Acute Kidney Injury Resolvin D Series and Protectin D1 Mitigate

The Journal of Immunology

Resolvin D Series and Protectin D1 Mitigate Acute Kidney Injury1 Jeremy S. Duffield,2,3* Song Hong,3† Vishal S. Vaidya,* Yan Lu,† Gabrielle Fredman,† Charles N. Serhan,2† and Joseph V. Bonventre* Omega-3 fatty acid docosahexaenoic acid is converted to potent resolvins (Rv) and protectin D1 (PD1), two newly identified families of natural mediators of resolution of inflammation. We report that, in response to bilateral ischemia/reperfusion injury, mouse kidneys produce D series resolvins (RvDs) and PD1. Administration of RvDs or PD1 to mice before the ischemia resulted in a reduction in functional and morphological kidney injury. Initiation of RvDs and RvD1 administration 10 min after reperfusion also resulted in protection of the kidney as measured by serum creatinine 24 and 48 h later. Interstitial fibrosis after ischemia/reperfusion was reduced in mice treated with RvDs. Both RvDs and PD1 reduced the number of infiltrating leukocytes and blocked TLR-mediated activation of macrophages. Thus, the renal production of Rv and protectins, a previously unrecognized endogenous anti-inflammatory response, may play an important role in protection against and resolution of acute kidney injury. These data may also have therapeutic implications for potentiation of recovery from acute kidney injury. The Journal of Immunology, 2006, 177: 5902–5911.

I

njury to the kidney, even of a relatively modest degree, is a major health burden accounting for many hospital admissions and high morbidity and mortality (1–3). Although there have been major advances in the management of acute kidney injury over the past 50 years, treatment is largely supportive. In many cases, there is complete failure or partial failure of resolution with important implications for mortality and, in some cases, progression to end stage renal disease (4). Over the last decade, it has been appreciated that acute kidney injury is an inflammatory disease, with inflammation contributing to small vessel congestion and the perpetuation of the functional deficiencies (5–9). Even with repair and regeneration following injury, emerging evidence indicates that postinflammatory scarring of the kidney may be a major factor contributing to chronic renal disease (10 –12). It has also been postulated that ischemic inflammatory kidney injury is a central component of many progressive diseases of the kidney (13–15). In murine models of ischemic and toxic acute kidney injury, the innate immune response plays a key role in disease progression as demonstrated by the protective effect of Abs that block integrin and integrin receptor interactions necessary for diapedesis of leukocytes as well as the protective effect of depletion of complement *Renal Division, Department of Medicine, and †Center for Experimental Therapeutics and Reperfusion Injury, Department of Anesthesiology, Perioperative and Pain Medicine, Brigham and Women’s Hospital and Harvard Medical School, Boston, MA 02115 Received for publication February 1, 2006. Accepted for publication August 15, 2006.

factors such as C5a (5, 7). Similar to the innate inflammatory response in other settings, neutrophils (polymorphonuclear neutrophils (PMNs))4 predominate in the first 24 h of injury, being replaced subsequently by monocytes/macrophages and T cells (5, 16 –18). To date, several compounds have been shown in animal models to ameliorate acute kidney injury, but in clinical trials these reagents have had limited efficacy (19 –22). One of the features of injury to the kidney may be redundancy in proinflammatory pathways. Therefore, blockade of one pathway and/or target alone may be ineffective. The focus of therapy has to date been primarily directed at proinflammatory factors and there has been little attention to the understanding of the endogenous factors normally involved in the resolution of inflammation or an attempt to facilitate this process therapeutically. Recently, it has been recognized that endogenous antiinflammatory lipid mediators can be generated which are shortlived autacoids derived from precursor omega-3 fatty acids (23, 24). Production of these mediators can be enhanced by aspirin. These novel families of compounds were termed resolvins (Rv) and protectins (25) because they derive from docosahexaenoic acid (DHA) and act to resolve inflammation (23). During the resolution of inflammation in mice treated with aspirin, endothelial-neutrophil (PMN) interactions were required for the generation of some of these compounds by human cells (23). PMNs generate both bioactive D series Rv (RvDs) and protectins from DHA and precursors present in exudates (23, 26).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 This work was supported in part by a Senior Fellowship from the National Kidney Research Fund (to J.S.D.), Grants DK-73299 (to J.S.D.), DK-39773 (to J.V.B.), DK38452 (to J.V.B.), GM-38765 (to C.N.S.), DK-074448 (to C.N.S.), and P50 DE016191 (to S.H., Y.L., and C.N.S.) from the National Institutes of Health, and Grant 0535492 (to V.S.V.) from the American Heart Association.

J.S.D. and S.H. and both laboratories contributed equally to this work.

4 Abbreviations used in this paper: PMN, polymorphonuclear neutrophil; DHA, docosahexaenoic acid; I/R, ischemia/reperfusion; PD1, protectin D1, 10R,17S-dihydroxy-docosa-4Z,7Z,11E,13E,15Z,19Z-hexaenoic acid, also termed neuroprotectin D1 (NPD1) when produced in neural tissues (see Ref. 49); PLP, paraformaledhydeL-lysine-periodate; AA, arachidonic acid; LDH, lactate dehydrogenase; LC, liquid chromatography; MS, mass spectrometer/spectrometry; PAS, periodic acid-Schiff; MPO, myeloperoxidase; Rv, resolvin; RvDs, Rv D series; RvD1, 7S,8,17S-trihydroxy-docosa-4Z,9E,11E,13Z,15E,19Z-hexanoic acid; RvD2, 7S,16,17S-trihydroxydocasa-4Z,8E,10Z,12E,14E,19Z-hexanoic acid; RvD3, 4S,11,17S-treihydroxy-docasa5,7E,9E,13Z,15E,19Z-hexanoic acid; RvD4, 4S,5,17S-trihydroxy-docosa-6E,8E, 10Z,13Z,15E,19Z-hexanoic acid; RvD5, 7S,17S-dihydroxy-docosa-4Z,8E,10Z, 13Z,15E,19Z-hexanoic acid; RvD6, 4S,17S-dihydroxy-docosa 5E,7Z,10Z,13Z, 15E,19Z-hexanoic acid; LDH, lactate dehydrogenase.

Copyright © 2006 by The American Association of Immunologists, Inc.

0022-1767/06/$02.00

2 Address correspondence and reprint requests to Dr. Jeremy S. Duffield, Renal Division, Brigham and Women’s Hospital, Harvard Institutes of Medicine, 5th Floor, 4 Blackfan Circle, Boston MA 02115; E-mail address: [email protected] or Dr. Charles N. Serhan, Center for Experimental Therapeutics and Reperfusion Injury, Brigham and Women’s Hospital, 75 Francis Street, Boston, MA 02115; E-mail address: [email protected] 3

The Journal of Immunology In this study, we document endogenous production of antiinflammatory RvDs and protectin D1 (PD1) in the kidney following ischemic acute kidney injury. We show that the precursor DHA is present in kidneys before and after ischemic acute kidney injury. Increased amounts of Rv and protectins were generated in the postischemic kidney, and administration of these compounds protected kidneys from ischemic injury, reducing leukocyte influx and the postischemic increase in serum creatinine as well as reducing postischemic kidney fibrosis. It is of particular interest that initiation of administration of RvD1 and RvDs after reperfusion also provides a functional protection of the kidney as measured by serum creatinine 24 and 48 h later. Based on these results, it is likely that endogenous anti-inflammatory compounds directed at resolution of the inflammatory response play an important role in the natural course of acute kidney injury (27). Dysregulation or inadequacy of this response may be responsible for delay in recovery or inability to recover from acute kidney injury as seen in many cases in humans. These newly identified DHA-derived lipid mediators may serve as a new paradigm for the design of effective therapeutics to treat patients with acute kidney injury and hasten their recovery improving morbidity and mortality in these patients.

