- Email: [email protected]
Monitoring changes in gene expression in renal ischemia-reperfusion in the rat
Kidney International, Vol. 61 (2002), pp. 1646–1654
CELL BIOLOGY – IMMUNOLOGY – PATHOLOGY
Monitoring changes in gene expression in renal ischemia-reperfusion in the rat TAKUMI YOSHIDA, MANJULA KURELLA, FRANCISCA BEATO, HYUNSUK MIN, JULIE R. INGELFINGER, ROBIN L. STEARS, RITA D. SWINFORD, STEVEN R. GULLANS, and SHIOW-SHIH TANG Pediatric Nephrology Unit, Massachusetts General Hospital, Renal Division, Department of Medicine, Brigham & Women’s Hospital, Harvard Medical School, Boston, Massachusetts, Division of Pediatric Nephrology Unit, University of Texas Medical Center in Houston, Texas, USA
Monitoring changes in gene expression in renal ischemia-reperfusion in the rat. Background. Although acute renal failure (ARF) is a relatively common disorder with major morbidity and mortality, its molecular basis remains incompletely defined. The present study examined global gene expression in the well-characterized ischemia-reperfusion model of ARF using DNA microarray technology. Methods. Male Wistar rats underwent bilateral renal ischemia (30 min) or sham operation, followed by reperfusion for 1, 2, 3 or 4 days. Plasma creatinine increased approximately fivefold over baseline, peaking on day 1. Renal total RNA was used to probe cDNA microarrays. Results. Alterations in expression of 18 genes were identified by microarray analysis. Nine genes were up-regulated (ADAM2, HO-1, UCP-2, and thymosin 4 in the early phase and clusterin, vanin1, fibronectin, heat-responsive protein 12 and FK506 binding protein in the established phase), whereas another nine were down-regulated (glutamine synthetase, cytochrome p450 IId6, and cyp 2d9 in the early phase and cyp 4a14, Xist gene, PPAR␥, ␣-albumin, uromodulin, and ADH B2 in the established phase). The identities of these 18 genes were sequenceverified. Changes in gene expression of ADAM2, cyp2d6, fibronectin, HO-1 and PPAR␥ were confirmed by quantitative real-time polymerase chain reaction (PCR). ADAM2, cyp2d6, and PPAR␥ have not previously been known to be involved in ARF. Conclusion. Using DNA microarray technology, we identified changes in expression of 18 genes during renal ischemiareperfusion injury in the rat. We confirmed changes in five genes (fibronectin, ADAM2, cyp 2d6, HO-1 and PPAR␥) by quantitative real-time PCR. Several genes, not previously been identified as playing a role in ischemic ARF, may have importance in this disease.
Key words: microarray, ischemia-reperfusion, gene expression. Received for publication August 16, 2001 and in revised form October 23, 2001 Accepted for publication November 19, 2001
2002 by the International Society of Nephrology
Acute renal failure (ARF), characterized by rapid decline in glomerular filtration rate (GFR), is potentially reversible but is associated with major morbidity and mortality. Despite new insights into the pathogenesis of ARF, neither the incidence nor mortality has declined in decades [1]. During the induction and course of ARF many intrarenal alterations occur, including changes in cell polarity [2–5], levels of signaling molecules [6–11] expression of adhesion molecules [8, 12–15] and patterns of apoptosis and necrosis [16]. Nevertheless, it is probable that additional factors participate in the evolution and repair of ARF but are yet to be appreciated. Discovering such factors may lead to improved strategies for preventing ARF and treating this serious disorder. Certain changes seen in the course of ARF involve infiltrating cells and adhesion molecules. Neutrophil infiltration is induced by adhesion molecules such as intercellular adhesion molecule-1 (ICAM-1) and E-selectin via the mediation of tumor necrosis factor-␣ (TNF␣) and nuclear factor-B (NF-B). Infiltrating neutrophils induce adhesion molecules such as integrins [12], E-selectin [13], P-selectin [14], intercellular adhesion molecule1 [8], and neural cell adhesion molecule [15]. Inhibition of NF-B using a decoy oligonucleotide has been reported to suppress the inflammatory response that follows ischemia-reperfusion injury [17]. Animals that express aberrant adhesion molecules may be protected from renal injury, as noted in ICAM-1–deficient mice that are resistant to ischemic renal injury [18]. Depolarization of epithelial cells occurs in the course of renal injury and repair, in part secondary to changes in adhesion molecules. For example, in ARF induction of transforming growth factor- (TGF-), deactivation of integrin, and increased expression of adhesion molecules is associated with disruption of the cytoskeleton in proximal tubular cells [19]. In this setting the actin microfilament network becomes redistributed into large
1646
Yoshida et al: ARF and DNA microarray
cytoplasmic aggregates [19], which leads to depolarization and exfoliation of epithelial cells. Newly expressed integrin appears to restore cell polarization, accompanied by increased expression of laminin and fibronectin [19]. Ischemic injury results in apoptosis as well as necrosis. Apoptosis is mediated through several signaling cascades, one of which is the mitogen-activated protein (MAP) kinase cascade. Members of this family, such as p38MAPK [10], JNK and ERK [11] have been previously reported to play a role in ARF. Park, Chen and Bonventre suggested that activation of JNK or p38 is involved in protection against ischemic injury in the kidney [11]. TNF-␣ and integrin also have been reported to mediate apoptosis [20, 21]. DNA microarray technology offers the advantage of analyzing thousands of genes simultaneously, with the potential to uncover new mechanisms of pathogenesis. Thus, microarray analysis coupled with bioinformation could detect changes in genes previously unknown to participate in renal injury and repair. Such an approach could lead, potentially, to a new paradigm through which ARF may be better understood. In the present study, we examined changes in gene expression induced by ischemia-reperfusion injury. A total of 2,100 mouse cDNAs spotted onto a glass slide were used for profiling gene expression. As expected, we found that many of the changes in expression occur in genes that encode proteins previously implicated in ischemia-reperfusion injury. However, and importantly, ADAM2 (a disintegrin domain and metalloproteinase), cytochrome p450 IID6, peroxisome proliferator-activated receptor gamma (PPAR␥) were newly identified to participate in ARF. METHODS Animals and RNA extraction Male Wistar rats (200 to 250 g, N ⫽ 16 ischemiareperfusion, N ⫽ 20 sham) were housed in a 12-hour light/dark cycle, and allowed free access to food and water. Control and treated animals were weight- and age-matched at the time of initiation of bilateral ischemic injury, and body weights were recorded at initiation and completion of experiments. The animals were allowed free access to standard rodent chow and water until overnight prior to surgery when food (but not water) was removed. Animals were anesthetized intraperitoneally with pentobarbital sodium, 65 mg/kg. Both renal arteries were exposed using a midline incision and then clamped for 30 minutes followed by reperfusion. Sham animals had an incision plus 30 minutes of waiting time without clamping. After ischemia or sham treatment, muscle layer incisions were sutured, and skin incisions closed with Michel clips. We gave a total of 6 mL of 0.9% saline solution IP to the rats during and after surgery to prevent dehydration. In the present study, we chose to analyze
1647
Fig. 1. Serum creatinine in sham-operated (䊊) and ischemia-reperfusion (䊉) rats at various times. Creatinine rose to a peak at 24 hours in the ischemia-reperfusion rats. Data are expressed as mean ⫾ SEM for 4 rats.
RNA from animals that showed more prominent creatinine changes. Animals were then followed for 1, 2, 3, or 4 days (at least N ⫽ 3 per group per time point). At sacrifice the kidneys were removed and snap-frozen in liquid nitrogen. The samples were stored at ⫺70⬚C until further study. The serum creatinine level over time is shown in Figure 1, and documents successful induction of ARF. Total RNA was isolated from kidney by Trizol (Gibco, Rockville, MD, USA) extraction. High-density DNA microarray preparation High-density cDNA microarrays were generated by standard methods [22]: Mouse expressed sequence tags (ESTs; Research Genetics, Huntsville, AL, USA) were amplified by the polymerase chain reaction (PCR), according to the manufacturer’s recommendations. Each sample was run in a 1% agarose gel for quality verification (data not shown). The sizes of PCR products ranged from approximately 300 bp to 1.2 kb. DNA microarrays, containing 2,100 distinct ESTs, were printed onto glass slides using a pin-style arrayer (BioRobotics, Cambridge, UK). The identities of all 18 genes were confirmed by DNA sequencing using an ABI PRISM 377 DNA sequencer (Applied Biosystems, Foster City, CA, USA) performed by the Genome Center staff at Brigham & Women’s Hospital. The sequence was confirmed using BLAST (http://www.ncbi.nlm.nih.gov/BLAST/). To make microarrays, sialylated glass slides (CEL Associates Inc., Houston, TX, USA) were cleaned for two hours in a solution of 2 N NaOH. After rinsing in distilled water, the slides were treated with a 1:5 dilution of polyl-lysine adhesive solution (Sigma, St. Louis, MO, USA) for one hour, and then dried for five minutes at 80⬚C in a vacuum oven. DNA samples from 100-L PCR reactions were precipitated with ethanol in 96-well microtiter plates. The resulting precipitates were resuspended in 20 L 3 ⫻ standard sodium citrate (SSC; 20 ⫻
1648
Yoshida et al: ARF and DNA microarray
SSC; 3.0 mol/L NaCl and 0.3 mol/L sodium citrate) and transferred to new plates for arraying. After printing, the microarrays were rehydrated for 10 seconds in a humidity chamber. The DNA was then ultraviolet-crosslinked to the surface by subjecting the slides to 60 mJ of energy (Stratagene Stratalinker, La Jolla, CA, USA). Directly after the blocking reaction, the bound DNA was denatured by a two-minute incubation in distilled water at ⬃95⬚C. The slides were then transferred into a bath of ⫺20⬚C ethanol, rinsed, and then spun dry in a clinical centrifuge. Slides were stored in a closed box at room temperature until used. cDNA probe labeling cDNA probes were synthesized as previously reported [23] to improve signal detection on high-density microarrays using a fluorescent oligonucleotide dendrimeric signal amplification system. Briefly, 10 g of total RNA was primed with 1.0 picomole of modified reverse transcription (RT) primer with capture sequence for Cy3 or Cy5 (Genisphere, Montvale, NJ, USA). The RT reaction was performed according to standard protocols (SuperScript RT II; Gibco), followed by the addition of 3.5 L of 0.5 mol/L NaOH/50 mmol/L ethylenediaminetetraacetic acid (EDTA), and heating at 65⬚C for ten minutes to denature the DNA/RNA hybrids. This reaction was then neutralized with 5 L of 1 mol/L Tris, pH 7.5. For double-labeled detection, RT reactions were mixed and precipitated using 3 mol/L sodium acetate and 100% ethanol, then washed once with 70% ethanol. The pellets were resuspended in 20 L distilled H2O. Hybridization cDNA microarrays were prehybridized with 500 L of preheated (65⬚C) ExpressHyb (Clontech, Palo Alto, CA, USA) for an hour in a microarray hybridization chamber (Grace Bio-Labs, Bend, OR, USA) in a rotisserie hybridization oven. Probe was put into the hybridization chamber and incubated overnight in a rotisserie hybridization oven. The next day the microarrays were washed for five minutes in 2 ⫻ SSC/0.2% sodium dodecyl sulfate (SDS) at 65⬚C, then in 2 ⫻ SSC and 0.2 ⫻ SSC for two minutes each. For dendrimer detection, 1.0 L of the Cy3- and Cy5-labeled dendrimers (Genisphere) were suspended in 15 L of 4 ⫻ SSC/2% SDS and layered on the array. The arrays were covered with a coverslip and then incubated at 65⬚C for eight hours in a microarray hybridization cassette (Telechem International, Sunnyvale, CA, USA). The chips were then washed as above. Data collection, normalization, and analysis Each time point was analyzed twice with reverse color labeling. The mean of the two measurements was used to determine the fold changes. As was stated, dye reversal was used. This analysis uses a log-normal error model.
