Renovascular Disease and Mitochondrial Dysfunction in Human Mesenchymal Stem Cells

Visual Abstract

Abstract

Key Points

• Renovascular disease impairs the capacity of human adipose tissue–derived mesenchymal stem/stromal cells to repair ischemic murine kidneys.

• miR-378h modulated the capacity of renovascular disease adipose tissue–derived mesenchymal stem/stromal cells to repair ischemic kidneys in vivo.

Background

Renovascular disease leads to renal ischemia, hypertension, and eventual kidney failure. Autologous transplantation of adipose tissue–derived mesenchymal stem/stromal cells (MSCs) improves perfusion and oxygenation in stenotic human kidneys, but associated atherosclerosis and hypertension might blunt their effectiveness. We hypothesized that renovascular disease alters the human MSC transcriptome and impairs their reparative potency.

Methods

MSCs were harvested from subcutaneous abdominal fat of patients with renovascular disease and healthy volunteers (n=3 each), characterized and subsequently injected (5×10⁵/200 μl) into mice 2 weeks after renal artery stenosis or sham surgery (n=6/group). Two weeks later, mice underwent imaging and tissue studies. MSCs from healthy volunteers and in those with renovascular disease were also characterized by mRNA/microRNA (miRNA) sequencing. Based on these, MSC proliferation and mitochondrial damage were assessed in vitro before and after miRNA modulation and in vivo in additional renal artery stenosis mice administered with MSCs from renovascular disease pretreated with miR-378h mimic (n=5) or inhibitor (n=4).

Results

MSCs engrafted in stenotic mouse kidneys. Healthy volunteer MSCs (but not renovascular disease MSCs) decreased BP, improved serum creatinine levels and stenotic-kidney cortical perfusion and oxygenation, and attenuated peritubular capillary loss, tubular injury, and fibrosis. Genes upregulated in renovascular disease MSCs versus healthy volunteer MSCs were mostly implicated in transcription and cell proliferation, whereas those downregulated encoded mainly mitochondrial proteins. Upregulated miRNAs, including miR-378h, primarily target nuclear-encoded mitochondrial genes, whereas downregulated miRNAs mainly target genes implicated in transcription and cell proliferation. MSC proliferation was similar, but their mitochondrial structure and reparative function both in vivo and in vitro improved after miR-378h inhibition.

Conclusions

Renovascular disease impaired the reparative capacity of human MSCs, possibly by dysregulating miR-378h that targets mitochondrial genes.

Podcast

This article contains a podcast at https://dts.podtrac.com/redirect.mp3/www.asn-online.org/media/podcast/JASN/2024_08_21_ASN0000000000000440.mp3

Introduction

Renovascular disease is characterized by the narrowing or blockage of renal arteries due to atherosclerosis.¹ Renovascular disease induces inflammation, microvascular remodeling, and fibrosis within the poststenotic kidney,² contributing to kidney failure particularly in the elderly population.³ Importantly, patients with renovascular disease have a higher risk for developing renovascular hypertension, which accelerates kidney injury and is associated with higher cardiovascular morbidity and mortality.⁴ Renal revascularization failed to demonstrate superior effectiveness over standard medical therapy for most patients,⁵ warranting development of novel therapeutic interventions to ameliorate kidney injury and halt its progression toward kidney failure.

Mesenchymal stem/stromal cells (MSCs) are adult stem cells, which can be isolated from several sources and can differentiate into multiple cell types.⁶ Endowment with important anti-inflammatory and proangiogenic features makes MSCs attractive for regenerative therapies.⁷ Indeed, MSCs constitute the primary cell type used for therapeutic administration in both preclinical and clinical trials.⁸ We have previously shown in swine renovascular disease that a single intrarenal infusion of adipose tissue–derived MSCs with or without renal revascularization ameliorates poststenotic kidney inflammation and microvascular damage.^9,10 Similarly, administration of autologous adipose tissue–derived MSCs in patients with renovascular disease increased kidney tissue oxygenation and cortical blood flow 3 months later,^11,12 underscoring their potential to preserve the ischemic kidney.

However, comorbidities and cardiovascular risk factors may negatively influence the biological properties and therapeutic potential of MSCs.¹³ Adipose tissue–derived MSCs from pigs with renovascular disease exhibit decreased angiogenesis and increased senescence,¹⁴ and adipose tissue–derived MSCs obtained from patients with renovascular disease have increased DNA damage, associated with reduced angiogenic and migratory function.¹⁵ However, whether renovascular disease impairs the reparative potency of human MSCs and the mechanisms underpinning such dysfunction remain unknown.

MicroRNAs (miRNAs) are small (approximately 22 nucleotide long) non-coding RNAs that modulate gene expression by binding to specific sites within their target mRNAs.¹⁶ In this study, we took advantage of a murine renal artery stenosis (RAS) model and of RNA sequencing (seq) to test the hypothesis that renovascular disease alters the mRNA and miRNA cargo of human adipose tissue–derived MSCs and impairs their reparative potency. We further explored whether its modulation could ameliorate renovascular disease–induced damage in human MSCs in vitro and restore their reparative capacity in murine poststenotic kidneys in vivo.

Methods

Study Cohorts

Studies were performed in MSCs harvested from abdominal subcutaneous fat collected from three renovascular disease and three age-matched, sex-matched, and body mass index–matched healthy volunteers as previously described.^11,12 Informed written consent was obtained after approval of the Institutional Review Board of the Mayo Clinic. Similar to the Cardiovascular Outcomes for Renal Atherosclerotic Lesions trial,⁵ entry criteria for patients with renovascular disease included hypertension (systolic BP ≥130 and/or diastolic BP ≥80 mm Hg) and renal artery Doppler ultrasound velocity acceleration (peak systolic velocity >200 cm/s) or magnetic resonance/computed tomography angiography with evident atherosclerotic stenosis >60% and/or poststenotic dilation. Exclusion criteria included serum creatinine >2.5 mg/dl, uncontrolled hypertension (systolic BP >180 mm Hg) despite antihypertensive therapy, diabetes requiring medications, recent cardiovascular event (myocardial infarction, stroke, congestive heart failure within 6 months), pregnancy, and kidney transplant. All patients with renovascular disease were treated with angiotensin-converting enzyme inhibitors or angiotensin receptor blockers at recommended daily doses. Entry criteria for healthy volunteers included older than 18 years, systolic BP <130 and diastolic BP <80 mm Hg, and healthy overall state. In all patients with renovascular disease and healthy volunteers, blood and urine samples were collected, and lipid profile, serum creatinine, and urine protein levels were assessed by standard procedures. eGFR was calculated using the Chronic Kidney Disease Epidemiology Collaboration formula.²⁰

