Overexpression of miR‐19a and miR‐20a in iPS‐MSCs preserves renal function of chronic kidney disease with acute ischaemia‐reperfusion injury in rat

Abstract This study tested the hypothesis that therapy with double overexpression of miR‐19a‐3p and miR‐20a‐5p (miRDOE) to human inducible pluripotent stem cell–derived mesenchymal stem cells (iPS‐MSCs) was superior to iPS‐MSCs alone for preserving renal function in rat with pre‐existing chronic kidney disease (CKD), followed by ischaemia‐reperfusion (IR) injury. In vitro study demonstrated that the protein expressions of oxidative stress (NOX‐1/NOX‐2/NOX4/oxidized protein/p22phox), inflammatory downstream signalling (TLR2&4/MyD88/TRAF6/IKK‐ß/p‐NFκB/IL‐1ß/IL‐6/MMP‐9) and cell apoptosis/death signalling (cleaved caspase‐3/mitochondrial Bax/p‐ERKs/p‐JNK/p‐p38) at time‐points of 24‐hour/48‐hour cell cultures were significantly increased in p‐Cresol‐treated NRK‐52E cells than in the control that was significantly reversed by miR‐19a‐3p‐transfected iPS‐MSC (all P < .001). Animals were categorized into group 1 (sham‐operated control), group 2 (CKD‐IR), group 3 (CKD‐IR + oligo‐miRDOE of iPS‐MSCs/6.0 ×105/intra‐renal artery transfusion/3 hours after IR procedure), group 4 (CKD‐IR + iPS‐MSCs) and group 5 (CKD‐IR + miRDOE of iPS‐MSCs/6.0 ×105/intra‐renal artery transfusion/3 hour after IR procedure). By day 35, the creatinine/BUN levels were lowest in group 1, highest in group 2 and significantly lower in group 5 than in groups 3 and 4 (all P < .0001) but they showed no difference between the latter two groups. The protein expressions of oxidative stress, inflammatory downstream signalling and cell apoptosis/death signalling exhibited an identical pattern of creatinine level among the five groups (all P < .00001). Also, the microscopic findings demonstrated that the kidney injury score/fibrotic area/number of inflammatory cells (CD14+/CD68+) exhibited an identical pattern of creatine level (all P < .0001). The miRDOE of iPS‐MSCs was superior to iPS‐MSCs for preserving the residual kidney function and architecture in CKD‐IR rat.


| INTRODUC TI ON
Acute kidney injury (AKI) with and without pre-existing chronic kidney disease (CKD), which is common in hospitalized patients, includes a group of clinical syndromes that primarily manifest as a rapid decline in renal function in association with the accumulation of metabolic waste. 1,2 Previous study has shown that about 17 million hospital admissions per year in the United States are complicated by AKI, resulting in additional costs up to $10 billion to the healthcare system. 3 Additionally, AKI has been estimated to occur in approximately 10%-15% of patients admitted to hospital, while its incidence in intensive care unit has been found to be in more than 50% of patients. 4,5 This situation, therefore, warrants the development of new therapeutic modalities for improving the kidney function after AKI, [6][7][8] especially in pre-existing CKD setting.
Summarizing from the literatures surveyed, we can find that the CKD, with variously divergent entities of known and unknown causal aetiologies clearly identified, [9][10][11][12][13][14] is still the major and rapidly growing contributor to healthcare burden worldwide. [14][15][16] Abundant data from clinical observational studies have revealed that CKD incurred high morbidity and mortality in hospitalized patients, especially in those CKD patients with co-existing cardiovascular disease (ie cardiorenal syndrome). 14,[17][18][19][20][21][22] Intriguingly, despite the state-of-the-art pharmaceutical strategies, such as the uses of angiotensin-converting enzyme inhibitor (ACEI), angiotensin II type I receptor blockade (ARB), and direct renin inhibitor (DRI), as well as good education, and renewed guideline for CKD precision management, progressive deterioration of residual renal function in the setting of CKD is commonly encountered, as a subsequence of leading to the adverse development of end-stage renal disease (ESRD). [23][24][25][26][27][28] Thus, current treatment of CKD is regrettably left much to be desired.
Intriguingly, plentiful studies [29][30][31][32][33][34] have demonstrated that mesenchymal stem cell (MSC) and endothelial progenitor cell (EPC) treatments for CKD are safe and efficacious maintenance of residual renal function in CKD setting. Additionally, studies have recently further revealed that human induced pluripotent stem cell (iPSC)derived MSCs (ie iPS-MSCs) exhibit multiple paracrine actions (ie including angiogenesis, immunomodulatory and anti-inflammatory factors) for organ repair and regeneration due to strong capacities of self-renewal and differentiation into most somatic cell lineages. 35,36 Furthermore, our previous study 37 also demonstrated that iPSCderived MSC therapy efficaciously safeguarded the rodent renal function and kidney architecture from acute ischaemia-reperfusion (IR) injury.
Recently, our phase I clinical trial demonstrated that intrarenal artery administration of autologous EPCs (ie CD34+ cell therapy) to CKD patients was safe and retained the renal function in a stable state at one-year follow-up. 33 In this study, we also found that the gene expressions of miR-374a-5p, miR-19a-3p, miR-106b-5p, miR-26b-5p and miR-20a-5p in peripheral blood mononuclear cells, five anti-apoptotic miRNAs, were significantly lower in CKD patients than in healthy individuals. 33 These findings 33,35,36 raised the hypothesis that transfected miR-19a-3p and miR-20a-5p (ie double miRNA overexpression) to human iPS-MSCs were superior to iPS-MSCs only for preserving the residual renal function in rat CKD.

