Molecular Changes in miRNA in Irradiated Rat Kidneys: Role of miR-34a and its Vascular Targets in the Notch Pathway.

MicroRNA libraries (K1–K13) from kidneys of rats 35 days postirradiation and age-matched, nonirradiated controls were prepared and analyzed by next-generation sequencing ( Supplementary Table S1 ; https://doi.org/10.1667/RADE-20-00078.1.S1 ). The most abundant mature microRNAs with read counts from each control rat >5,000 are listed in Supplementary Table S2 ; https://doi.org/10.1667/RADE-20-;00078.1.S1 . The mean read counts were calculated from the four libraries of control groups (K1–K4). Results were compared with the human kidney microRNA profile reported by Ben-Dov et al . ( 15 ). The top 16 most abundant microRNAs in rat kidney were also among the most abundant (top 0.02%) of microRNA reported in data for the human kidney. The exceptions were miR-3596a and miR-3557–5p, which are specific to rats only and have not been reported in humans ( Supplementary Table S2 ). Next-generation rat kidney sequence data were further analyzed by comparing the expression of individual microRNA between groups using t tests. The t tests were performed separately to compare nonirradiated and irradiated rats, and to evaluate the effects of lisinopril. Differences in expression of microRNA between irradiated rats with and without lisinopril were also determined. Although many sequences were changed ( P < 0.05) after irradiation or lisinopril, miR-34a-5p was the only mature microRNA among these with a sequence that had been previously confirmed and characterized in rats. The increased expression of this microRNA in irradiated rats compared to nonirradiated control rats is described in Table 1 . Sequence data were also used for Ingenuity Pathway Analysis (IPA). Comparisons were made between nonirradiated (0 Gy) and 13 Gy irradiated, between 0 Gy and 13 Gy plus lisinopril and between 13 Gy and 13 Gy plus lisinopril treatment groups. When comparison was done between 0 Gy and 13 Gy, miR-34a-5p is the only focus molecule (shown in red) in the network recognized by IPA ( Fig. 1A ). IPA results show that, in addition to miR-34a-5p, miR-431–5p and miR-494–3p are focus molecules, when comparing 0 Gy with 13 Gy plus lisinopril ( Fig. 1B ). When comparing 13 Gy with 13 Gy plus lisinopril, miR-431–5p is the only focus molecule in a small network that contains only five molecules ( Fig. 1C ). In these studies, we considered all the major networks with defined pathways. MiR-34a-5p is a focus molecule when comparison was made between controls and all irradiated rats with and without lisinopril, while miR-431–5p is a focus of both pathways between irradiation plus lisinopril versus either control or irradiated-only groups ( Fig. 1 ). Here we focused on miR-34a-5p, since it showed consistent read counts in all samples ( Table 1 ), while miR-431–5p was not expressed in some samples (data not shown). MiR-34a-5p was further tested using kidney RNA derived from rats used for the library preparation and sequencing. Real-time quantitative RT-qPCR results confirmed the increase of miR-34a-5p after irradiation ( Fig. 2 ). Another two independent batches of rat RNA were used to confirm these results. Though the ANOVA used to compare the three groups with multiple comparisons did not reach significance for a difference in miR-34a-5p between irradiated-only rats versus the irradiation plus lisinopril group, the two groups were different by Mann-Whitney rank sum test ( Table 2 ). After confirming the increase of miR-34a-5p after irradiation, we searched for its targets. MiR-34a-5p is a well-studied microRNA and according to the Ingenuity Knowledge Base, it has been associated with many diseases and molecules ( Table 3 ) including the Notch pathway. Since we are interested in vascular disease caused by radiation, we investigated the roles of miR-34a-5p in the Notch pathway, a pathway involved in angiogenesis, vessel regression and cardiovascular diseases ( 21 ). Information from the Ingenuity Knowledge Base shows that miR-34a-5p regulates Notch2 and binds Notch1 and Jagged1 ( Table 3 ). The pairings of the targets of interest based on complimentary sequence ( Jagged1 , Notch1 and Notch2 , with miR-34a-5p), were checked on targetscan.org and the predicted consequential