MSCs Harvesting and Characterization

MSCs were isolated from subcutaneous abdominal fat (2–4 g) under sterile conditions and subsequently digested in 2% collagenase-H for 90 minutes at 37°C, centrifuged at 400×g for 5 minutes, passed through a 70 μm cell strainer, and incubated in red blood cell lysis buffer. MSCs were then cultured for 3 weeks and expanded in culture for three passages in Advanced MEM with 5% platelet lysate²¹ and 2 mM L-glutamine and characterized using flow cytometry for expression of common MSC markers (CD73⁺, CD90⁺, CD105⁺, CD34⁻, CD45⁻).^9,22 MSC phenotype was also confirmed by their ability to differentiate into osteocytes, chondrocytes, and adipocytes, as previously described.^9,22

To further assess the impact of renovascular disease on MSC paracrine function, we characterized MSC protein profile by liquid chromatography–mass spectrometry proteomics at the Mayo Clinic Proteomics Core, as previously described.^23–26 Protein groups with a false discovery rate <0.05 and an absolute fold change (renovascular disease extracellular vesicles/healthy volunteer extracellular vesicles) >2 were classified as upregulated, whereas those with false discovery rate <0.05 and fold change <−2 were considered downregulated. Functional pathway analysis of differentially expressed proteins was performed using Gene Set Enrichment Analysis (GSEA, version 4.0.3, Broad Institute).²⁷

In addition, extracellular vesicles were collected from supernatants of 10×10⁶ human healthy volunteer MSC and renovascular disease MSC, cultured for 72 hours in advanced MEM medium without supplements, and ultracentrifuged twice, as shown.²² Concentration and size distribution of isolated extracellular vesicles were assessed by NanoSight Tracking Analysis using NanoSight NS300.

See Supplemental Material for detailed methods.

Mice Studies

Eleven-week-old 129-S1 mice (Jackson Lab, Bar Harbor, ME) were randomly assigned to one of four groups: Sham+vehicle, RAS+vehicle, RAS+healthy volunteer MSCs, or RAS+renovascular disease MSCs (n=6 mice/group, three male and three female, each) (Figure 1A). Mice were initially acclimated 2–3 days before the surgery and fed Picolab Rodent Diet 20-5053. At baseline, a 0.15 mm diameter plastic cuff was surgically inserted around the mouse right renal artery, whereas Sham mice underwent a similar procedure without cuff placement.^28–30 Animal technicians with extensive experience performed both Sham and RAS surgeries. Successful RAS was subsequently confirmed by a right/left kidney volume or weight ratio under 0.9. Two weeks after RAS or Sham, mice were anesthetized (1.5%–2.0% isoflurane inhalation), the left internal carotid artery was cannulated, and a plastic tube was advanced caudally for injections of MSCs or vehicle. The aorta was slowly infused with either PBS or far-red dye (Life Technologies) prelabeled MSCs (5×10⁵ cells in 200 μl PBS) from either those with renovascular disease or healthy volunteers, and mice were then allowed to recover. Two weeks later, mice underwent in vivo tests; they were then euthanized by exsanguination (blood collection), and their kidneys were harvested for ex vivo studies.³⁰ Serum creatinine (Arbor Assays, Cat KB02-H1) and plasma renin (Angiotensin-I RIA kit, ALPCO, Salem, NH) levels were measured by standard procedures, and right/left kidney weight was calculated. Body weight and systolic BP (tail-cuff) at baseline, 2 weeks, and 4 weeks were also recorded.

Figure 1.

Renovascular disease MSCs failed to decrease BP. (A) Experimental protocol. Adipose tissue–derived mesenchymal stem/stromal cells (MSCs) were isolated from subcutaneous abdominal fat of three patients with renovascular disease and three healthy volunteers and subsequently injected into 129-S1 mice 2 weeks after RAS or sham surgery (n=6/group). Two weeks later, mice underwent micro-MRI in vivo followed by ex vivo studies. (B) Body weight was consistently similar among the groups (n=6 each). (C) Systolic BP increased in all RAS groups 2 weeks after surgery but decreased at 4 weeks only in RAS+healthy volunteer MSCs (n=6 each). (D) Right kidney/left kidney weight was similarly lower in all RAS groups (n=6 each). (E) Plasma renin levels were elevated in all RAS groups but decreased only in mice treated with healthy volunteer MSCs (n=6 each). *P < 0.05 versus Sham+vehicle, †P < 0.05 versus RAS+vehicle, ‡P < 0.05 versus RAS+healthy volunteer MSCs. HV, healthy volunteer; MRI, magnetic resonance imaging; RAS, renal artery stenosis; RVD, renovascular disease; SBP, systolic BP.

Imaging Studies

Two weeks after MSC or vehicle injection, murine cortical and medullary perfusion and oxygenation were assessed using arterial spinning label and blood oxygen–dependent magnetic resonance imaging, respectively.^30,31 Images were analyzed and quantified using MATLAB R2015-a (MathWorks) and kidney volume assessed using Analyze software (version 12.0, Bio-medical Imaging Resource, Mayo Clinic, MN), as shown.³¹

Histological Studies

Tubular injury was scored (0–5) in 5 μm murine kidney cross-sections stained with hematoxylin and eosin.^32–34 Kidney cross-sections were imaged in a Nikon AX-R confocal microscopy system (Melville, NY). Tubular injury (dilation, atrophy, cast formation, cell detachment, or thickening of tubular basement membrane) was scored from 1 to 5 based on the % of tubules injured, 0 being normal tubules, 1: <10%, 2: 10%–25%, 3: 26%–50%, 4: 51%–75%, 5: >75% of tubules injured. Tubulointerstitial and glomerular fibrosis were assessed in Masson trichrome-stained slides and quantified over the entire tissue section using the threshold color plugin of ImageJ (https://imagej.nih.gov/ij/) (Wayne Rasband, National Institutes of Health), as shown.³⁰ To calculate MSC engraftment, prelabeled MSCs were first counted manually in the entire kidney section under confocal fluorescence microscopy. The total area of each cross-section was calculated using an image analysis program (MetaMorph, Meta Imaging Series 6.3.2, Allentown, PA) and the number of cells/mm² averaged and multiplied by the section thickness and then by the total kidney volume obtained by magnetic resonance imaging. This value, taken to represent the total number of MSC in the kidney, was divided by the number of injected cells to obtain a retention rate.³⁰ The ability of MSCs to incorporate into tubular and endothelial cells was also assessed in frozen poststenotic kidney sections stained with the proximal tubular marker phaseolus vulgaris erythroagglutinin (PHA-E, Vector Lab) and the endothelial marker CD31 (Cell Signaling, Cat 77699).