The cut off was based on at least twofold changes for the genes of interest. The chips were scanned in a ScanArray 5000 confocal laser scanner (GSI Lumonics, Marina Del Rey, CA, USA). The signal and background signals were then quantified with ImaGene 4.0 quantification software (BioDiscovery, Billerica, MA, USA). Signals were measured as the mean pixel intensity within each circumscribed spot and average background was measured using the mean pixel intensities of blank spots in the array (without cDNA) and as the mean pixel intensity outside the circumscribed spot diameter for calculating specific signals. Quantitative real time PCR Quantitation with real-time, one-step reverse-transcriptase polymerase chain reaction (RT-PCR) was performed with SYBR Green PCR Reagents and an ABI PRISM 7700 Sequence Detection System (PE Applied Biosystems, Foster City, CA, USA), according to the manufacturer’s instructions. Reactions were performed using 1.0 L of RNA at a concentration of 100 ng/L, in a reaction volume of 50 L. An RT reaction was performed at 48⬚C for 30 minutes, followed by PCR consisting of AmpliTaq activation for 10 minutes at 95⬚C, then 40 cycles with heating to 95⬚C for 15 seconds, and cooling to 60⬚C for one min. mRNA levels were normalized to levels of glyceraldehyde-3-phosphate dehydrogenase (GAPDH). We elected to confirm changes in five genes using real-time one-step quantitative RT-PCR. These five genes and PCR primers are as follows: Fibronectin. Forward primer, 5⬘-GTGGCTGCCTTCAACTTCTC-3⬘ Reverse primer, 5⬘-GTGGGTTGCAAACCTTCAAT-3⬘ HO-1. Forward primer, 5⬘-TCTATCGTGCTCGCATGAAC-3⬘ Reverse primer, 5⬘-CAGCTCCTCAAACAGCTCAA-3⬘ ADAM2. Forward primer, 5⬘-GGGATCTGTTTGGCACTGAT-3⬘ Reverse primer, 5⬘-ATTAGGTGCGCAGTCCTTGT-3⬘ PPAR␥. Forward primer, 5⬘-TGTGGACCTCTCTGTGATGG-3⬘ Reverse primer, 5⬘-CATTGGGTCAGCTCTTGTGA-3⬘ Cytochrome P450 IId6. Forward primer, 5⬘-AGCTTCAACACCGCTATGGT-3⬘ Reverse primer, 5⬘-CAGCAGTGTCCTCTCCATGA-3⬘ GAPDH. Forward primer, 5⬘-ACTCCCATTCCTCCACCTTT-3⬘ Reverse primer, 5⬘-TTACTCCTTGGAGGCCATGT-3⬘ All primer sets were designed according to manufacturer’s recommendations, and when feasible, to span one or more introns. All amplicons ranged between 142 and 148 base pairs. Electrophoretic analysis of expected product sizes was performed for all primer sets prior to one step, real time RT-PCR, to confirm the fidelity of the reaction.
Yoshida et al: ARF and DNA microarray
RESULTS cDNA microarray analysis Representative differential expression patterns are shown for sham (Fig. 2A) and ischemia-reperfusion samples (Fig. 2B) at day 1 on one glass slide. Ischemia-reperfusion (IR) samples were labeled with Cy3 and sham with Cy5. This figure demonstrates clear signals with little background and several instances of clear differential gene expression. Every spot is well rounded and has virtually the same diameter. Figure 2C shows an overlay picture of sham and IR such that green spots indicate up-regulation and red spots indicate down-regulation, and yellow spots indicate no change, respectively. The numbers 1, 2, 3, 4, and 5 indicate clusterin, ADAM2, HO-1, cyp2d6, PPAR␥, respectively. Of these, clusterin, ADAM2, and HO-1 are increased and cyp2d6 and PPAR␥ are decreased in IR. As shown in Figure 3A, a scatter plot of sham (Cy3) versus sham (Cy5) fell on a line of identity. When a scatter plot of sham (Cy5) versus IR (Cy3) was performed, some changes in individual genes were seen (Fig. 3B). Red spots, (identified as Adam2, HO-1, and clusterin) signify up-regulated genes in IR and green spots (identified as cyp2d6 and PPAR␥) signify down-regulated genes. To confirm these data, reverse labeling was performed. Instead of labeling IR with Cy3 and sham with Cy5, IR was labeled with Cy5 and sham with Cy3. In Figure 3C, a scatter plot of this reverse labeling is shown; now green spots (identified as Adam2, HO-1, and clusterin) signify genes decreased in IR and red spots (identified as cyp2d6 and PPAR␥) signify transcripts increased in IR. Thus, these reverse labeling experiments demonstrated consistent results. Furthermore, all 18 transcripts altered were sequence-verified to confirm their identities.
1649
days, consistent with other reports [12, 25]. ADAM2 was up-regulated by 3.3 fold and gradually normalized. Genes down-regulated by ischemia-reperfusion injury Figure 5 shows nine genes that were decreased in IR. Of these nine, four genes decreased acutely, including alpha-albumin protein, cytochrome P450 IID6 (cyp2d6), cytochrome P450 IId9 (cyp2d9), and glutamine synthetase (Fig. 5A). Cyp2d6 was markedly decreased, to approximately 20% of the sham value. Five genes were chronically down-regulated, including alcohol dehydrogenase-B2, cytochrome P450 IVa14 (cyp4a14), PPAR␥, Xist gene, and uromodulin (Fig. 5B). Uromodulin was decreased to approximately 20 to 25% of the sham value. Confirmation of change in expression Microarray data of five genes of interest were confirmed by real-time quantitative PCR (Fig. 6); others were not tested. Fibronectin was chosen because of its well-known increase in ischemic injury [12]. HO-1 was selected as increased on gene array and previously described to be increased in ARF [24, 26–28]. ADAM2 was chosen as a gene that was increased on gene array, and a previously undescribed finding in ARF. PPAR␥ and Cyp 2d6 were chosen as they were decreased on gene array, previously undescribed findings in ARF. Using real-time PCR, HO-1 and ADAM 2 were confirmed to be transitionally increased whereas cytochrome p450 IId6 and PPAR␥ were decreased in ischemia-reperfusion. Thus, there was excellent concordance between microarray data and real time PCR, though the magnitude of changes were more pronounced than those seen with microarray analysis.
Genes up-regulated by ischemia-reperfusion injury
DISCUSSION
Figure 4 shows the nine genes increased in IR. Of these, three increased acutely. These include heme oxygenase-1, thymosin 4, and uncoupling protein (UCP)-2 (Fig. 4A). Heme oxygenase-1 (HO-1) increased at day 1 followed by normalization, a finding previously reported in a rat model of ARF [24]. Thymosin 4 (Tb4) increased approximately twofold. Uncoupling protein (UCP)-2 mRNA level increased by 2.8-fold at day 1 and recovered to baseline at day 4. Six genes were elevated throughout the four-day time course of IR (Fig. 4B). These included ADAM2, clusterin, FK506 binding protein 1a (Fkbp1a), fibronectin, vanin1, and heat-responsive protein 12 (hrsp12). The expression of clusterin (also known as a TRPM-2 or sulfated glycoprotein 2) was increased approximately fivefold and maintained at that level until day 4. Fibronectin increased by about 2.5 fold and remained so for four
In the present study, we demonstrated that multiple gene transcripts were increased or decreased by ischemia-reperfusion, and we identified several genes previously unknown to be involved in renal injury and repair. For example, it has been previously reported that clusterin [16] and fibronectin [12, 29] are up-regulated by renal ischemia-reperfusion injury, and that HO-1 is up-regulated during ARF [24, 30, 31]. Uromodulin has previously been shown to decrease after renal ischemiareperfusion injury [32]. The transcript with the largest observed change was clusterin, known as a marker for cell injury in the tissues. In the present studies clusterin was increased from day 1 to day 4 after ischemic injury. Clusterin mRNA is overexpressed in apoptosis-resistant cells [33] and is required for survival of cells treated with TNF␣ or subjected to oxidative stress [34–36]. Thus, we infer from these find-
1650
Yoshida et al: ARF and DNA microarray
Fig. 2. Representative differential expression by cDNA array. Representative differential expression is shown by microarray of sham (A, Cy3) and ischemia-reperfusion (B, Cy5) at day 1 on one glass slide. (C ) Overlay picture of sham and IR microarray is shown. In this picture, each red spot indicates higher signal in sham and each green spot indicates higher in IR. Numbers denote: (1) clusterin, (2) ADAM2, (3) HO-1, (4) cyp2d6, (5) PPAR␥. Note that the signal of spots numbered 1, 2 and 3 are higher on the right compared to the left, indicating an increase. In contrast, the spots numbered 4 and 5 indicate a decrease in expression.
Fig. 3. Scatter plot graph of microarray data. (A) Sham labeled by Cy3 versus sham labeled by Cy5. All data points are on a 45⬚ angle, confirming the accuracy of this technique. (B) Scatter plot of sham labeled by Cy5 versus ischemia-reperfusion (IR) labeled by Cy3 at day 1 was shown. Red spots, (typed as ADAM2, HO-1, and clusterin) signify genes that are upregulated and green spots, (typed as cyp2d6 and PPAR␥) signify genes that are down-regulated. (C ) Scatter plot of IR labeled by Cy5 vs sham labeled by Cy3 as shown on day 1, as a reverse labeling of scatter plot B. Red spots, (typed as ADAM2, HO-1, and clusterin) signify up-regulated and green spots, (typed as cyp2d6 and PPAR␥), down-regulated. Thus, reverse labeling confirmed the data. Abbreviations are: ADAM2, a disintegrin domain and metalloproteinase; Clu, clusterin; HO1, heme oxygenase-1; Cyp2d6, cytochrome P450 2d6; PPAR␥, peroxisome proliferator activated receptor-gamma.