Materials and Methods In vivo experimental protocols Kidney ischemia/reperfusion model. Male 23- to 28-g BALB/c mice (Charles River Laboratories) fed on standard chow were anesthetized with pentobarbital (65 ␮g/g) i.p., shaved on both flanks, and prepared by cleaning skin with betidine. The animals were placed prone on temperaturecontrolled heating pads linked to a rectal probe (Harvard Apparatus). Core mouse temperature was stabilized between 36.7°C and 37.3°C. Mice were given reagents by tail vein injection. When compounds were administered i.v., the interval between injection and clamping of kidney vessels was 10 min. Longitudinal incisions were made over both kidneys in the mid scapular line, the muscle wall was divided by blunt dissection. The left kidney was exposed. A microaneurysm clamp (Roboz) was placed across the renal pedicle to occlude artery and vein. Dusking of the hue of the kidney was confirmed after which the clamped kidney was returned to the retroperitoneum, and the skin held closed. The right kidney was then exposed and the renal pedicle clamped within 1 min of the placement of the left kidney clamp. For both kidneys, the clamps were allowed to remain for precisely 30 min. Clamps were then removed and reperfusion confirmed visually. Kidneys were again returned to the retroperitoneum, and the skin wound closed with clips. Before closure of the right flank wound, a pocket was created by blunt dissection between the fascial planes separating dermis and muscle, pointing cranially toward the cervical end of the thoracic spine. Primed, warmed, preprepared osmotic minipumps were placed in this space with the exit point facing cranially, and the skin wound closed. Mice were transferred to a warmed cage until they had fully recovered from anesthesia and given 1 ml of PBS (37°C) i.p. All mice were given another 1 ml of PBS (37°C) i.p. 24 h later following tail bleeds. One hundred microliters of blood were collected by tail tip incisions and capillary tube withdrawal. Experimental protocols with DHA, Rv, and protectins. Mice (n ⫽ 5/ group (weight matched)) received 200–300 ␮l of DHA (125 ␮g/mouse), PD1 (3.5 or 35 ␮g/mouse), 17S-hydroxy DHA (17S-HDHA) (17.5 ␮g/ mouse), or vehicle (0.9% NaCl saline containing 0.01% delipidated BSA). Half of each solution was injected into the lateral tail vein 10 min before bilateral clamping of renal pedicles. The other half was infused s.c. via Alzet osmotic-pump (Durect) at a flow rate of 8 ␮l/h (model 2001D) for 24 h or 1 ␮l/h (model 1003D) for 72 h. Each treatment group was matched with a vehicle-treated control group. In a second series of experiments, cohorts of mice (n ⫽ 7/group) received PD1 (3.5 ␮g or 35 ␮g/mouse), RvD (3.5 or 35 ␮g/mouse), or vehicle (0.9% NaCl saline containing 0.01% BSA). Half of each solution was injected into the lateral tail vein 10 min before bilateral clamping of renal pedicles. Another half was infused s.c. via Alzet osmotic pump (model 2001D) at 8 ␮l/h until mice were euthanized at 24 h. In a third series of experiments, mice (n ⫽ 5/group) were treated identically to the second series with the exception that the reagents were administered by a 1003D minipump at a flow rate of 1 ␮l/h continuously over 48 h. Tail bleeds were performed at 24 h, and mice were euthanized at 48 h. In a fourth series of experiments, animals were treated as in the second and third except that the infusion via osmotic pump lasted

5903 for 72 h and the mice were allowed to recover from ischemia/reperfusion (I/R) injury for 15 days at which time they were euthanized. A fifth series of experiments was conducted to evaluate whether treatment postreperfusion would have any beneficial effects. Mice (n ⫽ 4/group (weight matched)) received 200 ␮l of RvDs (10 ␮g/mouse), RvD1 (10 ␮g/mouse), PD1 (10 ␮g/mouse), or vehicle. In these studies, half of each solution was injected i.p. 10 min after release of the bilateral clamps on the renal pedicles. The other half was infused s.c. via Alzet osmotic pump (1003D) as above. Sham-operated mice also received compounds administered identically. All studies were conducted in accordance with a protocol approved by the Harvard Center for Animal Resources and Comparative Medicine.

Lipidomic analysis of tissue RvDs and PD1 Blood was extracted with 2 vol of methanol (⬃4°C). Renal tissues were homogenized and extracted in methanol on ice. The tissue and blood extracts were cleaned with C18 solid phase extraction cartridge (300 mg; Alltech) (24, 28). The analyses were conducted using a liquid chromatography-UV-ion trap tandem mass spectrometer (LC-UV-MS/MS) (LCQ; ThermoFinnigan) equipped with a LUNA C18 column (100 mm ⫻ 2 mm ⫻ 5 ␮m) (23, 24, 29) The column was eluted at 0.2 ml/min. The mobile phase gradient was as follows: from 0 to 8 min, phase A (methanol: water:acetic acid ⫽ 65:35:0.01 by volume); from 8.01 to 30 min, linearly changing from A to methanol; and from 30.01 to 35 min, methanol. The photodiode-array UV detector (UV) scanned from 200 to 400 nm. The electrospray voltage for the mass spectrometer LCQ was 4.3 kV. The ion trap analyzer typically scanned from m/z 200 to 800 in MS mode and from m/z 95 to 380 in MS/MS mode.

Characterization of RvDs, PD1, and related products RvDs and 17S-hydroxy-DHA (HDHA) were generated using DHA (Cayman Chemical) and 15-lipoxygenase (24, 30, 31). The enzymatically derived preparations were isolated via C18 solid phase extraction followed by HPLC on a C18 column (150 mm ⫻ 2 mm ⫻ 5 ␮m) (Phenomenex), which was eluted with 70% methanol. RvD and 17S-HDHA were isolated using HPLC, collected, and characterized (24). PD1 was prepared by total organic synthesis and was quantified by both physiochemical and biological properties (32). The solutions were taken to dryness with N2 gas and suspended in ethanol as stock solutions. Immediately before I/R experiments, each stock solution was diluted with saline (0.9% NaCl) or saline containing delipidated endotoxin-free BSA (0.01% BSA) (33). The composition of the RvDs was 1:2:1 (RvD1:RvD2:RvD3), reflecting their relationship in vivo. These results were confirmed with RvD1 prepared by total organic synthesis, matched with enzymatic and biologically generated RvD1 that will be reported elsewhere (Y. P. Sun, J. Uddin, S. F. Oh, K. Gotlinger, E. Campbell, S. P. Colgan, N. A. Petasis, and C. N. Serhan, submitted for publication). The synthetic RvD1 and PD1 were provided by Prof. N. P. Petasis (Department of Chemistry, University of Southern California, Los Angeles, CA). The total organic synthesis of PD1 was reported (34).

Assessment of kidney injury Preparation of kidneys for analysis. Cohorts of mice were euthanized at 24 h, 48 h, or 15 days. Tissues were either flushed with ice-cold PBS to remove erythrocytes and circulating leukocytes or perfusion fixed in situ with paraformaldehyde-L-lysine-periodate (PLP) solution using techniques previously described (10). Unfixed kidneys were snap-frozen in liquid N2 and stored at ⫺80°C. Fixed tissues were transferred from PLP to 18% sucrose solution in PBS after 2 h. Eighteen hours later, they were frozen in OCT (optimal cutting temperature) compound and stored at ⫺80°C. Tissues for staining were fixed in 10% neutral buffered formalin for 12 h, transferred to 70% ethanol and embedded in paraffin wax. Immunofluorescence. Five-micrometer cryotome-cut PLP-fixed sagittal sections were preblocked with Fc-block (BD Pharmingen), then immunolabeled with anti-GR-1 Abs or anti-CD11b Abs (eBioscience) at 1/200 dilution, followed by affinity-purified goat anti-rat Cy3 fluorescent Abs (Jackson ImmunoResearch) and then by anti-CD68-FITC Abs (Serotec) in 10% rabbit serum for 2 h at room temperature. Sections were washed in PBS three times, then mounted with Vectashield including 4⬘,6⬘-diamidino-2-phenylindole (200 ng/ml). Sections were viewed by fluorescence microscopy (⫻200) and serial images captured using identical settings, covering the entire section. All images were assessed quantitatively for percentage area of kidney positive for a particular stain using methods previously described (35). Briefly, digital images were assessed using Fovea Pro software. A range of hues, saturations, and intensities were selected to selectively include the positively stained cells only. These settings were applied to each image, giving a percentage area of the image positive for