| Ethics
All animal procedures were approved by the Institute of Animal Care and Use Committee at Kaohsiung Chang Gung Memorial Hospital Frederick, MD, USA)-approved animal facility in our hospital with controlled temperature and light cycles (24°C and 12/12-hour light cycle). 35, the creatinine/BUN levels were lowest in group 1, highest in group 2 and significantly lower in group 5 than in groups 3 and 4 (all P < .0001) but they showed no difference between the latter two groups. The protein expressions of oxidative stress, inflammatory downstream signalling and cell apoptosis/death signalling exhibited an identical pattern of creatinine level among the five groups (all P < .00001). Also, the microscopic findings demonstrated that the kidney injury score/fibrotic area/number of inflammatory cells (CD14+/CD68+) exhibited an identical pattern of creatine level (all P < .0001). The miR DOE of iPS-MSCs was superior to iPS-MSCs for preserving the residual kidney function and architecture in CKD-IR rat.

K E Y W O R D S
chronic kidney disease, fibrosis, inflammation, iPS-MSCs, ischaemia-reperfusion injury, microRNAs, oxidative stress 2.2 | miR-19a-3p and miR-20a-5p were candidates for double overexpression (miR DOE ) in iPS-MSCs and treatment of CKD animals We have evaluated that miR-19a-3p and miR-20a-5p were the two most suitable candidates among the five miRNAs (ie miR-374a-5p/miR-19a-3p/miR-106b-5p/miR-26b-5p/miR-20a-5p) to be overexpressed (ie transfection) for the purpose of treatment of CKD + IR animals. In detail, transfections of miR-19a-3p and miR- TransIT-X2 reagent was mixed with miRNA mimics for 25 minutes at room temperature. The miRNA mimics-containing complexes were further distributed into cells. Two days later, expressions of miRNAs and related genes were validated by the real-time qPCR assay ( Figure S1).

| Creation of animal model of CKD
The procedure and protocol of CKD induction have been described in detail in our previous report. 29,31 Pathogen-free, adult male Sprague-Dawley (SD) rats (n = 40) weighing 320-350 g (Charles River Technology, BioLASCO Taiwan Co. Ltd.) were anaesthetized by inhalational 2.0% isoflurane and placed supine on a warming pad at 37°C for midline laparotomies. The shame-operated control (SC) rats received laparotomy only, while CKD was induced in all animals of the CKD groups by right nephrectomy plus arterial ligation of the upper two-third (upper and middle poles) blood supplies of the left kidney, leaving the lower third (lower pole) kidney with normal blood supply. This model allowed preservation of a limited amount of functioning renal parenchyma and simulation of CKD.