MSC Tracking and Immune Rejection Studies

To track the fate of human MSCs in murine kidneys over the course of 2 weeks, the aorta of additional Sham mice was slowly infused with either PBS (Control) or far-red dye prelabeled MSCs (5×10⁵ cells in 200 μl PBS); animals were euthanized 3, 7, and 14 days after delivery (n=4 each); and MSC retention rate calculated as described above. To investigate whether delivery of human MSCs in mice is associated with cellular rejection, immune cell levels in the blood and kidney tissue were assessed at each timepoint using flow cytometry and immunofluorescent staining, respectively.

For flow cytometry, cells were stained with fluorescein isothiocyanate anti-mouse CD45 antibody, phycoerythrin anti-mouse CD3 antibody, allophycocyanin anti-mouse/human CD11b antibody, and PerCP anti-mouse CD19 antibody. The cells were acquired with FlowSight Imaging flow cytometer (Amnis), and data were analyzed with Amnis IDEAS version 6.2.²⁶ CD45-positive cells were gated by histogram of expression of CD45 from live single cells. Next, T cells, B cells, and monocytes were gated from CD45⁺ cells on the basis of their expression of CD3, CD19, and CD11b, respectively.

In addition, renal immunofluorescent staining with antibodies against CD3 (Abcam, ab16669), CD19 (ab245235), and F4/80 (ab6640) was performed. Renal T cells, B cells, and macrophages were counted in the entire tissue section and expressed as their number per field.⁹

mRNA-Seq of Human MSCs

mRNA-seq was performed at the Mayo Clinic Genomic Analysis Core, as previously described.^22,35–37 Libraries were prepared and data analyzed using the MAPRSeq v.1.2.1 workflow.³⁸ Gene expression was normalized to 1 million reads (reads per kilobase per million mapped reads [RPKM]).³⁹ Genes with RPKM >0.1, fold change (renovascular disease MSCs/healthy volunteer MSCs) >1.4, and P < 0.05 were classified as upregulated and those with RPKM >0.1, fold change (renovascular disease MSCs/healthy volunteer MSCs) <0.7, and P < 0.05 as downregulated in renovascular disease MSCs versus healthy volunteer MSCs. Upregulated and downregulated genes were visualized in volcano plots, and heat maps were generated. Functional analysis (GSEA)²⁷ was performed to obtain a ranking of primary gene ontology categories of upregulated and downregulated genes.

miRNA-Seq and Integrated (mRNA-/miRNA-Seq/mRNA-Seq) Analysis of Human MSCs

miRNA-seq analysis was performed as previously described^22,40 using the comprehensive analysis pipeline for microRNA sequencing data workflow,⁴¹ Bowtie,⁴² and MiRDeep2.⁴³ Differentially expressed analysis was performed using edgeR.⁴⁴ miRNAs with fold change (renovascular disease MSCs/healthy volunteer MSCs) >1.4, and P < 0.05 were classified as upregulated and those with fold change (renovascular disease MSCs/healthy volunteer MSCs) <0.7 and P < 0.05 as downregulated in renovascular disease MSCs versus healthy volunteer MSCs. Differentially expressed miRNAs were visualized in volcano plots and heat maps, target prediction analysis was performed TargetScan,⁴⁵ and their functional analysis was performed using GSEA. To visualize miRNA target genes downregulated in renovascular disease MSCs and corresponding mRNAs upregulated in renovascular disease MSCs, or vice versa, Venn diagrams were constructed using VENNY 2.1.

Role of miR-378h

In Vitro Studies in Human MSCs

Our RNA-seq studies identified dysregulated genes implicated in cell proliferation and mitochondrial function. We focused on miR-378h, which was the only mitochondrial-related miRNA (mitomiR) among the top miRNAs upregulated in renovascular disease MSCs. mitomiRs represent a select group of miRNAs commonly expressed in mitochondrial tissue fractions that are critical modulators of mitochondrial homeostasis and function.^46,47 Furthermore, miR-378 targets important genes implicated in cell proliferation (e.g., Cell Division Cycle 20B [CDC20B]),⁴⁸ a prominent functional category of genes upregulated in renovascular disease MSCs versus healthy volunteer MSCs. To investigate whether miRNA modulation alters MSC proliferation and mitochondrial function, we subsequently treated renovascular disease MSC and healthy volunteer MSC with a miR-378h inhibitor (anti-miR-378h, ThermoFisher, Minneapolis, MN, MH21446) or mimic (miR-378h mimic, ThermoFisher, MC21446) and a transfection agent (Lipofectamine, RNAiMAX; Invitrogen, Life Technologies, Grand Island, NY) for 48 hours, following the manufacturer's instructions.

Cell proliferation was then assessed using a Cell Imaging Multimode Reader (Cytation-5, Bio Tek Santa Clara, CA). Cells were seeded in a 24-well plate (5×10⁴ per well) and kept at 37°C with 5% CO₂. Cell confluence was captured every hour, monitored for 48 hours, and data analyzed using Gen5 software (Bio-Tek).⁴⁹

MSC mitochondrial structure was assessed by digital Transmission Electron Microscopy (Phillips CM10). Cells were preserved in Trump's fixative solution, mounted on mesh grids, and stained with aqueous uranyl acetate and lead citrate. Representative MSCs (n=20) were randomly selected, and mitochondrial area (nm²) and matrix density (1/mean gray values) were determined in ten randomly selected mitochondria per cell, as previously described.^50,51 Mitochondrial reactive oxygen species production was assessed by Mito-SOX (ThermoFisher, Cat: M36008)⁵² and membrane potential by tetramethylrhodamine ethyl ester (Cat: T669).⁵³

Expression of miR-378h, ATP Synthase, H+ Transporting, Mitochondrial F0 Complex, Subunit F2 (ATP5J2), ATP Synthase F1 Subunit Gamma (ATP5F1C), B-Cell chronic lymphocytic leukemia/Lymphoma 2-Like 12 (BCL2L12), Fumarylacetoacetate Hydrolase Domain Containing 2A (FAHD2A), Cytochrome-C Oxidase Subunit 5A (COX5A), Mitofusin-1 (MFN1), nicotinamide adenine dinucleotide:Ubiquinone Oxidoreductase Subunit A1 (NDUFA1), and Translocase of Outer Mitochondrial Membrane-7 (TOMM7) was assessed by real-time quantitative PCR (qPCR) using the ΔΔCt method.^54,55 miR-378h expression was normalized by RNU6B and expression of the remaining genes by GAPDH.

