Kidney International, Vol. 64 (2003), pp. 480–492

CELL BIOLOGY–IMMUNOLOGY–PATHOLOGY

Modification of the transcriptomic response to renal ischemia/ reperfusion injury by lipoxin analog

NIAMH E. KIERAN,1 PETER P. DORAN,1 SUSAN B. CONNOLLY,MARIE-CLAIRE GREENAN, DEBRA F. HIGGINS,MARTIN LEONARD,CATHERINE GODSON,CORMAC T. TAYLOR, ANNA HENGER,MATTHIAS KRETZLER,MELISSA J. BURNE,HAMID RABB, and HUGH R. BRADY

Human Genomics and Bioinformatics Research Unit, Department of Medicine and Therapeutics, Conway Institute of Biomolecular and Biomedical Research, University College Dublin, Mater Misericordiae Hospital, Dublin 7 and the Dublin Molecular Medicine Centre, Ireland; Medizinische Poliklinik, Ludwig-Maximilians-Universitat, Munchen, Germany; and Nephrology Division, Johns Hopkins University Hospital, Baltimore, MD, USA

Modification of the transcriptomic response to renal ischemia/ (e.g., aquaporin-1) and the zinc metalloendopeptidase meprin- reperfusion injury by lipoxin analog. 1␤ implicated in renal remodeling. Background. Lipoxins are lipoxygenase-derived eicosanoids Conclusion. Treatment with the lipoxin analog 15-epi-16- with anti-inflammatory and proresolution bioactivities in vitro (FPhO)-LXA4-Me prior to injury modified the expression of and in vivo. We have previously demonstrated that the stable many differentially expressed pathogenic mediators, including synthetic LXA4 analog 15-epi-16-(FPhO)-LXA4-Me is reno- cytokines, growth factors, adhesion molecules, and proteases, protective in murine renal ischemia/reperfusion injury, as suggesting a renoprotective action at the core of the pathophys- gauged by lower serum creatinine, attenuated leukocyte infil- iology of acute renal failure (ARF). Importantly, this lipoxin- tration, and reduced morphologic tubule injury. modulated transcriptomic response included many genes ex- Methods. We employed complementary oligonucleotide mi- pressed by renal parenchymal cells and was not merely a reflection croarray and bioinformatic analyses to probe the transcripto- of a reduced renal mRNA load resulting from attenuated leu- mic events that underpin lipoxin renoprotection in this setting. kocyte recruitment. The data presented herein suggest a frame- Results. Microarray-based analysis identified three broad work for understanding drivers of kidney injury in ischemia/ categories of genes whose mRNA levels are altered in response reperfusion and the molecular basis for renoprotection by li- to ischemia/reperfusion injury, including known genes previously poxins in this setting. implicated in the pathogenesis of ischemia/reperfusion injury [e.g., intercellular adhesion molecule-1 (ICAM-1), p21, KIM-1], known genes not previously associated with ischemia/reper- Ischemic acute renal failure (ARF) remains a formida- fusion injury, and cDNAs representing yet uncharacterized genes. Characterization of expressed sequence tags (ESTs) dis- ble clinical problem that is associated with high morbid- played on microarrays represents a major challenge in studies ity and mortality [1, 2]. The pathophysiology of ARF is of global gene expression. A bioinformatic annotation pipeline complex and multipronged and includes persistent intra- successfully annotated a large proportion of ESTs modulated renal vasoconstriction, hypoxic injury to microvascular during ischemia/reperfusion injury. The differential expression endothelial cells and tubule epithelial cells, and leuko- of a representative group of these ischemia/reperfusion injury– modulated genes was confirmed by real-time polymerase chain cyte-mediated cytotoxicity [1–3]. Despite the impressive reaction. Prominent among the up-regulated genes were clau- efficacy of agents that target these events in experimental din-1, -3, and -7, and ADAM8. Interestingly, the former re- models, none has proved as effective as monotherapy in sponse was claudin-specific and was not observed with other randomized, controlled clinical trials [1, 2]. Accordingly, claudins expressed by the kidney (e.g., claudin-8 and -6) or attention has shifted toward therapeutic interventions indeed with other components of the renal tight junctions (e.g., occludin and junctional adhesion molecule). Noteworthy among that simultaneously target two or more of the aforemen- the down-regulated genes was a cluster of transport proteins tioned pathophysiologic events. Lipoxins are lipoxygenase-derived arachidonate me- tabolites that are formed predominantly through cell- 1 Dr. Kieran and Dr. Doran contributed equally to this work. cell interactions in many human and experimental in- Key words: lipoxins, acute renal failure, ischemia, microarrays, gene flammatory, hypersensitivity, and vascular diseases [4, 6]. chips, bioinformatics, claudins, meprin, ADAM8. The spectrum of bioactivities reported for lipoxins in Received for publication October 14, 2002 vitro and in vivo suggests that these lipid mediators may and in revised form January 14, 2003 be protective in renal ischemia/reperfusion injury. Lipox- Accepted for publication March 21, 2003 ins are potent intrarenal vasodilators that inhibit poly-  2003 by the International Society of Nephrology morphonuclear cells (PMNs) chemotaxis, adhesion and

480 Kieran and Doran et al: Transcriptomic response to renal ischemia-reperfusion injury 481 migration across the endothelium and epithelium, pro- METHODS mote clearance of apoptotic PMNs, and modulate several Experimental ischemia/reperfusion injury cytokine responses [4–13]. Cyclooxygenase-2 (COX-2), ARF was induced in National Institutes of Health when acetylated by aspirin, catalyzes the generation of (NIH) Swiss mice (25 to 35 g) by clamping both renal 15R-HETE from arachadonic acid [4]. 15R-HETE is pedicles for 30 minutes, followed by 24 hours of reperfu- then converted by leukocytes to 15R epimers of lipoxins sion [21]. Animals received a 15 ␮g single bolus injection

[aspirin-triggered lipoxins (ATLs)] during cell interac- of 15-epi-16-(FPhO)-LXA4-Me or an equivalent volume tions [3]. ATLs share many of the anti-inflammatory and of its vehicle into the inferior vena cava 10 minutes prior proresolution properties of native lipoxins in vitro [4]. to clamping. Sham-operated animals served as controls.

