reperfusion injury

http://www.kidney-international.org

original article

& 2011 International Society of Nephrology

see commentary on page 10

Inhibition of 20-HETE synthesis and action protects the kidney from ischemia/reperfusion injury Uwe Hoff1,7, Ivo Lukitsch1,7, Lyubov Chaykovska1, Mechthild Ladwig2, Cosima Arnold3, Vijay L. Manthati4, T. Florian Fuller5, Wolfgang Schneider6, Maik Gollasch1, Dominik N. Muller3, Bert Flemming2, Erdmann Seeliger2, Friedrich C. Luft6, John R. Falck4, Duska Dragun1,7 and Wolf-Hagen Schunck3,7 1

Nephrology and Intensive Care Medicine, Campus Virchow and Center for Cardiovascular Research, Charite´ Medical Faculty, Berlin, Germany; 2Institute for Vegetative Physiology, Charite´ Medical Faculty, Berlin, Germany; 3Max Delbrueck Center for Molecular Medicine, Berlin, Germany; 4University of Texas Southwestern, Dallas, Texas, USA; 5Urology Department, Campus Benjamin Franklin, Charite´ Medical Faculty of the Charite´, Berlin, Germany and 6Experimental and Clinical Research Center, Berlin, Germany

20-Hydroxyeicosatetraenoic acid (20-HETE) production is increased in ischemic kidney tissue and may contribute to ischemia/reperfusion (I/R) injury by mediating vasoconstriction and inflammation. To test this hypothesis, uninephrectomized male Lewis rats were exposed to warm ischemia following pretreatment with either an inhibitor of 20-HETE synthesis (HET0016), an antagonist (20-hydroxyeicosa-6(Z),15(Z)-dienoic acid), an agonist (20-hydroxyeicosa-5(Z),14(Z)-dienoic acid), or vehicle via the renal artery and the kidneys were examined 2 days after reperfusion. Pretreatment with either the inhibitor or the antagonist attenuated I/R-induced renal dysfunction as shown by improved creatinine clearance and decreased plasma urea levels, compared to controls. The inhibitor and antagonist also markedly reduced tubular lesion scores, inflammatory cell infiltration, and tubular epithelial cell apoptosis. Administering the antagonist accelerated the recovery of medullary perfusion, as well as renal medullary and cortical re-oxygenation, during the early reperfusion phase. In contrast, the agonist did not improve renal injury and reversed the beneficial effect of the inhibitor. Thus, 20-HETE generation and its action mediated kidney injury due to I/R. Whether or not these effects are clinically important will need to be tested in appropriate human studies. Kidney International (2011) 79, 57–65; doi:10.1038/ki.2010.377; published online 20 October 2010 KEYWORDS: acute kidney injury; eicosanoids; ischemia/reperfusion

Correspondence: Wolf-Hagen Schunck, Max-Delbrueck Center for Molecular Medicine, Robert-Roessle-Str. 10, Berlin 13125, Germany. E-mail: [email protected] 7

These authors contributed equally to this work.

Received 19 October 2009; revised 23 June 2010; accepted 27 July 2010; published online 20 October 2010 Kidney International (2011) 79, 57–65

Ischemia/reperfusion (I/R)-induced acute kidney injury (AKI) leads to increased morbidity and mortality, particularly after cardiovascular surgery and kidney transplantation.1–3 I/R-induced mechanisms include persistent vasoconstriction, inflammation, endothelial dysfunction, and tubular injury.4,5 20-Hydroxyeicosatetraenoic acid (20-HETE), the product of cytochrome P450 (CYP)-catalyzed o-hydroxylation of arachidonic acid, modulates renal vascular and tubular function.6,7 Major sites of 20-HETE generation are preglomerular microvessels, where 20-HETE mediates constriction by inhibiting calciumactivated potassium channels, and the renal tubule, where 20-HETE promotes salt excretion by inhibiting Na þ -K þ -ATPase in proximal tubules and the Na þ -K þ -2Cl cotransporter in the thick ascending loop of Henle. 20-HETE is produced by CYP4A/ 4F isoforms, which require molecular oxygen, NADPH, and free arachidonic acid to catalyze the o-hydroxylation reaction.8 Thus, 20-HETE synthesis proceeds under normoxic conditions and is dependent on extracellular signal-induced activation of phospholipases A2, which release arachidonic acid from membrane phospholipids. Once produced, 20-HETE is partially re-esterified into phospholipids generating a membrane pool of preformed 20-HETE that is also accessible to phospholipases A2.9 These features link the biosynthesis, storage, and 20-HETE release to various signaling pathways triggered by vasoactive hormones and growth factors.7 On the other hand, the same features may be responsible for an excessive 20-HETE production under I/R conditions, which stimulate phospholipase A2 activities in the heart, brain, and kidney.10–12 Inhibition of 20-HETE synthesis and action reduced I/R injury in the heart and brain.13,14 20-HETE overproduction mediated the cytotoxic and proapoptotic effects of hypoxia on renal tubular epithelial cells.15 Vascular CYP4A overproduction caused endothelial dysfunction and hypertension by enhanced vasoconstriction, uncoupling of endothelial nitric oxide (NO) synthase, and nuclear factor (NF)-kB activation.16–18 These studies suggest that overproduction of 20-HETE has the potential to have a crucial role in the whole series of events leading to postischemic AKI. 57

original article

20-HETE (ng/ml)

