Genome Engineering of Human Urine-Derived Stem Cells to Express Lactoferrin and Deoxyribonuclease

Abstract

Urine-derived stem cells (USCs) are adult kidney cells that have been isolated from a urine sample and propagated in tissue culture on gelatin-coated plates. Urine is a practical and completely painless source of cells for gene and cell therapy applications. We have isolated, expanded, and optimized transfection of USCs to develop regenerative therapies based on piggyBac transposon modification. USCs from a healthy donor sample were isolated according to established protocols. Within 2 months, 10 clones had been expanded, analyzed, and frozen. Fluorescence-activated cell sorting analysis of individual clones revealed that all 10 clones expressed characteristic USC markers (97–99% positive for CD44, CD73, CD90, and CD146; negative for CD31, CD34, and CD45). The isolated USCs were successfully differentiated along the osteogenic, adipogenic, and chondrogenic lineages, suggesting multipotent differentiation capacity. Additionally, the USCs were differentiated into podocytes positive for NEPHRIN (NPHS1), podocalyxin, and Wilms tumor 1 (WT1). Transfection of USCs with a strongly expressing Green fluorescent protein plasmid was optimized to achieve 61% efficiency in live cells using several commercially available lipophilic reagents. Transgene promoters were compared in five luciferase-expressing piggyBac transposons by live animal imaging. The CMV promoter produced the highest luciferase signal, followed by EF1-α. Finally, HEK-293 and USCs were transfected with piggyBac transposons expressing lactoferrin and DNase1 for treatment of acute kidney injury associated with rhabdomyolysis. We found that both proteins were expressed in USCs and that lactoferrin was successfully secreted into the cell culture media. In conclusion, USCs represent a clinically relevant cell type that can express nonviral transgenes.

Impact statement

Acute kidney injury (AKI) affects over 13 million people worldwide each year, with hospitalization rates on the rise. There are no therapies that directly regenerate the kidney after AKI. Each human kidney contains approximately one million nephrons that process ∼100 L of urinary filtrate each day. Thousands of kidney cells become detached and are excreted in the urine. A small percentage of these cells can be clonally derived into urine-derived stem cells. We have optimized methods for genome engineering of adult human urine-derived stem cells for future applications in regenerative approaches to treat kidney injury.

Introduction

Urine-derived stem cells (USCs) can be easily isolated, exhibit multilineage differentiation capabilities, and have a high capacity for self-renewal and expansion.¹ There are ∼1 million nephrons per kidney that process ∼100 L of glomerular filtrate per day. Some of these live cells become detached and can be collected in the urine.² USCs represent a specific subpopulation of cells shed in the urine that do not grow well in media for bone marrow-derived mesenchymal stromal cells (MSCs) or have similar morphology yet express some markers in common with MSC.³ These markers include ITGB1 (Integrin Beta 1, CD29), CD44 (Hyaluronate Receptor, Phagocytic Glycoprotein 1), NT5E (5′-Nucleotidase Ecto, CD73), THY1 (cell surface glycoprotein CD90), ENG (Endoglin, CD105), and ALCAM (Activated Leukocyte Cell Adhesion Molecule, CD166). USCs do not express most hematopoietic stem cell markers, except Platelet and Endothelial Cell Adhesion Molecule 1 (PECAM1, CD31), Major Histocompatibility Complex, Class 1 (MHC-1), and Class 2 (HLA-DR).³

They generally express pericyte marker Melanoma Cell Adhesion Molecule (MCAM, CD146), podocyte and embryonic stem cell markers Stage-Specific Embryonic Antigen 4 (SSEA4), as well as the monoclonal antibodies TRA-1-60 and TRA-1-81 recognizing the type 1 lactosamine epitopes of podocalyxin (PODXL).^3,4 They are capable of multipotent differentiation³ and have been successfully differentiated into endothelial,⁵ urothelial,⁶ neuronal,⁷ alveolar,⁸ and renal cell types.⁹ Cells isolated from the urine have been reprogrammed into induced pluripotent stem cells (iPSCs) for a large array of disease modeling approaches,^10–16 suggesting that USCs can be isolated from patients with many kinds of diseases. USCs may be utilized for tissue engineering approaches, including repair of the ureter.^17,5