target regions are shown in Table 4 . Targets of interest were tested at transcriptional (mRNA) and translational (protein) levels by RT-qPCR and Western blotting, respectively ( Supplementary Table S3 ; https://doi.org/10.1667/RADE-20-00078.1.S1 ). Jagged1 protein was decreased after irradiation but not in irradiated rats given lisinopril. Irradiated-only rats were also different from the irradiation plus lisinopril group as indicated by Mann-Whitney rank sum test, confirming mitigation ( Fig. 3 , Table 2 ). No decrease in Jagged1 mRNA was observed after irradiation ( Fig. 3 ). Notch2 mRNA was decreased after irradiation only but not in the irradiated rats given lisinopril. There was also no difference between the irradiated-only and the irradiation plus lisinopril group when these two groups were compared to each other only ( Fig. 3 and Table 2 ). Notch2 was not shown to be downregulated at the protein level ( Fig. 3 ). Notch1 did not change at either mRNA or protein level after irradiation (data not shown). To confirm that vascular regression occurred in the kidney after irradiation and determine if these changes were mitigated by lisinopril, we stained kidney sections with a vascular endothelial-specific antibody. Immunohistochemistry staining for endothelial cells by RECA-1 in kidneys at 100 days postirradiation (during radiation nephropathy) confirmed a decrease in endothelial cells in irradiated rats that was mitigated by lisinopril ( Fig. 4 and Table 2 ). In the renal medulla there were more endothelial cells marking peritubular vasculature in control rats than irradiated rats. In the cortex, the control rats and lisinopril-treated rats showed more endothelial cells in glomeruli, while the glomeruli in irradiated-only rats had fewer endothelial cells. Results of quantification of the images using ImageJ software showed a decrease in endothelial cells after irradiation and mitigation of the decrease by lisinopril in both the medulla and the cortex.

Female WAG/RijCmcr rats (11–12 weeks old) were irradiated and housed in an isolated barrier in the animal facilities of the Medical College of Wisconsin (MCW; Wauwatosa, WI). Rats were irradiated as described previously ( 4 ) with 13 Gy to the whole body, except for one hind limb that was externalized from the chamber and shielded (leg-out PBI). Dosimetry was conducted as previously reported elsewhere by us ( 16 ). All animal procedures were performed at MCW under protocols approved by the Institutional Animal Care and Use Committee. MCW is accredited by Association for the Assessment and Accreditation of Laboratory Animal Care (AAALAC). Rats were randomized into three groups: 1. nonirradiated controls (0 Gy); 2. irradiation-only group (13 Gy); and 3. irradiation with lisinopril group (13 Gy + lisinopril). Lisinopril was added to the drinking water at a dose of ~24 mg m −2 day −1 , which is comparable to that approved for clinical use. The drug was started from day 7 postirradiation and continued until termination of the study. Rats were also given daily subcutaneous injection of saline (40 ml kg −1 day −1 ) from days 3–7 and enrofloxacin (10 mg kg −1 day −1 ) from 1–14 days postirradiation as supportive care ( 4 ). Experiments were conducted in batches, each batch consisting of irradiated, irradiated with lisinopril and age-matched nonirradiated rats. The first batch was used for microRNA-seq. MicroRNA and targets were examined at day 35 postirradiation, which is before the onset of radiation pneumonitis or nephropathy. Data from the sequence result were validated in the same as well as in two other experiments using rats that were independent of the sequence study. Kidneys were harvested from rats that were anesthetized with isoflurane. The abdominal aorta was isolated and cannulated, and the kidneys were perfused with saline before they were removed ( 17 , 18 ). Kidneys from multiple batches were used for Western blotting. An independent batch was euthanized at 100 days when nephropathy had developed, and these kidneys were used for histological staining of vascular endothelium. Fresh kidney tissue (~100 mg) was cut and weighed. Total RNA was isolated with TRIzol ® reagent (Ambion ® /Life Technologies, Austin, TX) following the protocol provided by the manufacturer. RNA was