0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0

#

In a first series of experiments, we used isolated perfused kidneys to analyze the effect of I/R on renal 20-HETE production. The experiments were performed in the absence (0.1% dimethylsulfoxide (DMSO) as vehicle control) or presence of the CYP4A inhibitor HET0016 (200 nmol/l); n ¼ 6 per group. Under basal conditions, vehicle-perfused kidneys constantly released low amounts of 20-HETE (0.16± 0.06 ng/ml) corresponding to about 0.8 ng/min at a flow rate of 5 ml/min. Twenty min of global ischemia induced a significant overproduction of 20-HETE (Figure 1a). 20-HETE release was increased more than fourfold during the first 5 min of reperfusion to 0.69±0.12 ng/ml (Po0.001 vs basal levels) and remained elevated until the end of the experiment (0.40±0.08 ng/ml). HET0016 significantly reduced I/R-induced 20-HETE overproduction to o50% (0.32±0.08 ng/ml at 5 min; Po0.05 vs vehicle control) and mediated a complete decline to basal levels after 30 min of reperfusion (Figure 1a). In a second series of experiments, we analyzed the release of 20-HETE in uninephrectomized animals under the conditions described below for inducing I/R injury in vivo (Figure 1b). Untreated control kidneys contained about 30 ng 20-HETE per g wet weight, of which more than 90% became accessible only after alkaline hydrolysis. These data show that the overwhelming part of 20-HETE was present in normal kidneys esterified to membrane phospholipids. Forty-five min of warm ischemia caused a strong increase of free 20-HETE in the kidneys pretreated with vehicle or 6,15-20HEDE (Po0.05 vs non-ischemic control, n ¼ 5–6 per group). In contrast, pretreatment with HET0016 resulted in significantly lower free 20-HETE levels in the ischemic kidneys (Po0.05 vs the vehicle and 6,15-20-HEDE groups, n ¼ 5–6 per group). Reperfusion was associated with a strong decrease in free 20-HETE and a concomitant increase of esterified 20-HETE in all groups, but most pronounced after HET0016 pretreatment (Figure 1b). 58

IR+vehicle IR+HET0016

#

#*

#*

* 5 min 10 min 15 min 30 min Time after ischemia

60

#

# #

40

* Esterified Free

*

20

$

$

$ $

0 NC

RESULTS I/R-induced 20-HETE release

# #

–5 min

20-HETE (ng/g)

We tested the hypothesis that targeting 20-HETE synthesis and action may protect the kidney against I/R-induced AKI. The experiments were performed in uninephrectomized Lewis rats to partially mimic the situation after renal transplantation, in which novel pharmacological interventions are sorely needed to promote early graft function and long-term survival. We used an inhibitor of CYP4A/4F enzymes to block 20-HETE synthesis (HET0016)19 and synthetic analogs to antagonize or mimic 20-HETE action.20 The analogs lack two of the four double bonds present in 20-HETE, rendering them more resistant against autoxidation and metabolism by cyclooxygenases and lipoxygenases. The agonist, 20-hydroxyeicosa-5(Z),14(Z)-dienoic acid (5,14-20-HEDE) and the antagonist, 20-hydroxyeicosa6(Z),15(Z)-dienoic acid (6,15-20-HEDE) differ only in double-bond positions and otherwise share identical physicochemical properties.

U Hoff et al.: 20-HETE and ischemia/reperfusion

VI

AR VR AI Animal group

HI

HR

Figure 1 | Ischemia/reperfusion (I/R)-induced release of 20-hydroxyeicosatetraenoic acid (20-HETE). (a) In isolated perfused kidneys, I/R induced an increased release of 20-HETE that was partially blocked by HET0016. Data are given as mean±s.e.m. (n ¼ 6 per group). #Po0.05 vs basal 20-HETE release 5 min before ischemia; *Po0.05 for vehicle vs HET0016. (b) In uninephrectomized Lewis rats, 45 min of warm ischemia induced increased renal levels of free 20-HETE after pretreating the animals with vehicle (VI group) or the 20-HETE antagonist, 6,15-20-HEDE (AI group) compared with non-ischemic controls (NC group). Ischemia in the presence of HET0016 resulted in significantly lower free 20-HETE levels (HI group). Free 20-HETE levels were decreased and esterified 20-HETE levels were increased 2 h after reperfusion (groups VR, AR, and HR). Statistically significant differences with Po0.05 were observed as indicated: * ¼ vs free 20-HETE in NC; # ¼ vs esterified 20-HETE in NC; $ ¼ vs free 20-HETE in VI and AI.

Role of 20-HETE in I/R-induced renal functional impairment

Increased serum creatinine and urea levels as well as decreased creatinine clearance indicate a significant decline in renal function, representative for our model of AKI (Figure 2). In vehicle-treated animals, I/R induced an eightfold increase of the serum creatinine and urea levels and reduced creatinine clearance to 20% compared with shamoperated uninephrectomized controls (0.26±0.02 vs 1.37± 0.04, Po0.001, n ¼ 8 per group). The vehicle itself (100 ml of saline containing 1% DMSO) had no effect compared with an animal group that did not receive any intrarenal infusion (creatinine: 2.41±0.23 vs 2.54±0.77 mg/dl; urea: 287.8± 34.8 vs 303.2±76.13 mg/dl; creatinine clearance: 0.26±0.02 vs 0.25±0.09 ml/min; all comparisons were not significant; eight vehicle treated vs six untreated animals). I/R-induced impairment in renal function (Figure 2) was significantly ameliorated by pretreatments with HET0016 (50 mg injection via the renal artery) or 6,15-20-HEDE (20 mg injection). The creatinine clearance was improved to 0.63±0.12 ml/min with HET0016 (Po0.01 vs vehicle; n ¼ 8 per group) and to 0.76±0.12 ml/min with 6,15-20-HEDE Kidney International (2011) 79, 57–65

original article

*xy

3

*

2

*#+

1

*#+y *#+

0 x

e 6 6 icl DE DE 01 01 0 eh T0 -HE -HE T v E + E 0 0 IR -2 +H -2 +H IR ,15 14 DE , +6 R+5 HE IR I 0-2 14 , +5 IR

-N

i Un

400 350 300 250 200 150 100 50 0

*xy *xy

* *#+

*#+

x le E E 16 16 i-N hic 00 HED HED 00 T Un ve T E 0+ E 0IR -2 +H -2 +H IR ,15 14 DE , +6 R+5 HE I IR 0-2 14 , +5 IR

Creatinine clearance (ml/min)

4

Serum urea (mg/dl)

Serum creatinine (mg/dl)

