Am J Physiol Renal Physiol 307: F251–F262, 2014. First published June 4, 2014; doi:10.1152/ajprenal.00578.2013.

CALL FOR PAPERS

Novel Therapeutics in Renal Disease

Chronic bilateral renal denervation attenuates renal injury in a transgenic rat model of diabetic nephropathy Yimin Yao,1,4 Ingrid C. Fomison-Nurse,1 Joanne C. Harrison,1 Robert J. Walker,2 Gerard Davis,3 and Ivan A. Sammut1 1

Department of Pharmacology, University of Otago, Dunedin, New Zealand; 2Department of Medicine, University of Otago, Dunedin, New Zealand; 3Department of Physiology, University of Otago, Dunedin, New Zealand; and 4Australian School of Advanced Medicine, Macquarie University, New South Wales, Australia Submitted 1 November 2013; accepted in final form 2 June 2014

Yao Y, Fomison-Nurse IC, Harrison JC, Walker RJ, Davis G, Sammut IA. Chronic bilateral renal denervation attenuates renal injury in a transgenic rat model of diabetic nephropathy. Am J Physiol Renal Physiol 307: F251–F262, 2014. First published June 4, 2014; doi:10.1152/ajprenal.00578.2013.—Bilateral renal denervation (BRD) has been shown to reduce hypertension and improve renal function in both human and experimental studies. We hypothesized that chronic intervention with BRD may also attenuate renal injury and fibrosis in diabetic nephropathy. This hypothesis was examined in a female streptozotocin-induced diabetic (mRen-2)27 rat (TGR) shown to capture the cardinal features of human diabetic nephropathy. Following diabetic induction, BRD/sham surgeries were conducted repeatedly (at the week 3, 6, and 9 following induction) in both diabetic and normoglycemic animals. Renal denervation resulted in a progressive decrease in systolic blood pressure from first denervation to termination (at 12 wk post-diabetic induction) in both normoglycemic and diabetic rats. Renal norepinephrine content was significantly raised following diabetic induction and ablated in denervated normoglycemic and diabetic groups. A significant increase in glomerular basement membrane thickening and mesangial expansion was seen in the diabetic kidneys; this morphological appearance was markedly reduced by BRD. Immunohistochemistry and protein densitometric analysis of diabetic innervated kidneys confirmed the presence of significantly increased levels of collagens I and IV, ␣-smooth muscle actin, the ANG II type 1 receptor, and transforming growth factor-␤. Renal denervation significantly reduced protein expression of these fibrotic markers. Furthermore, BRD attenuated albuminuria and prevented the loss of glomerular podocin expression in these diabetic animals. In conclusion, BRD decreases systolic blood pressure and reduces the development of renal fibrosis, glomerulosclerosis, and albuminuria in this model of diabetic nephropathy. The evidence presented strongly suggests that renal denervation may serve as a therapeutic intervention to attenuate the progression of renal injury in diabetic nephropathy. renal denervation; diabetic nephropathy; (mRen-2)27 rat; fibrosis

leading cause of end-stage renal disease worldwide, and effective interventional strategies are desperately required (49). The characteristics of diabetic nephropathy manifest as albuminuria, glomerular basement membrane (GBM) thickening, mesangial hypertrophy, and tubulointerstitial fibrosis (8, 18, 44). Hypertension is often

DIABETIC NEPHROPATHY IS THE

Address for reprint requests and other correspondence: I. A. Sammut, Dept. of Pharmacology and Toxicology, Univ. of Otago School of Medical Sciences, PO Box 56, Dunedin, Otago 9016, NZ (e-mail: [email protected]). http://www.ajprenal.org

present, forming a significant factor in the progression of diabetic nephropathy (56). Elevation of sympathetic nerve activity (SNA) has been shown to occur in patients with diabetic nephropathy (47) and is established as a major contributor in the onset of hypertension and renal failure (30, 58). Furthermore, the pharmacological inhibition of SNA at doses which do not affect hypertension has been shown to effectively reduce the progression of renal damage in both clinical diabetic nephropathy (52) and in experimental diabetic nephropathy models (1). These findings appear to indicate that increased SNA may instigate and even accelerate the development of diabetic renal injury independently of its hypertensive effect. As renal nerves play a significant role in controlling SNA (48), diseased kidneys are therefore able to exert an excitatory effect on the SNA in patients with chronic renal failure (11, 26). Earlier studies using rat models of experimental nephropathy have demonstrated that renal sensory nerve activity from the diseased kidneys to the brain is raised, resulting in an increased posterior hypothalamic noradrenaline turnover and renal SNA elevation (RSNA) (9, 10, 62– 64). As the kidneys are richly supplied by renal sympathetic nerves, an increase in RSNA will further deteriorate the function of the diseased kidneys by increasing renin synthesis and release from juxtaglomerular cells (28), thereby provoking an increase in renal Na⫹ and water reabsorption (14) and renal vascular resistance (15), in what is described as a vicious cycle (60). By effectively interrupting this excitatory cycle, bilateral renal denervation (BRD) has reemerged as an effective therapy shown to delay the progression of hypertension. In several clinical trials, BRD was demonstrated to provide long-term blood pressure reduction in patients with difficult-to-control and resistant hypertension (16, 31, 38, 43, 51, 53). The unexpected failings of the latest clinical study (SYMPLICITY HTN-3) to show efficacy have, unfortunately, cast some doubt on the value of renal denervation in resistant hypertension (55). However, BRD may have renoprotective effects in patients beyond simple blood pressure control. BRD has been found to improve glucose tolerance and insulin sensitivity (43). In a rat model of early type 1 diabetes, BRD was also found to abolish glomerular hyperfiltration and prevent glomerular hypertrophy at 14 days after diabetic onset (42). Taken together, these findings suggested that renal nerves are involved in the early renal pathological changes occurring during the development of diabetic nephropathy. However, the chronic effects of BRD

1931-857X/14 Copyright © 2014 the American Physiological Society

F251

F252

RENAL DENERVATION IN DIABETIC NEPHROPATHY

on the progression of diabetic nephropathy have not been fully examined. In this study, we examined the hypothesis that BRD conducted after the onset of hyperglycemia will preserve renal morphology and function and reduce the severity of hypertension and expression of renal fibrotic markers formed as a consequence of diabetic induction. Furthermore, as previous reports have shown inconsistencies in the level of intrarenal distribution of angiotensin II type 1 receptors (AT1R), particularly as renin and angiotensin-converting enzyme content in the rodent nephron may be variably altered in diabetes (35), this study also examined the effect of denervation on renal AT1R in diabetic nephropathy. The underlying features of the hypertensive transgenic (mREN-2)27 rat (TGR) model include elevated prorenin levels and high tissue renin-angiotensin system activity that are critical to the development of diabetic nephropathy in humans (4). Previous reports have shown that streptozotocin (STZ)-induced diabetes in female heterozygous TGRs results in a rapid progression of the principal pathological features of human diabetic nephropathy without the prominent effects of hypertension on renal pathology (3, 4, 32). Consequently, diabetic nephropathy was studied in this female TGR model over a 12-wk period following hyperglycemic onset at 6 wk of age. The present studies were performed to investigate the therapeutic benefit of BRD on the progression of diabetic nephropathy in the female heterozygous, reninoverexpressing TGR model of STZ-induced hyperglycemia. METHODS

Animal model and research design. All experimental procedures involved in this project were conducted under the approval of the University of Otago Animal Ethics Committee and in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals. Female heterozygous TGRs were obtained from the CARA Transgenic Facility at the University of Otago. Rats received either an intravenous bolus of a citrate-buffered vehicle (0.05 M) or STZ (55 mg/kg) to induce diabetes at 6 wk of age as described (32). Blood glucose levels were monitored and maintained at 20 –25 mM by daily administration of isophane (sc, 1– 4 units). Animals in both normoglycemic and diabetic groups were randomized to either BRD or sham denervation surgeries performed repetitively in each animal at postnatal weeks 9, 12, and 15 (Fig. 1). Systolic blood pressure (SBP) and heart rate were measured weekly following habituation, using tail-cuff plethysmography and 24-h urine collections obtained at specific intervals. Upon termination at postnatal week 18, the animals were anesthetized (halothane-N2O2-O2 mix), and the kidneys were harvested and stored for further histological and biochemical processing. In a separate set of treatment groups, the glomerular filtration rate (GFR) was measured using the inulin clearSTZ/vehicle

6

BRD/Sham

9

12

Termination

15

18 weeks (age)