stored at −80°C for future use. RNA concentration was measured using NanoDrop 2000 (Thermo Scientific ™ , Rockford, IL). Small RNA libraries (13 in total; 0 Gy=4; 13 Gy=4; 13 Gy + lisinopril=5) were generated from each rat using the TruSeq ® Small RNA Library Prep kit (Illumina ® , San Diego, CA) ( 19 ). Libraries were visualized and quantitated with the Agilent Technologies 2100 Bioanalyzer (Santa Clara, CA), using DNA-1000 Chip. Next-generation sequencing was performed at the Human and Molecular Genetics Center Sequencing Core at MCW ( 19 ). Small RNA sequencing data were analyzed following protocol as described elsewhere ( 19 ). Differential expression of microRNAs was analyzed between 0 Gy and 13 Gy, between 0 Gy and 13 Gy with lisinopril and between 13 Gy and 13 Gy with lisinopril with P = 0.05 as cut-off. Enrichment analysis of differentially expressed microRNAs was conducted using Ingenuity ® Pathway Analysis (IPA ® ; QIAGEN ® , Redwood City, CA). Targets of miR-34a-5p were predicted using TargetScan with default mammalian parameters. The targets of interest were prioritized by Ingenuity Knowledge Base using IPA and by relevance to our study of the biology of vascular regression. Validation of changes in RNA expression ( 19 ) of selected markers was performed by reverse transcription followed by quantitative PCR (RT-qPCR) from rats used for miRNA-seq (first batch) as well as from independent experimental batches. miRCURY LNATM Universal RT microRNA PCR kit (for miRNA, Exiqon, Vedbaek, Denmark) and iScript ™ cDNA synthesis kit (for mRNA, Bio-Rad ® Laboratories Inc., Hercules, CA) were used for reverse transcription. LNA-primers (for miRNA, Exiqon), IDT PrimeTime qPCR primers [for mRNA, Integrated DNA Technologies (IDT ® ), Coralville, IA] and SYBR ® Green Mastermix (BioTools Inc., Jupiter, FL) were used for qPCR. The relative expression of target RNA to references (miRNA to miR-16–5p and mRNA to b-actin) were calculated by the formula: 2 −ΔCT (ΔCT=CT (target-reference)). Data were then normalized to the mean relative expression of controls, to determine the fold change after irradiation. Harvested kidney tissue (~80 mg) was weighed and homogenized with RIPA buffer (MilliporeSigma, St. Louis, MO) containing protease and phosphatase inhibitors. A total of 40 μg protein from each homogenized sample was boiled for 5 min in Laemmli sample buffer (Bio-Rad), separated on 7.5% polyacrylamide gel electrophoreses (PAGE) (Bio-Rad), and transferred to a nitrocellulose membrane. The blots were blocked for 1 h at room temperature with 5% bovine serum albumin (BSA), developed with appropriate primary antibodies and matched secondary antibodies conjugated to horseradish peroxidase and visualized by enhanced chemiluminescence. At 100 days postirradiation, kidneys were harvested, fixed and embedded in paraffin ( 20 ). Whole mount kidney sections (4 μm thick) were stained with RECA-1 (Novus Biologicals LLC, Littleton, CO), a rat endothelial cell-specific antibody. Embedding, cutting and staining were done by the Histology Core of the Children’s Hospital of Wisconsin (Wauwatosa, WI). Slides were scanned by the Tissue Bank at MCW. Four randomly selected fields, two from the medulla and two from the cortex, were imaged in each kidney. The brown and blue colors were quantitated using ImageJ software (National Institutes of Health, Bethesda, MD). Endothelial cells (brown) were normalized by nuclei (blue) and expressed as endothelial cells. For microRNA expression levels from the next-generation sequence data, a t test was used for comparisons between two groups. Results from RT-qPCR, Western blot and histological staining were obtained by one-way analysis of variance (ANOVA) or Kruskal-Wallis one-way ANOVA on ranks when either the normality or equal variance test failed. All pairwise multiple comparisons were performed to determine statistical significance by Dunn’s or Holm-Sidak method of post hoc analyses. In addition, summarized changes for miRNA, mRNA and protein expression were also compared by t test or Mann-Whitney rank sum tests to address two specific questions: 1. Does radiation change gene expression? 2. Is the gene expression in irradiated rats that receive lisinopril different from that of irradiated rats that do not receive lisinopril?