U Hoff et al.: 20-HETE and ischemia/reperfusion

1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0

*#+

*#+ *#+xy

* x

-N

i Un

e

E E 16 16 00 HED HED 00 T E 0 0 -2 -2 + +H IR ,15 14 DE , +6 R+5 HE I IR 0-2 4 ,1 +5 IR

eh +v

IR

*xy

icl

T HE

Figure 2 | Role of 20-hydroxyeicosatetraenoic acid (20-HETE) in ischemia/reperfusion (I/R)-induced loss of renal function. Serum levels of creatinine (a) and urea (b) and the creatinine clearance (c) were determined 2 days after reperfusion. HET0016 and 6,15-20-HEDE significantly improved I/R-induced renal dysfunction, whereas 5,14-20-HEDE was not protective and partially reversed the beneficial effect of HET0016. Data are given as mean±s.e.m. (n ¼ 8 per group). Statistically significant differences with Po0.01 were observed as indicated: * ¼ vs Uni-Nx, # ¼ vs IR þ vehicle, þ ¼ vs IR þ 5,14-20-HEDE, x ¼ vs IR þ HET0016 and y ¼ vs IR þ 6,15-20-HEDE.

(Po0.01 vs vehicle; n ¼ 8 per group). In contrast, 5,14-20HEDE (20 mg injection) did not improve renal dysfunction. The 20-HETE agonist partially reversed the protective effect of the CYP4A inhibitor as revealed by the animal group receiving a combined treatment of HET0016 and 5,14-20HEDE. This reversal was significant for both serum urea levels and creatinine clearance (each Po0.01 vs HET0016 alone; n ¼ 8 per group). Role of 20-HETE in I/R-induced renal tubular injury

I/R induced severe renal tubular damage as indicated by the occurrence of flattened tubular epithelium, exfoliated tubular epithelial cells, widened tubular lumina, hyaline cast formation, and necrotic tubules (Figure 3a). The degree of necrotic renal injury was significantly higher in the outer medulla (acute tubular necrosis (ATN) score: 2.50±0.20) compared with renal cortex (1.33±0.12); Figure 3b and c. The ATN scores were significantly reduced by administration of HET0016 (outer medulla: 1.04±0.31, Po0.001 vs vehicle; cortex: 0.58±0.27, Po0.05 vs vehicle). Also the 20-HETE antagonist 6,15-20-HEDE exerted a strong protective effect and reduced the ATN scores to 0.92±0.21 in the outer medulla and to 0.41±0.16 in the renal cortex (both Po0.001 vs vehicle). In contrast, tubular damage was as severe as in the vehicle control after infusing the 20-HETE agonist. The combined administration of 5,14-20-HEDE and HET0016 trended to result in enhanced ATN scores compared with HET0016 alone, however, without reaching statistical significance (Figure 3). Role of 20-HETE in I/R-induced apoptosis of tubular epithelial cells

I/R-induced renal injury resulted in a strong increase of apoptotic tubular epithelial cells (Figure 4a). Semiquantitative evaluation showed that apoptosis was more pronounced in the outer medulla (4.10±0.26% per field of view; FOV) than in the renal cortex (1.87±0.27% per FOV; Figure 4c and b). Kidney International (2011) 79, 57–65

The number of apoptotic cells was strongly reduced by HET0016 (outer medulla: 1.37±0.27%, cortex: 0.53±0.12% per FOV; both Po0.001 vs vehicle control) or 6,15-20-HEDE (outer medulla: 0.94±0.42%, cortex 0.52±0.21% per FOV; both Po0.001 vs vehicle). In contrast, animals pretreated with 5,14-20-HEDE showed a high degree of renal tubular apoptosis not different from the vehicle control. 5,14-20HEDE reversed the antiapoptotic effect of HET0016 comparing animals pretreated with HET0016 alone or with a combination of both drugs (Po0.01; Figure 4). Role of 20-HETE in the postischemic inflammatory response

I/R induced pronounced renal inflammation as indicated by a dense infiltration of monocytes/macrophages into the damaged zones of the outer medulla and renal cortex (Figure 5). Monocytes/macrophages were predominantly located in the tubulointerstitium adjacent to injured tubules and around small arteries (Figure 5a). Morphometric quantification displayed higher cell counts in the outer medulla (53.4±1.6 cells/high power field (HPF)) compared with renal cortex (46.8±2.0 cells/HPF) of vehicle-treated animals (Figure 5b and c). Application of HET0016 attenuated the response by about 50% (outer medulla: 24.7±4.7 and cortex: 15.2±1.7 cells/ HPF, both Po0.001 vs vehicle control). The 20-HETE antagonist, 6,15-20-HEDE, also markedly reduced ED1 þ infiltrates to 19.0±5.9 in the outer medulla and to 15.5±4.4 cells/HPF in the cortex (both Po0.001 vs vehicle). The 20HETE agonist, 5,14-20-HEDE, did not inhibit cell infiltration and rather trended to further aggravate inflammation in the renal cortex (51.7±2.4 cells/HPF, P ¼ 0.132) and significantly in the outer medulla (61.5±2.0 cells/HPF, Po0.05 vs vehicle control). 5,14-20-HEDE clearly reversed the anti-inflammatory effect of CYP4A inhibition (outer medulla: 42.6±1.9 and cortex 33.8±1.6 cells/HPF, both Po0.05 compared with HET0016 alone). 59

