BASIC RESEARCH www.jasn.org

MicroRNA-489 Induction by –Inducible Factor–1 Protects against Ischemic Injury

† † ‡ † ‡ Qingqing Wei,* Yong Liu, Pengyuan Liu, Jielu Hao,* Mingyu Liang, Qing-sheng Mi, § | Jian-Kang Chen,* and Zheng Dong*

*Department of Cellular Biology and Anatomy, Medical College of Georgia at Augusta University and Charlie Norwood Veterans Affairs Medical Center, Augusta, Georgia; †Department of Physiology, Medical College of Wisconsin, Milwaukee, Wisconsin; ‡Kidney Institute, Changzheng Hospital of Second Military Medical University, Shanghai, China; §Departments of Dermatology and Internal Medicine, Henry Ford Health System, Detroit, Michigan; and |The Second Xiangya Hospital, Central South University, Changsha, China

ABSTRACT MicroRNAs have been implicated in ischemic AKI. However, the specific microRNA species that regulates ischemic kidney injury remains unidentified. Our previous microarray analysis revealed microRNA-489 induction in kidneys of mice subjected to renal -reperfusion. In this study, we verified the induction of microRNA-489 during ischemic AKI in mice and further examined the underlying mechanisms. Hypoxia– inducible factor–1a deficiency associated with diminished microRNA-489 induction in cultured rat proximal tubular cells subjected to hypoxia and kidney tissues of mice after renal ischemia-. Moreover, genomic analysis revealed that microRNA-489 is intronic in the calcitonin receptor gene, and chromatin immunoprecipitation assays showed increased binding of hypoxia–inducible factor–1 to a specificsitein the calcitonin receptor gene promoter after hypoxia. Inhibition of microRNA-489 increased in renal tubular cells after ATP depletion injury in vitro, whereas microRNA-489 mimics mediated protection. In mice, inhibition of microRNA-489 enhanced tubular cell death and ischemic AKI without significantly affecting tubular cell proliferation. Deep sequencing identified 417 mRNAs that were recruited to the RNA–induced silencing complex by microRNA-489. Of the identified mRNAs, 127 contain microRNA-489 targeting sites, and of those, 18 are involved in the cellular stress response, including the poly(ADP-ribose) polymerase 1 gene implicated in ischemic kidney injury. Sequence analysis and in vitro studies validated poly(ADP-ribose) poly- merase 1 as a microRNA-489 target. Together, these results suggest that microRNA-489 is induced via hypoxia–inducible factor–1 during ischemic AKI to protect kidneys by targeting relevant genes.

J Am Soc Nephrol 27: 2784–2796, 2016. doi: 10.1681/ASN.2015080870

Renal ischemia-reperfusion is a main cause of AKI, important role played by miRs in various cellular whichisassociatedwithhighmortality.1 Recent processes ranging from cell growth and prolifera- work further indicates that AKI may be an impor- tion to cell death. Notably, dysregulation of miRs tant contributing factor in the development and progression of CKD.2,3 Decades of research have Received August 7, 2015. Accepted January 25, 2016. gained significant insights into the pathogenesis of ischemic AKI, including tubular cell injury and Published online ahead of print. Publication date available at death, vascular damage, and inflammatory re- www.jasn.org. sponse.4–7 However, effective approaches for the Correspondence: Dr. Zheng Dong, The Second Xiangya Hospital, diagnosis and treatment of AKI are still lacking.8 Central South University, 139 Middle Renmin Road, Changsha, Hunan 410011, China; or Dr. Qingqing Wei, Department of Cel- MicroRNAs (miRs) are a group of noncoding lular Biology and Anatomy, Medical College of Georgia at Augusta small RNAs of 21–25 nucleotides that regulates tar- University, 1459 Laney Walker Boulevard, Augusta, GA 30912. get gene expression mainly by blocking their Email: [email protected] or [email protected] mRNA translation.9,10 Recent work has shown an Copyright © 2016 by the American Society of Nephrology

2784 ISSN : 1046-6673/2709-2784 JAmSocNephrol27: 2784–2796, 2016 www.jasn.org BASIC RESEARCH has emerged as a critical pathogenic factor in a variety of hu- this induction, we measured miR-489 by quantitative real– man diseases.9,10 In kidneys, miRs have been shown to play time PCR (Figure 1). Compared with sham-operated con- indispensable roles in kidney development.11 For example, trol, miR-489 increased 1.4460.17-fold (n=3; P,0.05) and deletion of Dicer (a key enzyme for the production of func- 1.6960.19-fold (n=6; P,0.01) in kidney cortex and outer tional miRs) from podocytes and juxtaglomerular cells led to medulla after 30 minutes of ischemia and 12 or 48 hours of the loss of and function of these cells, resulting in reperfusion, respectively (Figure 1A). Interestingly, kidney defects.12–16 Furthermore, specificmiRs,suchas miR-489 induction was also detected during hypoxic incu- miR-192 and miR-29, were shown to be important regulators bation of cultured rat proximal tubular cells (RPTCs). As 17–20 21 in diabetic nephropathy. In 2010, we and Goldwin shown in Figure 1B, hypoxia (1% O2) induced miR-489 et al.22 revealed significant changes of miR expression profiles time dependently, reaching 1.8460.65-fold (n=5; P,0.05) during ischemic AKI in mice. Furthermore, we showed that at 6 hours and then decreasing toward the basal level (Figure ischemic AKI was markedly attenuated by the deletion of 1B). The in vitro datasuggestthatmiR-489islikelytobe Dicer and consequent depletion of miRs in proximal tubular induced in renal tubules by hypoxia in ischemic AKI. To cells.21 Although these studies showed a role of miRs in the localize miR-489 induction in kidneys after ischemia- pathogenesis of ischemic AKI, the specificpathogenicmiR reperfusion injury, we conducted in situ hybridization species remained unknown. There are .1000 miR species in (Figure 1C). Low levels of miR-489 were detected in most humans. In the last couple of years, several miRs have been renal tubules in control kidneys. After 30 minutes of ische- implicated in the regulation of ischemic AKI.23–26 For exam- mia and 48 hours of reperfusion, there was an overall in- ple, miR-24 was shown to promote renal tubular cell injury crease of miR-489 in renal tubules, and the increase was by targeting H2A histone family, member X, and heme more in injured tubules mainly in the outer stripe of outer oxygenase 1,23 whereas miR-21 was upregulated by ischemic medulla (arrows in Figure 1C). Interestingly, we also de- preconditioning to protect kidneys against subsequent injury tected some strong interstitial signals that appeared in by inhibiting programmed cell death protein 4.25 blue, but the interstitial staining did not change significantly miR-489 is an miR that has been reported to act as a tumor after ischemic injury. It was suggested that miR-489 was suppressor and play roles in muscle stem cell quiescence and mainly induced in renal tubular cells. cardiac hypertrophy. It is also upregulated in clear cell papillary renal cell carci- noma.27–30 In ischemic AKI, our miR profiling suggested miR-489 induction in kidney tissues after renal ischemia- reperfusion.21 However, the mechanism of miR-489 induction in these disease con- ditions was unclear, and the role of miR-489 in ischemic AKI was unknown. In this study, we show that miR-489 is in- duced via hypoxia–inducible factor–1 (HIF-1) during hypoxia of tubular cells in vitro and ischemic AKI in mice. We further show a renoprotective role of miR-489 under these conditions. Deep RNA sequencing (RNA-seq) has further identi- fied multiple target genes of miR-489 that may be responsible for its protective effect.

