Effects of voluntary wheel running on the kidney at baseline and after ischaemia–reperfusion-induced acute kidney injury: a strain difference comparison


N. C. Moningka: Department of Cellular & Molecular Physiology, Yale University, PO Box 208026, New Haven, CT, USA. Email: natasha.moningka@yale.edu

Key points

  • The impact of exercise on functional proteins critical for proper blood vessel health in the kidney is not well understood.

  • Using rats of different genetic backgrounds, we studied how chronic exercise affects the abundance of specific blood vessel proteins in the kidney, and whether this has any impact on the rat's susceptibility to acute kidney injury.

  • We found that the renal response to exercise is dependent on genetic background, and that in one strain exercise rendered the kidney more vulnerable to acute kidney injury.

  • The vulnerable rats exhibited exercise-dependent loss of the protective renal proteins critical for proper blood vessel health, while the protected strain showed increases in these protective proteins.

  • Our findings are particularly relevant regarding exercise prescription to kidney disease patients.

Abstract  Exercise-induced vascular endothelial adaptations in the kidney are not well understood. Therefore, we investigated the impact of voluntary wheel running (VWR) on the abundance of endothelial nitric oxide synthase (eNOS) and extracellular superoxide dismutase (EC SOD), in kidney and lung, and other SOD isoforms and total antioxidant capacity (TAC), in kidney. We also determined whether VWR influences susceptibility to acute kidney injury (AKI). Male Sprague–Dawley and Fisher 344 rats, VWR or sedentary for 12 weeks, were subjected to AKI (uninephrectomy (UNX) and 35 min of left kidney ischaemia–24 h reperfusion, IR). We measured glomerular filtration rate (GFR) and renal plasma flow (RPF), and analysed renal structural injury. Running was comparable between strains and VWR reduced body weight. In Sprague–Dawley rats, VWR reduced eNOS and EC SOD, but increased Mn SOD in kidney. Similar changes were seen after 6 weeks of VWR in Sprague–Dawley rats. In Fisher 344 rats, VWR increased eNOS, all SOD isoforms and TAC in kidney. Both strains increased eNOS and EC SOD in lung with VWR. Compared to UNX alone, UNX-IR injury markedly reduced renal function for both strains; however, in the Sprague–Dawley rats, VWR exacerbated falls in GFR and RPF due to UNX-IR, whereas in the Fisher 344 rats, GFR was unaffected by VWR. Some indices of renal structural injury due to UNX-IR tended to be worse in SD vs. F344. Our study demonstrates that genetic background influences the effect of exercise on kidney eNOS and EC SOD, which in turn influence the susceptibility to AKI.


acute kidney injury


extracellular superoxide dismutase


endothelial nitric oxide synthase




filtration fraction


Fisher 344 rat


glomerular filtration rate




nitric oxide synthase


plasma creatinine


renal blood flow


renal plasma flow


Sprague–Dawley rat






voluntary wheel running


Exercise improves cardiovascular health and many benefits result from metabolic adaptations, including reductions in plasma triglycerides, increases in the high-density lipoprotein to low-density lipoprotein ratio and improved insulin sensitivity (Sasaki & Gisele, 2006). There are also direct vascular benefits of exercise (Padilla et al. 2011), including shear stress-dependent increases in vasodilatory nitric oxide (NO) and the antioxidant extracellular superoxide dismutase (EC SOD) (Fukai et al. 2000; Davis et al. 2003; Harrison et al. 2006). These exercise-induced vascular changes have been primarily investigated in the endothelium of tissues where blood flow increases during exercise (i.e. exercising skeletal muscle, heart, lung; Padilla et al. 2011). In contrast, there are a number of tissues where blood flow remains unchanged or decreases during exercise; for example, renal blood flow (RBF) falls during exercise in a process that is partly dependent on increased sympathetic outflow (Poortmans, 1990; Tidgren et al. 1991; Mueller et al. 1998; Padilla et al. 2011). In kidney the impact of exercise on intravascular shear is difficult to predict as shear stress is determined not only by blood flow but also by vessel radius and blood viscosity (Padilla et al. 2011) and also because the renal vascular architecture is very complex (Moffatt & Fourman, 1963).

There is little and conflicting information on the impact of exercise on renal shear- regulated endothelial enzymes and endothelium-dependent relaxation. In rabbits, exercise (treadmill) training improved endothelium-dependent and endothelium-independent whole kidney vasodilation via NO (de Moraes, 2004). In contrast, an acute bout of high-intensity treadmill exercise decreased the renal endothelial NO synthase (eNOS) and NOS activity in exercise-trained rats whereas eNOS and NOS activity in lung (where blood flow increases with exercise) was enhanced (Miyauchi et al. 2003; Maeda 2004). Exhaustive exercise in the rat also decreased renal cortical NOS and SOD activity and caused significant renal tubular damage (Lin et al. 2010).