original article

U Hoff et al.: 20-HETE and ischemia/reperfusion

Control

IR+vehicle

IR+6,15-20-HEDE

3

#+y

2 #+

#+

1 0

xy

2

#+

#+

1

#+

0 e icl

eh +v

IR

IR+5,14-20-HEDE+HET0016

3

xy ATN-score - cortex

ATN-score - outer medulla

IR+HET0016

IR+5,14-20-HEDE

16 00

ET

+H

IR

0-

-2

5 ,1 +6

IR

DE

HE

0-

HE

-2

4 ,1 +5

IR

0-

T HE

+

E ED

H

-2

14 5,

+

IR

e

6 01

icl

16 00

DE

eh +v

IR

E

ED

0 ET

+H

IR

H 0-

-2

-2

5 ,1 +6

IR

ED

H 0-

4 ,1 +5

IR

16 00

E T HE

+ DE E H

0-

-2

4 ,1

+5

IR

Figure 3 | Role of 20-hydroxyeicosatetraenoic acid (20-HETE) in ischemia/reperfusion (I/R)-induced tubular injury. (a) Representative images of hematoxylin- and eosin-stained sections of the outer medulla of kidneys harvested from the different treatment groups 2 days after inducing renal I/R injury. Original magnification 100  . (b, c) Evaluation of tubular necrosis in the outer medulla and renal cortex using an acute tubular necrosis (ATN) score. HET0016 and 6,15-20-HEDE significantly ameliorated tubular damage, whereas 5,14-20-HEDE was not protective. Significant differences with Po0.05 were observed as indicated: # ¼ vs IR þ vehicle, þ ¼ vs IR þ 5,14-20-HEDE, x ¼ vs IR þ HET0016 and y ¼ vs IR þ 6,15-20-HEDE.

20-HETE, I/R-induced renal hemodynamic changes, and intrarenal oxygenation

Renal artery and vein occlusion caused an immediate decrease in total renal blood flow, as well as of the Laserfluxes that characterize regional perfusion in the renal cortex and medulla (Figure 6a–c). Moreover, cortical and medullary pO2 instantly dropped to about 0.1 mm Hg followed by a slow increase to about 2 mm Hg during 45 min of ischemia (Figure 6d and e). Upon reperfusion, total renal blood flow steeply increased to almost 50% and then slowly recovered to about 90% of the baseline flow. Systemic mean arterial blood pressure remained stable throughout the entire observation period (baseline average 97.6±3.5 mm Hg). These features were not significantly different, comparing rats receiving intrarenal injections of vehicle or 6,15-20-HEDE before inducing ischemia. 60

However, the two animal groups showed marked differences in the rate and extent of intrarenal reoxygenation during the reperfusion phase (Figure 6d and e). In the renal cortex and medulla of 6,15-20-HEDE-pretreated animals, oxygen pressure rapidly reached and then even slightly exceeded the baseline levels before ischemia. In the vehicle group, the intrarenal pO2 recovered only partially and reached 2 h after reperfusion about 80% (cortex) and 60% (medulla) of the baseline values. In particular, reoxygenation in medulla occurred with significant delay as compared with the 6,15-20-HEDE group (Figure 6d). 6,15-20-HEDE also significantly accelerated the recovery of medullary perfusion that remained, however, below the baseline levels (Figure 6c). The effect of the 20-HETE antagonist was less pronounced in the cortex, in which the difference in regional Laser-fluxes did not reach significance compared with vehicle (Figure 6b). Kidney International (2011) 79, 57–65

original article

U Hoff et al.: 20-HETE and ischemia/reperfusion

Control

IR+vehicle

IR+5,14-20-HEDE

IR+HET0016

IR+6,15-20-HEDE

IR + 5,14-20-HEDE+HET0016

*xy

5

*

4

*#+xy

3

*#+ *#+

2 1

TUNEL + area (% of view field)

4

*xy

3

*

2

*#+xy *#+ *#+

1

U ni -N x +v eh IR i + c IR H ET le +6 00 ,1 516 IR 20 IR +5 -H + ,1 5 ED 4,1 420 E 20 -H -H ED E E+ D E H ET 00 16 IR

U ni -N x +v eh IR i cl +H e IR ET +6 00 ,1 16 IR IR 5-2 +5 0+5 ,1 H , 14 4ED 20 -2 0- E -H H ED ED E+ E H ET 00 16

0

0

IR

TUNEL + area (% of view field)

5

Figure 4 | Role of 20-hydroxyeicosatetraenoic acid (20-HETE) in ischemia/reperfusion (I/R)-induced apoptosis of tubular epithelial cells. (a) Representative images of outer medullary sections stained using the TdT-mediated dUTP nick end labeling assay to evaluate the presence of apoptotic nuclei. Original magnification 400  . (b, c) Quantification of apoptosis in the outer medulla and renal cortex by counting the percentage of apoptotic cells per high power view field. HET0016 and 6,15-20-HEDE significantly protected against I/R-induced apoptosis, whereas 5,14-20-HEDE did not. 5,14-20-HEDE significantly reversed the antiapoptotic effect of HET0016. Significant differences with Po0.01 were observed as indicated: * ¼ vs Uni-Nx; # ¼ vs IR þ vehicle; þ ¼ vs IR þ 5,14-20-HEDE; x ¼ vs IR þ HET0016; y ¼ vs IR þ 6,15-20-HEDE.

DISCUSSION

We found that inhibiting the generation and action of 20-HETE can significantly ameliorate experimental AKI. Treatments with a CYP4A/4F-inhibitor (HET0016) or a 20-HETE antagonist (6,15-20-HEDE) protected against I/Rinduced vascular inflammation, tubular injury, and loss of renal function. In contrast, the structurally related 20-HETE agonist (5,14-20-HEDE) did not improve renal I/R injury but instead partially reversed the beneficial effects of HET0016-mediated inhibition of 20-HETE synthesis. Our data on the renal production of 20-HETE demonstrate that this CYP-eicosanoid is indeed overproduced under I/R conditions. Probably, both phospholipase-mediated release of stored 20-HETE and de novo synthesis contributed to the large increase in free 20-HETE that reached calculated global concentrations of 50–80 nM in the ischemic kidneys. Kidney International (2011) 79, 57–65

20-HETE efficiently uncouples endothelial NO synthase and activates nuclear factor-kB at concentrations as low as 1 nM.17,18 Moreover, 20-HETE significantly constricts rat renal afferent arterioles with a threshold concentration of 1 nM.21 Thus, the amounts of free 20-HETE present in ischemic kidneys appear more than sufficient to trigger important early events in the development of AKI, such as endothelial dysfunction, inflammation, and vasoconstriction. Moreover, studies in the heart indicated that 20-HETE acts as an inhibitor of adenosine triphosphate (ATP)-sensitive potassium (KATP)-channels.22 Opening of mitochondrial and plasmalemmal KATP channels is a key component of survival mechanisms in several cell types23 and also improves I/R-induced renal injury.24 Our pharmacological interventions aimed at inhibiting 20-HETE synthesis and blocking 20-HETE action were also 61