5904

RvDs AND PD1 ATTENUATE ISCHEMIC RENAL FAILURE

the stain. For each kidney, the average for the each whole sagittal section was obtained by recording the area for each captured image. Staining of tissue sections. Three-micrometer paraffin sagittal sections were stained with periodic acid-Schiff (PAS) or the Masson’s trichrome method. PAS-stained sections from each kidney were assessed blindly for histological severity of disease using the following established scale (5): 1⫹, normal; 2⫹, mitosis and necrosis of individual cells; 3⫹, necrosis of all cells in adjacent proximal tubules, with survival of surrounding tubules; 4⫹, necrosis confined to the distal third of the proximal tubules with a band of necrosis extending across the inner cortex; 5⫹, necrosis affecting all three segments of the proximal tubules. Each kidney was ascribed a disease severity value. The quantity of blue-stained collagen deposition was assessed in kidneys 15 days after ischemic injury using quantitative morphometry measurements as described above. The quantity of collagen was determined as the percentage area of the total kidney area. Quantitative myeloperoxidase (MPO) assay. Snap-frozen kidney samples were homogenized at 4°C in potassium phosphate buffer (pH 6.0) containing 0.5% hexadecyltrimethylammonium bromide, sonicated, freezethawed three times, then sonicated again. The suspension was centrifuged at 12,000 ⫻ g for 15 min, and 20 ␮l of supernatant was added to 900 ␮l of potassium phosphate buffer (pH 6.0) containing 0.167 mg/ml o-dianisidine dihydrochloride (Sigma-Aldrich) and 0.0006% hydrogen peroxide (Sigma-Aldrich). The rate of change in absorbance at 460 nm was monitored at 23°C at 90-s intervals (36). Creatinine measurements. Plasma was obtained by either tail bleeds as described or cardiac puncture into preheparinized tubes. Plasma was separated by centrifugation (800 ⫻ g for 10 min). Thirty-five-microliter aliquots of plasma in duplicate were used for measurement with a Beckmann II Creatinine analyzer (37). Mouse activity score. Mice were assessed 24 h following I/R for wellbeing using a standardized assay ranging from 1 to 3 (1⫹, lazy, slow movement; 2⫹, intermediate level of activity; 3⫹, active movement or searching) (38). This time reflects the peak of the plasma creatinine.

Results

In vitro assays Tubule cell injury. LLC-PK1, swine kidney epithelial cells with proximal tubule characteristics, were obtained from American Type Culture Collection and passaged in DMEM (Invitrogen Life Technologies) with glutamine and 10% FCS. For the experiments cells were grown to confluence in 24-well plates (Corning) in 1.0 ml of medium containing 1% FCS. Experiments were conducted in triplicate. Cells were treated with 500 ␮M H2O2 and simultaneously one of the active reagents, DHA, PD1, RvDs, or the essential fatty acid, arachidonic acid (AA), in one of the following molar concentrations: 0.5, 50, and 500 ␮M. In previous experiments we had determined that 500 ␮M H2O2 resulted in ⬃50% lactate dehydrogenase (LDH) release 6 h later (39). Thus, after 6 h of incubation, 100 ␮l of supernatant was collected and cells lysed in the remaining medium with 0.1% Triton X-100. The proportion of total cellular LDH released into the supernatant was used as a measure of injury and LDH was quantitated using a standard colorimetric LDH-specific assay (39). Macrophage TNF-␣ production. Primary bone marrow-derived macrophages were cultured as previously described (40). Day 7 mature macrophages were plated into 24-well plates (2.5 ⫻ 105/well). Each well had 500 ␮l of DMEM/F12 containing 10% FCS. Each experiment was performed in triplicate. A total of 100 ng/ml LPS (Escherichia coli 0127:B8; SigmaAldrich) was added to each well together with simultaneous addition of one of the active reagents, DHA, PD1, or RvD in one of the following molar concentrations: 0.05, 0.5, 50, or 500 nM in PBS 0.01% delipidated BSA or an equal volume of vehicle alone. Macrophages were cultured for 24 h, supernatants harvested, and assayed for TNF-␣ generation using a mouse TNF-␣ sandwich ELISA (R&D Systems). The concentration of TNF-␣ was normalized for the total protein in each well of macrophages which was determined by dissolving cells in 100 ␮l of complete lysis buffer (Roche) containing 0.1% Triton X-100. Two microliters was diluted 50fold; the sample was mixed with nine parts protein assay solution (BioRad) and protein was quantitated colorimetrically at 595 nm and compared with an albumin control.

Statistics Values are expressed as mean ⫾ SEM. Differences among groups were assessed by ANOVA and between groups by Students’ t test. Significant differences between groups are denoted: ⴱ, p ⬍ 0.05, ⴱⴱ, p ⬍ 0.01.

I/R acute kidney injury results in biosynthesis and release of D series resolvins and protectins To explore the role of endogenously generated omega-3 DHA, and lipids derived from exogenously introduced omega-3 DHA, on recovery after injury to the kidney, DHA (62.5 ␮g/mouse) or vehicle was administered as a bolus to mice before ischemia and then infused by osmotic pump (dose 62.5 ␮g/mouse) over the succeeding 24 h. The impact of renal I/R in the presence and absence of concurrent administration of DHA on the biosynthesis and release of D series Rv and protectins was studied using LC-UV-MS/MSbased informatics (Fig. 1). In contrast to sham-treated mice, ischemia followed by 24 h reperfusion triggered in kidney tissue the endogenous biosynthesis and/or release of the precursor DHA into the plasma (Fig. 1A). The DHA-derived lipid mediator, PD1, and its biosynthetic intermediate, 17S-HDHA, and to a lesser extent RvD1 and RvD3, were also generated by the postischemic kidney, but not by the sham-operated kidney (Fig. 1, C and F). In plasma, endogenous DHA was markedly elevated by I/R (Fig. 1A) and there was an increase in the intermediate 17S-HDHA. In contrast to kidney tissue, however, there was no increase in plasma levels of PD1 or RvD1 in the absence of added DHA, whereas both RvD2 and RvD4 and to a lesser extent RvD3 levels were increased in plasma (Fig. 1, D and E). Both RvD5 and RvD6 were also identified at low levels in plasma after I/R in the vehicle group (Fig. 1D). Increased levels of 17S-HDHA were found in postischemic kidneys in both the vehicle and DHA-treated groups, suggesting that I/R induced activity of the enzyme 15-lipoxygenase (Fig. 1, A and C), which is instrumental in transforming DHA to 17S-HDHA in vivo in mouse tissues (24, 33). Exogenous administration of DHA to mice exposed to kidney I/R not only increased the amounts of unesterified DHA in both plasma and renal tissue (Fig. 1A), but increased the tissue levels of RvD6 and RvD2. In addition, the level of the precursor 17SHDHA was increased, as was the level of RvD1 and to a lesser extent RvD3. PD1 was notably not increased by DHA infusion (Fig. 1C). Exogenous administration of DHA following ischemic kidney injury led to de novo generation of PD1 and RvD1 in plasma. RvD2 concentration was increased in plasma as were the levels of other Rv including RvD3, RvD5, and RvD6 (Fig. 1C). Of interest, administration of DHA did not directly change RvD4 levels in plasma or in kidney tissue within this interval. The levels of 17S-HDHA were also increased in plasma with administration of DHA (Fig. 1D). The identification of the DHA-derived bioactive mediators described above was conducted using a LC-MS-MS informatics approach generating ion chromatograms from LC-MS-MS (Fig. 1, E and F). The MS/MS spectrum at m/z 375 acquired at a LC retention time of 4.8 min on a representative plasma sample from mice 24 h following I/R is consistent with the structure of RvD2 with diagnostic ions at m/z 109 (129-2H2O-2H), 123 (142-H2O-H), 131 (129 ⫹ 2H), 203 (247-CO2-H), 209 (263-3H2O), 217 (233 ⫹ 2HH2O), 227 (263-2H2O), 233, 241 (276-2H2O⫹H), 247 (246⫹H), 261 (263-2H), 277 (276⫹H), 287 (306-H2O-H), 293 (M-H-2H2OCO2-2H), 313 (M-H-H2O-CO2), 331 (M-H-CO2), 339 (M-H2H2O), 357 (M-H-H2O-CO2), and 375 (M-H) (Fig. 1E) (24). The complete stereochemistry of neuroprotectin D1/PD1 was recently established (41). PD1 was identified in mouse kidney 24 h following ischemia based on the MS/MS spectrum at m/z 359, which possesses diagnostic ions at m/z 359 (M-H), 341 (M-H-H2O), 323 (M-H-2H2O), 315 (M-H-CO2), 297 (M-H-H2O-CO2), 289, and 277 (M-H-2H2O-CO2-2H). Ions consistent with the carbon 10 and carbon 17 alcohol-containing positions were observed at m/z 153,