We report that renal microRNA profiles are altered as late as 35 days after a single exposure to 13 Gy, within a latent period when no signs or symptoms of renal dysfunction are detected. There is a marked increase in miR-34a-5p which could potentially regulate multiple molecular signaling pathways including the Notch pathway. To our knowledge, these are the first data that show miR-34a-5p changes in kidney tissue after irradiation. Most postirradiation microRNA studies have focused on irradiation of cultured cancer cells or tissues. This study is one of the few such studies conducted in irradiated kidneys without tumors, and provides evidence that microRNA is an important marker to identify radiation injuries in the kidney. Expression of the Notch ligand protein, Jagged1, is downregulated in the irradiated kidney. Jagged1 protein is not decreased in irradiated kidney in the presence of the mitigator of radiation nephropathy, lisinopril. The mitigation of radiation effects by lisinopril, as well as other ACE inhibitors, has been studied for decades ( 4 , 20 , 22 – 24 ), but the mechanisms remain unknown. This may be the first time that an ACE inhibitor is described to regulate radiation-induced changes in molecular mediators of the Notch pathway. Jagged1 is known to participate in vascular remodeling, embryonic lethality and cardiovascular defects ( 25 ). Jagged1 had a unidirectional effect on endothelial cells and vascular smooth muscle cells to inhibit arterial lesion formation after vascular injury ( 26 ), indicating protection of the vasculature by Jagged1. Since radiation decreased Jagged1 expression in the kidney at 35 days and lisinopril mitigated this effect, it is possible that partial loss of Jagged1 protein reduces protection to the vasculature in the kidney, which then results in damage and vessel regression by radiation at 100 days. Together, the findings suggest that miR-34a-5p may be involved in radiation-induced vascular regression through a Jagged1-mediated process. Our studies also indicate complex regulation of gene expression by microRNA, which may occur at transcriptional or translational levels, based on mechanisms that remain to be identified. In this study, we used a rat PBI model by shielding only part of one hind limb (leg-out PBI). The marrow in the unexposed leg is able to repopulate the marrow compartment and help the rats survive from radiation-induced bone marrow toxicity. After receiving a 13 Gy dose with supportive care in the first few weeks postirradiation, rats survived the GI and hematopoietic toxicities. In this model, lethal radiation nephropathy as measured by blood urea nitrogen (>120 mg/dl) occurs after 90 days in rats that survive radiation pneumonitis, which occurs between 42–80 days. We selected 35 days as the time point in this study, for the following reasons. 1. At this time, survivors have recovered from the acute radiation syndrome (ARS) comprising GI and bone marrow injuries. The animals are in a relatively stable latent period before manifestation of DEARE. 2. This is the latest time point that we have tested ACE inhibitors as a countermeasure to mitigate radiation pneumonitis ( 27 ). Furthermore, radiation nephropathy is also mitigated by countermeasures such as ACE inhibitors started after this time ( 28 ). Investigations into the molecular events in the lungs and kidneys at this crucial time may reveal insights into the mechanism of DEARE. The high percentage of the mapping rate (≥90%) we achieved by next-generation sequencing, indicates a high quality of library preparation ( Supplementary Table S1 ; https://doi.org/10.1667/RADE-20-00078.1.S1 ). The most abundant microRNAs in rat kidney were ranked very high in abundance in the human kidney, indicating that the two profiles are similar ( Supplementary Table S2 ). Thus, the rat kidney microRNA expression may serve as a surrogate for humans and will be valuable since similar human studies are not feasible. The most abundant microRNA in the nonirradiated rat kidneys is miR-10a-5p ( Supplementary Table S2 ; https://doi.org/10.1667/RADE-20-00078.1.S1 ). The high expression of this microRNA in kidney tissues has also been reported in humans ( 15 , 29 ) and other species ( 29 , 30 ). In fact, miR-10a-5p was identified as a diagnostic serum biomarker for acute kidney injury in patients ( 31 ). However, in the current study, we did not observe a change in expression of this microRNA in kidneys at day 35 postirradiation. Since kidney dysfunction, as measured by increase in blood urea nitrogen (BUN), is not detected until after day 70 postirradiation in the rat, 35 days may be too early to observe a change in miR-10a-5p ( 20 , 24 ). Future studies following this marker in the blood and kidney may