Urine is a fast, simple, practical, and completely painless cell source. Stem cells from other sources, such as iPSCs, require long, expensive culture periods, whereas USCs can be obtained within 2 weeks.¹⁸ USCs have unique promise in terms of the number of cells that can be produced in a short period, even after storage of urine at 4°C for 24 h.² USCs have a higher proliferation rate, enhanced colony-forming ability, and decreased immunogenicity compared with adipose-derived MSCs.¹⁹ Of the USC clones tested, 60% had long telomeres and expressed telomerase, suggestive of a potential for many future cell divisions, yet the USC did not form teratomas in nude mice.³ USCs have also been found to be capable of robust immune cell inhibition.¹⁹ In USCs isolated from female transplant recipients of male kidneys, the Y-chromosome was found by fluorescence in situ hybridization and polymerase chain reaction (PCR) analysis, suggesting that the cells came primarily from either the kidney or the transplanted portion of the ureter.³

Furthermore, the presence of the renal markers CD13, CD44, NR3C2, CD146, CD224, PAX2, and PAX8 in cultured USCs also imply a renal origin.³ Therefore, USCs likely originate from transitional cells within the glomerulus at the parietal cell–podocyte interface, similar to adult parietal epithelial multipotent progenitors.²⁰

Human USCs have been tested in multiple animal models of disease. Incorporation of USCs significantly accelerated wound healing in rabbits, resulting in increased cytokeratin area and microvessel density as compared with rabbits treated with polycaprolactone/gelatin nanofibrous membrane alone.²¹ Neocartilage formation was stimulated by injection of USCs combined with hyaluronic acid hydrogels into rabbit knee joints.²² In streptozotocin-induced diabetic mice, USC-isolated exosomes accelerated wound closure and angiogenesis.²³ Human USCs were also found to prolong median survival time, improve blood glucose, and increase CD31 expression in pancreatic islets.²⁴ In a rat model of ischemic kidney injury, human USCs were shown to positively contribute to repair.²⁵ Green fluorescent protein (GFP)-transduced USCs embedded in hydrogel were able to integrate into tubular epithelium 1 week postinjection. USCs engrafted into the injured tubule and lowered blood urea nitrogen, creatinine, and inflammatory markers in a rat model of acute kidney injury (AKI).²⁵ Therefore, USCs show promise for repair and regeneration in multiple organs.

Extracellular traps (ETs) composed of genomic DNA fibers and granule proteins are extruded from neutrophils, monocytes, and macrophages,.^26,27 In mouse models of rhabdomyolysis, macrophages were found to release macrophage ETs (METs).²⁸ Extracellular DNA release was decreased in these mice when macrophages and platelets were depleted, but not neutrophils, suggesting that macrophages and platelets are involved in rhabdomyolysis-induced ET release and AKI. These ETs are the result of a process called “ETosis,” whereby macrophages release a net-like material of cellular DNA and accompanying histones. The purpose of released DNA is to inhibit the growth of microorganisms during infection; however, in the absence of infection, such a reaction may instead cause a pathologic response through autoinflammatory processes. METs were shown to be responsible for much of the kidney damage associated with rhabdomyolysis.²⁸ Accelerated enzymatic digestion of extracellular DNA may be achieved through administration of DNase1, a human enzyme in the kidney that degrades DNA.²⁹

In addition, the natural iron-binding glycoprotein, lactoferrin, or lactotransferrin (LTF), has antiviral, antibacterial, anti-inflammatory, and antioxidant properties.³⁰ Bovine lactoferrin has been previously shown to inhibit METs through either direct inhibition or chelation of iron.^31,32 LTF has an isoelectric point of 8.0–8.5, therefore it possesses a positive charge at a physiological pH of 7.4.³³ This positive charge may prevent the spread of ETs through charge–charge interactions.³¹ Therefore, in this work we sought to develop transfected USCs as protein factories expressing the potentially therapeutic proteins DNase1 and lactoferrin from piggyBac transposons for stable genome integration.