Yoshida et al: ARF and DNA microarray
Fig. 4. Time course of up-regulated genes. The relative fold change of ischemia-reperfusion/sham is shown in a time-dependent manner. Abbreviations are: IR, ischemia-reperfusion; ADAM2, a disintegrin domain and metalloproteinase; Clu, clusterin; Fkbp1a, FK506 binding protein 1a; Fn, fibronectin; HO1, heme oxygenase-1; Hrsp12, heatresponsive protein 12; Tmsb4, thymosin beta-4; Ucp2, UCP2; Vnn1, vanin 1. Symbols in A are: (䊊) HO1; (䊉) Tmsb4; (䊐) Ucp2. Symbols in B are: (䊏) Adam2; (䉱) Clu; (䉭) Fkbp1a; (䉬) FN; (䉫) Vnn1; (䊊) Hrsp12.
ings that clusterin may play role in protection from ischemia-induced apoptosis in kidney. In the present study, HO-1 increased at day 1, decreased gradually at day 2 and normalized by day 4. HO-1 has already been shown to be increased in the kidney in various models of ARF [24, 26–28]. Shimizu et al reported that HO-1 conferred a protective effect in ischemic acute renal failure, in that when HO activity was blocked by tin mesoporphyrin, increases in microsomal heme concentration, and serum creatinine concentration were observed, along with extensive tubular epithelial cell injury [31]. Uromodulin (Tamm-Horsfall protein) was greatly decreased in our model. This protein, one of the most abundant in the renal tubule, is exclusively localized to
1651
Fig. 5. Time course of down-regulated genes. Relative fold change of ischemia-reperfusion/sham is shown in time-dependent manner. Abbreviations are: IR; ischemia-reperfusion; Aalb, alpha-albumin protein; Adhb2, alcohol dehydrogenase-B2; Cyp2d6, cytochrome P450 2d6; Cyp2d9, cytochrome P450 2d9; Cyp4a14, cytochrome P450, 4a14 (Cyp4a14); PPAR␥, peroxisome proliferator activated receptor gamma; Glns-ps1, intronless glutamine synthetase gene; Xist, Xist gene, P0 region; Umod, uromodulin. Symbols in A are: (䊊) Adhb2; (䊉) Cyp4a14; (䊐) PPAR␥; (䊏) Xist; (䉭) Umod. Symbols in B are: (䉱) Aalb; (䉫) Cyp2d6; (䉬) Cyp2d9; (䊊) Glns-ps1.
the thick ascending limb of Henle’s loop [37]. Safirstein et al and Bachmann et al reported that uromodulin mRNA declined to undetectable levels by 24 hours after renal ischemia-reperfusion [32, 37]. We confirm their observation on day 1 after ischemia injury. In addition, our data show that uromodulin gene expression was continuously inhibited up to day 4 post-injury. The microarray data showed that ADAM2 was increased on day 1 after IR injury and gradually returned to normal on day 4. The real-time PCR confirmed that there is an increase of ADAM2 on day 1. However, the PCR results further showed no changes or decrease after day 2. We do not know the reasons for this discrepancy. Nevertheless, the PCR results did confirm that ADAM2 was transiently up-regulated on day 1. Novel observations included increased ADAM2 in ARF. This gene, also known as fertilin beta, is a well-characterized mole-
1652
Yoshida et al: ARF and DNA microarray
Fig. 6. Time course of relative fold change of mRNA by real-time quantitative polymerase chain reaction (PCR). Relative fold change of ischemia-reperfusion/sham is shown in time-dependent manner. Abbreviations are: IR, ischemia-reperfusion; ADAM2, a disintegrin domain and metalloproteinase; Cyp2d6, cytochrome P450 2d6; Fn, fibronectin; HO1, heme oxygenase-1; PPAR␥, peroxisome proliferator activated receptor gamma. Symbols are: (䊏) Adam2; (䊐) Cyp2d6; (䉬) FN; (䊊) HO-1; (䊐) PPAR␥. The values are expressed by log 2. Data are expressed as mean ⫾ SEM; N ⫽ 4.
cule that mediates cell-cell communication in sperm-egg interactions. ADAM2 has not previously been known to be involved in ARF. Interestingly, this protein increases in the central nervous system in cases of Alzheimer’s disease, along with alteration in cell-matrix contact in the brain [38]. We thus infer that this molecule plays a role in cell-matrix interactions within the kidney. Several members of the ADAM family have the ability to cleave transmembrane proteins, such as growth factors, as well as their receptors, cell adhesion molecules, and cytokines, resulting in soluble forms of those substances, a phenomenon that is termed ectodomain shedding [39]. For example, ADAM-17 (TNF␣ converting enzyme, TACE), a member of this family, cleaves transmembrane proTNF␣ to produce mature form of TNF␣. Tissue inhibitor of metalloproteases-3 (TIMP-3) inhibits these proteins that are involved in ectodomain shedding [40–43]. It may be that ADAM2 plays a role as a shedder, during the regeneration needed during tissue repair after renal damage. Thus, we speculate that ADAM2 plays an important role in repair after tissue injury with in the kidney. We also noted that thymosin 4 (Tb4) was increased in ischemia-reperfusion injury. The beta-thymosins constitute a family of actin monomer-sequestering proteins widely distributed among vertebrates, and Tb4 is most abundant protein in this family. Tb4 has been reported to bind to actin monomers and to enhance wound healing [44–46]. Additionally, it is up-regulated following focal brain ischemia [47]. Our findings lead us to hypothesize that Tb4 has a general role in repair of cytoskeleton and
in re-polarization of renal tubular epithelial cells after IR injury. UCP-2 was increased by IR injury, a previously unreported finding. This gene is involved in mitochondrial hydrogen peroxide generation [48]. UCP-2 mRNA has been reported to be up-regulated approximately threefold in fatty liver, during which hydrogen peroxide production by mitochondria was increased. It has been speculated that this increase may result from chronic apoptotic stress and may play a role in adaptation to such stress [49]. Li et al reported that UCP-2 is involved in cellular defense against oxidative stress (hydrogen peroxide 200 mmol/L) in clonal beta-cells that overexpress UCP-2 [50], further supporting this theory. We speculate that UCP-2 plays an analogous role in renal injury. Vanin-1 increased by about 2.7-fold and remained elevated throughout the study period. Vanin-1, also called pantetheinase, hydrolyzes pantethein to cysteamine (a powerful antioxidant) and pantothenic acid [51]. Therefore, we hypothesize that vanin-1 acts as an endogenous anti-oxidant in renal repair after ischemia-reperfusion injury [52]. We noted PPAR␥ to be down-regulated in ischemiareperfusion injury, a novel finding. PPARs (peroxisome proliferator-activated receptors), a family of transcription factors, have multiple roles that are of great interest. Although etomoxir, a PPAR␣ agonist, was reported to protect the kidney in ischemia-reperfusion injury [53], to our knowledge the present work is the first to find changes in PPAR␥ in this model. PPAR␥ is a nuclear receptor that serves as a transcription factor and thus regulates several physiologic pathways, particularly those involved in lipid metabolism [54]. Fahmi reported that PPAR␥ inhibited interleukin-1 induced matrix metalloproteinase 13 [54]. Furthermore, Yee et al reported that the stromelysin-1 (MMP-3) gene, which is involved in tissue remodeling, contains a PPAR element in its promotor [55]. Therefore, it seems likely that PPAR␥ plays a role in tissue repair, especially in control of MMP. Nakajima et al observed that heterozygous PPAR␥-deficient mice subjected to intestinal ischemia suffered more pronounced injury than control mice [56]. Furthermore, when such mice were treated with a PPAR␥-activating ligand, BRL-49653, they were partially reconstituted and suffered less damage. This report supports our hypothesis that PPAR␥ may act through one of its pathways related to apoptosis and/or tissue repair after ischemic injury. Cytochrome p450 IId6, IId9, and IVa14, all of which are involved with oxidation, were suppressed. When these enzymes are active, they lead to consumption of oxygen. Hence, when tissue is subjected to severe ischemic injury, these enzymes are likely suppressed in an attempt to minimize ischemic injury. We speculate that these enzymes are suppressed during ischemic renal injury in an
Yoshida et al: ARF and DNA microarray
attempt to minimize oxidation injury. Suppression of these cytochrome p450s may occur during the ischemic state to slow down oxidative metabolism, which would lead to cytoprotection. The present analysis of the global gene expression profile in ischemia-reperfusion used whole kidney. Thus, the expression pattern within various intrarenal structures is not distinguishable from the whole. It also is possible that certain of the observations could be the result of inflammatory cell infiltration or injury to the microvasculature. Nonetheless, given the early stage and high cost of array technology, we chose not to focus on dissected cell types (such as, glomerulus, proximal tubule, thick ascending loop of the limb of Henle, distal tubule, collecting duct, etc.) in an initial study of renal ischemiareperfusion. Once routine amplicfication technology is established, pursuing such a goal will be more feasible. In conclusion, using DNA microarray we have identified 18 genes that in mRNA expression lead in the course of ischemia-reperfusion injury. The identities of these 18 genes were confirmed by DNA sequencing. Changes in five genes (fibronectin, ADAM2, cytochrome p450 IId6, HO-1 and PPAR␥) were further confirmed by quantitative real-time PCR. Several genes that have not previously been identified in ischemic ARF may have important implications concerning the evolution of this disease.
9. 10. 11.
12. 13. 14. 15. 16.
17.
18.
19.
ACKNOWLEDGMENTS This study was supported by National Institutes of Health Grants DK 50836 (to S.S. Tang), HL 48455 and DK 58950 (to J.R. Ingelfinger), DK 36031, DK 51606, and DK58849 (to S.R. Gullans). Microarray experiments were performed in the BWH NIDDK DNA microarray Biotechnology Center. We thank Roderick Jensen for valuable discussions.
20.
Reprint requests to Shiow-Shih Tang, Ph.D., Pediatric Nephrology Unit, Massachusetts General Hospital, 55 Fruit Street, BHX 410, Boston, Massachusetts 02114, USA. E-mail: [email protected]
22.
REFERENCES
24.