Fig. 1. Experimental design and confirmation of successful diabetic induction and chronic bilateral renal denervation (BRD). Streptozotocin (STZ; iv) was administered to 6-wk-old female transgenic rats (TGRs) allocated to the diabetic group. BRD or sham surgeries were performed, repetitively, at 9, 12, and 15 wk of age. Weekly systolic blood pressure recordings (Œ) and 24-h urine collections (o) were performed. At 18 wk of age, animals were euthanized, and kidneys were collected for immunohistochemical and Western blot analysis.

ance method and corrected to the left kidney weight as previously described (24). BRD. As renal nerves in rats have been shown to fully reinnervate the kidney within 3– 6 wk after surgical denervation, it was necessary to conduct repeat denervation surgeries (34, 45). At week 3 after STZ administration, a midline dorsal lumbar incision was made bilaterally under general anesthesia (halothane-N2O2-O2 mix), and each kidney exposed by retracting the abdominal muscles. Positive identification of the renal nerves was confirmed using a stimulator (10 V/10 Hz, 1-ms square wave pulse) to produce temporary renal blanching. Subsequently, the renal arteries were stripped back of the surrounding nerve bundles surrounding the renal arteries and painted with 10% phenol in absolute ethanol to destroy any remnant nerves. Subsequent denervation surgeries were similarly conducted. Sham renal denervation in the intact animals was performed using identical procedures, with the renal nerves left intact. Inulin clearance studies. Each animal was anesthetized with a bolus injection of pentobarbitone sodium (70 mg/kg ip; National Veterinary Supplies). The left femoral vein was cannulated and infused with 0.9% saline (18 ml·kg⫺1·h⫺1). The left carotid artery was also cannulated to allow access for blood sampling and mean arterial pressure monitoring during the experiment. A midline laparotomy was conducted to allow cannulation of the left kidney ureter for urine collection. On completion of the preparation, the 0.9% saline infusion was replaced by a priming dose (2 ml) and then a sustained infusion (18 ml·kg⫺1·h⫺1) of inulin (15 mg/ml; Sigma, St. Louis, MO) and pentobarbitone sodium (0.44 mg/ml) in 0.9% saline. Following a 2-h equilibration period, urine was collected from the left ureter in two 15-min periods. Blood plasma samples were collected from the carotid artery at the beginning and end of each period. Inulin concentration in the plasma and urine samples was determined using a modification of the protocol described by Bojesen (7). Plasma and urine samples (50 ␮l) were deproteinized in 250 ␮l of NaOH (0.15 M) and neutralized in 1,700 ␮l of an H2SO4 (3.68 mM) and ZnSO4 (850 mM) solution. The precipitated protein was removed by centrifugation (500 g, 15 min, 4°C). Subsequently, 500 ␮l of the supernatant was incubated (100°C for 15 min) in a final ratio of 1:22.66 (vol/vol) to 2.5 ml of the color reagent [20% diphenylamine in absolute ethanol (wt/vol) diluted in a 3:10 (vol/vol) H2SO4 (0.125 M) solution in absolute ethanol]. The absorbance of each sample was read (␭ ⫽ 655 nm) at room temperature, and inulin clearance is expressed as a plasma-to-urine ratio. Histopathology. Freshly harvested kidneys were fixed in 10% neutral buffered formalin, routinely processed, embedded in paraffin, and sectioned using standard protocols. Periodic acid-Schiff’s staining was used to identify changes in basement membrane architecture and glycogen deposition in the kidneys. The glomerulosclerotic index (GSI) for each animal was determined by examining 150 –200 glomeruli in each section and scored in a double-blind manner using trained observers as previously described (32, 50): grade 0, normal; grade 1, sclerotic area up to one-quarter of total glomerular area; grade 2, sclerotic area more than one-quarter and up to half of total glomerular area; grade 3, sclerosis of more than half and up to three-quarters of the glomerular area; and grade 4, sclerosis of more than three-quarters of total glomerular area. Glomerulosclerosis was defined as glomerular basement membrane thickening, mesangial expansion, and capillary occlusion. The GSI was calculated using the following formula: GSI ⫽ (1 ⫻ N1 ⫹ 2 ⫻ N2 ⫹ 3 ⫻ N3 ⫹ 4 ⫻ N4)/(N0 ⫹ N1 ⫹ N2 ⫹ N3 ⫹ N4), where NX is the number of glomeruli in each grade of sclerosis. Immunohistochemistry. Staining was performed on 5-␮m microwave-fixed, paraffin-embedded sections using standard protocols. After a 2-h incubation with an animal-free blocker (Vector Laboratories, Burlingame, CA), the sections were incubated (15 h at 4°C) with the following primary antibodies diluted in Tris-buffered saline: mouse anti-rat ␣-smooth muscle actin (SMA; 1:400; Sigma), rabbit anti-rat collagen I (1:100; Rockland, Gilbertsville, PA), rabbit anti-rat colla-

AJP-Renal Physiol • doi:10.1152/ajprenal.00578.2013 • www.ajprenal.org

F253

RENAL DENERVATION IN DIABETIC NEPHROPATHY

gen IV (1:100; Abcam, Cambridge, UK), rabbit anti-rat NPHS2 (podocin; 1:40; Abcam), rabbit anti-rat transforming growth factor (TGF)-␤1 (1:100; Santa Cruz Biotechnology, Santa Cruz, CA), mouse monoclonal AT1R (ab9391; 1:20, Abcam), and sheep anti-rat tyrosine hydroxylase (1:200; Novus Biologicals, Littleton, CO). The sections were then washed in Tris-buffered saline and incubated with either fluorescein (Molecular Probes, Leiden, The Netherlands)- or peroxidase (Pierce, Rockford, IL)-conjugated secondary antibodies. Immunohistochemical labeling for each protein of interest was photographed in 10 separate fields/section using a Zeiss Axioplan 2 microscope with dedicated Axiovision v4.2 software (Carl Zeiss Vision). Semiquantitative analysis of DAB staining was conducted on the micrographs using Adobe Photoshop CS5 software (Adobe Systems, San Jose, CA) and calculated by the amount of pixels contained within a color spectrum using the “color range tool” preoptimized specifically for DAB staining against a fixed intensity cut-off threshold in which the ratio of DAB stained pixels was calculated against background in each field of interest (21, 39). Fluorescent labeling of tyrosine hydroxylase was photographed (⫻40 objective) using a Zeiss LSM 710 confocal microscope. Western blot analysis. Frozen kidney cortical tissues were homogenized, and protein concentrations were determined as previously published (41) using a Lowry DC assay kit (Bio-Rad, Hercules CA). Sample protein concentrations were extrapolated using a protein standard curve with a Pearson correlation coefficient of ⱖ0.99.

A

B 310

30 STZ/Vehicle

STZ/Vehicle

280 Body weight (g)

Blood glucose (mM)

Sample preparation, gel loading, electrophoresis, and transblotting procedures were carried out using a Tris-glycine-buffered system (Invitrogen Life Technologies, Carlsbad, CA) according to the manufacturer’s instructions. Immunoblotting was performed using standard protocols with the following primary antibodies: rabbit anti-rat collagen I (1:400; Rockland), rabbit anti-rat collagen IV (1:1,000; Abcam), and rabbit anti-rat TGF-␤1 (1:400; Santa Cruz Biotechnology). To confirm equal protein loading, blots were reprobed with rabbit anti-rat ␤-tubulin (1:5,000; Sigma). A horseradish peroxidase-conjugated secondary antibody (1:1,000; Pierce) was used to label the protein band, and blots were visualized using chemiluminescence (SuperSignal West Dura Extended Duration Substrate; Pierce). Densitometric scanning of protein bands of interest was performed using a GS-710 Calibrated Imaging Densitometer (BioRad) and quantified by Quantity One software (Bio-Rad). Each specific protein band density was ratioed against the ␤-tubulin protein density quantified in the same sample. Renal biochemical studies. The Na⫹ concentration of the urine samples was analyzed using a FP20 flame photometer (SEAC, Florence, Italy) against a 100 mM NaCl standard solution, diluted (1:200) in a Li2CO3 (15 meqV) reference solution. Urinary albumin concentration was measured using an Albumin Blue Fluorescent Assay Kit (Active Motif, Carlsbad, CA) performed according to the manufacturer’s instructions. Renal cortical norepinephrine levels were quantified at the terminal study point (12 wk post-diabetic induction) using

20

10

* **

250 220

NGInx (n=7) NGDnx (n=7) DBInx (n=8) DBDnx (n=7)