In Vivo Studies in RAS Mice

To mechanistically establish the role of miR-378h in modulating the capacity of human MSCs to repair ischemic kidneys in vivo, two additional groups of RAS mice were treated with renovascular disease MSCs pretreated with a miR-378h inhibitor (n=4) or mimic (n=5). Two weeks after MSC injection, murine kidneys were studied in vivo and ex vivo, as described above.

Statistical Analysis

Statistical analysis was performed using JMP version 14 (SAS Institute, Cary, NC). Results were expressed as mean±SD. The Shapiro-Wilk test was used to test for deviation from normality. Parametric (ANOVA/Student t test) and nonparametric (Wilcoxon/Kruskal–Wallis) tests were used as appropriate, and significance accepted for P < 0.05.

Results

Patient Characteristics

Age, sex, body mass index, BP, and cholesterol fractions for participants with renovascular disease and healthy volunteer groups are presented in Table 1. Patients with renovascular disease exhibited significant bilateral stenoses, associated with higher serum creatinine levels and lower eGFR compared with healthy volunteers. Urinary protein excretion did not differ among the groups.

Table 1.

Clinical, laboratory, and demographic data of healthy volunteers and patients with atherosclerotic renovascular disease (n=3 each)

Parameters	Healthy Volunteers	Renovascular Disease
Demographics
Age, yr	60±12	66±13
Sex (female/male)	2/1	2/1
BMI, kg/m2	27.6±4.6	31.6±3.4
Related laboratory measures
Systolic BP, mm Hg	125±10	148±22
Diastolic BP, mm Hg	79±8	72±8
Mean BP, mm Hg	94±2	97±3
Total cholesterol, mg/dl	155±28	168±11
Triglycerides, mg/dl	142±57	156±56
HDL, mg/dl	47±9	55±11
LDL, mg/dl	80±18	82±13
Concomitant medication
Antihypertensives	0/3	3/3
Statins/lipid-lowering drugs	0/3	3/3
Kidney function
Renal artery velocity, cm/sa	—	257.0±26.2b
Serum creatinine, mg/dl	1±0	2±0b
eGFR-MDRD, ml/min per 1.73 m2	83±3	40±15b
Proteinuria, mg/24 h	87±22	570±880

BMI, body mass index; MDRD, modification of diet in renal disease.

^aDoppler ultrasound assessment of the more stenotic renal artery.

^bP ≤ 0.05 versus healthy volunteer.

Animal Studies

Body weight was similar among the groups (Figure 1B). Systolic BP was elevated in all RAS mice 2 weeks after surgery but decreased 2 weeks later only in RAS+healthy volunteer MSCs (Figure 1C). Right/left kidney weight ratio was equally lower in all RAS versus Sham mice (Figure 1D). Contrarily, plasma renin that increased in all RAS groups, decreased only in mice treated with healthy volunteer MSCs (Figure 1E). MSCs were detected in the poststenotic kidney parenchyma 2 weeks after injection, colocalizing with the proximal tubular cell marker PHA-E, and near CD31⁺ endothelial cells (Figure 2A, Supplemental Figures 1–4, and Supplemental Video 1, and Supplemental Video 2). We found that 19.0%±3.6%, 12.0%±3.0%, and 7.7%±1.5% of injected MSCs remained in the kidneys of Sham mice at 3, 7, and 14 days after delivery, respectively (Figure 2B). A similar proportion of healthy volunteer MSC and renovascular disease MSC engrafted into mouse poststenotic kidneys (7.6%±1.0% and 8.3%±1.1%, respectively, P = 0.33). Neither the percentage of circulating cells nor the number of kidney T cells, B cells, or monocyte/macrophages differ between Sham mice treated with human MSCs and those treated with PBS (controls) at 3, 7, or 14 days (Supplemental Figures 5 and 6). Serum creatinine levels at 4 weeks were higher in RAS+vehicle compared with Sham+vehicle decreased to normal levels in RAS+healthy volunteer MSCs, but remained elevated in RAS+renovascular disease MSCs (Figure 2C).

Figure 2.

Renovascular disease MSCs failed to improve kidney function and oxygenation. (A) Representative kidney sections of MSC-treated mice showing immunofluorescent prelabeled MSCs (far-red, arrow) incorporated into renal tubules marked with PHA-E (green) and in the vicinity of endothelial cells (CD31, green) 2 weeks after injection. (B) Engraftment of injected human MSCs (calculated as a percentage of the injected dose) observed in mouse kidneys at 3, 7, and 14 days after delivery (n=4 each) gradually declined over time yet remained appreciable. (C) Serum creatinine levels were elevated in RAS compared with Sham, improved after healthy volunteer MSCs, and less after renovascular disease MSCs (n=6 each). (D) Kidney atrophy observed in all RAS groups improved only in healthy volunteer MSC–treated mice (n=6 each). Cortical perfusion was lower (E) and hypoxia higher (F) in RAS kidneys versus Sham and improved only in RAS+healthy volunteer MSCs, whereas medullary perfusion and oxygenation remained lower in all RAS mice (n=6 each). PHA-E, phaseolus vulgaris erythroagglutinin.

Poststenotic Kidney Volume, Perfusion, and Oxygenation

Poststenotic kidney volume at 4 weeks was reduced in all RAS groups and significantly improved only in RAS+healthy volunteer MSCs (Figure 2D). Poststenotic kidneys exhibited lower cortical and medullary perfusion compared with Sham+vehicle (Figure 2E). RAS+healthy volunteer MSCs improved cortical but not medullary perfusion compared with RAS+vehicle, whereas both remained blunted in RAS+renovascular disease MSCs. Cortical and medullary hypoxia were higher in all RAS groups compared with Sham+vehicle (Figure 2F). Healthy volunteer MSCs improved cortical (but not medullary) oxygenation, whereas in RAS+renovascular disease MSCs had no effect on either.