Methods

Isolation and characterization of USCs

Sterile USC media (Dulbecco's modified Eagle's medium [DMEM] with 1 g/L glucose, L-glutamine, sodium pyruvate (Corning, Glendale, AZ) supplemented with 2% HyClone fetal bovine serum (FBS; Thermo Fisher Scientific, Waltham, MA), 10 ng/mL human epidermal growth factor (hEGF; PeproTech, Cranbury, NJ), 2 ng/mL platelet-derived growth factor (PDGF-BB GF149; Millipore, Burlington, MA), 1 ng/mL transforming growth factor-B (Human TGFB 100-21; PeproTech), 2 ng/mL basic fibroblast growth factor (bFGF; Sigma-Aldrich, Burlington, MA), 0.5 μM hydrocortisone (Sigma), 25 μg/mL recombinant human insulin (Sigma), 20 μg/mL transferrin (Calbiochem/Millipore), 549 ng/mL adrenaline (Sigma), 50 ng/mL triiodothyronine (T3; Sigma), and 1% penicillin–streptomycin (Mediatech/CellGro-Corning)) was prepared without filtering based on Guan.³⁴

The Vanderbilt Institutional Review Board (IRB) approved this study prior to initiation of the research. Urine samples were obtained with informed consent from healthy 18–25-year-old male donors and refrigerated for 24 h, then equilibrated to room temperature and subjected to urine dipstick analysis with Fisherbrand 10SG Urine Reagent Strips (Thermo Fisher). Urine was centrifuged at 400 × g and washed twice with phosphate-buffered saline without calcium and magnesium (PBS; Mediatech/CellGro-Corning) at RT. The pellet from the 24-h urine sample was distributed across one gelatin-coated 24-well plate. Gelatin coating was performed through incubation of a standard TC-treated multiwell plate with 0.1% gelatin diluted from 2% gelatin solution (Sigma) at 37°C for 30 min. Cells were cultured at 37°C, 5% CO₂. Media were changed twice per week during expansion up to 20 passages. Individual clones were characterized by flow cytometry analysis in the VA TVHS Flow Cytometry Laboratory for expression of CD31, CD34, CD44, CD45, CD73, CD90, and CD146 (BD Biosciences, East Rutherford, NJ).

In vitro multilineage differentiation assays of USCs and analysis

Following differentiation along the osteogenic, chondrogenic, and adipogenic lineages, the multipotent differentiation efficiency of USCs was assessed. Osteocytes were generated by treating the USCS with mesenchymal stem cell osteogenic differentiation medium (Cyagen Biosciences, Santa Clara, California). Adipocytes were also generated the same way using mesenchymal stem cell adipogenic differentiation medium (Cyagen Biosciences). For chondrocytes, the USCs were dissociated and aggregated by centrifuging at 500 × g for 5 min. The pellets were grown on transwells (Corning) for a total of 28 days in chondrogenic differentiation medium (Cyagen Biosciences). Oil Red O (Sigma-Aldrich) staining was used for adipocytes, whereas Alizarin Red (Sigma-Aldrich) staining was used to measure mineralization in osteocytes. Chondrocytes were stained with Safranin O (Sigma-Aldrich) and Alcian Blue (Sigma-Aldrich), and the results of all the staining were examined under a phase-contrast microscope.

Podocyte differentiation was carried by treating USCs with podocyte differentiation VRAD media³⁵ composed of DMEM/F12 (Sigma-Aldrich) supplemented with 10% FBS (Fisher Scientific), 100 nM vitamin D3 (Fisher Scientific), and 100 mM retinoic acid (STEMCELL Technologies, Vancouver, Canada) for 10 days. The resulting podocytes were stained for podocyte markers PODXL, Wilms tumor 1 (WT1), and NEPHRIN (NPHS1).

Transcriptome comparison of USC clones

RNA was harvested from six clones of isogenic USCs by the RNeasy Kit (Qiagen, Germantown, MD). The samples were quantified on NanoDrop (NanoDrop/Thermo Fisher) and Qubit (Invitrogen/Thermo Fisher, Waltham, MA) machines. The RNA was processed and sequenced (Novogene, Bejing, China) according to standard workflow practices of total RNA qualification, mRNA enrichment, double-stranded cDNA synthesis, end repair, poly-A and adaptor addition, fragment selection and PCR, library quality assessment, and sequencing (Illumina, San Diego, CA). Data were mapped against the hg19 human reference genome assembly. Pathway analysis was performed by Novogene.