Renal injury due to exercise has also been reported in humans. Exercise-induced acute kidney injury (AKI) reflects renal ischaemia caused by rhabdomyolysis (the breakdown of muscle fibres) and/or severe volume depletion (Seedat et al. 1990; Bosch et al. 2009; Yan et al. 2010). Exercise-associated AKI is usually rare and associated with high-intensity exercise such as marathon running (Clarkson, 2007). However, subjects with renal hypouricaemia (a condition where the plasma antioxidant uric acid is low) frequently develop exercise-induced AKI (Ishikawa, 2002; Yan et al. 2010; Saito et al. 2011). Together, these studies suggest that in vulnerable individuals, exercise may increase susceptibility to an oxidative stress-mediated insult such as ischaemia–reperfusion (IR)-induced AKI, possibly connected to reduced renal eNOS/EC SOD. Furthermore, although exercise-induced AKI has been reported in humans, the effects of exercise on AKI outcomes is not known.

Therefore, we conducted the present study to determine whether 12 weeks of voluntary wheel running (VWR) protects or increases vulnerability against uninephrectomy (UNX) IR-induced AKI. We sought to determine (1) the impact of varying exposures to chronic VWR (6 vs 12 weeks) on abundance of renal cortex eNOS and EC SOD and (2) the impact of VWR on IR- induced AKI, an injury associated with reduced renal NO production and increased oxidative stress (Noiri et al. 2001). Our hypotheses were that (1) VWR would reduce renal eNOS and EC SOD secondary to exercise-induced reductions in renal vascular shear stress, and (2) loss of renal eNOS and EC SOD would render the exercised kidney more vulnerable to IR-induced AKI. We studied two different rat strains, as predisposition to age-dependent kidney damage (related to oxidative stress) is greater in the Sprague–Dawley (SD) vs. the Fisher 344 (F344) (Erdely et al. 2003; Moningka et al. 2011a, b). We speculated that the exercised SD rats might be more vulnerable to exercise exacerbation of IR-induced AKI than F344 rats. Studies were conducted in male SD rats allowed 6 or 12 weeks of VWR and F344 rats allowed 12 weeks of VWR, compared to sedentary (SED), age-matched controls. We measured renal cortex eNOS and EC SOD abundance in control kidneys of all rats with additional measurements of the other SOD's and total kidney cortex antioxidant capacity in the 12-week control kidneys. We also assessed the functional and histological responses to IR-UNX-induced AKI in both strains after 12 weeks of VWR.


Animal procedures

All animal handling was in accordance with and approved by the University of Florida's Institutional Animal Care and Use Committee, and conforms to the principles of UK regulations, as described in Drummond (2009). Male SD (n= 60) and F344 (n= 47) rats were purchased from Harlan Laboratories (Indianapolis, IN, USA) at 10–12 weeks of age. All rats were singly housed in a temperature- and light-controlled environment with ad libitum access to standard rat chow and water. Body weight (BW) was measured at weekly intervals in all rats throughout the studies.

In the first series of studies SD rats were allowed access to a VWR apparatus (Lafayette Instruments, Model 80859, Lafayette, IN, USA), available 24/7, in their home cages for 6 weeks (n= 7) and compared to age-matched controls (SED, n= 5). VWR activity was measured by attached odometers and acquired using the Activity Wheel Monitor Software (Lafayette Instruments). The 6-week VWR and SED rats had been implanted with telemetry probes ∼10 days prior to randomization into groups (see below) and mean blood pressure (BP) and heart rate (HR) were measured at baseline and then once a week for 6 weeks with continual recording over a 24 h period. Rats were removed from running wheels and killed under isoflurane anaesthesia within 30 min and the soleus muscle and kidney cortex were harvested and flash-frozen in liquid nitrogen. A further three SD rats were allowed access to VWR for 6 weeks and were compared to an additional SED SD group (n= 3) for kidney immunohistochemistry. At the end of the 6-week period, these rats were anaesthetized with isoflurane, the abdomen was opened and the aortic bifurcation cannulated and a blood sample was withdrawn; the kidneys were then perfused with cold PBS and the left kidney was removed and the cortex flash-frozen in liquid nitrogen. The perfusate was then switched to a 2% paraformaldehyde–lysine–periodate (PLP) and the right kidney perfused for 5 min. A slice of the perfused kidney was placed in 2% PLP for 24 h at 4°C, then transferred into PBS at 4°C until further immunohistochemical analysis (see below).