Analogs of LXA4 and LXB4 (the major mammalian li- The influence of the lipoxin analog on changes in renal poxin) and of ATLs have been synthesized that are rela- function, morphology, and PMN infiltration in these tively resistant to degradation [4]. mice has been previously reported [21]. As a second Importantly, they display many of the biologic proper- model of ARF, murine folic acid–induced acute tubular ties of native lipoxins and ATLs in vitro and are potential necrosis (ATN) was used. This model, as previously de- prototype therapeutics [14]. Native LXA4, ATLs, and scribed [22], involved administering 250 mg/kg body several synthetic lipoxin analogs have already been dem- weight of folic acid by intraperitoneal injection. Animals onstrated to have impressive anti-inflammatory and pro- were sacrificed at 0, 3, 6, 24, and 72 hours. Serum creati- resolution activity in experimental dermal inflammation, nine increased approximately four- to five-fold over the glomerulonephritis, and hind limb–induced second or- course of the experiment [22]. gan injury [15–20]. Taqman-PCR We have previously demonstrated the protective effect Total RNA was isolated from murine kidneys and of a stable analog of aspirin-triggered 15-epi-LXA , 15- 4 spleens using Trizol solution (Gibco, Paisley, UK) ac- epi-16-(FPhO)-LXA -Me in experimental murine ARF 4 cording to the manufacturer’s instructions. For real-time in vivo as manifested by significant functional and mor- polymerase chain reaction (PCR) analysis of murine tis- phologic protection and by blunted chemokine and cyto- sue, chromosomal DNA was removed from total RNA kine responses [21]. The latter experiments focused on using DNase I. DNAse-treated RNA (2 ␮g) was tran- a relatively small number of mediators that have pre- scribed to cDNA using standard procedures. Real-time viously been implicated in the pathogenesis of ischemia/ PCR was performed on a TaqMan ABI 7700 Sequence reperfusion injury or have been identified as players in Detection System (Applied Biosystems, Weiterstadt, lipoxin-mediated cytoprotection in other systems. While Germany) using heat-activated TaqDNA polymerase such targeted approaches can shed valuable new light on (Amplitaq Gold; Applied Biosystems, Weiterstadt, Ger- molecular mechanisms of disease, they are, by definition, many) [23]. After an initial hold of 2 minutes at 50ЊC biased by and dependent on, previous work in the bio- and 10 minutes at 95ЊC, the samples were cycled 40 times logic system of interest or in related systems. With ad- at 95ЊC for 15 seconds and 60ЊC for 60 seconds. Target vances in transcriptomics, it is now apparent that even gene forward and reverse primers and probes were de- simple perturbations of single-cell systems induce com- signed using Primer Express 1.5 software (Applied Bio- plex and highly coordinated changes in gene expression. systems, Foster City, CA, USA). Commercially available By applying these emerging technologies to well-charac- predeveloped TaqMan assay reagents (PDARs) were terized disease models it should be possible to decipher, in used for the internal standards human glycerin-aldehyde- 3-phosphate-dehydrogenase (GAPDH) and 18S ribo- an unbiased fashion, the full spectrum of genomic events somal RNA (18S rRNA). All primers and probes were that underpin organ dysfunction and the complex sym- obtained from Applied Biosystems, Weiterstadt, Ger- phony of interrelationships within these disease-related many. The primers were cDNA-specific, not amplifying gene expression networks, ultimately identifying the ge- genomic DNA. The following sequence of oligonucleo- nomic switches that direct these responses toward resolu- tide primers (300 nmol) and probes (100 nmol) were tion and repair. Against this background, we have em- used: (1) ADAM8 (forward primer: 5Ј-GCCCCTTGA ployed transcriptome profiling using oligonucleotide ACGCTCCTT-3Ј; reverse primer: 5Ј-TTCCATCCAT microarrays and computational gene annotation in the GCAAACCTTTC-3Ј; and probe: 5Ј-TATTGCAGGG present study to identify the profile of genes whose ex- CACCAAGTGCGAGG-3Ј); (2) claudin-1 (forward pression is altered in experimental murine ischemia/ primer: 5Ј-GATGTGGATGGCTGTCATTGG -3Ј; re- reperfusion injury, and to probe ischemia/reperfusion verse primer: 5Ј-CCATGCTGTGGCCACTAATGT-3Ј; injury–modulated genes whose expression levels are fur- and probe: 5Ј-CGCCAGACCTGAAAT-3Ј); (3) clau- ther influenced by 15-epi-16-(FPhO)-LXA4-Me. din-3 (forward primer: 5Ј-CATCACGGCGCAGATC 482 Kieran and Doran et al: Transcriptomic response to renal ischemia-reperfusion injury

AC-3Ј; reverse primer: 5Ј-TGCTCTGCACCACGCA tures such as introns, exons, and promoter regions [26]. GTT-3Ј; and probe: 5Ј-CATCCACAGGCCCTC-3Ј); (4) Having determined the sequence of the gene, the amino claudin-7 (forward primer: 5Ј-ATGATGAGCTGCAA acid sequence of corresponding proteins was deduced by AATGTACGA-3Ј; reverse primer: 5Ј-CACCAGGGA open reading frame analysis. The amino acid sequences CACCACCATTAA-3Ј; and probe: 5Ј-CCCTGCAGGC were thus assembled further by homology searching of CACT-3Ј); and (5) meprin-1␤ (forward primer: 5Ј-CAC the existing protein databases and structural modeling AAGAGATGGCCACATACCA-3Ј; reverse primer: of the protein to identify conserved structural motifs or 5Ј-CGATAGCGCTCAAAAGCATTG-3Ј; and probe: structural domains shared by other proteins. For protein 5Ј-CAGCTTGGAAATGAAT-3Ј). structural analysis, the structure prediction meta server Target and housekeeper cDNA templates (GAPDH was employed [27]. This is an integrated platform for and 18S rRNA) were quantified by standard curves of di- the utilization of a number of high-quality structure pre- luted standard cDNA. The expression of each target gene diction servers and permits identification of conserved was normalized to both different housekeeping tran- sequences by comparison with the Structural Classifica- scripts. All measurements were performed in duplicate. tion of Proteins (SCOP) database [28].

Controls consisting of ddH2O were negative in all runs.

Oligonucleotide microarray (gene chip) analysis RESULTS AND DISCUSSION Oligonucleotide microarrays identify claudin-1, -3, Twelve ␮g of total RNA were annealed to the T7-(dt)24 primers. First-strand cDNA synthesis was completed by and -7, and ADAM8 among 226 known genes whose incubating the mixture with 2 ␮L of Superscript II RT mRNA levels are increased in response to (20,000 units) at 42ЊC for 1 hour. Second-strand synthesis experimental murine renal ischemia/reperfusion injury was carried out in a total volume of 150 ␮L. Following To reduce the impact of background physiologic inter- purification, the cDNA pellet was precipitated with am- animal variation on mRNA levels, total RNA was pooled monium acetate and resuspended in RNAse-free water. from three individual animals from sham-operated and Synthesis of cRNA was performed by in vitro transcrip- ischemia/reperfusion injury animal groups. Oligonucleo- tion. The amplified cRNA was purified with an affinity tide microarray analyses were performed in duplicate resin column. The cRNA was fragmented and hybridized with a correlation coefficient between experiments of to the Affymetrix GeneChip Murine U74Av2 array 0.92. Of a total of 12,488 genes displayed on the microar- (Santa Clara, CA, USA), permitting the assessment of ray, 6795 (54%) were expressed in the kidney under the abundance of approximately 12,500 mRNA tran- basal conditions. Of these, 1016 were altered following scripts representing known genes and ESTs. Hybridiza- ischemia/reperfusion injury, 445 of which had expression tion proceeded overnight. Subsequent washing and stain- changes of two-fold or greater. Two hundred twenty-six ing of the arrays was carried out using the Affymetrix known genes were found to be up-regulated. Table 1 GeneChip fluidics station protocol EukGE-WS2 (Santa shows changes in several functional classes of genes im- Clara, CA, USA). Following washing and staining, the plicated in the pathogenesis of ischemic ARF that are probe arrays were scanned twice at 3 ␮m resolution candidate targets for lipoxin-mediated renoprotection, using the Affymetrix GeneChip System confocal scanner including chemoattractants, cytokines, chemokines and (Santa Clara, CA, USA). chemokine receptors (e.g., interleukin-6, GRO-1, inter- Data analysis was performed using Affymetrix Gene- feron-␥, and C-C receptor 1); growth factors and related Chip software (Santa Clara, CA, USA). The criteria molecules (e.g., thrombospondin-1 and early growth re- used to identify significant differential regulation of the sponse gene 1); adhesion molecules (e.g., intercellular transcripts was a greater than 2.0-fold change in expres- adhesion molecule-1); cell cycle and apoptosis-related sion in replicate experiments. genes (e.g., 14-3-3-␸ and annexin III); transcription fac- tors (e.g., ATFx); proteases (e.g., lipocalin-2); and cy- Bioinformatic annotation of unknown sequences toskeletal proteins (e.g., ␤-tubulin). cDNA sequences which did not show homology with Among the molecules previously implicated in the known genes by Basic Local Alignment Search Tool pathogenesis of ischemic ARF whose mRNA levels were (BLAST) search against the nonredundant database increased with microarray analysis are the adhesion mol- were used as input into an annotation pipeline. Briefly, ecules ICAM-1 and vascular cellular adhesion mole- the annotation pipeline consisted of serial stages of ho- cule-1 (VCAM-1), the cell cycle inhibitors p21 and 14-3- mology searching, using BLAST and position-specific 3-sigma, and an EST homologous to rat kidney inducible iteration (PSI)-BLAST searches of the high-throughput molecule-1 (KIM-1) [29–31]. These served as positive genomic sequence databases to identify genomic clones controls for the system. Of interest was the coordinated [24, 25]. The identified clones were used in GENSCAN- up-regulation of three members of the claudin family, based sequence topography analysis to identify gene fea- claudins 1, 3, and 7. Claudins are tight junction integral Kieran and Doran et al: Transcriptomic response to renal ischemia-reperfusion injury 483

Table 1. Identification of claudin-1, -3, and -7, and ADAM8 among other genes whose mRNA levels are increased in response to experimental renal ischemia/reperfusion injury, as determined by oligonucleotide microarray analysis