RESULTS Figure 1. miR-489 is induced in ischemic AKI and hypoxic renal tubular cells. (A) miR-489 Is Induced in Ischemic AKI C57BL/6 mice were subjected to 30 minutes of bilateral renal ischemia with 12 or 48 hours of reperfusion or sham operation. miR-489 in the kidney cortex and outer and Hypoxic Renal Tubular Cells fi medulla was detected by real-time PCR. (B) RPTCs were treated with 1% oxygen, and Our microarray analysis identi ed miR-489 miR-489 level was examined by quantitative real–time PCR (n=3–6). *P,0.05 compared fi as one of the miRs that was signi cantly with sham operation or no hypoxia controls; **P,0.01 compared with sham. (C) C57BL/6 induced in kidney tissues after 30 minutes mice were subjected to 30 minutes of bilateral ischemia with 48 hours of reperfusion of renal ischemia and 12–48 hours of or sham operation. miR-489 in the kidney was detected by in situ hybridization. The reperfusion in mice.21 To fur ther confirm arrows show the significant induction of miR-489 in injured tubules. Scale bar, 0.2 mm.

J Am Soc Nephrol 27: 2784–2796, 2016 MicroRNA-489 in Ischemic Kidney Injury 2785 BASIC RESEARCH www.jasn.org

HIF-1 Contributes to miR-489 Induction in Ischemic AKI with previous work,32–34 HIF-1a was also induced in kidney tis- and Hypoxic Incubation sues during renal ischemia-reperfusion in our study (Figure 2B). Because we observed miR-489 induction by hypoxia in RPTCs, To determine the role of HIF-1 in miR-489 induction we hypothesized that miR-489 induction during renal cell during renal hypoxia and ischemia-reperfusion, we initially hypoxia and ischemic AKI may involve HIF-1, the master established stable human embryonic kidney (HEK293) cell regulator of gene expression in hypoxia.31 HIF-1 consists of lines with HIF-1a knockdown by two different shRNAs a-andb-subunits. Although HIF-1b is constitutively (shRNA-A and shRNA-B). Compared with scrambled expressed, HIF-1a is induced by hypoxia and forms a sequence–transfected cells, HIF-1 inhibitory shRNA– heterodimer with HIF-1b to translocate to the nucleus to in- transfected cells showed a lower HIF-1a induction during duce hypoxic gene expression. Accordingly, HIF-1a induction hypoxic incubation (Figure 2C). Importantly, 6 hours of hyp- is a good indication of HIF-1 activation. In our study, HIF-1a oxia induced a 1.7760.36-fold increase of miR-489 in the was induced during hypoxic incubation of RPTCs, starting scrambled cells, which was completely blocked in siHIF-1– from 3 hours and lasting until 9 hours (Figure 2A). In line transfected cells (Figure 2E).

Figure 2. HIF-1 contributes to miR-489 induction in ischemic AKI and hypoxic incubation. (A) The Western blot of HIF-1a in RPTC whole–cell lysate with 1% oxygen treatment. Cyclophilin B was used as the internal loading control. (B) C57BL/6 mice were subjected to 30 minutes of bilateral renal ischemia with 12 or 48 hours of reperfusion or sham operation. The HIF-1a level was detected by Western blot, and b-actin was used as the internal loading control. (C) HEK293 stable cells with scramble shRNA or HIF-1a inhibition shRNA-A or -B were treated with 1% oxygen for 6 hours. The HIF-1a level was examined by Western blot with b-actin as the internal loading control. (D) PT-HIF-1 wild-type (WT) or PT-HIF-1 knockout (KO) mice were subjected to 30 minutes of bilateral renal ischemia and 12 hours of reperfusion or sham operation. The HIF-1a level was detected by immunohistochemical staining, and the renal tubules with HIF-1a were shown by arrows. (E) HEK293 stable cells with scramble shRNA or HIF-1a inhibition shRNA-A or -B were treated with 1% oxygen for 6 hours. miR-489 level was detected by real-time PCR. *P,0.05 compared with the scramble group. (F) PT-HIF-1 WT or PT-HIF-1 KO mice were subjected to 30 minutes of bilateral renal ischemia and 48 hours of reperfusion or sham operation. miR-489 level in the cortical and outer medulla total RNA was detected by real-time PCR (n=3–5). *P,0.05 compared with sham group.

2786 Journal of the American Society of Nephrology J Am Soc Nephrol 27: 2784–2796, 2016 www.jasn.org BASIC RESEARCH

We further examined the HIF-1 dependence of miR-489 Experimentally, we examined the binding of HIF-1a to these induction in vivo. To this end, we generated renal proximal sites by CHIP assay. After hypoxia, significantly more DNA tubule–specific HIF-1a [PT-HIF-1(2/2)] knockout mice by fragments from calcr site 1 were pulled down by anti–HIF-1a crossing floxed HIF-1a mice with PEPCK-CRE transgenic (Figure 3B), suggesting that HIF may directly bind to calcr site mice. PT-HIF-1(2/2) and PT-HIF-1(+/+) mice showed sim- 1 to regulate miR-489 transcription. ilar renal ischemia-reperfusion injury (Supplemental Figure 1). Immunohistochemical staining showed the induction of Anti–miR-489 Increases Renal Tubular Cell Apoptosis HIF-1a in the nuclei of many renal tubular cells in PT-HIF-1 after ATP Depletion (+/+) kidneys, and this inductive response was suppressed in To examine the role of miR-489 in kidney injury, we initially PT-HIF-1 knockout mice, verifying HIF-1a deletion in kidney tested the in vitro model of ATP depletion, which recapitulates tubules in these animals (Figure 2D). Importantly, renal some of the key features of renal ischemia-reperfusion (some- ischemia-reperfusion induced miR-489 time dependently in times called chemical hypoxia/ischemia). To inhibit miR-489, wild-type mice, which was partially suppressed in PT-HIF-1 we transfected RPTCs with anti–miR-489, a locked nucleic knockout mice (Figure 2F). Together, these in vivo and in vitro acid (LNA) oligonucleotide that is specifically against data indicate that miR-489 is induced via HIF-1 during renal miR-489. The transfection of anti–miR-489 or scrambled se- hypoxia and ischemia-reperfusion. quence did not induce noticeable injury difference in RPTCs In the genome, the miR-489 gene is localized at the intronic under control conditions as shown by low levels of apoptosis region of the calcinotin receptor (calcr) gene, suggesting that and activity (Figure 4). ATP depletion induced 21%6 miR-489 may be an intronic miR of calcr. Using the JASPAR 3% apoptosis in scrambled sequence transfected cells, which database (http://jaspar.genereg.net), we analyzed the pro- was increased to 43%65% by anti–miR-489 (Figure 4, A and moter of calcr and the genomic region immediately upstream B). Consistently, caspase activation after ATP depletion was of the miR-489 sequence. The analysis identified three putative also markedly increased by anti–miR-489 (Figure 4C). The HIF-1 binding sites: one located immediately upstream of the injury-enhancing effect of anti–miR-489 suggests that miR-489 gene and two in the promoter of calcr (Figure 3A). miR-489 may play a protective role in renal tubular cells.