In the second series, SD and F344 rats were randomly divided into sedentary control (SED; SD n= 12, F344 n= 15) or voluntary exercise (VWR; SD n= 13, F344 n= 14) groups. Two days before the end of the 12-week period, rats were placed on a low nitrate, nutritionally complete diet (AIN-76C; MP Biomedicals, Solon, OH, USA). At the end of the 12 weeks rats were placed in a metabolic cage for 24 h of urine collection and ∼24 h after running cessation they were prepared for recovery surgery under fully sterile conditions and under general anaesthesia using isoflurane (5% induction and 1–2% maintenance dose). The left kidney was located through a lateral subcostal incision and the renal pedicle was clamped for 35 min. During the 35 min ischaemic period, we removed the right (normal, control) kidney. Cortical sections of the right kidney were frozen in liquid nitrogen and stored at −80°C for further analyses (see below). At 35 min the clamp was removed from the left kidney, incisions were closed and the rats were allowed to recover with left kidney reperfusion for 24 h. This produced a UNX-IR model of AKI. To control for the right UNX, additional studies were performed in all four groups with right UNX only to produce SD SED UNX (n= 10), SD VWR UNX (n= 7), F344 SED UNX (n= 9) and F344 VWR UNX (n= 9). All rats subjected to recovery surgery were given buprenorphine subcutaneously (0.01–0.05 mg kg−1) for analgesia.

Rats were prepared for inulin clearance studies 24 h after UNX-IR or UNX surgery for determination of glomerular filtration rate (GFR) and renal plasma flow (RPF). Rats were anaesthetized with Inactin (intraperitoneal injection, 120 mg kg−1 BW) and placed on a heated table to maintain a body temperature of 37 ± 1°C. The trachea was cannulated with PE-240 tubing and exposed to a constant flow of oxygen. Using PE-50 tubing filled with heparinized saline, the femoral artery was cannulated for measurement of BP and for collection of blood samples. A baseline BP was taken as well as a blood sample of 250 μl for measurement of creatinine. The femoral vein was cannulated and a 0.5 ml bolus of FITC-inulin (final concentration 2 mg ml−1 0.9% NaCl; Sigma) was infused; thereafter, FITC-inulin was infused at a rate of 1.2 ml (100 g BW)−1 h−1. To ensure that the rat was euvolaemic, artificial plasma containing 5% bovine serum albumin/γ-globulin was infused at a rate of 1% BW h−1 for the first 15 min, and at 0.15% 100 g BW h−1 for the remainder of the study. Next, the abdominal cavity was opened (midline incision) to expose the left kidney and bladder. A non-occluding catheter was placed in the left renal vein to obtain an arterial–venous inulin extraction for calculation of RPF. We chose this approach rather than the traditional para-aminohippuric acid method as AKI reduces para-aminohippuric acid extraction, which leads to an underestimation of RPF (Corrigan et al. 1999; Di Gusto et al. 2008). The bladder was cannulated with flanged PE-50 tubing for collection of urine and for gravimetric determination of urine volume and flow rate. Levels of anaesthesia, BP and body temperature were monitored throughout. After a 60 min stabilization period, two 20 min urine collections were made with mid-point collections of femoral arterial and renal venous blood for plasma FITC-inulin measurement. Urine samples were also analysed for FITC-inulin which together with plasma inulin concentrations were used to calculate GFR (from inulin clearance), RPF (from [urine inulin concentration/plasma arterous–venous inulin extraction]/urine flow) and filtration fraction (FF, from [GFR/RPF]). The rat was killed by exsanguination under anaesthesia, and the lung and left kidney were removed. A portion of the kidney was prepared for histology (see below), and the remainder was divided into sections of cortex, which together with lung were frozen in liquid nitrogen and stored at −80°C for further analyses.


PLP-perfused kidneys were embedded in polyester wax and 5 μm sections (2–4 per kidney) were mounted onto glass slides, dewaxed and peroxidase-blocked for 45 min then washed with distilled water. To reduce background, sections were steamed in antigen retrieval solution (DAKO, Carpinteria, CA, USA) for 30 min, cooled for 20 min, then blocked with protein blocker (DAKO) for 15 min. Sections were placed in a humidified tray and incubated overnight with the mouse monoclonal eNOS antibody (1: 5000; BD Transduction, Franklin Lakes, NJ, USA) at 4°C, washed with PBS, incubated for 30 min with one drop of MACH2 mouse horseradish peroxidase (HRP)-polymer secondary antibody (Biocare Medical, Concord, CA, USA), washed again with PBS, and then incubated with diaminobenzidine for 5 min. Sections were then dehydrated with xylene, mounted onto cover slips using Eukitt mounting medium (Sigma, St Louis, MO, USA), and allowed to dry flat prior to observation by light microscopy.