190 160 130 0

0 5

6

7

8

9

10

11

12

13

14

15

16

17

5

18

6

7

8

9

Age (weeks)

D

Baseline

240 220

+ ***

200

+

+

180

+

++

160

++

++

+++

+ +++

+++

+++

+++

*** **

+++

140 0

16

17

18

*** ***

450

400

350

5

6

7

8

9

10

11

12

13

14

15

16

17

0

18

5

6

7

8

9

8

*** ***

6 4 2 0 5

6

7

8

9

10

11

12

Age (weeks)

13

14

15

16

17

18

Urine flow (ml/24h/kg)

F STZ/Vehicle

10

10

11

12

13

14

15

16

17

18

Age (weeks)

12 UNaV (mmol/24h/kg)

15

500

Age (weeks)

E

14

STZ/Vehicle

550

STZ/Vehicle

Heart rate (BPM)

Systolic blood pressure (mmHg)

C

10 11 12 13 Age (weeks)

600 STZ/Vehicle 450 300 ***

150

***

0 5

6

7

8

9

10 11 12 13 Age (weeks)

14

15

16

17

18

Fig. 2. Physiological parameters in (mRen-2)27 TGRs from 5 to 18 wk of age. Values are means ⫾ SE. Vertical dotted line represents the time point when BRD/sham surgery was performed. Œ, Normoglycemic sham (NGInx); , diabetic sham (DBInx); 䊐, normoglycemic denervated (NGDnx); , diabetic denervated (DBDnx). A: mean weekly blood glucose levels. B: absolute body weight change, C: systolic blood pressure. D: heart rate. E: absolute Na⫹ excretion (UNaV). F: urine flow. In A, B, D–F: **P ⬍ 0.01 or ***P ⬍ 0.001 DBInx vs. NGInx, and *P ⬍ 0.05 or ***P ⬍ 0.001 DBDnx vs. NGDnx. In C: ***P ⬍ 0.001 NGInx vs. NGDnx, ***P ⬍ 0.001 DBInx vs. DBDnx and **P ⬍ 0.01 DBDnx vs. NGDnx. ⫹P ⬍ 0.01, ⫹⫹P ⬍ 0.01, or ⫹⫹⫹P ⬍ 0.001 vs. systolic blood pressure at baseline (8.5 wk) within the group. AJP-Renal Physiol • doi:10.1152/ajprenal.00578.2013 • www.ajprenal.org

F254

RENAL DENERVATION IN DIABETIC NEPHROPATHY

A g

g a

g

a

B Noradrenaline conc. (ng/mg protein)

50

***

40 30 20 10

ψψψ

***

0

NGInx

NGDnx

groups, which was maintained over the 9-wk denervation period (NGInx vs. NGDnx, P ⬍ 0.001; DBInx vs. DBDnx, P ⬍ 0.001, Fig. 2C). Denervation, however, produced a significantly greater decrease in SBP in diabetic animals compared with the normoglycemic group (DBDnx vs. NGDnx, P ⬍ 0.01). Diabetic TGRs also had a lower heart rate compared with normoglycemic groups in both sham-operated (DBInx vs. NGInx, P ⬍ 0.001) and denervated groups (DBDnx vs. NGDnx, P ⬍ 0.001, Fig. 2D). Both diabetic groups exhibited an increase in UNaV (Fig. 2E) and urine flow (Fig. 2F) following diabetic induction which persisted until termination (DBInx vs. NGInx, P ⬍ 0.001; DBDnx vs. NGDnx, P ⬍ 0.001). Repeated BRD had no impact on these parameters in either normoglycemic or diabetic animals.

DBInx

DBDnx

A

Fig. 3. Confirmation of successful chronic BRD. A: confirmation of surgical denervation in kidneys isolated from diabetic TGRs. Left: innervated kidney section from diabetic TGR subjected to sham surgery showing tyrosine hydroxylase-positive fluorescent immunostaining (green) indicated by arrows within the adventitia of intrarenal arterioles, tubules, and glomeruli. Middle: denervated kidney section from a diabetic TGR subjected to repeated chronic BRD. Right: a negative control kidney section without tyrosine hydroxylase primary antibody. Nuclei were stained with 4,6-diamidino-2-phenylindole (blue). Scale bars ⫽ 50 ␮m. a, arteriole; g, glomerulus. B: renal cortical tissue norepinephrine levels analyzed by ELISA. Values are means ⫾ SE. ***P ⬍ 0.001 vs. NGInx. ⌿⌿⌿P ⬍ 0.001 vs. DBInx.

RESULTS

Physiological parameters. Diabetic onset resulted in elevated blood glucose levels (Fig. 2A) and reduced body weight gain (DBInx vs. NGInx, P ⬍ 0.01; DBDnx vs. NGDnx, P ⬍ 0.05, Fig. 2B) in both sham-operated and denervated TGRs. Repeated BRD had no impact on these parameters in either normoglycemic or diabetic groups. As expected, SBP was elevated in the TGRs after week 6 with no significant difference seen between the sham-operated normoglycemic and diabetic groups over the duration of the experiment. As shown in Fig. 2C, BRD intervention in both normoglycemic and diabetic animals produced a progressive, significant decrease in SBP compared with the sham-operated

B

3

Glomerulosclerotic index

a commercial norepinephrine ELISA kit (BA E-5200, Labor Diagnostika Nord) performed according to the manufacturer’s instructions. Statistics. All values are expressed as means ⫾ SE of n ⫽ 7– 8 individual experiments/group. Statistical analysis was conducted to examine differences in responses between the four groups: normoglycemic innervated (NGInx), normoglycemic denervated (NGDnx), diabetic innervated (DBInx), and diabetic denervated (DBDnx). For systolic blood pressure (SBP), heart rate, body weight, urine flow, and absolute sodium secretion (UNaV), a two-way repeated measures ANOVA was used to provide a comparison between each of the two treatment paradigms: NGInx vs. NGDnx, DBInx vs. DBDnx, NGInx vs. DBInx, and NGDnx vs. DBDnx. Significant interaction between two groups in SPB and heart rate comparisons was further analyzed using one-way repeated measures ANOVA followed by a Dunnett post hoc comparison against an internal baseline measured at 8.5 wk post-STZ administration (presurgical values). All other comparisons were made with two-way ANOVA followed by a Bonferroni post hoc comparison. Statistics were performed using SigmaStat v3.5 software (Systat Software, San Jose, CA). Differences in group means were considered statistically significant if P ⬍ 0.05.

NGInx (n=7) NGDnx (n=7) DBInx (n=6) DBDnx (n=7)

***

2

1

0 NGInx

NGDnx

DBInx

DBDnx

Fig. 4. Renal morphological changes in TGRs. A: morphological changes in cortex and inner stripe of outer medulla in kidney sections stained with periodic acid-Schiff. In NGInx and NGDnx kidneys, glomeruli appearance and tubular structure in the cortex and medulla appeared normal with no remarkable features. In the kidneys from the DBInx TGRs, most glomeruli showed a severely thickened glomerular basement membrane (GBM) and mesangial hypertrophy; many cortical tubules were dilated with fibrotic casts deposited in the lumen; inflammatory cell infiltration was seen in some areas of the kidney (arrows). In the medulla, severe fibrosis was observed extensively across the interstitium and tubules (arrow). In DBDnx kidneys, the occurrence of glomeruli with a thickened GBM and mesangial hypertrophy was reduced to normoglycemic levels; a few cortical tubules were still dilated but with significantly less fibrotic casts appearing in the lumen; inflammatory infiltration was not noticeable. In the medulla, some interstitial fibrosis was still seen (arrows); however, the frequency and the severity of fibrosis were reduced. B: glomerulosclerotic index in kidneys from TGRs subjected to diabetes and renal denervation. Values are means ⫾ SE. ***P ⬍ 0.001 vs. NGInx. ⌿P ⬍ 0.05 vs. DBInx.