Fold change in mRNA levels GenBank Accession Gene OM1a OM2a Chemoattractants, cytokines, chemokines and chemokine receptors X54542 Interleukin-6 2.5 5.0 U49513 Small inducible cytokine A9 (MIP-1␣) 5 11.5 J04596 GRO-1 4.3 11.3 U41341 Endothelial monocyte-activating polypeptide-1 3.3 3.5 V00755 IFN-␤ fibroblast 3 2.8 X16834 Mac-2 antigen 10.4 11.4 U29678 C-C receptor 1 12.5 5.0 X59769 IL-1 receptor type 2 3.6 7.0 X57796 TNF receptor member 1␣ 2.4 14.1 U69599 IFN gamma receptor type 2 6.3 11.6 Growth factors and related molecules D25540 TGF␤ type 1 receptor 2.4 2.2 M62470 Thrombospondin-1 4.3 4.9 M28845 Early growth response gene 1 7.8 9.3 Adhesion molecules M90551 ICAM-1 3.2 4.3 M94487 VCAM-1 3.7 10.0 AF072127 Claudin-1 4.8 5.9 AF095905 CPETR2 (claudin-3) 4.5 2.7 AF087825 Claudin-7 7.1 5.2 AA986114 KIM-1 (rat homologue) 14.3 15 X14432 Thrombomodulin 4.3 5.6 Cell cycle, apoptosis, heat shock proteins, and other mediators of oxidative stress AF058798 14-3-3 protein sigma 9.9 19.5 AF017128 Fos-like antigen 1 4.4 14.1 M12571 Heat shock protein 70 kD 3 11.3 9.1 AB020886 SSeCKS 15.3 35.2 U09507 Cyclin-dependent kinase inhibitor1A (p21) 9.8 16.8 V00835 Metallothionein 1 5.2 3.9 X68273 CD68 6.4 4.6 D14077 Clusterin 5.3 3.5 X13333 CD14 5.6 5.1 M69260 Lipocortin1 5 4.9 AJ001633 Annexin III 11.3 22.5 L31958 pMAT1 (clone) 15.2 41.7 Transcription factors and other modulators of cell signaling AB012276 ATFx 4.1 4.0 U19118 LRG-21 5.0 4.5 Proteases X81627 Lipocalin-2 64.8 42.9 Cytoskeletal X04663 Beta-tubulin, Mbeta 5 50.4 42.4 X13297 Vascular smooth muscle, ␣-actin 12.4 11.3 M14044 Calpactin 1 heavy chain (p36) 8.1 5.9 Matrix proteins or components and modulators of cell matrix X56304 Tenascin C 8 8.7 Z68618 Transgelin 3.9 4.8 U03715 Procollagen, type XVIII, alpha 1 9.9 4.8 Z31362 (Balb/C) Tx 01 8.3 7.2 Miscellaneous U77630 Adrenomedullin 3.8 2.6 X13335 ADAM8 5.5 13.3 Abbreviations are: IFN-␤, interferon beta; IL-1, interleukin-1; TNF, tumor necrosis factor; ICAM-1, intercellular adhesion molecule-1; VCAM-1, vascular cellular adhesion molecule-1; CPETR2, clostridium perfringens enterotoxin receptor 2; KIM-1, kidney inducible molecule-1; SSeCKS, SrC-suppressed C kinase substrate; ATFx, expression of a specific mouse gene; ADAM8, a disintegrin and metalloproteinase domain protein 8. a Oligonucleotide microarrays (OM) analyzed RNA pooled from 3 sham and 3 ischemia/reperfusion injury animals, fold change reflecting expression in ischemia/ reperfusion injury animals using sham as baseline for comparison 484 Kieran and Doran et al: Transcriptomic response to renal ischemia-reperfusion injury

junctional adhesion molecule) expressed by the kidney were not perturbed during ischemia/reperfusion injury (data not shown), while mRNA for claudin-2 was sup- pressed [Ϫ1.9-fold on both oligonucleotide microarrays (OM), not reaching cutoff levels used for this analysis], suggesting differential regulation of gene expression in this setting. The finding of higher mRNA levels for clau- din-1, -3, -7 in the context of ischemia/reperfusion injury is intriguing given the dramatic changes in epithelial po- larity and adhesion that typifies ischemic ATN [3]. In- deed, recent studies have suggested additional dynamic roles for claudins as modulators of cell signaling and the actin cytoskeleton [32]. Claudin proteins, including claudin-1 perturbed during ischemia/reperfusion injury in the present study, have been shown to recruit mem- brane-type matrix metalloproteinases (MT-MMPs), in- cluding promatrix metalloproteinase-2, enhancing acti- vation of pro-MMP-2 in vitro [35]. Together these Fig. 1. Confirmation by real-time polymerase chain reaction (RT- observations raise the possibility of a role for claudins PCR) of differential expression of oligonucleotide microarray-identified genes in experimental ischemia/reperfusion injury in murine kidneys in tissue remodeling in ischemic ARF. in vivo and modification of expression by 15-epi-16-(FPhO)-LXA4-Me. In light of the perturbations in claudin expression de- RT-PCR confirms up-regulation of claudin-1, claudin-3, claudin-7, and scribed above and their potential role in regeneration ADAM8, and down-regulation of meprin-1␤ in murine kidneys har- vested from mice with experimental ischemia/reperfusion injury ( )as and repair, it is noteworthy that renal mRNA levels for compared to sham-operated controls (᭿) and modification of expression a disintegrin and metalloproteinase domain protein 8 ᮀ by 15-epi-16-(FPhO)-LXA4-Me ( ). Expression data is shown as ratio (ADAM8) were also enhanced in experimental IRI. This to 18SrRNA (y axis shows ratio to rRNA). finding was confirmed by real-time PCR analysis (Fig. 1). The ADAM gene family is composed of proteins that have transmembrane and cytoplasmic domains com- membrane proteins [32]. To date, 24 members of the posed of MMP-like, disintegrin-like, cysteine-rich, and claudin gene family have been described [32]. Within epidermal growth factor (EGF)-like (snake venom the kidney, claudins form an important part of tight MMP-like) domains. ADAM family members have been junctions, as vividly highlighted with the reporting that implicated in cell fusion, cell-to-matrix adhesion, and mutations in claudin-16 (paracellin-1), which is exclu- proteolytic degradation of extracellular matrix, [36] and sively expressed in the thick ascending limb of Henle, can modulate cell signaling systems which are key for are associated with hereditary hypomagnesemia [33]. normal cellular processes such as cell morphogenesis, Damage to tight junctions resulting in impaired cell-cell wound healing, and tumor cell invasion [37]. adhesion and disruption of the permeability barrier are Changes in renal mRNA levels following ischemia/ features of epithelial cell damage in ischemic acute renal reperfusion injury could represent ischemia/reperfusion failure [34]. injury–triggered changes in the renal parenchymal tran- To confirm the differential expression of these key scriptome, infiltration of renal parenchyma by leuko- modulators of barrier integrity, real-time PCR was per- cytes, or a combination of these events. In this regard, formed with RNA extracted from the kidneys of animals it is noteworthy that neither claudin-1 nor ADAM8 were with ischemia/reperfusion injury (Fig. 1). Under basal expressed in splenic tissue harvested from sham-oper- conditions mouse kidneys expressed little mRNA for ated and ischemia/reperfusion injury animals, as assessed claudin-1, whereas mRNA levels were significantly en- by RT-PCR, suggesting that the increase in their renal hanced following ischemia/reperfusion injury. Similarly, mRNA levels in ischemia/reperfusion injury reflected low expression of claudin-3 was seen in sham-operated perturbation of the transcriptome of resident renal cells. animals, whereas claudin-3 mRNA levels were signifi- cantly enhanced in ischemia/reperfusion injury. Clau- Oligonucleotide microarrays identify epidermal growth ␤ din-7 expression was also induced in the latter setting. factor and meprin-1 among 112 known genes whose Real-time PCR demonstrated fold increases of 6.97, 4.71, mRNA levels are decreased in response to and 9.24 for claudin-1, -3, and -7, respectively, confirming experimental ischemia/reperfusion injury the data obtained with oligonucleotide microarrays. Of the 445 transcripts (25%) whose mRNA levels were Messenger RNA levels for several other claudins (clau- altered two-fold or greater in experimental ischemia/ din-8 and -6) and tight junction proteins (occludin and reperfusion injury, 112 were down-regulated. Among the Kieran and Doran et al: Transcriptomic response to renal ischemia-reperfusion injury 485