Preparation for telemetry

In a preliminary operation, under isoflurane anaesthesia and using a fully sterile technique, a catheter was fed under the skin by trocar and introduced into the left femoral artery. The catheter was tied into position, and the C40 transmitter unit was sutured to the internal abdominal wall. Rats were singly housed and allowed to recover for ∼10 days. BP and HR were measured using Data Sciences International (New Brighton, MN, USA) equipment and software.

Renal pathology

As previously described (Moningka et al. 2011a, b), kidney tissues were fixed in 10% formalin, processed, embedded in paraffin wax, cut into 5 μm sections and stained with periodic acid Schiff and counterstained with haematoxylin and eosin (Sigma). Tissue sections were scored blinded for acute tubular necrosis (by B.P.C.) with specific evaluation of tubule cell swelling, brush border loss, nuclear condensation, karyolysis (dissolution of the nucleus), regeneration, capillaritis (endothelial inflammation) and cell sloughing. Each category was given a score as follows: A grade of ‘0’ signified that none of the tubules was involved, 1 =≤10%, 2 = 11–25%, 3 = 26–50%, 4 = 51–75% and 5 = 76–100% were involved.

Western blot

The relative protein abundance of eNOS (BD Transduction; 1:250), SOD isoforms (Stressgen Reagents; EC SOD 1:250, CuZn SOD 1:2000 and Mn SOD 1:2000) in renal cortical tissue and whole lung were measured by Western blot as previously described (Moningka et al. 2012). Briefly, 200 μg of homogenized tissue was separated by electrophoresis (7.5 or 12% acrylamide gel, 140 V, 65 min), and then transferred onto nitrocellulose membranes (GE Healthcare, Little Chalfont, UK) for 60 min at 0.18 A. Membranes were stained with Ponceau Red (Sigma) to check for transfer efficiency/uniformity and equal loading, incubated in blocking solution for 60 min and washed in TBS + 0.05% Tween before overnight primary antibody incubation at 4°C. Membranes were then incubated with the appropriate HRP-tagged secondary antibody for 1 h, washed and developed with enhanced chemiluminescent reagents (Thermo Scientific, Waltham, MA, USA). Bands were quantified by densitometry using the VersaDoc Imaging System and One Analysis Software (BioRad, Hercules, CA, USA). Protein abundance was calculated as integrated optical density (IOD) of the protein of interest, factored for Ponceau Red stain (to determine total protein loaded) and positive control. As the relative densitometry units are arbitrary, the SED control values were set at 100.

Analytical methods

FITC-inulin concentrations (mg ml−1) in urine and plasma were measured by fluorescence (excitation and emission wavelengths, 485 and 530 nm, respectively) as previously described (Knight et al. 2007). Creatinine concentrations (mg dl−1) in plasma were measured by high-performance liquid chromatography (Sasser et al. 2009). To measure the total antioxidant capacity levels in kidney cortex, we used Cayman's Total Antioxidant Assay kit. Urine NOx (NO3+ NO2) content was measured by Griess assay, and urine H2O2 was measured using Molecular Probes’ Amplex Red Hydrogen Peroxide/Peroxidase Assay Kit according to the manufacturer's recommendations (Carlsbad, CA, USA) and as previously described (Moningka et al. 2012). Urine protein concentration was measured by the Bradford method (BioRad). Citrate synthase activities in soleus tissue in units of μm min−1 g−1 wet tissue weight were based on methods adapted by Srere (1969).

Statistical analyses

All data are presented as means ± standard error and analysed using GraphPad Prism software (San Diego, CA, USA). Running activity, body weight and protein abundance data were analysed by Student's t test. Functional and urinary analytical data were analysed by two-way analysis of variance (ANOVA), followed by the Bonferonni's correction method as the post hoc test to correct for multiple comparisons. Histological analyses were by the non-parametric Mann–Whitney U test. Significance was defined as P < 0.05.


Running activity in both strains is given in Fig. 1 for the series 2 experiments where rats were exposed to 12 weeks of VWR. For SD rats, running increased gradually over the first 3 weeks of exposure to VWR and by week 4 a maximum value was reached (Fig. 1A). In series 2, running activity for SD rats was unchanged for the remainder of the study. As shown in Supplemental Fig. S1A, the running activity of the SD rats allowed 6 weeks of VWR (Series 1) followed a similar pattern to that shown in the series 2 SD rats (Fig. 1A). The majority of activity occurred during the dark cycle in both strains (Supplemental Fig. S2). As also shown in Fig. 1, BW increased with age in all rats but in both strains the rate of rise was attenuated with VWR when compared to SED rats. The rate of rise of BW was also attenuated in the 6 week VWR vs. SED SD rats (Supplemental Fig. S1B). There was an elevation in soleus muscle citrate synthase activity (19.1 ± 1.2 vs. 24.3 ± 0.6 μm min−1 g−1 wet weight in SED vs. VWR; P < 0.05), indicating that a training effect was achieved with 6 weeks of VWR. No significant change in BP was noted between SED and VWR rats but there was a significant reduction in HR following 6 weeks of VWR (Supplemental Fig. S3).