AJP-Renal Physiol • doi:10.1152/ajprenal.00578.2013 • www.ajprenal.org

F255

RENAL DENERVATION IN DIABETIC NEPHROPATHY

Confirmation of renal denervation. Positive tyrosine hydroxylase staining was found within cortical sections of shamoperated TGR kidneys indicating the presence of sympathetic innervation (Fig. 3A). Successful denervation was confirmed by the absence of tyrosine hydroxylase staining and by the reduction of total norepinephrine levels in denervated renal cortical tissues (Fig. 3B). Denervation reduced cortical norepinephrine content significantly from 23 ⫾ 1.6 to 1.0 ⫾ 0.1 ng/mg protein in the NGInx vs. NGDnx groups, respectively (P ⬍ 0.001), and from 37 ⫾ 2.6 to 1.7 ⫾ 0.6 ng/mg protein in the DBInx vs. the DBDnx groups, respectively (P ⬍ 0.001). Renal norepinephrine content was also raised as a consequence of diabetic induction in the sham-operated animals (P ⬍ 0.001). Renal morphological and functional parameters. Glomerular structure appeared normal in the normoglycemic kidneys (Fig. 4A). In contrast, most glomeruli were sclerosed in kidneys from diabetic sham-operated animals. This appearance was attenuated in denervated diabetic kidneys. The glomerular pathology was reflected by the significantly higher GSI recorded in kidneys isolated from sham-operated diabetic TGRs compared with the normoglycemic groups (1.94 ⫾ 0.47 DBInx vs. 0.56 ⫾ 0.10 NGInx, P ⬍ 0.001, Fig. 4B). This elevated GSI was attenuated in the diabetic TGRs subjected to BRD (0.68 ⫾ 0.14 DBDnx vs. 1.94 ⫾ 0.47 DBInx, P ⬍ 0.001). Normal renal medullary structure was observed in the normoglycemic TGR kidneys (Fig. 4A). In contrast, in diabetic sham-operated TGR kidneys, matrix expansion, GBM thickening, and lumen expansion were observed among the renal tubules distributed at the outer medullary region. Denervation of the diabetic TGR kidneys greatly reduced the extent of structural injury. Podocin staining was diminished within the glomeruli of sham-operated diabetic animals compared with the normoglycemic groups (Fig. 5, A and B). Diabetic-induced podocin loss was prevented in TGRs subjected to BRD. Denervation did not have any effect, however, on podocin staining in these hypertensive TGRs animals in the absence of diabetic insult. The functional significance of this podocyte injury was demonstrated by the presence of albuminuria in the diabetic TGRs.

A

B

Urinary absolute albumin levels assessed after 12 wk of hyperglycemia were raised in the diabetic sham-operated group compared with the normoglycemic groups (10.7 ⫾ 2.5 vs. 1.6 ⫾ 0.2 mg·24 h⫺1·kg⫺1, respectively, P ⬍ 0.001, Fig. 5A). Diabetic animals subjected to BRD showed a significantly reduced urinary albumin level compared with the diabetic sham-operated group (5.0 ⫾ 1.1 vs 10.7 ⫾ 2.5 mg·24 h⫺1·kg⫺1, respectively, P ⬍ 0.05). Left kidney weight recorded at termination was raised in the diabetic TGRs compared with the normoglycemic TGRs (P ⬍ 0.001, Fig. 6A). GFR (corrected for kidney weight) in the diabetic TGRs was significantly lower (P ⬍ 0.001, Fig. 6B) compared with the normoglycemic TGRs. BRD had no impact on kidney weight or GFR in either normoglycemic or diabetic TGRs. Renal fibrotic and profibrotic markers. Weak collagen I immunolabeling was seen distributed in both renal cortex and medulla in the normoglycemic TGRs (Fig. 7A). In the diabetic sham-operated group, elevated collagen I expression was found extracellularly at the GBM and in the interstitial space of the outer medulla. Intervention with BRD after the onset of diabetes reduced collagen I expression in both the cortex and medulla. Differences in collagen I expression were confirmed by Western blot analysis showing a twofold higher expression in the renal cortex of diabetic sham-operated TGRs (Fig. 7B), and this elevation was reduced in diabetic TGRs subjected to BRD. Collagen IV was weakly labeled in the normoglycemic TGR kidneys (Fig. 8A). Kidneys from the diabetic shamoperated group show elevated collagen IV expression within the glomeruli and in the interstitial space throughout the cortex and medulla. BRD intervention also successfully attenuated the extent of collagen IV expression found in diabetic TGRs. Protein analysis confirmed increased collagen IV expression in the cortex of diabetic sham-operated TGRs compared with the normoglycemic groups (Fig. 8B), and this elevated expression was also reduced in the denervated diabetic TGRs. Very low levels of ␣-SMA were detected in the medulla of normoglycemic TGR kidneys (Fig. 9, A and B). In the diabetic sham-operated group, ␣-SMA was extensively expressed

C

Urinary albumin (mg/24h/kg)

15

***

10

ψ 5

0 NGInx

NGDnx

DBInx

DBDnx

Fig. 5. Glomerular structural and functional integrity in TGRs. Podocin localization (A) and densitometric quantification of podocin positively stained regions (B) within glomeruli of renal cortical sections immunolabeled with anti-NPHS2 (podocin) antibody and counterstained with hematoxylin are shown. Strong podocin immunohistochemical staining (indicated by arrows) is seen in the glomeruli of both normoglycemic sham-operated and denervated TGR kidneys. Glomerular staining intensity of podocin in sham-operated diabetic TGR was markedly reduced. Podocin labeling within glomeruli of diabetic TGRs subjected to renal denervation was noticeably greater than in glomeruli isolated from the sham-operated diabetic TGRs. Podocyte injury was also confirmed by analyzing 24-h urinary absolute albumin excretion (C) in TGRs before termination. Groups are defined as in Fig. 2. Values are means ⫾ SE. **P ⬍ 0.01, ***P ⬍ 0.001 vs. NGInx. ⌿P ⬍ 0.05 vs. DBInx. AJP-Renal Physiol • doi:10.1152/ajprenal.00578.2013 • www.ajprenal.org

F256

RENAL DENERVATION IN DIABETIC NEPHROPATHY NGInx (n=7) NGDnx (n=6) DBInx (n=6) DBDnx (n=6)

A

Left kidney weight (g)

1.5

***

1.0

0.5

0.0 NGInx

Glomerular filtration rate (ml/min/g k.w.)

B

NGDnx

DBInx

DBDnx

***

1.5

tural damage seen in diabetic nephropathy. BRD intervention after pancreatic ␤-cell destruction and diabetic onset attenuated the progression of diabetic nephropathy in this transgenic model of renin-overexpressing hypertensive rats. As expected, BRD reduced the severity of hypertension and renal catecholamine levels in the normoglycemic animals. These parameters were also successfully attenuated within the diabetic animals. However, while BRD had no effect on renal morphology or injury markers in normoglycemic hypertensive groups, this denervation protocol clearly provided protection when hypertension was compounded by the presence of diabetes. Both diabetic-induced albuminuria and podocin loss but not hyperglycemia were decreased by BRD intervention. BRD also reduced intrarenal AT1R expression within the collecting ducts and attenuated renal fibrosis in diabetic TGRs by reducing the expression of the profibrotic inducer TGF-␤1. This reduction was supported by the attendant reduction in ␣-SMA and collagens I and IV. This study assesses for the first time the

1.0

A 0.5

0.0 NGInx

NGDnx

DBInx

DBDnx

Fig. 6. Left kidney weight (A) and glomerular filtration rate (B) in TGRs at termination. Groups are defined as in Fig. 2. Values are means ⫾ SE. ***P ⬍ 0.001 between indicated groups.

DISCUSSION

The current study is the first to demonstrate that renal nerves play an important role in the long-term development of struc-

B

NGInx NGDnx DBInx DBInx

~160 kDa Collagen I 55 kDa -tubulin

*

-tubulin

1.0

Collage n I /

within the renal tubule and interstitial space at the outer medullary region (Fig. 9, C and E), and this diabetic-induced elevation was attenuated in the medulla of diabetic TGRs subjected to BRD (Fig. 9, D and F). TGF-␤1 was elevated in both the cortex and medulla of the sham-operated diabetic TGR kidneys compared with normoglycemic TGRs (Fig. 10A). This expression was reduced in the diabetic TGR kidneys subjected to BRD. Western blot analysis showed a twofold higher expression in the renal cortex of the diabetic sham-operated group compared with normoglycemic TGRs (Fig. 10B), and this elevation was reduced in diabetic TGRs subjected to BRD. AT1R immunohistochemical studies in the normoglycemic kidneys produced faint staining within the proximal tubules and cortical and medullary collecting ducts (Fig. 11A). AT1R immunostaining became more pronounced within epithelial cells of the dilated cortical and outer medullary collecting tubules in the diabetic kidneys, producing a significant increase in AT1R immunostaining intensity relative to normoglycemic innervated kidneys (Fig. 11B). Conversely, AT1R staining in the rest of the nephron was reduced, particularly within regions of extracellular matrix expansion. While chronic renal denervation failed to affect the low AT1R staining intensity within the normoglycemic group, the procedure resulted in an abrupt loss of AT1R intensity in the diabetic animals.