Table 2. Identification of meprin-1␤ among other genes whose expression is suppressed in response to experimental murine renal ischemia/reperfusion injury

Fold change in mRNA levels GenBank Accession Gene OM1a OM2a Chemoattractants, cytokines, chemokines and chemokine receptors L12030 Stromal cell-derived factor 1 (CXCL12) Ϫ2 Ϫ2.1 D17444 Leukemia inhibitor factor receptor Ϫ3.8 Ϫ2.5 Growth factors U22516 Angiogenin Ϫ3.0 Ϫ3.8 M30641 Fibroblast growth factor-1 Ϫ2.7 Ϫ2.3 V00741 Epidermal growth factor Ϫ8.8 Ϫ5.4 M34094 Midkine Ϫ2.6 Ϫ2.6 Adhesion molecules AB010144 Mitsugumin 29 Ϫ2.2 Ϫ2.1 Cell cycle, apoptosis, heat shock proteins, and other mediators of oxidative stress X98055 Glutathione-S-transferase theta 1 Ϫ6.3 Ϫ11.6 AF159298 Nucleotide phosphodiesterase (Pde1a7) Ϫ4.1 Ϫ3.5 Proteases and dipeptidases L15193 Meprin-1␤Ϫ18.9 Ϫ10.7 AJ131851 Cathepsin F Ϫ3.5 Ϫ2.7 D13139 Dipeptidase 1 Ϫ4 Ϫ3.8 Cytoskeletal M21828 Growth arrest specific Ϫ4.7 Ϫ3.7 U04354 Adseverin Ϫ3.2 3.0 Matrix proteins Z35167 Procollagen, type IV, ␣ 4 Ϫ2.4 Ϫ12.8 Transport proteins U38652 Solute carrier family 22, member 1 Ϫ3.0 Ϫ2.5 X77241 Solute carrier family 17, member 1 Ϫ2.2 Ϫ2.0 D88533 NBAT Ϫ2.1 Ϫ2.3 L02914 Aquaporin 1 Ϫ2.5 Ϫ2.6 U73521 Solute carrier family 1, member 1 Ϫ2.5 Ϫ2.0 AF058054 Solute carrier family 16, member 7 Ϫ2.1 Ϫ2.4 AF075261 Orphan transporter (Xtrp3) Ϫ2.7 Ϫ2.4 AJ012754 Solute carrier family 7, member 7 Ϫ2.7 Ϫ2.8 Miscellaneous U42443 MECA39 Ϫ4.8 Ϫ5.2 AB005141 Klotho Ϫ2.8 Ϫ2.8 a Oligonucleotide microarrays (OM) were performed with RNA pooled from 3 sham and 3 ischemia/reperfusion injury animals, fold change reflecting expression in ischemia/reperfusion injury using sham as baseline for comparison

major families of genes whose mRNA levels were sup- peptidases expressed by renal proximal tubules and brush pressed were chemoattractants, cytokines, chemokines border membranes of the intestine [39]. Meprin A is com- and chemokine receptors (e.g., stromal cell derived fac- posed of homodimers of ␣ subunits, whereas meprin B tor-1 and leukemia inhibitory factor receptor; growth is a heterodimer of ␣ and ␤ subunits. Meprins ␣ and ␤ factors [e.g., EGF, angiogenin, and fibroblast growth fac- subunits have markedly different substrate and peptide tor]; adhesion molecules (e.g., mitsugumin 29); cell cycle- bond specificity, meprin ␣ preferring substrates with and apoptosis-related genes (e.g., glutathione-S-trans- small or hydrophobic residues while meprin ␤ selects for ferase ␭1); proteases (e.g., meprin-1␤; and cytoskeletal acidic amino acids, such as those found in the peptide proteins (e.g., adseverin) (Table 2). hormone gastrin [40]. Meprin A inhibition has been de- In keeping with previous reports, renal expression of scribed as protective in ischemia/reperfusion injury [41], EGF was transiently decreased after ischemia/reperfusion while little is known about meprin B in ischemia/reperfu- injury, further validating the approach [38]. A noteworthy sion injury. Meprins are the only known endopeptidases finding was down-regulation of meprin-1␤ (meprin B). in brush border membranes that degrade proteins. Inter- Suppression of meprin-1␤ was confirmed by real-time estingly, down-regulation of meprin-1␤ has been recently PCR analysis of RNA from the kidneys of sham-oper- reported in experimental hydronephrosis, suggesting that ated and ischemia/reperfusion injury animals (Fig. 1). diverse renal insults modulate the expression of this en- Normal mouse kidney displays strong detectable meprin dopeptidase in vivo [42]. transcript levels, the expression of which is blunted in the Real-time PCR analysis of mRNA levels in mice with injured kidney. Meprins (A and B) are zinc metalloendo- folic acid–induced ATN also demonstrated increased 486 Kieran and Doran et al: Transcriptomic response to renal ischemia-reperfusion injury

Fig. 3. Serum creatinine levels assessed after 24 hours of reperfusion.

N ϭ 3 for sham-operated controls, 15-epi-16-(FPhO)-LXA4-Me (15 ␮g/ mouse), and vehicle-treated acute renal failure (ARF). Modified from Leonard et al, J Am Soc Nephrol 13:1657–1662, 2002.

computational gene annotation pipeline as described in the experimental procedures. These studies have led to the identification of a further 24 genes whose expression is altered in response to renal ischemia/reperfusion in- jury, a representative cohort of which are detailed in Table 3. The unknown sequences identified in the mi- croarray experiment were used as input to the annotation pipeline. Typical of such analysis was the EST AI836034, found to be up-regulated 2.1- and 2.3-fold on oligonucleo- tide microarrays 1 and 2, respectively. First, the sequence was retrieved using the ENTREZ sequence retrieval sys- Fig. 2. Confirmation by real-time polymerase chain reaction of differential expression of oligonucleotide microarray-identified genes in experimental tem at the National Centre for Biotechnology Informa- acute tubular necrosis in murine kidneys in vivo. (A through D) RT- tion. This EST, UI-M-AQ0-aad-f-06-0-UI.s2, when com- PCR confirms up-regulation of claudin-1, claudin-3, claudin-7, and pared to the nonredundant nucleotide databases this ADAM8 and (E) down-regulation of meprin-1␤ in murine kidneys (N ϭ 3) harvested from mice with folic acid induced acute tubular necrosis sequence produced no significant homology to any known (ATN). A time course with 0, 3, 6, 24, and 72 hours after folic acid is gene. However this sequence was 99% identical to a shown. Expression data is shown as ratio to 18S rRNA (y axis shows 998-nucleotide cDNA clone from adult mouse brain, ratio to rRNA). clone:0710007K04. To determine the genomic structure of this sequence, GENSCAN analysis of the clone topog- raphy was completed. This algorithm identifies features levels for claudin-1, -3, and -7, and ADAM8, and de- such as introns, exons, and promoter regions in genomic creased meprin-1␤ mRNA levels, suggesting that these sequence. Analysis of cDNA clone:0710007K04 demon- events are common to both ischemic and nephrotoxic strated the presence of a 218 amino acid single exon ATN (Fig. 2). protein. The EST sequence used as input was contained within the predicted open-reading frame of the cDNA, Bioinformatic annotation of cDNAs corresponding to positioned in the 5Ј untranslated region adjacent to the transcripts without homology to known genes amino acid sequence. Pairwise alignment was performed identifies a further cohort of genes whose expression with the EST and cDNA sequences to confirm the pre- is altered in experimental ischemia/reperfusion injury dicted location. The amino acid sequence thus generated Of the 445 renal transcripts altered in response to was then compared to the nonredundant protein data- ischemia/reperfusion injury, 107 corresponded to ESTs base, where a substantial homology was found to a hu- without homology to known genes, as determined by man protein similar to enigma (LIM domain protein) standard database search strategies of the human and (GenBank accession no. AAH14521). Enigma proteins murine genome databases. The annotation of these tran- are a family of cytoplasmic proteins that possess a PDZ scripts remains a major obstacle to the streamlined analy- domain at the amino terminal and one to three Lun-11 sis of data derived from commercial microarrays. To Isle-1 mec-3 (LIM) domains at the carboxyl terminal. By further characterize these transcripts we employed a virtue of these two protein interacting domains, enigma Kieran and Doran et al: Transcriptomic response to renal ischemia-reperfusion injury 487