U Hoff et al.: 20-HETE and ischemia/reperfusion

Control

IR+vehicle

IR+5,14-20-HEDE

IR+HET0016

IR+6,15-20-HEDE

IR+5,14-20-HEDE+HET0016

* *#+xy

50 40 30

*#+

*#+

20 10 U ni -N IR x +v eh IR ic le +H IR ET +6 00 ,1 516 IR 2 I R 0+5 +5 H ,1 ED ,1 4420 E 20 -H -H ED ED E+ E H ET 00 16

0

*xy

60

*

50

*#+xy

40 30

*#+

20

*#+

10 0 U ni -N x +v eh IR ic le +H IR ET +6 00 ,1 516 IR 20 IR +5 + H 5, ,1 ED 14 420 E -2 0-H H ED ED E+ E H ET 00 16

60

70

*#xy

IR

70

ED1+ cells (cells/HPF)

ED1+ cells (cells/HPF)

original article

Figure 5 | Role of 20-hydroxyeicosatetraenoic acid (20-HETE) in renal inflammation. (a) Representative images of outer medullary sections stained for ED1 þ monocytes/macrophages to evaluate inflammatory cell infiltration as induced by renal ischemia/ reperfusion (I/R) injury in the different animal groups. Original magnification 100  . (b, c) Quantification of inflammatory cell infiltration in the outer medulla and renal cortex by counting the number of ED1 þ cells per high power view field (HPF). HET0016 and 6,15-20-HEDE significantly decreased I/R-induced inflammatory cell infiltration. In contrast, 5,14-20-HEDE significantly increased the inflammatory response in the outer medulla and reversed the anti-inflammatory effect of HET0016 both in the outer medulla and renal cortex. Significant differences with Po0.01 were observed as indicated: * ¼ vs Uni-Nx; # ¼ vs IR þ vehicle; þ ¼ vs IR þ 5,14-20-HEDE; x ¼ vs IR þ HET0016; y ¼ vs IR þ 6,15-20-HEDE.

effective in preserving renal function. The data on necrosis, inflammation, and apoptosis suggest that blockade of 20-HETE action may be more promising than inhibition of 20-HETE synthesis in preventing renal injury. This difference may be due to the fact that HET0016 only mediates inhibition of 20-HETE de novo synthesis, whereas 6,15-20HEDE has the capacity to antagonize the action of total free 20-HETE including preformed 20-HETE released from membrane stores under ischemia. We found that inhibition of 20-HETE actions significantly accelerates and improves medullary and cortical reoxygenation in the early reperfusion phase. In this way, 20-HETE blockade largely reduces the time that these regions are exposed to severe tissue hypoxia, 62

despite the onset of reperfusion. Intrarenal oxygenation is regulated by a variety of factors including local perfusion, local oxygen consumption, and arterial-to-venous oxygen shunting.25 In our model, inhibition of 20-HETE action indeed promotes the recovery of medullary blood flow. This finding is in agreement with the vasoconstrictor role of 20-HETE in renal arterioles21 and with the capacity of 20-HETE to reduce the formation of vasodilatory NO,17 as discussed above. As NO inhibits CYP enzymes, NO-deficiency leads to enhanced 20-HETE synthesis.26 These 20-HETE/NO interactions may prompt a vicious cycle that progressively impairs endothelial function and results in persistent vasoconstriction during I/R. Surprisingly, we found Kidney International (2011) 79, 57–65

original article

U Hoff et al.: 20-HETE and ischemia/reperfusion

1.0 0.8 0.6 0.4 Vehicle 6,15-20-HEDE

0.0 0

20 40 60 80 100 120 140 160 Time (min)

Cortical pO2 (mm Hg)

d

16 14 12 10 8 6 4 2 0

1.4 1.2 1.0 0.8 0.6 0.4 0.2

Vehicle 6,15-20-HEDE

0.0 0

1.2 *

1.0

*

0.8 0.6 0.4 0.2 *

0.0

20 40 60 80 100 120 140 160 Time (min)

0

Vehicle 6,15-20-HEDE P < 0.05

20 40 60 80 100 120 140 160 Time (min)

e *

Vehicle 6,15-20-HEDE P < 0.05

*

Medullary pO2 (mm Hg)

0.2

c

1.6

Medullary flux (rel. change)

b

1.2

Cortical flux (rel. change)

Renal blood flow (rel. change)

a

12 10

*

* Vehicle 6,15-20-HEDE P < 0.05

*

*

*

8 6 4 2 0

0

20 40 60 80 100 120 140 160 Time (min)

0

20 40 60 80 100 120 140 160 Time (min)

Figure 6 | Role of 20-hydroxyeicosatetraenoic acid (20-HETE) in ischemia/reperfusion (I/R)-induced renal hemodynamic changes and local tissue oxygenation. (a–c) Time course of relative changes in total renal blood flow (average baseline 2.9±0.3 ml/min), local cortical Laser-flux, and local medullary Laser-flux. (d, e) Time course of local cortical pO2 and local medullary pO2. Renal ischemia (from time 0–45 min) was induced 5 min after pretreating the animals by intrarenal injections of vehicle or 6,15-20-HEDE followed by 2 h of reperfusion (n ¼ 7 animals per group). Note the significantly accelerated and improved recovery of medullary perfusion (c) and of the cortical and medullary oxygenation (d and e) in the 6,15-20-HEDE compared with the vehicle-treated animals.