provide a biomarker for radiation nephropathy. MiR-34a-5p was reported to be highly expressed in normal tissues of rat, including testis and lung, but its expression in kidney was not high ( 32 ). The findings in the current study are consistent with that of Dutta et al . Even though the basal expression of miR-34a-5p in nonirradiated kidneys is relatively low ( Table 1 ), it is significantly increased (approximately sevenfold) after irradiation ( Table 1 , Fig. 2 ). As shown in Table 3 , miR-34a is a well-studied microRNA. With regards to radiation response, miR-34a was reported to be increased after irradiation in cancer and normal cells ( 33 ), but few of these studies included investigations at the RNA and protein level, and none evaluated the effect of lisinopril. In addition, miR-34a has been reported as a stress ( 34 , 35 ) and fibrosis marker ( 36 , 37 ). Pulmonary ( 27 , 38 ), cardiac ( 22 , 39 – 41 ) and renal ( 42 , 43 ) fibrosis were commonly seen in different radiation models. In other models, increased expression of miRNA-34a was observed in renal tubular epithelial cells of patients with renal fibrosis ( 36 ), unilateral ureteral obstruction (UUO) ( 36 , 44 ) and after renal injury induced by oxidative stress ( 45 ). The miR-34a overexpression was accompanied by downregulation of Klotho in two of these studies ( 36 , 45 ). Nephrotoxic activities exerted by Cyclosporine A (CsA), a widely used immunosuppressive drug, increased tubulointerstitial fibrosis and inflammation in the kidney, which was again correlated with higher expression of miR-34a ( 34 ). Also, under high glucose conditions, expression of renal fibrosis markers was upregulated through the miR-34a-5p/Sirt1/HIF-1a pathway ( 37 ). The miR-34a expression was modulated in ischemia/reperfusion-induced acute kidney injury (AKI) ( 35 ). It will be interesting to investigate if miR-34a is a key factor for the development of fibrosis after irradiation in multiple organs. When comparing results of microRNA expression in rats after 0 Gy and 13 Gy irradiation by IPA, miR-34a-5p was the only focus molecule in the identified network ( Fig. 1A ). This result shows that miR-34a-5p may play an important role in kidneys after irradiation. When the comparison was done between 0 Gy and 13 Gy irradiation plus lisinopril, in addition to miR-34a-5p, miR-431–5p and miR-494–3p were also focus molecules ( Fig. 1B ). However, within this network, miR-34a-5p is still the central molecule, while miR-431–5p and miR-494–3p interact with fewer molecules ( Fig. 1B ). These data support the role of miR-34a-5p in radiation nephropathy, which remains a focus even in the presence of a mitigating agent. When a comparison was performed between 13 Gy irradiation only and with lisinopril treatment, miR-431–5p is the only microRNA identified ( Fig. 1C ). Taking into consideration the networks and the interactions shown in Fig. 1B and C , we hypothesize that miR-431–5p may be involved in mitigation of radiation nephropathy by lisinopril via protein argonaute2 (AGO2) or nuclear receptor subfamily 0 group B member 2 (NR0B2). AGO2 is a component of RNA-induced silencing complex and drives mRNA cleavage in the RNA-mediated gene silencing ( 46 ). NR0B2 functions as a transcriptional inhibitor of gene expression by binding directly to multiple nuclear receptors ( 47 ). Further studies are needed to investigate this hypothesis. Jagged1 is a ligand that interacts with Notch receptors, like Notch1, 2, 3 and 4 ( 48 ). It is known that microRNA can regulate gene expression by binding to complementary target mRNAs ( 49 ). The luciferase reporter assays have already been performed, confirming that Jagged1 is a direct target of miR-34a ( 50 – 53 ). The binding sites of miR-34a in the Jagged1 3′ untranslated regions (3′-UTR) have also been described elsewhere ( 50 ). An additional pairing sequence and site predicted by TargetScan is included in Table 4 . In the current study, we found that the protein level of Jagged1 is decreased after irradiation in the kidney as determined by Western blots normalized for β-actin ( Fig. 3 ). A decrease in Jagged1 expression would be expected by an increase of miR-34a ( Fig. 2 ). However, Jagged1 mRNA is not altered after irradiation in the kidney ( Fig. 3 ), suggesting that the downregulation of Jagged1 by miR-34a after irradiation in rat kidney does not occur at the transcriptional level. This result is concordant with the general understanding of microRNA described to regulate gene expression post-transcriptionally as well as repress protein production ( 54 ). The mechanisms of how microRNA represses protein production are not fully understood, but it is often believed that it is through microRNA-induced translational repression and mRNA