Table 3. Bioinformatic annotation of cDNAs corresponding to transcripts without homology to known genes by BLAST search whose expression is altered in response to experimental murine renal ischemia/reperfusion injury

Fold change in mRNA level GenBank Accession Gene annotation OM1a OM2a Induced transcripts AI132207 Eukaryotic translation initiation factor 1A 4.2 2.1 AW122030:d Phosphoserine aminotransferase 2.3 2.1 AA270365 Beta-1,4 mannosyltransferase 2.2 5.5 AW122864 RNA polymerase II, polypeptide H 2 2.1 AW045413 DNA polymerase delta subunit 3 6 AI845934 EBNA1 binding protein 2 2.1 2.1 AI509617 Mitochondrial ribosomal protein S18b 2.1 2.7 AW060179 Zw10 interactor 2.1 2.0 AI836034 Enigma (LIM domain protein) 2.3 2.1 AI841914 Acyl CoA binding protein 2.8 2.3 Suppressed transcripts AW050325 Crystallin, lamda 1 Ϫ2 Ϫ2.3 AA656014 Transmembrane 7 superfamily, member 1 Ϫ2.1 Ϫ2.2 AW124932 3-oxoacid CoA transferase Ϫ2.4 Ϫ2.1 AA675075 Kidney and liver proline oxidase 1 Ϫ2.8 Ϫ3.7 NP011467 Sepiapterin reductase Ϫ2.4 Ϫ2.6 AI845337 Quinoid dihydropteridine reductase Ϫ2.1 Ϫ2.0 AI840094 Androgen-inducible aldehyde reductase Ϫ2.1 Ϫ2.2 AW211207 Cgi-18 protein (asc-1 complex subunit p50) Ϫ3 Ϫ2.0 AV334517 Biphenyl hydrolase-related Ϫ2.3 Ϫ2.3 AI842259 Pyruvate dehydrogenase kinase, isoenzyme 3 Ϫ2.4 Ϫ2.4 AW121801 Adenosine kinase Ϫ2.4 Ϫ2.1 AI314227 Aspartoacylase aspa Ϫ3.2 Ϫ2.1 AI852646 Aminoacylase-1 Ϫ2.7 Ϫ2.0 AW123061 Never in mitosis gene A-related kinase 2 Ϫ2.5 Ϫ2.4 a Oligonucleotide microarrays (OM) were performed with RNA pooled from 3 sham and 3 ischemia/reperfusion injury animals