5905

FIGURE 1. Endogenous DHA, Resolvin D series, and PD1 in murine renal tissues and plasma 24 h after ischemic injury. A, Concentration of DHA in kidney (left panel) and plasma (right panel) from sham-operated mice, mice 24 h following I/R kidney injury (vehicle), and mice 24 h following I/R injury with an infusion of DHA. B, The molecular structures of DHA-derived 17S-HDHA, PD1, RvD1 and RvD2, and RvD3. C, DHA-derived products in kidney from sham-operated mice, mice 24 h post-kidney I/R injury, and mice 24 h post-I/R injury with an infusion of DHA. D, DHA-derived compounds in plasma from sham-operated mice, mice 24 h post-kidney I/R injury, and mice 24 h post-I/R injury with an infusion of DHA. Identification and quantification were conducted via LC-UV-MS-MS-based informatics. Mice received 5 ␮g of DHA/g body weight or an equivalent volume of its vehicle. Details are shown in Materials and Methods. Sham-operated mice served as control. Data represent the mean from two independent experiments. E, MS/MS spectrum at m/z 375 showing the production of RvD2 in mouse plasma 24 h following I/R. F, MS/MS spectrum at m/z 359 showing endogenous generation of PD1 in mouse kidney 24 h after ischemia.

163 (182-H-H2O), 181 (182-H), 205, 217 (216-CO2-H), 243 (261H2O), 245 (261-H2O ⫹ 2H), and 261 (Fig. 1F) (24). Because RvDs and PD1 display activity of anti-inflammatory autacoids in inflammatory settings (32), we next considered whether RvDs and PD1 might function as anti-inflammatory agents in I/R-induced kidney injury. Administration of RvDs and PD1 protects kidneys from ischemic acute kidney injury Because local inflammation is a prominent component in the pathophysiology of acute kidney injury (8) and DHA is the precursor of RvDs and PD1, we evaluated whether DHA was an effective therapy for ischemic kidney injury. Mice were given DHA by i.p. injection immediately before induction of bilateral I/R in-

jury. At reperfusion, DHA therapy was continued over 25 h by s.c. osmotic-pump infusion. At 24 h after reperfusion, levels of plasma creatinine (a quantitative marker of renal failure) were measured (Fig. 2A). Neither creatinine levels nor kidney histology (examined at 24 h, data not shown) were different in the DHA-treated mice compared with the vehicle-treated mice. Furthermore, when histological injury was determined by a semiquantitative injury severity score determined by viewing PAS-stained kidney sections, there was no significant difference between DHA-treated and vehicle-treated sections (4.2 ⫾ 0.3 vs 4.0 ⫾ 0.2). The RvDs display potent anti-inflammatory activity (30, 33, 42). We administered these compounds to mice in amounts related to their endogenous formation (1:2:1, RvD1:RvD2:RvD3; vide supra) 10 min before ischemia of both kidneys and during the

5906

FIGURE 2. Kidneys are partially protected from ischemic injury with ProtectinD1, Resolvin D series, but not the precursor DHA. A, Mice with renal I/R injury were treated with vehicle (䡺), with DHA (125 ␮g/mouse), PD1 (3.5 ␮g/mouse) or the intermediate 17S-HDHA (17.5 ␮g/mouse), and plasma creatinine was measured 24 h following injury. Note that DHA does not protect kidneys from renal injury. B, Mice were treated with PD1 (3.5 ␮g (shaded) or 35 ␮g/mouse (f) (24)), RvDs (3.5 ␮g (shaded) or 35 ␮g/mouse (f) (24)) and plasma creatinine measured 24 h later. Note a dose-dependent increase in protection with both compounds. C, Mice were treated with PD1 (3.5 ␮g (shaded) or 35 ␮g/mouse (f) (24)), RvDs (3.5 ␮g (shaded) or 35 ␮g/mouse (f) (24)), and plasma creatinine was measured 48 h later. Note that at high concentrations both compounds result in more rapid declines in plasma creatinine from 24 to 48 h than vehicle, but at lower doses, RvDs-treated mice show similar plasma creatinine levels at 24 and 48 h after injury, albeit in both cases lower than vehicle-treated animals (ⴱ, p ⬍ 0.05, ⴱⴱ, p ⬍ 0.01 compared with vehicle-treated controls).

RvDs AND PD1 ATTENUATE ISCHEMIC RENAL FAILURE subsequent 48-h interval via a s.c. mini pump. Half of the total dose of RvDs was given by infusion pump (Alzet). At both a total dose of 3.5 ␮g (140 ng/g body weight) or 35 ␮g (1.4 ␮g/g body weight), RvDs limited functional renal injury at 24 h as reflected by a lower plasma creatinine when compared with vehicle-treated animals (Fig. 2B). RvDs had no effect on plasma creatinine levels in sham-operated mice (data not shown). PD1 administered to mice as above also protected kidneys. Both RvDs and PD1 not only attenuated peak creatinine levels at 24 h, but also appeared to have additional actions on the renal resolution process between 24 and 48 h. Plasma creatinine levels returned to near normal at 48 h in PD1-treated mice and in mice treated with 35 ␮g of RvDs, whereas in vehicle-treated mice plasma creatinine levels remained markedly elevated (Fig. 2C). A lower total dose of 3.5 ␮g (140 ng/g) had less efficacy compared with the higher dose of 35 ␮g (1.4 ␮g/g) at 24 h. Furthermore, PAS-stained sections of postischemic kidney showed enhanced tubule cell survival, decreased renal inflammation, and decreased capillary occlusion in mice treated with 35 ␮g of either PD1 or RvDs (Fig. 3, Table I). From a behavioral perspective, mice with acute kidney injury secondary to ischemia exhibited reduced activity during the first 24 h of reperfusion. Mice treated with either RvDs or PD1 exhibited increased activity (Table II) compared with vehicle-treated mice, providing further evidence for the protective actions of these mediators.

RvDs and PD1 reduce leukocyte accumulation Because we and others have strongly implicated leukocytes in the pathophysiology of ischemic kidney disease (5, 9, 12), we used two independent methods to quantify the actions of PD1 and RvDs on leukocyte involvement in kidney injury. Tissue MPO activity in whole kidney was assessed and leukocytes were stained with specific Abs for immunocytochemistry. Kidney MPO activity was decreased in mice treated with DHA, 17S-HDHA, PD1, and RvDs compared with vehicle-treated mice at both 24 and 48 h (Fig. 4). Both exogenous PD1 and RvDs resulted in a ⬎50% reduction in MPO activity 48 h into the recovery from I/R. High-dose RvDs reduced MPO activity by 80% compared with vehicle (Fig. 4B).