Kidney Tissue Injury

Tubular injury and tubulointerstitial and glomerular fibrosis were higher and peritubular capillary density lower in all RAS groups compared with Sham+vehicle (Figure 3). Healthy volunteer MSCs ameliorated peritubular capillary loss, glomerular and interstitial fibrosis, and tubular injury, which remained altered in mice treated with renovascular disease MSCs.

Figure 3.

Renovascular disease MSCs did not attenuate kidney tissue injury. (A) Representative H&E and Masson trichrome staining of Sham, RAS, RAS+healthy volunteer MSCs, and RAS+renovascular disease MSCs. (B) Tubular injury and tubulointerstitial and glomerular fibrosis were higher in all RAS groups compared with Sham but decreased only in RAS+healthy volunteer MSCs (n=6 each). Peritubular capillary loss that was evident in all RAS versus Sham kidneys improved exclusively in healthy volunteer MSC–treated mice (n=6 each). H&E, hematoxylin and eosin.

Human MSC Studies

Human MSC Characterization and Differentiation

Cultured human adipose tissue–derived MSCs expressed common MSC markers (CD73⁺, CD90⁺, CD105⁺, CD34⁻, CD45⁻) (Supplemental Figure 7) and differentiated into osteocytes, chondrocytes, and adipocytes (Supplemental Figure 8), in agreement with our previous observations.^9,22

Paracrine Function

Liquid chromatography–mass spectrometry proteomic analysis identified 7977 distinct proteins in MSCs, of which seven were upregulated and seven downregulated in renovascular disease MSCs compared with healthy volunteer MSCs (Supplemental Figure 9A, Supplemental Tables 1 and 2, and Supplemental File 1). Functional analysis revealed that these 14 differentially expressed proteins are primarily involved in important cellular processes, including vesicle-mediated transport, apoptosis, and cell adhesion (Supplemental Figure 9B).

NanoSight Tracking Analysis showed healthy volunteer MSC and renovascular disease MSC release similar numbers of extracellular vesicles with comparable size distribution (Supplemental Figure 10).

mRNA-Seq

Quality measures for all mRNA-seq samples are provided in Supplemental Figures 11–16. In both renovascular disease MSC and healthy volunteer MSC, the top 100 highly expressed genes account for 31% of all mapped transcripts (Supplemental Figure 17A). These most abundant mRNAs encode mostly proteins involved in translation, including ribosomal and extracellular matrix proteins (Supplemental Figure 17B). Approximately 40% of all annotated genes (n=23,398) are expressed at levels >1 RPKM at similar proportion in renovascular disease MSC and healthy volunteer MSC (n=10,614 and n=11,007, respectively) (Supplemental Figure 17, C and D). In addition, the proportion of top 100 highly expressed genes out of all other genes, and the distribution of ribosomal and extracellular matrix genes was similar between the groups (Supplemental Figure 17D).

Differential expression analysis showed 168 genes upregulated and 513 downregulated in renovascular disease MSCs compared with healthy volunteer MSCs (Supplemental Figure 18). Functional clustering analysis revealed that upregulated genes are primarily implicated in cellular proliferation and translational activity (Supplemental Figure 19), whereas downregulated genes mainly encode for mitochondrial proteins (Supplemental Figure 20).

miRNA-Seq and Integrated (mRNA-/miRNA-Seq/mRNA-Seq) Analysis

Total and aligned reads for all miRNA-seq samples are provided in Supplemental Table 3. miRNA-seq identified 53 miRNAs upregulated and 27 downregulated in renovascular disease MSCs compared with healthy volunteer MSCs (Figure 4A and Supplemental Tables 4 and 5). Among top upregulated miRNAs is the mitomiR miR-378h (fold change=10.0, P = 0.03), which targets several mitochondria-related genes (Supplemental File 2) and modulates important mitochondrial functions, including apoptosis, lipid metabolism, and metabolic processes (Supplemental Figure 21). qPCR studies in four additional human renovascular disease MSC samples (total n=7 renovascular disease MSCs) confirmed that its expression is higher in renovascular disease MSCs versus healthy volunteer MSCs (Figure 4B), consistent with the RNA-seq findings. Integrated analysis identified 120 targets of downregulated miRNAs that overlapped with upregulated mRNAs (Supplemental Figure 22A), primarily implicated in cell proliferation and transcriptional activity (Supplemental Figure 22, B–D). By contrast, 297 overlapping targets of upregulated miRNAs and downregulated mRNAs (Supplemental Figure 23A) were mostly involved in mitochondrial functions (Supplemental Figure 23, B–D).

Figure 4.

Renovascular disease alters miRNA expression levels in MSCs. (A) Volcano plot and heatmaps of 53 upregulated (right, fold change >1.4, P < 0.05) and 27 downregulated (left, fold change <0.7, P < 0.05) miRNAs in renovascular disease MSCs versus healthy volunteer MSCs (n=3 each). The y axis corresponds to the mean expression value of −log 2 (P value), whereas the x axis displays the log2 fold change (renovascular disease MSCs/healthy volunteer MSCs) value. miRNAs with higher and lower expression in renovascular disease MSCs versus healthy volunteer MSCs are indicated with red and blue colors, respectively. (B) Elevated expression of miR-378h was confirmed by real-time qPCR in seven human renovascular disease MSCs samples (four additional renovascular disease MSC samples were tested) compared with the three healthy volunteer MSCs samples. (C–F) Treatment of human MSCs with a mimic or inhibitor of miR-378h. Expression of miR-378h (C) and the mitochondrial genes ATP synthase, H+ transporting, mitochondrial F0 complex, subunit F2 (ATP5J2) (D), B-cell chronic lymphocytic leukemia/lymphoma 2-like-12 (BCL2L12) (E), and fumarylacetoacetate hydrolase domain containing-2A (FAHD2A) (F) in renovascular disease MSC and healthy volunteer MSCs untreated or treated with lipofectamine (control), anti-miR-378h, or miR-378h mimic (n=3 each). miRNA, microRNA; qPCR, quantitative PCR.