mir-489 Mimic Protects against Apoptosis without Inducing Cell Quiescence We further tested the effect of miR-489 mimics in RPTCs. The cells in control conditions did not show obvious apoptosis with very low levels of caspase activity. After ATP depletion, the cells with negative control mimics had 42%68% apoptosis, whereas those transfected with miR-489 mimics had 22%67% apoptosis (Figure 5, A and B). Consistently, miR-489 mimics signifi- cantly reduced caspase activity (Figure 5C). miR-489 was reported to maintain the quiescent status of muscle stem cells.27 How- ever,inRPTCs,miR-489didnotsignifi- cantly attenuate cell proliferation as shown by Ki67 staining (Figure 5A), suggesting that miR-489 is not critical to the proliferation or Figure 3. HIF-1 binds to Calcr/miR-489 gene promoter. (A) Predicted HIF-1 binding quiescence in renal tubular cells. sites in mouse miR-489 and calcr gene promoter regions. mmu-miR-489 (miR-489)and mouse calcr gene upstream regions were analyzed for HIF-1 binding sites by JASPAR Anti–miR-489 Worsens Renal . online analysis (http://jaspar.genereg.net). The potential binding sites of 90% Function in Ischemic AKI chance (relative score) to bind HIF-1 are listed. (B) The induction of HIF-1 binding to its To determine the role of miR-489 in ische- binding sites after hypoxia. BUMPT cells were treated with 1% oxygen for 6 hours. The – a mic AKI in vivo, mice were injected with chromatin samples were sonicated and immunoprecipitated with anti HIF-1 anti- – body. The induced HIF-1 binding to the specific sites by hypoxia compared with anti miR-489 and then subjected to normoxia was examined by real-time PCR to detect the DNA fragments containing the 25 minutes of bilateral renal ischemia binding sites (n=3–4). H/N, the ratio of HIF-1 binding under hypoxia over HIF-1 with reperfusion (Figure 6A). We first ver- binding under normoxia. *P,0.05comparingHIF-1bindingefficiency with the spe- ified the inhibition of miR-489 in kidney cific site after hypoxia and normoxia treatment. tissues by anti–miR-489. As shown in

J Am Soc Nephrol 27: 2784–2796, 2016 MicroRNA-489 in Ischemic Kidney Injury 2787 BASIC RESEARCH www.jasn.org

induced by renal ischemia-reperfusion in anti–miR-489–injected mice. In the kidneys of these animals, 22% of renal tubules showed the signs of injury or damage (Figure 7B). Moreover, these tissues showed significantly more apoptotic cells (73/mm2)thanthe scrambled oligo group (39/mm2) (Figure 7D).

Anti–miR-489 Does Not Affect Tubular Proliferation after Ischemic AKI In our study, the maximal BUN was de- tected at 24–48 hours of reperfusion after 25 minutes of bilateral renal ischemia. By 72 hours of reperfusion, BUN started to decrease (Figure 6D), suggesting renal re- covery as a result of kidney repair. Anti– miR-489 had notable effects on BUN and renal histopathology at 72 hours of reperfusion (Figures 6D and 7), suggesting that anti–miR-489 may adversely affect kidney repair. Tubular cell proliferation Figure 4. Anti–miR-489 increases renal tubular cell apoptosis after ATP depletion. RPTCs were transfected with 200 nM scramble LNA or anti–miR-489 LNA followed by and resultant tubular repair are key 5 ATP depletion and recovery. (A) The morphology of cells and nuclear. Scale bar, processes in kidney recovery after AKI. 0.2 mm. (B) The apoptosis percentage by morphology estimation. (C) The caspase Thus, we examined tubular cell prolifera- activity (n=3–4). AFC, 7-amino-4-trifluoromethyl coumarin. *P,0.05 compared with tion by Ki67 staining.35 As expected, sham– thescramblegroup. operated control kidneys only showed a few Ki67-positive cells, whereas significant amounts of Ki67-positive cells were in- Figure 6B, real–time PCR analysis showed that miR-489 in duced by 25 minutes of ischemia with 72 hours of reperfusion kidney cortex and outer medulla was reduced to ,5% com- (Figure 8A). Of note, cell counting indicated that anti–miR- pared with scrambled sequence–injected mice (Figure 6B). In 489 did not affect cell proliferation (Figure 8B). These results this 25-minute ischemia model, significant miR-489 induc- suggest that, in ischemic AKI, miR-489 mainly contributes to tion was detected after 24 hours of reperfusion (Figure 6C). kidney protection and does not contribute to kidney repair. We then monitored renal function of the mice after ischemic AKI daily by measuring BUN. Compared with scrambled Identification of miR-489 Targets by Argonaute-2 oligo–injected mice, anti–miR-489–treated mice consistently Immunoprecipitation and RNA-seq showed a higher BUN after renal ischemia (Figure 6D). Anti– In cells, an miR represses target gene expression viathe microRNA– miR-489–treated mice also showed marginally higher serum induced silencing complex (miRISC). Argonaute (Ago) proteins at 72 hours of reperfusion (Figure 6E). are the key components of miRISC, where an miR and its target gene mRNAs associate. Therefore, to identify the targets of Anti–miR-489 Increases Kidney Tubular Damage and miR-489, we analyzed the mRNAs that were associated with Apoptosis in Ischemic AKI miRISC in the presence or absence of miR-489 by deep RNA- We further examined the effect of anti–miR-489 on kidney seq. Specifically, FLAG-tagged Ago-2 and miR-489 mimics (or tissue injury. Kidney tissue sections was stained by hematox- negative control RNA oligos) were cotransfected into HEK cells ylin and eosin and terminal deoxynucleotidyl transferase– to collect lysate for immunoprecipitation (IP) with an anti- mediated digoxigenin-deoxyuridine nick–end labeling assay FLAG antibody. Successful IP of FLAG-Ago-2 was verified by to reveal the tissue damage and apoptosis, respectively. As immunoblotting (Figure 9A). Moreover, real–time PCR analysis expected, sham–operated control kidneys showed a healthy showed that miR-489 transfection led to a 483-fold increase of histology without tubular damage or apoptosis (Figure 7, A miR-489 in transfected cells, which was drastically increased to and C). Twenty-five minutes of renal ischemia with 72 hours 359,129-fold in FLAG-Ago-2 immunoprecipitate (Figure 9B). of reperfusion induced a moderate renal tubular damage in This marked accumulation of miR-489 in FLAG-Ago-2 immu- scrambled oligo–treated mice as shown by dilation, loss of brush noprecipitate was consistent with miR association in miRISC. boarder, and as well as apoptosis in some renal tubules We then analyzed the total RNAs and the RNAs in FLAG- (Figure 7, A and C). In contrast, much more severe AKI was Ago-2 immunoprecipitate by RNA-seq. The analysis detected

2788 Journal of the American Society of Nephrology J Am Soc Nephrol 27: 2784–2796, 2016 www.jasn.org BASIC RESEARCH

RPTCs with or without ATP depletion in- jury (Figure 10C). Together, the results suggestthatPARP1maybeadirecttarget of miR-489.