USC transfection reagent comparison

After expansion for 8 weeks, USCs were transfected at 40% confluency in 6-well tissue culture dishes (Falcon/Fisher Scientific, Waltham, MA) with 500 ng/μL ZEOpTCMV-GFP using a variety of commercially available lipophilic transfection reagents according to the manufacturer's protocols (as shown in Fig. 4). All conditions were tested in parallel on the same clone 1J. Endotoxin-free plasmid DNA was prepared according to the manufacturer's instructions (Qiagen) and eluted in water or Buffer QR (Mirus Bio, Madison, WI) for all transfection experiments. USCs were examined for GFP expression by imaging on the ZOE fluorescent imager (Bio-Rad, Hercules, CA) in the Vanderbilt O'Brien VCKD Genome Engineering Core at 24- and 48 h post-transfection. After imaging, cells were trypsinized with 0.25% trypsin ethylenediaminetetraacetic acid ( Mediatech/CellGro/Corning, Corning, NY) until single cells were visible, then collected and centrifuged at 150 × g for 5 min. Cell pellets were washed once with PBS and resuspended in PBS and Propidium Iodide (PI; Sigma-Aldrich; 60 μl of 50 μg/mL PI in 3 mL of PBS). Control samples for PI calibration were prepared in PBS alone.

FIG. 4.

USC transfection with various commercial lipophilic reagents. USCs from the same clone were transfected in parallel with the commercially available reagents shown at the indicated ratios of DNA to reagent as suggested by the manufacturer's protocols. (A) Representative merged brightfield and green fluorescence images were taken on the ZOE fluorescent imager. (B) Quantification of flow cytometry analysis of the percentage of total cells that were Live GFP+ (green) and the percentage of cells within the live cell gate that were GFP+ (blue). PI was used to assess viability. (C) Flow cytometry results for each condition. At least 1,000 cells were counted. Cells were gated as intact (black); viable (blue); and GFP+ (green). GFP, green fluorescent protein; PI, Propidium Iodide.

Fluorescence-activated cell sorting (FACS) analysis was performed in the TVHS VA Flow Cytometry Laboratory.

USC promoter comparison

To evaluate various promoters for the expression of a transgene in USCs, a pool of USCs from different clones were plated in 6-well dishes. The cells were transfected at ∼75–90% confluence using Lipofectamine 2000 ( Invitrogen/Thermo Fisher) with plasmid DNAs expressing enhanced firefly luciferase: pT-effLuc-Thy1.1³⁶; pTpBCAGLuc^37,38; pTEeL; pTPeL; and pTGeL^39,40 from the promoters CMV, CAG,⁴¹ EF1-α,⁴² podocin,⁴³ and γGT,⁴⁴ respectively. Control cells were transfected with pMAX-GFP (Lonza, Basel, Switzerland). Transfections were carried out at a ratio of 5 μl reagent to 2.5 μg plasmid DNA. The following day, 10 μl of Xenolight D-luciferin potassium salt (PerkinElmer, Waltham, MA) diluted in PBS to a concentration of 30 mg/mL⁴⁵ was added to each well. Cells were imaged for luciferase expression using an In Vivo Imaging System (IVIS) spectrum bioluminescent and fluorescent imaging system (PerkinElmer) in the Vanderbilt Institute for Small Animal Imaging core. Bioluminescent images were analyzed using Living Image Software (PerkinElmer).

Optimization of culture conditions post-transfection

USCs were transfected when 75–90% confluent with GFP plasmid using the Lipofectamine Kits (Stem, 2000, and LTX) according to the manufacturer's instructions (Invitrogen/Thermo Fisher). At 30 min, 1 h, 2 h, or 4 h post-transfection, medium in each well was changed to fresh medium containing 25% USC-conditioned medium saved from previous cultures of USC. The USC-conditioned medium provides secreted factors and exosomes from healthy USC to the culture. Transfected and untransfected control wells with no media change were included. Images were obtained on the ZOE at 24 and 48 h post-transfection.