FIGURE 3. Histological demonstration of protection from ischemic injury by PD1 or RvDs. Forty-eight hours following ischemic injury, kidneys were sectioned and stained by the PAS method. At higher magnification (⫻20) (upper panels) of the outer medullary region, vehicle-treated kidneys show widespread proximal tubular necrosis, with complete loss of tubular cells in some areas, lumina filled with debris and marked flattening of tubular cells in distal tubules and thick ascending limbs. In addition, many apoptotic cells are seen in the distal tubules, and interstitial cell infiltrate is visible. At lower power (⫻4) (lower panels), the whole outer medullar area shows confluent necrosis as distinguished by pale pink (arrows), and more proximally many tubules show proteinaceous debris (intense pink). By comparison, kidneys from mice treated with both PD1 and RvDs show much milder disease. Many fewer proximal tubules show necrosis (⫻20), many tubules have an intact brush border and those with necrosis show small amounts of intratubular debris alongside intact surviving tubule cells. Distal tubules and ascending limbs are normal in appearance without tubule cell flattening and with fewer apoptotic cells. At low power (⫻4), there is little evidence of necrotic debris or proteinaceous debris.


5907

Table I. Histological injury score for kidneys of mice treated with vehicle, PD1, or RvDsa

Score

Vehicle

PD1

RvDs

4.25 ⫾ 0.17*

3.25 ⫾ 0.29

3.5 ⫾ 2.4

The dose of PD1 was 35 ␮g and RvDs was 35 ␮g. Scores were determined at 48 h postinjury (see Materials and Methods for details). Sham-operated mice had histological injury score of 1.0 irrespective of treatment; *, p ⫽ 0.017 (ANOVA). a

Furthermore, PD1 reduced MPO activity by 67% 24 h after ischemia. It is important to note that, while high-dose DHA reduced MPO activity by 41% (Fig. 4A), it did not prevent the rise in creatinine seen with I/R as discussed above (Fig. 2). The degree of monocyte and neutrophil infiltration was assessed quantitatively by morphometry using Abs to CD11b, which is expressed by both PMNs and monocytes, Abs to GR-1, which is expressed by PMNs, and Abs to CD68, which is expressed by monocytes (Fig. 5). Both PD1 and RvDs treatments resulted in quantitatively less area of CD11b, GR-1, and CD68 immunostaining objectively assessed by morphometry (35). Thus, tissue PMNs and monocytes were reduced by both PD1 and RvDs (Fig. 5). Administration of RvDs or RvD1 after onset of ischemic injury is protective Our novel compounds demonstrated efficacy in protecting kidneys from severe injury and promoting resolution. In the above experiments, however, the injured kidneys were exposed to the compounds prior and subsequent to onset of injury. To determine whether Rv and protectins retained efficacy when reaching the kidney after the onset of injury, further studies were designed. In these, Rv and protectins were administered after ischemic injury. Furthermore, the compounds were given i.p. rather than i.v. (followed by s.c. infusion) to slow the rate at which they might gain access to the kidneys. Kidneys treated with RvDs or synthesized RvD1 exhibited marked protection from development of acute renal failure (Fig. 6). PD1 given after ischemic injury did not significantly alter the course of acute renal failure compared with vehicle-treated mice. RvDs and PD1 limit leukocyte activation but do not directly protect proximal tubule cells from oxidative agents To study the impact of these novel renoprotective DHA-derived mediators in more detail, we addressed their actions in vitro. The generation of oxygen free radicals and hydrogen peroxide are implicated in the pathogenesis of both ischemic and toxic acute kidney injury (43, 44). To model tubular cell injury observed in vivo, we tested whether RvDs and PD1 protected cultured proximal tubular epithelial cells from H2O2 injury using an established assay for cell injury (39). The effects of AA and DHA were examined in parallel. None of the anti-inflammatory compounds (RvDs, PD1, DHA, or AA) at concentrations as high as 500 nM, had any effect on the H2O2-induced release of LDH from cultured tubular cells (Fig. 7A). In each case, despite the presence of relatively high

Table II. Mouse activity score following ischemia in mice treated with PD1 or RvDsa

Score

Vehicle

PD1

RvDs

2.1 ⫾ 0.2*

2.8 ⫾ 0.3

2.7 ⫾ 0.4

a The dose of PD1 was 35 ␮g and RvDs was 35 ␮g. Activity scores were determined at 24 h. Sham-operated mice had activity scores of 3.0 irrespective of treatment; ⴱ, p ⫽ 0.03 (ANOVA)

FIGURE 4. RvDs and PD1 reduce MPO activity in whole kidney 24 and 48 h after ischemic injury. A, Percentage of MPO activity, 24 h following I/R, in kidneys treated with 17S-HDHA (17.5 ␮g/mouse), PD1 or parent compound DHA, compared with vehicle-treated mice undergoing I/R (䡺). B, Relative percentage of MPO activity in kidneys 48 h following I/R treated with PD1 (3.5 ␮g/mouse (shaded) or 35 ␮g/mouse (f) (24)) or RvDs (3.5 ␮g/mouse (shaded) or 35 ␮g/mouse (f) (24)). All values represent mean ⫾ SEM from three or four different mice. When MPO levels for RvDs, PD1, or DHA treatments were compared with vehicle-treated disease control there was a significant reduction (ⴱⴱ, p ⬍ 0.01).

concentrations of RvDs and PD1, ⬃25% of total LDH was released from cells by H2O2 exposure. By contrast, in cells not treated with H2O2, 7.3 ⫾ 1.9% of total LDH was released in control conditions. LDH release from cells not treated with H2O2 was not affected by these compounds (data not shown). These results suggest that the actions of these mediators in the kidney were not the direct result of tubule cell protection from oxidative influences. In addition to H2O2, the monocyte/macrophage proinflammatory cytokine TNF-␣ has also been implicated in the pathogenesis of ischemic acute kidney injury (45, 46). We prepared cultured macrophages derived from bone marrow and tested whether DHA, RvDs, or PD1, administered simultaneously with a proinflammatory stimulus LPS, limited TNF-␣ generation in response to cell activation (Fig. 7B). At low concentrations, DHA itself had no suppressive effect on TNF-␣ release in macrophages by LPS. At higher concentrations, DHA reduced TNF-␣ release by 22.0 ⫾ 1.6% (Fig. 7B). By contrast, RvD1 afforded a dose-dependent reduction in LPS-induced TNF-␣ release, and had notable actions on TNF-␣ release at concentrations 100 times lower than those required with its precursor, DHA. PD1 also lowered LPS-induced TNF-␣ release, although this was less marked than RvD1, suggesting that both active compounds have anti-inflammatory activities directed at monocytes/macrophages. RvDs and PD1 limit postischemic interstitial kidney fibrosis Even though there is recovery of the kidney functionally following renal pedicle clamping and reperfusion, as measured by serum creatinine, histology reveals incomplete resolution with progressive interstitial fibrosis over weeks (10). This fibrosis may result

5908

FIGURE 5. Both neutrophils and monocytes are reduced by PD1 and RvDs in the kidney following ischemic injury. Kidney sections 48 h following I/R injury were immunolabeled with anti-CD11b (myeloid leukocytes), anti-CD68 (macrophages), or anti-GR1 Abs (neutrophils). The entire outer medullary region of sagittal sections was photographed at ⫻200 in serial images and the percentage area occupied by red or green color was measured by computerized morphometry. Compared with vehicle-treated kidneys, both PD1- and Resolvin D-treated kidneys showed many fewer myeloid leukocytes at 48 h following injury, as detected by CD11b (upper panels (⫻20) and upper left graph). To determine whether this was due to a reduction in recruitment of neutrophils or monocytes, sections were labeled for monocytes/macrophages (green) detecting CD68 and neutrophils (red) detecting GR-1 (lower two immunocytochemistry panels). At 48 h, both types of leukocyte were present in increased numbers in the vehicletreated kidneys, and anti-CD68 Ab-labeled interstitial cells predominated slightly over anti-GR-1-labeled cells. When sections were assessed by morphometry, the area of neutrophils, stained with anti-GR-1 Abs, was markedly reduced in kidneys treated with both compounds (lower graph) as was the area of monocytes stained with anti-CD68 Abs (upper graph). Thus, both neutrophil and monocyte recruitment is reduced by both RvDs and PD1. ⴱⴱ, p ⬍ 0.01, ⴱ, p ⬍ 0.05 compared with vehicle-treated controls.