that reoxygenation in the presence of the 20-HETE antagonist leads to intrarenal pO2 values that even exceeded those measured under baseline conditions. Although the actual mechanisms are unclear, we were interested to note that 20-HETE-mediated changes in NO bioavailability could possibly also affect tubular oxygen consumption, as NO may inhibit sodium reabsorption and increase the efficiency of nephron oxygen utilization.25 The results of our study are paralleled by previous studies in the heart13,22 and brain,14 suggesting that inhibition of 20-HETE generation and action may provide a general therapeutic option for the prevention and treatment of ischemic organ injury. However, Roman and associates recently reported that HET0016 exacerbated and 20-HETE agonists attenuated I/R-induced renal injury.27 These findings that obviously contradict our results may be related to the use of different models (Sprague–Dawley rats with bilateral clamping of two kidneys vs uninephrectomized Lewis inbred rats in our study) and treatment forms (systemic high-dose compared with local low-dose drug application). Moreover, Regner et al.27 solely used 20-HETE agonists and did not compare their effects with those of 20-HETE antagonists as performed in our study. A specific concern associated with high systemic HET0016 levels may arise from the potential of this CYP-inhibitor to inhibit not only 20-HETE synthesis, but also the o-hydroxylation and inactivation of leukotriene B4 that functions as neutrophil chemoattractant, and thereby aggravates ischemic renal damage.28 Kidney International (2011) 79, 57–65

However, the contradictory findings most probably reflect the complex vascular and tubular roles of 20-HETE in the kidney. This very problem also led to apparently contradictory findings on the pro- and antihypertensive roles of 20-HETE. Spontaneously hypertensive rats and salt-sensitive Dahl rats develop hypertension that is associated with either overproduction or deficiency of 20-HETE.29,30 This apparent contradiction has been largely resolved recognizing that vascular overproduction of 20-HETE induces vasoconstriction, endothelial dysfunction, and hypertension, whereas tubular deficiency of 20-HETE impairs salt excretion and thus also causes hypertension. This dual and site-specific role of 20-HETE is similarly reflected by opposite effects of HET0016 in different models, namely, reduction of blood pressure in androgen- and cyclosporin A-hypertensive rats vs induction of salt-sensitive hypertension in normal Sprague–Dawley rats.31–33 We propose that 20-HETE may have different roles during AKI development. We intentionally injected low doses (mg/kg range) of HET0016 and of the 20-HETE analogs directly into the renal artery before occlusion, to achieve effective drug levels during ischemia with rapid washout upon reperfusion. This strategy blocked the effects of 20-HETE overproduction. Regner et al.27 used high doses of HET0016 and 5,14-20HEDE (mg/kg range) subcutaneously, resulting in high systemic drug levels that acted not only during ischemia but also during reperfusion. Under these conditions, 20-HETE agonists may protect the kidney against I/R injury 63

original article

primarily in the reperfusion phase by inhibiting sodium reabsorption, thereby limiting tubular oxygen consumption. Roman and associates also reported that HET0016 blocked, whereas 20-HETE exacerbated, the cytotoxic and proapoptotic effects of hypoxia on tubular epithelial cells.15 These in vitro data are in agreement with the results of our in vivo study. Taken together, the findings suggest a rather strict therapeutic window in terms of dosage and timing that determines the outcome of pharmacological interventions targeting the complex CYP/20–HETE system in the kidney. We believe our data could offer a therapeutic perspective. I/R-induced AKI is common after cardiovascular and kidney transplant surgery, clinical situations that could be amenable to the experimental approaches used here. Noteworthy, a recent clinical study demonstrated that the extent of 20-HETE release occurring within the first 5 min of allograft reperfusion is a negative predictor of post-transplant allograft function.34 However, further studies are required to strictly define the therapeutic window for 20-HETE antagonists or agonists. Moreover, cell-type-specific drug delivery could avoid potential adverse effects that are clearly predictable recognizing the site-specific and partially opposite effects of 20-HETE in the kidney. These conclusions may apply in general to all therapeutic strategies aimed at targeting 20-HETE synthesis and action, as this CYP-eicosanoid contributes to the development of endothelial dysfunction, vascular inflammation, uncontrolled angiogenesis, hypertension, and ischemic organ injury, but is nonetheless required to maintain renal tubular function.35 METHODS Animal model Inbred male Lewis (Lew, RT1) rats, aged 10 weeks, were kept under standard conditions. Local authorities approved all experimental procedures according to American Physiological Society guidelines. After removal of the right kidney, HET0016 (50 mg), 6,15-20-HEDE (20 mg), 5,14-20-HEDE (20 mg), a combination of HET0016 (50 mg) and 5,14-20-HEDE (20 mg) or vehicle (n ¼ 8 per group) were administered in the left renal artery. HET0016 was purchased from Cayman Chemicals (Biomol GmbH, Hamburg, Germany) and the 20-HETE analogs were synthesized as described previously.20 Stock solutions were prepared in DMSO and further diluted 100-fold with 0.9% NaCl solution to give the indicated amounts in a volume of 100 ml for intrarenal injection. A quantity of 100 ml of 0.9% NaCl containing 1% DMSO served as vehicle control. The remaining left kidney was exposed to 45 min of warm ischemia by clamping renal artery and vein 5 min after intrarenal drug administration. One day after I/R, animals were set into metabolic cages for urine collection over a period of 24 h. Organs were harvested and blood samples were taken 2 days post-I/R. 20-HETE release In isolated perfused kidneys. The renal artery of the left isolated kidney was cannulated and the organ was perfused in an organ chamber at 37 1C at constant flow (5 ml/min).36 The oxygenated (95% O2–5% CO2) perfusion buffer contained either vehicle (0.1% DMSO) or HET0016 (200 nmol/l); n ¼ 6 kidneys per group. After an equilibration period of 20 min, perfusion was 64