DISCUSSION

By microarray analysis, we identified 13 miRs that were significantly upregulated in kidney tissues during bilateral renal ischemia-reperfusion.21 In that analysis, miR-489 showed a greater than fourfold induction in kidney cortex and outer medulla. Our study has further validated miR-489 induction in ischemic AKI by quantitative real–time PCR and in situ hybridization (Figure 1, A and C). Figure 5. miR-489 mimic decreases tubular cell apoptosis after ATP depletion without miR-489 induction shown by real-time changing cell proliferation. RPTCs were transfected with 200 nM negative control or miR-489 mimics and then subjected to ATP depletion and recovery. (A) The cell PCR is not as high as that shown by morphology, nuclear, and Ki67 staining. Scale bar, 0.2 mm. (B) The percentage of previous microarray analysis. The discrep- apoptosis. (C) Caspase activity (n=4–6). AFC, 7-amino-4-trifluoromethyl coumarin. ancy between these two analyses, according *P,0.05 compared with the negative control group. to past experience, was mainly caused by the use of different normalization criteria. Specifically, global normalization of 49,116 hits, of which 905 showed significant (over twofold) snoRNA202, snoRNA135, and U87 was used for normaliza- increases in FLAG-Ago-2 immunoprecipitate after miR-489 tion in the microarray analysis, whereas only snoRNA202 was expression. In these 905 hits, 417 were for protein-coding used in our real–time PCR quantification. Nonetheless, these genes, whereas387 were for noncoding genes, and 101hits were analyses showed miR-489 induction mainly in renal tubular for un-named genes (Figure 9C, Supplemental Tables 1 and 2). cells after ischemic AKI. We further analyzed the 39-untranslated regions (39-UTRs) of Mechanistically, we showed evidence for the involvement of the 417 coding genes and found that 127 of them contained HIF-1 in miR-489 induction during renal tubular cell hypoxia miR-489–targeting seed sequence. Functional classification in vitro and renal ischemia-reperfusion in vivo (Figure 2). using The Database for Annotation, Visualization and Inte- First, HIF-1 activation and miR-489 induction showed a grated Discovery indicated that these 127 genes may partici- good temporal correlation in both in vivo and in vitro models. pate in the regulation of several cellular processes or activities Second, miR-489 induction during hypoxia was attenuated in (Supplemental Table 3). Especially, 18 of the identified genes HIF-1a knockdown cells. Third, compared with wild-type were shown to be cellular stress response related (Figure 9D). littermates, PT-HIF-1a knockout mice showed lower Together, the results suggest that miR-489 may regulate kidney miR-489 induction during renal ischemia-reperfusion. It is injury by targeting multiple genes, including many involved in noteworthy that the suppression of miR-489 induction in stress response. PT-HIF-1a knockout kidneys was not complete. At least two features of the model contributed to the incomplete effect. It is Verification of PARP1 as miR-489 Target known that PEPCK-Cre leads to a significant (70%–80%) but Among the 18 stress-responsive genes, PARP1 is a known cell incomplete gene deletion in kidney proximal tubules.37 Con- death mediator that is activated during ischemic AKI to con- sistently, immunostaining showed that HIF-1a was deleted tribute to renal tubular cell death and tissue damage.36 Thus, from the majority (approximately 80%) but not all of the we examined PARP1 regulation by miR-489 to verify the proximal tubules in our PT-HIF-1a knockout mice (Figure RNA–induced silencing complex (RISC) IP/RNA-seq results. 2D). In addition, in this model, HIF-1a was specifically ab- By sequence analysis with Segal Lab of Computational Biology lated from proximal tubular cells and not ablated from other online software, we identified a potential miR-489 binding site kidney cells, such as the cells of glomeruli and distal tubules. in the 39-UTR of mouse PARP1 gene (Figure 10A). When The proximal tubular cells and other cell types in kidneys that inserted into a luciferase reporter plasmid, this 39-UTR se- were not ablated of HIF-1a could maintain a partial miR-489 quence suppressed luciferase expression in the presence of induction in PT-HIF-1a knockout mice during renal miR-489 (Figure 10B). Furthermore, miR-489 suppressed, ischemia-reperfusion. Our CHIP assay (Figure 3) further pro- whereas anti–miR-489 induced PARP1 protein expression in vides the evidence for direct binding of HIF-1 to one of the

J Am Soc Nephrol 27: 2784–2796, 2016 MicroRNA-489 in Ischemic Kidney Injury 2789 BASIC RESEARCH www.jasn.org

may have long-lasting effects.44 In our study, LNAs were delivered intravenously, and the effect on other organs cannot be excluded. However, structurally, anti-miR LNAs are oligonucleotides, which specifi- cally accumulate in the kidney tubular cells along with the liver.45,46 We further verified that the anti–miR-489 LNA targeted kidney tissues by showing the suppression of miR-489 in kidney tissues (Figure 6B). Functionally, anti–miR-489 LNA increased apoptosis after ATP depletion (Figure 4). Moreover, it enhanced ischemic AKI in mice. These injury enhancing effects of anti–miR-489 support a protective role of miR-489 in hypoxic/ischemic injury. How does miR-489 protect cells and tissues? The key to answering this question is the identification of target genes of miR-489. It is understood that miRs mainly Figure 6. Anti–miR-489 worsens renal function in ischemic AKI. C57BL/6 male mice function by binding to the 39-UTR of were treated with 20 mg/kg scramble or anti–miR-489 LNA and subjected to 25 minutes of bilateral kidney ischemia. (A) Schematics showing the timeline of LNA mRNAs of their target genes, resulting in administration and kidney ischemia injury. I.V., intravenous. (B) Real-time PCR of the the formation of RISC to inhibit mRNA miR-489 level in mouse kidney cortex and outer medulla. (C) Real-time PCR of the translation. In this study, we performed miR-489 level in mouse kidneys cortex and outer medulla without LNA treatment at RISC IP followed by RNA-seq to analyze 24 and 48 hours of reperfusion. (D) The BUN of the mice at different reperfusion times. the targets of miR-489. This method has (E) Serum creatinine at 72 hours of reperfusion (n=3–7). *P,0.05 comparing the remarkable merits for miR target identifi- scramble group with the anti–miR-489 group. cation.47 Specifically, the association of RNAs to RISC induced by a specificmiR is a reliable indication of direct targeting of putative sites in the Calcr/miR-489 gene promoter region. By the RNAs by this miR. In addition, it generates unbiased in- its genomic location, miR-489 seems to be an intronic gene of formation of the global target profile of the miR. By using this Calcr. Together, these in vitro and in vivo studies support a role method, we identified 417 mRNAs that were recruited to RISC of HIF-1 in miR-489 induction during hypoxic/ischemic kid- by miR-489. Additional bioinformatics analysis of these ney injury. MiR-489 may be regulated directly by HIF-1 under mRNAs indicated that 127 of them contain miR-489– these conditions. targeting seed sequence in their 39-UTR. On the basis of these HIF-1 is known to be activated in ischemic AKIand has been observations, we propose that the genes of these 127 mRNAs reported to play a regulatory role in the underlying pathogen- are the targets of miR-489. esis.32,38,39 However, it remains unclear as to which HIF-1– In the 127 target genes of miR-489, 18 are stress response regulated genes are critical to the regulation. In addition to the related. These genes are of particular interest to this study, classic genes that facilitate oxygen delivery, , and because they may account for the observed protective effect of anaerobic metabolism or glycolysis,31,33 recent studies have miR-489. Among the 18 stress-responsive genes, we verified documented HIF-1 regulation of miRs, such as miR-210, PARP1 as one of the miR-489 targets (Figure 10), and the pro- miR-21, miR-29, and miR-127.25,40–42 This study has further tective effect of miR-489 to renal cell apoptosis is associated shown miR-489 induction via HIF-1. Importantly, by using with the reduction of PARP1 level (Figures 5 and 10C). PARP1 anti–miR-489, we have shown a protective role of miR-489 in is a well documented substrate of , and its cleavage is renal tubular cells, which may contribute to HIF-1–mediated considered as a biochemical hallmark of apoptosis. Petrilli regulation of ischemic AKI and kidney injury in general. et al.48 further showed that noncleavable PARP1 knock–in The anti–miR-489 used in our study is in the form of LNA, mice were resistant to ischemic AKI, suggesting a pathogenic in which the ribose ring is locked by a methylene bridge role of PARP1 cleavage. Moreover, a role of PARP1 in ischemic connecting the 29 Oatomandthe49 C atom. This modifica- AKI has been shown by studies using pharmacologic PARP1 tion stabilizes the binding of LNA to its target miR to achieve inhibitors and gene knockout models.49–51 PARP1 activation an optimal inhibition. Sequence-based LNAs have been used may accelerate ATP depletion, induce DNA damage, and in- recently to determine the function of specificmiRsin vivo, hibit glycolysis to induce tubular cell injury and death.49–51 In including the kidneys.23,43 LNA–mediated miR inhibition light of these findings, our results suggest that miR-489 may