from persistence of inflammatory leukocytes, in particular interstitial macrophages (47). Interstitial fibrosis and persistent leukocyte infiltration (chronic inflammation) are the harbingers of scarring and chronic renal failure. As many as 40% of patients with acute kidney injury are left with worsening of their baseline kidney status and chronic disease after recovery from the acute injury (3, 48). We assessed whether RvDs and PD1 could limit the deposition of interstitial collagens that contribute to fibrosis. Kidney sections stained for collagen by Gomori’s trichrome method were assessed by quantitative computerized morphometry to assess the


FIGURE 6. D series resolvins and synthesized Resolvin D1, administered after the onset of ischemic injury partially protect kidneys from ischemic injury. A, Mice with renal I/R injury were treated with vehicle, PD1 (10.0 ␮g), RvDs (10.0 ␮g), or RvD1 (10.0 ␮g) 10 min after both clamps were removed from kidneys and plasma creatinine were measured 24 h later. B, Mice with renal I/R injury were treated with vehicle, PD1 (10.0 ␮g), RvDs (10.0 ␮g), or RvD1 (10.0 ␮g) 10 min after both clamps were removed from kidneys, and plasma creatinine was measured 48 h later. Note at this dose RvDs-treated mice experienced lower peak creatinine at 24 h compared with vehicle-treated mice and synthesized RvD1-treated mice had more marked protection at both 24 and 48 h. PD1 treated mice at this dose showed no evidence of significant protection at 24 or 48 h (ⴱ, p ⬍ 0.05 compared with vehicle-treated controls).

area of collagen deposition 15 days after I/R. RvDs treatment for 72 h after ischemia resulted in 44 ⫾ 17% less deposition of collagen, whereas PD1 was less effective (a nonsignificant reduction of 21 ⫹ 12% in reducing scarring at 15 days (Fig. 8)).

Discussion We demonstrate for the first time that two families of newly identified anti-inflammatory mediators, D series Rv and PD1, are generated in the kidney in response to I/R. These mediators are generated by stereospecific enzymatic pathways both in the presence and absence of exogenous administration of precursor DHA. When infused peripherally in mice, they markedly attenuate ischemic kidney injury as well as reduce fibrosis. One of the actions of these compounds in the kidney is to limit the influx of leukocytes and another is to limit activation of leukocytes. Therefore, the pathways described are likely to counterregulate the proinflammatory, cellular and molecular signaling that results from I/R. This represents a new paradigm in the pathophysiology of acute kidney injury, which often leads to acute renal failure. Hence, these antiinflammatory and counterregulatory influences may be critical to the resolution of damage and limitation of chronic fibrosis resulting from this damage. More PD1 than RvD is produced with I/R, especially in the absence of exogenously administered DHA. RvDs and PD1 displayed profound protective effects within 24 h of I/R (PMN dominated) that persisted for 48 h (monocyte dominated). RvDs were apparently more effective than PD1 in reducing scarring 15 days


FIGURE 7. D series Rv and PD1 do not protect tubular epithelial cells from toxic injury, but can limit macrophage activation. A, Confluent wells of cultured LLC-PK1 (swine proximal tubule cells) treated with H2O2 (500 ␮M) for 6 h released similar amounts of LDH into supernatants regardless of prior administration of docosanoids or arachidonic acid (AA) at concentrations ranging from 0.5 to 500 nM. Baseline proportionate release of LDH was 0.073 ⫾ 0.019. B, Cultured bone marrow-derived macrophages were activated with LPS, 100 ng/ml, for 10 min followed by administration of docosanoids in doses ranging from 0.05 to 50 nM, for 24 h. Each lipid (DHA, RvD1, and PD1) gave significant reduction in TNF-␣ release into the medium when applied at a concentration of 50 nM (ⴱ, p ⬍ 0.001). However, only RvD1 and PD1 showed a dose-dependent reduction in TNF-␣ release, suggesting that only at very high concentrations is DHA able to stop macrophage activation. Control LPS-activated macrophage cultures (no docosanoids) generated TNF-␣ at a concentration of 1762 pg/ml.

after ischemia and were much more effective at reducing injury when initially administered after reperfusion. Experiments with epithelial cells in vitro suggest that protection in vivo is not the result of a direct effect on the tubular epithelial cells of the kidney but more likely related to the ability of these mediators to downregulate components of the inflammatory response involving infiltrating cells. We have now shown that endogenously generated RvDs and PD1 protect from ischemic injury in the kidney. PD1, also known as NPD1 when generated in neuronal tissues (41, 49), is protective in the brain (33). There is currently a need for effective therapies for diseases in humans resulting from ischemia to the kidney as well as brain, both of which are characterized by uncontrolled local inflammation, which is believed to contribute to acute and chronic functional impairment (2, 33). RvDs and PD1 both limit infiltration of leukocytes and also limit activation of leukocytes in both postischemic organs. It is possible that both compounds may have additional cellular sites of actions in the kidney, i.e., on the endothelium and vascular tone, as well as interstitial fibroblasts because they are also antifibrotic.

5909

FIGURE 8. D series Rv and PD1 limit interstitial fibrosis that persists following recovery from I/R kidney injury. A, Fifteen days following I/R injury, with and without administration of compounds, kidneys were assessed for interstitial fibrosis in the outer medulla in 3-␮m sections stained with Gomori’s trichrome. Note turquoise-stained material in the interstitium corresponding to fibrosis, seen here as pale-stained material between deeply stained tubules. B, The entire outer medulla of each kidney was imaged sequentially and the area of kidney sections occupied by turquoisestained material was assessed objectively and blinded analyses were performed using computerized morphometry. ⴱ, p ⬍ 0.05 compared with vehicle.

In an experimental model of stroke, there was active generation of both PD1 and RsD1, the former peaking 10 h after injury, and the latter peaking 24 h after injury (33). Infusion of PD1 into ventricles of the brain following stroke markedly attenuated stroke area as measured 48 h after ischemia (33). RvDs and PD1 in sterile peritonitis are generated during the resolution phase, and administration of both RvDs and PD1 not only limited neutrophil influx but also limited both chemokines and proinflammatory cytokines in the inflammatory exudates (26). It is notable that we found that the endogenous level of the biosynthetic precursor of RvDs and PD1, DHA, was increased in response to injury. This likely reflects an activation of cellular lipases (i.e., PLA2) enzymes, which can cleave DHA from phospholipids. Because increased levels of DHA in tissue and plasma alone does not account for the increased generation of D series Rv and PD1 found with I/R, renal ischemic injury may up-regulate the local formation of RvDs and PD1 (24, 33). Endogenously generated anti-inflammatory mediators with or without exogenous DHA administration may be present in insufficient quantities in normal mice during acute kidney injury to provide adequate protection of the organ from innate immune-mediated injury. Our studies also demonstrate that RvDs and synthesized RvD1 are effective in attenuating renal injury when administered after the insult. These findings indicate that these “proresolution-of-inflammation” compounds may not simply block the activation of the inflammatory response pathways but that they are bioactive during the acute injury phase and can actively counteract inflammation