Figure 1.

Running activity patterns versus body weight 
Running activity patterns over 12 weeks and its impact on body weight in (A) Sprague–Dawley (SD; SED n= 22, VWR n= 21) and (B) Fisher 344 (F344; SED n= 21, VWR n= 23) rats. For each strain, the total average distance (km) run per day steadily increased over the first few weeks, and remained relatively constant for the remainder of the study. Voluntary wheel running (VWR) effectively reduced body weight in both strains compared to sedentary (SED) controls. #P < 0.05 compared to baseline; *P < 0.05 between the two groups.

As shown in Fig. 2 the kidney cortex abundance of both eNOS and EC SOD was greatly reduced by 6 weeks of VWR vs. the SED SD. By immunohistochemistry, eNOS was detected only in the endothelial lining of blood vessels of the kidney (Fig. 3) and 6 weeks of VWR did not change eNOS localization, although eNOS staining decreased in intensity in SD rats with VWR.

Figure 2.

Six weeks of voluntary exercise on endothelial nitric oxide and extracellular superoxide dismutase in kidney cortex 
Effect of 6 weeks voluntary exercise (VWR) on endothelial nitric oxide (eNOS) and extracellular superoxide dismutase (EC SOD) in kidney cortex of Sprague–Dawley (SD) rats (SED, n= 5; VWR, n= 7). Six weeks of VWR significantly reduced kidney cortex protein abundance levels of eNOS and EC SOD in the SD rat. *P < 0.05 compared to SED.

Figure 3.

Six weeks of voluntary wheel running on immunoreactivity of endothelial nitric oxide synthase in kidney 
Effect of 6 weeks voluntary wheel running (VWR) on the immunoreactivity of endothelial nitric oxide synthase (eNOS) in the kidney of Sprague–Dawley rats (SED, n= 3 (A); VWR, n= 3 (B)). For both groups, eNOS predominately localizes to the endothelium of vessel walls (indicated by arrowheads). Localization was unchanged after 6 weeks of VWR in VWR vs. SED.

As shown in Fig. 4A, both eNOS and EC SOD abundance in kidney cortex were also decreased by 12 weeks VWR vs. SED SD. The magnitude of the decline was less than in the 6 week VWR rats, perhaps reflecting the 24 h delay after cessation of exercise in 12 week rats. As shown in Fig. 4B, the kidney cortex eNOS and EC SOD response to 12 weeks of VWR in F344 rats was opposite to that seen in SD rats, with substantial increases occurring in the abundance of both enzymes. In contrast to strain-related differences in the kidney cortex, homogenate of lung showed a significant rise in both eNOS and EC SOD abundance from VWR compared to SED in both SD and F344 rats (Fig. 5). Of note, the F344 rats showed a similar 12 week VWR pattern to the SD rats (Fig. 1) and although F344 rats weighed less than SD rats, the rate of rise in BW with age was again attenuated with VWR, suggesting a training effect.

Figure 4.

Twelve weeks voluntary exercise on endothelial nitric oxide and extracellular superoxide dismutase in kidney cortex 
Effect of 12 weeks voluntary exercise (VWR) on endothelial nitric oxide (eNOS) and extracellular superoxide dismutase (EC SOD) in kidney cortex of (A) Sprague–Dawley (SD; SED n= 12, VWR n= 12) and (B) Fisher 344 (F344; SED n= 12, VWR n= 12) rats. Twelve weeks of VWR significantly reduced kidney cortex eNOS and EC SOD protein abundance in the SD rat, whereas in the F344 rat, both proteins markedly increased. *P < 0.05 compared to SED.

Figure 5.

Twelve weeks of voluntary exercise and renal uninephrectomy and ischaemia–reperfusion injury on endothelial nitric oxide synthase and extracellular superoxide dismutase in lung 
Effect of 12 weeks of voluntary exercise (VWR) and renal uninephrectomy (UNX) and ischaemia–reperfusion (IR) injury on endothelial nitric oxide synthase (eNOS) and extracellular superoxide dismutase (EC SOD) in lung of (A) Sprague–Dawley (SD; SED n= 6–12, VWR n= 6–12) and (B) Fisher 344 (F344; SED n= 6–12, VWR n= 6–12) rats. For both strains, VWR significantly increased eNOS and EC SOD in lung tissues harvested after 24 h of UNX-IR. *P < 0.05 between the two groups.