DNase1 and lactoferrin transposon plasmids

Transposons were made by VectorBuilder (Chicago, IL). The transposon expressing human DNase1, pPB[Exp]-EF1A>hDNASE1[NM_001351825.1] is catalog number VB190418-1107wss. The transposon expressing human lactoferrin, pPB[Exp]-EF1A>hLTF[NM_002343.5]* is catalog number VB190418-1238puv.

Western blotting

HEK-293 cells were cultured in Eagle's MEM containing 10% FBS to 50–70% confluency and transfected using FuGene 6 according to the manufacturer's instructions (Promega). USCs were cultured in the USC media described above. Transposon plasmids expressing DNase 1 or lactoferrin were used for transfection of both cell types. Cells were harvested in cold 1 × PBS and whole cell lysates were prepared in RIPA buffer (Sigma) supplemented with mammalian protease inhibitor mix (Sigma) and phosphatase inhibitor mix (Phos-Stop, Roche). Cell lysates with 10 μg total protein were run on 4–12% NuPage Bis–Tris gels (Invitrogen) with 3-(N-morpholino)propanesulfonic acid (MOPS) buffer followed by transfer to nitrocellulose for immunoblotting. Blots were incubated in anti-DNase1 antibody at 1:1000 (Abcam); anti-lactoferrin antibody at 1:1000 (Abcam); or mouse β-actin (Novus) at 1:10,000 at 4°C overnight and detected using IRD-labeled secondary antibodies (LICOR) with an Odyssey Infrared Imaging System (LICOR).

Immunocytochemistry

Podocytes generated were fixed with 4% paraformaldehyde in PBS for 15 min and were washed with PBS. Then cells were permeabilized and blocked in PBS-BT (6 g bovine serum albumin, 10% Triton X-100 and 2% Sodium Azide in PBS). After blocking, cells were incubated with primary antibodies for 45 min to 2 h in RT followed with secondary antibody staining for at least 45 min. The following antibodies were used: rabbit anti-PODXL (1:200, 18150-1-AP, ProteinTech), rabbit anti-WT1 (1:200, 12609-1-AP; ProteinTech), sheep anti-NEPHRIN (1:200, AF4269-SP; R&D Systems). Images were taken using the ZOE fluorescent microscope (Bio-Rad, California).

Experiment

USCs were obtained from urine samples by centrifugation and plating of the pellet onto gelatin-coated multiwell plates (Fig. 1A, B). USCs from a healthy 23-year-old male donor were individually expanded as clones and subjected to analysis to examine interclone variability of USCs derived from a single donor. USCs expressed markers, such as CD44 and CD73, by immunostaining (Fig. 1C). Analysis of individual clones by FACS characterization revealed consistent expression of characteristic USC markers (CD44, CD73, CD90, and CD146 positive; CD31, CD34, and CD45 negative; Fig. 1D, E).

FIG. 1.

Isolation of human USCs. (A) Urine was collected from healthy donors, centrifuged to obtain a cell pellet, and plated into gelatin-coated 24-well dishes. Wells with clones growing were expanded to four 10 cm plates before cell marker analysis and cryobanking. (B) Brightfield images showing USC isolation by culture on gelatin-coated plates in heavily supplemented media. Day 1, debris and crystals were present in the 24-h urine pellet. Day 15, debris has been washed away and cells divide. Day 24, colonies formed. Day 60, after passage, the cells were cryobanked and expanded. Images are representative and were acquired on the ZOE fluorescent imager (Bio-Rad) (n = 3). (C) Immunofluorescent staining for USC markers CD44 (red) and CD73 (green), costained with DAPI (blue) and merged. (D) USCs express a typical USC marker profile by FACS. (E) Percentage of live cells staining positively for each marker (n = 10 ± SD). Scale bars, 100 μm. DAPI, 4′,6-diamidino-2-phenylindole; FACS, fluorescence-activated cell sorting; SD, standard deviation; USC, urine-derived stem cell.