1. Thadhani R, Pascual M, Bonventre JV: Acute renal failure. N Engl J Med 334:1448–1460, 1996 2. Molitoris BA, Wagner MC: Surface membrane polarity of proximal tubular cells: Alterations as a basis for malfunction. Kidney Int 49:1592–1597, 1996 3. Molitoris BA, Chan LK, Shapiro JI, et al: Loss of epithelial polarity: A novel hypothesis for reduced proximal tubule Na⫹ transport following ischemic injury. J Membr Biol 107:119–127, 1989 4. Molitoris BA, Hoilien CA, Dahl R, et al: Characterization of ischemia-induced loss of epithelial polarity. J Membr Biol 106:233– 242, 1988 5. Molitoris BA, Wilson PD, Schrier RW, et al: Ischemia induces partial loss of surface membrane polarity and accumulation of putative calcium ionophores. J Clin Invest 76:2097–2105, 1985 6. Raman N, Atkinson SJ: Rho controls actin cytoskeletal assembly in renal epithelial cells during ATP depletion and recovery. Am J Physiol 276:C1312–C1324, 1999 7. Donnahoo KK, Meldrum DR, Shenkar R, et al: Early renal ischemia, with or without reperfusion, activates NFkappaB and increases TNF-alpha bioactivity in the kidney. J Urol 163:1328– 1332, 2000 8. Okusa MD, Linden J, Huang L, et al: A(2A) adenosine receptor-
21.
23.
25.
26.
27.
28.
29. 30.
31.
1653
mediated inhibition of renal injury and neutrophil adhesion. Am J Physiol (Renal Physiol) 279:F809–F818, 2000 Sakai M, Tsukada T, Harris RC: Oxidant stress activates AP-1 and heparin-binding epidermal growth factor-like growth factor transcription in renal epithelial cells. Exp Nephrol 9:28–39, 2001 Faccio L, Chen A, Fusco C, et al: Mxi2, a splice variant of p38 stress-activated kinase, is a distal nephron protein regulated with kidney ischemia. Am J Physiol (Cell Physiol) 278:C781–C790, 2000 Park KM, Chen A, Bonventre JV: Prevention of kidney ischemia/ reperfusion-induced functional injury and JNK, p38, and MAPK kinase activation by remote ischemic pretreatment. J Biol Chem 276:11870–11876, 2001 Zuk A, Bonventre JV, Brown D, et al: Polarity, integrin, and extracellular matrix dynamics in the postischemic rat kidney. Am J Physiol (Cell Physiol) 275:C711–C731, 1998 Singbartl K, Ley K: Protection from ischemia-reperfusion induced severe acute renal failure by blocking E-selectin. Crit Care Med 28:2507–2514, 2000 Singbartl K, Green SA, Ley K: Blocking P-selectin protects from ischemia/reperfusion-induced acute renal failure. FASEB J 14:48– 54, 2000 Abbate M, Brown D, Bonventre JV: Expression of NCAM recapitulates tubulogenic development in kidneys recovering from acute ischemia. Am J Physiol 277:F454–F463, 1999 Schumer M, Colombel MC, Sawczuk IS, et al: Morphologic, biochemical, and molecular evidence of apoptosis during the reperfusion phase after brief periods of renal ischemia. Am J Pathol 140: 831–838, 1992 Vos IH, Govers R, Grone HJ, et al: NFkappaB decoy oligodeoxynucleotides reduce monocyte infiltration in renal allografts. FASEB J 14:815–822, 2000 Kelly KJ, Williams WW, Jr, Colvin RB, et al: Intercellular adhesion molecule-1-deficient mice are protected against ischemic renal injury. J Clin Invest 97:1056–1063, 1996 Kellerman PS, Clark RA, Hoilien CA, et al: Role of microfilaments in maintenance of proximal tubule structural and functional integrity. Am J Physiol 259:F279–F285, 1990 Daemen MA, van de Ven MW, Heineman E, et al: Involvement of endogenous interleukin-10 and tumor necrosis factoralpha in renal ischemia-reperfusion injury. Transplantation 67:792– 800, 1999 Kumar CC: Signaling by integrin receptors. Oncogene 17:1365– 1373, 1998 DeRisi JL, Iyer VR, Brown PO: Exploring the metabolic and genetic control of gene expression on a genomic scale. Science 278:680–686, 1997 Stears RL, Getts RC, Gullans SR: A novel, sensitive detection system for high-density microarrays using dendrimer technology. Physiol Genomics 3:93–99, 2000 Maines MD, Mayer RD, Ewing JF, et al: Induction of kidney heme oxygenase-1 (HSP32) mRNA and protein by ischemia/reperfusion: possible role of heme as both promotor of tissue damage and regulator of HSP32. J Pharmacol Exp Ther 264:457–462, 1993 Walker PD: Alterations in renal tubular extracellular matrix components after ischemia-reperfusion injury to the kidney. Lab Invest 70:339–345, 1994 Nath KA, Balla G, Vercellotti GM, et al: Induction of heme oxygenase is a rapid, protective response in rhabdomyolysis in the rat. J Clin Invest 90:267–270, 1992 Agarwal A, Balla J, Alam J, et al: Induction of heme oxygenase in toxic renal injury: A protective role in cisplatin nephrotoxicity in the rat. Kidney Int 48:1298–1307, 1995 Vogt BA, Shanley TP, Croatt A, et al: Glomerular inflammation induces resistance to tubular injury in the rat. A novel form of acquired, heme oxygenase-dependent resistance to renal injury. J Clin Invest 98:2139–2145, 1996 Rosenberg ME, Paller MS: Differential gene expression in the recovery from ischemic renal injury. Kidney Int 39:1156–1161, 1991 Horikawa S, Ito K, Ikeda S, et al: Induction of heme oxygenase-1 in toxic renal injury: Mercuric chloride-induced acute renal failure in rat. Toxicol Lett 94:57–64, 1998 Shimizu H, Takahashi T, Suzuki T, et al: Protective effect of
1654
32. 33. 34.
35. 36. 37. 38. 39. 40. 41.
42.
43.
Yoshida et al: ARF and DNA microarray
heme oxygenase induction in ischemic acute renal failure. Crit Care Med 28:809–817, 2000 Safirstein R, Megyesi J, Saggi SJ, et al: Expression of cytokinelike genes JE and KC is increased during renal ischemia. Am J Physiol 261:F1095–F1101, 1991 Koch-Brandt C, Morgans C: Clusterin: A role in cell survival in the face of apoptosis? Prog Mol Subcell Biol 16:130–149, 1996 Sintich SM, Steinberg J, Kozlowski JM, et al: Cytotoxic sensitivity to tumor necrosis factor-alpha in PC3 and LNCaP prostatic cancer cells is regulated by extracellular levels of SGP-2 (clusterin). Prostate 39:87–93, 1999 Humphreys D, Hochgrebe TT, Easterbrook-Smith SB, et al: Effects of clusterin overexpression on TNFalpha- and TGFbetamediated death of L929 cells. Biochemistry 36:15233–15243, 1997 Viard I, Wehrli P, Jornot L, et al: Clusterin gene expression mediates resistance to apoptotic cell death induced by heat shock and oxidative stress. J Invest Dermatol 112:290–296, 1999 Bachmann S, Metzger R, Bunnemann B: Tamm-Horsfall proteinmRNA synthesis is localized to the thick ascending limb of Henle’s loop in rat kidney. Histochemistry 94:517–523, 1990 Gerst JL, Raina AK, Pirim I, et al: Altered cell-matrix associated ADAM proteins in Alzheimer’s disease. J Neurosci Res 59:680–684, 2000 Hooper NM, Karran EH, Turner AJ: Membrane protein secretases. Biochem J 321:265–279, 1997 Amour A, Slocombe PM, Webster A, et al: TNF-alpha converting enzyme (TACE) is inhibited by TIMP-3. FEBS Lett 435:39– 44, 1998 Hargreaves PG, Wang F, Antcliff J, et al: Human myeloma cells shed the interleukin-6 receptor: Inhibition by tissue inhibitor of metalloproteinase-3 and a hydroxamate-based metalloproteinase inhibitor. Br J Haematol 101:694–702, 1998 Smith MR, Kung H, Durum SK, et al: TIMP-3 induces cell death by stabilizing TNF-alpha receptors on the surface of human colon carcinoma cells. Cytokine 9:770–780, 1997 Borland G, Murphy G, Ager A: Tissue inhibitor of metalloproteinases-3 inhibits shedding of L-selectin from leukocytes. J Biol Chem 274:2810–2815, 1999
44. De La Cruz EM, Ostap EM, Brundage RA, et al: Thymosinbeta(4) changes the conformation and dynamics of actin monomers. Biophys J 78:2516–2527, 2000 45. Malinda KM, Sidhu GS, Mani H, et al: Thymosin beta4 accelerates wound healing. J Invest Dermatol 113:364–368, 1999 46. Young JD, Lawrence AJ, MacLean AG, et al: Thymosin beta 4 sulfoxide is an anti-inflammatory agent generated by monocytes in the presence of glucocorticoids. Nat Med 5:1424–1427, 1999 47. Vartiainen N, Pyykonen I, Hokfelt T, et al: Induction of thymosin beta(4) mRNA following focal brain ischemia. Neuroreport 7:1613– 1616, 1996 48. Negre-Salvayre A, Hirtz C, Carrera G, et al: A role for uncoupling protein-2 as a regulator of mitochondrial hydrogen peroxide generation. FASEB J 11:809–815, 1997 49. Yang S, Zhu H, Li Y, et al: Mitochondrial adaptations to obesityrelated oxidant stress. Arch Biochem Biophys 378:259–268, 2000 50. Li LX, Skorpen F, Egeberg K, et al: Uncoupling protein-2 participates in cellular defense against oxidative stress in clonal betacells. Biochem Biophys Res Commun 282:273–277, 2001 51. Pitari G, Malergue F, Martin F, et al: Pantetheinase activity of membrane-bound Vanin-1: Lack of free cysteamine in tissues of Vanin-1 deficient mice. FEBS Lett 483:149–154, 2000 52. Frankel D, Mehindate K, Schipper HM: Role of heme oxygenase-1 in the regulation of manganese superoxide dismutase gene expression in oxidatively-challenged astroglia. J Cell Physiol 185:80–86, 2000 53. Portilla D, Dai G, Peters JM, et al: Etomoxir-induced PPARalpha-modulated enzymes protect during acute renal failure. Am J Physiol (Renal Physiol) 278:F667–F675, 2000 54. Fahmi H, Di Battista JA, Pelletier JP, et al: Peroxisome proliferator–activated receptor gamma activators inhibit interleukin1beta-induced nitric oxide and matrix metalloproteinase 13 production in human chondrocytes. Arthritis Rheum 44:595–607, 2001 55. Yee J, Kuncio GS, Bhandari B, et al: Identification of promoter activity and differential expression of transcripts encoding the murine stromelysin-1 gene in renal cells. Kidney Int 52:120–129, 1997 56. Nakajima A, Wada K, Miki H, et al: Endogenous PPAR gamma mediates anti-inflammatory activity in murine ischemia-reperfusion injury. Gastroenterol 120:460–469, 2001
CELL BIOLOGY – IMMUNOLOGY – PATHOLOGY
Monitoring changes in gene expression in renal ischemia-reperfusion in the rat TAKUMI YOSHIDA, MANJULA KURELLA, FRANCISCA BEATO, HYUNSUK MIN, JULIE R. INGELFINGER, ROBIN L. STEARS, RITA D. SWINFORD, STEVEN R. GULLANS, and SHIOW-SHIH TANG Pediatric Nephrology Unit, Massachusetts General Hospital, Renal Division, Department of Medicine, Brigham & Women’s Hospital, Harvard Medical School, Boston, Massachusetts, Division of Pediatric Nephrology Unit, University of Texas Medical Center in Houston, Texas, USA
Monitoring changes in gene expression in renal ischemia-reperfusion in the rat. Background. Although acute renal failure (ARF) is a relatively common disorder with major morbidity and mortality, its molecular basis remains incompletely defined. The present study examined global gene expression in the well-characterized ischemia-reperfusion model of ARF using DNA microarray technology. Methods. Male Wistar rats underwent bilateral renal ischemia (30 min) or sham operation, followed by reperfusion for 1, 2, 3 or 4 days. Plasma creatinine increased approximately fivefold over baseline, peaking on day 1. Renal total RNA was used to probe cDNA microarrays. Results. Alterations in expression of 18 genes were identified by microarray analysis. Nine genes were up-regulated (ADAM2, HO-1, UCP-2, and thymosin 4 in the early phase and clusterin, vanin1, fibronectin, heat-responsive protein 12 and FK506 binding protein in the established phase), whereas another nine were down-regulated (glutamine synthetase, cytochrome p450 IId6, and cyp 2d9 in the early phase and cyp 4a14, Xist gene, PPAR␥, ␣-albumin, uromodulin, and ADH B2 in the established phase). The identities of these 18 genes were sequenceverified. Changes in gene expression of ADAM2, cyp2d6, fibronectin, HO-1 and PPAR␥ were confirmed by quantitative real-time polymerase chain reaction (PCR). ADAM2, cyp2d6, and PPAR␥ have not previously been known to be involved in ARF. Conclusion. Using DNA microarray technology, we identified changes in expression of 18 genes during renal ischemiareperfusion injury in the rat. We confirmed changes in five genes (fibronectin, ADAM2, cyp 2d6, HO-1 and PPAR␥) by quantitative real-time PCR. Several genes, not previously been identified as playing a role in ischemic ARF, may have importance in this disease.