proteins are capable of protein-protein interactions. It fusion injury kidneys in the presence and absence of has been proposed that enigma proteins may act as 15-epi-16-(FPhO)-LXA4-Me. For the purposes of this adapters between kinases and the cytoskeleton [43]. This analysis, ischemia/reperfusion injury–associated gene ex- general approach was used to identify 10 up-regulated pression levels served as basal levels and lipoxin-medi- and 14 down-regulated genes in experimental renal isch- ated changes were expressed as fold change relative to emia/reperfusion injury. this baseline parameter. Pretreatment with lipoxin ana- log resulted in a blunting of ischemia/reperfusion injury– Modulation of the renal transcriptomic response to associated gene induction (Tables 4 and 5). This effect ischemia/reperfusion injury by was observed in all of the functional classes of genes 15-epi-16-(FPhO)-LXA4-Me induced by ischemia/reperfusion injury, including in- We have previously demonstrated a protective effect flammatory mediators such as chemoattractants, cyto- of a lipoxin-stable analog 15-epi-16-(FPhO)-LXA4-Me kines, chemokines and chemokine receptors, growth fac- in renal ischemia/reperfusion injury, establishing the tors, adhesion molecules, and other genes. This marked therapeutic potential of lipoxin and ATL analogs in this attenuation in both the magnitude and spectrum of gene setting (Fig. 3) [21]. The renoprotective effect of 15-epi- expression changes applied to genes typically expressed 16-(FPhO)-LXA4-Me was associated with reduced PMN by infiltrating leukocytes (e.g., CD14 and lipocortin 1) infiltration and modulated intrarenal cytokine expres- and to genes whose expression is most abundant in resi- sion. The finding of reduced mRNA levels for interleu- dent epithelial cells (e.g., claudins, KIM-1). kin-1 (IL-1), IL-6 and GRO-1 in lipoxin-treated animals In addition to attenuating proinflammatory responses in association with increased expression of suppressors induced by a range of important mediators, lipoxins, ATLs, of cytokine signaling (SOCS) -1 and SOCS-2 suggests and their analogs actively promote resolution of inflamma- that a complex interplay of transcriptional events un- tion by promoting macrophage clearance of apoptotic derpin the renoprotective effect of lipoxin in experimen- neutrophils and by stimulating the release of cytokines tal ischemia/reperfusion injury [21]. To further define and growth factors implicated in tissue remodeling and the molecular events associated with renoprotection we repair [e.g., transforming growth factor (TGF) beta] [3, 4]. performed oligonucleotide microarray-based compari- In a second analysis we investigated whether lipoxin son of mRNA expression profiles in the ischemia/reper- treatment was associated with modulation of the expres- 488 Kieran and Doran et al: Transcriptomic response to renal ischemia-reperfusion injury c ) Ϫ 5.1 Ϫ 4.5 Ϫ 2.2 Ϫ 1.1 Ϫ 2.7 Ϫ 1.2 Ϫ 2.3 Ϫ 3.9 Ϫ 1.7 Ϫ 1.4 Ϫ 2.7 Ϫ 4.4 Ϫ 1.5 Ϫ 1.0 Ϫ 1.4 Ϫ 3.6 Ϫ 2.4 Ϫ 4.0 Ϫ 3.1 Ϫ 2.8 Ϫ 6.7 Ϫ 3.5 Ϫ 2.9 Ϫ 1.5 Ϫ 1.5 Ϫ 1.1 Ϫ 2.4 Ϫ 2.6 Ϫ 1.6 Ϫ 5.9 Ϫ 1.1 Ϫ 1.8 Ϫ 2.3 Ϫ 1.5 Ϫ 2.1 Ϫ 12.9 ( Continued reperfusion injury a OM2 b 5.0 3.5 2.8 5.0 7.0 2.2 4.9 9.3 4.3 5.9 2.7 5.2 5.6 9.1 3.9 4.6 3.5 5.1 4.9 2.0 4.0 4.5 2.1 11.3 11.4 14.1 11.6 10.0 19.5 14.1 35.2 16.8 22.5 41.7 42.9 injury/Sham c -Me on mRNA levels of genes whose levels were increased 4 Fold change in mRNA levels Ϫ 1.4 Ϫ 2.7 Ϫ 1.2 Ϫ 1.5 Ϫ 1.1 Ϫ 2.2 Ϫ 3.9 Ϫ 1.3 Ϫ 2.0 Ϫ 2.3 Ϫ 2.4 Ϫ 3.4 Ϫ 1.0 Ϫ 1.7 Ϫ 1.2 Ϫ 2.6 Ϫ 2.8 Ϫ 1.8 Ϫ 3.3 Ϫ 5.0 Ϫ 7.8 Ϫ 3.7 Ϫ 2.3 Ϫ 1.8 Ϫ 1.5 Ϫ 1.2 Ϫ 2.5 Ϫ 2.1 Ϫ 1.6 Ϫ 1.5 Ϫ 2.1 Ϫ 3.1 Ϫ 3.2 Ϫ 1.5 Ϫ 2.9 reperfusion injury a OM1 b 2.4 6.3 injury/Sham Ischemia/reperfusion Lipoxin/ischemia/ Ischemia/reperfusion Lipoxin/ischemia/ during ischemia/reperfusion injury Gene fibroblast 3 type 1 receptor 2.4 IFN gamma receptor type 2 The transcriptomic response associated with lipoxin-mediated renoprotection: Influence of 15-epi-16-(FPhO)-LXA X54542 Interleukin-6 2.5 U49513J04596U41341V00755X16834U29678 SmallX59769 inducible cytokine A9X57796 GRO-1 EndothelialU69599 monocyte-activating polypeptide IFN- ␤ Mac-2 antigen C-C receptor 1 IL-1 receptor type 2 TNF receptor member 1 ␣ 3.3 5 10.4 4.3 3.6 12.5 3.5 11.5 M62470M28845M94487AF072127 Thrombospondin-1AF095905 Early growth responseAF087825 gene 1AA986114X14432 VCAM-1 Claudin-1 CPETR2 (claudin-3) Claudin-7 KIM-1 (rat homologue) Thrombomodulin 7.8 4.3 4.5 14.3 4.8 3.7 4.3 7.1 1.2 15 1.5 D25540M90551 TGF ␤ ICAM-1AF058798 14-3-3 protein sigma 3.2 9.9 AF017128M12571AB020886U09507V00835 Fos-likeX68273 antigen 1D14077 Heat SSeCKSX13333 shock protein 70 kDM69260 3 Cyclin-dependentAJ001633 kinase inhibitor1A (p21) MetallothioneinL31958 1 CD68AW060179 Clusterin CD14 Lipocortin Annexin 1 III Zw10 interactor pMAT1(clone) 9.8 11.3 4.4 15.3 5.2 5.3 6.4 11.3 5 2.1 15.2 5.6 AB012276 ATFx 4.1 U19118AI132207 LRG-21 Eukaryotic translation initiation factor 1A 4.2 5.0 X81627 Lipocalin-2 64.8 Table 4. GenBank Ac- Chemoattractants, cytokines, chemokines and chemokine receptors cession Growth factors and related molecules Adhesion molecules Cell cycle, apoptosis, heat shock proteins, and other mediators of oxidative stress Transcription factors and other modulators of cell signaling Proteases Kieran and Doran et al: Transcriptomic response to renal ischemia-reperfusion injury 489 c c ) 1 1.4 2.9 1.7 1.2 2.9 1.9 1.2 Ϫ 1.7 Ϫ 1.4 Ϫ 1.1 Ϫ 2.9 Ϫ 1.4 Ϫ 1.8 Ϫ 2.7 Ϫ 2.4 Ϫ 1.3 Ϫ 1.4 Ϫ 1.2 Ϫ 1.6 Ϫ 1.3 Ϫ 1.4 Ϫ 2.0 Ϫ 10.5 ( Continued reperfusion injury reperfusion injury a a OM2 OM2 b b 5.9 8.7 4.8 4.8 7.2 2.6 2.1 5.5 2.1 6 2.1 2.7 2.1 2.3 Ϫ 2.1 Ϫ 2.5 Ϫ 3.8 Ϫ 2.3 Ϫ 5.4 Ϫ 2.6 Ϫ 2.1 42.4 11.3 13.3 injury/Sham injury/Sham c c -Me on mRNA levels of genes which are suppressed following 4 1.2 2.5 3.0 1.4 2.5 3.5 2.0 1.4 Fold change in mRNA levels Fold change in mRNA levels Ϫ 1.5 Ϫ 1.2 Ϫ 1.2 Ϫ 2.2 Ϫ 1.3 Ϫ 1.4 Ϫ 2.8 Ϫ 1.9 Ϫ 9.7 Ϫ 1.3 Ϫ 1.9 Ϫ 1.5 Ϫ 1.5 Ϫ 1.5 Ϫ 1.2 Ϫ 1.7 reperfusion injury reperfusion injury a a OM1 OM1 b b 8.1 8 9.9 3.9 8.3 3.8 5.5 2.2 2.3 2.1 2 3 2.1 2.3 2.8 Ϫ 2 Ϫ 3.8 Ϫ 3.0 Ϫ 2.7 Ϫ 8.8 Ϫ 2.6 Ϫ 2.2 50.4 12.4 injury/Sham injury/Sham Table 4. (Continued) Ischemia/reperfusion Lipoxin/ischemia/ Ischemia/reperfusion Lipoxin/ischemia/ Ischemia/reperfusion Lipoxin/ischemia/ Ischemia/reperfusion Lipoxin/ischemia/ experimental murine ischemia/reperfusion injury 1 ␣ -actin ␣ Gene Gene Vascular smooth muscle, Calpactin 1 heavy chain (p36) Tenascin C Transgelin Procollagen, type XVIII, (Balb/C) Tx 01 Adrenomedullin ADAM 8 Beta-1,4 mannosyltransferase EBNA1 binding protein 2 RNA polymerase II, polypeptideDNA H polymerase delta subunit Mitochondrial ribosomal protein S18b Enigma (LIM domain protein) Acyl CoA binding protein Leukemia inhibitor factor receptor Angiogenin Fibroblast growth factor-1 Epidermal growth factor Midkine Mitsugumin 29 The transcriptomic response associated with lipoxin-mediated renoprotection: Effect of 15-epi-16-(FPhO)-LXA Fold change reflecting expression in ischemia/reperfusion injury animals using sham as baseline for comparison Oligonucleotide microarrays (OM) were performed with RNA pooled from 3 ischemia/reperfusion injury animals and 3 animals pretreated with lipoxin prior to induction of ischemia/reperfusion injury Fold change reflecting expression in lipoxin-treated animals using ischemia/reperfusion injury as baseline for comparison For abbreviations, seea Table 1. b c X04663 Beta-tubulin, M ␤ eta 5 L12030 Stromal cell-derived factor 1 (CXCL12) X13297 M14044 D17444 U22516 M30641 V00741 M34094 AB010144 X56304 Z68618 U03715 Z31362 U77630 X13335 AW122030:dAA270365 Phosphoserine aminotransferase AI845934 AW122864 AW045413 AI509617 AI836034 AI841914 Table 5. GenBank Ac- Cytoskeletal GenBank Chemoattractants, cytokines, chemokines and chemokine receptors cession Accession Growth factors Adhesion molecules Matrix proteins or components and modulators of cell matrix Miscellaneous 490 Kieran and Doran et al: Transcriptomic response to renal ischemia-reperfusion injury c 6.5 1.9 2.8 2.3 1.2 1.6 1.9 2.2 1.9 1.4 1.5 1.2 1.4 1.3 1.6 1.5 1.8 1.5 1.5 1.2 1.2 1.3 1.8 1.3 1.5 1.3 1.3 1.3 1.6 1.5 1.6 Ϫ 1.2 reperfusion injury a OM2 b Ϫ 3.5 Ϫ 3.7 Ϫ 2.7 Ϫ 3.8 Ϫ 3.7 Ϫ 3.0 Ϫ 2.5 Ϫ 2.0 Ϫ 2.3 Ϫ 2.6 Ϫ 2.0 Ϫ 2.4 Ϫ 2.4 Ϫ 2.8 Ϫ 5.2 Ϫ 2.8 Ϫ 2.3 Ϫ 2.2 Ϫ 2.1 Ϫ 2.6 Ϫ 2.0 Ϫ 2.2 Ϫ 2.0 Ϫ 2.3 Ϫ 2.4 Ϫ 2.1 Ϫ 2.1 Ϫ 2.0 Ϫ 2.4 Ϫ 11.6 Ϫ 10.7 Ϫ 12.8 injury/Sham c 9.1 2.1 1.9 3.0 1.2 1.6 2.1 3.7 2.3 1.4 1.5 1.2 1.9 1.0 1.6 1.6 1.6 1.5 1.5 1.3 1.4 1.2 2.0 1.2 1.6 1.3 1.3 1.5 2.2 1.5 2.2 Fold change in mRNA levels Ϫ 1.1 reperfusion injury a OM1 b 18.9 Ϫ 6.3 Ϫ 4.1 Ϫ 2.8 Ϫ 3.5 Ϫ 4.0 Ϫ 4.7 Ϫ 3.2 Ϫ 2.4 Ϫ 3.0 Ϫ 2.2 Ϫ 2.1 Ϫ 2.5 Ϫ 2.5 Ϫ 2.1 Ϫ 2.7 Ϫ 2.7 Ϫ 4.8 Ϫ 2.8 Ϫ 2.0 Ϫ 2.1 Ϫ 2.4 Ϫ 2.4 Ϫ 2.1 Ϫ 2.1 Ϫ 3 Ϫ 2.3 Ϫ 2.4 Ϫ 2.4 Ϫ 3.2 Ϫ 2.7 Ϫ 2.5 injury/Sham Table 5. (Continued) Ischemia/reperfusion Lipoxin/ischemia/ Ischemia/reperfusion Lipoxin/ischemia/ 4 ␣ Gene Kidney and liver proline oxidase 1 Meprin-1 ␤Ϫ Cathepsin F Dipeptidase 1 Growth arrest specific Procollagen, type IV, Solute carrier family 22, member 1 Adseverin Solute carrier family 17,NBAT member 1 Aquaporin 1 Solute carrier family 1,Solute member carrier 1 family 16,Orphan member transporter 7 (Xtrp3) Solute carrier family 7, member 7 MECA39 Klotho Transmembrane 7 superfamily, member 1 Sepiapterin reductase Quinoid dihydropteridine reductase Androgen-inducible aldehyde reductase Biphenyl hydrolase-related Pyruvate dehydrogenase kinase, isoenzyme 3 Aspartoacylase aspa Aminoacylase-1 Fold change reflecting expression in ischemia/reperfusion injury animals using sham as baseline for comparison Oligonucleotide microarrays (OM) were performed with RNA pooled from 3 ischemia/reperfusion injury animals and 3 animals pretreated with lipoxin prior to induction of ischemia/reperfusion injury Fold change reflecting expression in lipoxin-treated animals using ischemia/reperfusion injury as baseline for comparison 42443 For abbreviations, seea Table 1. b c X98055 Glutathione-S-transferase theta 1 AF159298 Nucleotide phosphodiesterase (Pde1a7) AA675075 L15193 AJ131851 D13139 M21828 Z35167 U38652 U04354 X77241 D88533 L02914 U73521 AF058054 AF075261 AJ012754 AB005141 AW050325 Crystallin, lambda 1 AA656014 AW124932 3-oxoacid CoA transferase NP011467 AI845337 AI840094 AW211207 Cgi-18 protein (asc-1 complex subunit p50) AV334517 AI842259 AW121801 Adenosine kinase AI314227 AI852646 AW123061 Never in mitosis gene A-related kinase 2 GenBank Cell cycle, apoptosis, heat shock protein, and other mediators of oxidative stress Accession Proteases and dipeptidases Cytoskeletal Matrix proteins Transport proteins Miscellaneous Kieran and Doran et al: Transcriptomic response to renal ischemia-reperfusion injury 491