U Hoff et al.: 20-HETE and ischemia/reperfusion

stopped for 20 min to mimic warm ischemia followed by 30 min of reperfusion. The total renal outflow was collected in 5 min intervals before and up to 30 min after inducing ischemia. 20-HETE-d6 (Cayman Chemicals) was added to the perfusate samples as internal standard. Solid phase extraction and quantification of 20-HETE by liquid chromatography tandem mass spectrometry (Lipidomix GmbH, Berlin, Germany) were performed as described previously.37 In vivo during I/R. Rats were uninephrectomized, pretreated with vehicle, 6,15-20-HEDE or HET0016, and ischemia was induced in the remaining left kidney as described above. The ischemic kidneys were removed and immediately frozen in liquid nitrogen either directly after 45 min of ischemia or 2 h after reperfusion (n ¼ 5–6 animals per treatment group and time point). Total kidneys were then pulverized in liquid nitrogen and 30 mg aliquots were used for subsequent liquid chromatography tandem mass spectrometry analysis.37 To differentiate between esterified and free 20-HETE, the samples were extracted with or without previous alkaline hydrolysis. Renal pathology, immunohistochemistry, and morphometry Formalin-fixed and paraffin-embedded renal sections (2 mm) stained with hematoxylin and eosin and periodic acid-Schiff (PAS) were graded according to previously described score for ATN.38 Renal apoptosis was examined in fluorescein-stained formalin-fixed renal sections using the in situ Cell Death Detection Kit (Roche Diagnostics, Indianapolis, IN). TdT-mediated dUTP nick end labeling-positive cells were visualized under a fluorescence microscope (Zeiss Axio Imager A1, Jena, Germany) and quantified using a digital imaging system (Zeiss Axiocam HR with Axiovision 4.4 software). The positively immunostained area was binarized and calculated as percent per FOV at 400  magnification in renal cortex and outer medulla. Monocyte staining was performed with the monoclonal mouse antibody ED-1 (Serotec, Oxford, UK) and visualized using the alkaline-phosphatase-anti-alkaline-phosphatase method. The number of ED1 þ cells was scored in 10 randomly chosen FOVs in cortex and outer medulla. Results were expressed as mean±s.e.m. of cells/FOV at 400  magnification HPF. In vivo measurement of renal hemodynamics and tissue oxygenation Urethane intraperitoneally (1.2 g/kg body mass; Sigma, Taufkirchen, Germany) caused anesthesia. Rats were maintained at 37 1C and breathed spontaneously. After right nephrectomy, a non-obstructive catheter was placed into the abdominal aorta (tip at the level of the remaining left renal artery) for recording of arterial blood pressure (Gould, Valley View, OH) and for intrarenal drug administration. A soft ribbon was loosely slung around the left renal artery and vein. An ultrasound transit time flow probe (Transonic Systems, Ithaca, NY) was positioned around the left renal artery to measure renal blood flow. Two combined optical Laser-Doppler-flux- and pO2-probes attached to an OxyLite/OxyFlo-apparatus (Oxford Optronics, Oxford, UK) were used to measure local flux of erythrocytes (Laser-Doppler-signal) and local pO2. As described previously,39 by means of micromanipulators, one of the combined probes was inserted into the kidney to a depth of approximately 2 mm, thus providing values for the subjacent cortical tissue, whereas the tip of the other probe was advanced to approximately 4 mm, thus providing values for the subjacent medullary tissue. After obtaining baseline values, intrarenal bolus injections of either 6,15-20-HEDE (20 mg) or vehicle (n ¼ 7 rats per group) were Kidney International (2011) 79, 57–65

original article

U Hoff et al.: 20-HETE and ischemia/reperfusion

performed. Five min later, renal ischemia was initiated by tightening the ribbon around the left renal artery and vein, and ischemia was maintained for 45 min. All parameters were continuously recorded at a sampling rate of 50 Hz until 2 h after the end of ischemia time. Statistical analysis Statistical analysis was performed using non-parametric Mann– Whitney U-test with paired comparisons for more than two groups. Po0.05 was considered significant. All statistical tests were performed using SPSS 11.5 for Windows (SPSS, Chicago, IL).

16.

17.

18.

19.

20.

DISCLOSURE

All the authors declared no competing interests.

21.

ACKNOWLEDGMENTS

22.

We thank Christel Andre´e and Ramona Zummach for excellent technical assistance. We are grateful to Michael Rothe and Katrin Blossey, Lipidomix GmbH (Berlin, Germany) for performing the liquid chromatography tandem mass spectrometry analysis. This study was supported in part by grants of the Werner Jacksta¨dt-Stiftung, Wuppertal, Germany to UH, of the BMWi, Germany (KF0568201UL7) to DD and WHS, and by USPHS NIH (GM31278) and the Robert A. Welch Foundation to JRF. REFERENCES 1. Lameire N, Van Biesen W, Vanholder R. Acute renal failure. Lancet 2005; 365: 417–430. 2. Aydin Z, van Zonneveld AJ, de Fijter JW et al. New horizons in prevention and treatment of ischaemic injury to kidney transplants. Nephrol Dial Transplant 2007; 22: 342–346. 3. Karkouti K, Wijeysundera DN, Yau TM et al. Acute kidney injury after cardiac surgery: focus on modifiable risk factors. Circulation 2009; 119: 495–502. 4. Sutton TA, Fisher CJ, Molitoris BA. Microvascular endothelial injury and dysfunction during ischemic acute renal failure. Kidney Int 2002; 62: 1539–1549. 5. Bonventre JV, Zuk A. Ischemic acute renal failure: an inflammatory disease? Kidney Int 2004; 66: 480–485. 6. McGiff JC, Quilley J. 20-HETE and the kidney: resolution of old problems and new beginnings. Am J Physiol 1999; 277: R607–R623. 7. Roman RJ. P-450 metabolites of arachidonic acid in the control of cardiovascular function. Physiol Rev 2002; 82: 131–185. 8. Capdevila JH, Falck JR. Biochemical and molecular properties of the cytochrome P450 arachidonic acid monooxygenases. Prostaglandins Other Lipid Mediat 2002; 68–69: 325–344. 9. Carroll MA, Balazy M, Huang DD et al. Cytochrome P450-derived renal HETEs: storage and release. Kidney Int 1997; 51: 1696–1702. 10. Jenkins CM, Cedars A, Gross RW. Eicosanoid signalling pathways in the heart. Cardiovasc Res 2009; 82: 240–249. 11. Phillis JW, O’Regan MH. A potentially critical role of phospholipases in central nervous system ischemic, traumatic, and neurodegenerative disorders. Brain Res Brain Res Rev 2004; 44: 13–47. 12. Nakamura H, Nemenoff RA, Gronich JH et al. Subcellular characteristics of phospholipase A2 activity in the rat kidney. Enhanced cytosolic, mitochondrial, and microsomal phospholipase A2 enzymatic activity after renal ischemia and reperfusion. J Clin Invest 1991; 87: 1810–1818. 13. Gross GJ, Falck JR, Gross ER et al. Cytochrome P450 and arachidonic acid metabolites: role in myocardial ischemia/reperfusion injury revisited. Cardiovasc Res 2005; 68: 18–25. 14. Miyata N, Seki T, Tanaka Y et al. Beneficial effects of a new 20hydroxyeicosatetraenoic acid synthesis inhibitor, TS-011 [N-(3-chloro-4morpholin-4-yl) phenyl-N’-hydroxyimido formamide], on hemorrhagic and ischemic stroke. J Pharmacol Exp Ther 2005; 314: 77–85. 15. Nilakantan V, Maenpaa C, Jia G et al. 20-HETE-mediated cytotoxicity and apoptosis in ischemic kidney epithelial cells. Am J Physiol Renal Physiol 2008; 294: F562–F570.