USCs were differentiated for more than 2 weeks on gelatin-coated coverslips by culturing in media along osteogenic, adipogenic, and chondrogenic lineages according to commercially available protocols. Differentiation and staining of the USCs was performed using commercial protocols originally developed for MSCs (Fig. 2A). Note that although limited adipogenic differentiation as compared with adipose-derived MSCs, the Oil Red O stain shown in this study is consistent with previous results for USCs.¹⁹ Human USCs were capable of osteogenic, adipogenic, and chondrogenic differentiation under the right induction conditions, according to the results of trilineage differentiation induction. Next, we added VRAD medium to the USCs and observed morphologic changes, including foot-like extensions consistent with a podocyte phenotype (Fig. 2B). These cells also stained positive for the podocyte markers NEPHRIN, PODXL, and WT1 (Fig. 2C). These data confirmed the potential for differentiation of the USCs along multiple lineages, as expected based on the literature.

FIG. 2.

USCs have multipotent differentiation potential. (A) USC differentiation into osteogenic, adipogenic, and chondrogenic lineages in vitro. USCs differentiated along the osteogenic lineage were stained for Alizarin Red. USCs differentiated along the adipogenic lineage were stained for Oil Red O. USCs differentiated along the chondrogenic lineage were stained for Alcian Blue and Safranin-O. (B) USCs were differentiated into podocytes with VRAD (podocyte differentiation medium) medium. Brightfield images show morphologic changes, including the development of characteristic foot processes. (C) Immunofluorescent staining of differentiated USC-podocytes for podocyte markers, including PODXL, NEPHRIN (NPHS1), and WT1 Scale bars, 100 μm. PODXL, podocalyxin; WT1, Wilms tumor 1.

RNA was harvested from six clones and subjected to RNA sequencing to determine the differences between transcriptomes of individual clones. The Pearson correlation coefficients were all larger than 0.92 (Fig. 3A) indicating a highly similar overall expression pattern. Venn diagram analysis of five clones showed over 10,000 genes expressed in common between the clones (Fig. 3B). Cluster analysis showed that although some genes are different between the clones, there are not clusters of genes that are shared by more than one clone (Fig. 3C). Gene ontology comparison (Fig. 3D) and Kyoto Encyclopedia of Genes and Genomes pathway analysis (Fig. 3E) suggest that clones may have modest differences in extracellular matrix, metabolism, adhesion, and cytokine signaling. Taken together, these data indicate that the clones isolated from the same urine sample are similar overall in their gene expression pattern.

FIG. 3.

RNA-sequencing of individual USC clones. Six clones of USCs were subjected to RNAseq comparison. (A) Pearson correlation analysis. (B) Venn diagram of gene expression for five of the six clones showing >10k genes in common for the different clones. (C) Cluster analysis showed the differences between the six clones were irregular, suggestive of individual origins. (D) Gene ontology comparison revealed a theme of differences in extracellular matrix pathways, including collagen catabolism. (E) KEGG pathway analysis for clone A versus clone K found the most differences in cytokine–cytokine pathway interaction. KEGG, Kyoto Encyclopedia of Genes and Genomes.

As lipophilic reagents are simpler to use, more cost-effective, and often provide higher survival rates than electroporation, we chose to focus on these reagents to optimize transfection of the USCs. Live imaging was performed to visualize the GFP-positive (GFP+) cells (Fig. 4A) followed by flow cytometry analysis for GFP using PI to determine viability. Shown are the percentages of live GFP+ cells and the percentage of live cells that were also GFP+ from transfection with each reagent (Fig. 4B, C). We found that the Lipofectamine suite of reagents (Lipofectamine STEM, Lipofectamine 2000, and Lipofectamine LTX Plus; Invitrogen Life Technologies, Carlsbad, CA) had superior performance as compared with the other reagents tested, including the FuGene suite of reagents.

Next, we checked various promoters to determine which produced optimal protein expression from the USCs. We compared several enhanced firefly luciferase-expressing piggyBac transposons following transfection of pooled clones of USCs (Fig. 5). Cells transfected with an equivalent quantity of plasmid expression enhanced green fluorescent protein (EGFP) were used as a negative control. We found that plasmids expressing luciferase from cell-type specific kidney promoters podocin (pT-PeL) or gamma-glutamyl transferase (pT-GeL) had no detectable luciferase expression, similar to the negative control of EGFP-transfected cells. In contrast, the constitutive promoters CMV (ZEOpT-effLuc-Thy1.1), CAG (pT-CAG-Luc), and EF1-α (pT-EeL) all produced levels exceeding the negative control. This suggested that efficient USC transfection could be achieved with any of these three promoters (Fig. 5).