NGInx (n=4) NGDnx (n=6) DBInx (n=5) DBDnx (n=5)

0.8 0.6 0.4 0.2 0.0 NGInx

NGDnx

DBInx

DBDnx

Fig. 7. Collagen I localization and expression in the cortex and inner stripe of outer medulla of kidney sections immunolabeled with collagen I antibody and counterstained with hematoxylin. A: in normoglycemic innervated and denervated kidney sections, collagen I was weakly labeled and occasionally seen in the glomerular basement membrane and medullary tubules. In the diabetic innervated kidney sections, intense collagen I staining was found in the GBM (arrow) in most of the glomeruli; strong collagen I staining was also found extensively in the interstitium (arrow) of the inner stripe of outer medullary tubules. BRD in the diabetic TGRs reduced collagen I labeling to normoglycemic levels in the glomeruli and medullary tubules. Scale bars ⫽ 50 ␮m. B: Western blot analysis indicates an increased collagen I protein expression in the diabetic innervated TGR kidneys. This elevation is attenuated in the diabetic TGR kidneys subjected to renal denervation. Values are means ⫾ SE. *P ⬍ 0.05 vs. NGInx. ⌿P ⬍ 0.05 vs. DBInx.

AJP-Renal Physiol • doi:10.1152/ajprenal.00578.2013 • www.ajprenal.org

RENAL DENERVATION IN DIABETIC NEPHROPATHY

A

B

NGInx NGDnx DBInx DBInx

~160 kDa Collagen IV 55 kDa -tubulin

Collage n IV /

-tubulin

1.5

*

NGInx (n=4) NGDnx (n=6) DBInx (n=5) DBDnx (n=5)

1.0

0.5

0.0 NGInx

NGDnx

DBInx

DBDnx

Fig. 8. Collagen IV localization and expression in the cortex and inner stripe of outer medulla of kidney sections immunolabeled with collagen IV antibody and counterstained with hematoxylin. A: in sections from normoglycemic innervated and denervated kidneys, collagen IV was weakly labeled in the glomeruli and medullary tubules. In the diabetic innervated kidney sections, intense collagen IV staining was found within the glomeruli; strong collagen IV staining was also found extensively in the interstitium of the medullary tubules. BRD in the diabetic TGRs reduced collagen IV labeling to normoglycemic levels in the glomeruli and medullary tubules. Scale bars ⫽ 50 ␮m. B: Western blot analysis indicates increased collagen IV protein expression in the diabetic innervated TGR kidneys. This elevation was attenuated in the diabetic TGR kidneys subjected to renal denervation. Values are means ⫾ SE. *P ⬍ 0.05 vs. NGInx. ⌿P ⬍ 0.05 vs. DBInx.

effect of renal denervation on the individual contribution of diabetes as well as hypertension in diabetic nephropathy. The induction of diabetes in the hypertensive TGR animals resulted in a reduced rate of growth compared with the normoglycemic animals. This pattern of body weight difference recorded in the intact normoglycemic and diabetic TGRs is consistent with earlier published data gained from the same animal (32). While this reduction in body weight gain has also been observed in other diabetic rat models following STZ administration, our results indicate that the surgical intervention procedure itself had minimal impact on the animals’ health and growth status (12). The 12-wk time course of this experimental model was carefully selected to ensure that the development of renal injury did not result solely from the effects of chronic hypertension but as a consequence of the synergistic effects of hyperglycemic induction. Applehoff and coworkers (3) clearly demonstrated that both diabetes and the ren-2 gene may contribute to the glomerular pathology described at 4 mo

F257

in the male rat. Conversely, female TGRs develop lower levels of hypertension (30 – 40 mmHg less) than their male littermates and show no evidence of morphological changes in the nephron before 4 – 6 mo of age (4). Consequently, while SBP was elevated in the intact female heterozygous TGR animals over the 12-wk period, no evidence of renal morphological injury or proteinuria was seen in the normoglycemic female TGRs. We found that BRD intervention reduced SBP and renal cortical catecholamine levels in the normoglycemic reninoverexpressing hypertensive rats. While heart rate was unaffected by denervation in either normo- or hyperglycemic groups, BRD produced a more pronounced decrease in SBP in the diabetic than in the normoglycemic animals, suggesting that RSNA is highly important in maintaining elevated blood pressure in the diabetic TGRs with renal injury. Indeed, renal catecholamine levels were significantly raised in the diabetic fibrosed kidneys compared with the hypertensive only group, strongly supporting the hypothesis that diabetic renal injury can elevate RSNA. In support of this premise, afferent renal nerve activity has also been shown to be elevated in the rat renal injury model (10). Consequently, in innervated diabetic hypertensive TGRs, the amplified afferent signals originating from the kidneys may serve to sustain the elevation in systemic SNA and maintain an increased blood pressure, possibly to compensate for a potential fall in cardiac output resulting from diabetic cardiomyopathy. In contrast, in denervated diabetic TGRs, permanent removal of renal afferent nerves suppressed SNA, resulting in a fall in SBP. Studies conducted by Campese and colleagues (5, 9, 63) in the nephrectomized hypertensive rat renal injury model have shown that selective renal afferent denervation can abolish the severity of hypertension and reduce posterior hypothalamic noradrenaline content. Similarly, BRD was found to reduce blood pressure and attenuate the hypothalamic norepinephrine turnover in a neurogenic hypertensive rat model (62– 64). These prior studies support the suggestion that elevated afferent impulses from injured kidneys can stimulate central sympathetic activity at the hypothalamic level, thereby evoking hypertension and may explain the antihypertensive effects of BRD observed in the present study. Further work on this model should help to clarify the specific role of afferent renal nerve activity in the progression of hypertension and renal failure. Following diabetic onset, marked hyperglycemia and albuminuria, combined with a reduction in the podocyte-specific protein podocin, were observed in the innervated diabetic TGRs. The association between glomerular podocin loss and albuminuria has also been found in both experimental models and in human diabetic nephropathy (37, 61). The reduction of the diabetes-induced albuminuria with BRD can be partially explained by the reduction in blood pressure consequent with this intervention (54). However, studies in cultured podocytes have shown that the ␣-adrenoceptor agonist phenylephrine can increase intracellular Ca2⫹ concentration to alter podocyte shape and reduce foot processes, thus directly regulating podocyte functions (29). These in vitro studies strongly suggest that elevated cortical norepinephrine levels recorded in the diabetic nephropathy animals may directly mediate podocyte injury, consequently permitting the development of albuminuria. This supports and may partially explain the mechanism through which BRD attenuated podocyte injury in the diabetic kidneys. Indeed, BRD also reduced the severity of glomerulosclerosis

AJP-Renal Physiol • doi:10.1152/ajprenal.00578.2013 • www.ajprenal.org

F258

RENAL DENERVATION IN DIABETIC NEPHROPATHY

NGInx

NGDnx

DBInx

DBDnx

Fig. 9. ␣-Smooth muscle actin (SMA) localization and expression in the inner stripe of outer medulla of kidney sections immunolabeled with anti-␣-SMA antibody and counterstained with hematoxylin. Minimal ␣-SMA staining was seen in the medullary tubules of normoglycemic shamoperated (A) and denervated (B) kidneys. In the diabetic innervated TGR kidneys, intense ␣-SMA staining was extensively expressed in the interstitium of the medullary tubules (C and magnified in E). BRD in the diabetic TGR kidneys markedly decreased medullary ␣-SMA deposition (D and magnified in F). Scale bars ⫽ 50 ␮m.