In Vitro Studies

Cell proliferation did not differ between renovascular disease and healthy volunteer MSCs over 48 hours and remained unaltered after miR-378h modulation (Supplemental Figure 24). Mitochondrial area was similar between MSC groups, but matrix density, which was lower in renovascular disease MSCs versus healthy volunteer MSCs, was restored in renovascular disease MSCs coincubated with anti-miR-378h (Supplemental Figure 25). Contrarily, the miR-378h mimic blunted matrix density in healthy volunteer MSCs but did not affect renovascular disease MSCs. MitoSOX was higher and tetramethylrhodamine ethyl ester lower in renovascular disease MSCs compared with healthy volunteer MSCs, but both reversed in renovascular disease MSCs treated with anti-miR-378h (Supplemental Figure 26), whereas the miR-378h mimic again had reciprocal effects. Expression of miR-378h that decreased in healthy volunteer MSCs treated with anti-miR-378h increased in healthy volunteer MSCs co-incubated with miR-378h mimic (Figure 4C). Expression of this miRNA, higher in renovascular disease MSCs compared with healthy volunteer MSCs, decreased in renovascular disease MSCs treated with anti-miR-378h. Expression of the miR-378h targets ATP5J2, BCL2L12, FAHD2A, ATP5F1C, COX5A, MFN1, NDUFA, and TOMM7 decreased in healthy volunteer MSCs treated with miR-378h mimic and were lower in renovascular disease MSCs compared with healthy volunteer MSCs (Figure 4, C–F, and Supplemental Figure 27), corresponding with the RNA-seq results. Notably, co-incubation of renovascular disease MSCs with anti-miR-378h restored the expression of ATP5J2, BCL2L12, FAHD2A, COX5A, NDUFA, and TOMM7 genes to untreated healthy volunteer MSC levels.

miR-378h Modulation In Vivo

At 4 weeks, poststenotic kidney volume and cortical perfusion that were lower in RAS+vehicle and RAS+renovascular disease MSCs compared with Sham+vehicle remained reduced in RAS+renovascular disease MSCs pretreated with a miR-378h mimic, but improved in those pretreated with a miR-378h inhibitor (Supplemental Figure 28, A and B). Neither renovascular disease MSCs pretreated with a miR-378h mimic nor inhibitor improved medullary perfusion (Supplemental Figure 28, A–C). Cortical and medullary hypoxia that were higher in RAS+vehicle and RAS+renovascular disease MSCs compared with Sham+vehicle remained elevated in RAS+renovascular disease MSCs+miR-378h mimic (Supplemental Figure 28D). Renovascular disease MSCs+anti-miR-378h improved cortical (but not medullary) oxygenation.

Tubular injury and tubulointerstitial and glomerular fibrosis that were higher in RAS+vehicle and RAS+renovascular disease MSCs compared with Sham+vehicle remained elevated in RAS+renovascular disease MSCs pretreated with a miR-378h mimic, but decreased in those treated with a miR-378h inhibitor (Supplemental Figure 29). Peritubular capillary density, lower in RAS+vehicle and RAS+renovascular disease MSCs compared with Sham+vehicle, remained reduced in RAS+renovascular disease MSCs+miR-378h mimic, but improved in RAS+renovascular disease MSCs+anti-miR-378h.

Discussion

This study shows that human renovascular disease is associated with changes in both the mRNA and miRNA profiles of adipose tissue–derived MSCs and impairs their reparative capacity. Specifically, healthy volunteer MSCs decreased BP, improved poststenotic kidney function, and attenuated peritubular capillary loss, tubular injury, and fibrosis. However, these salutary effects were abrogated in mice treated with renovascular disease MSCs. RNA-seq revealed upregulated genes in renovascular disease MSCs primarily implicated in cell proliferation and transcriptional activity and downregulated genes encoding for proteins responsible for important mitochondrial functions like oxidative phosphorylation and oxidoreductase activity. Integrated miRNA-seq/mRNA-seq analysis showed that upregulated miRNAs, including the mitomiR miR-378h, primarily target mitochondrial genes, whereas downregulated miRNAs mainly target genes implicated in transcription and cell proliferation. Congruently, inhibition of miR-378h restored expression of its mitochondrial gene targets, attenuated mitochondrial impairment of human renovascular disease MSCs in vitro, and restored their ability to repair murine poststenotic kidneys in vivo. Therefore, our observations demonstrate maladaptive kidney repair by human renovascular disease MSCs and implicate miR-378h in their mitochondrial damage.

Adipose tissue–derived MSCs have been gaining considerable attention among health care professionals, possibly because of their abundance, ease of harvesting, and regenerative properties.⁵⁶ However, cardiovascular risk factors can compromise autologous MSC integrity and function. Renovascular hypertension is often accompanied by cardiovascular manifestations, which might alter the characteristics and potency of endogenous MSCs.⁵⁷ Our group has previously shown decreased proangiogenic factor expression and higher senescence in adipose tissue–derived MSCs from renovascular disease compared with normal pigs.¹⁴ Similarly, adipose tissue–derived MSCs from patients with renovascular disease show increased DNA damage, reduced migration, and lower secretion of proangiogenic factors than those from healthy kidney donors.¹⁵ This study extends those observations and demonstrates blunted renoprotective effects in human renovascular disease adipose tissue–derived MSCs.

In agreement with our previous observation,²⁸ 2 weeks after delivery, approximately 8% of injected MSCs were retained in mouse kidneys, some costaining with PHA-E, suggesting engraftment in proximal tubules. Although MSCs did not engraft into CD31⁺ endothelial cells, they were detected in their vicinity, suggesting that small vessels might benefit from paracrine effects of MSCs. Importantly, longitudinal analysis of systemic blood and kidney tissue from human MSCs–treated mice argued against their rejection.

Healthy volunteer MSCs decreased systolic BP, plasma renin, and serum creatinine, corroborating previous findings.³⁰ Furthermore, healthy volunteer MSCs decreased poststenotic kidney hypoxia, likely due to preservation of cortical perfusion and peritubular capillary density, and blunted tubular injury and tubulointerstitial and glomerular fibrosis, underscoring the reparative potential of healthy MSCs. Contrarily, many of these renoprotective effects were blunted in mice treated with renovascular disease MSCs.

To investigate potential underlying mechanisms, we probed the transcriptomic and miRNA profile of renovascular disease and healthy volunteer adipose tissue–derived MSCs using RNA-seq and identified 168 genes upregulated in renovascular disease MSCs versus healthy volunteer MSCs. Functional annotation clustering analysis linked these genes to transcriptional activity and cell proliferation, including transcription factors, protein kinases, cell differentiation markers, and growth factors. Among them are the HMG-box transcription factor SRY-Box Transcription Factor-9 (SOX9)⁵⁸ and cyclin-c (CCNC),⁵⁹ which promote cell proliferation.