FIG. 5.

Promoter optimization and selection. USCs were transiently transfected using Lipofectamine 2000 with 2.5 μg of the transposon plasmid DNA indicated. Promoter name is in parenthesis below the transposon plasmid name. The first well was transfected with an EGFP transposon plasmid as a negative control for luciferase expression (EGFP; neg ctrl). Cells were analyzed for luciferase expression after addition of luciferin to each well on the IVIS spectrum. As a measure of the amount of luciferase present, the radiance (p/sec/cm²/sr) is shown on the right as a heatmap color scale with a minimum radiance of 1.74e5 depicted in purple and a maximum radiance value of 3.24e6, depicted in red. EGFP, enhanced green fluorescent protein; IVIS, In Vivo Imaging System.

We noticed a loss of transgene expression and altered cell morphology in USCs at 48 h post-transfection with Lipofectamine reagents, so we sought to optimize our transfection protocol by changing the media after transfection. USCs transfected with an EGFP-expressing plasmid were changed to 25% USC-conditioned medium 30 min, 1 h, 2 h, or 4 h post-transfection. Transfected and untransfected control wells were included with no media change (Fig. 6). GFP+ cells were no longer present 2 days after transfection in wells that did not get a media change, whereas in the wells that received a media change, GFP-positive cells were retained, with optimum expression in wells changed 1–2 h post-transfection. Based on these results, we changed the USC media 1–2 h after transfection with a Lipofectamine reagent in subsequent experiments.

FIG. 6.

Optimization of USCs transfected by Lipofectamine and PLUS reagent with pCMV-GFP. USC medium was changed to 25% conditioned medium at the time points indicated following transfection. The efficiency of GFP transfection was assessed by imaging. Scale bar represents 100 μm.

To create USCs capable of stable long-term expression of proteins that could assist in recovery following rhabdomyolysis, we designed two piggyBac transposons that expressed either human DNase1 or lactoferrin following the EF1-α promoter (Fig. 7A). We first tested expression from these transposons in HEK-293 cells transfected using FuGene and performed western blot analysis of cell lysates (Fig. 7B). We found that both DNase1 and lactoferrin were expressed in the HEK-293 cells from these transposons. We transfected USCs with the transposons according to our optimized transfection protocol and analyzed both cell lysates and media from transfected cells by western blot, using mock-transfected cells as a negative control and beta-actin as a loading control. We found that USCs express DNase1 (Fig. 7C), but could not detect DNase1 in the conditioned media. USCs also express lactoferrin from the piggyBac transposon, but at detectable levels both in the cell lysate and in the media. As expected, it appeared that higher levels of secreted lactoferrin were detected when the media were concentrated (Fig. 7C).

FIG. 7.

Expression of DNase1 and lactoferrin in USCs. (A) Vector maps of the DNase1 (yellow) and lactoferrin (orange) transposons. (B) Western blot for DNase1 (green, left) or lactoferrin (green, right) and β-actin loading control (red) in HEK-293 RIPA lysates transfected as indicated. (C) Western blot with antibodies to DNase1 or lactoferrin (top panel, green; source image shown in black in the bottom panel) and β-actin loading control (top panel, red) on transfected USC lysates and conditioned media with or without concentration, confirming expression of DNase1 and lactoferrin in stem cells and secretion of lactoferrin into the media.