and preserved glomerular structure in diabetic TGRs. Selective renal afferent denervation was shown to reduce blood pressure and glomerulosclerosis severity in earlier studies of a hypertensive rat model of chronic renal failure conducted by Campese and colleagues (10). While glomerular protection may in part stem from the reduction in blood pressure consequent to BRD intervention, it is important to note, however, that in the absence of hyperglycemia, none of the hypertensive animals used in this study showed any evidence of proteinuria or renal morphological damage. Evidence in support of this antihypertensive-independent effect comes from other studies by Nagasu and coworkers (46) showing that removal of sympathetic influence by acute renal denervation can directly protect glomerular structure in the absence of a blood pressurelowering effect. Furthermore, the inhibition of RSNA using pharmacological sympatholytic agents such as prazosin, timolol, guanethidine, and moxonidine at subantihypertensive concentrations has also been shown to preserve glomerular structure in rat models of hypertensive renal failure (2, 22). The presence of ␤1-adrenoceptors within glomerular mesangial cells (6) suggests that the RSNA may play a direct role in mesangial dysfunction. Consequently, removal of this elevated sympathetic influence on diabetic kidneys may protect the structural and functional integrity of the glomerular mesangium. The presence of renal cortical and medullary fibrosis as evidenced by the extensive deposition of glycogen and accompanied by tubule dilation in the diabetic kidneys was confirmed by the elevated expression of renal fibrotic markers TGF-␤1, ␣-SMA, and collagens I and IV. Repeated BRD at 3 wk following diabetic onset suppressed the levels of these fibrotic markers. Multiple mechanistic pathways may be involved in mediating the protective effect of BRD against diabetic renal fibrosis and warrants further investigation. Previous work conducted in other models of renal injury has shown that suppression of renal sympathetic innervation by surgical denervation or moxonidine administration may be effective in suppressing renal fibrotic markers such as TGF-␤1 (2, 23, 57). Renal fibrogenesis induced by ureteric obstruction in a rat model has similarly been shown to be attenuated by denervation and by

selective ␣2-adrenoreceptor agonist administration; this protection was reversed, however, by the infusion of a calcitonin gene-related peptide and norepinephrine, elegantly demonstrating that both renal afferent and efferent nerves are involved in the formation of renal fibrosis (33). Inflammatory responses are also involved in mediating the progression of diabetic renal injury (19). As renal denervation in the anti-Thy-1.1-induced glomerulonephritis model was shown to produce anti-inflammatory and antifibrotic effects (57), it is conceivable that the protection afforded by BRD against diabetic nephropathy in the TGRs could partially be due to an anti-inflammatory effect. Systemic renin angiotensin system activity can also be effectively reduced by a BRD abrogation of renin release from the juxtaglomerular apparatus (20). Elevated norepinephrine release from renal sympathetic nerve terminals as seen in the diabetic TGRs group can induce TGF-␤ expression, thereby enhancing the extracellular matrix accumulation and profibrotic actions of intrarenal angiotensin II at the renal tubules (13). BRD intervention in this study may consequently inhibit the involvement of the intrarenal renin-angiotensin system in the TGF-␤-mediated profibrotic pathways observed in the diabetic kidneys. Unfortunately, this study was not set up to tease out the effects of denervation on the catabolic action of proinflammatory matrix metalloproteinases and other proteolytic enzymes on pathogenic extracellular matrix accumulation (40). The potential impact of BRD on metalloproteinases specific to each stage of albuminuria development remains to be clarified. While various studies have demonstrated the interstitial and intratubular localization of the intrarenal renin-angiotensin system, only a few studies have attempted to resolve the activation status of this pathway in diabetes. Induction of moderate hyperglycemia using STZ in a rat model has been shown to result in elevated systemic angiotensin II and increased AT1R immunostaining in the renal collecting ducts within 20 days (25, 35). Examination of our animal kidneys subjected to a longer time course of hyperglycemia (12 wk post STZ) shows that diabetic induction also had the effect of increasing AT1R expression in the cortical and outer medullary collecting duct of these renin-overexpressing TGRs. No

AJP-Renal Physiol • doi:10.1152/ajprenal.00578.2013 • www.ajprenal.org

RENAL DENERVATION IN DIABETIC NEPHROPATHY

A

B

NGInx NGDnx DBInx DBInx

~25 kDa TGF- 1 55 kDa -tubulin

*

NGInx (n=4) NGDnx (n=6) DBInx (n=7) DBDnx (n=5)

0.2

TGF-

1

/

-tubulin

0.3

0.1

0.0 NGInx

NGDnx

DBInx

DBDnx

Fig. 10. Transforming growth factor (TGF)-␤1 localization and expression in the cortex and medulla of the kidney sections immunolabeled with antiTGF-␤1 antibody and counterstained with hematoxylin. A: in sections from normoglycemic innervated and denervated kidneys, TGF-␤1 staining was weakly labeled. In the diabetic innervated kidneys, intense staining was found within the glomerular mesangium (thin arrows) and in the cytoplasm and luminal surface of the proximal and distal tubules (wide arrow); in the medulla, TGF-␤1 was found in the interstitial space within the cytoplasm of the medullary tubules. BRD in the diabetic TGRs reduced TGF-␤1 labeling to normoglycemic levels in the glomeruli and medullary tubules. Scale bars ⫽ 50 ␮m. B: Western blot analysis shows increased TGF-␤1 protein expression in the cortex of diabetic innervated TGR kidneys. This elevation is attenuated in the diabetic TGRs subjected to renal denervation. Values are means ⫾ SE. *P ⬍ 0.05 vs. NGInx. ⌿P ⬍ 0.05 vs. DBInx.

change in AT1R immunostaining of the nephron was evident following BRD in the normoglycemic animals, suggesting that while both denervation and reduction in hypertensive status may not have had any effect on the weak AT1R staining, AT1R localization was not raised as a consequence of the hypertension but as a result of diabetic induction. Significantly, BRD intervention in the diabetic TGR abrogated the AT1R staining of collecting tubules. This finding is consistent with the other data presented here demonstrating that renal morphology and fibrotic makers are not markedly altered by denervation in the hypertensive only heterozygous female TGR. The significance of the increased AT1R expression within the collecting tubules may be explained as an angiotensin II-mediated conservation of the Na⫹/water balance in the face of an increased urine output in diabetes (25). However, while BRD reduced AT1R expression, BRD did not restore UNaV and reduce output in this TGR model.

F259

Diabetic induction resulted in the development of renal hypertrophy and a decreased GFR after 12 wk in the present model, indicating that renal function was at the declining phase of the diabetic disease progression. However, both GFR reduction (32) and elevation (59) have been reported at this same time point following diabetic onset in the same diabetic TGR model. Type 1 diabetic induction in the TGR increased GFR at 4 wk but reduced it by ⬎50% after 12 wk of diabetes in TGRs. These studies suggest that the initial point of GFR decline is variable, occurring at slightly different time points during the disease progression in different batches of animals. In contrast to the renal morphological protection afforded by BRD in diabetic TGRs, no similar effect was observed with the GFR indices. It should be noted that GFR in the normoglycemic hypertensive TGRs was also unaffected by BRD intervention. Studies conducted in a type 1 diabetic rat have indicated that BRD can reverse GFR changes if conducted early (2 wk) within the disease progression (42). While this discrepancy is hard to explain, our observations are consistent with a recent clinical study, which found that GFR was unchanged by BRD treatment in patients with moderate to severe chronic kidney disease (27). In conclusion, the present study conducted in renin-overexpressing TGRs substantiates the significance of renal sympathetic nerves in the pathogenesis of diabetic nephropathy. While denervation clearly had no effect on renal morphology or fibrotic marker status in hypertensive only animals, BRD substantially reduced albuminuria and attenuated the renal fibrosis and glomerulosclerosis consequences of hyperglycemia. BRD was conducted over a 12-wk period starting 3 wk after the onset of early diabetic injury (36). Further work will be needed to clarify whether this procedure can revert established diabetic nephropathy and show benefit over current pharmacological interventions. Early experimental studies using mesenchymal cell transplantation in type 1 diabetic mice have proved promising, indicating that both microalbuminuria and renal structural injury may be reversed; however, this therapy is still experimental (17). While reducing the progress of the disease, angiotensin-converting enzyme inhibitors fail to prevent angiotensin escape pathways from eventually provoking the renal manifestations of hyperglycemia. By attenuating the development of diabetic nephropathy, renal denervation may consequently prove valuable as an additive to the current practice of renin-angiotensin-aldosterone blockade. The superiority of this combination therapy over monotherapy with quinapril has already been demonstrated in a renal ablation model (23). We conclude that the protection of renal indices and structure reported in this study provides strong scientific support for the application of renal denervation procedures in diabetic nephropathy. ACKNOWLEDGMENTS The authors thank Prof. Zoltan Endre (Kidney Research Group, University of Otago, Christchurch, NZ) for valuable discussions on and use of the transgenic (mRen-2)27 colony. The authors also thank Amanda Fisher (University of Otago, Department of Pathology) and Andrew McNaughton (University of Otago, University of Otago, Department of Anatomy) for technical assistance with immunostaining and confocal microscopy. GRANTS These studies were supported by the Otago School of Medical Sciences, University of Otago.