2790 Journal of the American Society of Nephrology J Am Soc Nephrol 27: 2784–2796, 2016 www.jasn.org BASIC RESEARCH

protect against ischemic AKI at least partly by suppressing PARP1 expression. Nonethe- less, several other target genes may also me- diate the effect of miR-489. For example, NFAM1 can activate the expression of in- flammatory cytokine TNF-a,52,53 which has been implicated in ischemic AKI.54 In addition, DNAJC3 encodes the procell death protein P58, which is involved in endothe- lium reticulum stresses,55 apotentialpatho- genicresponseinischemicAKI.56 By suppressing these genes, miR-489 may directly promote cell survival in injured kid- neys. In addition, several targets of miR-489 are proinflammatory genes, such as CXCL2 and CXCL10. Therefore, miR-489 may also block the expression of these genes, resulting in a suppression of inflammation and inflammatory tissue damage. Figure 7. Anti–miR-489 increases kidney tubular damage and apoptosis in ischemic In conclusion, this study has shown the – AKI. C57BL/6 mice were treated with scramble LNA or anti miR-489 LNA followed by induction of miR-489 during ischemic AKI 25 minutes of bilateral renal ischemia and 72 hours of reperfusion. (A) The repre- and hypoxic incubation of kidney cells. sentative hematoxylin and eosin images of kidneys. (B) The injured tubular percentage miR-489 induction under these conditions after 25 minutes of ischemia and 72 hours of reperfusion. (C) The representative ter- minal deoxynucleotidyl transferase–mediated digoxigenin-deoxyuridine nick–end la- is mediated by HIF-1. After being induced, beling (TUNEL) images of kidneys. (D) The apoptotic tubular cell numbers in kidneys miR-489 protects kidney cells and tissues from TUNEL images (n=4–6). Scale bar, 0.2 mm. *P,0.05 compared with the scramble against injury. The protective effect of group. miR-489 may involve the 129 target genes identified in this study, especially the 18 cell stress–related genes, including PARP1.

CONCISE METHODS

Animals All mice were housed and used at Charlie Norwood Veterans Affairs Medical Center following a protocol approved by the Institutional Animal Care and Use Committees. C57BL/6J and Hif1atm3Rsjo floxed breeder mice were purchased from The Jackson Laboratory (Bar Harbor, ME). PEPCK-CRE mice were originally obtained from Volker Haase (Vanderbilt University).37 Hif1a tm3Rsjo floxed mice were crossed with PEPCK- CRE mice to obtain animals with PT-HIF-1a knockout mice and wild-type littermates follow- ing an established breeding protocol.21

RISC IP RISC IP was performed according to that in Figure 8. Anti–miR-489 does not affect tubular proliferation after ischemic AKI. previous studies,57–59 with some modifications. C57BL/6 mice were treated with scramble LNA or anti–miR-489 LNA followed by 25 minutes of bilateral renal ischemia and 72 hours of reperfusion. (A) Represen- HEK293T cells were seeded in 100-mm dishes m tative images of immunohistochemistry of Ki67. Lower panels are the magnified and cotransfected with 10 g FLAG-Ago-2 plas- images of the boxed areas from upper panels. Scale bar, 0.2 mm. (B) The Ki67– mid (Addgene, Cambridge, MA) or control positive cell numbers in kidneys with 25 minutes of ischemia and 72 hours of plasmid (p3xFLAG-Myc-CMV-24 Expression reperfusion (n=4). Vector;Sigma-Aldrich,St.Louis,MO)and

J Am Soc Nephrol 27: 2784–2796, 2016 MicroRNA-489 in Ischemic Kidney Injury 2791 BASIC RESEARCH www.jasn.org

Figure 9. Identification of miR-489 targets by Ago-2 IP and RNA-seq. (A) HEK293 cells were transfected with Flag-Ago-2 expression plasmid or control plasmid. The lysates were subjected to IP with anti-Flag antibody. Western blot of Ago-2 was performed to verify the successful IP of Ago-2. (B–D) HEK293 cells were cotransfected with Flag-Ago-2 and miR-489 RNA mimic (489) or negative control RNA oligos (NC). The lysates were subjected to IP with anti-Flag. The total RNA directly from cells or the RNAs in the IP products were extracted for RNA-seq. (B) Real-time PCR to show the enrichment of miR-489 in the IP products. The miR-489 levels were compared between the 489 group and the NC group (n=3). (C) Diagram showing the different types of genes identified by RNA-seq. (D) List of the 127 genes as miR-489 targets. These coding genes were recruited to Ago-2 RISC by miR-489 and contained predicted miR-489 target sequence in their 39-UTR. The genes are categorized in the table according to the ratio between their mRNA levels in the presence of miR-489 or the negative control oligo. *Eighteen genes related to stress response.