In the renal cortex, 12 weeks of VWR in the SD rats had no effect on abundance of cytosolic (CuZn) SOD but produced a marked rise in mitochondrial (Mn SOD) abundance. Total antioxidant capacity (TAC) showed no change with VWR (Fig. 6A) in the SD rats, presumably reflecting offsetting changes in the renal cortex EC SOD and Mn SOD abundances. In the F344 kidney cortex 12 weeks of VWR led to increases in both CuZn and Mn SOD and TAC increased (Fig. 6B), reflecting increases in the abundance of all three SOD isoforms with exercise. There were no differences in total NO production (indicated by 24 h urinary excretion of NOx), in total H2O2 production (inline image) or in the urinary protein excretion (UprotV) between SED and 12 weeks of VWR rats of either strain, or between strains (Supplemental Tables S1 and S2).

Figure 6.

Twelve weeks of voluntary wheel running on CuZn superoxide dismutase, manganese SOD and total antioxidant capcity in the kidney cortex 
Effect of 12 weeks voluntary wheel running (VWR) on CuZn superoxide dismutase (CuZn SOD), manganese SOD (Mn SOD) and total antioxidant capcity (TAC) in the kidney cortex of (A) Sprague–Dawley (SD; SED n= 6–12, VWR n= 6–12) and (B) Fisher 344 (F344; SED n= 6–12, VWR n= 6–12) rats. Twelve weeks of VWR increased CuZn SOD in the F344 rats only and increased Mn SOD for both strains. No differences in TAC were detected in SD rats; however, in F344 rats, 12 weeks of VWR increased TAC. *P < 0.05 between SED and VWR.

As illustrated in Table 1, the renal haemodynamic parameters to 24 h UNX in the SD rats were similar between SED and VWR groups. When IR was superimposed on UNX, much greater falls in GFR and RPF occurred in both SED and VWR SD groups, and plasma creatinine (PCr) rose markedly. Of note, the fall in GFR and rise in PCr were exacerbated in the VWR SD rats subjected to UNX-IR compared to SED UNX-IR SD rats, due to greater falls in RPF and FF in the VWR group.

Table 1.  Renal haemodynamic responses of Sprague–Dawley rats
UNX (n= 7)UNX-IR (n= 8)UNX (n= 6)UNX-IR (n= 7)
  1. GFR, glomerular filtration rate; RPF, renal plasma flow; FF, filtration fraction; PCr, plasma creatinine. *P < 0.05 vs. respective UNX, †P < 0.05 vs. SED UNX-IR.

GFR (ml min−1 (100 g BW)−1)0.47 ± 0.080.17 ± 0.04*0.45 ± 0.030.04 ± 0.01*†
RPF (ml min−1 (100 g BW)−1)2.74 ± 0.820.96 ± 0.23*1.83 ± 0.260.30 ± 0.15*†
FF0.24 ± 0.020.19 ± 0.030.28 ± 0.030.12 ± 0.02*
PCr (mg ml−1)0.28 ± 0.021.07 ± 0.27*0.27 ± 0.032.42 ± 0.35*†

In the F344 rat, as in the SD rat, the renal haemodynamic response to UNX was similar in the SED and VWR groups (Table 2). Again, combination UNX and IR led to increases in PCr and marked falls in GFR, RPF and FF compared to UNX alone. In contrast to SD rats, however, there was no difference in the severity of the falls in GFR and RPF, and an increase in PCr in the VWR vs. SED F344 rats.

Table 2.  Renal haemodynamic responses of Fisher 344 rats
UNX (n= 8)UNX-IR (n= 8)UNX (n= 6)UNX-IR (n= 5)
  1. GFR, glomerular filtration rate; RPF, renal plasma flow; FF, filtration fraction; PCr, plasma creatinine. *P < 0.05 vs. respective UNX.

GFR (ml min−1 (100 g BW)−1)0.47 ± 0.030.06 ± 0.01*0.58 ± 0.040.16 ± 0.06*
RPF (ml min−1 (100 g BW)−1)1.84 ± 0.230.40 ± 0.10*2.53 ± 0.261.02 ± 0.35*
FF0.28 ± 0.020.20 ± 0.02*0.25 ± 0.020.21 ± 0.05
PCr (mg ml−1)0.25 ± 0.041.77 ± 0.19*0.29 ± 0.021.78 ± 0.28*

The combination of UNX and IR led to structural changes in kidney cortex characteristic of AKI, and in the SD rats there was significantly greater brush border loss in the VWR vs. SED group (Fig. 7A). In the F344 rat, UNX-IR generally affected SED and VWR groups to the same degree, although there was less cell sloughing in the VWR rats (Fig. 7B).

Figure 7.

