported by our recent study (9) demonstrating that in the stenotic kidney of hypertensive pigs, chronic administration of antioxidants can achieve greater improvement in renal hemodynamics than acute infusion of antioxidants.
Indeed, benefits of chronic inhibition of XO in cardiovascular disease continue to emerge (19). For example, a recent study (34) showed a protective blood pressure-independent effect of the blockade of XO in a rat model of hypertension-induced renal hypertrophy, suggesting a role for locally synthesized XO in the development of hypertension-associated end-organ damage. On the other hand, a 4-wk allopurinol administration did not affect endothelial function in forearm microcirculation in hypercholesterolemic patients (41), suggesting that the effects of chronic XO inhibition on cardiovascular and renal function may be variable and may depend on the model and on the treatment regimen.
In summary, we observed that a 12-wk experimental HC was associated with increased plasma level of uric acid and renal expression of XO. Short-term inhibition of XO using intrarenal oxypurinol infusion significantly improved renal perfusion response to an endothelium-dependent challenge, suggesting that oxypurinol may be renoprotective under conditions of increased oxidative stress. Therefore, this study suggests an important role for XO in the regulation of renal endothelial function in HC, although chronic tissue injury and other sources of ROS likely contribute to renal vascular dysfunction as well. Further studies will be needed to assess the relative contribution of these different systems to regulation of renal hemodynamics, function, and structure in HC.
This study was partly supported by National Institutes of Health Grants HL-63282 and HL-77131, the American Heart Association, and by the University of Study of Pisa, Italy.
The authors are grateful to the staff of the EBCT for their invaluable technical assistance with performance of scans.
↵* Both authors contributed equally to the manuscript.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2006 the American Physiological Society