AJP-Renal Physiol • doi:10.1152/ajprenal.00578.2013 • www.ajprenal.org

F260

RENAL DENERVATION IN DIABETIC NEPHROPATHY

Fig. 11. A: ANG II type 1 receptor (AT1R) immunolabeling and localization in the cortex and outer medulla of normoglycemic and diabetic sections counterstained with hematoxylin. Faint AT1R labeling (brown DAB stain) was observed within the proximal and distal tubules of normoglycemic or diabetic TGRs. AT1R labeling was also weak within the collecting duct segments and connecting tubules of normoglycemic kidney sections but appeared prominent (see arrows) within the epithelial layer of the collecting tubules obtained from the intact DBInx group. Note the lack of DAB staining in the abnormally expanded matrix around the dilated collecting tubules of the DBInx. AT1R labeling was strikingly reduced by denervation in the diabetic animals. Scale bars ⫽ 50 ␮m. B: percentage of tissue positively stained for DAB against background, recorded in 10 randomly selected cortical and medullary fields. Values are means ⫾ SE (n ⫽ 7 - 8 individual animals/group). ***P ⬍ 0.0001 vs. NGInx.

DISCLOSURES No conflicts of interest, financial or otherwise, are declared by the authors.

3.

AUTHOR CONTRIBUTIONS Author contributions: Y.Y., J.C.H., R.J.W., G.D., and I.A.S. provided conception and design of research; Y.Y., I.C.F.-N., and J.C.H. performed experiments; Y.Y., I.C.F.-N., and G.D. analyzed data; Y.Y., J.C.H., G.D., and I.A.S. interpreted results of experiments; Y.Y. prepared figures; Y.Y. and I.A.S. drafted manuscript; Y.Y., I.C.F.-N., J.C.H., R.J.W., G.D., and I.A.S. approved final version of manuscript; J.C.H. and I.A.S. edited and revised manuscript.

4.

5.

6.

REFERENCES 1. Amann K, Koch A, Hofstetter J, Gross ML, Haas C, Orth SR, Ehmke H, Rump LC, Ritz E. Glomerulosclerosis and progression: effect of subantihypertensive doses of alpha and beta blockers. Kidney Int 60: 1309 –1323, 2001. 2. Amann K, Rump LC, Simonaviciene A, Oberhauser V, Wessels S, Orth SR, Gross ML, Koch A, Bielenberg GW, Van Kats JP, Ehmke H, Mall G, Ritz E. Effects of low dose sympathetic inhibition on glomeru-

7. 8. 9.

losclerosis and albuminuria in subtotally nephrectomized rats. J Am Soc Nephrol 11: 1469 –1478, 2000. Appelhoff RJ, Hill JV, Findon G, Frampton CM, Perry E, Ponnamperuma D, Endre ZH. Differential contribution of diabetes and the Ren2 gene to glomerular pathology in diabetic (mREN-2)27 rats. Lab Invest 90: 1225–1235, 2010. Bachmann S, Peters J, Engler E, Ganten D, Mullins J. Transgenic rats carrying the mouse renin gene–morphological characterization of a lowrenin hypertension model. Kidney Int 41: 24 –36, 1992. Bigazzi R, Kogosov E, Campese VM. Altered norepinephrine turnover in the brain of rats with chronic renal failure. J Am Soc Nephrol 4: 1901– 1907, 1994. Boivin V, Jahns R, Gambaryan S, Ness W, Boege F, Lohse MJ. Immunofluorescent imaging of beta 1- and beta 2-adrenergic receptors in rat kidney. Kidney Int 59: 515–531, 2001. Bojesen E. A method for determination of inulin in plasma and urine. Acta Med Scand Suppl 266: 275–282, 1952. Brosius F. New insights into the mechanisms of fibrosis and sclerosis in diabetic nephropathy. Rev Endocr Metab Disord 9: 245–254, 2008. Campese VM, Kogosov E. Renal afferent denervation prevents hypertension in rats with chronic renal failure. Hypertension 25: 878 –882, 1995.

AJP-Renal Physiol • doi:10.1152/ajprenal.00578.2013 • www.ajprenal.org

RENAL DENERVATION IN DIABETIC NEPHROPATHY 10. Campese VM, Kogosov E, Koss M. Renal afferent denervation prevents the progression of renal disease in the renal ablation model of chronic renal failure in the rat. Am J Kidney Dis 26: 861–865, 1995. 11. Converse R, Jacobsen TN, Toto RD, Jost CM, Cosentino F, FouadTarazi F, Victor RG. Sympathetic overactivity in patients with chronic renal failure. N Engl J Med 327: 1912–1918, 1992. 12. Cooper ME, Allen TJ, O’Brien RC, Macmillan PA, Clarke B, Jerums G, Doyle AE. Effects of genetic hypertension on diabetic nephropathy in the rat: Functional and structural characteristics. J Hypertens 6: 1009 – 1016, 1988. 13. DiBona GF. Peripheral and central interactions between the renin-angiotensin system and the renal sympathetic nerves in control of renal function. Ann NY Acad Sci 940: 395–406, 2001. 14. DiBona GF, Sawin LL. Role of renal alpha 2-adrenergic receptors in spontaneously hypertensive rats. Hypertension 9: 41–48, 1987. 15. DiSalvo J, Fell C. Changes in renal blood flow during renal nerve stimulation. Proc Soc Exp Biol Med 136: 150 –153, 1971. 16. Esler MD, Krum H, Sobotka PA, Schlaich MP, Schmieder RE, Bohm M. Renal sympathetic denervation in patients with treatment-resistant hypertension (The Symplicity HTN-2 Trial): a randomised controlled trial. Lancet 376: 1903–1909, 2010. 17. Ezquer FE, Ezquer ME, Parrau DB, Carpio D, Yanez AJ, Conget PA. Systemic administration of multipotent mesenchymal stromal cells reverts hyperglycemia and prevents nephropathy in type 1 diabetic mice. Biol Blood Marrow Transplant 14: 631–640, 2008. 18. Fioretto P, Mauer M. Histopathology of diabetic nephropathy. Semin Nephrol 27: 195–207, 2007. 19. Fornoni A, Ijaz A, Tejada T, Lenz O. Role of inflammation in diabetic nephropathy. Curr Diabetes Rev 4: 10 –17, 2008. 20. Friis UG, Jensen BL, Sethi S, Andreasen D, Hansen PB, Skott O. Control of renin secretion from rat juxtaglomerular cells by cAMP-specific phosphodiesterases. Circ Res 90: 996 –1003, 2002. 21. Gaspard GJ, Pasumarthi KB. Quantification of cardiac fibrosis by colour-subtractive computer-assisted image analysis. Clin Exp Pharmacol Physiol 35: 679 –686, 2008. 22. Gattone VH, Evan AP, Overhage JM, Severs WB. Developing renal innervation in the spontaneously hypertensive rat: evidence for a role of the sympathetic nervous system in renal damage. J Hypertens 8: 423–428, 1990. 23. Hamar P, Kokeny G, Liptak P, Krtil J, Adamczak M, Amann K, Ritz E, Gross ML. The combination of ACE inhibition plus sympathetic denervation is superior to ACE inhibitor monotherapy in the rat renal ablation model. Nephron Exp Nephrol 105: e124 –e136, 2007. 24. Hammad FT, Wheatley AM, Davis G. Long-term renal effects of unilateral ureteral obstruction and the role of endothelin. Kidney Int 58: 242–250, 2000. 25. Harrison-Bernard LM, Imig JD, Carmines PK. Renal AT1 receptor protein expression during the early stage of diabetes mellitus. Int J Exp Diabetes Res 3: 97–108, 2002. 26. Hausberg M, Kosch M, Harmelink P, Barenbrock M, Hohage H, Kisters K, Dietl KH, Rahn KH. Sympathetic nerve activity in end-stage renal disease. Circulation 106: 1974 –1979, 2002. 27. Hering D, Mahfoud F, Walton AS, Krum H, Lambert GW, Lambert EA, Sobotka PA, Bohm M, Cremers B, Esler MD, Schlaich MP. Renal denervation in moderate to severe CKD. J Am Soc Nephrol 23: 1250 – 1257, 2012. 28. Holmer S, Rinne B, Eckardt KU, Le HM, Schricker K, Kaissling B, Riegger G, Kurtz A. Role of renal nerves for the expression of renin in adult rat kidney. Am J Physiol Renal Fluid Electrolyte Physiol 266: F738 –F745, 1994. 29. Huber TB, Gloy J, Henger A, Schollmeyer P, Greger R, Mundel P, Pavenstadt H. Catecholamines modulate podocyte function. J Am Soc Nephrol 9: 335–345, 1998. 30. Jyothirmayi GN, Alluru I, Reddi AS. Doxazosin prevents proteinuria and glomerular loss of heparan sulfate in diabetic rats. Hypertension 27: 1108 –1114, 1996. 31. Katholi RE, Rocha-Singh KJ. Catheter-based renal sympathetic denervation reduces systolic blood pressure by 32 mmHg in people with treatment-resistant hypertension. Evid Based Med 16: 109 –110, 2011. 32. Kelly DJ, Wilkinson-Berka JL, Allen TJ, Cooper ME, Skinner SL. A new model of diabetic nephropathy with progressive renal impairment in the transgenic (mRen-2)27 rat (TGR). Kidney Int 54: 343–352, 1998. 33. Kim J, Padanilam BJ. Renal nerves drive interstitial fibrogenesis in obstructive nephropathy. J Am Soc Nephrol 24: 229 –242, 2013.