decay ( 49 , 54 , 55 ). Our results of no change in Jagged1 mRNA but decrease in protein after irradiation suggests that radiation-induced Jagged1 decreases are through miR-34a-5p-mediated translation repression instead of mRNA decay. Jagged1 has been described to interact with receptors of the Notch signaling pathway ( Table 5 ), which play critical roles in vasculogenesis and angiogenesis ( 56 ). These findings suggest that downregulation of Jagged1 protein induces vascular regression that has been observed after irradiation ( 9 , 10 ). Expression of the Notch receptors Notch1 and Notch2 were tested in this study since they are also closely associated with miR-34a ( Table 3 ). Notch2 was decreased after irradiation at the mRNA level but the protein expression was not changed ( Fig. 3 ). It is possible that radiation-induced miR-34a-5p causes Notch2 mRNA decay. Though it is not a widely believed mechanism, it is reported that in specific situations microRNA activates rather than represses translation ( 49 ). Whether this is the reason for unchanged Notch2 protein and decrease in mRNA in this study needs to be investigated. Notch1, as found in the Ingenuity Knowledge Base ( Table 3 ), was also described as a direct target of miR-34a ( 50 ). However, although miR-34a-5p was increased after irradiation, we did not detect a change in Notch1 mRNA or protein. These results match the IPA pathway projection based on the microRNA profile from sequence data, in which only Jagged1 and Notch2, but not Notch1, were predicted to be involved in the network with miR-34a-5p as the core ( Fig. 1A ). MiR-494–3p and miR-431–5p were also identified by IPA when nonirradiated controls versus 13 Gy plus lisinopril treatment groups were compared ( Fig. 1B ). While we recognize the role of miR-34a-5p in inducing radiation nephropathy, it is possible that lisinopril mitigates renal injury via miR-431–5p-mediated interactions ( Fig. 1C ). Future investigations will be needed to determine further details of the molecular mechanism of radiation nephropathy and its mitigation by lisinopril. It has been reported that radiation activates Notch pathways in studies of irradiation in cancer models ( 12 , 13 ). In this study, Jagged1 of Notch pathway was inhibited instead of activated. Jagged1 inhibition is not in conflict with the general understanding that Notch signaling plays key roles in carcinogenesis and tumor progression ( 57 , 58 ). The Notch signaling pathway is one of the most commonly activated signaling pathways in cancer, even without irradiation ( 57 ). Taken together, these studies show the different responses in Notch signaling after irradiation in normal tissue versus tumors. It is important to note that only three members of Notch pathway, Jagged1, Notch1 and Notch2, were studied here, since they were the only three predicted to be regulated by or to bind to miR-34a-5p ( Table 3 ). While only inhibition of Jagged1 protein was found in this study, it is possible that other members of the Notch pathway that did not share homology with miR-34a-5p and were therefore not studied in this investigation, such as DLL (Delta-like ligand), Notch3 and Notch4, may change after irradiation. Future studies of more members of the Notch pathway, regardless of their association with miR-34a-5p, will help to confirm regulation of different members and overall activity of the Notch pathway after irradiation in normal tissue. Histological analyses confirmed changes in vascular endothelium 100 days postirradiation to accompany renal dysfunction (quantitated by measuring BUN) that we have detected at 90 days (64 mg/dl in irradiated rats versus <25 mg/dl in nonirradiated controls) but not at 35 days ( 4 ) postirradiation. Thus, it is not surprising that a pathway, which signals vascular development, sprouting and pruning ( 11 , 59 ), may be altered by radiation since vascular regression is a well-recognized outcome of radiation exposure ( 9 , 10 , 60 , 61 ). Despite great effort, we were not able to identify an antibody that could detect rat endothelial-specific protein in kidney by Western analyses, to confirm the vascular regression after irradiation and the mitigation by lisinopril observed by histology ( Fig. 4 ). We are conducting studies to develop alternative techniques, such as vascular metabolic imaging or perfusion CT, to verify vessel regression in this model. In summary, renal microRNA profiles are altered by radiation, with an increase in miR-34a-5p after a single exposure to 13 Gy in conjunction with downregulation of the Notch ligand protein, Jagged1. Lisinopril protects against loss of Jagged1 and radiation-associated decrease in endothelial cells, potentially linking Jagged1 to vascular regression after radiation.