Key words: microarray, ischemia-reperfusion, gene expression. Received for publication August 16, 2001 and in revised form October 23, 2001 Accepted for publication November 19, 2001
2002 by the International Society of Nephrology
Acute renal failure (ARF), characterized by rapid decline in glomerular filtration rate (GFR), is potentially reversible but is associated with major morbidity and mortality. Despite new insights into the pathogenesis of ARF, neither the incidence nor mortality has declined in decades [1]. During the induction and course of ARF many intrarenal alterations occur, including changes in cell polarity [2–5], levels of signaling molecules [6–11] expression of adhesion molecules [8, 12–15] and patterns of apoptosis and necrosis [16]. Nevertheless, it is probable that additional factors participate in the evolution and repair of ARF but are yet to be appreciated. Discovering such factors may lead to improved strategies for preventing ARF and treating this serious disorder. Certain changes seen in the course of ARF involve infiltrating cells and adhesion molecules. Neutrophil infiltration is induced by adhesion molecules such as intercellular adhesion molecule-1 (ICAM-1) and E-selectin via the mediation of tumor necrosis factor-␣ (TNF␣) and nuclear factor-B (NF-B). Infiltrating neutrophils induce adhesion molecules such as integrins [12], E-selectin [13], P-selectin [14], intercellular adhesion molecule1 [8], and neural cell adhesion molecule [15]. Inhibition of NF-B using a decoy oligonucleotide has been reported to suppress the inflammatory response that follows ischemia-reperfusion injury [17]. Animals that express aberrant adhesion molecules may be protected from renal injury, as noted in ICAM-1–deficient mice that are resistant to ischemic renal injury [18]. Depolarization of epithelial cells occurs in the course of renal injury and repair, in part secondary to changes in adhesion molecules. For example, in ARF induction of transforming growth factor- (TGF-), deactivation of integrin, and increased expression of adhesion molecules is associated with disruption of the cytoskeleton in proximal tubular cells [19]. In this setting the actin microfilament network becomes redistributed into large
1646
Yoshida et al: ARF and DNA microarray
cytoplasmic aggregates [19], which leads to depolarization and exfoliation of epithelial cells. Newly expressed integrin appears to restore cell polarization, accompanied by increased expression of laminin and fibronectin [19]. Ischemic injury results in apoptosis as well as necrosis. Apoptosis is mediated through several signaling cascades, one of which is the mitogen-activated protein (MAP) kinase cascade. Members of this family, such as p38MAPK [10], JNK and ERK [11] have been previously reported to play a role in ARF. Park, Chen and Bonventre suggested that activation of JNK or p38 is involved in protection against ischemic injury in the kidney [11]. TNF-␣ and integrin also have been reported to mediate apoptosis [20, 21]. DNA microarray technology offers the advantage of analyzing thousands of genes simultaneously, with the potential to uncover new mechanisms of pathogenesis. Thus, microarray analysis coupled with bioinformation could detect changes in genes previously unknown to participate in renal injury and repair. Such an approach could lead, potentially, to a new paradigm through which ARF may be better understood. In the present study, we examined changes in gene expression induced by ischemia-reperfusion injury. A total of 2,100 mouse cDNAs spotted onto a glass slide were used for profiling gene expression. As expected, we found that many of the changes in expression occur in genes that encode proteins previously implicated in ischemia-reperfusion injury. However, and importantly, ADAM2 (a disintegrin domain and metalloproteinase), cytochrome p450 IID6, peroxisome proliferator-activated receptor gamma (PPAR␥) were newly identified to participate in ARF. METHODS Animals and RNA extraction Male Wistar rats (200 to 250 g, N ⫽ 16 ischemiareperfusion, N ⫽ 20 sham) were housed in a 12-hour light/dark cycle, and allowed free access to food and water. Control and treated animals were weight- and age-matched at the time of initiation of bilateral ischemic injury, and body weights were recorded at initiation and completion of experiments. The animals were allowed free access to standard rodent chow and water until overnight prior to surgery when food (but not water) was removed. Animals were anesthetized intraperitoneally with pentobarbital sodium, 65 mg/kg. Both renal arteries were exposed using a midline incision and then clamped for 30 minutes followed by reperfusion. Sham animals had an incision plus 30 minutes of waiting time without clamping. After ischemia or sham treatment, muscle layer incisions were sutured, and skin incisions closed with Michel clips. We gave a total of 6 mL of 0.9% saline solution IP to the rats during and after surgery to prevent dehydration. In the present study, we chose to analyze
1647
Fig. 1. Serum creatinine in sham-operated (䊊) and ischemia-reperfusion (䊉) rats at various times. Creatinine rose to a peak at 24 hours in the ischemia-reperfusion rats. Data are expressed as mean ⫾ SEM for 4 rats.
RNA from animals that showed more prominent creatinine changes. Animals were then followed for 1, 2, 3, or 4 days (at least N ⫽ 3 per group per time point). At sacrifice the kidneys were removed and snap-frozen in liquid nitrogen. The samples were stored at ⫺70⬚C until further study. The serum creatinine level over time is shown in Figure 1, and documents successful induction of ARF. Total RNA was isolated from kidney by Trizol (Gibco, Rockville, MD, USA) extraction. High-density DNA microarray preparation High-density cDNA microarrays were generated by standard methods [22]: Mouse expressed sequence tags (ESTs; Research Genetics, Huntsville, AL, USA) were amplified by the polymerase chain reaction (PCR), according to the manufacturer’s recommendations. Each sample was run in a 1% agarose gel for quality verification (data not shown). The sizes of PCR products ranged from approximately 300 bp to 1.2 kb. DNA microarrays, containing 2,100 distinct ESTs, were printed onto glass slides using a pin-style arrayer (BioRobotics, Cambridge, UK). The identities of all 18 genes were confirmed by DNA sequencing using an ABI PRISM 377 DNA sequencer (Applied Biosystems, Foster City, CA, USA) performed by the Genome Center staff at Brigham & Women’s Hospital. The sequence was confirmed using BLAST (http://www.ncbi.nlm.nih.gov/BLAST/). To make microarrays, sialylated glass slides (CEL Associates Inc., Houston, TX, USA) were cleaned for two hours in a solution of 2 N NaOH. After rinsing in distilled water, the slides were treated with a 1:5 dilution of polyl-lysine adhesive solution (Sigma, St. Louis, MO, USA) for one hour, and then dried for five minutes at 80⬚C in a vacuum oven. DNA samples from 100-L PCR reactions were precipitated with ethanol in 96-well microtiter plates. The resulting precipitates were resuspended in 20 L 3 ⫻ standard sodium citrate (SSC; 20 ⫻
1648
Yoshida et al: ARF and DNA microarray
SSC; 3.0 mol/L NaCl and 0.3 mol/L sodium citrate) and transferred to new plates for arraying. After printing, the microarrays were rehydrated for 10 seconds in a humidity chamber. The DNA was then ultraviolet-crosslinked to the surface by subjecting the slides to 60 mJ of energy (Stratagene Stratalinker, La Jolla, CA, USA). Directly after the blocking reaction, the bound DNA was denatured by a two-minute incubation in distilled water at ⬃95⬚C. The slides were then transferred into a bath of ⫺20⬚C ethanol, rinsed, and then spun dry in a clinical centrifuge. Slides were stored in a closed box at room temperature until used. cDNA probe labeling cDNA probes were synthesized as previously reported [23] to improve signal detection on high-density microarrays using a fluorescent oligonucleotide dendrimeric signal amplification system. Briefly, 10 g of total RNA was primed with 1.0 picomole of modified reverse transcription (RT) primer with capture sequence for Cy3 or Cy5 (Genisphere, Montvale, NJ, USA). The RT reaction was performed according to standard protocols (SuperScript RT II; Gibco), followed by the addition of 3.5 L of 0.5 mol/L NaOH/50 mmol/L ethylenediaminetetraacetic acid (EDTA), and heating at 65⬚C for ten minutes to denature the DNA/RNA hybrids. This reaction was then neutralized with 5 L of 1 mol/L Tris, pH 7.5. For double-labeled detection, RT reactions were mixed and precipitated using 3 mol/L sodium acetate and 100% ethanol, then washed once with 70% ethanol. The pellets were resuspended in 20 L distilled H2O. Hybridization cDNA microarrays were prehybridized with 500 L of preheated (65⬚C) ExpressHyb (Clontech, Palo Alto, CA, USA) for an hour in a microarray hybridization chamber (Grace Bio-Labs, Bend, OR, USA) in a rotisserie hybridization oven. Probe was put into the hybridization chamber and incubated overnight in a rotisserie hybridization oven. The next day the microarrays were washed for five minutes in 2 ⫻ SSC/0.2% sodium dodecyl sulfate (SDS) at 65⬚C, then in 2 ⫻ SSC and 0.2 ⫻ SSC for two minutes each. For dendrimer detection, 1.0 L of the Cy3- and Cy5-labeled dendrimers (Genisphere) were suspended in 15 L of 4 ⫻ SSC/2% SDS and layered on the array. The arrays were covered with a coverslip and then incubated at 65⬚C for eight hours in a microarray hybridization cassette (Telechem International, Sunnyvale, CA, USA). The chips were then washed as above. Data collection, normalization, and analysis Each time point was analyzed twice with reverse color labeling. The mean of the two measurements was used to determine the fold changes. As was stated, dye reversal was used. This analysis uses a log-normal error model.