F261

34. Kline RL, Mercer PF. Functional reinnervation and development of supersensitivity to NE after renal denervation in rats. Am J Physiol Regul Integr Comp Physiol 238: R353–R358, 1980. 35. Kobori H, Kamiyama M, Harrison-Bernard LM, Navar LG. Cardinal role of the intrarenal renin-angiotensin system in the pathogenesis of diabetic nephropathy. J Investig Med 61: 256 –264, 2013. 36. Kobori H, Prieto-Carrasquero MC, Ozawa Y, Navar LG. AT1 receptor mediated augmentation of intrarenal angiotensinogen in angiotensin II-dependent hypertension. Hypertension 43: 1126 –1132, 2004. 37. Koop K, Eikmans M, Baelde HJ, Kawachi H, De HE, Paul LC, Bruijn JA. Expression of podocyte-associated molecules in acquired human kidney diseases. J Am Soc Nephrol 14: 2063–2071, 2003. 38. Krum H, Schlaich M, Whitbourn R, Sobotka PA, Sadowski J, Bartus K, Kapelak B, Walton A, Sievert H, Thambar S, Abraham WT, Esler M. Catheter-based renal sympathetic denervation for resistant hypertension: a multicentre safety and proof-of-principle cohort study. Lancet 373: 1275–1281, 2009. 39. Lehr HA, Mankoff DA, Corwin D, Santeusanio G, Gown AM. Application of Photoshop-based image analysis to quantification of hormone receptor expression in breast cancer. J Histochem Cytochem 45: 1559 – 1565, 1997. 40. Lods N, Ferrari P, Frey FJ, Kappeler A, Berthier C, Vogt B, Marti HP. Angiotensin-converting enzyme inhibition but not angiotensin II receptor blockade regulates matrix metalloproteinase activity in patients with glomerulonephritis. J Am Soc Nephrol 14: 2861–2872, 2003. 41. Lowry O, Rosebrough N, Farr A, Randall R. Protein measurement with the Folin phenol reagent. J Biol Chem 193: 265–275, 1951. 42. Luippold G, Beilharz M, Muhlbauer B. Chronic renal denervation prevents glomerular hyperfiltration in diabetic rats. Nephrol Dial Transplant 19: 342–347, 2004. 43. Mahfoud F, Schlaich M, Kindermann I, Ukena C, Cremers B, Brandt MC, Hoppe UC, Vonend O, Rump LC, Sobotka PA, Krum H, Esler M, Bohm M. Effect of renal sympathetic denervation on glucose metabolism in patients with resistant hypertension: a pilot study. Circulation 123: 1940 –1946, 2011. 44. Mogensen CE, Christensen CK, Vittinghus E. The stages in diabetic renal disease. With emphasis on the stage of incipient diabetic nephropathy. Diabetes 32, Suppl 2: 64 –78, 1983. 45. Mulder J, Hokfelt T, Knuepfer MM, Kopp UC. Renal sensory and sympathetic nerves reinnervate the kidney in a similar time dependent fashion following renal denervation in rats. Am J Physiol Regul Integr Comp Physiol 304: R675–R682, 2013. 46. Nagasu H, Satoh M, Kuwabara A, Yorimitsu D, Sakuta T, Tomita N, Kashihara N. Renal denervation reduces glomerular injury by suppressing NAD(P)H oxidase activity in Dahl salt-sensitive rats. Nephrol Dial Transplant 25: 2889 –2898, 2010. 47. Nielsen FS, Hansen HP, Jacobsen P, Rossing P, Smidt UM, Christensen NJ, Pevet P, Vivien-Roels B, Parving HH. Increased sympathetic activity during sleep and nocturnal hypertension in Type 2 diabetic patients with diabetic nephropathy. Diabet Med 16: 555–562, 1999. 48. Phillips JK. Pathogenesis of hypertension in renal failure: role of the sympathetic nervous system and renal afferents. Clin Exp Pharmacol Physiol 32: 415–418, 2005. 49. Ritz E, Rychlik I, Locatelli F, Halimi S. End-stage renal failure in type 2 diabetes: a medical catastrophe of worldwide dimensions. Am J Kidney Dis 34: 795–808, 1999. 50. Saito T, Sumithran E, Glasgow EF, Atkins RC. The enhancement of aminonucleoside nephrosis by the co-administration of protamine. Kidney Int 32: 691–699, 1987. 51. Schlaich MP, Sobotka PA, Krum H, Lambert E, Esler MD. Renal sympathetic-nerve ablation for uncontrolled hypertension. N Engl J Med 361: 932–934, 2009. 52. Strojek K, Grzeszczak W, Gorska J, Leschinger MI, Ritz E. Lowering of microalbuminuria in diabetic patients by a sympathicoplegic agent: novel approach to prevent progression of diabetic nephropathy? J Am Soc Nephrol 12: 602–605, 2001. 53. Symplicity HTN1 Investigators. Catheter-based renal sympathetic denervation for resistant hypertension: durability of blood pressure reduction out to 24 months. Hypertension 57: 911–917, 2011. 54. Thomas MC, Atkins RC. Blood pressure lowering for the prevention and treatment of diabetic kidney disease. Drugs 66: 2213–2234, 2006. 55. Todoran TM, Basile JN, Zile MR. Renal sympathetic denervation for blood pressure control: a review of the current evidence and ongoing studies. J Clin Hypertens (Greenwich) 16: 331–341, 2014.

AJP-Renal Physiol • doi:10.1152/ajprenal.00578.2013 • www.ajprenal.org

F262

RENAL DENERVATION IN DIABETIC NEPHROPATHY

56. Van Buren PN, Kim C, Toto R, Inrig JK. Intradialytic hypertension and the association with interdialytic ambulatory blood pressure. Clin J Am Soc Nephrol 6: 1684 –1691, 2011. 57. Veelken R, Vogel EM, Hilgers K, Amann K, Hartner A, Sass G, Neuhuber W, Tiegs G. Autonomic renal denervation ameliorates experimental glomerulonephritis. J Am Soc Nephrol 19: 1371–1378, 2008. 58. Weinrauch LA, Kennedy FP, Gleason RE, Keough J, D’Elia JA. Relationship between autonomic function and progression of renal disease in diabetic proteinuria: clinical correlations and implications for blood pressure control. Am J Hypertens 11: 302–308, 1998. 59. Wilkinson-Berka JL, Kelly DJ, Koerner SM, Jaworski K, Davis B, Thallas V, Cooper ME. ALT-946 and aminoguanidine, inhibitors of advanced glycation, improve severe nephropathy in the diabetic transgenic (mREN-2)27 rat. Diabetes 51: 3283–3289, 2002.

60. Winternitz SR, Oparil S. Importance of the renal nerves in the pathogenesis of experimental hypertension. Hypertension 4: III108 –III114, 1982. 61. Wu Y, Dong J, Yuan L, Liang C, Ren K, Zhang W, Fang F, Shen J. Nephrin and podocin loss is prevented by mycophenolate mofetil in early experimental diabetic nephropathy. Cytokine 44: 85–91, 2008. 62. Ye S, Gamburd M, Mozayeni P, Koss M, Campese VM. A limited renal injury may cause a permanent form of neurogenic hypertension. Am J Hypertens 11: 723–728, 1998. 63. Ye S, Ozgur B, Campese VM. Renal afferent impulses, the posterior hypothalamus, and hypertension in rats with chronic renal failure. Kidney Int 51: 722–727, 1997. 64. Ye S, Zhong H, Yanamadala V, Campese VM. Renal injury caused by intrarenal injection of phenol increases afferent and efferent renal sympathetic nerve activity. Am J Hypertens 15: 717–724, 2002.

AJP-Renal Physiol • doi:10.1152/ajprenal.00578.2013 • www.ajprenal.org