5910


and injury. The fact that equimolar administration of RvD1 appears more potent in protecting kidneys from injury than the mixture of RvDs in the postinjury administration of compounds suggests that RvD1 might be more potent than RvD2 and RvD3. Further studies will be required to determine whether RvD1 is more potent at blocking neutrophil activation or chemotaxis. Our studies suggest that PD1 in the kidney is not efficacious in promoting resolution once the injury is established. This may reflect the reduced efficacy of PD1 after established disease onset, and point to a distinct mechanism of action from the RvDs series, or possibly reflects inadequate bioavailability. In the studies in which PD1 is effective, doses of both 3.5 and 35 ␮g were effective, whereas in the postinjury studies, a single dose of 10 ␮g was chosen due to limitation of availability of the compound. It is possible that after injury onset higher doses of PD1 are required to achieve therapeutic levels in the injured kidneys. Further studies will be required. Earlier studies focused on the actions of RvDs and PD1 on PMN activity and chemotaxis (24, 27, 33, 41). In the present study, we also implicate these compounds in limiting macrophage activity showing, for the first time, that LPS-induced activation of macrophages is reduced by PD1 and to a lesser extent by RvDs even when administered at the same time as LPS-induced activation, without the necessity for pretreatment. Because LPS acts through TLR4 and to a lesser extent TLR2 it is likely that these mediators impinge on TLR-activation pathways (50, 51). Further studies are required to determine whether blockage of macrophage activation has effects on many pathways or is limited to TLR-mediated pathways. In summary, we demonstrated activation of newly described DHA-biosynthetic pathways in the kidney resulting in the local appearance of anti-inflammatory mediators following ischemic injury. Administration of D series Rv and PD1 sharply reduced renal injury in a model of acute kidney injury. Thus, a previously unrecognized endogenous anti-inflammatory response to injury may play an essential role in resolution of acute kidney injury. Moreover, we uncovered an antifibrotic action of these mediators and found that they also act on macrophages. Dysregulation of these response(s) may underlie a delay in recovery or inability to recover in many cases of acute renal failure in humans. Administration of mimetics of these endogenously produced biotemplates retaining anti-inflammatory and antifibrotic activities may be of therapeutic importance in treating acute kidney injury.

6. Heyman, S. N., W. Lieberthal, P. Rogiers, and J. V. Bonventre. 2002. Animal models of acute tubular necrosis. Curr. Opin. Crit. Care 8: 526 –534. 7. de Vries, B., J. Kohl, W. K. Leclercq, T. G. Wolfs, A. A. van Bijnen, P. Heeringa, and W. A. Buurman. 2003. Complement factor C5a mediates renal ischemiareperfusion injury independent from neutrophils. J. Immunol. 170: 3883–3889. 8. Bonventre, J. V., and A. Zuk. 2004. Ischemic acute renal failure: an inflammatory disease? Kidney Int. 66: 480 – 485. 9. Rabb, H., F. Daniels, M. O’Donnell, M. Haq, S. R. Saba, W. Keane, and W. W. Tang. 2000. Pathophysiological role of T lymphocytes in renal ischemiareperfusion injury in mice. Am. J. Physiol. 279: F525–F531. 10. Park, K. M., J. Y. Byun, C. Kramers, J. I. Kim, P. L. Huang, and J. V. Bonventre. 2003. Inducible nitric-oxide synthase is an important contributor to prolonged protective effects of ischemic preconditioning in the mouse kidney. J. Biol. Chem. 278: 27256 –27266. 11. Hirschberg, R. 2005. Wound healing in the kidney: complex interactions in renal interstitial fibrogenesis. J. Am. Soc. Nephrol. 16: 9 –11. 12. Basile, D. P., D. Donohoe, K. Roethe, and J. L. Osborn. 2001. Renal ischemic injury results in permanent damage to peritubular capillaries and influences longterm function. Am. J. Physiol. 281: F887–F899. 13. Mazzali, M., J. A. Jefferson, Z. Ni, N. D. Vaziri, and R. J. Johnson. 2003. Microvascular and tubulointerstitial injury associated with chronic hypoxia-induced hypertension. Kidney Int. 63: 2088 –2093. 14. Kang, D. H., J. Kanellis, C. Hugo, L. Truong, S. Anderson, D. Kerjaschki, G. F. Schreiner, and R. J. Johnson. 2002. Role of the microvascular endothelium in progressive renal disease. J. Am. Soc. Nephrol. 13: 806 – 816. 15. Eddy, A. A. 1994. Experimental insights into the tubulointerstitial disease accompanying primary glomerular lesions. J. Am. Soc. Nephrol. 5: 1273–1287. 16. Burne, M. J., F. Daniels, A. El Ghandour, S. Mauiyyedi, R. B. Colvin, M. P. O’Donnell, and H. Rabb. 2001. Identification of the CD4⫹ T cell as a major pathogenic factor in ischemic acute renal failure. J. Clin. Invest. 108: 1283–1290. 17. Clish, C. B., J. A. O’Brien, K. Gronert, G. L. Stahl, N. A. Petasis, and C. N. Serhan. 1999. Local and systemic delivery of a stable aspirin-triggered lipoxin prevents neutrophil recruitment in vivo. Proc. Natl. Acad. Sci. USA 96: 8247– 8252. 18. Cailhier, J. F., M. Partolina, S. Vuthoori, S. Wu, K. Ko, S. Watson, J. Savill, J. Hughes, and R. A. Lang. 2005. Conditional macrophage ablation demonstrates that resident macrophages initiate acute peritoneal inflammation. J. Immunol. 174: 2336 –2342. 19. Lovis, C., F. Mach, Y. R. Donati, J. V. Bonventre, and B. S. Polla. 1994. Heat shock proteins and the kidney. Renal Fail. 16: 179 –192. 20. Maestri, M., D. C. Dafoe, G. A. Adams, A. Gaspari, F. Luzzana, F. Innocente, J. Rademacher, P. Dionigi, A. Barbieri, F. Zonta, et al. 1997. Insulin-like growth factor-I ameliorates delayed kidney graft function and the acute nephrotoxic effects of cyclosporine. Transplantation 64: 185–190. 21. de Vries, B., R. A. Matthijsen, A. A. van Bijnen, T. G. Wolfs, and W. A. Buurman. 2003. Lysophosphatidic acid prevents renal ischemia-reperfusion injury by inhibition of apoptosis and complement activation. Am. J. Pathol. 163: 47–56. 22. Kourembanas, S., L. P. McQuillan, G. K. Leung, and D. V. Faller. 1993. Nitric oxide regulates the expression of vasoconstrictors and growth factors by vascular endothelium under both normoxia and hypoxia. J. Clin. Invest. 92: 99 –104. 23. Serhan, C. N., S. Hong, K. Gronert, S. P. Colgan, P. R. Devchand, G. Mirick, and R. L. Moussignac. 2002. Resolvins: a family of bioactive products of omega-3 fatty acid transformation circuits initiated by aspirin treatment that counter proinflammation signals. J. Exp. Med. 196: 1025–1037. 24. Hong, S., K. Gronert, P. R. Devchand, R. L. Moussignac, and C. N. Serhan. 2003. Novel docosatrienes and 17S-resolvins generated from docosahexaenoic acid in murine brain, human blood, and glial cells: autacoids in anti-inflammation. J. Biol. Chem. 278: 14677–14687. 25. Serhan, C. N., K. Gotlinger, S. Hong, and M. Arita. 2004. Resolvins, docosatrienes, and neuroprotectins, novel omega-3-derived mediators, and their aspirintriggered endogenous epimers: an overview of their protective roles in catabasis. Prostaglandins Other Lipid Mediat. 73: 155–172. 26. Bannenberg, G. L., N. Chiang, A. Ariel, M. Arita, E. Tjonahen, K. H. Gotlinger, S. Hong, and C. N. Serhan. 2005. Molecular circuits of resolution: formation and actions of resolvins and protectins. J. Immunol. 174: 4345– 4355. 27. Ariel, A., P. L. Li, W. Wang, W. X. Tang, G. Fredman, S. Hong, K. H. Gotlinger, and C. N. Serhan. 2005. The docosatriene protectin D1 is produced by TH2 skewing and promotes human T cell apoptosis via lipid raft clustering. J. Biol. Chem. 280: 43079 – 43086. 28. Gronert, K., C. B. Clish, M. Romano, and C. N. Serhan. 1999. Transcellular regulation of eicosanoid biosynthesis. In Eicosanoid Protocols. E. A. Lianos, ed. Humana Press, Totowa, NJ, pp. 119 –144. 29. Lu, Y., S. Hong, E. Tjonahen, and C. N. Serhan. 2005. Mediator-lipidomics: databases and search algorithms for PUFA-derived mediators. J. Lipid Res. 46: 790 – 802. 30. Serhan, C. N., C. B. Clish, J. Brannon, S. P. Colgan, N. Chiang, and K. Gronert. 2000. Novel functional sets of lipid-derived mediators with antiinflammatory actions generated from omega-3 fatty acids via cyclooxygenase 2-nonsteroidal antiinflammatory drugs and transcellular processing. J. Exp. Med. 192: 1197–1204. 31. Bazan, N. G. 2003. Synaptic lipid signaling: significance of polyunsaturated fatty acids and platelet-activating factor. J. Lipid Res. 44: 2221–2233. 32. Serhan, C. N. 2005. Novel omega-3-derived local mediators in anti-inflammation and resolution. Pharmacol. Ther. 105: 7–21. 33. Marcheselli, V. L., S. Hong, W. J. Lukiw, X. H. Tian, K. Gronert, A. Musto, M. Hardy, J. M. Gimenez, N. Chiang, C. N. Serhan, and N. G. Bazan. 2003.