Table 6. Genes whose mRNA levels were not altered in response to experimental renal ischemia/reperfusion injury but were altered significantly by lipoxin analog

Fold change in mRNA levels Ischemia/reperfusion Lipoxin vs. ischemia/ GenBank injury vs. Shamb reperfusion injuryc Accession Gene OM1a OM2a OM1a OM2a M15268 Aminolevulinic acid synthase 2, erythroid 1.8/1.1 Ϫ4.9/Ϫ3.9 U49915 Adipocyte specific protein adipoQ 1.2/1.2 Ϫ4.2/Ϫ3.1 U96684 Paired-Ig-like receptor A3 1.8/1.5 2.8/Ϫ3.4 V00722 Mouse gene for ␤-1-globin 1.5/1.5 Ϫ3.0/Ϫ2.4 X04673 Adipsin 1.4/1.2 Ϫ7.0/Ϫ19.1 D67076 ADAMTS1 1.6/1.9 Ϫ2.2/Ϫ2.2 AB017615 Eos protein Ϫ1.7/Ϫ1.0 Ϫ3.1/Ϫ2.2 a Oligonucleotide microarrays (OM) analyzed RNA pooled from 3 sham and 3 ischemia/reperfusion injury animals b Fold change reflecting expression in ischemia/reperfusion injury animals using sham as baseline for comparison c Fold change reflecting expression in lipoxin-treated using ischemia/reperfusion injury as baseline for comparison

sion of a cluster of genes that was not perturbed by CONCLUSION ischemia/reperfusion injury alone and could represent The ischemia/reperfusion injury–associated and reno- a lipoxin-stimulated renoprotective genomic program. protective transcriptomic changes described herein will To this end, 5779 of the total number of 6795 renal- provide the template for further dissection of the molecu- expressed genes displayed on the microarray (80%) were lar component of renal injury by ischemia/reperfusion in- not perturbed by ischemia/reperfusion injury alone. Of jury and other insults in vitro and for the elucidation of these, 7 known genes (0.12%) were perturbed by two- novel mechanisms of renoprotection in renal disease. fold or more during lipoxin treatment on both micro- arrays, 6 genes being induced and 1 gene being suppressed ACKNOWLEDGMENTS (Table 6). The relatively small number of lipoxin-respon- The expert technical assistance of Karin Frach and Sandra Irrgang sive, ischemia/reperfusion injury–independent genes iden- is gratefully acknowledged. These studies were supported by grants tified in this analysis, in comparison to the striking modu- from the Irish Government’s Programme for Research in Third Level lation of ischemia/reperfusion injury–associated genes, Institutions, the Health Research Board of Ireland (N.K., S.C., C.G., suggests that the lipoxin analog conferred renoprotec- H.B.), the Wellcome Trust (C.G., H.B., C.T.), the Punchestown Kidney Renal Research Fund (P.D.), EU fifth framework (H.R.B., C.G., P.D.), tion either by inhibiting a pathogenic event, yet to be de- and the Science Foundation of Ireland (C.T.), and from the National fined at the core of ischemia/reperfusion injury, or that the Institutes of Health and the National Kidney Foundation, USA (M.B. and analog had dramatic effects on gene expression by virtue H.R.) and the German National Genome Project (DHGP01KW9922/2) (M.K.). We thank Berlex, San Francisco, California, for the generous of its ability to influence several key pathophysiologic gift of the stable synthetic LXA4 analog 15-epi-16-(FphO)-LXA4-Me. events in parallel, such as vascular tone, leukocyte recruit- ment and clearance, and cytokine responses. Most lipoxin- Reprint requests to Dr. Catherine Godson, Department of Medicine and Therapeutics, Mater Misericordiae Hospital, 44 Eccles Street, Dub- mediated effects reported to date are mediated through lin 7, Ireland. engagement of specific G protein–coupled lipoxin recep- E-mail: [email protected] tors [5]. LXA4 also modulates activation and transactiva- tion of growth factor receptors [13]. Our data do not REFERENCES exclude a role for these events at different time points 1. Brady HR, Brenner BM, Clarkson M, et al: Acute renal failure, throughout the course of ischemia/reperfusion injury. in The Kidney,6th ed, edited by Brenner BM, Philadelphia, W.B. It was striking in the present study that the lipoxin Saunders, 2000, pp. 1201–1262 2. Thadhani R, Pascual M, Bonventre JV: Acute renal failure. analog prevented ischemia/reperfusion injury–induced N Engl J Med 334:1448–1460, 1996 changes in mRNA levels for most genes and was not 3. Ashworth SL, Molitoris BA: Pathophysiology and functional limited to a specific family or families of genes. This obser- significance of apical membrane disruption during ischemia. Curr vation suggests that the lipoxin analog modulated a single Opin Nephrol Hypertens 8:449–458, 1999 4. Serhan CN: Lipoxins and novel aspirin-triggered 15-epi-lipoxins pathophysiologic event at the core of ATN and related (ATL): A jungle of cell-cell interactions or a therapeutic opportu- downstream processes or modulated a number of impor- nity? Prostaglandins 53:107–137, 1997 tant events in parallel. The latter would appear most likely, 5. McMahon B, Mitchell S, Brady HR, et al: Lipoxins: Revelations on resolution. Trends Pharmacol Sci 22:391–395, 2001 given the compelling evidence from in vitro systems that 6. Badr KF, DeBoer DK, Schwartzberg M, et al: Lipoxin A4 antago- lipoxin can influence a variety of pathobiologic functions nizes cellular and in vivo actions of leukotriene D4 in rat glomerular that are relevant to ischemic ATN, including vascular mesangial cells: Evidence for competition at a common receptor. Proc Natl Acad Sci USA 86:3438–3442, 1989 tone, epithelial cell injury, cytokine release, and leukocyte 7. Lee TH, Horton CE, Kyan-Aung U, et al: Lipoxin A4 and lipoxin recruitment and clearance (vide supra). B4 inhibit chemotactic responses of human neutrophils stimulated 492 Kieran and Doran et al: Transcriptomic response to renal ischemia-reperfusion injury