Twelve weeks of voluntary exercise and renal uninephrectomy ischaemia–reperfusion injury on indices of acute renal structural injury in kidney cortex 
Effect of 12 weeks voluntary exercise (VWR) and renal uninephrectomy (UNX) ischaemia–reperfusion (IR) injury on indices of acute renal structural injury in kidney cortex of (A) Sprague–Dawley (SD; SED n= 12, VWR n= 12) and (B) Fisher 344 (F344; SED n= 12, VWR n= 12) rats. Analyses were conducted on IR kidneys only as minimal to no injury was present with UNX alone. In the SD rat, exercise and IR tended to increase all markers of acute injury and significantly increased the level of brush border loss. In the F344 rat, IR affected SED and VWR groups to the same degree; however, 12 weeks of VWR ameliorated the level of sloughing against IR. *P < 0.05 compared to SED IR.


The main novel findings in this study are that whereas chronic VWR causes directionally opposite changes in the normal kidney renal cortex abundance of eNOS and EC SOD in SD vs. F344 rats, in the lung, increases in eNOS and EC SOD are seen in both strains. The reductions in eNOS and EC SOD that occur in the normal kidney of SD rats correlates with an exercise-induced exacerbation of UNX-IR-induced falls in GFR. In contrast, the F344 rats exhibit increased eNOS and EC SOD in the normal kidney cortex and in these rats there is no exercise-induced exacerbation of the IR injury. These observations support the hypothesis that select strains or groups of individuals may be susceptible to renal damage when exposed to prolonged exercise.

There is substantial evidence that exercise provides benefit in skeletal muscle and coronary blood vessels, and protects the heart against IR-induced injury (Fogarty et al. 2004; Powers et al. 2007; Sindler et al. 2009). Several mechanisms are involved in the cardiac protection, including the up-regulation of myocardial antioxidant capacity, which helps combat the increased content of reactive oxygen species, lipid peroxides, protein oxidation and protein nitration associated with myocardial IR (Powers et al. 2007). Pulmonary blood flow (and presumably shear stress) increases with exercise, and as shown in the present study, eNOS and EC SOD abundance in the lung increase with exercise in both SD and F344 rats. In the kidney, however, blood flow is reduced during exercise, primarily due to increased renal sympathetic nerve activity (Tidgren et al. 1991) via activation of the α-adrenergic receptors (Mueller et al. 1998). Angiotensin II, endothelin-1 and vasopressin are also involved (Stebbins & Symons 1993, 1995; Ahlborg et al. 1995). At low intensities, GFR is well maintained due to compensatory increases in filtration fraction; however, at high intensities, RBF falls markedly and GFR falls (Poortmans & Vanderstraeten, 1994).

If a fall in RBF causes a reduction in shear stress within the renal vasculature, falls in renal eNOS and EC SOD protein abundance are expected. Indeed, Miyauchi et al. (2003) report a fall in renal eNOS mRNA, protein abundance and enzyme activity with an acute bout of treadmill exercise in trained Wistar rats. Exhaustive exercise leading to functional and structural kidney damage in the untrained SD rat also decreased renal cortical NOS and SOD activity (Lin et al. 2010). In the present study, we observed a reduction in renal eNOS and EC SOD in the SD rat after 6 and 12 weeks of VWR. This was in contrast to the F344 rat, where these beneficial enzymes increased with 12 weeks of VWR. This finding in F344 rats confirmed our previous report of increased kidney cortex eNOS and EC SOD in the F344 rat using forced treadmill running as an alternative exercise modality (Moningka et al. 2011a, b).

The difference in the eNOS and EC SOD renal cortical response to exercise between normal, young adult males of two strains was a surprising finding. It is unlikely to reflect different psychological stress responses (which could be a concern for forced exercise) as VWR is voluntary. Both SD and F344 rats exhibit intensity-dependent falls in RBF with treadmill running (Armstrong & Laughlin, 1984; McAllister, 1998), although there has been no direct comparison between the two strains. If the F344 rat has a more efficient cardiac output response to low-intensity exercise vs. SD rats, they may not exhibit a fall in RBF with low-intensity VWR; however, this remains to be determined. Alternatively, both strains may undergo renal vasoconstriction and falls in RBF with VWR, but the pattern of intrarenal shear stress may vary, as shear stress depends on vessel radius and local viscosity as well as flow. The renal circulation is very complex with intricate branching patterns (Moffat & Fourman, 1963), and there may be architectural differences within the renal vasculature of the two strains that create different local shear responses. These are intriguing possibilities that merit further study as genetic background may also determine the renal eNOS and EC SOD responses to exercise in humans. If so, knowledge of the renal exercise phenotype would be important in determining recommended exercise intensity and modality.