Radiation nephropathy is one of the lethal organ injuries induced by radiation. It affects cancer patients after radiotherapy ( 1 , 2 ) and occurs in animal models of radiation mimicking terrorism or accidental exposures ( 3 ). Molecular pathways to determine the mechanism(s) of radiation nephropathy have not been well studied, however, patients develop radiation nephropathy months to years after exposure compared to acute injuries to other organs, such as bone marrow, gastrointestinal (GI) tract and lungs, complicating cause and effect. Animal models have been predominantly used in studies of high-dose, non-tumor (normal) tissue irradiation, because conducting such studies in humans is not feasible. However, an ideal animal model should reflect human biology. We developed a model of partial-body irradiation (PBI) in rat to study radiation-induced nephropathy and to develop countermeasures ( 4 ). In this model, 11–12-week-old female Wistar rats received 13 Gy whole-body irradiation with one hind leg placed outside the field (13 Gy leg-out PBI) to minimize bone marrow toxicity. The rats were given supportive care consisting of subcutaneous saline injection in the first week and antibiotics in the drinking water to help improve survival from GI toxicity by radiation ( 4 ). Surviving rats developed delayed effects of acute radiation exposure (DEARE) including radiation pneumonitis between 35–80 days followed by lethal radiation nephropathy after 100 days. This rat model is one of only two known animal models of radiation injury to manifest all major sequelae relevant to survival after exposure. The second model, in nonhuman primates ( 5 ), demonstrates only mild kidney injury, possibly because it may take years to develop severe radiation nephropathy, which we observe within 120 days in the rat. We also studied the angiotensin converting enzyme inhibitor (ACEi) lisinopril as a mitigator in this model. At a molecular level, lisinopril suppresses signaling by the renin-angiotensin system (RAS) to lower levels of the vasoconstrictor, angiotensin II. Lisinopril is used in the clinic to treat hypertension and heart failure, as well as myocardial infarction and diabetic nephropathy prophylaxis ( 6 ). In the current study, lisinopril was started 7 days postirradiation in drinking water and continued to the end of the experiment in our studies. Lisinopril improves survival through the pneumonitis and nephropathy sequelae ( 4 ). MicroRNAs regulate gene expressions and are altered during development, pathological conditions and radiation ( 7 ). MicroRNAs are under development as powerful molecular markers for diagnostics as well as treatment for radiation injuries ( 8 ), since specific microRNAs may target key molecules and signaling pathways that are associated with these injuries. Studying microRNAs and the target gene changes may help to find the molecular mechanism(s) and treatment strategies for radiation-induced multi-organ injuries, including late renal injuries. Vascular injuries, especially vascular regression and endothelial injury after irradiation, contribute to the development of organ failure ( 9 , 10 ). Understanding molecular events that cause vascular regression in early stages before the symptoms occur in the animal model will facilitate designing of preventative strategies for radiation nephropathy in the clinic. The Notch signaling pathway is known to mediate vascular structure and function at a molecular level ( 11 ) and is itself regulated by exposure to radiation ( 12 , 13 ). It is also one of very few pathways that has been implicated in vascular regression in any model and this is reported to be mediated by the RAS ( 14 ). Therefore, the focus of this study was to elucidate the molecular mechanisms of radiation vascular dropout by investigating microRNAs and their vascular targets in the Notch pathway that are altered in a rat model after irradiation to multiple organs including the kidneys. We used next-generation miRNA-seq technology to characterize the complete rat renal microRNA profile before symptoms of radiation nephropathy set in. Most importantly, we confirmed the translational applicability of our rat model with humans. The two show similarities in the renal microRNA profiles after comparing our rat data with similar human sequence reported by Ben-Dov et al. ( 15 ).

Table S1. Library preparation and sequence quality. Table S2. Most abundant mature unique microRNA found in kidneys of rat and human. Table S3. Targets of interest tested: a. Primers used for qPCR; b. Antibodies used for Western blots and immunohistochemistry.