The cut off was based on at least twofold changes for the genes of interest. The chips were scanned in a ScanArray 5000 confocal laser scanner (GSI Lumonics, Marina Del Rey, CA, USA). The signal and background signals were then quantified with ImaGene 4.0 quantification software (BioDiscovery, Billerica, MA, USA). Signals were measured as the mean pixel intensity within each circumscribed spot and average background was measured using the mean pixel intensities of blank spots in the array (without cDNA) and as the mean pixel intensity outside the circumscribed spot diameter for calculating specific signals. Quantitative real time PCR Quantitation with real-time, one-step reverse-transcriptase polymerase chain reaction (RT-PCR) was performed with SYBR Green PCR Reagents and an ABI PRISM 7700 Sequence Detection System (PE Applied Biosystems, Foster City, CA, USA), according to the manufacturer’s instructions. Reactions were performed using 1.0 L of RNA at a concentration of 100 ng/L, in a reaction volume of 50 L. An RT reaction was performed at 48⬚C for 30 minutes, followed by PCR consisting of AmpliTaq activation for 10 minutes at 95⬚C, then 40 cycles with heating to 95⬚C for 15 seconds, and cooling to 60⬚C for one min. mRNA levels were normalized to levels of glyceraldehyde-3-phosphate dehydrogenase (GAPDH). We elected to confirm changes in five genes using real-time one-step quantitative RT-PCR. These five genes and PCR primers are as follows: Fibronectin. Forward primer, 5⬘-GTGGCTGCCTTCAACTTCTC-3⬘ Reverse primer, 5⬘-GTGGGTTGCAAACCTTCAAT-3⬘ HO-1. Forward primer, 5⬘-TCTATCGTGCTCGCATGAAC-3⬘ Reverse primer, 5⬘-CAGCTCCTCAAACAGCTCAA-3⬘ ADAM2. Forward primer, 5⬘-GGGATCTGTTTGGCACTGAT-3⬘ Reverse primer, 5⬘-ATTAGGTGCGCAGTCCTTGT-3⬘ PPAR␥. Forward primer, 5⬘-TGTGGACCTCTCTGTGATGG-3⬘ Reverse primer, 5⬘-CATTGGGTCAGCTCTTGTGA-3⬘ Cytochrome P450 IId6. Forward primer, 5⬘-AGCTTCAACACCGCTATGGT-3⬘ Reverse primer, 5⬘-CAGCAGTGTCCTCTCCATGA-3⬘ GAPDH. Forward primer, 5⬘-ACTCCCATTCCTCCACCTTT-3⬘ Reverse primer, 5⬘-TTACTCCTTGGAGGCCATGT-3⬘ All primer sets were designed according to manufacturer’s recommendations, and when feasible, to span one or more introns. All amplicons ranged between 142 and 148 base pairs. Electrophoretic analysis of expected product sizes was performed for all primer sets prior to one step, real time RT-PCR, to confirm the fidelity of the reaction.
Yoshida et al: ARF and DNA microarray
RESULTS cDNA microarray analysis Representative differential expression patterns are shown for sham (Fig. 2A) and ischemia-reperfusion samples (Fig. 2B) at day 1 on one glass slide. Ischemia-reperfusion (IR) samples were labeled with Cy3 and sham with Cy5. This figure demonstrates clear signals with little background and several instances of clear differential gene expression. Every spot is well rounded and has virtually the same diameter. Figure 2C shows an overlay picture of sham and IR such that green spots indicate up-regulation and red spots indicate down-regulation, and yellow spots indicate no change, respectively. The numbers 1, 2, 3, 4, and 5 indicate clusterin, ADAM2, HO-1, cyp2d6, PPAR␥, respectively. Of these, clusterin, ADAM2, and HO-1 are increased and cyp2d6 and PPAR␥ are decreased in IR. As shown in Figure 3A, a scatter plot of sham (Cy3) versus sham (Cy5) fell on a line of identity. When a scatter plot of sham (Cy5) versus IR (Cy3) was performed, some changes in individual genes were seen (Fig. 3B). Red spots, (identified as Adam2, HO-1, and clusterin) signify up-regulated genes in IR and green spots (identified as cyp2d6 and PPAR␥) signify down-regulated genes. To confirm these data, reverse labeling was performed. Instead of labeling IR with Cy3 and sham with Cy5, IR was labeled with Cy5 and sham with Cy3. In Figure 3C, a scatter plot of this reverse labeling is shown; now green spots (identified as Adam2, HO-1, and clusterin) signify genes decreased in IR and red spots (identified as cyp2d6 and PPAR␥) signify transcripts increased in IR. Thus, these reverse labeling experiments demonstrated consistent results. Furthermore, all 18 transcripts altered were sequence-verified to confirm their identities.
1649
days, consistent with other reports [12, 25]. ADAM2 was up-regulated by 3.3 fold and gradually normalized. Genes down-regulated by ischemia-reperfusion injury Figure 5 shows nine genes that were decreased in IR. Of these nine, four genes decreased acutely, including alpha-albumin protein, cytochrome P450 IID6 (cyp2d6), cytochrome P450 IId9 (cyp2d9), and glutamine synthetase (Fig. 5A). Cyp2d6 was markedly decreased, to approximately 20% of the sham value. Five genes were chronically down-regulated, including alcohol dehydrogenase-B2, cytochrome P450 IVa14 (cyp4a14), PPAR␥, Xist gene, and uromodulin (Fig. 5B). Uromodulin was decreased to approximately 20 to 25% of the sham value. Confirmation of change in expression Microarray data of five genes of interest were confirmed by real-time quantitative PCR (Fig. 6); others were not tested. Fibronectin was chosen because of its well-known increase in ischemic injury [12]. HO-1 was selected as increased on gene array and previously described to be increased in ARF [24, 26–28]. ADAM2 was chosen as a gene that was increased on gene array, and a previously undescribed finding in ARF. PPAR␥ and Cyp 2d6 were chosen as they were decreased on gene array, previously undescribed findings in ARF. Using real-time PCR, HO-1 and ADAM 2 were confirmed to be transitionally increased whereas cytochrome p450 IId6 and PPAR␥ were decreased in ischemia-reperfusion. Thus, there was excellent concordance between microarray data and real time PCR, though the magnitude of changes were more pronounced than those seen with microarray analysis.
Genes up-regulated by ischemia-reperfusion injury
DISCUSSION
Figure 4 shows the nine genes increased in IR. Of these, three increased acutely. These include heme oxygenase-1, thymosin 4, and uncoupling protein (UCP)-2 (Fig. 4A). Heme oxygenase-1 (HO-1) increased at day 1 followed by normalization, a finding previously reported in a rat model of ARF [24]. Thymosin 4 (Tb4) increased approximately twofold. Uncoupling protein (UCP)-2 mRNA level increased by 2.8-fold at day 1 and recovered to baseline at day 4. Six genes were elevated throughout the four-day time course of IR (Fig. 4B). These included ADAM2, clusterin, FK506 binding protein 1a (Fkbp1a), fibronectin, vanin1, and heat-responsive protein 12 (hrsp12). The expression of clusterin (also known as a TRPM-2 or sulfated glycoprotein 2) was increased approximately fivefold and maintained at that level until day 4. Fibronectin increased by about 2.5 fold and remained so for four
In the present study, we demonstrated that multiple gene transcripts were increased or decreased by ischemia-reperfusion, and we identified several genes previously unknown to be involved in renal injury and repair. For example, it has been previously reported that clusterin [16] and fibronectin [12, 29] are up-regulated by renal ischemia-reperfusion injury, and that HO-1 is up-regulated during ARF [24, 30, 31]. Uromodulin has previously been shown to decrease after renal ischemiareperfusion injury [32]. The transcript with the largest observed change was clusterin, known as a marker for cell injury in the tissues. In the present studies clusterin was increased from day 1 to day 4 after ischemic injury. Clusterin mRNA is overexpressed in apoptosis-resistant cells [33] and is required for survival of cells treated with TNF␣ or subjected to oxidative stress [34–36]. Thus, we infer from these find-
1650
Yoshida et al: ARF and DNA microarray
Fig. 2. Representative differential expression by cDNA array. Representative differential expression is shown by microarray of sham (A, Cy3) and ischemia-reperfusion (B, Cy5) at day 1 on one glass slide. (C ) Overlay picture of sham and IR microarray is shown. In this picture, each red spot indicates higher signal in sham and each green spot indicates higher in IR. Numbers denote: (1) clusterin, (2) ADAM2, (3) HO-1, (4) cyp2d6, (5) PPAR␥. Note that the signal of spots numbered 1, 2 and 3 are higher on the right compared to the left, indicating an increase. In contrast, the spots numbered 4 and 5 indicate a decrease in expression.
Fig. 3. Scatter plot graph of microarray data. (A) Sham labeled by Cy3 versus sham labeled by Cy5. All data points are on a 45⬚ angle, confirming the accuracy of this technique. (B) Scatter plot of sham labeled by Cy5 versus ischemia-reperfusion (IR) labeled by Cy3 at day 1 was shown. Red spots, (typed as ADAM2, HO-1, and clusterin) signify genes that are upregulated and green spots, (typed as cyp2d6 and PPAR␥) signify genes that are down-regulated. (C ) Scatter plot of IR labeled by Cy5 vs sham labeled by Cy3 as shown on day 1, as a reverse labeling of scatter plot B. Red spots, (typed as ADAM2, HO-1, and clusterin) signify up-regulated and green spots, (typed as cyp2d6 and PPAR␥), down-regulated. Thus, reverse labeling confirmed the data. Abbreviations are: ADAM2, a disintegrin domain and metalloproteinase; Clu, clusterin; HO1, heme oxygenase-1; Cyp2d6, cytochrome P450 2d6; PPAR␥, peroxisome proliferator activated receptor-gamma.