2792 Journal of the American Society of Nephrology J Am Soc Nephrol 27: 2784–2796, 2016 www.jasn.org BASIC RESEARCH

RNA-seq and Data Analyses Three sets of RNA samples from HEK293 cells with or without RISC IP were subjected to RNA- seq as described previously.60 Briefly, mRNA (cDNA) libraries were built with 5 mgRNA followed by cluster generation and sequencing. The RNA-seq data were analyzed by an in-house pipeline, which included reads mapping and alignment, transcript construction, quantifica- tion of transcript abundance, and identification of differential expression using Bowtie (http:// bowtie-bio.sourceforge.net/index.shtml), Tophat v2 (http://tophat.cbcb.umd.edu/), and Cufflinks (http://cufflinks.cbcb.umd.edu/). Dif- ferential expression of transcripts was identified using false discover rate ,0.05. The mRNA reads that showed a more than twofold increase in the hsa-miR-489 group compared with the negative control group in the precipitated RNA samples Figure 10. Confirmation of PARP1 as an miR-489 target. (A) Prediction of mmu-miR- were considered to be potential miR-489 binding 9 489 binding site in the 3 -UTR of PARP1. The potential binding site (bold) was mRNAs. predicted by the Segal Lab of Computational Biology with the input of the indicated 39-UTR of mouse PARP1 and mmu-miR-489–3p sequences. (B) Target reporter lucif- erase assay. HEK293 cells were cotransfected with miR mimics and the pMIR-REPORT Chromatin Immunoprecipitation miRNA Expression Reporter Vector with or without the insert of mouse PARP1 39-UTR. BUMPT cells with or without 6 hours of hypoxia The luciferase activity ratios (mmu-miR-489 mimic-to-negative control mimic ratios) treatment were fixed with 0.75% formaldehyde were calculated to indicate the fold change, and the ratio with empty vector and then neutralized with glycine. Then, the cells cotransfection was used for normalization (n=4). *Statistically significant difference were collected in cold PBS and pelleted with comparing the mmu-miR-489 group with the negative control group. (C) Effects of 10003g centrifugation for 5 minutes at 4°C. The miR-489 and anti–miR-489 on PARP1 expression. RPTCs were transfected with 200 nM cell pellet was resuspended and lysed in the son- miR-489 mimic, LNA, or control sequences and then subjected to ATP depletion and ication buffer (5 mM HEPES, pH 7.9, 140 mM recovery. Full shows the full-length band of PARP1, and cleaved shows the cleaved NaCl, 1 mM EDTA, 1% Triton X-100, 0.1% Na- band of PARP1. Cyclophilin B was used as the loading control. AZ, ATP depletion with deoxycholate, and 0.1% SDS) for 10 minutes at 10 mM azide treatment; Con, control conditions without ATP depletion; ddG, the 4°C, and then, it was sonicated for 15 seconds difference between dGduplex and dGopen; dGduplex, the energy score to form duplex between a miRNA and its target; dGopen, the energy score without duplex each six times to shear the DNA. After centrifu- 3 forming; NC, negative control miR; Scr, scramble LNA. gation at 8000 g for 30 seconds at 4°C, the chromatin samples were collected from the supernatant; 50-ml aliquots of the chromatin 100 nM scramble RNA oligos or hsa-miR-668 mimic (Dharmacon samples were used for INPUT DNA extraction. After incubation Research, Inc., Pittsburgh, PA). After 48 hours, cells were used for with 1 ml 1 mg/ml RNase A (Qiagen, Valencia, CA) at 65°C overnight, either total RNA extraction as described above or RISC IP. the INPUT DNA was extracted with the QIAquick PCR Purification For RISC IP,cells were lysed in 800 ml lysis buffer (150 mM KCl, 25 mM Kit (Qiagen). According to the DNA concentration of the INPUT Tris-HCl,pH7.4,5mMEDTA,and0.5%NP-40withfreshlyadded5mM samples, the chromatin samples were diluted with sonication buffer dithiothreitol, 1:1000 proteinase cocktail, and 100 U/ml SUPERase; Life to 1 or 2 mg/ml DNA and incubated at 4°C with 1 mg/ml BSA and Technologies, Grand Island, NY) for 30 minutes at 4°C. The cell lysates 40 ml protein A agarose beads (Santa Cruz Biotechnology, Santa Cruz, were centrifuged for 20 minutes at 14,0003g at 4°C to collect the super- CA), which were prebalanced with sperm DNA and BSA in 1 ml natant, which was further filtered through a 0.45-mmsyringefilter. reaction buffer. After 2 hours of incubation, the beads were removed Anti–FLAG-resin (Sigma-Aldrich) was equilibrated according to by centrifugation, and 4 mganti–HIF-1a antibody was added to the the manufacturer’s instruction in lysis buffer, mixed with the col- reaction for overnight incubation. Finally, the antibody was precip- lected cell lysate at the ratio of 300 mlresinto800ml lysate, and then itated with protein A agarose bead. After serial washes in sonication incubated at 4°C for 4 hours with rotation. The resin beads were then buffer, buffer A (50 mM HEPES, pH 7.9, 500 mM NaCl, 1 mM EDTA, washed twice with 103 volume lysis buffer for 5 minutes each. At the 1% Triton X-100, 0.1% Na-deoxycholate, and 0.1% SDS), buffer B last wash, about 5% beads were saved to elute proteins with SDS lysis (20 mM Tris-HCl, pH 8.0, 1 mM EDTA, 250 mM LiCl, 0.5% NP-40, buffer for immunoblotting. The other beads were suspended in and 0.5% Na-deoxycholate), and TE buffer, the DNA was eluted with 600 ml RNA lysis buffer from the miRVana miRNA Isolation Kit 50mMTris-HCl,pH8.0,1mMEDTA,1%SDS,and50mM for RNA extraction according to the manufacturer’s instruction. NaHCO3. The eluted DNA was purified with the QIAquick PCR

J Am Soc Nephrol 27: 2784–2796, 2016 MicroRNA-489 in Ischemic Kidney Injury 2793 BASIC RESEARCH www.jasn.org