| Animal grouping, procedure and protocol for acute kidney ischaemia-reperfusion (IR) injury in preexisting CKD and therapeutic strategy
The procedure and protocol of acute kidney IR injury have been previously described. 37,38 Briefly, by day 28 after CKD induction, those of CKD animals were again anaesthetized by inhalational 2.0% isoflurane and placed supine on a warming pad at 37°C for midline laparotomies. SC animals underwent laparotomy only, while acute kidney IR injury of left kidney was induced in all animals in groups 2 to 5 by clamping the renal pedicles for 50 minutes using noncrushing vascular clips. Followed by IR procedure, the animals were divided into group 1 [sham-operated control (SC)], group 2 (CKD + IR), group 3 (CKD + IR + oligo-miR DOE of iPS-MSCs by intra-renal artery transfusion 3 hours after IR procedure), group 4 [CKD + IR + iPS-MSCs (6.0 × 10 5 cells) by intra-renal artery transfusion 3 hours after IR procedure] and group 5 [CKD + IR + miR DOE of iPS-MSCs (6.0 × 10 5 cells) by intra-renal artery transfusion 3 hours after IR procedure], respectively.
The animals in each group were killed and kidney specimens collected for individual study by day 7 after the acute kidney IR induction, that is the total study period was 35 days (CKD was 28 days + 7 days after IR procedure = 35 days). The time-point of iPS-MSC administration to the animals at 3 hours after AK-IR was based on our recent reports. 37,39 Additionally, the dosage and the route of intra-renal artery administration were based on our previous report 32 with minimal modification. Furthermore, based on the concept that the intravenous administration of stem cell will frequently be trapped into the lung parenchyma, the purpose of intra-renal artery administration of cells in the present study was to make sure the cells were precisely into renal circulation rather were retained in the lung parenchyma for an achievement of great therapeutic effect.

| Procedure and protocol of cell culture for differentiation of human iPS into iPS-MSCs
The procedure and protocol of human iPSC culture for differentiation into iPS-MSCs were as per the manufacturer's instructions and have been described in our previous report. 37,39 In detail, by day 1, human iPSCs (mTeSR™1; StemCell, #28315) were first washed by 5 mL PBS, followed by 2 mL Accutase (Gibco, #A1110501; Accutase: PBS = 1:1); the incubator reaction was continuous for 1 minute. Additionally, 2 mL KO DMEM/F12 (Gibco, #12660012) was added and the cells were collected in 15-mL centrifuge tubes for 5-minute centrifugation (200 g).
The cells were then cultured in a 10-cm dish for 24 hours in mTeSR™1 culture medium. By day 2, the cells (mTeSR™1) were collected and washed with 5 mL PBS. STEMdiff™-ACF Mesenchymal Induction Medium (StemCell, #05241) was added and the incubator culture was continued for 24 hours. The STEMdiff™-ACF Mesenchymal Induction Medium was exchanged once/day from days 1 to 3. This procedure was repeated on days 3-6. On days 7-21, the procedure was repeated but the culture medium was refreshed every 3 days. As the standard method of iPS-MSC culture, the descriptions were similar to our previous studies. 37,39

| In vitro study design for assessment of oxidative stress effect of p-Cresol-on inducing renal tubular epithelial cells damage
To elucidate the impact of oxidative stress on renal tubular epithelial cells (ie the NRK-52E cells), these cells were co-cultured with p-Cresol (25 μmol/L, ie uraemic toxin producing oxidative stress to mimic the situation in advanced CKD). The cells (1 × 10 5 ) per mL, which were utilized in this in vitro study, were first cultured in DMEM-Low plus 10% FBS. For Western blotting, the cells were co-cultured with designed drug in 10-cm dish with 1.0 × 10 6 cells for 24 hours. All the cells were then collected for individual assays. The dosage of p-Cresol in the culture medium was based on our previous report. 40

| The time-points of measurement of blood urine nitrogen (BUN) and creatinine levels
The blood samples were serially collected before and after the CKD procedure (ie at days 14 and 35 prior to and immediately before the animals to be killed). Serum levels of creatinine and BUN were measured in duplicate using standard laboratory equipment. The mean intra-assay coefficient of variance for BUN and creatinine was less than 4.0%.