Yoshida et al: ARF and DNA microarray
Fig. 4. Time course of up-regulated genes. The relative fold change of ischemia-reperfusion/sham is shown in a time-dependent manner. Abbreviations are: IR, ischemia-reperfusion; ADAM2, a disintegrin domain and metalloproteinase; Clu, clusterin; Fkbp1a, FK506 binding protein 1a; Fn, fibronectin; HO1, heme oxygenase-1; Hrsp12, heatresponsive protein 12; Tmsb4, thymosin beta-4; Ucp2, UCP2; Vnn1, vanin 1. Symbols in A are: (䊊) HO1; (䊉) Tmsb4; (䊐) Ucp2. Symbols in B are: (䊏) Adam2; (䉱) Clu; (䉭) Fkbp1a; (䉬) FN; (䉫) Vnn1; (䊊) Hrsp12.
ings that clusterin may play role in protection from ischemia-induced apoptosis in kidney. In the present study, HO-1 increased at day 1, decreased gradually at day 2 and normalized by day 4. HO-1 has already been shown to be increased in the kidney in various models of ARF [24, 26–28]. Shimizu et al reported that HO-1 conferred a protective effect in ischemic acute renal failure, in that when HO activity was blocked by tin mesoporphyrin, increases in microsomal heme concentration, and serum creatinine concentration were observed, along with extensive tubular epithelial cell injury [31]. Uromodulin (Tamm-Horsfall protein) was greatly decreased in our model. This protein, one of the most abundant in the renal tubule, is exclusively localized to
1651
Fig. 5. Time course of down-regulated genes. Relative fold change of ischemia-reperfusion/sham is shown in time-dependent manner. Abbreviations are: IR; ischemia-reperfusion; Aalb, alpha-albumin protein; Adhb2, alcohol dehydrogenase-B2; Cyp2d6, cytochrome P450 2d6; Cyp2d9, cytochrome P450 2d9; Cyp4a14, cytochrome P450, 4a14 (Cyp4a14); PPAR␥, peroxisome proliferator activated receptor gamma; Glns-ps1, intronless glutamine synthetase gene; Xist, Xist gene, P0 region; Umod, uromodulin. Symbols in A are: (䊊) Adhb2; (䊉) Cyp4a14; (䊐) PPAR␥; (䊏) Xist; (䉭) Umod. Symbols in B are: (䉱) Aalb; (䉫) Cyp2d6; (䉬) Cyp2d9; (䊊) Glns-ps1.
the thick ascending limb of Henle’s loop [37]. Safirstein et al and Bachmann et al reported that uromodulin mRNA declined to undetectable levels by 24 hours after renal ischemia-reperfusion [32, 37]. We confirm their observation on day 1 after ischemia injury. In addition, our data show that uromodulin gene expression was continuously inhibited up to day 4 post-injury. The microarray data showed that ADAM2 was increased on day 1 after IR injury and gradually returned to normal on day 4. The real-time PCR confirmed that there is an increase of ADAM2 on day 1. However, the PCR results further showed no changes or decrease after day 2. We do not know the reasons for this discrepancy. Nevertheless, the PCR results did confirm that ADAM2 was transiently up-regulated on day 1. Novel observations included increased ADAM2 in ARF. This gene, also known as fertilin beta, is a well-characterized mole-
1652
Yoshida et al: ARF and DNA microarray
Fig. 6. Time course of relative fold change of mRNA by real-time quantitative polymerase chain reaction (PCR). Relative fold change of ischemia-reperfusion/sham is shown in time-dependent manner. Abbreviations are: IR, ischemia-reperfusion; ADAM2, a disintegrin domain and metalloproteinase; Cyp2d6, cytochrome P450 2d6; Fn, fibronectin; HO1, heme oxygenase-1; PPAR␥, peroxisome proliferator activated receptor gamma. Symbols are: (䊏) Adam2; (䊐) Cyp2d6; (䉬) FN; (䊊) HO-1; (䊐) PPAR␥. The values are expressed by log 2. Data are expressed as mean ⫾ SEM; N ⫽ 4.
cule that mediates cell-cell communication in sperm-egg interactions. ADAM2 has not previously been known to be involved in ARF. Interestingly, this protein increases in the central nervous system in cases of Alzheimer’s disease, along with alteration in cell-matrix contact in the brain [38]. We thus infer that this molecule plays a role in cell-matrix interactions within the kidney. Several members of the ADAM family have the ability to cleave transmembrane proteins, such as growth factors, as well as their receptors, cell adhesion molecules, and cytokines, resulting in soluble forms of those substances, a phenomenon that is termed ectodomain shedding [39]. For example, ADAM-17 (TNF␣ converting enzyme, TACE), a member of this family, cleaves transmembrane proTNF␣ to produce mature form of TNF␣. Tissue inhibitor of metalloproteases-3 (TIMP-3) inhibits these proteins that are involved in ectodomain shedding [40–43]. It may be that ADAM2 plays a role as a shedder, during the regeneration needed during tissue repair after renal damage. Thus, we speculate that ADAM2 plays an important role in repair after tissue injury with in the kidney. We also noted that thymosin 4 (Tb4) was increased in ischemia-reperfusion injury. The beta-thymosins constitute a family of actin monomer-sequestering proteins widely distributed among vertebrates, and Tb4 is most abundant protein in this family. Tb4 has been reported to bind to actin monomers and to enhance wound healing [44–46]. Additionally, it is up-regulated following focal brain ischemia [47]. Our findings lead us to hypothesize that Tb4 has a general role in repair of cytoskeleton and
in re-polarization of renal tubular epithelial cells after IR injury. UCP-2 was increased by IR injury, a previously unreported finding. This gene is involved in mitochondrial hydrogen peroxide generation [48]. UCP-2 mRNA has been reported to be up-regulated approximately threefold in fatty liver, during which hydrogen peroxide production by mitochondria was increased. It has been speculated that this increase may result from chronic apoptotic stress and may play a role in adaptation to such stress [49]. Li et al reported that UCP-2 is involved in cellular defense against oxidative stress (hydrogen peroxide 200 mmol/L) in clonal beta-cells that overexpress UCP-2 [50], further supporting this theory. We speculate that UCP-2 plays an analogous role in renal injury. Vanin-1 increased by about 2.7-fold and remained elevated throughout the study period. Vanin-1, also called pantetheinase, hydrolyzes pantethein to cysteamine (a powerful antioxidant) and pantothenic acid [51]. Therefore, we hypothesize that vanin-1 acts as an endogenous anti-oxidant in renal repair after ischemia-reperfusion injury [52]. We noted PPAR␥ to be down-regulated in ischemiareperfusion injury, a novel finding. PPARs (peroxisome proliferator-activated receptors), a family of transcription factors, have multiple roles that are of great interest. Although etomoxir, a PPAR␣ agonist, was reported to protect the kidney in ischemia-reperfusion injury [53], to our knowledge the present work is the first to find changes in PPAR␥ in this model. PPAR␥ is a nuclear receptor that serves as a transcription factor and thus regulates several physiologic pathways, particularly those involved in lipid metabolism [54]. Fahmi reported that PPAR␥ inhibited interleukin-1 induced matrix metalloproteinase 13 [54]. Furthermore, Yee et al reported that the stromelysin-1 (MMP-3) gene, which is involved in tissue remodeling, contains a PPAR element in its promotor [55]. Therefore, it seems likely that PPAR␥ plays a role in tissue repair, especially in control of MMP. Nakajima et al observed that heterozygous PPAR␥-deficient mice subjected to intestinal ischemia suffered more pronounced injury than control mice [56]. Furthermore, when such mice were treated with a PPAR␥-activating ligand, BRL-49653, they were partially reconstituted and suffered less damage. This report supports our hypothesis that PPAR␥ may act through one of its pathways related to apoptosis and/or tissue repair after ischemic injury. Cytochrome p450 IId6, IId9, and IVa14, all of which are involved with oxidation, were suppressed. When these enzymes are active, they lead to consumption of oxygen. Hence, when tissue is subjected to severe ischemic injury, these enzymes are likely suppressed in an attempt to minimize ischemic injury. We speculate that these enzymes are suppressed during ischemic renal injury in an
Yoshida et al: ARF and DNA microarray
attempt to minimize oxidation injury. Suppression of these cytochrome p450s may occur during the ischemic state to slow down oxidative metabolism, which would lead to cytoprotection. The present analysis of the global gene expression profile in ischemia-reperfusion used whole kidney. Thus, the expression pattern within various intrarenal structures is not distinguishable from the whole. It also is possible that certain of the observations could be the result of inflammatory cell infiltration or injury to the microvasculature. Nonetheless, given the early stage and high cost of array technology, we chose not to focus on dissected cell types (such as, glomerulus, proximal tubule, thick ascending loop of the limb of Henle, distal tubule, collecting duct, etc.) in an initial study of renal ischemiareperfusion. Once routine amplicfication technology is established, pursuing such a goal will be more feasible. In conclusion, using DNA microarray we have identified 18 genes that in mRNA expression lead in the course of ischemia-reperfusion injury. The identities of these 18 genes were confirmed by DNA sequencing. Changes in five genes (fibronectin, ADAM2, cytochrome p450 IId6, HO-1 and PPAR␥) were further confirmed by quantitative real-time PCR. Several genes that have not previously been identified in ischemic ARF may have important implications concerning the evolution of this disease.
9. 10. 11.
12. 13. 14. 15. 16.
17.
18.
19.
ACKNOWLEDGMENTS This study was supported by National Institutes of Health Grants DK 50836 (to S.S. Tang), HL 48455 and DK 58950 (to J.R. Ingelfinger), DK 36031, DK 51606, and DK58849 (to S.R. Gullans). Microarray experiments were performed in the BWH NIDDK DNA microarray Biotechnology Center. We thank Roderick Jensen for valuable discussions.
20.
Reprint requests to Shiow-Shih Tang, Ph.D., Pediatric Nephrology Unit, Massachusetts General Hospital, 55 Fruit Street, BHX 410, Boston, Massachusetts 02114, USA. E-mail: [email protected]
22.
REFERENCES
24.