by leukotriene B4 and N-formyl-L-methionyl-L-leucyl-L-phenylal- 24. Madden SF, Lappin DW, Murphy M, et al: Computational gene anine. Clin Sci Lond 77:195–203, 1989 annotation in diabetic nephropathy. J Am Soc Nephrol 13:118A, 8. Papayianni A, Serhan CN, Brady HR: Lipoxin A4 and B4 inhibit 2002 leukotriene-stimulated interactions of human neutrophils and en- 25. Altschul SF, Madden TL, Schaffer AA, et al: Gapped BLAST dothelial cells. J Immunol 156:2264–2272, 1996 and PSI-BLAST: A new generation of protein database search 9. Colgan SP, Serhan CN, Parkos CA, et al: Lipoxin A4 modulates programs. Nucleic Acids Res 25:3389–3402, 1997 transmigration of human neutrophils across intestinal epithelial 26. Burge C, Karlin S: Prediction of complete gene structures in monolayers. J Clin Invest 92:75–82, 1993 human genomic DNA. J Mol Biol 268:78–94, 1997 10. Gronert K, Gewirtz A, Madara JL, et al: Identification of a 27. Bujnicki JM, Elofsson A, Fischer D, Rychlewski L: Structure human enterocyte lipoxin A4 receptor that is regulated by interleu- prediction meta server. Bioinformatics 17:750–751, 2001 kin (IL)-13 and interferon gamma and inhibits tumor necrosis 28. Lo Conte L, Ailey B, Hubbard TJ, et al: SCOP: A structural factor alpha-induced IL-8 release. J Exp Med 187:1285–1294, 1998 classification of proteins database. Nucleic Acids Res 28:257–259, 11. Goh J, Baird AW, O’Keane C, et al: Lipoxin A(4) and aspirin- 2000 triggered 15-epi-lipoxin A(4) antagonize TNF-alpha-stimulated 29. Megyesi J, Andrade L, Vieira JM, Jr, et al: Coordination of the neutrophil-enterocyte interactions in vitro and attenuate TNF- cell cycle is an important determinant of the syndrome of acute alpha-induced chemokine release and colonocyte apoptosis in hu- renal failure. Am J Physiol Renal Physiol 283:F810–F816, 2002 man intestinal mucosa ex vivo. J Immunol 167:2772–2780, 2001 30. Ichimura T, Bonventre JV, Bailly V, et al: Kidney injury mole- 12. Godson C, Mitchell S, Harvey K, et al: Cutting edge: Lipoxins cule-1 (KIM-1), a putative epithelial cell adhesion molecule con- stimulate nonphlogist phagocytosis of apoptotic neutrophils by taining a novel immunoglobulin domain, is up-regulated in renal monocyte-derived macrophages. J Immunol 164:1663–1667, 2000 cells after injury. J Biol Chem 273:4135–4142, 1998 13. McMahon B, Mitchell D, Shattock R, et al: Lipoxin, leukotriene, 31. Yoshida T, Tang SS, Hsiao LL, et al: Global analysis of gene and PDGF receptors cross-talk to regulate mesangial cell prolifera- expression in renal ischemia-reperfusion in the mouse. Biochem tion. FASEB J 16:1817–1819, 2002 Biophys Res Commun 291:787–794, 2002 14. Serhan CN, Maddox JF, Petasis NA, et al: Design of lipoxin A4 32. Tsukita S, Furuse M, Itoh M: Multifunctional strands in tight stable analogs that block transmigration and adhesion of human junctions. Nat Rev Mol Cell Biol 2:285–293, 2001 neutrophils. Biochemistry 34:14609–14615, 1995 33. Simon DB, Lu Y, Choate KA, et al: Paracellin-1, a renal tight 15. Takano T, Clish CB, Gronert K, et al: Neutrophil-mediated junction protein required for paracellular Mg2ϩ resorption. Sci- changes in vascular permeability are inhibited by topical applica- ence 285:103–106, 1999 tion of aspirin-triggered 15-epi-lipoxin A4 and novel lipoxin B4 34. Kwon O, Nelson WJ, Sibley R, et al: Backleak, tight junctions, stable analogues. J Clin Invest 101:819–826, 1998 and cell- cell adhesion in postischemic injury to the renal allograft. 16. Clish CB, O’Brien JA, Gronert K, et al: Local and systemic J Clin Invest 101:2054–2064, 1998 delivery of a stable aspirin-triggered lipoxin prevents neutrophil 35. Miyamori H, Takino T, Kobayashi Y, et al: Claudin promotes recruitment in vivo. Proc Natl Acad Sci USA 96:8247–8252, 1999 activation of pro-matrix metalloproteinase-2 mediated by mem- 17. Hachicha M, Pouliot M, Petasis NA, Serhan CN: Lipoxin brane-type matrix metalloproteinases. J Biol Chem 276:28204– (LX)A4 and aspirin-triggered 15-epi-LXA4 inhibit tumor necrosis 28211, 2001 factor 1alpha-initiated neutrophil responses and trafficking: Regu- 36. Wolfsberg TG, Primakoff P, Myles DG, White JM: ADAM, a lators of a cytokine-chemokine axis. J Exp Med 189:1923–1930, novel family of membrane proteins containing a disintegrin and 1999 metalloprotease domain: Multipotential functions in cell-cell and 18. Papayianni A, Serhan CN, Phillips ML, et al: Transcellular bio- cell-matrix interactions. J Cell Biol 131:275–278, 1995 synthesis of lipoxin A4 during adhesion of platelets and neutrophils 37. Bohm BB, Aigner T, Gehrsitz A, et al: Up-regulation of MDC15 in experimental immune complex glomerulonephritis. Kidney Int (metargidin) messenger RNA in human osteoarthritic cartilage. 47:1295–1302, 1995 Arthritis Rheum 42:1946–1950, 1999 19. Munger KA, Montero A, Fukunaga M, et al: Transfection of rat 38. Safirstein R, Price PM, Saggi SJ, Harris RC: Changes in gene kidney with 15-lipoxygenase suppresses inflammation and pre- expression after temporary renal ischemia. Kidney Int 37:1515– serves function in experimental glomerulonephritis. Proc Natl 1521, 1990 Acad Sci USA 96:13375–13380, 1999 39. Bond JS, Rojas K, Overhauser J, et al: The structural genes, 20. Chiang N, Gronert K, Clish CB, et al: Leukotriene B4 receptor MEP1A and MEP1B, for the alpha and beta subunits of the metal- transgenic mice reveal novel protective roles for lipoxins and aspi- loendopeptidase meprin map to human chromosomes 6p and 18q, rin-triggered lipoxins in reperfusion. J Clin Invest 104:309–316, respectively. Genomics 25:300–303, 1995 1999 40. Bertenshaw GP, Turk BE, Hubbard SJ, et al: Marked differences 21. Leonard MO, Hannan K, Burne MJ, et al: 15-epi-16-(para-fluor- between metalloproteases meprin A and B in substrate and peptide ophenoxy)-lipoxin A(4)-methyl ester, a synthetic analogue of 15- bond specificity. J Biol Chem 276:13248–13255, 2001 epi-lipoxin A(4), is protective in experimental ischemic acute renal 41. Carmago S, Shah SV, Walker PD: Meprin, a brush-border en- failure. J Am Soc Nephrol 13:1657–1662, 2002 zyme, plays an important role in hypoxic/ischemic acute renal 22. Imgrund M, Grone E, Grone HJ, et al: Re-expression of the tubular injury in rats. Kidney Int 61:959–966, 2002 developmental gene Pax-2 during experimental acute tubular ne- 42. Ricardo SD, Bond JS, Johnson GD, et al: Expression of subunits crosis in mice. Kidney Int 56:1423–1431, 1999 of the metalloendopeptidase meprin in renal cortex in experimen- 23. Cohen CD, Frach K, Schlondorff D, Kretzler M: Quantitative tal hydronephrosis. Am J Physiol 270:F669–F676, 1996 gene expression analysis in renal biopsies: A novel protocol for a 43. Guy PM, Kenny DA, Gill GN: The PDZ domain of the LIM high-throughput multicenter application. Kidney Int 61:133–140, protein enigma binds to beta-tropomyosin. Molec Biol Cell 10: 2002 1973–1984, 1999