| The time courses of collection of 24-hour urine for the ratio of urine protein to urine creatinine
The procedure and protocol were based on our previous reports. 30,32,39 For the collection of 24-hour urine in individual study, each animal was put into a metabolic cage [DXL-D, space: 190 × 290 × 550 mm 3 , Suzhou Fengshi Laboratory Animal Equipment Co. Ltd.] for 24 hours with free access to food and water. Urine in 24 hours was collected in all animals prior to and at days 14 and 35 after CKD induction for determining the ratio of urine protein to urine creatinine.

| Western blot analysis of left kidney specimens
The procedure and protocol have been described in our previous reports. 30,32,39 In details, primary antibodies against tumour necro- Immunoreactive bands were visualized by enhanced chemiluminescence (ECL; Amersham Biosciences), which was then exposed to Biomax L film (Kodak). For quantification, ECL signals were digitized using Labwork software (UVP). For oxyblot protein analysis, a standard control was loaded on each gel.

| Immunofluorescent studies
The procedures and protocols for immunofluorescent (IF) examinations have been described in our previous reports. 30,32,39 In details,

| Histopathological analysis for quantification of kidney fibrosis at day 35 after CKD induction
Masson's trichrome staining was utilized for investigating the fibrosis in kidney parenchyma. Three serial sections of kidney in each animal were prepared at 3 µm thickness by microtome (Leica RM2235).
The integrated area (µm 2 ) of fibrosis on each section was calculated using the Image Tool 3 (IT3) image analysis software (University of Texas, Health Science Center, San Antonio, UTHSCSA; Image Tool for Windows, Version 3.0). Three randomly selected high-power fields (HPFs) (100×) were analysed in each section. The numbers of pixels in fibrotic area obtained from three HPFs were summated. The procedure was repeated in two other sections for each animal. The mean pixel number per HPF for each animal was then determined by summating all pixel numbers and dividing by 9. The mean integrated area (µm 2 ) of fibrosis in kidney per HPF was obtained using a conversion factor of 19.24 (1 µm 2 corresponded to 19.24 pixels).

| Statistical analysis
Quantitative data are expressed as mean ± SD. Statistical analyses were performed using SAS statistical software for Windows version 8.2 (SAS Institute). Statistical differences between two groups were analysed with Student's t test was used. A probability value <.05 was considered statistically significant.

| The protein expressions of inflammatory downstream signalling, oxidative stress and cell apoptosis/death signalling at 24/72 hours after oxidative stress stimulation in renal tubular epithelial cells
In the in vitro study, NRK-52E cells were co-cultured with p-Cresol

| The time courses of circulating levels of creatinine and blood urine nitrogen (BUN) and ratio of urine protein to urine creatinine prior to and after CKD-IR procedures
By day 0, the circulating levels of creatinine, BUN and the ratio of urine protein to urine creatinine did not differ among the five groups ( Figure 4A-C). However, by day 14 after CKD induction, these parameters were significantly lower in group 1 (SC) than in group 2 (CKD + IR), group 3 (CKD + IR + oligo-miR DOE of iPS-MSCs), group 4 (CKD + IR + iPS-MSCs) and group 5 (CKD + IR + miR DOE of iPS-MSCs). On the other hand, these parameters did not differ among groups 2 to 5.
By day 35 after CKD induction, theses parameters were lowest in group 1, highest in group 2, significantly lower in group 5 than in groups 3 and 4, but they did not differ between groups 3 and 4 ( Figure 4D-F).

| The protein expressions of apoptosis/cell death biomarkers in renal parenchyma at day 35 after CKD-IR procedure
The protein expressions of cleaved caspase-3, mitochondrial Bax (ie apoptotic markers), p-JNK, p-EKR and p-38 (ie cell proliferation/ death biomarkers) were lowest in group 1, highest in group 2, significantly lower in group 5 than in groups 3 and 4, but they did not differ between groups 3 and 4 ( Figure 4N-Q, T).