Purification Kit after RNase A treatment. The precipitated DNA sam- ACKNOWLEDGMENTS ples were quantified by real-time PCR with INPUT DNA for normal- ization. The following primers were used for HIF-1 binding sites This study was supported by National Natural Science Foundation of detection: miR-489 F: CAA GAG GTT CCA GAC ATA CAT AGA T; China grants 81430017 and 81370791, American Heart Associa- miR-489 R: GCC ATA TAG CCT AGC AGT CTT AC; calcr site 1 tion grant 12SDG8270002, National Institutes of Health grants F: CTT GGG CAT CCC ACA GTA TT; calcr site 1 R: GTG TAC 2R01DK058831 and 1R01DK087843, and Department of Veterans AAC TCC GTG GCT TTA; calcr site 2 F: CGG AGT TGT ACA Administration grant 5I01BX000319. CGA CTT AGA TAA A; calcr site 2 R: GCC TAT GCT TAC ACC ACA CA; VEGF F: CCC ACC TCA CAA ACA CAC TAC; and VEGF R: GGT GTG CAT AAT GTA GTC ACT AGG. DISCLOSURES None. Luciferase Activity Assay to Evaluate miR Targets The 39-UTR of the mouse PARP1 gene was synthesized 9 (Genescript, Piscataway, NJ) and inserted into the 3 -UTR of the REFERENCES luciferase gene of the pMIR-REPORT Luciferase plasmid (Life Technologies). The plasmids with or without the insert were 1. Lameire NH, Bagga A, Cruz D, De Maeseneer J, Endre Z, Kellum JA, Liu cotransfected with pMIR-REPORT b-gal Control Plasmid (Life KD, Mehta RL, Pannu N, Van Biesen W, Vanholder R: Acute kidney Technologies) and 200 nM miR mimics (negative control versus injury: An increasing global concern. Lancet 382: 170–179, 2013 mmu-miR-489 mimics; Life Technologies) into HEK293 cells. One 2. Okusa MD, Chertow GM, Portilla D; Advisory Group of the American Society of Nephrology: The nexus of acute day after the transfection, the lysate was collected in reporter lysis kidney injury, chronic kidney disease, and World Kidney Day 2009. Clin buffer from the Luciferase Assay System (Promega, Madison, WI). J Am Soc Nephrol 4: 520–522, 2009 The luciferase activity was measured by the Luciferase Assay 3. Venkatachalam MA, Griffin KA, Lan R, Geng H, Saikumar P, Bidani AK: System, and the b-galactosidase activity was measured by the Acute kidney injury: A springboard for progression in chronic kidney b-Galactosidase Enzyme Assay System (Promega). The luciferase disease. Am J Physiol Renal Physiol 298: F1078–F1094, 2010 activity was normalized with b-galactosidase activity. The lucifer- 4. Sharfuddin AA, Molitoris BA: Pathophysiology of ischemic acute kidney injury. Nat Rev Nephrol 7: 189–200, 2011 ase ratio between mmu-miR-489 and control empty vector was 5. Bonventre JV, Yang L: Cellular pathophysiology of ischemic acute considered as one, and the ratio of the pMIR-REPORT Luciferase kidney injury. JClinInvest121: 4210–4221, 2011 plasmid to the PARP1 39-UTR insert was compared with evaluate 6. Kinsey GR, Okusa MD: Pathogenesis of acute kidney injury: Foundation the inhibitory effect of mmu-miR-489. for clinical practice. Am J Kidney Dis 58: 291–301, 2011 7. Linkermann A, Chen G, Dong G, Kunzendorf U, Krautwald S, Dong Z: – In Situ Regulated cell death in AKI. JAmSocNephrol25: 2689 2701, 2014 Hybridization 8. Molitoris BA: Therapeutic translation in acute kidney injury: The fi Mice were perfuse xed with 4% paraformaldehyde, and the kidney epithelial/endothelial axis. JClinInvest124: 2355–2363, 2014 samples were further fixedovernightat4°C.Afterbalancedin20%su- 9. Mendell JT, Olson EN: MicroRNAs in stress signaling and human crose in PBS overnight at 4°C, the kidney samples were frozen embedded disease. Cell 148: 1172–1187, 2012 for 7-mm cryosections. The fresh cryosections were dried for 20 seconds 10. Bartel DP: MicroRNAs: Target recognition and regulatory functions. – in air and fixed in 4% paraformaldehyde for 20 minutes at room tem- Cell 136: 215 233, 2009 m 11. Ho J, Kreidberg JA: MicroRNAs in renal development. Pediatr Nephrol perature. After three PBS washes, the specimens were treated with 1 g/ml 28: 219–225, 2013 proteinase K for 10 minutes followed with a PBS rinse. Then, miR-489 12. Nagalakshmi VK, Ren Q, Pugh MM, Valerius MT, McMahon AP, Yu J: was detected with the IsHyb In Situ Hybridization Kit (Biochain Institute, Dicer regulates the development of nephrogenic and ureteric com- Inc., Newwark, CA) following the manufacturer’s instruction. Briefly, the partments in the mammalian kidney. Kidney Int 79: 317–330, 2011 slides were incubated with prehybridization solution at 70°C for 3 hours. 13. Ho J, Ng KH, Rosen S, Dostal A, Gregory RI, Kreidberg JA: Podocyte- specific loss of functional microRNAs leads to rapid glomerular and The hybridization mixture was then prepared by mixing digoxigenin– tubular injury. JAmSocNephrol19: 2069–2075, 2008 labeled LNA probe and hybridization solution to detect mmu-miR-489 14. Patel V, Hajarnis S, Williams D, Hunter R, Huynh D, Igarashi P: (Exiqon, Inc., Woburn, MA). The hybridization mixture was heated at MicroRNAs regulate renal tubule maturation through modulation of 65°C for 5 minutes and cooled on ice. After overnight hybridization with Pkd1. J Am Soc Nephrol 23: 1941–1948, 2012 the hybridization mixture at 65°C, the specimens were washed with serial 15. Sequeira-Lopez ML, Weatherford ET, Borges GR, Monteagudo MC, Pentz ES, Harfe BD, Carretero O, Sigmund CD, Gomez RA: The dilutions of SSC buffer and incubated with antidigoxigenin – fi microRNA-processing enzyme dicer maintains juxtaglomerular cells. alkaline phosphatase conjugated antibody. The nal signal JAmSocNephrol21: 460–467, 2010 was detected by NBT/BCIP solution and analyzed by light microscopy. 16. Shi S, Yu L, Chiu C, Sun Y, Chen J, Khitrov G, Merkenschlager M, Holzman LB, Zhang W, Mundel P, Bottinger EP: Podocyte-selective deletion of dicer induces proteinuria and glomerulosclerosis. JAmSoc Other Analytical Methods – Animal and cell models, RNA extraction, immunoblotting, immu- Nephrol 19: 2159 2169, 2008 fl 17. Chung AC, Huang XR, Meng X, Lan HY: miR-192 mediates TGF-beta/ nohistochemical and immuno uorescence staining, caspase activity Smad3-driven renal fibrosis. JAmSocNephrol21: 1317–1325, 2010 measurement, and data analyses are described in Supplemental 18. Kato M, Zhang J, Wang M, Lanting L, Yuan H, Rossi JJ, Natarajan R: Material. MicroRNA-192 in diabetic kidney glomeruli and its function in