Contrastingly, genes downregulated in renovascular disease MSCs mainly encode for mitochondria-related genes, implicated in oxidative phosphorylation, electron transfer, intracellular transport, and oxidoreductase activity. These include subunits of electron transport chain respiratory complexes, such as complex-I nicotinamide adenine dinucleotide:ubiquinone oxidoreductase subunit-A1 (NDUFA1), complex-IV cytochrome-C oxidase subunit-5A (COX5A), and complex-V ATP synthase-F1 subunit gamma (ATP5F1C), as well as the mitochondrial fusion marker mitofusin-1 (MFN1), the protein import protein translocase of outer mitochondrial membrane-7 (TOMM7), and the anti-apoptotic BCL2L12. Because mitochondria regulate MSC viability, plasticity, proliferative, and differentiation,^17–19 such downregulation might impair MSC functionality and regenerative potential.

Renovascular disease might alter the transcriptome of MSCs by dysregulating miRs. We identified 53 miRNAs upregulated and 27 downregulated in renovascular disease MSCs compared with healthy volunteer MSCs. Targets of downregulated miRNAs included many upregulated genes implicated in translational activity and cellular proliferation. Contrarily, targets of upregulated miRNAs coincided with mitochondria-regulating genes that were downregulated in renovascular disease MSCs, including miR-27a (dysregulates mitochondrial homeostasis⁶⁰), miR-30a (fission and apoptosis⁶¹), and miR-150b (mitochondrial translation⁶²). Therefore, renovascular disease–imposed changes in miRNA expression might alter proliferation-related and mitochondria-related genes in human MSCs.

Renovascular disease–induced changes might affect MSC paracrine function because proteomic analysis identified in renovascular disease MSCs dysregulated proteins involved in important cellular processes (e.g., vesicle-mediated transport, apoptosis, and cell adhesion), including mitochondrial import proteins like TIMP3 and TOMM7.^63,64 Nevertheless, healthy volunteer MSC and renovascular disease MSC release similar numbers and sizes of extracellular vesicles, suggesting preservation of this paracrine functional aspect.

Conspicuously, miR-378h was ten-fold upregulated in renovascular disease MSCs versus healthy volunteer MSCs. miR-378 family members participate in aging and cancer biology^65,66 and modulate angiogenesis.⁶⁷ Moreover, miR-378h targets key genes involved in cell proliferation and mitochondrial function, including apoptosis, lipid metabolism, and metabolic processes, like CDC20B⁴⁸ and MT-ATP6.⁶⁸ Indeed, a miR-378h mimic decreased in healthy volunteer MSCs mitochondrial matrix density and membrane potential and increased mitochondrial reactive oxygen species production due to membrane depolarization.⁶⁹ Conversely, anti-miR-378h protected renovascular disease MSC mitochondria. Basal expression of ATP5F1C, COX5A, MFN1, NDUFA1, TOMM7, and BCL2L12 was downregulated in renovascular disease MSCs, validating RNA-seq findings, but anti-miR-378h upregulated most, and mitochondrial superoxide production fell. Interestingly, cell proliferation that in RNA-seq distinguished renovascular disease MSC was functionally similar to healthy volunteer MSCs and remained unchanged after miRNA modulation. Possibly, upregulation of these genes might be counteracted by other regulatory mechanisms (e.g., post-translational modifications) or manifest on an injurious stimulus.

Importantly, anti-miR-378h pretreatment restored the renal reparative capacity of renovascular disease MSCs. Specifically, kidney volume, perfusion, and oxygenation improved in renovascular disease MSCs pretreated with anti-miR-378h, as did tubular injury, capillary loss, and tubulointerstitial and glomerular fibrosis. Overall, our observations underscore the important role of this mitomiR in mediating renovascular disease–induced mitochondrial damage and dysfunction in MSC.

Our study is limited by shorter duration of murine RAS compared with human renovascular disease. Yet, this model recapitulates many important features of human poststenotic kidneys.^28,70 Although xenotransplantation may elicit rejection and compromise MSC engraftment, we found showed no evidence of cellular rejection.⁹ The increase in kidney volume in vivo but not weight ex vivo in renovascular disease MSCs might be due to improved kidney function and flow. mRNA-seq and miRNA-seq analyses involved a relatively small sample size (which is not uncommon in omics studies). Although patients with renovascular disease were treated with medications that may attenuate the effect of hypertension and atherosclerotic disease, clear alterations in mRNA/miRNA expression remained detectable in MSCs. MSCs were harvested adhering to routine protocols^9,11 and expanded for only three passages to avoid long-term culture senescence.⁷¹ Changes in transcriptomic and miRNA profiles were further validated by qPCR. Additional regulatory mechanisms could have also affected gene expression in renovascular disease MSCs, and the mechanisms of miRNA dysregulation remain to be elucidated. Finally, CKD, hypertension, or atherogenic factors may all dysregulate the mRNA and miRNA cargo and mitochondrial function of renovascular disease MSCs. Thus, additional studies are needed to confirm our findings in a larger cohort of patients with renovascular disease and to define the role of miR-378h in renovascular disease MSCs.

In summary, our study shows that renovascular disease impairs the capacity of human adipose tissue–derived MSCs to repair murine poststenotic kidneys, partly by post-transcriptional regulation of mitochondria-related genes through miR-378h. These observations may reflect a key functional deficit in endogenous MSCs that compromises their reparative capacity in individuals with renovascular disease. Taken together, our findings support developing novel strategies to preserve the therapeutic efficacy of MSCs in patients with renovascular disease.

Disclosures

Disclosure forms, as provided by each author, are available with the online version of the article at http://links.lww.com/JSN/E773.

Funding

L.O. Lerman: NHLBI Division of Intramural Research (HL158691), National Institute of Diabetes and Digestive and Kidney Diseases (DK120292, DK122734, and DK100081), National Institute on Aging (AG062104). A. Eirin: National Institute of Diabetes and Digestive and Kidney Diseases (DK129240) and Regenerative Medicine Minnesota (RMM-091620-DS-004).

Author Contributions

Conceptualization: Alfonso Eirin, Yamei Jiang, Lilach O. Lerman.

Data curation: Alfonso Eirin, Autumn G. Hughes, Sara Kazeminia, Bo Lu, Brandon Lu, Sarosh Siddiqi, Hui Tang, Stephen C. Textor, Li Xing, Ailing Xue, Xiang Y. Zhu.

Formal analysis: Alfonso Eirin, Autumn G. Hughes, Yamei Jiang, Sara Kazeminia, Bo Lu, Brandon Lu, Sarosh Siddiqi, Hui Tang, Li Xing, Ailing Xue, Xiang Y. Zhu.