| The protein expressions of inflammatory biomarkers in renal parenchyma at day 35 after CKD-IR procedure
To elucidate whether expression of inflammatory signalling was similar in vivo to that of the in vitro, Western blot analysis was performed for the harvested kidney specimen. The protein expressions of TLR-2, TLR-4, MyD88, TRAF6, p-NF-κB, IL-1ß, IL-6 and MMP-9, eight indicators of inflammatory biomarkers, were lowest in group 1, highest in group 2, significantly lower in group 5 than in groups 3 and 4, but they did not differ between groups 3 and 4. On the other hand, the protein expression of IKB-ß, a protein for propagating the inflammatory signalling from the upper to the lower pathway, revealed an opposite pattern of TLR-4/MyD88 among the five groups ( Figure 5).

| The kidney injury score and fibrotic area in renal parenchyma at day 35 after CKD-IR procedure
The microscopic finding of HE stain demonstrated that the kidney injury score was lowest in group 1, highest in group 2, significantly lower in group 5 than in groups 3 and 4, but it did not differ between groups 3 and 4. Additionally, the Masson's trichrome stain identified that the expression of fibrotic area was identical to the pattern of kidney injury score among the five groups ( Figure 6).
Our previous study has shown that the microRNA expressions of miR-19a-3p and miR-20a-5p retained anti-apoptotic property. 33 Intriguingly, a cardinal finding in the present study was that the double overexpression of these two microRNAs was superior to iPS-MSCs alone for preserving the residual kidney function and integrity of kidney architecture in CKD + IR in rat through F I G U R E 6 The kidney injury score and fibrotic area in kidney parenchyma at day 35 after CKD-IR procedure. A-E, Light microscopic findings (200×) of HE stain showing significantly higher degree of loss of brush border in renal tubules (yellow arrows), tubular necrosis (green arrows), tubular dilatation (red asterisk) protein cast formation (black asterisk) and dilatation of Bowman's capsule (blue arrows) in CKD-IR group than in other groups. F, Analytical result of kidney injury score, *P < .01; **P < .001; and ***P < .0001. G-K, Illustrating histological finding (200×) of Masson's trichrome stain for fibrotic area of kidney parenchyma (blue colour). L, Analytical result of fibrotic area, *P < .01; **P < .001; and ***P < .0001. Scale bars in right lower corner represent 50 µm. Student's t test was used (n = 6 for each group). ns = no significant difference. SC = sham-operated control; CDK-IR = chronic kidney disease + ischaemia-reperfusion; iPS-MSC = human inducible pluripotent stem cell-derived mesenchymal stem cell; and MSC miR/DO = miR-19a-3p and miR-20a-5p overexpression (ie double miRNA overexpression) in iPS-MSCs It is well recognized that the residual renal function is a crucial factor for predicting the long-term outcome in CKD patients.
Regrettably, the residual renal function in those CKD patients is more vulnerable to be damaged by various harmful factors (ie called AKI in setting of CKD) which frequently results in poorer prognostic outcome than in those AKI patients with a pre-existing normal renal function. Importantly, there is still lacking an effective treatment for these patients which, as a consequence, leads the majority of the patients are waiting for haemodialysis. In this way, our findings may be the potentially therapeutic candidate for these patients, especially those are refractory to the conventional therapy.

| Study limitations
This study has limitation. First, the duration of IR study period in preexisting CKD setting was only 7 days. Thus, the long-term outcome after IR injury was not investigated in the present study. Second, although extensive works had been done in the present study, the crystal-clear mechanism of how iPS-MSC and miR DOE of iPS-MSCs safeguarded the residual renal function and the integrity of kidney parenchyma remained uncertain. Based on the results of our study, we had schematically proposed the underlying mechanisms on how the miR DOE of iPS-MSCs affects and preserves the outcome in rodent in setting of CKD + IR injury in Figure 5. Third, this study did not measure the time courses of the parameters which were related to the serial changes of renal function in blood and urine after IR procedure. However, the final results (ie by day 7 after IR procedure) were still attractive and promising, suggesting that our study is still clinically relevant.

| CON CLUS IONS
In conclusion, the miR DOE of iPS-MSCs provided remarkably additional benefits than iPS-MSCs alone for recovering the residual kidney function and integrity of kidney architecture in CKD-IR in rat through down-regulating the inflammatory, oxidative stress and cell apoptotic/death signalling pathways as well as safeguarding the mitochondrial function.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding authors upon reasonable request.