Falls in NO bioavailability can render the kidney susceptible to injury as NO is critical for normal renal function and its deficiency leads to chronic kidney disease (Baylis, 2007). In addition, endothelial injury contributes to IR-induced AKI. There is recruitment of local inflammatory signals, which stimulate reactive oxygen species leading to reductions in NO bioavailability and development of renal dysfunction (Le Dorze et al. 2009). Several studies have reported cases of exercise-induced renal ischaemia-induced AKI (Seedat et al. 1990; Bosch et al. 2009; Yan et al. 2010) and rhabdomyolysis, which leads to myoglobin-induced AKI (Bosch et al. 2009). Therefore, we conducted the present study to determine whether 12 weeks of VWR protects or increases vulnerability against UNX-IR-induced AKI.

Despite comparable running activities and BW responses to chronic VWR in the two rat strains, exercise in the SD rat rendered the kidney susceptible to IR-induced AKI with exacerbation of the falls in GFR and RPF, and increased PCr compared to SED SD rats. Thus, the exercise-induced loss of renal eNOS and EC SOD in this strain was associated with an amplified AKI in response to UNX-IR. In contrast, in the F344 rat, 12 weeks of VWR increased renal eNOS and EC SOD abundance and VWR had no impact on the decline in renal function with IR in these rats, although there was minimal protection against structural damage. A limitation to this study is the potential confounding influence UNX may create on the pathogenesis of IR. In addition, it is also possible that clamping of the left kidney may create hyperfiltration to the contralateral, right kidney, whose purpose was to serve as the normal, control kidney.

The strain-dependent differences in renal cortex eNOS and EC SOD with exercise are specific to the kidney as lung enzymes showed increases with exercise in both strains and there was no difference in total systemic NO or H2O2 production. The net renal response to exercise in SD rats was a reduction in eNOS and no change in TAC, whereas in F344 rats there was increased eNOS and increased TAC. Thus, these data suggest that the exercise-induced exacerbation of IR AKI in the SD rat is related to the reduced eNOS and unchanged TAC in the kidney cortex prior to the acute insult.

In conclusion, this study provides evidence that genetic background influences the effect of exercise on kidney eNOS and EC SOD, which in turn influence the susceptibility to IR-induced AKI. Although UNX-IR injury impaired renal function in both SED and VWR groups of both strains, in the SD rat VWR exacerbated falls in GFR, RPF and PCr. Important questions remain: What causes the different renal endothelial response to exercise in the two strains? There could be differences in the renal haemodynamic responses or perhaps different renal metabolic responses to exercise. Also, do these variable renal endothelial responses to exercise also occur in humans and, if so, how can they be predicted? In addition, there could be differences in sympathetic activation in response to exercise, which suggest that the impact of exercise on renal outcomes can be manipulated by denervation. Future studies should address these areas.

Translational perspective

Renal failure afflicts a growing population in the US and according to the 2010 US Renal Data System Annual Report these numbers are expected to grow. Benefits of physical activity are critical for this population given the numerous cardiovascular-related benefits. For patients with kidney disease, the National Kidney Foundation recommends to exercise at least 3 days a week for ∼30 min a session, and that it must consist of either continuous movement of large muscle groups (i.e. walking, swimming, cycling and aerobic dancing), or low-level weight-bearing exercise (i.e. high repetition of low weight lifting). They also recommend that participation in any sort of exercise programme must consider type, duration, frequency and intensity of exercise. Indeed, potential risks with exercise should be considered. The data in this body of work are provoking in that we have discovered certain settings where exercise can have negative impacts on renal injury. Genetic background is clearly a major factor, which is highly clinically relevant as there is huge influence of genetics in development of renal injury in humans. It is also important to note that the purpose of these studies was not to support discontinuation of exercise prescription for the renal disease population, but to underline the importance of better understanding the risk versus benefit of exercise in renal disease patients.


Author contributions

Corresponding author N.C.M was involved in data collection, analysis, interpretation and manuscript preparation, and provided intellectual content. Studies conducted on 6 week SD rats was a collaboration between L.F.H, J.K.A, and J.W.V., and all experiments were conducted in their laboratories. Co-authors L.F.H., J.W.V., M.W.C., C.A.W, J.K.A. and M.S. were also involved in some data collection, provided intellectual content and helped with manuscript preparation. Co-author B.C. evaluated renal pathology, provided intellectual content and revised the manuscript critically prior to submission. C.B. was responsible for the conception and design of the study, and provided critical assessment of all aspects of the data analyses and preparation of the manuscript.


We thank Harold Snellen, Bruce Cunningham and Florence Whitehill (University of Florida, College of Medicine Electron Microscopy Core Facility) for excellent technical assistance. This work was funded by NIH RO1 DK56843 (C.B.) and NIH RO1 HL76518 (L.F.H). N.C.M. was supported by NIH T32DK076541, M.W.C by NIH T32DK751825, J.K.A. by a Postdoctoral Fellowship from the American Heart Association and M.S. by 5R25HL103181.