Kidney International (2011) 79, 57–65

23. 24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

Wang JS, Singh H, Zhang F et al. Endothelial dysfunction and hypertension in rats transduced with CYP4A2 adenovirus. Circ Res 2006; 98: 962–969. Cheng J, Ou JS, Singh H et al. 20-hydroxyeicosatetraenoic acid causes endothelial dysfunction via eNOS uncoupling. Am J Physiol Heart Circ Physiol 2008; 294: H1018–H1026. Ishizuka T, Cheng J, Singh H et al. 20-Hydroxyeicosatetraenoic acid stimulates nuclear factor-kappaB activation and the production of inflammatory cytokines in human endothelial cells. J Pharmacol Exp Ther 2008; 324: 103–110. Sato M, Ishii T, Kobayashi-Matsunaga Y et al. Discovery of a N’hydroxyphenylformamidine derivative HET0016 as a potent and selective 20-HETE synthase inhibitor. Bioorg Med Chem Lett 2001; 11: 2993–2995. Alonso-Galicia M, Falck JR, Reddy KM et al. 20-HETE agonists and antagonists in the renal circulation. Am J Physiol 1999; 277: F790–F796. Imig JD, Zou AP, Stec DE et al. Formation and actions of 20hydroxyeicosatetraenoic acid in rat renal arterioles. Am J Physiol 1996; 270: R217–R227. Gross ER, Nithipatikom K, Hsu AK et al. Cytochrome P450 omegahydroxylase inhibition reduces infarct size during reperfusion via the sarcolemmal KATP channel. J Mol Cell Cardiol 2004; 37: 1245–1249. Rodrigo GC, Standen NB. ATP-sensitive potassium channels. Curr Pharm Des 2005; 11: 1915–1940. Sun Z, Zhang X, Ito K et al. Amelioration of oxidative mitochondrial DNA damage and deletion after renal ischemic injury by the KATP channel opener diazoxide. Am J Physiol Renal Physiol 2008; 294: F491–F498. Evans RG, Gardiner BS, Smith DW et al. Intrarenal oxygenation: unique challenges and the biophysical basis of homeostasis. Am J Physiol Renal Physiol 2008; 295: F1259–F1270. Alonso-Galicia M, Sun CW, Falck JR et al. Contribution of 20-HETE to the vasodilator actions of nitric oxide in renal arteries. Am J Physiol 1998; 275: F370–F378. Regner KR, Zuk A, Van Why SK et al. Protective effect of 20-HETE analogues in experimental renal ischemia reperfusion injury. Kidney Int 2009; 75: 511–517. Noiri E, Yokomizo T, Nakao A et al. An in vivo approach showing the chemotactic activity of leukotriene B(4) in acute renal ischemicreperfusion injury. Proc Natl Acad Sci USA 2000; 97: 823–828. Sacerdoti D, Escalante B, Abraham NG et al. Treatment with tin prevents the development of hypertension in spontaneously hypertensive rats. Science 1989; 243: 388–390. Roman RJ, Alonso-Galicia M, Wilson TW. Renal P450 metabolites of arachidonic acid and the development of hypertension in Dahl saltsensitive rats. Am J Hypertens 1997; 10: 63S–67S. Singh H, Cheng J, Deng H et al. Vascular cytochrome P450 4A expression and 20-hydroxyeicosatetraenoic acid synthesis contribute to endothelial dysfunction in androgen-induced hypertension. Hypertension 2007; 50: 123–129. Blanton A, Nsaif R, Hercule H et al. Nitric oxide/cytochrome P450 interactions in cyclosporin A-induced effects in the rat. J Hypertens 2006; 24: 1865–1872. Hoagland KM, Flasch AK, Roman RJ. Inhibitors of 20-HETE formation promote salt-sensitive hypertension in rats. Hypertension 2003; 42: 669–673. Dolegowska B, Blogowski W, Domanski L. Is it possible to predict the early post-transplant allograft function using 20-HETE measurements? A preliminary report. Transpl Int 2009; 22: 546–553. Doggrell SA. Taking the 20-HETE out of the cardiovascular system: the potential of 20-HETE synthesis inhibitors. Curr Opin Investig Drugs 2005; 6: 901–906. Mederos y Schnitzler M, Storch U, Meibers S et al. Gq-coupled receptors as mechanosensors mediating myogenic vasoconstriction. EMBO J 2008; 27: 3092–3103. Kulas J, Schmidt C, Rothe M et al. Cytochrome P450-dependent metabolism of eicosapentaenoic acid in the nematode Caenorhabditis elegans. Arch Biochem Biophys 2008; 472: 65–75. Dragun D, Hoff U, Park JK et al. Prolonged cold preservation augments vascular injury independent of renal transplant immunogenicity and function. Kidney Int 2001; 60: 1173–1181. Seeliger E, Flemming B, Wronski T et al. Viscosity of contrast media perturbs renal hemodynamics. J Am Soc Nephrol 2007; 18: 2912–2920.

65