1. Thadhani R, Pascual M, Bonventre JV: Acute renal failure. N Engl J Med 334:1448–1460, 1996 2. Molitoris BA, Wagner MC: Surface membrane polarity of proximal tubular cells: Alterations as a basis for malfunction. Kidney Int 49:1592–1597, 1996 3. Molitoris BA, Chan LK, Shapiro JI, et al: Loss of epithelial polarity: A novel hypothesis for reduced proximal tubule Na⫹ transport following ischemic injury. J Membr Biol 107:119–127, 1989 4. Molitoris BA, Hoilien CA, Dahl R, et al: Characterization of ischemia-induced loss of epithelial polarity. J Membr Biol 106:233– 242, 1988 5. Molitoris BA, Wilson PD, Schrier RW, et al: Ischemia induces partial loss of surface membrane polarity and accumulation of putative calcium ionophores. J Clin Invest 76:2097–2105, 1985 6. Raman N, Atkinson SJ: Rho controls actin cytoskeletal assembly in renal epithelial cells during ATP depletion and recovery. Am J Physiol 276:C1312–C1324, 1999 7. Donnahoo KK, Meldrum DR, Shenkar R, et al: Early renal ischemia, with or without reperfusion, activates NFkappaB and increases TNF-alpha bioactivity in the kidney. J Urol 163:1328– 1332, 2000 8. Okusa MD, Linden J, Huang L, et al: A(2A) adenosine receptor-
21.
23.
25.
26.
27.
28.
29. 30.
31.
1653
mediated inhibition of renal injury and neutrophil adhesion. Am J Physiol (Renal Physiol) 279:F809–F818, 2000 Sakai M, Tsukada T, Harris RC: Oxidant stress activates AP-1 and heparin-binding epidermal growth factor-like growth factor transcription in renal epithelial cells. Exp Nephrol 9:28–39, 2001 Faccio L, Chen A, Fusco C, et al: Mxi2, a splice variant of p38 stress-activated kinase, is a distal nephron protein regulated with kidney ischemia. Am J Physiol (Cell Physiol) 278:C781–C790, 2000 Park KM, Chen A, Bonventre JV: Prevention of kidney ischemia/ reperfusion-induced functional injury and JNK, p38, and MAPK kinase activation by remote ischemic pretreatment. J Biol Chem 276:11870–11876, 2001 Zuk A, Bonventre JV, Brown D, et al: Polarity, integrin, and extracellular matrix dynamics in the postischemic rat kidney. Am J Physiol (Cell Physiol) 275:C711–C731, 1998 Singbartl K, Ley K: Protection from ischemia-reperfusion induced severe acute renal failure by blocking E-selectin. Crit Care Med 28:2507–2514, 2000 Singbartl K, Green SA, Ley K: Blocking P-selectin protects from ischemia/reperfusion-induced acute renal failure. FASEB J 14:48– 54, 2000 Abbate M, Brown D, Bonventre JV: Expression of NCAM recapitulates tubulogenic development in kidneys recovering from acute ischemia. Am J Physiol 277:F454–F463, 1999 Schumer M, Colombel MC, Sawczuk IS, et al: Morphologic, biochemical, and molecular evidence of apoptosis during the reperfusion phase after brief periods of renal ischemia. Am J Pathol 140: 831–838, 1992 Vos IH, Govers R, Grone HJ, et al: NFkappaB decoy oligodeoxynucleotides reduce monocyte infiltration in renal allografts. FASEB J 14:815–822, 2000 Kelly KJ, Williams WW, Jr, Colvin RB, et al: Intercellular adhesion molecule-1-deficient mice are protected against ischemic renal injury. J Clin Invest 97:1056–1063, 1996 Kellerman PS, Clark RA, Hoilien CA, et al: Role of microfilaments in maintenance of proximal tubule structural and functional integrity. Am J Physiol 259:F279–F285, 1990 Daemen MA, van de Ven MW, Heineman E, et al: Involvement of endogenous interleukin-10 and tumor necrosis factoralpha in renal ischemia-reperfusion injury. Transplantation 67:792– 800, 1999 Kumar CC: Signaling by integrin receptors. Oncogene 17:1365– 1373, 1998 DeRisi JL, Iyer VR, Brown PO: Exploring the metabolic and genetic control of gene expression on a genomic scale. Science 278:680–686, 1997 Stears RL, Getts RC, Gullans SR: A novel, sensitive detection system for high-density microarrays using dendrimer technology. Physiol Genomics 3:93–99, 2000 Maines MD, Mayer RD, Ewing JF, et al: Induction of kidney heme oxygenase-1 (HSP32) mRNA and protein by ischemia/reperfusion: possible role of heme as both promotor of tissue damage and regulator of HSP32. J Pharmacol Exp Ther 264:457–462, 1993 Walker PD: Alterations in renal tubular extracellular matrix components after ischemia-reperfusion injury to the kidney. Lab Invest 70:339–345, 1994 Nath KA, Balla G, Vercellotti GM, et al: Induction of heme oxygenase is a rapid, protective response in rhabdomyolysis in the rat. J Clin Invest 90:267–270, 1992 Agarwal A, Balla J, Alam J, et al: Induction of heme oxygenase in toxic renal injury: A protective role in cisplatin nephrotoxicity in the rat. Kidney Int 48:1298–1307, 1995 Vogt BA, Shanley TP, Croatt A, et al: Glomerular inflammation induces resistance to tubular injury in the rat. A novel form of acquired, heme oxygenase-dependent resistance to renal injury. J Clin Invest 98:2139–2145, 1996 Rosenberg ME, Paller MS: Differential gene expression in the recovery from ischemic renal injury. Kidney Int 39:1156–1161, 1991 Horikawa S, Ito K, Ikeda S, et al: Induction of heme oxygenase-1 in toxic renal injury: Mercuric chloride-induced acute renal failure in rat. Toxicol Lett 94:57–64, 1998 Shimizu H, Takahashi T, Suzuki T, et al: Protective effect of
1654
32. 33. 34.
35. 36. 37. 38. 39. 40. 41.
42.
43.
Yoshida et al: ARF and DNA microarray
heme oxygenase induction in ischemic acute renal failure. Crit Care Med 28:809–817, 2000 Safirstein R, Megyesi J, Saggi SJ, et al: Expression of cytokinelike genes JE and KC is increased during renal ischemia. Am J Physiol 261:F1095–F1101, 1991 Koch-Brandt C, Morgans C: Clusterin: A role in cell survival in the face of apoptosis? Prog Mol Subcell Biol 16:130–149, 1996 Sintich SM, Steinberg J, Kozlowski JM, et al: Cytotoxic sensitivity to tumor necrosis factor-alpha in PC3 and LNCaP prostatic cancer cells is regulated by extracellular levels of SGP-2 (clusterin). Prostate 39:87–93, 1999 Humphreys D, Hochgrebe TT, Easterbrook-Smith SB, et al: Effects of clusterin overexpression on TNFalpha- and TGFbetamediated death of L929 cells. Biochemistry 36:15233–15243, 1997 Viard I, Wehrli P, Jornot L, et al: Clusterin gene expression mediates resistance to apoptotic cell death induced by heat shock and oxidative stress. J Invest Dermatol 112:290–296, 1999 Bachmann S, Metzger R, Bunnemann B: Tamm-Horsfall proteinmRNA synthesis is localized to the thick ascending limb of Henle’s loop in rat kidney. Histochemistry 94:517–523, 1990 Gerst JL, Raina AK, Pirim I, et al: Altered cell-matrix associated ADAM proteins in Alzheimer’s disease. J Neurosci Res 59:680–684, 2000 Hooper NM, Karran EH, Turner AJ: Membrane protein secretases. Biochem J 321:265–279, 1997 Amour A, Slocombe PM, Webster A, et al: TNF-alpha converting enzyme (TACE) is inhibited by TIMP-3. FEBS Lett 435:39– 44, 1998 Hargreaves PG, Wang F, Antcliff J, et al: Human myeloma cells shed the interleukin-6 receptor: Inhibition by tissue inhibitor of metalloproteinase-3 and a hydroxamate-based metalloproteinase inhibitor. Br J Haematol 101:694–702, 1998 Smith MR, Kung H, Durum SK, et al: TIMP-3 induces cell death by stabilizing TNF-alpha receptors on the surface of human colon carcinoma cells. Cytokine 9:770–780, 1997 Borland G, Murphy G, Ager A: Tissue inhibitor of metalloproteinases-3 inhibits shedding of L-selectin from leukocytes. J Biol Chem 274:2810–2815, 1999
44. De La Cruz EM, Ostap EM, Brundage RA, et al: Thymosinbeta(4) changes the conformation and dynamics of actin monomers. Biophys J 78:2516–2527, 2000 45. Malinda KM, Sidhu GS, Mani H, et al: Thymosin beta4 accelerates wound healing. J Invest Dermatol 113:364–368, 1999 46. Young JD, Lawrence AJ, MacLean AG, et al: Thymosin beta 4 sulfoxide is an anti-inflammatory agent generated by monocytes in the presence of glucocorticoids. Nat Med 5:1424–1427, 1999 47. Vartiainen N, Pyykonen I, Hokfelt T, et al: Induction of thymosin beta(4) mRNA following focal brain ischemia. Neuroreport 7:1613– 1616, 1996 48. Negre-Salvayre A, Hirtz C, Carrera G, et al: A role for uncoupling protein-2 as a regulator of mitochondrial hydrogen peroxide generation. FASEB J 11:809–815, 1997 49. Yang S, Zhu H, Li Y, et al: Mitochondrial adaptations to obesityrelated oxidant stress. Arch Biochem Biophys 378:259–268, 2000 50. Li LX, Skorpen F, Egeberg K, et al: Uncoupling protein-2 participates in cellular defense against oxidative stress in clonal betacells. Biochem Biophys Res Commun 282:273–277, 2001 51. Pitari G, Malergue F, Martin F, et al: Pantetheinase activity of membrane-bound Vanin-1: Lack of free cysteamine in tissues of Vanin-1 deficient mice. FEBS Lett 483:149–154, 2000 52. Frankel D, Mehindate K, Schipper HM: Role of heme oxygenase-1 in the regulation of manganese superoxide dismutase gene expression in oxidatively-challenged astroglia. J Cell Physiol 185:80–86, 2000 53. Portilla D, Dai G, Peters JM, et al: Etomoxir-induced PPARalpha-modulated enzymes protect during acute renal failure. Am J Physiol (Renal Physiol) 278:F667–F675, 2000 54. Fahmi H, Di Battista JA, Pelletier JP, et al: Peroxisome proliferator–activated receptor gamma activators inhibit interleukin1beta-induced nitric oxide and matrix metalloproteinase 13 production in human chondrocytes. Arthritis Rheum 44:595–607, 2001 55. Yee J, Kuncio GS, Bhandari B, et al: Identification of promoter activity and differential expression of transcripts encoding the murine stromelysin-1 gene in renal cells. Kidney Int 52:120–129, 1997 56. Nakajima A, Wada K, Miki H, et al: Endogenous PPAR gamma mediates anti-inflammatory activity in murine ischemia-reperfusion injury. Gastroenterol 120:460–469, 2001