Acknowledgments We thank Mary Halm Small for skillful assistance in manuscript preparation, K. H. Gotlinger, Eileen O’Leary, and Xiao Ming Sun for expert technical support.

Disclosures The authors have no financial conflict of interest.

References 1. Bonventre, J. V., and J. M. Weinberg. 2003. Recent advances in the pathophysiology of ischemic acute renal failure. J. Am. Soc. Nephrol. 14: 2199 –2210. 2. Thadhani, R., M. Pascual, and J. V. Bonventre. 1996. Acute renal failure. N. Engl. J. Med. 334: 1448 –1460. 3. Duffield, J. S., and J. V. Bonventre. 2004. Acute renal failure from Bench to Bedside. In Chronic Kidney Disease, Dialysis and Transplant 42, 2nd Ed. B. J. G. Pereira, M. H. Sayegh and P. Blake, eds. Elsevier Saunders, Philadelphia, pp 765–786. 4. Liano, F., and J. Pascual. 1996. Epidemiology of acute renal failure: a prospective, multicenter, community-based study. Madrid Acute Renal Failure Study Group. Kidney Int. 50: 811– 818. 5. Kelly, K. J., W. W. Williams, Jr., R. B. Colvin, S. M. Meehan, T. A. Springer, J. C. Gutierrez-Ramos, and J. V. Bonventre. 1996. Intercellular adhesion molecule-1-deficient mice are protected against ischemic renal injury. J. Clin. Invest. 97: 1056 –1063.


34. 35.

36.

37.

38.

39.

40.

41.

42.

Novel docosanoids inhibit brain ischemia-reperfusion-mediated leukocyte infiltration and pro-inflammatory gene expression. J. Biol. Chem. 278: 43807– 43817. Yang, R. 2005. Synthetic study of lipid mediators. In Department of Chemistry. University of Southern California, Los Angeles. Duffield, J. S., S. J. Forbes, C. Constandinou, M. Clay, M. Partilona, S. Vuthoori, S. Wu, R. L. Lang, and J. P. Iredale. 2005. Depletion of macrophages reveals distinct and opposing functions during liver injury and repair. J. Clin. Invest. 115: 56 – 65. Bradley, P. P., D. A. Priebat, R. D. Christensen, and G. Rothstein. 1982. Measurement of cutaneous inflammation: estimation of neutrophil content with an enzyme marker. J. Invest. Dermatol. 78: 206 –209. Duffield, J. S., K. M. Park, L.-L. Hsiao, V. Kelley, D. T. Scadden, T. Ichimura, and J. V. Bonventre. 2005. Restoration of tubular epithelial cells in the postischemic kidney occurs independently of bone marrow-derived stem cells. J. Clin. Invest. 115: 1743–1755. Castellani, S., and P. M. Adams. 1981. Acute and chronic phencyclidine effects on locomotor activity, stereotypy and ataxia in rats. Eur. J. Pharmacol. 73: 143–154. Hung, C. C., T. Ichimura, J. L. Stevens, and J. V. Bonventre. 2003. Protection of renal epithelial cells against oxidative injury by endoplasmic reticulum stress preconditioning is mediated by ERK1/2 activation. J. Biol. Chem. 278: 29317–29326. Duffield, J. S., L. P. Erwig, X. Wei, F. Y. Liew, A. J. Rees, and J. S. Savill. 2000. Activated macrophages direct apoptosis and suppress mitosis of mesangial cells. J. Immunol. 164: 2110 –2119. Serhan, C. N., K. Gotlinger, S. Hong, Y. Lu, J. Siegelman, T. Baer, R. Yang, S. P. Colgan, and N. A. Petasis. 2006. Anti-inflammatory actions of neuroprotectin D1/Protectin D1 and its natural stereoisomers: assignments of dihydroxycontaining docosatrienes. J. Immunol. 176: 1848 –1859. Bannenberg, G., R. L. Moussignac, K. Gronert, P. R. Devchand, B. A. Schmidt, W. J. Guilford, J. G. Bauman, B. Subramanyam, H. D. Perez, J. F. Parkinson, and

5911

43.

44. 45.

46.

47.

48.

49.

50. 51.

C. N. Serhan. 2004. Lipoxins and novel 15-epi-lipoxin analogs display potent anti-inflammatory actions after oral administration. Br. J. Pharmacol. 143: 43–52. Doi, K., Y. Suzuki, A. Nakao, T. Fujita, and E. Noiri. 2004. Radical scavenger edaravone developed for clinical use ameliorates ischemia/reperfusion injury in rat kidney. Kidney Int. 65: 1714 –1723. McCord, J. M. 1985. Oxygen-derived free radicals in postischemic tissue injury. N. Engl. J. Med. 312: 159 –163. Daemen, M. A., M. W. van de Ven, E. Heineman, and W. A. Buurman. 1999. Involvement of endogenous interleukin-10 and tumor necrosis factor-␣ in renal ischemia-reperfusion injury. Transplantation 67: 792– 800. Kelly, K. J., W. W. Williams, Jr., R. B. Colvin, and J. V. Bonventre. 1994. Antibody to intercellular adhesion molecule 1 protects the kidney against ischemic injury. Proc. Natl. Acad. Sci. USA 91: 812– 816. Duffield, J. S., P. S. Tipping, M. Clay, J. F. Cailhier, T. Kipari, J. V. Bonventre, R. L. Lang, and J. H. Hughes. 2005. Conditional ablation of macrophages halts progression of crescentic glomerulonephritis. Am. J. Pathol. 167: 1207–1219. Morgera, S., A. K. Kraft, G. Siebert, F. C. Luft, and H. H. Neumayer. 2002. Long-term outcomes in acute renal failure patients treated with continuous renal replacement therapies. Am. J. Kidney Dis. 40: 275–279. Mukherjee, P. K., V. L. Marcheselli, C. N. Serhan, and N. G. Bazan. 2004. Neuroprotectin D1: a docosahexaenoic acid-derived docosatriene protects human retinal pigment epithelial cells from oxidative stress. Proc. Natl. Acad. Sci. USA 101: 8491– 8496. Kawai, T., O. Adachi, T. Ogawa, K. Takeda, and S. Akira. 1999. Unresponsiveness of MyD88-deficient mice to endotoxin. Immunity 11: 115–122. Tapping, R. I., S. Akashi, K. Miyake, P. J. Godowski, and P. S. Tobias. 2000. Toll-like receptor 4, but not toll-like receptor 2, is a signaling receptor for Escherichia and Salmonella lipopolysaccharides. J. Immunol. 165: 5780 –5787.