2794 Journal of the American Society of Nephrology J Am Soc Nephrol 27: 2784–2796, 2016 www.jasn.org BASIC RESEARCH

TGF-beta-induced collagen expression via inhibition of E-box repres- 38. Conde E, Alegre L, Blanco-Sánchez I, Sáenz-Morales D, Aguado-Fraile sors. Proc Natl Acad Sci U S A 104: 3432–3437, 2007 E, Ponte B, Ramos E, Sáiz A, Jiménez C, Ordoñez A, López-Cabrera M, 19. Krupa A, Jenkins R, Luo DD, Lewis A, Phillips A, Fraser D: Loss of del Peso L, de Landázuri MO, Liaño F, Selgas R, Sanchez-Tomero JA, MicroRNA-192 promotes fibrogenesis in diabetic nephropathy. JAm García-Bermejo ML: Hypoxia inducible factor 1-alpha (HIF-1 alpha) is Soc Nephrol 21: 438–447, 2010 induced during reperfusion after renal ischemia and is critical for 20. Long J, Wang Y, Wang W, Chang BH, Danesh FR: MicroRNA-29c is a cell survival. PLoS One 7: e33258, 2012 signature microRNA under high glucose conditions that targets 39. Hill P, Shukla D, Tran MG, Aragones J, Cook HT, Carmeliet P, Maxwell Sprouty homolog 1, and its in vivo knockdown prevents progression of PH: Inhibition of hypoxia inducible factor hydroxylases protects against diabetic nephropathy. JBiolChem286: 11837–11848, 2011 renal ischemia-reperfusion injury. J Am Soc Nephrol 19: 39–46, 2008 21. Wei Q, Bhatt K, He HZ, Mi QS, Haase VH, Dong Z: Targeted deletion of 40. Devlin C, Greco S, Martelli F, Ivan M: miR-210: More than a silent player Dicer from proximal tubules protects against renal ischemia-reperfusion in hypoxia. IUBMB Life 63: 94–100, 2011 injury. J Am Soc Nephrol 21: 756–761, 2010 41. Fang Y, Yu X, Liu Y, Kriegel AJ, Heng Y, Xu X, Liang M, Ding X: miR-29c 22. Godwin JG, Ge X, Stephan K, Jurisch A, Tullius SG, Iacomini J: Iden- is downregulated in renal interstitial fibrosis in humans and rats and tification of a microRNA signature of renal ischemia reperfusion injury. restored by HIF-a activation. Am J Physiol Renal Physiol 304: Proc Natl Acad Sci U S A 107: 14339–14344, 2010 F1274–F1282, 2013 23. Lorenzen JM, Kaucsar T, Schauerte C, Schmitt R, Rong S, Hübner A, 42. Aguado-Fraile E, Ramos E, Sáenz-Morales D, Conde E, Blanco-Sánchez Scherf K, Fiedler J, Martino F, Kumarswamy R, Kölling M, Sörensen I, I, Stamatakis K, del Peso L, Cuppen E, Brüne B, Bermejo ML: miR-127 Hinz H, Heineke J, van Rooij E, Haller H, Thum T: MicroRNA-24 an- protects proximal tubule cells against ischemia/reperfusion: Identifi- tagonism prevents renal ischemia reperfusion injury. JAmSocNephrol cation of kinesin family member 3B as miR-127 target. PLoS One 7: 25: 2717–2729, 2014 e44305, 2012 24. Kaucsár T, Révész C, Godó M, Krenács T, Albert M, Szalay CI, Rosivall L, 43. Bhatt K, Wei Q, Pabla N, Dong G, Mi QS, Liang M, Mei C, Dong Z: Benyó Z, Bátkai S, Thum T, Szénási G, Hamar P: Activation of the miR-17 MicroRNA-687 induced by hypoxia-inducible factor-1 targets phos- family and miR-21 during murine kidney ischemia-reperfusion injury. phatase and tensin homolog in renal ischemia-reperfusion injury. JAm Nucleic Acid Ther 23: 344–354, 2013 Soc Nephrol 26: 1588–1596, 2015 25. Xu X, Kriegel AJ, Liu Y, Usa K, Mladinov D, Liu H, Fang Y, Ding X, Liang 44. Elmén J, Lindow M, Schütz S, Lawrence M, Petri A, Obad S, Lindholm M: Delayed ischemic preconditioning contributes to renal protection M, Hedtjärn M, Hansen HF, Berger U, Gullans S, Kearney P, Sarnow P, by upregulation of miR-21. Kidney Int 82: 1167–1175, 2012 Straarup EM, Kauppinen S: LNA-mediated microRNA silencing in non- 26. Lan YF, Chen HH, Lai PF, Cheng CF, Huang YT, Lee YC, Chen TW, Lin H: human primates. Nature 452: 896–899, 2008 MicroRNA-494 reduces ATF3 expression and promotes AKI. JAmSoc 45. Rappaport J, Hanss B, Kopp JB, Copeland TD, Bruggeman LA, Nephrol 23: 2012–2023, 2012 Coffman TM, Klotman PE: Transport of phosphorothioate oligonucle- 27. Cheung TH, Quach NL, Charville GW, Liu L, Park L, Edalati A, Yoo B, otides in kidney: Implications for molecular therapy. Kidney Int 47: Hoang P, Rando TA: Maintenance of muscle stem-cell quiescence by 1462–1469, 1995 microRNA-489. Nature 482: 524–528, 2012 46. Hamar P, Song E, Kökény G, Chen A, Ouyang N, Lieberman J: Small 28. Kikkawa N, Hanazawa T, Fujimura L, Nohata N, Suzuki H, Chazono H, interfering RNA targeting Fas protects mice against renal ischemia- Sakurai D, Horiguchi S, Okamoto Y, Seki N: miR-489 is a tumour- reperfusion injury. Proc Natl Acad Sci U S A 101: 14883–14888, 2004 suppressive miRNA target PTPN11 in hypopharyngeal squamous cell 47. Kuzembayeva M, Chiu YF, Sugden B: Comparing proteomics and RISC carcinoma (HSCC). Br J Cancer 103: 877–884, 2010 immunoprecipitations to identify targets of Epstein-Barr viral miRNAs. 29. Munari E, Marchionni L, Chitre A, Hayashi M, Martignoni G, Brunelli M, PLoS One 7: e47409, 2012 Gobbo S, Argani P, Allaf M, Hoque MO, Netto GJ: Clear cell papillary 48. Pétrilli V, Herceg Z, Hassa PO, Patel NS, Di Paola R, Cortes U, Dugo L, renal cell carcinoma: Micro-RNA expression profiling and comparison Filipe HM, Thiemermann C, Hottiger MO, Cuzzocrea S, Wang ZQ: with clear cell renal cell carcinoma and papillary renal cell carcinoma. Noncleavable poly(ADP-ribose) polymerase-1 regulates the inflammation Hum Pathol 45: 1130–1138, 2014 response in mice. J Clin Invest 114: 1072–1081, 2004 30. Wang K, Liu F, Zhou LY, Long B, Yuan SM, Wang Y, Liu CY, Sun T, Zhang 49. Devalaraja-Narashimha K, Padanilam BJ: PARP-1 inhibits glycolysis in XJ, Li PF: The long noncoding RNA CHRF regulates cardiac hypertro- ischemic kidneys. JAmSocNephrol20: 95–103, 2009 phy by targeting miR-489. Circ Res 114: 1377–1388, 2014 50. Zheng J, Devalaraja-Narashimha K, Singaravelu K, Padanilam BJ: Poly 31. Semenza GL: Oxygen sensing, hypoxia-inducible factors, and disease (ADP-ribose) polymerase-1 gene ablation protects mice from ischemic pathophysiology. Annu Rev Pathol 9: 47–71, 2014 renal injury. Am J Physiol Renal Physiol 288: F387–F398, 2005 32. Rosenberger C, Mandriota S, Jürgensen JS, Wiesener MS, Hörstrup JH, 51. Chatterjee PK, Zacharowski K, Cuzzocrea S, Otto M, Thiemermann C: Frei U, Ratcliffe PJ, Maxwell PH, Bachmann S, Eckardt KU: Expression of Inhibitors of poly (ADP-ribose) synthetase reduce renal ischemia- hypoxia-inducible factor-1alpha and -2alpha in hypoxic and ischemic reperfusion injury in the anesthetized rat in vivo. FASEB J 14: 641–651, rat kidneys. J Am Soc Nephrol 13: 1721–1732, 2002 2000 33. Haase VH: Mechanisms of hypoxia responses in renal tissue. JAmSoc 52. Ohtsuka M, Arase H, Takeuchi A, Yamasaki S, Shiina R, Suenaga T, Nephrol 24: 537–541, 2013 Sakurai D, Yokosuka T, Arase N, Iwashima M, Kitamura T, Moriya H, 34. Lieberthal W, Nigam SK: Acute renal failure. II. Experimental models of Saito T: NFAM1, an immunoreceptor tyrosine-based activation motif- acute renal failure: Imperfect but indispensable. Am J Physiol Renal bearing molecule that regulates B cell development and signaling. Physiol 278: F1–F12, 2000 Proc Natl Acad Sci U S A 101: 8126–8131, 2004 35. Yang L, Besschetnova TY, Brooks CR, Shah JV, Bonventre JV: Epithelial 53. Yang J, Hu G, Wang SW, Li Y, Martin R, Li K, Yao Z: Calcineurin/nuclear cell cycle arrest in G2/M mediates kidney fibrosis after injury. Nat Med factors of activated T cells (NFAT)-activating and immunoreceptor 16: 535–543, 2010 tyrosine-based activation motif (ITAM)-containing protein (CNAIP), a 36. Martin DR, Lewington AJ, Hammerman MR, Padanilam BJ: Inhibition of novel ITAM-containing protein that activates the calcineurin/NFAT- poly(ADP-ribose) polymerase attenuates ischemic renal injury in rats. signaling pathway. JBiolChem278: 16797–16801, 2003 Am J Physiol Regul Integr Comp Physiol 279: R1834–R1840, 2000 54. Donnahoo KK, Meng X, Ayala A, Cain MP, Harken AH, Meldrum DR: 37. Rankin EB, Tomaszewski JE, Haase VH: Renal cyst development in mice Early kidney TNF-alpha expression mediates infiltration and with conditional inactivation of the von Hippel-Lindau tumor suppres- injury after renal ischemia-reperfusion. Am J Physiol 277: R922–R929, sor. Cancer Res 66: 2576–2583, 2006 1999

J Am Soc Nephrol 27: 2784–2796, 2016 MicroRNA-489 in Ischemic Kidney Injury 2795 BASIC RESEARCH www.jasn.org

55. Petrova K, Oyadomari S, Hendershot LM, Ron D: Regulated association 59. Karginov FV, Conaco C, Xuan Z, Schmidt BH, Parker JS, Mandel G, of misfolded lumenal proteins with P58/ Hannon GJ: A biochemical approach to identifying microRNA targets. DNAJc3. EMBO J 27: 2862–2872, 2008 Proc Natl Acad Sci U S A 104: 19291–19296, 2007 56. Yang JR, Yao FH, Zhang JG, Ji ZY, Li KL, Zhan J, Tong YN, Lin LR, He YN: 60. Liu Y, Liu P, Yang C, Cowley AW Jr., Liang M: Base-resolution maps Ischemia-reperfusion induces renal tubule pyroptosis via the CHOP- of 5-methylcytosine and 5-hydroxymethylcytosine in Dahl S rats: caspase-11 pathway. Am J Physiol Renal Physiol 306: F75–F84, 2014 Effect of salt and genomic sequence. 63: 827–838, 57. Beitzinger M, Meister G: Experimental identification of microRNA tar- 2014 gets by immunoprecipitation of Argonaute protein complexes. Methods Mol Biol 732: 153–167, 2011 58. HendricksonDG,HoganDJ,Herschlag D, Ferrell JE, Brown PO: Systematic identification of mRNAs recruited to argonaute 2 by specificmicroRNAsand This article contains supplemental material online at http://jasn.asnjournals. corresponding changes in transcript abundance. PLoS One 3: e2126, 2008 org/lookup/suppl/doi:10.1681/ASN.2015080870/-/DCSupplemental.

2796 Journal of the American Society of Nephrology J Am Soc Nephrol 27: 2784–2796, 2016