Funding acquisition: Lilach O. Lerman.

Investigation: Alfonso Eirin, Autumn G. Hughes, Yamei Jiang, Sara Kazeminia, Amir Lerman, Lilach O. Lerman, Brandon Lu, Hui Tang, Xiang Y. Zhu.

Methodology: Alfonso Eirin, Yamei Jiang, Sara Kazeminia, Amir Lerman, Bo Lu, Brandon Lu, Hui Tang, Stephen C. Textor, Li Xing, Xiang Y Zhu.

Software: Alfonso Eirin, Autumn G. Hughes, Yamei Jiang, Sara Kazeminia, Brandon Lu, Sarosh Siddiqi, Hui Tang, Stephen C. Textor, Li Xing, Xiang Y. Zhu.

Supervision: Lilach O. Lerman.

Validation: Autumn G. Hughes, Amir Lerman, Lilach O. Lerman, Hui Tang, Stephen C. Textor, Li Xing, Ailing Xue.

Visualization: Amir Lerman, Bo Lu, Sarosh Siddiqi, Stephen C. Textor.

Writing – original draft: Alfonso Eirin.

Writing – review & editing: Autumn G. Hughes, Yamei Jiang, Sara Kazeminia, Amir Lerman, Lilach O. Lerman, Bo Lu, Brandon Lu, Sarosh Siddiqi, Hui Tang, Stephen C. Textor, Li Xing, Ailing Xue, Xiang Y. Zhu.

Data Sharing Statement

The data supporting these findings are available at Figshare: https://doi.org/10.6084/m9.figshare.24059709.v1, https://doi.org/10.6084/m9.figshare.24060099.v1, https://doi.org/10.6084/m9.figshare.26086225.v1.

Supplemental Material

This article contains the following supplemental material online at http://links.lww.com/JSN/E768, http://links.lww.com/JSN/E769, http://links.lww.com/JSN/E770, http://links.lww.com/JSN/E774, http://links.lww.com/JSN/E782.

Supplemental Methods.

Supplemental Table 1. Proteins upregulated in human renovascular disease MSCs compared with healthy volunteer MSCs.

Supplemental Table 2. Proteins downregulated in human renovascular disease MSCs compared with healthy volunteer MSCs.

Supplemental Table 3. Quality measures of miRNA-seq analysis of human renovascular disease MSCs and healthy volunteer MSCs.

Supplemental Table 4. miRNAs upregulated in human renovascular disease MSCs compared with healthy volunteer MSCs.

Supplemental Table 5. miRNAs downregulated in human renovascular disease MSCs compared with healthy volunteer MSCs.

Supplemental Figure 1. Human MSCs were detectable in mouse kidneys 2 weeks after delivery.

Supplemental Figure 2. Human MSCs were detectable in mouse kidneys 2 weeks after delivery.

Supplemental Figure 3. Human MSCs were not detectable in mouse kidney glomeruli 2 weeks after delivery.

Supplemental Figure 4. Human MSCs were detectable in mouse kidney interstitium 2 weeks after delivery.

Supplemental Figure 5. Delivery of human MSCs did not elicit an evident systemic immune response in mice.

Supplemental Figure 6. Delivery of human MSCs in mice was not associated with kidney tissue immune rejection.

Supplemental Figure 7. Human adipose tissue–derived MSC characterization.

Supplemental Figure 8. Human adipose tissue–derived MSC trilineage differentiation.

Supplemental Figure 9. Proteomic profile of healthy volunteer MSC and renovascular disease MSC.

Supplemental Figure 10. Concentrations and size distribution of healthy volunteer MSC–derived and renovascular disease MSC–derived EVs. Exosome nanoparticle tracking analysis showing similar concentration and size distribution of healthy volunteer MSC–derived and renovascular disease MSC–derived EVs.

Supplemental Figure 11. Quality measures of mRNA-seq: Sample 1.

Supplemental Figure 12. Quality measures of mRNA-seq: Sample 2.

Supplemental Figure 13. Quality measures of mRNA-seq: Sample 3.

Supplemental Figure 14. Quality measures of mRNA-seq: Sample 4.

Supplemental Figure 15. Quality measures of mRNA-seq: Sample 5.

Supplemental Figure 16. Quality measures of mRNA-seq: Sample 6.

Supplemental Figure 17. mRNA-seq of renovascular disease MSC and healthy volunteer MSC.

Supplemental Figure 18. Renovascular disease alters mRNA expression levels in MSCs.

Supplemental Figure 19. Genes upregulated in renovascular disease MSCs are largely implicated in transcription and cell proliferation.

Supplemental Figure 20. Genes downregulated in renovascular disease MSCs primarily encode mitochondrial proteins.

Supplemental Figure 21. Functional analysis of miR-378h targets.

Supplemental Figure 22. miRNAs downregulated in renovascular disease MSCs modulate transcription and cell proliferation.

Supplemental Figure 23. miRNAs upregulated in renovascular disease MSCs modulate mitochondrial function.

Supplemental Figure 24. Renovascular disease does not alter MSC proliferation.

Supplemental Figure 25. miR-378h inhibition improves mitochondrial structure in renovascular disease MSCs.

Supplemental Figure 26. miR-378h inhibition improves mitochondrial function in renovascular disease MSCs.

Supplemental Figure 27. miR-378h inhibition upregulates mitochondrial genes in renovascular disease MSCs.

Supplemental Figure 28. miR-378h inhibition improves the functional renal reparative capacity of renovascular disease MSCs.

Supplemental Figure 29. miR-378h inhibition improves the structural renal reparative capacity of renovascular disease MSCs.

Supplemental File 1. Proteomics raw data. Individual protein expression profiles of human renovascular disease and healthy volunteer MSCs (n=6 each).

Supplemental File 2. miR-378h mRNA targets. List of all mRNA targets of miR-378h obtained from miRWalk 3.0.

Supplemental Video 1. Human MSCs were detectable in mouse kidneys 2 weeks after delivery. Representative video of confocal microscopy kidney sections of MSC-treated mice showing several immunofluorescent prelabeled MSCs (far-red) surrounding renal tubules marked with PHA-E (green).

Supplemental Video 2. Human MSCs were not detectable in mouse kidney glomeruli 2 weeks after delivery. Representative video of confocal microscopy CD31 (green)-stained kidney sections of MSC-treated mice, showing that immunofluorescent prelabeled MSCs (far-red, arrow) were not observed in the glomerulus 2 weeks after injection.