Discussion

Rhabdomyolysis is a clinical syndrome cause by damaged muscles.^46,47 Physical overexertion, certain prescribed and illicit drugs, infections, heat stroke, crush injuries, transportation accidents, gunshot wounds, explosions, and other traumatic injuries may cause severe muscle damage.^48,49 Recently, it has been appreciated that the strong association between rhabdomyolysis and mortality is attributable to acute kidney injury that is currently only treated with supportive care.⁴⁸ During the injury, products are released by the breakdown of muscle into the bloodstream, which eventually leads to acute kidney injury.⁵⁰ As the kidney clears the degraded muscle fibers containing myoglobin, toxic effects on renal cells result in high rates of cell death and tubular injury. During AKI the macrophages release ETs containing DNA fibers and granule proteins into the kidney, designed to clear pathogens from a site of infection.²⁸ However, in most cases, rhabdomyolysis is not associated with infection and the ETs play an inflammatory role. The effectiveness of DNase1 to digest ETs has been reported previously.^51–54

After ET formation, the DNA becomes decondensed and binds to antimicrobial components such as elastase and lactoferrin in granular vesicles.⁵⁵ Lactoferrin has high affinity for iron and has been investigated for its NET fiber inhibition ability in addition to its antimicrobial properties.²⁸ Inhibition by treatment with DNase1 or Lactoferrin at the time of induction of injury resulted in a complete prevention of acute kidney injury.²⁸ Taking into consideration these two reported therapeutic proteins, we investigated the possibility of engineering cells with piggyBac transposons to produce DNase1 and lactoferrin as potential therapeutic treatments. In this study, we optimized transfection of stem cells collected from human urine samples called USCs and expressed multiple proteins in these cells.

The methods used to isolate USCs and the resulting cell type is similar to renal progenitor cells isolated from urine.^56,35 Differences in the isolation methods are likely to result in phenotypic changes. Renal progenitor cells have been transfected with laboratory-synthesized VIPER reagent.⁵⁷ To enhance the applicability of our nonviral transfection method, we chose to optimize transfection with a number of lipophilic reagents according to the manufacturer's protocols.

This work is the first step toward modifying USCs to inhibit kidney inflammation during rhabdomyolysis. Others have demonstrated that treatment with DNase1 and lactoferrin at the time of rhabdomyolysis induction stopped acute kidney injury in mice.²⁸ Our modified stem cells expressed DNase1 and lactoferrin. DNase1 was detected in cell lysate but was not secreted into the media in amounts detectable by immunoblot, for unknown reasons. Lactoferrin was present in both cell lysate and media aliquots. This study provides proof of concept that the USCs can be “protein factory” vehicles. In the future, the USCs present an intriguing cell type to become a “protein factory” for the manufacture of proteins of interest for cell and gene therapy.

Future work will determine if therapies based on lactoferrin and/or DNase1 USCs may have the ability to stop kidney damage in mouse models of rhabdomyolysis-associated acute kidney injury. We believe that administration of allogeneic genome-engineered USCs that survive for a short time following injury and secrete helpful substances at the site of injury should be tested as a proof-of-concept gene and cell therapy.

Authors' Contributions

Validation and investigation: Z.K., R.C.W., J.B., F.M.W., I.J., T.K.T., and L.E.W. Conceptualization: L.E.W. Methodology: Z.K., R.A.V., and L.E.W. Resources, project administration, funding acquisition, and supervision: M.H.W. and L.E.W. Writing—original draft and visualization: Z.K. and L.E.W. Writing—review and editing: Z.K., J.B., R.A.V., I.J., M.H.W., and L.E.W.

Disclosure Statement

The Woodard Laboratory and Wilson Laboratory receive unrelated research funding from Bayer and SalioGen Therapeutics. L.E.W. and M.H.W. provide service to the American Society of Gene and Cell Therapy.

Funding Information

L.E.W. (BX002797 and BX004845) and M.H.W. (BX004258 and BX002190) were supported by awards from the Department of Veterans Affairs. This work is the result of a pilot award to L.E.W. (P30 DK114809) from the National Institutes of Health (NIH), National Institute of Digestive and Diabetes and Kidney Diseases (NIDDK) and the Vanderbilt Center for Kidney Disease. M.H.W. was supported by a NIH award from the NIDDK (DK093660). Flow cytometry experiments were performed in the VA Flow Cytometry Laboratory with assistance from Catherine Alford. The VUIIS Center for Small Animal Imaging is supported in part by Vanderbilt Ingram Cancer Center Support Grant (P30 CA68485) and the IVIS spectrum was purchased with an NIH grant (S10 OD021804).